The design, synthesis, and evaluation of two universal doxorubicin-linkers: Preparation of conjugates that retain topoisomerase II activity

Article abstract of DOI:10.1016/j.bmcl.2005.09.046

The design, synthesis, and evaluation of two N-alkylmaleimide aldehydes have been achieved, which upon reductive alkylation with the C3′-amino group of doxorubicin (DOX) permits the preparation of DOX conjugates via Michael addition of thiol-containing vectors. This method enables the mild, facile, and high-throughput preparation of DOX conjugates that retain the basic C3′-nitrogen, a pre-requisite for topoisomerase II inhibition. Seven DOX-amino acid conjugates were prepared, each displaying similar inhibitory activity as the parent drug.

Full text of DOI:10.1016/j.bmcl.2005.09.046

Bioorganic & Medicinal Chemistry Letters 16 (2006) 104–107
The design, synthesis, and evaluation of two universal
doxorubicin-linkers: Preparation of conjugates that
retain topoisomerase II activity
Chengzao Sun,a,* Simon E. Aspland,a Carlo Ballatore,a Rosario Castillo,a
Amos B. Smith, IIIb,* and Angelo J. Castellinoa,*
aAcidophil, LLC, Suite 250, 10835 Road to the Cure, San Diego, CA 92121, USA
bDepartment of Chemistry, University of Pennsylvania, 231 South 34th Street, Philadelphia, PA 19104-6323, USA
Received 27 August 2005; revised 13 September 2005; accepted 15 September 2005
Available online 18 October 2005
Abstract—The design, synthesis, and evaluation of two N-alkylmaleimide aldehydes have been achieved, which upon reductive
alkylation with the C30-amino group of doxorubicin (DOX) permits the preparation of DOX conjugates via Michael addition of
thiol-containing vectors. This method enables the mild, facile, and high-throughput preparation of DOX conjugates that retain
the basic C30-nitrogen, a pre-requisite for topoisomerase II inhibition. Seven DOX–amino acid conjugates were prepared, each
displaying similar inhibitory activity as the parent drug.
Ó 2005 Elsevier Ltd. All rights reserved.
Anthracycline cytotoxic agents, in particular doxorubi-
cin (DOX) and daunorubicin, have been employed in
the area of cancer chemotherapy for more than 30 years.
Side effects, such as cardiotoxicity, in conjunction with
the development of spontaneous and acquired resis-
tance, however, limit the effectiveness of these drugs.
To improve efficacy, various tumor targeting approach-
es have been explored wherein the anthracycline is
attached to a vector.1
amines have also been prepared, which are stable to
proteolysis and retain the basic nitrogen. One example
of a proteolytically stable conjugate, which does not
contain a basic C30-nitrogen, is a DOX–D-penetratin
conjugate prepared as a C30-amide. This conjugate lost
5-fold of the cytotoxic activity compared to the parent
drug in K562 cells, but importantly displayed a 20-fold
improvement in MDR cells. Such a toxicity profile how-
ever may be ascribed to a difference in bio-distribution.6
Examples of anthracycline conjugates incorporating tu-
mor-specific or selective vectors have been reported
which contain hydrazone linkages at the C13 carbonyl,2
an ester linkage at the C14 hydroxyl,3 and/or amide4 or
alkylamine5 linkages at the C30-amine. Most reported
C30-amine conjugates comprise peptides or proteins at-
tached via amide bonds, which are later removed
exploiting prodrug strategies such as proteolysis. Such
prodrug approaches have been taken in order to release
the free drug with a basic anthracycline C30-nitrogen,
known to be a pre-requisite for topoisomerase II activi-
ty. Conjugates at the C30-nitrogen employing alkyl-
Alkylamine derivatives at the C30-nitrogen of DOX
have been prepared by reductive alkylation.4a,b,7 How-
ever, the reaction requires an excess (2–3 equiv) of alde-
hyde as the vector input to ensure complete reaction.
Even then only poor, or at best modest, yields are often
observed.
To avoid laborious syntheses and to permit parallel
attachment of a series of vectors to DOX while retaining
the topoisomerase II inhibition, we have designed, syn-
thesized, and evaluated two potentially universal
DOX-linkers (Scheme 1), which feature a basic C30-
nitrogen and a maleimide terminus that would permit
direct attachment of a wide variety of vectors through
Keywords: Anthracycline; Doxorubicin; Conjugate; Topoisomerase II;
Maleimide.
formation of
a maleimido–thioether linkage. The
structures of the universal DOX-linkers (1a and b) are
illustrated in Scheme 1, in conjunction with a general
reaction scheme for attachment of thiol-bearing vectors.
*
Corresponding authors. Tel.: +1 760 471 7390; fax: +1 858 459
6603; e-mail addresses: csun@acidophil.com; acastellino@acidophil.
com
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.09.046

C. Sun et al. / Bioorg. Med. Chem. Lett. 16 (2006) 104–107
105
O
OH
O
O
OH
O
OH
OH
OH
OH
SH-vector
OCH3
O
OH
O
OCH3
O
OH
O
O
O
O
O
HN
HN
X
X
OH
OH
vector
N
N
S
O
O
1a, X = -CH2-
1b, X = -CH2OCH2-
DOX-linker-vector
Scheme 1. General reaction scheme for the preparation of DOX-linker–vector conjugates via maleimide–thioether linkage.
The prospective DOX-30-aminoalkylmaleimide deriva-
tives (1a and b) were prepared via the route depicted
O
HO
OMe
O
O
HN
O
in Scheme 2. The respective maleimide alcohols 3a and
HO
N
H
O
b were constructed via the literature protocols,8 involv-
ing the condensation of amino alcohols 2a and b with
N-methoxycarbonyl maleimide in saturated aqueous
NaHCO3. The resultant alcohols were then oxidized
with Dess–Martin periodinane9 to furnish the corre-
sponding aldehydes 4a and b, respectively. Aldehyde
4a had been reported to be unstable due to rapid poly-
merization10 and therefore not readily isolable in pure
state. We were however able to alleviate this problem
by using 4a and b immediately in the next step after
purification by silica gel column chromatography. A
small sample of 4a was diverted for characterization
by NMR.11
i
OMe
+
6
NH2.HCl
O
N
O
4a
5
Scheme 3. Reagents: (i) NaCNBH3, MeOH, 88%.
(Scheme 3). As expected the mono-alkylated product 6
was produced in 88% yield after HPLC purification.
Topoisomerase II catalyzes the formation of relaxed
conformations of DNA from the super-coiled plasmid.13
The mechanism of action of DOX is believed to involve
inhibition of topoisomerase II activity, which leads to
the eventual breakage of genomic DNA.
Treating a mixture of DOX and freshly prepared alde-
hydes 4a and b under reductive alkylation conditions,
employing NaCNBH3, provided 1a and b in modest
yield (ca. 20–30%).12 The major side products, observed
in both cases, were tentatively assigned to be the corre-
sponding dihydroquinones, based on the observation of
parent ions with two higher mass units (LC/MS) than
those of the DOX conjugates. To verify that the side
reactions were not associated with the aminosugar or
with the maleimido aldehydes, daunosaminide 5 was
condensed with aldehyde 4a via reductive alkylation
With the prospective universal DOX–maleimide linkers
in hand, we prepared a small set of DOX–amino acid
conjugates to evaluate their ability to inhibit topoiso-
merase II activity. We selected amino acid conjugates
of Asp, Lys, Ser, and Cys based on the following consid-
erations. First, given that many targeting vectors will be
peptides, a prodrug approach using the DOX-linkers
described herein will likely result in an amino acid
conjugate. Second, by employing amino acids with
common and representative peptide functionalities, such
as amino, carboxyl, and hydroxyl groups, we would at
the same time examine the compatibility of the Michael
addition between thiols and maleimides in the presence
of these common functional groups.
O
O
O
ii
i
N
+
N
H2N
X
OH
OMe
X
OH
O
O
Seven DOX–amino acid conjugates were prepared in a
parallel fashion by mixing, under neutral conditions,
the DOX–maleimides 1a or b with cysteine or related
amino acids pre-equipped with a thiol functional group
(Scheme 4). The coupling reactions in general proved to
be fast (completed within 30–60 min), with high conver-
sion rates. Amino, carboxyl, and hydroxyl groups do
not appear to interfere with the reaction. Yields after
HPLC purifications are listed in Table 1.
2a, X = -CH2-
2b, X = -CH2OCH2-
3a,b
O
OH
OH
O
O
OH
N
H
OH
iii
X
O
O
OCH3
O
O
O
1a,b
4a,b
O
HN
OH
X
N
The DOX derivatives 9a–g were then directly compared
to DOX for their inhibitory effect in a standard topoiso-
merase II activity assay, at 10, 3, 1, 0.3, 0.1, and
0.03 lmol concentrations. All compounds revealed first
inhibition effects at 1–10 lmol levels, similar to that of
the parent drug.
O
Scheme 2. Reagents: (i) satd aq NaHCO3, (a) 47%, (b) 61%; (ii)
Dess–Martin periodinane, CH2Cl2, (a) 43%, (b) 41%; (iii) doxorubicin
hydrochloride, 1 M NaCNBH3, cat. AcOH, MeCN/H2O, (a) 21%, (b)
21%.

106
C. Sun et al. / Bioorg. Med. Chem. Lett. 16 (2006) 104–107
O
O
OH
OH
O
R2HN
OH
R3
iii
OH
R1
7a, R1 = CH2COOtBu, R2 = Boc, R3 = OH
7b, R1 = CH2OH, R2 = Boc, R3 = OH
7c, R1 = H, R2 =(CH2)4NHBoc, R3 = OMe
OCH3
O
O
O
O
9a-g
i or ii
8a, R1 = CH2CO2H, R2 = H, R3 = NH(CH2)2SH
8b, R1 = CH2OH, R2 = H, R3 = NH(CH2)2SH
8c, R1 = (CH2)4NH2, R2 = CO(CH2)2SH, R3 = OMe
8d, cysteine
NH
HO
X
N
Y
O
Scheme 4. Reagents: (i) for 8a and b, HS(CH2)2NH2ÆHCl, BOP, DIEA, DMSO; (ii) for 8c, HS(CH2)2CO2H, BOP, DIEA, DMSO; and (iii) 1a or b,
DMSO.
Table 1. Examples of DOX–amino acid conjugates prepared via 1a,b
Compound
X
Y
Conversion (%)b
Yield (HPLC, %)
First effective concentration for
topo II inhibition (lmol)
DOX
9a
9b
1–3
1–3
–CH2–
–CH2–
–CH2–
H-Asp-NHCH2CH2S-
–SCH2CH2CO-Lys-OH
H-Cys(ꢀ)-OHa
>95
>95
80c
46
38
19
44
43
49
26
1–3
9c
9d
3–10
3–10
1–3
–CH2OCH2–
–CH2OCH2–
–CH2OCH2–
–CH2OCH2–
H-Asp-NHCH2CH2S–
–SCH2CH2CO-Lys-OH
H-Ser-NHCH2CH2S–
H-Cys(ꢀ)-OHa
>95
>95
>95
85c
9e
9f
9g
1–3
3–10
a The cysteine conjugates are linked through the side-chain thiol.
b A mixture of 1a or b and 9a,b,d–f was stirred in DMSO for 30 min before LC/MS determination of the conversion rate and HPLC purification.
c Cysteine was suspended in DMSO with 5% H2O and stirred with 1a or b for 1 h to ensure complete consumption of the latter compounds.
In summary, two new DOX-30-aminoalkylmaleimide
Jones, R. E. Nat. Med. 2000, 6, 1248; (e) Pasqualini, R.;
Koivunen, E.; Kain, R.; Lahdenranta, J.; Sakamoto, M.;
derivatives designed to be versatile DOX-linkers for
Stryhn, A.; Ashmun, R. A.; Shapiro, L. H.; Arap, W.;
Ruoslahti, E. Cancer Res. 2000, 60, 722; (f) Liu, C.; Sun,
C.; Huang, H.; Janda, K.; Edgington, T. Cancer Res.
2003, 63, 2957.
the preparation of stable conjugates and/or prodrugs
have been designed and synthesized. When conjugated
with a series of amino acids pre-equipped with thiol
groups, the conjugates retained the parent drugꢀs activi-
ty toward topoisomerase II. The availability of this
5. (a) Bakina, E.; Wu, Z.; Rosenblum, M.; Farquhar, D. J.
Med. Chem. 1997, 40, 4013; (b) Fenick, D. J.; Taatjes, D.
methodology holds the promise for high-throughput
synthesis of DOX conjugates, as well as conjugates of
other anthracycline drugs.
J.; Koch, T. H. J. Med. Chem. 1997, 40, 2452; (c) Kasiotis,
K. M.; Magiatis, P.; Pratsinis, H.; Skaltsounis, A.; Abadji,
V.; Charalambous, A.; Moutsatsou, P.; Haroutounian, S.
A. Steroids 2001, 66, 785.
6. Mazel, M.; Clair, P.; Rousselle, C.; Vidal, P.; Scherrmann,
J.-M.; Mathieu, D.; Temsamani, J. Anti-Cancer Drugs
2001, 12, 107.
References and notes
7. Allart, B.; Lehtolainen, P.; Yla-Herttuala, S.; Martin, J.
F.; Selwood, D. L. Bioconjug. Chem. 2003, 14, 187.
8. (a) Keller, O.; Rudinger, J. Helv. Chim. Acta 1975, 58, 531;
(b) Dean, R. T.; Boutin, R. H.; Weber, R. W. U.S. Patent
5,144,043, 1992.
9. Dess, D. B.; Martin, J. C. J. Org. Chem. 1983, 48, 4155.
10. Bott, K. Chem. Commun. 1969, 22, 1304.
11. Synthetic procedure and proton NMR for 4a: 1-(3-
1. For reviews of recent efforts to improve anthracyclines see
(a) Monneret, C. Eur. J. Med. Chem. 2001, 36, 483; (b)
Minotti, G.; Menna, P.; Salvatorelli, E.; Cairo, G.;
Gianni, L. Pharmacol. Rev. 2004, 56, 185.
2. Trail, P. A.; Willner, D.; Lasch, S. J.; Henderson, A. J.;
Hofstead, S.; Casazza, A. M.; Firestone, R. A.; Hellstrom,
I.; Hellstrom, K. E. Science 1993, 261, 212.
3. Nagy, A.; Plonowski, A.; Schally, A. V. Proc. Natl. Acad.
Sci. U.S.A. 2000, 97, 829.
hydroxypropyl)-1H-pyrrole-2,5-dione
(3a,
200 mg,
1.29 mmol) was dissolved in 5 mL CH2Cl2. DMP (15%
wt in CH2Cl2, 4 mL, 1.93 mmol) was added in one portion
and the mixture was stirred for 2 h. 2-Propanol (3 mL) was
added and the solution was stirred for 30 min. The
resulting solution was filtered through a silica gel pad
eluted with EtOAc and the filtrate was concentrated. The
crude product was purified by silica gel chromatography
eluting with EtOAc–hexane (2/1) providing 3-(2,5-dioxo-
2,5-dihydro-pyrrole-1-yl)-propionaldehyde (4a, 110.0 mg,
4. For a few examples see (a) Bosslet, K.; Straub, R.;
Blumrich, M.; Czech, J.; Gerken, M.; Sperker, B.; Kro-
emer, H. K.; Gesson, J. P.; Koch, M.; Monneret, C.
Cancer Res. 1998, 58, 1195; (b) Denmeade, S. R.; Nagy,
A.; Gao, J.; Lilja, H.; Schally, A. V.; Isaacs, J. T. Cancer
Res. 1998, 58, 2537; (c) Trouet, A.; Masquelier, M.;
Baurain, R.; Deprez-De Campeneere, D. Proc. Natl. Acad.
Sci. U.S.A. 1982, 79, 626; (d) DeFeo-Jones, D.; Garsky, V.
M.; Wong, B. K.; Feng, D. M.; Bolyar, T.; Haskell, K.;
Kiefer, D. M.; Leander, K.; McAvoy, E.; Lumma, P.;
Wai, J.; Senderak, E. T.; Motzel, S. L.; Keenan, K.; Van
Zwieten, M.; Lin, J. H.; Freidinger, R.; Huff, J.; Oliff, A.;
1
0.72 mmol, 55.7% yield), which was used immediately. H
NMR (CDCl3, 300 MHz) d 9.74 (t, J = 1.2 Hz, 1H), 6.69
(s, 2H), 3.84 (t, J = 6.9 Hz, 2H), 2.77 (dt, J = 1.2, 6.9 Hz,

C. Sun et al. / Bioorg. Med. Chem. Lett. 16 (2006) 104–107
107
2H). Although NMR provided a clean spectrum, we
observed the formation of precipitate 15 min after the
sample was dissolved in CDCl3 at a concentration of
30 mg/mL.
(m, 3H), 3.32–2.98 (m, 2H), 2.76 (m, 1H), 2.54 (m, 2H),
2.37 (m, 1H), 2.15 (m, 1H), 1.85–1.54 (m, 4H), 1.37 (d,
J = 7.0 Hz, 3H). Electrospray (ESI) m/z 681.2 (M+H+,
C34H36N2O13 required 681.2).
12. Synthetic procedure and selected spectroscopic data for
1a: to a stirred solution of doxorubicin hydrochloride
(100 mg, 0.172 mmol), 4a (68.2 mg, 0.446 mmol), and
glacial AcOH (20 lL, 195 mol %) in CH3CN–H2O (2/1,
5 mL) was added a 1 M solution of NaCNBH3 in THF
(57 lL, 0.33 mol %). The mixture was stirred under
nitrogen atmosphere in the dark at rt for 1 h. The solution
was then concentrated in vacuo to give a residue, which
was diluted with an aqueous 5% NaHCO3 solution and
extracted with CH2Cl2. Concentration of the organic
solution and purification of the resulting residue by silica
gel chromatography eluting with CH2Cl2–CH3OH (20/1)
provided 26.0 mg of N-3-maleimidopropyl Doxorubicin
13. All DOX conjugates were assayed for their ability to
inhibit topoisomerase II (topo II) using the topo II Drug
Screening Kit (TopoGEN Inc, Columbus, OH). Specifi-
cally the kit was used to assay whether DOX derivatives
alter the ability of topo II to catalyze the formation of
relaxed conformation DNA from a super-coiled plasmid.
The substrate DNA is a plasmid called pRYG that
contains a 54 base pair series of alternating purine/
pyrimidine DNA specifically acted on by topo II. In the
absence of any drug topo II produces a relaxed confor-
mation DNA from super-coiled starting DNA. The two
forms of DNA can be distinguished on an agarose gel: the
super-coiled DNA appearing as a single band and the
relaxed DNA as a series of usually three bands close
together. DOX usually inhibited this process at about 1–
3 lmol. DOX derivatives were compared directly to DOX
at 10, 3, 1, 0.3, 0.1, and 0.03 lmol concentrations.
1
1a (21.4% yield). H NMR (CDCl3, 300 MHz) d 8.03 (d,
J = 8.4 Hz, 1H), 7.79 (t, J = 8.4 Hz, 1H), 7.41 (d,
J = 8.4 Hz, 1H), 6.68 (s, 2H), 5.51 (m, 1H), 5.32 (m,
1H), 4.82–4.76 (m, 2H), 4.09 (s, 3H), 3.96 (m, 1H), 3.58