Design and optimization of 20-O-linked camptothecin glycoconjugates as anticancer agents

Article abstract of DOI:10.1021/jm010893l

To improve the biological profile of 20(S)-camptothecin, a novel class of 20-O-linked camptothecin glycoconjugates has been designed for preferential cellular uptake into tumor cells by an active transport mechanism. Such conjugates have been optimized for enhanced solubility, stabilization of the camptothecin lactone ring, sufficient hydrolytic and proteolytic stability, and for an overall improvement in tumor selectivity. The constitution of the peptide spacer has a major impact on stability and biological activity of the conjugates both in vitro and in vivo. Glycoconjugates 17-22 with valine residues at the linkage position to camptothecin are sufficiently stable and show good antitumor activity in vitro against HT29 and other tumor cell lines. Fluorescence microscopy and flow cytometry experiments indicate that glycoconjugates such as 19 are taken up into lysosomal compartments of the tumor cell line HT29 by an active transport mechanism. The steric configuration of the particular amino acid residues linked to the camptothecin moiety has a major impact on the in vivo activity of the corresponding glycoconjugates in the breast cancer xenograft MX-1 model. Inhibiting tumor growth by >96%, the glycoconjugates 19 and 21 show the best activity in this particular model and have been investigated more extensively. The glycoconjugate 19 compares favorably to topotecan 4 and glycoconjugate 21 with respect to toxicity against hematopoietic stem cells and hepatocytes. Based on its profile, 19 has been selected for clinical trials.

Full text of DOI:10.1021/jm010893l

4186
J . Med. Chem. 2001, 44, 4186-4195
Design a n d Op tim iza tion of 20-O-Lin k ed Ca m p toth ecin Glycocon ju ga tes a s
An tica n cer Agen ts
Hans-Georg Lerchen,*,§ J oerg Baumgarten,§ Karsten von dem Bruch,§ Thomas E. Lehmann,§ Michael Sperzel,†
Grazyna Kempka,† and Heinz-Herbert Fiebig‡
Bayer AG, Central Research, Life Sciences, 51368 Leverkusen, Germany, Bayer AG, Pharma Research Center,
42096 Wuppertal, Germany, and Institute for Experimental Oncology, Oncotest GmbH, 79108 Freiburg, Germany
Received April 4, 2001
To improve the biological profile of 20(S)-camptothecin, a novel class of 20-O-linked camptothecin
glycoconjugates has been designed for preferential cellular uptake into tumor cells by an active
transport mechanism. Such conjugates have been optimized for enhanced solubility, stabilization
of the camptothecin lactone ring, sufficient hydrolytic and proteolytic stability, and for an overall
improvement in tumor selectivity. The constitution of the peptide spacer has a major impact
on stability and biological activity of the conjugates both in vitro and in vivo. Glycoconjugates
17-22 with valine residues at the linkage position to camptothecin are sufficiently stable and
show good antitumor activity in vitro against HT29 and other tumor cell lines. Fluorescence
microscopy and flow cytometry experiments indicate that glycoconjugates such as 19 are taken
up into lysosomal compartments of the tumor cell line HT29 by an active transport mechanism.
The steric configuration of the particular amino acid residues linked to the camptothecin moiety
has a major impact on the in vivo activity of the corresponding glycoconjugates in the breast
cancer xenograft MX-1 model. Inhibiting tumor growth by >96%, the glycoconjugates 19 and
21 show the best activity in this particular model and have been investigated more extensively.
The glycoconjugate 19 compares favorably to topotecan 4 and glycoconjugate 21 with respect
to toxicity against hematopoietic stem cells and hepatocytes. Based on its profile, 19 has been
selected for clinical trials.
In tr od u ction
compounds is the opening of the lactone E-ring and the
formation of an equilibrium between the ring closed
lactone form and the open carboxylate form (Scheme 1)
which is pH- and also species-dependent. Particularly,
human serum albumin preferentially binds to the
carboxylate form of camptothecin derivatives, shifting
the equilibrium in favor of the carboxylate.7 The lactone
form is more stable in mouse serum, leading to the
marked efficacy of these compounds in nude mouse
xenografts,8 which, however, might not be predictive for
the same level of clinical activity for these compounds.
A number of research programs focused on improved
analogues are ongoing. So far, several camptothecin
derivatives such as 9-nitrocamptothecin (RFS-2000),
exatecan (DX-8951), lurtotecan (NX-211), BNP-1350,
diflomotecan (BN80915), and CKD-602 have entered
clinical trials.9
Furthermore, a number of prodrugs and delivery
systems have been developed. 20-O-Alkyl esters of
camptothecin10 and simple amino acid prodrugs11 have
been prepared to increase lactone ring stability and
water solubility of camptothecins. Polymeric delivery
systems such as poly(ethylene glycol) (PEG) conjugates12
and N-2-hydroxypropylmethacrylamide polymer (HPMA)
conjugates13 have been developed to improve the phar-
macokinetic profile of camptothecins.
Camptothecin 1 is a pentacyclic alkaloid isolated by
Wall et al. in the early 1960s from the Chinese tree
Camptotheca acuminata.1 It showed impressive activity
against leukemias and a variety of solid tumors. Clinical
trials were begun in 1970 with the more soluble sodium
salt 2. Despite some early evidence for activity, toxicities
such as myelosuppression, vomiting, and diarrhea were
severe, and clinical trials were suspended.2 Neverthe-
less, due to its unique activity camptothecin still
represents an important lead structure for extensive
research programs directed towards novel anticancer
compounds.
Important milestones in this field have been the
elucidation of the mechanism of action in the mid 1980s
and the approval and clinical success of the first
representatives of the camptothecin class of compounds,
irinotecan (3) (CPT-11; Yakult Honsha)3 and topotecan
(4) (SmithKline Beecham),4 another decade later.
The camptothecins act by binding to the topo-
isomerase I-DNA complex, leading to an accumulation
of DNA strand breaks upon replication, ultimately
causing cell death.5 The closed lactone E-ring is es-
sential for activity based on the established mechanism
of action.6 A major issue of the camptothecin class of
In previous studies we optimized modified fucoside
residues for preferential receptor-mediated uptake into
tumor cells compared to liver cells utilizing neoglyco-
conjugates of bovine serum albumin (BSA).14 In these
model systems it has been shown that a chemical
modification of fucoside residues at the 3-hydroxy group
* To whom correspondence should be addressed: Bayer AG
Leverkusen, Central Research, Bldg. Q18, 51368 Leverkusen,
Germany. Tel: +49 214 30 24426. Fax: +49 214 30 46410. E-mail:
hans-georg.lerchen.hl@bayer-ag.de.
§ Bayer AG, Central Research.
† Bayer AG, Pharma Research Center.
‡ Oncotest GmbH.
10.1021/jm010893l CCC: $20.00 © 2001 American Chemical Society
Published on Web 10/26/2001

20-O-Linked Camptothecin Glycoconjugates
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24 4187
Sch em e 1. Structures of 20(S)-Camptothecin, Irinotecan, and Topotecan
led to a decreased uptake of the respective neoglyco-
conjugates of BSA into liver cells. In contrast, the
efficient uptake of such BSA neoglycoconjugates by
SW480 tumor cells was not negatively affected by
modifications of the 3-hydroxy group. As preferred
carbohydrate residues mediating a differential uptake
into tumor cells vs liver cells 3-O-methyl-fucoside and
3-O-carboxymethyl-fucoside residues have been identi-
fied.14 Our next goal was to improve the tumor selectiv-
ity and solubility of camptothecin by the synthesis of
glycoconjugates designed for preferential cellular uptake
into tumor cells by an active transport mechanism.
Therefore, previously optimized fucoside derivatives
such as p-aminophenyl 3-O-methyl-â-L-fucosides have
been attached to the 20-hydroxy group of the camptoth-
ecin molecule via ester linked dipeptide spacers to
obtain a novel class of low molecular weight glyco-
conjugates with improved water solubility and with
increased stability of the lactone ring. With respect to
the overall profile of such low molecular weight camp-
tothecin glycoconjugates the optimization of the spacer
group turned out to be of particular importance. Cellular
uptake is not solely triggered by the carbohydrate
residue, and, furthermore, the spacer group has a major
impact on the cleavability of the conjugates which
directly contributes to the cytotoxic activity. Such glyco-
conjugates should be sufficiently stable in culture
medium and in serum and at the same time unfold their
cytotoxic activity preferably by release of the toxophore
inside the target cell after cellular uptake.
efficient method utilizing urethane-protected N-carboxy
anhydrides16 (UNCAs). The camptothecin amino acid
conjugates 5-7 have been obtained by acylation of the
20-hydroxy group with UNCAs and subsequent depro-
tection of the terminal amino group as described previ-
ously.15 Attachment of appropriately protected residues
of basic amino acids 8-10 in the presence of N-ethyl-
N′-(3-dimethylaminopropyl) carbodiimide hydrochloride
(EDCI) and hydroxybenzotriazole (HOBt) yields the
protected dipeptide conjugates 11a -h (Scheme 2). In
the case of the attachment of the histidine moiety 8, no
protection of the side chain functionality is required,
whereas in the cases of lysine 9 fluorenyl-9-methoxy-
carbonyl (Fmoc) and arginine 10 bis-tert-butoxycarbonyl
(Boc), protection of the side chain is required. The tert-
butoxycarbonyl (Boc)-protecting group at the amino
terminus is removed with trifluoroacetic acid, affording
the dipeptide conjugates 12b, g, h carrying lysine with
a Fmoc-protected side chain at the N-terminus and the
fully deprotected dipeptide conjugates carrying free
histidine (12a , e, f) and arginine residues (12c, d ) at
the N-terminus, respectively.
The carbohydrate residues should be attached to the
R-amino terminus of the dipeptide conjugates to en-
hance proteolytic stability of the glycoconjugates, whereas
the basic side chains should contribute to water solubil-
ity. p-Aminophenyl 3-O-methyl-â-L-fucopyranoside was
one of the preferred carbohydrate residues with promise
for tumor targeting.14 It is transformed into the isothio-
cyanate 13 ready for linkage to the dipeptide conjugates
12a -h . The coupling reactions proceed smoothly in the
presence of N-ethyl diisopropylamine. It is worth men-
tioning that in the case of dipeptide conjugates 12a and
12c-f with terminal arginine and histidine residues,
respectively, the coupling reactions proceed with high
regioselectivity at the R-amino group to yield the glyco-
conjugates 14a and 14c-f without any side chain
protection required. In the case of the partially protected
lysine-containing glycoconjugates 14b, g, h , after the
carbohydrate attachment the Fmoc group is removed
with piperidine in dimethyl formamide. Finally, the
glycoconjugates are treated with 1 equiv of 0.1 N
aqueous hydrochloric acid to provide the water soluble
glycoconjugate hydrochlorides 15-22.
We have reported previously that a bulky side chain
of the amino acid linked to camptothecin via an ester
bond is crucial for hydrolytic stability of such 20-O-
linked glycoconjugates.15
Here we describe the optimization of those camptoth-
ecin glycoconjugates according to criteria such as ef-
ficacy, stability, solubility, and reduced toxicity against
hematopoietic stem cells and hepatocytes. The lead
compound 19 (BAY 38-3441) has been selected for
clinical trials. Investigations toward cellular uptake of
19 will also be described.
Resu lts a n d Discu ssion
Ch em istr y. For the challenging acylation of the
tertiary and unreactive 20-hydroxy group, particularly
with bulky amino acid residues, we have developed an
Solu bility, Sta bility, a n d An titu m or a l Activity
of Ca m p toth ecin Glycocon ju ga tes 15-22. As hy-

4188 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24
Lerchen et al.
Sch em e 2. Synthesis of Glycoconjugates of 20(S)-Camptothecin
Ta ble 1. Stability and in Vitro Activity of Camptothecin
Glycoconjugates
drochlorides, the exemplified glycoconjugates 15-22
exhibit good water solubility [∼5 mg/mL at pH 4-5].
Furthermore, due to the R-acylation the lactone ring of
the camptothecin moiety is markedly stabilized in these
glycoconjugates. Even under drastic conditions such as
addition of 2 equiv of aqueous sodium hydroxide, which
leads to a quantitative ring opening in the case of
topotecan 4, the glycoconjugates 19 and 21 remain
intact with less than 2% lactone ring opening.
stability in culture
medium of HT29
after 24 h % CPT/
% conjugate
activity against cancer
cell lines [IC50/nM]
compd
R1
R2
HT29 SW480 B16F10
15
16
17
18
19
20
21
22
Gly
Gly
Val
D-Val Arg
Val His
D-Val His
Val Lys
D-Val Lys
His
Lys
Arg
95/0
68/23
6/88
6/89
7/84
5/86
3/91
4/86
10
10
100
100
60
160
150
120
15
30
30
50
170
110
100
230
170
150
400
400
250
600
700
600
According to our design criteria, camptothecin glyco-
conjugates should have reasonable hydrolytic and pro-
teolytic stability in the culture medium of tumor cell
lines. In this case the observed cytotoxic activity should
mainly be due to a cellular uptake of glycoconjugates
rather than an efficient release of the toxophore outside
the cell. Therefore, in parallel to measuring cytotoxic
activity against tumor cell lines (Table 1, column 5-7)
using an MTT growth inhibition assay, the stability of
the conjugates in the culture medium of HT29 tumor
cells was assessed by HPLC analysis (Table 1, column
4).
The side chain of the amino acid residue linked to
camptothecin has a significant impact on both stability
and cytotoxic activity of the glycoconjugates. Glyco-
conjugates 15 and 16 with a glycine residue linked to
camptothecin are readily cleaved in the culture medium
with an almost quantitative camptothecin release within
24 h (Table 1, column 4). Therefore, the cytotoxic activity
observed with the conjugates 15 and 16 is most probably
due to the amount of camptothecin cleaved in the
culture medium. The cytotoxic activity of camptothecin
1 in these assays is in the same range (IC50 values are
5, 10, and 20 nM against HT29, SW480, and B16F10,
respectively). Topotecan 4 in these systems displayed
cytotoxic activity with IC50s of 15, 20, and 150 nM.
In contrast, a bulky side chain R1 of the amino acid
residue linked to the hydroxy group of camptothecin
significantly increases the stability of respective glyco-
conjugates in the supernatant of HT29 tumor cells
(Table 1, column 4, conjugates 17-22). At the same time
these conjugates exhibit good antitumoral activity in
vitro against HT29 and other tumor cell lines, which
cannot be fully explained by the amount of camptothecin
released extracellularly (Table 1, column 5-7). In a
cellular assay the addition of 10 mol% of free camptoth-
ecin to glycoconjugate 19 does not significantly alter its
cytotoxic activity against HT29 in vitro (Figure 1).

20-O-Linked Camptothecin Glycoconjugates
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24 4189
F igu r e 1. Effect of the addition of 10% camptothecin on the
cytotoxic activity of glycoconjugate 19 against HT29 cell
cultures.
F igu r e 3. Hemopoietic activity in mice. Measurement of
colony forming units counts (CFU-c) of progenitor stem cells
per femur, 4 and 24 h after 3 days of treatment at the
respective MTDs of topotecan 4, 19, and 21.
F igu r e 2. In vivo activity of glycoconjugates in the MX-1
breast cancer model with daily i.v. administration for 3 days
at their MTD.
most active compounds 19 and 21 in comparison to
topotecan (4), C57BL/6 mice have been treated intra-
peritoneally daily for 3 days. Subsequently, 4 and 24 h
after the last treatment hemopoietic stem cells have
been removed from the femur and the ability for colony
formation in vitro has been determined for stem cells
isolated from the treated groups compared to the control
groups. The counts of colony forming units (CFU-C) per
femur is shown in Figure 3 for the untreated control
group and for the groups treated with equitoxic doses
of 19 and 21 and the reference compound 4. Equitoxic
doses have been estimated from a broader panel of
pharmacology models.
The steric configuration of the valine residue and the
particular constitution of the side chain of the basic
amino acid residue R2 only have a minor impact on
stability and cytotoxic activity against tumor cell lines
in vitro.
Glycoconjugates with sufficient stability have been
investigated in vivo against human tumor xenografts,
such as the breast cancer MX-1, grown subcutaneously
in mice. Glycoconjugates have been administered in-
travenously for the first 3 days after randomization at
dose ranges between 8 and 32 mg/kg/day. The activities
of the compounds at their respective maximum tolerated
doses (MTDs) are exemplified in Figure 2.
The glycoconjugate 19 exhibits the lowest myelo-
toxicity in this model. Four hours after the third
treatment with 18 and 27 mg/kg/day, respectively, the
count of colony forming units is only moderately de-
creased compared to the untreated control group. The
effect is reversible, and mice appeared completely
recovered 24 h after the third treatment. In contrast,
glycoconjugate 21 and topotecan 4 exhibit significantly
higher myelotoxicity in this model.
The antitumor activities of compounds 17-22 against
tumor cell lines in vitro are in the same range (Table
1). In contrast, the in vivo activities of these compounds
in the MX-1 breast cancer model are significantly
different. Particularly, a strong dependency on the
configuration of the valine residue is observed when
comparing the pairs of diastereoisomers 17/18 and 21/
22. The glycoconjugates 19 and 21 show the highest
activity in this model and effected tumor remissions
with optimal T/C values of 1.8 and 3.4 at their MTDs of
32 and 16 mg/kg/day, respectively. Topotecan in the
same model effected tumor remissions with a T/C of 12.7
at an approximate MTD of 2.5 mg/kg/day with a 3-day
therapy schedule. Therefore, further biological evalua-
tions have been focused on the glycoconjugates 19 and
21 in comparison to topotecan.
Cytotoxicity on P r im a r y Hep a tocytes fr om Dif-
fer en t Sp ecies in Vitr o. Liver damage might be a
possible limiting toxicity for camptothecin derivatives,
particularly for glycoconjugates. An in vitro model using
primary hepatocytes from rat, dog, and man cultivated
in a “sandwich model” was used for the determination
of the hepatotoxic potential. As endpoints, aspartate
aminotransferase (AST), alanine aminotransferase (ALT),
and lactate dehydrogenase (LDH) levels in the cellular
supernatant have been determined, and the results
obtained with compounds 19, 21, and 4 are exemplified
in Table 2.
Eva lu a tion of Hem a top oietic Activity of Con ju -
ga t es 19 a n d 21. To assess the myelotoxicity of the

4190 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24
Lerchen et al.
F igu r e 4. Uptake of glycoconjugates into the cellular compartments of HT29 cells determined by confocal laser scan microscopy.
Ta ble 2. No Effect Concentrations (NOEC) with Respect to
in an analogous manner as described for camptothecin
Elevated Levels of ALT,a AST,band LDHc in the Supernatant of
conjugates in Scheme 2.
Cultivated Hepatocytes
Incubation of HT29 cells for 6 h with the nonconju-
gated fluorescence marker 24 effects an efficient cellular
rat NOEC [µM] dog NOEC[µM] human NOEC [µM]
compd ALT AST LDH ALT AST LDH ALT AST LDH
staining at concentrations as low as 0.1 µM. Confocal
laser scan microscopy reveals a diffuse staining of the
whole cell, only sparing the nucleus (Figure 4, column
C). In contrast, the glycoconjugate 23 displays cellular
staining at 0.5 µM, which is predominantly localized in
the lysosomal compartments, the same pattern as
observed for the camptothecin glycoconjugate 19 at 100
µM.
To further investigate the mode of cellular uptake,
FACS studies have been performed with the fluores-
cence marker 24 and its respective glycoconjugate 23
at 4 °C and at 37 °C (Figure 5). As indicated by cellular
staining and in consistency with an unspecific, passive
membrane penetration, the cellular uptake of the
fluorescence marker 24 does not show any temperature
dependency. In contrast, the cellular uptake of the
glycoconjugate 23 is markedly higher at 37 °C compared
to 4 °C, which is almost at control level. This temper-
ature dependency of cellular staining indicates an active
transport of 23 into the cell. Taken together, these
results support a cellular uptake of the camptothecin
glycoconjugate 19 and its fluorescent analogue 23 into
lysosomal compartments of HT29 tumor cells by an
active transport mechanism.
19
21
4
10
30
0.1 0.1
10
30
10
30
1
10 10
10 30
10
30
1
30
30
0.1
30
30
1
30
30
10
1
0.1
a
ALT: alanine aminotransferase. b AST: aspartate aminotrans-
ferase. c LDH: lactate dehydrogenase.
The glycoconjugates 19 and 21 exhibit a markedly
decreased toxicity in vitro against cultivated hepatocytes
compared to topotecan.
Stu d ies on Cellu la r Up ta k e of Glycocon ju ga tes.
To investigate the mode of cellular uptake of glyco-
conjugates into the tumor cell line HT29, confocal laser
scan fluorescence microscopy and fluorescence activated
cell sorting (FACS) technologies have been employed.
Due to the inadequate fluorescence of the camptothecin
molecule, high concentrations (100 µM) and long incu-
bation times (24 h) with camptothecin glycoconjugates
such as 19 are required to effect detectable cellular
staining (Figure 4, column A). After 24 h incubation of
HT29 cells with 100 µM of glycoconjugate 19, a cellular
staining consisting predominantly of lysosomal com-
partments is observed.
To verify these results in the concentration range of
the IC50 values of the glycoconjugate 19, a model system
was employed, where the poor fluorophor camptothecin
was substituted by the fluorescence dye 24, to obtain
the highly fluorescent glycoconjugate 23. The fluoro-
phore 24 was chosen to closely reassemble the molecular
wheight, size of the aromatic ring system, and polarity
of camptothecin. The glycoconjugate 23 was synthesized
Con clu sion s
To improve the biological profile of the potent anti-
cancer compound camptothecin, glycoconjugates have
been synthesized. A 3-O-methyl-fucoside residue, which
has previously been shown to mediate a preferential
uptake of neoglycoconjugates of bovine serum albumin

20-O-Linked Camptothecin Glycoconjugates
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24 4191
breast cancer xenograft MX-1 model. The glycoconju-
gates 19 and 21 showed the best activity in this
particular model and have been investigated more
extensively.
The glycoconjugate 19 shows an excellent activity in
vivo against the MX-1 xenograft, inhibiting tumor
growth by >96%. This activity has been confirmed in a
variety of tumor models.17 Furthermore, it shows a
favorable profile compared to topotecan 4 and 21 with
respect to toxicity against hematopoietic stem cells and
hepatocytes. Fluorescence microscopic and flow cytom-
etry experiments suggest that it is taken up into
lysosomal compartments of the tumor cell line HT29 by
an active transport mechanism. From this novel class
of anticancer agents, the glycoconjugate 19 (BAY 38-
3441) has been selected for clinical development.
Exp er im en ta l Section
Gen er a l Meth od s. Melting points were determined in open
capillaries on
a Bu¨chi melting point apparatus and are
uncorrected. Flash chromatography was carried out on E.
Merck silica gel, 230-400 mesh. TLC analysis was carried out
on E. Merck silica gel plates 60F254. NMR spectra were
recorded at 400 MHz with equipment from Bruker. Chemical
shifts and coupling constants are given in ppm and Hz,
respectively. FAB-MS was performed on a Finnigan MAT 900
double focusing sector instrument. HPLC analysis was per-
formed on
a Waters Alliance 2690 instrument with UV
detection at 257 nM or 356 nM; column: E. Merck LiChrospher
100, RP-18 (5 µm) 250 × 4 mm.
Ch em istr y. Amino acid derivatives 8-10 have been pur-
chased from Bachem. Amino acid conjugates of camptothecin
5-7 and p-isocyanatophenyl-3-O-methyl-â-L-fucopyranoside 13
were synthesized as described earlier.14,15
P r ep a r a tion of Dip ep tid e Con ju ga tes of 20(S)-Ca m p -
toth ecin 12a -h w ith Eith er F r ee or F m oc-P r otected
Sid e Ch a in F u n ction a lity (Gen er a l P r oced u r e). NR-(tert-
Butoxycarbonyl)-protected amino acid derivatives 8-10 (0.047
mol) are dissolved in 800 mL of anhydrous DMF and cooled
to 0 °C. 1-Hydroxybenzotriazole (HOBt) (0.07 mol) and 0.056
mol of N-ethyl-N′-(3-dimethylaminopropyl)carbodiimide hy-
drochloride (EDCI) are added, and the mixture is stirred for
30 min at 0 °C. Subsequently, 0.039 mol of the trifluoroacetate
of the respective 20-hydroxy-linked amino acid conjugate of
camptothecin 5-7 and finally 24.3 mL N-ethyl diisopropyl-
amine are added. The mixture is stirred for 16 h at ambient
temperature. The solvent is evaporated at 25 °C. After addition
of 1.5 L of water the residue is stirred for 15 min and solidifies.
The precipitate is filtered and washed with water. It is
dissolved in 800 mL of dichloromethane, and the remaining
water is removed in a separation funnel. Alternatively, 1.5 L
of dichloromethane is added to the aqueous suspension, and
the product mixture is distributed between the two layers. The
organic layer is concentrated to 150 mL. This solution is
dropped to 2 L of diethyl ether at 0 °C with stirring. The
precipitate is filtered, washed with diethyl ether, and dried
in vacuo. If required, an additional precipitation from dichloro-
methane/methanol with diethyl ether or purification by flash
chromatography using acetonitrile as eluent is also possible.
The resulting protected intermediates from series 11 a -h
are dissolved in 600 mL of dichloromethane, and the solution
is cooled to 5 °C. Two hundred milliliters of trifluoroacetic acid
is added during stirring. The reaction mixture is allowed to
warm to ambient temperature and is stirred for additional 2
h. The solvent and trifluoroacetic acid are evaporated in vacuo
at 25 °C bath temperature. Diethyl ether is added to the
residue, and the mixture is stirred for 10 min. The precipitate
is filtered, washed with diethyl ether, and dried in vacuo
overnight.
F igu r e 5. Temperature dependency of cellular uptake deter-
mined by FACS studies.
into SW480 tumor cells vs liver cells,14 has been
attached to the 20-hydroxy group of camptothecin via
dipeptide spacers.
Such 20-O-linked camptothecin glycoconjugates have
been optimized for enhanced stability of the lactone ring
and for sufficient hydrolytic and proteolytic stability.
The constitution of the peptide spacer has a major
impact on stability and biological activity of the conju-
gates both in vitro and in vivo. A bulky side chain R2 of
the amino acid residue linked to the hydroxy group of
camptothecin significantly increases the stability in the
supernatant of HT29 tumor cells. Glycoconjugates 17-
22, with valine residues at the linkage position to
camptothecin, are sufficiently stable and show good
antitumor activity in vitro against HT29 and other
tumor cell lines, which cannot be fully explained by the
extracellular camptothecin release. The configuration
of this particular R-amino acid residue linked to the
camptothecin moiety does not have a significant impact
on stability and cytotoxic activity in vitro. In contrast,
its stereochemistry has a major impact on the in vivo
activity of the corresponding glycoconjugates in the
20(S)-20-O-[Hist id ylglycyl]ca m p t ot h ecin Bist r iflu or -
a ceta te (12a ). The activated histidine derivative Boc-His(Boc)-

4192 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24
Lerchen et al.
OSu was utilized in the coupling reaction instead of EDCI/
HOBT. Yield: 68% (over two steps); 1H NMR (400 MHz,
CD3OD/CD2Cl2) δ 4.18, 4.45 (2 d, J ) 18 Hz, 2H, CH2 Gly),
4.19 (t, J ) 7 Hz, 1H, R-CH His) 7.22 (s, 1H, imidazole) 7.40
(s, 1H, CPT D-ring), 7.74(dd, J ) 7 Hz, 1H, CPT A-ring), 7.90
(dd, J ) 7 Hz, 1H, CPT A-ring), 8.08 (d, J ) 8 Hz, 1H, CPT
A-ring),), 8.18 (d, J ) 8 Hz, 1H, CPT A-ring), 8.4 (s, 1H,
imidazole), 8.6 (s, 1H, CPT B-ring).
20(S)-20-O-[Nꢀ-(F lu or en yl-9-m et h oxyca r b on yl)lysyl-
glycyl]ca m p t ot h ecin Tr iflu or a cet a t e (12b ). Boc-Lys-
(Fmoc)-OH was used as starting material. Yield: 73% (over
two steps); TLC: Rf ) 0.56 (acetonitrile/water/glacial acetic
acid 5/1/0.2).
20(S)-20-O-[Ar gin ylva lyl]ca m p t ot h ecin Bist r iflu or -
a ceta te (12c). Boc-Arg(Boc)2-OH was used as starting mate-
rial. The product of the coupling reaction was purified by flash
chromatography using ethyl acetate as eluent. Yield: 56%
(over two steps). TLC: Rf ) 0.22 (acetonitrile/water/glacial
acetic acid 10/2.5/1.2).
20(S)-20-O-[Ar gin yl-D-va lyl]ca m p toth ecin Bistr iflu or -
a ceta te (12d ). Boc-Arg(Boc)2-OH was used as starting mate-
rial. The product of the coupling reaction was purified by flash
chromatography using ethyl acetate as eluent. Yield: 56%
(over two steps); TLC: Rf ) 0.1 (acetonitrile/water/glacial
acetic acid 5/1/0.2); 1H NMR (400 MHz, CD3OD/CD2Cl2) δ 3.18
(t, J ) 7 Hz, 2H, Ν-CH2 Arg), 3.88 (t, J ) 6 Hz, 1H, R-CH
Arg), 4.72 (d, J ) 5 Hz, 1H, R-CH Val), 7.41 (s, 1H, CPT
D-ring), 7.73 (dd, J ) 7 Hz, 1H, CPT A-ring), 7.89 (dd, J ) 7
Hz, 1H, CPT A-ring), 8.06 (d, J ) 8 Hz, 1H, CPT A-ring), 8.2
(d, J ) 8 Hz, 1H, CPT A-ring), 8.6 (s, 1H, CPT B-ring).
20(S)-20-O-[H ist id ylva lyl]ca m p t ot h ecin Bist r iflu or -
a ceta te (12e). Boc-His-OH was used as starting material.
Yield: 86% (over two steps); TLC: Rf ) 0.2 (acetonitrile/water/
glacial acetic acid 5/1/0.2); 1H NMR (400 MHz, CD3OD/CD2Cl2)
δ 3.28 and 3.42 (2dd, J ) 7 Hz, 15 Hz; 2H, Ν-CH2 His), 4.68
(d, J ) 5 Hz, 1H, R-CH Val), 7.20 (s, 1H, imidazole), 7.32 (s,
1H, CPT D-ring), 7.72 (dd, J ) 7 Hz, 1H, CPT A-ring), 7.86
(dd, J ) 7 Hz, 1H, CPT A-ring), 8.04 (d, J ) 8 Hz, 1H, CPT
A-ring), 8.16 (d, J ) 8 Hz, 1H, CPT A-ring), 8.45 (s, 1H,
imidazole), 8.56 (s, 1H, CPT B-ring).
20(S)-20-O-[Histid yl-D-va lyl]ca m p toth ecin Bistr iflu or -
a ceta te (12f). Boc-His-OH was used as starting material. The
product of the coupling reaction was purified by flash chro-
matography using acetonitrile/water 20/1 as eluent. Yield:
79% (over two steps); TLC: Rf ) 0.18 (acetonitrile/water/glacial
acetic acid 5/1/0.2).
20(S)-20-O-[Nꢀ-(F lu or en yl-9-m et h oxyca r b on yl)lysyl-
va lyl]ca m p toth ecin Tr iflu or a ceta te (12 g). Boc-Lys(Fmoc)-
OH was used as starting material. Yield: 86% (over two steps);
TLC: Rf ) 0.54 (acetonitrile/water/glacial acetic acid 5/1/0.2);
1H NMR (400 MHz, CD3OD/CD2Cl2 1:1) δ 3.13 (t, J ) 5 Hz,
2H, Ν-CH2 Lys), 4.15 (t, J ) 7 Hz, 1H, CH Fmoc), 4.81 (d, J
) 7 Hz, 2H, CH2 Fmoc), 7.70 (dd, J ) 7 Hz, 1H, CPT A-ring),
7.86 (dd, J ) 7 Hz, 1H, CPT A-ring), 8.02 (d, J ) 8 Hz, 1H,
CPT A-ring), 8.18 (d, J ) 8 Hz, 1H, CPT A-ring), 8.51 (s, 1H,
CPT B-ring).
petrolether, and dried in vacuo. 8.9 g (82%) are obtained; TLC
(acetonitrile/water 10:1): Rf ) 0.7.
P r ep a r a tion of Glycocon ju ga te Hyd r och lor id es of
20(S)-Ca m p toth ecin 14a -h (Gen er a l P r oced u r e). Camp-
tothecin dipeptide conjugates 12a -h (0.027 mol) are dissolved
in 1 L of anhydrous DMF and cooled to 0 °C. N-Ethyl
diisopropylamine (18.9 mL, 0.11 mol) and 9.4 g (0.03 mol) of
p-isothiocyanato-phenyl 3-O-methyl-â-L-fucopyranoside 13 are
added. After stirring for 2 h the reaction mixture is allowed
to warm to room temperature and stirring is continued for 16
h. After evaporating the solvent in vacuo at 25 °C, 1.5 L of
water is added to the residue, and the mixture is stirred for
30 min. During this step the product solidifies. The precipitate
is filtered, washed with water, and dried in vacuo. If required,
the crude product is purified by flash chromatography (aceto-
nitrile/water 30:1). The corresponding fractions are collected,
concentrated, and dissolved in 300 mL of dichloromethane.
Methanol is added dropwise until complete solution occurs.
This solution is added to 1.5 L of diethyl ether and stirred for
30 min. The glycoconjugates 14a -h precipitate and are
filtered, washed with diethyl ether, and dried in vacuo. The
fully deprotected glycoconjugates 14a and 14c-f are ready for
the transformation into corresponding hydrochlorides. From
14b, g, h prior to this the Fmoc group is removed.
For removal of the Fmoc-group, 0.021 mol of the Fmoc-
protected intermediates 14b, g, h is dissolved in 700 mL of
DMF, the solution is cooled to 0 °C, and 36 mL of piperidine
is added. The reaction mixture is allowed to warm to room
temperature, and stirring is continued for 1 h. The solvent is
evaporated, and the residue is dissolved in 200 mL of dichloro-
methane and precipitated with 600 mL of diethyl ether. The
precipitate is collected by filtration, dissolved in 400 mL of
methanol, and stirred for 1 h. Four hundred milliliters of
diethyl ether is added, and stirring is continued for 30 min.
The precipitating glycoconjugates 14b′, g′, h ′ are filtered,
washed with diethyl ether, and then dried in vacuo.
Subsequently, 0.016 mol of the fully deprotected conjugates
14a , 14c-f, and 14b′, g′, h ′ are suspended in 1.4 L of distilled
water. One equiv of a 0.1 M aqueous hydrochloric acid is added
in small portions. The mixture is ultrasonified until complete
solution. It is lyophilized, and the glycoconjugate hydro-
chlorides are obtained as white to yellowish solids.
20(S)-20-O-{NR-[4-(3-O-Meth yl-â-L-fu cop yr a n osyloxy)-
p h en yla m in oth ioca r bon yl]h istid ylglycyl}ca m p toth ecin
Hyd r och lor id e (15). Yield: 77% (over two steps); TLC: Rf
) 0.33 (acetonitrile/water/glacial acetic acid 5/1/0.2); 1H NMR
(400 MHz, CD3OD/CD2Cl2) δ 1.03 (t, J ) 7 Hz, 3H, CH3 CPT),
1.31 (d, J ) 6 Hz, 3H, CH3 Fuc), 3.1 (dd, J ) 6 Hz, J ) 15 Hz,
1H, Ν-CH2 His), 3.23 (dd, J ) 3 Hz, J ) 10 Hz, 1H, 3-CH
Fuc), 3.4 (dd, J ) 6 Hz, J ) 15 Hz, 1H, Ν-CH2 His), 3.5 (s,
3H, OCH3 Fuc), 3.72 (dq, J ) 1 Hz, J ) 5 Hz, 1H, 5-CH Fuc),
3.80 (dd, J ) 9 Hz, J ) 8 Hz, 1H, 2-CH Fuc), 3.9 (d, J ) 3 Hz,
1H, 4-CH Fuc), 4.82 (d, J ) 8 Hz, 1H, 1-CH Fuc), 7.0 (d, J )
9 Hz, 2H, oxy-phenylamino-), 7.2 (s, 1H, imidazole), 7.18 (d, J
) 9 Hz, 2H, oxy-phenylamino-), 7.41 (s, 1H, CPT D-ring), 7.73
(dd, J ) 7 Hz, 1H, CPT A-ring), 7.89 (dd, J ) 7 Hz, 1H, CPT
A-ring), 8.05 (d, J ) 8 Hz, 1H, CPT A-ring), 8.18 (d, J ) 8 Hz,
1H, CPT A-ring), 8.54 (s, 1H, CPT B-ring), 8.6 (s, 1H,
20(S)-20-O-[Nꢀ-(F lu or en yl-9-m eth oxyca r bon yl)lysyl-D-
va lyl]ca m p toth ecin Tr iflu or a ceta te (12h ). Boc-Lys(Fmoc)-
OH was used as starting material. Yield: 55% (over two steps);
TLC: Rf ) 0.60 (acetonitrile/water/glacial acetic acid 5/1/0.2).
p -Isot h iocya n a t op h en yl 3-O-Met h yl-â-L-fu cop yr a n o-
sid e (13). p-Aminophenyl 3-O-methyl-â-L-fucopyranoside (9.43
g, 0.035 mol) is dissolved in 0.9 L of dioxane/water 1:1 (v/v).
Thiophosgene (3.7 mL, 0.048 mol) is added, and the mixture
is stirred for 10 min at ambient temperature. N-Ethyl diiso-
propylamine (35 mL, 0.2 mol) is added, and the mixture is
stirred for an additional 15 min. Dioxane is evaporated in
vacuo at a bath temperature of maximum 25 °C. The aqueous
layer is extracted twice with 200 mL of dichloromethane. The
combined dichloromethane layers are washed twice with 200
mL of water, dried upon sodium sulfate, and evaporated. The
residue is stirred with 200 mL of diethyl ether. After 10 min
200 mL of petrolether is added, and the mixture is stirred for
additional 10 min. The precipitate is filtered, washed with
imidazole); MS (FAB) m/e 854 (M + H)+. Anal. (C42H43N7O11
× HCl × 4H2O) C, H, N, Cl.
S
20(S)-20-O-{NR-[4-(3-O-Meth yl-â-L-fu cop yr a n osyloxy)-
ph en yla m in oth ioca r bon yl]lysylglycyl}ca m p toth ecin Hy-
d r och lor id e (16). Yield: 67% (over three steps); TLC: Rf )
0.25 (acetonitrile/water/glacial acetic acid 5/1/0.2); 1H NMR
(400 MHz, CD3OD/CD2Cl2) δ 1.01 (t, J ) 7 Hz, 3H, CH3 CPT),
1.31 (d, J ) 6 Hz, 3H, CH3 Fuc), 2.86 (t, J ) 7 Hz, 2H, Ν-CH2
Lys), 3.23 (dd, J ) 3 Hz, J ) 9 Hz, 1H, 3-CH Fuc), 3.5 (s, 3H,
OCH3 Fuc), 3.73 (dq, J ) 1 Hz, J ) 5 Hz, 1H, 5-CH Fuc), 3.8
(dd, J ) 9 Hz, J ) 8 Hz, 1H, 2-CH Fuc),), 3.9 (d, J ) 3 Hz, 1H,
4-CH Fuc), 4.82 (d, J ) 8 Hz, 1H, 1-CH Fuc), 7.02 (d, J ) 9
Hz, 2H, oxy-phenylamino-), 7.23 (d, J ) 9 Hz, 2H, oxy-
phenylamino-), 7.41 (s, 1H, CPT D-ring), 7.72 (dd, J ) 7 Hz,
1H, CPT A-ring), 7.89 (dd, J ) 7 Hz, 1H, CPT A-ring), 8.06
(d, J ) 8 Hz, 1H, CPT A-ring), 8.2 (d, J ) 8 Hz, 1H, CPT

20-O-Linked Camptothecin Glycoconjugates
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24 4193
A-ring), 8.6 (s, 1H, CPT B-ring); MS (ESI) m/e 845 (M + H)+.
Anal. (C42H48N6O11S × HCl × 4H2O) C, H, N, Cl.
Val), 1.18 (d, J ) 6 Hz, 3H, CH3 Fuc), 3.27 (dd, J ) 3 Hz, J )
10 Hz, 1H, 3-CH Fuc), 2.83 (t, J ) 7 Hz, 2H, N-CH2 Lys), 3.4
(s, 3H, OCH3 Fuc), 3.66 (dd, J ) 10 Hz, J ) 8 Hz, 1H, 2-CH
Fuc), 3.74 (dq, J ) 1 Hz, J ) 6 Hz, 1H, 5-CH Fuc), 3.91 (d, J
) 3 Hz, 1H, 4-CH Fuc), 4.55 (d, J ) 5 Hz, 1H, R-CH Val), 4.83
(d, J ) 8 Hz, 1H, 1-CH Fuc), 7.02 (d, J ) 9 Hz, 2H,
oxy-phenylamino-), 7.21 (d, J ) 9 Hz, 2H, oxy-phenylamino-),
7.3 (s, 1H, CPT D-ring), 7.70 (dd, J ) 7 Hz, 1H, CPT A-ring),
7.87 (dd, J ) 7 Hz, 1H, CPT A-ring), 8.03 (d, J ) 8 Hz, 1H,
CPT A-ring), 8.12 (d, J ) 8 Hz, 1H, CPT A-ring), 8.58 (s, 1H,
20(S)-20-O-{NR-[4-(3-O-Met h yl-â-L-fu cop yr a n osyloxy)-
ph en ylam in oth iocar bon yl]ar gin ylvalyl}cam ptoth ecin Hy-
d r och lor id e (17). Yield: 83%; TLC: Rf ) 0.3 (acetonitrile/
water/glacial acetic acid 5/1/0.2); 1H NMR (400 MHz; CD3OD/
CD2Cl2) δ 0.95 (t, J ) 7 Hz, 3H, CH3 CPT), 1.32 (d, J ) 6 Hz,
3H, CH3 Fuc), 3.23 (dd, J ) 3 Hz, J ) 10 Hz, 1H, 3-CH Fuc),
3.51 (s, 3H, OCH3 Fuc), 3.74 (dq, J ) 1 Hz, J ) 5 Hz, 1H,
5-CH Fuc), 3.81 (dd, J ) 9 Hz, J ) 8 Hz, 1H, 2-CH Fuc), 3.9
(d, J ) 3 Hz, 1H, 4-CH Fuc), 4.83 (d, J ) 8 Hz, 1H, 1-CH Fuc),
7.4 (d, J ) 9 Hz, 2H, oxy-phenylamino-), 7.26 (d, J ) 9 Hz,
2H, oxy-phenylamino-), 7.37 (s, 1H, CPT D-ring), 7.72 (dd, J
) 7 Hz, 1H, CPT A-ring), 7.86 (dd, J ) 7 Hz, 1H, CPT A-ring),
8.06 (d, J ) 8 Hz, 1H, CPT A-ring), 8.18 (d, J ) 8 Hz, 1H,
CPT A-ring), 8.56 (s, 1H, CPT B-ring); MS (FAB) m/e 915 (M
+ H)+. Anal. (C45H54N8O11S × HCl × 5H2O) C, H, N, Cl.
CPT B-ring); MS (FAB) m/e 887 (M + H)+. Anal. (C45H54N6O11
× HCl × 3H2O) C, H, N, Cl.
S
20(S)-20-O-{NR-[4-(3-O-Met h yl-â-L-fu cop yr a n osyloxy)-
p h en yla m in ot h ioca r b on yl]-lysyl-D-va lyl}ca m p t ot h ecin
Hyd r och lor id e (22). Yield: 53%; TLC: Rf ) 0.30 (aceto-
nitrile/water/glacial acetic acid 5/1/0.2); [R]22 -74.8° (c 0.12
D
DMF); 1H NMR (400 MHz; CD3OD/CD2Cl2) δ 1.0 (t, J ) 7 Hz,
3H, CH3 CPT), 1.06 (d, J ) 7 Hz, 6H, CH3 Val), 1.3 (d, J ) 6
Hz, 3H, CH3 Fuc), 2.86 (t, J ) 7 Hz, 2H, N-CH2 Lys), 3.24
(dd, J ) 3 Hz, J ) 10 Hz, 1H, 3-CH Fuc), 3.5 (s, 3H, OCH3
Fuc), 3.72 (dq, J ) 1 Hz, J ) 6 Hz, 1H, 5-CH Fuc), 3.8 (dd, J
) 10 Hz, J ) 8 Hz, 1H, 2-CH Fuc), 3.90 (d, J ) 3 Hz, 1H,
4-CH Fuc), 4.81 (d, J ) 8 Hz, 1H, 1-CH Fuc), 6.9 (d, J ) 9 Hz,
2H, oxy-phenylamino-), 7.08 (d, J ) 9 Hz, 2H, oxy-phenyl-
amino-), 7.4 (s, 1H, CPT D-ring), 7.70 (dd, J ) 7 Hz, 1H, CPT
A-ring), 7.86 (dd, J ) 7 Hz, 1H, CPT A-ring), 8.02 (d, J ) 8
Hz, 1H, CPT A-ring), 8.2 (d, J ) 8 Hz, 1H, CPT A-ring), 8.56
(s, 1H, CPT B-ring). Anal. (C45H54N6O11S × HCl × 4H2O) C,
H, N, Cl.
20(S)-20-O-{NR-[4-(3-O-Met h yl-â-L-fu cop yr a n osyloxy)-
p h en yla m in oth ioca r bon yl]a r gin yl-D-va lyl}ca m p toth ecin
Hyd r och lor id e (18). Yield: 77%; TLC: Rf ) 0.27 (aceto-
nitrile/water/glacial acetic acid 5/1/0.2); 1H NMR (400 MHz;
CD3OD/CD2Cl2) δ 1.01 (t, J ) 7 Hz, 3H, CH3 CPT), 1.32 (d, J
) 6 Hz, 3H, CH3 Fuc), 3.24 (dd, J ) 3 Hz, J ) 10 Hz, 1H,
3-CH Fuc), 3.50 (s, 3H, OCH3 Fuc), 3.72 (dq, J ) 1 Hz, J ) 5
Hz, 1H, 5-CH Fuc), 3.8 (dd, J ) 9 Hz, J ) 8 Hz, 1H, 2-CH
Fuc), 3.9 (d, J ) 3 Hz, 1H, 4-CH Fuc), 4.82 (d, J ) 8 Hz, 1H,
1-CH Fuc), 6.9 (d, J ) 9 Hz, 2H, oxy-phenylamino-), 7.13 (d,
J ) 9 Hz, 2H, oxy-phenylamino-), 7.4 (s, 1H, CPT D-ring), 7.68
(dd, J ) 7 Hz, 1H, CPT A-ring), 7.84 (dd, J ) 7 Hz, 1H, CPT
A-ring), 8.0 (d, J ) 8 Hz, 1H, CPT A-ring), 8.19 (d, J ) 8 Hz,
1H, CPT A-ring), 8.52 (s, 1H, CPT B-ring). Anal. (C45H54N8O11S
× HCl × 6H2O) C, H, N, Cl.
Deter m in a tion of Glycocon ju ga te Sta bility by HP LC
An a lysis. La cton e Rin g Sta bility. Topotecan 4 and glyco-
conjugates 19 and 21 were dissolved in acetonitrile/water
1:1 at concentrations of 1 mg/10 mL. Subsequently 2 equiv of
1 N sodium hydroxide was added. After 1 h, lactone ring
opening was quantified by HPLC analysis (UV detetection at
257 nm).
Ca m p toth ecin Relea se in th e Su p er n a ta n t of HT29.
HT29 tumor cells (200 000 cells/mL) were grown in RPMI 1640
medium supplemented with 10% FCS for 24 h. Aqueous
solutions of the glycoconjugates 15-22 in concentrations of
10 µM were added and incubated for another 6 and 24 h,
respectively. Fifty microliters of the cellular supernatant was
injected into HPLC using an RP18 (5 µm) column with 70%
HClO4/water (0.4% v/v) as eluent A and acetonitrile as eluent
B (UV detection at 365 nm). Both glycoconjugates and released
camptothecin were quantified by UV detection at 356 nm.
Biology. MTT Assa y for Deter m in a tion of Cytotoxic
Activity Aga in st Tu m or Cell Lin es. The human colon
cancer cell lines SW480 and HT29 and the murine melanoma
cell line B16F10 were grown in RPMI 1640 medium supple-
mented with 10% FCS. Confluent cells were trypsinized and
resuspended in RPMI at 50 000 cells/mL (SW480 and HT29)
and 20 000 cells/mL (B16F10). One hundred microliters of
these cell suspensions was transferred to each well of a 96 well
microtiter plate and incubated for 1 day. Subsequently, 100
µL of RPMI and 1 µL of a solution of camptothecin derivatives
either in DMSO or water were added. Growth inhibition was
assessed 6 days after incubation using the MTT assay. Twenty
microliters of MTT solution (3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide) of a starting concentration of 5
mg per mL of H2O was added and incubated at 37 °C for 5 h.
Subsequently, the medium was removed, 100 µL of 2-propanol
per well was added, and 30 min later an additional 100 µL of
H2O was added. The extinction was measured at 595 nm using
an ELISA-Multiplate Reader.
Deter m in a tion of th e in Vivo Activity in th e Br ea st
Ca n cer Mod el MX-1. Six to eight week old athymic nude mice
(NMRI nu/nu strain) grown in the breeding facilities of the
Drug Development Laboratory Oncotest GmbH, Freiburg,
Germany were used. Human tumors established in serial
passage in nude mice were implanted subcutaneously in both
flanks of female animals. Treatment was started as soon as
the tumors reached a diameter of 5-7 mm depending on the
doubling time. Mice were randomly assigned to treatment
20(S)-20-O-{NR-[4-(3-O-Met h yl-â-L-fu cop yr a n osyloxy)-
p h en yla m in ot h ioca r b on yl]h ist id ylva lyl}ca m p t ot h ecin
Hyd r och lor id e (19). Yield: 78%; mp 180 °C (dec); TLC: Rf
) 0.27 (dichloromethane/methanol/ammonia 17% 15/2/0.2);
[R]22D -41.5° (c 0.2 DMF); 1H NMR (400 MHz; CD3OD/CD2Cl2)
δ 1.0 (t, J ) 7 Hz, 3H, CH3 CPT), 0.92 and 1.08 (2d, J ) 7 Hz,
6H, CH3 Val), 1.33 (d, J ) 6 Hz, 3H, CH3 Fuc), 3.24 (dd, J )
3 Hz, J ) 10 Hz, 1H, 3-CH Fuc), 3.23 and 3.42 (2dd, J ) 6
Hz, J ) 15 Hz, 2H, CH2 His), 3.50 (s, 3H, OCH3 Fuc), 3.75
(dq, J ) 1 Hz, J ) 6 Hz, 1H, 5-CH Fuc), 3.8 (dd, J ) 9 Hz, J
) 8 Hz, 1H, 2-CH Fuc), 3.91 (d, J ) 3 Hz, 1H, 4-CH Fuc), 4.85
(d, J ) 8 Hz, 1H, 1-CH Fuc), 7.05 (d, J ) 9 Hz, 2H,
oxy-phenylamino-), 7.2 (s, 1H, imidazole), 7.21 (d, J ) 9 Hz,
2H, oxy-phenylamino-), 7.36 (s, 1H, CPT D-ring), 7.72 (dd, J
) 7 Hz, 1H, CPT A-ring), 7.86 (dd, J ) 7 Hz, 1H, CPT A-ring),
8.05 (d, J ) 8 Hz, 1H, CPT A-ring), 8.14 (d, J ) 8 Hz, 1H,
CPT A-ring), 8.52 (s, 1H, CPT B-ring), 8.6 (s, 1H, imidazole);
MS (FAB) m/e 896 (M + H)+. Anal. (C45H49N7O11S × HCl ×
4H2O) C, H, N, Cl.
20(S)-20-O-{NR-[4-(3-O-Met h yl-â-L-fu cop yr a n osyloxy)-
ph en ylam in oth iocar bon yl]h istidyl-D-valyl}cam ptoth ecin
Hyd r och lor id e (20). Yield: 75%; TLC: Rf ) 0.36 (aceto-
nitrile/water/glacial acetic acid 5/1/0.2); 1H NMR (400 MHz;
CD3OD/CD2Cl2) δ 1.1 (t, J ) 7 Hz, 3H, CH3 CPT), 1.05 and
1.06 (2d, J ) 7 Hz, 6H, CH3 Val), 1.31 (d, J ) 6 Hz, 3H, CH3
Fuc), 3.24 (dd, J ) 3 Hz, J ) 10 Hz, 1H, 3-CH Fuc), 3.1 and
3.32 (2dd, J ) 7 Hz, J ) 15 Hz, 2H, CH2 His), 3.51 (s, 3H,
OCH3 Fuc), 3.72 (dq, J ) 1 Hz, J ) 6 Hz, 1H, 5-CH Fuc), 3.81
(dd, J ) 9 Hz, J ) 8 Hz, 1H, 2-CH Fuc), 3.9 (d, J ) 3 Hz, 1H,
4-CH Fuc), 4.8 (d, J ) 8 Hz, 1H, 1-CH Fuc), 6.88 (d, J ) 9 Hz,
2H, oxy-phenylamino), 7.01 (d, J ) 9 Hz, 2H, oxy-phenylamino-
), 7.2 (s, 1H, imidazole), 7.41 (s, 1H, CPT D-ring), 7.71 (dd, J
) 7 Hz, 1H, CPT A-ring), 7.84 (dd, J ) 7 Hz, 1H, CPT A-ring),
8.03 (d, J ) 8 Hz, 1H, CPT A-ring), 8.18 (d, J ) 8 Hz, 1H,
CPT A-ring), 8.55 (s, 1H, CPT B-ring), 8.61 (s, 1H, imidazole).
Anal. (C45H49N7O11S × HCl × 5H2O) C, H, N.
20(S)-20-O-{NR-[4-(3-O-Met h yl-â-L-fu cop yr a n osyloxy)-
p h en yla m in oth ioca r bon yl]lysylva lyl}ca m p toth ecin Hy-
d r och lor id e (21). Yield: 63%; mp 200 °C (dec); TLC: Rf )
0.36 (acetonitrile/water/glacial acetic acid 5/1/0.2); [R]22D -40.6°
(c 0.17 DMF); 1H NMR (400 MHz; CD3OD/CD2Cl2) δ 0.93 (t, J
) 7 Hz, 3H, CH3 CPT), 0.9 and 1.0 (2d, J ) 7 Hz, 6H, CH3

4194 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24
Lerchen et al.
groups and control groups (5 mice per group bearing 8-10
evaluable tumors). Tumor size was measured twice weekly by
two-dimensional measurement with calipers. Volumes were
calculated according to the formula a × b2/2 with a and b
representing two perpendicular tumor diameters. Relative
tumor volume values were calculated for each single tumor
by dividing the tumor size day × by the tumor size day 0 at
the day of randomization. Median RTV values were used for
further evaluation. Tumor volume inhibition (volume of test
group/control group T/C) was the main end point. Furthermore,
body weight changes and mortality were assessed as param-
eters for toxicity. Glycoconjugates were dissolved in water (c
) 1-3 mg/mL) and were administered daily for the first 3 days
intravenously.
Deter m in a tion of Hem op oietic Activity (Colon y F or -
m a tion of P r ogen itor Cells). C57bl/6 mice (n ) 3) were
treated intraperitoneally once a day for 3 consecutive days with
the glycoconjugates 19 or 21 in comparison to topotecan 4 at
different doses in the range of the respective MTDs calculated
from the nude mice models. At 4 and 24 h after the last
treatment, bone marrow cells were isolated from both femurs
and cultured (2 × 105/culture) on semisolid, CSF-supplemented
medium. After 7 days incubation (37 °C, 5.5% CO2) colony
formation (number of colonies with >50 cells/colony), indicat-
ing proliferation and differentiation of myeloid progenitor stem
cells, was determined.
Deter m in a tion of Cytotoxicity Aga in st Hep a tocytes.
Liver was first perfused with P-I (perfusion buffer I, consisting
of ethylene glycol-bis(â-aminoethyl ether)-N,N,N′,N′-tetra-
acetic acid (EGTA, 0.5 mM) and HEPES (10 mM) in Hank’s
Balanced Salt Solution (HBSS, Gibco, without Ca2+ and Mg2+))
at a flow rate of 20-30 mL/(min × cannula) for 20 min and
subsequently with P-II (perfusion buffer II, consisting of
HEPES (10 mM, pH 7.4) in HBSS, CaCl2 (5.0 mM), and 1 mg/
mL collagenase type IV). Cells were suspended in HBSS and
centrifuged at 80 g for 3 min. After washing, cells were counted
in a hemocytometer after staining with trypan blue. Viability
was estimated to be about 80-90%.
Hepatocytes were cultured in collagen sandwich gels ac-
cording to described methods.18,19 Briefly, 6-well plates
(Greiner, Solingen) were coated with 1.0 mL per well of rat
tail tendon collagen dissolved in William’s Medium E.20
Hepatocytes (2 × 106) were seeded into each well and allowed
to attach overnight. After removal of the unattached cells, the
cells were coated with 1 mL of the same collagen solution
which was previously used to prepare the lower part of the
sandwich. Culture medium (2.0 mL) was added (supplemented
with insulin [0.16 U/mL], glucagon [0.014 µg/mL], prednisolone
[9.6 µg/mL], glutamine [1%], penicillin/streptomycin [200
U/mL], and 10% fetal calf serum) and renewed daily. After 48
h of adaptation cells were incubated with the test compounds
at the indicated concentrations for up to 7 days.
resuspended in PBS at 2 × 106 cells/mL. Flow cytometric
measurements were performed using a FACStarPlus, Becton
Dickinson, Ca., U.S.A. The cells were excited at 488 nm using
an argon ion laser. The emitted fluorescence was measured
using a 530 BP30 band-passfilter (Standard green fluores-
cence). Viability staining was done by using propidiumiodide
dye (Molecular Probes, NL).
Ack n ow led gm en t. We thank Anna-Maria DiBetta,
Andrea Felder, Xanthula Kavazoudi, Cornelia Steidle,
Dirk Wolter, Manfred Hoffmann, and Michael Kazinski
for their skillful technical assistance. Furthermore we
thank Dr. Eberhard Kuckert for providing the fluores-
cence marker, Drs. Horst Antonicek and Walter Weichel
for their contributions to the cellular uptake studies,
and Drs. Daniel Gondol, Thomas Fa¨cke, and J oachim
Wesener for performing NMR and mass spectrometric
analysis.
Refer en ces
(1) Wall, M. E.; Wani, M. C.; Cook, C. E.; Palmer, K. H.; McPhail,
A. T.; Sim, G. A. Plant antitumor agents. I. The isolation and
structure of camptothecin, a novel alkaloidal leukemia and
tumor inhibitor from Camptotheca acuminata. J . Am. Chem. Soc.
1966, 88, 3888-3890.
(2) Moertel, C. G.; Schutt, A. J .; Reitemeier, R. J .; Hahn, R. G. Phase
II study of camptothecin (NSC-100880) in the treatment of
advanced gastrointestinal cancer. Cancer Chemother. Rep. 1972,
56, 95-101.
(3) Sawada, S.; Okajima, S.; Aiyama, R.; Nokata, K.; Furuta, T.;
Yokokura, T.; Sugino, E. Yamaguchi, K.; Miyasaka, T. Synthesis
and antitumor activity of 20(S)-camptothecin derivatives:
Carbamate-linked, water soluble derivatives of 7-ethyl-10-
hydroxycamptothecin. Chem. Pharm. Bull. 1991, 39, 1446-
1454.
(4) Kingsbury, W. D.; Boehm, J . C.; J akas, D. R.; Holden, K. G.;
Hecht, S. M.; Gallagher, G.; Caranfa, M. J .; McCabe, F. L.;
Faucette, L. F.; J ohnson, K. J .; Hertzberg, R. P. Synthesis of
water-soluble (aminoalky)camptothecin analogues: Inhibition of
topoisomerase I and antitumor activity. J . Med. Chem. 1991,
34, 98-107.
(5) Hsiang, Y. H.; Hertzberg, R.; Hecht, S.; Liu, L. F. Camptothecin
induces protein-linked DNA breaks via mammalian DNA topo-
isomerase I. J . Biol. Chem. 1985, 260, 14873-14878.
(6) J axel, C.; Kohn, K. W.; Wani, M. C.; Wall, M. E. Structure
activity study of the actions of camptothecin derivatives on
mammalian topoisomerase I, evidence for a specific receptor site
and for a relation to antitumor activity. Cancer Res. 1989, 49,
1465-1469.
(7) Burke, T. G.; Mi, Z. Preferential binding of the carboxylate form
of camptothecin by human serum albumin. Anal. Biochem. 1993,
212, 285-287.
(8) Liehr, J . G.; Ahmed, A. E.; Giovanella, B. C. Pharmacokinetics
of camptothecins administered orally. In The Camptothecins;
From discovery to the patient; Ann. N.Y. Acad. Sci., 1996; pp
157-163.
Stock solutions of test compounds in DMSO (3 mM) were
further diluted in culture medium to the desired concentration.
The primary hepatocytes were treated for 1-7 days, and the
cell medium containing the test compound was changed every
day.
The cytotoxic effects of the compound on the cells were
determined by measuring release of aspartate aminotrans-
ferase (AST), alanine aminotransferase (ALT), and lactate
dehydrogenase (LDH) in the cell supernatants.
(9) Lerchen, H.-G. Camptothecin antitumor agents. IDrugs 1999,
2, 896-906.
(10) Cao, C.; Harris, N.; Kozielski, A.; Vardeman, D.; Stehlin, J . S.;
Giovanella, B. Alkyl esters of camptothecin and 9-nitro-
camptothecin: Synthesis, in vitro pharmacokinetics, toxicity and
antitumor activity. J . Med. Chem. 1998, 41, 31-37.
(11) Vishnuvajjala, B. R.; Cradock, J . C.; Garzon-Aburbeh, A. Water
soluble camptothecin prodrugs. Pharm. Res. 1986, 3, 22S.
(12) Conover, C. D.; Greenwald, R. B.; Pendri, A.; Gilbert, K. W.;
Shum, K. L. Camptothecin delivery systems. Enhanced efficacy
and tumor accumulation of camptothecin following its conjuga-
tion to poly(ethylene glycol) via a glycine linker. Cancer Chemo-
ther. Pharmacol. 1998, 42, 407-414.
In addition, the mitochondrial dehydrogenase activity was
measured in an MTTassay.
Con foca l La ser Sca n F lu or escen ce Micr oscop y. HT29
cells were grown for 48 h on coverslips and then incubated
with test compounds for 6 and 24 h at temperatures of 37 °C.
The camptothecin glycoconjugate 19 was tested at 100 µM,
whereas the fluorescence dye conjugate 23 was tested at 0.5
µM and the fluorescence dye 24 at 0.1 µM. The confocal images
were generated using an inverted LSM 410 UV, Zeiss, Ger-
many.
(13) Caiolfa, V. R.; Zamai, M.; Fiorino, A.; Frigerio, E.; Pellizzoni,
C.; d'Argy, R.; Ghiglieri, A.; Castelli, M. G.; Farao, M.; Pesenti,
E.; Gigli, M.; Angelucci, F.; Suarato, A. Polymer-bound camp-
tothecin: initial biodistribution and antitumor activity studies.
J . Controlled Release 2000, 65, 105-119.
(14) Lerchen, H.-G.; Baumgarten, J .; Piel, N.; Kolb-Bachofen, V.
Lectin-mediated drug-targeting: Discrimination of carbohydrate
mediated cellular uptake between tumor and liver cells with
neoglycoconjugates carrying fucose epitopes regioselectively
modified in the 3-position. Angew. Chem., Int. Ed. 1999, 38,
3680-3683.
F low Cytom etr y. HT29 cells were incubated with 0.1 µM
of the glycoconjugates for 6 h at either 37 °C or 4 °C. Cells
were isolated with accutase, washed twice with PBS, and

20-O-Linked Camptothecin Glycoconjugates
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24 4195
(15) Lerchen, H.-G.; Von Dem Bruch, K. Synthesis of 20-O-linked
20(S)-camptothecin glycoconjugates: Impact of the side chain
of the ester-linked amino acid on epimerization during the
acylation reaction and on hydrolytic stability of the final
glycoconjugates. J . Prakt. Chem. 2000, 342, 753-760.
(16) Fuller, W. D.; Cohen, M. P.; Shabankareh, M.; Blair, R. K.;
Goodman, M.; Naider, F. R. Urethane-protected amino acid
N-carboxy anhydrides and their use in peptide synthesis. J . Am.
Chem. Soc. 1990, 112, 7414-7416.
Long-term culture sandwich configuration. FASEB 1998, 3,
174-177.
(19) Elsdale, T.; Bard, J . Collagen substrates for studies on cell
behaviour. J . Cell Biol. 1972, 54, 626-637.
(20) Bader, A.; Knop, E.; Kern, A.; Bo¨ker, K.; Fru¨hauf, N.; Crome,
O.; Esselmann, H.; Pape, C.; Kempka, G.; Sewing, K. F. 3-D
coculture of hepatic sinusoidal cells with primary hepatocytes-
design of an organotypical model. Exp. Cell Biol. 1996, 226,
223-233.
(17) Carter, C.; et al. unpublished results.
(18) Dunn, J . C. Y.; Yarmush, M. L.; Koebe, H. G.; Tompkins, R. G.
Hepatocyte function and extracellular matrix geometry:
J M010893L