Synthesis and evaluation of Taxol-folic acid conjugates as targeted antineoplastics

Article abstract of DOI:10.1016/S0968-0896(02)00019-6

A series of Taxol derivatives tethered at C2′ and C-7 to glutamate and folate have been synthesized for evaluation as prodrugs which release Taxol via hydrolytic lability of their α-alkoxy and α-amino esters. The half-time for hydrolysis of these materials was determined in pH 7 and pH 5 buffer. The in vitro cytotoxicity has been assessed in cell culture against A-549 lung cancer, MCF-7 breast cancer, and HT-29 colon cancer. Selected agents were further screened for folate binding and competitive binding with free folic acid. One agent (54), further evaluated in animal studies was found to increase the lifespan in mice, but was less effective than Taxol itself.

Full text of DOI:10.1016/S0968-0896(02)00019-6

Bioorganic & Medicinal Chemistry 10 (2002) 2397–2414
Synthesis and Evaluation of Taxol–Folic Acid Conjugates
as Targeted Antineoplasticsy
Jae Wook Lee, June Y. Lu, P. S. Low and P. L. Fuchs*
Department of Chemistry, Purdue University, West Lafayette, IN 47907, USA
Received 4 June 2001; accepted 19November 2001
Abstract—A series of Taxol derivatives tethered at C20 and C-7 to glutamate and folate have been synthesized for evaluation as
prodrugs which release Taxol via hydrolytic lability of their a-alkoxy and a-amino esters. The half-time for hydrolysis of these
materials was determined in pH 7 and pH 5 buffer. The in vitro cytotoxicity has been assessed in cell culture against A-549lung
cancer, MCF-7 breast cancer, and HT-29colon cancer. Selected agents were further screened for folate binding and competitive
binding with free folic acid. One agent (54), further evaluated in animal studies was found to increase the lifespan in mice, but was
less effective than Taxol itself. # 2002 Elsevier Science Ltd. All rights reserved.
Introduction
The successes achieved with high-dose chemotherapy in
the refractory metastatic patient group are driving oncol-
ogists toward application of highly aggressive treatment
regimens at the earliest point of diagnosis. These proto-
cols are limited by drug toxicity and severe physi-
ological effects and patient fatalities are not uncommon.
This situation has caused several members in the medi-
cal community to question whether the benefit/risk
boundary has been exceeded with the agents currently
available.11 Clearly, enhancement of the differential
specificity of anticancer agents by selective targeting
mechanisms could potentially diminish such toxicity.
Although Taxol (paclitaxel) has demonstrated some
extremely encouraging clinical responses,2 its poor
water solubility causes formulation problems and mani-
fests side effects due to the necessity of using cremophor
EL (polyethoxylated castor oil) and ethanol as a vehicle.
Advances in providing more soluble active derivatives
and prodrugs have been recently reported.3 From
structure–activity relationships, it has been established
that the 20- and/or the 7-hydroxyl groups in Taxol are
suitable for linking ester derivatives designed to both
improve water solubility and ultimately foster the
release of Taxol itself.4
One low molecular weight ligand that may avoid many
of the limitations associated with antibody-mediated
targeting12 is folic acid. Folic acid and its reduced
counterparts enter cells via two unrelated pathways.13
Covalent conjugation of folic acid to another molecule
provides a construct that is strongly favored to enter
cells via folate receptor mediated endocytosis. Based on
this advantage, we are exploring the use of folic acid as
a targeting agent for delivery of therapeutic drugs and
imaging agents to tumors. As will be documented
below, folate, when attached via its g-carboxylate to any
molecule, retains its ability to bind to its receptor with
normal affinity and thereby enter receptor-bearing cell
by endocytosis. In vitro and in vivo, the targeting is
highly tumor specific with cell binding constants near
10À10 M. The number of molecules internalized can be
very large (>106 per h), the pathway is nonharmful to
the cell, and entry is via a nondegradative, nonlysosomal
The current cancer treatment protocols of surgery,
radiation, hormone treatment,5 autologous (self-donor)
bone marrow transplant,6 infusions of colony-stimulat-
ing factors,7 and/or interferon8 are routinely supple-
mented by an increasingly aggressive adjuvant program
featuring high-dose multiple drug chemotherapy.9
Details of the high-dose clinical trials have been pub-
lished over the past 5 or 6 years.10 Although exciting
new drugs, such as Taxol will occasionally continue to
be added to the pharmacopoeia, it seems likely that
further progress will only occur in small increments in
the absence of innovative new strategies.
*Corresponding author. Tel.: +1-765-494-5242; fax: +1-765-494-797;
e-mail: pfuchs@purdue.edu
ySee ref 1.
0968-0896/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(02)00019-6

2398
J. W. Lee et al. / Bioorg. Med. Chem. 10 (2002) 2397–2414
pathway. Furthermore, folate is stable during storage,
relatively easy to ligate to other molecules, non-
immunogenic, and sufficiently small in size to not inter-
fere with extravasation and intercellular diffusion.
6, followed by alkylation of amide 7 with trimethylsilyl
bromoacetate. As expected, deprotection of the tri-
fluoroacetyl group of 7 using basic conditions such as
hydrazine and potassium carbonate in allyl alcohol, the
cyclization reaction occurred to give piperazinone 9
(Scheme 2).
In conjunction with our program directed at the design
and synthesis of covalently-tethered folic acid drug
conjugates for targeted delivery of chemotherapeutic
drugs, we elected to initially prepare reagents for the
efficient introduction of tri and tetraallyl esters of
EDTA and DTPA mono carboxylic acids. In this paper
we report the syntheses of Taxol–EDTA and Taxol–
DTPA derivatives and determine their in vitro anti-
neoplastic properties and their aqueous solubility. The
latter property was held to be a prime determinant in
selecting which agents to affix to folic acid for future
testing as targeted anticancer agents.
Because manipulation of 7 was not especially efficient,
we prepared DTPA-mono acid 12 employing the same
strategy employed for EDTA-mono acid 4 (Scheme 3).
Boc mono-protection of diethylene triamine 6 and sub-
sequent trialkylation were carried out to afford com-
pound 10 which was transformed to pentaester 11 by
Boc deprotection and dialkylation with tert-butyl bro-
moacetate. When the alkylation reaction was attempted
after deprotection of compound 10, it was again found
that DMF and excess alkylating agent were required to
reduce the formation of piperazin-2-one derivatives
13a,b. Removal of the tert-butyl groups of 11 with TFA
and the reaction of the resulting dicarboxylic acid with
DCC generated the anhydride intermediate, which was
quenched with allyl alcohol to give DTPA-mono acid
12.
Synthesis of Folate-Linked Taxol Derivatives
As shown Scheme 1, compound 2 was prepared by
mono protection of ethylene diamine 1 using (Boc)2O,
followed by bisalkylation of the residual amine moiety
with allyl bromoacetate in a yield of 86%. Boc depro-
tection of compound 2 using HCl followed by dialkyl-
ation with excess tert-butyl bromoacetate gave
compound 3 in 58% yield. Alkylation of 2 required the
use of DMF and excess alkylating reagent to reduce the
formation of piperazin-2-one compounds 5a and 5b.
When acetonitrile was used as solvent with 2 equiv of
tert-butyl bromoacetate, 5a and 5b were isolated in 20
and 50% yields, respectively. The yield of 5a is partially
reduced from losses in the aqueous wash solution. For-
tunately, using DMF with 5 equiv of tert-butyl bromo-
acetate decreases formation of 5a and 5b to around 5–
10%. The tert-butyl groups of compound 3 were
removed with TFA and the resulting dicarboxylic acid
was reacted with DCC to afford the anhydride inter-
mediate, followed by quenching with allyl alcohol to
give the desired EDTA mono acid 4 in 57% yield.
To achieve the preparation of C7-substituted derivatives
of Taxol, it was necessary to first protect the more
Scheme 2. Synthesis of DTTA mono acid. Reagents and conditions:
(a) (i) ethyl trifluoroacetate, CH2Cl2, 2 h at 0 ꢀC and then 5 h at 25 ꢀC;
(ii) allyl bromoacetate (3.3 equiv), DIEA, CH3CN, 25 ꢀC, 20 h, 52%;
(b) NaH, DMF, trimethylsilyl bromoacetate, 25 ꢀC, 12 h, 70% (98%
BRSM).
DTTA mono acid 8 was prepared by the tri-
fluoroacetylation and trialkylation of diethylene triamine
Scheme 1. Synthesis of EDTA mono acid. Reagents and conditions:
(a) (i) (Boc)2O, MeOH, 2 h at 0 ꢀC and then 2 h at 25 ꢀC; (ii) allyl
bromoacetate (2 equiv), DIEA, CH3CN, 25 ꢀC, 20 h, 86%; (b) (i) HCl,
EtOAc, 25 ꢀC, 10 h; (ii) t-butyl bromoacetate (>5 equiv), DIEA,
DMF, 25 ꢀC, 12 h, 58%; (c) (i) TFA, CH2Cl2, 25 ꢀC, 10 h; (ii) DCC,
CH2Cl2, 25 ꢀC, 5 h; (iii) allyl alcohol, Et3N, DMAP (cat.), 25 ꢀC, 15 h,
57%.
Scheme 3. Synthesis of DTPA mono acid. Reagents and conditions:
(a) (i) (Boc)2O, MeOH, 2 h at 0 ꢀC and then 10 h at 25 ꢀC; (ii) allyl
bromoacetate (3.3 equiv), DIEA, CH3CN, 25 ꢀC, 20 h, 52%; (b) (i)
HCl, EtOAc, 25 ꢀC, 10 h; (ii) t-butyl bromoacetate (>5 equiv), DIEA,
DMF, 25 ꢀC, 12 h, 60%; (c) (i) TFA, CH2Cl2, 25 ꢀC, 10 h, (ii) DCC,
CH2Cl2, 25 ꢀC, 5 h; (iii) allyl alcohol, Et3N, DMAP (cat.), 25 ꢀC, 15 h,
68%.

J. W. Lee et al. / Bioorg. Med. Chem. 10 (2002) 2397–2414
2399
reactive C-20 hydroxyl moiety. The alloc group was
selected as protecting group for the 20-OH of Taxol
which was projected to be deprotected at the end of the
synthesis using Pd [0] catalyzed cleavage of a collection
of allyl esters. 20-Alloc Taxol 15 was obtained by
the reaction of Taxol 14 with allyl chloroformate in
methylene chloride in the presence of DIEA in 98%
yield (Scheme 4).
In order to evaluate the previously prepared poly-
aminocarboxylates as cleavable linkers, we undertook
the reaction of 20-alloc Taxol 15 with EDTA-mono acid
4 and DTPA-mono acid 12. Reaction in methylene
chloride in the presence of DCC with catalytic DMAP
provided the fully protected Taxol–EDTA 16 and
Taxol–DTPA 18 in yields of 94–95%. Deprotection
of the entire collection of allyl groups was initially
assessed with compound 16 using 10% Pd(PPh3)4 and
Et2NH (40 equiv)14 in CH2C12. Unfortunately, these
conditions provided large amounts of transamidated
materials resulting from the excess diethylamine.
15
Substitution of PhSiH3 (1–2 equiv) for the diethyl-
amine afforded excellent results. Taxol–EDTA 17 and
Taxol–DTPA 19 were smoothly obtained in 70 and
75% yield, respectively, from deprotection of com-
pounds 16 and 18 using Pd(PPh3)4 and PhSiH3 in
CH2Cl2. The aqueous solubility of Taxol–EDTA 17
and Taxol–DTPA 19 was estimated using HPLC
methodology16 and was calculated to be 0.19and 0.23
mg/mL, respectively.
Scheme 4. Reagents and conditions: (a) allyl chioroformate, DIEA,
CH2Cl2, 25 ꢀC, 10 h, 98%; (b) EDTA mono acid 4, DCC, DMAP,
CH2Cl2, 25 ꢀC, 10 h, 95%; (c) DTPA mono acid 12, DCC, DMAP,
CH2Cl2, 25 ꢀC, 10 h, 94%; (d) PhSiH3, Pd(PPh3)4, CH2Cl2, 25 ꢀC, 1 h.
In order to provide substantially increased water solu-
bility and to anticipate the need for the hydrolytic
release of the drug, we elected to investigate inductively
Scheme 5. Reagents and conditions: (a) anhydride 21 or 22, CH2Cl2, 25 ꢀC, 36 h; (b) PEG-carboxylic acid 23 or 24, DCC, DMAP, CH2Cl2, 25 ꢀC,
10 h; (c) SnCl2, PhSH, Et3N, CH3CN, 25 ꢀC, 30 min; (d) DTPA dianhydride 29, DMSO, 25 ꢀC, 3 h; (e) Et2NH, Pd(PPh3)4, CH2C12, 25 ꢀC, 30 min.

2400
J. W. Lee et al. / Bioorg. Med. Chem. 10 (2002) 2397–2414
Scheme 6. Reagents and conditions: (a) DCC, DMAP, CH2Cl2, 15 h, 25 ꢀC; (b) PPh3, THF (H2O), 20 h, 25 ꢀC; (c) 36 (10 equiv), isopropanol, 5 h,
reflux; (d) Pd(PPh3)4, CH2Cl2, 15 mm, 25 ꢀC; (e) PteN3, MTBD, DMSO, 2 h, 25 ꢀC.
activated a-alkoxy and a-amino esters as the C-7 con-
necting function. We also added a short amine termi-
nated PEG linker to improve the water solubility and
facilitate amide bond formation. The easily prepared
azido-PEG-amine 2017,18 was reacted with anhydrides
21 and 22 to provide PEG-carboxylic acids 23 and 24 in
>98% yield. Coupling these acids with 20-alloc Taxol
15 in the presence of DCC and DMAP afforded 20-alloc
Taxol–PEG azides 25 and 26 in>94% yield. Reduction
of azides 25 and 26 was smoothly performed using
SnCl2/PhSH/Et3N18 to give primary amines 27 and 28
in 90 and 97% yields, respectively. These substrates
were reacted with excess DTPA dianhydride 29 to
afford 20-alloc Taxol–PEG-DTPA esters 30 and 31 in 80
and 85% yield. Deprotection of the 20-alloc groups on
30 and 31 using Pd(PPh3)4 and Et2NH in CH2Cl2
yielded Taxol–PEG-DTPA polyacids 32 and 33 in 95
and 90% yields, respectively. The aqueous solubility of
Taxol–PEG-DTPA 32 and 33 were determined to be 25
and 27 mg/mL, respectively (Scheme 5)
38 in 82% yield with concomitant deprotection of the
C20 methoxyacetyl functionality. Taxol–7-carbamoyl-
PEG-glutamic acid 39 was obtained from the deprotec-
tion of 38 using Pd(PPh3)4 and Et2NH in CH2Cl2 in
87% yield. Taxol–7-carbamoyl-PEG-Folate 40 was
generated in 55% yield by reaction of Taxol–7-carb-
amoyl-PEG-glutamic acid 39 and pteroyl azide20 in
DMSO in the presence of MTBD. Unfortunately, com-
pounds 39 and 40 showed no activity (see Table 2) and
the C-7 carbonate group was impervious to in situ
hydrolysis such that Taxol was not released in the aque-
ous environment (see Table 1). While there are C-7
derivatives of Taxol that are active without release,21
folate-tethered compound 40 is not among them (Table
2, Scheme 6).
In an effort to introduce linkers capable of releasing
Taxol under the hydrolytic conditions, it was decided to
employ ester groups as cleavable linkers at the C20 and
C7 alcohols of Taxol. Since the C-20 methoxyacetyl
moiety was readily hydrolyzed, it was decided to retain
similar inductive activation in the tethers to be
employed. Therefore, a-alkoxy and a-amino esters were
utilized as the connecting function in anticipation of
ultimate hydrolytic release of the drug itself.
Based upon the above preliminary results, we decided to
prepare C-7 linked conjugate 40 by coupling of the ver-
satile (and water soluble) amino-PEG-azide 2017,18 with
protected glutamic acid 34 in the presence of DCC and
DMAP (cat.), followed by reduction of the azide moiety
of 35 to afford 36 in an overall yield of 93%. Reaction
of 20-methoxyacetyl-protected-Taxol bearing an acyl-
imidazole at C-7 3719 requires a large excess of amino-
PEG-glutamate 36 in anhydrous isopropyl alcohol at
reflux, but delivers Taxol–7-carbamoyl-PEG-glutamate
Preparation of C20-tethered Taxol–folates are detailed
in Scheme 7. As a point of departure, amino-PEG-glu-
tamate 36 was reacted with anhydrides 21 and 22 to
provide glutamate-PEG-carboxylic acids 41 and 42 in
95 and 94% yields, respectively. Coupling these of

J. W. Lee et al. / Bioorg. Med. Chem. 10 (2002) 2397–2414
2401
Scheme 7. Reagents and conditions: (a) DMAP (cat.), CH2Cl2, 36 h, 25 ꢀC; (b) DIPC, DMAP (cat.), CH2Cl2, 12 h, 25 ꢀC; (c) Pd(PPh3)4, PhSiH3,
CH2Cl2, 1 h, 25 ꢀC; (e) PteN3, (i-Pr)2NEt, DMSO, 44 h, 25 ꢀC.
Table 1. Hydrolysis rate of folates and glutamates at 37 ꢀC
47 and 48 were obtained by the reaction of Taxol–PEG-
t1/2 at 37 ꢀC
glutamic acids 45 and 46 with pteroyl azide in DMSO in
the presence of DIEA in 25 and 30% yields, respectively.
The low yields in the coupling/purification sequence
were partially a reflection of the desired lability.
Compd Compd Linker type
type
Linked
at
pH 7
pH 5
39
40
45
46
47
48
51
52
53
54
61
63
Glu PEG-3
Fol PEG-3
Glu PEG-3+O anyd
C-7
C-7
C-20
130 (none)
130 (none)
300 (none)
300 (none)
0.8
3.3
1
938
32
38
40
10919
144
104
19
20
17
As a complement and comparison to the C20 deriva-
tives, it was deemed prudent to also prepare the corre-
sponding C-7 folate derivatives (Scheme 8). Preparation
of C7-substitued derivatives of Taxol required blocking
the more reactive C-20 hydroxyl moiety. The selected
alloc group was projected to be deprotected at the end
of the synthesis during Pd(0) catalyzed cleavage of all
allyl esters.
Glu PEG-3+NMe anhyd C-20
Fol
PEG-3+O anhyd
PEG-3+NMe anhyd C-20
C-7
Glu PEG-3+NMe anhyd C-7
C-20
Fol
Glu PEG-3+O anhyd
>300
60
>300
Fol
Fol
Fol
Fol
PEG-3+O anhyd
C-7
PEG-3+NMe anhyd C-77
Y PEG-3 azide
Y PEG-3 tetraacid
C-7
C-7
>260
170
Substrate dissolved in phosphate buffer solution (pH 7 or pH 5) by the aid
of DMSO was incubated at 37 ꢀC and analyzed using HPLC.
In order to further evaluate glutamate-PEG-carboxylic
acids 41 and 42 as cleavable linkers, we undertook the
reaction of 20-alloc Taxol 15 (Scheme 4) with glutamate-
PEG-carboxylic acids 41 and 42 in methylene chloride
in the presence of DIPC with catalytic DMAP. This
provided 20-alloc Taxol–7-PEG glutamates 49 and 50 in
52% (94% based upon recovered starting material) and
57% (98% BRSM) yields, respectively (Scheme 8).
Deprotection of the entire collection of allyl groups of
glutamate-PEG-carboxylic acids with taxol 14 in the
presence of DIPC and DMAP afforded Taxol–PEG
glutamates 43 (93%) and 44 (95%). Universal depro-
tection15 of 43 and 44 using Pd(PPh3)4 and PhSiH3 in
CH2Cl2 yielded Taxol–PEG-glutamic acids 45 and 46 in
92 and 93% yields, respectively. Taxol–20-PEG-Folates

2402
J. W. Lee et al. / Bioorg. Med. Chem. 10 (2002) 2397–2414
Scheme 8. Reagents and conditions: (a) DIPC, DMAP (cat.), CH2Cl2, 12 h, 25 ꢀC; (b) Pd(PPh3)4, Et2NH, CH2C12, 1 h, 25 ꢀC; (c) PteN3, i-Pr2NEt,
DMSO, 44 h, 25 ꢀC.
Table 2. In vitro cell culture ED50 (mg/mL) of selected compounds
with Taxol, folate, and the polycarboxylic acid DTPA.
For this purpose, the triply linked system azido-PEG-
iminodiacetic acid anhydride 57 was prepared (Scheme
9). Azido-PEG-amine 20 was reacted with allyl bromo-
acetate in acetonitrile in the presence of K2CO3 to give
iminodiallylester 55 which was subsequently depro-
tected using Pd(PPh3)4 and PhSiH3 in CH2Cl2 to afford
azido-PEG-iminodiacetic acid 56 in 92% yield. Azido-
PEG-iminodiacetic acid 56 was initially treated with
DCC at 25 ꢀC to generate anhydride 57. Addition of
amino-PEG-glutamate 36 to preformed anhydride 57
followed by reaction for an additional 12 h at 25 ꢀC
smoothly provided monoacid 58. Reaction of 20-alloc
Taxol 15 with carboxylic acid 58 in the presence of DCC
and a catalytic amount of DMAP in methylene chloride
for 12 h at 25 ꢀC afforded Y-shaped intermediate 59 in
95% yield after purification. Global deprotection of the
allyl groups of glutamate 59 using Pd(PPh3)4 and
Et2NH in CH2Cl2 generated Y-shaped-glutamic acid 60
in 96% yield. Y-Shaped-folate 61 was obtained in 43%
yield by the reaction of 60 with pteroyl azide in DMSO
in the presence of MTBD. Reduction of azide 61 was
effectively performed with SnCl2/PhSH/Et3N in DMF
to give amine 62 in 95% yield. Installation of the poly-
acid functionality was accomplished by treatment of 62
with 7 equiv of DTPA dianhydride in DMSO for 3 h at
25 ꢀC. Purification by MPLC followed by lyopholiza-
tion afforded 63 as a yellow solid in 80% yield.
Compd
A-549lung
MCF-7 breast
HT-29colon
15
17
19
32
33
39
40
45
46
47
48
51
52
53
54
5.5Â10À2
1.3Â10À4
9.3Â10À1
6.5Â10À2
1.9Â10À1
>100,000
>100,000
1.7Â10À3
1.5Â10À3
1.6Â10À3
1.2Â10À3
1.4Â10À3
2.5Â10À3
1.7Â10À3
1.6Â10À3
6.5Â10À2
120
6.6
8.2Â10À1
30
3.2Â10À2
1.8Â10À3
24
1.6
10
4.1Â10À2
4.3Â10À1
>100,000
>100,000
1.3Â10À3
1.1Â10À3
1.2Â10À3
1.0Â10À3
1.2Â10À3
3.3Â10À3
2.9Â10À3
1.1Â10À3
4.0Â10À2
110
>100,000
>100,000
7.5Â10À3
9.0Â10À3
3.1Â10À3
2.4x 10À3
4.9Â10À3
8.0Â10À3
3.3Â10À3
2.6Â10À3
1.6
61
63a
14
930
7.0Â10À2
Æ5.0Â10À2
1.4Â10À2
8.0Â10À2
(Taxol)b
Æ2.0Â10À2
Æ3.0Â10À2
Determined by the Purdue Cancer Center Cell Culture Laboratory.
aAverage of three independent runs.
bAverage of six independent runs.
49 and 50 using Pd(PPh3)4 and Et2NH in CH2Cl2 yielded
Taxol–7-PEG-glutamic acids 51 and 52 in 93 and 96%
yields, respectively. Taxol–7-PEG-Folates 53 and 54
were obtained by the reaction of Taxol–7-PEG-glutamic
acids 51 and 52 with pteroyl azide in DMSO in the pres-
ence of DIEA in 39and 35% yields, respectively.
An additional set of constructs bearing a short, amine-
terminated PEG linker was prepared with a view toward
with increased water solubility and facilitated amide
bond formation with DTPA dianhydride. This series
involved synthesis of a Y-shaped intermediate connected
Hydrolysis and In Vitro Biological Activity
The Taxol–folate conjugates and their glutamate pre-
cursors were surveyed for hydrolysis at 37 ꢀC both at
pH 7 and pH 5 (Table 1). As can be seen in the table,

J. W. Lee et al. / Bioorg. Med. Chem. 10 (2002) 2397–2414
2403
Scheme 9. Reagents and conditions: (a) allyl bromoacetate (2.2 equiv), K2CO3, CH3CN, 24 h, 25 ꢀC; (b) Pd(PPh3)4, PhSiH3, CH2Cl2, 1 h, 25 ꢀC; (c)
DCC, CH2Cl2, 5 h, 25 ꢀC; (d) 36, CH2Cl2, 12 h, 25 ꢀC; (e) 15, DCC, DMAP (cat.), CH2Cl2, 12 h, 25 ꢀC; (f) Pd(PPh3)4, Et2NH, CH2Cl2, 1 h, 25 ꢀC;
(g) PteN3, MTBD, DMSO, 2 h, 25 ꢀC; (h) SnCl2, PhSH, Et3N, DMF, 1 h, 25 ꢀC; (i) excess DTPA dianhydride 62, 3 h, 25 ꢀC.
attachment of glutamate (39) or folate (40) at C-7 via a
urethane moiety gives derivatives that, as expected, do
not suffer hydrolysis even after extended reaction times.
Hydrolytically stable, C-7 functionalized Taxol deriva-
tives have been shown to retain high bioactivity,21 as
exemplified by simple tri and tetracarboxylic acids 17
and 19, yet both glutamate (39) and folate (40) are
essentially inactive (Table 2).
alkoxy and a-methylamino reagents 41, 42 which were
prepared in the course of our investigation. Simply
attachment of these materials to C-20 alloc-protected
Taxol 15 followed by standard processing (Scheme 8)
provided the C-7 tethered glutamates and folates 51–54
in sufficient quantities for extensive testing. By this time,
we had developed our intuition to the point that we
expected folates 53 and 54 to only have importance as
structural controls for Y-shaped folate-tetracarboxylic
acid 63 (Scheme 9).
We next turned to C-20 for attachment of inductively-
activated a-alkoxy and a-methylamino esters 45–48
(Scheme 7). While these materials underwent facile
hydrolysis (Table 1) and exhibited high in vitro activity
(Table 2), the very ease of hydrolysis severely detracted
from our ability to isolate acceptable yields of folates 47
and 48 from the DMSO reaction mixture.
Therefore, it came as a surprise that Y-shaped folate-
tetracarboxylic acid 63 (three separate runs) was about
105 less active than I-shaped folates 53 and 54, which do
not bear the pendant tetracarboxylic acid (Table 2). It
should be noted that Y-shaped folate PEG-3 azide 61
(the precursor of 63) is far more active than 63, but still
ꢁ50Â less active than 53 and 54. Faced with data that
did not reveal any clear trends, we elected to deepen our
Although C-20 study did not generate any useful drug
species, we were pleased with the PEG-3 glutamate-a-

2404
J. W. Lee et al. / Bioorg. Med. Chem. 10 (2002) 2397–2414
testing environment in an attempt to define whether the
folate was selectively targeting any of the agents to the
tumor site.
required a ꢁ4Â greater concentration than free folate
3
to achieve a similar degree of competition with H folic
acid for receptor binding, indicating that 54 retained
most of the native affinity of folic acid for the folate
receptor (Fig. 1B).
Selection of an Agent for an In Vivo Trial
Although both 63 and 54 demonstrated their strong
3
Based upon our initial rationale in conjunction with the
hydrolysis and cell culture studies (Tables 1 and 2), we
elected to compare the abilities of four C-7 linked
folates (53, 54, 61, and 63) with that of free folic acid to
ability to compete with H folic acid for receptor bind-
ing (Fig. 1A), previous in vitro bioanalysis yielded a
high ED50 value (1–10Â10À4) with 63 in all cell lines
tested (Table 2). Because of this apparently poor cyto-
toxicity, a large quantity of 63 would be required for
animal trials. We therefore chose to first evaluate the
‘control’ analogue 54 prior to proceeding further with
63. To look for any receptor-mediated specific cyto-
toxicity, in vitro comparison of 54 with Taxol 14 was
undertaken in the folate-receptor positive human KB
tumor cell line. At equivalent molar concentrations,
Taxol 14 was 50-fold more toxic than 54 (Fig. 2). A
concurrent competition experiment with 54 plus 500-
fold molar excess of free folic acid did not decrease the
toxicity of 54, suggesting that the folate receptor was
not responsible for the cellular entry of 54.
3
displace H folic acid from cell surface folate receptors
in a competitive binding study. In comparison to non-
radioactive folate in receptor-positive murine M109
tumor cells, the target folate tetraacid 63 and its trun-
cated N-methyl analogue 54 (lacking the PEG-3 spacer
and the tetraacid) were far superior to 61 (the azide
precursor to 63) in their relative affinities for the folate
receptor (Fig. 1A). It was also interesting to note that
replacement of the N-methyl unit in 54 with an oxygen
atom (53) decreased the ability to displace folic acid by
at least an order of magnitude. Similarly, in another
folate receptor-positive human KB tumor cell line, 54
Comparison of Taxol 14 and 54 at a constant drug
concentration (2Â10À7 M) was next undertaken in three
tumor cell lines (KB, M109, A549). The first two lines
express high levels of the folate receptor.22 The A549cell
line, however, expresses negligible levels of functional
folate receptors.23 With the exception of Taxol 14 being
uniformly more potent, little difference was seen with
respect to folate receptor involvement (Fig. 3). There-
fore, there is no evidence of selective cytotoxicity of 54
towards folate receptor positive tumor cell lines in vitro.
The first note of encouragement came from a compara-
tive general toxicity study in healthy tumor-free mice.
At an equivalent Taxol 14 dosage of 25 mg/kg/day
(ꢁ46 mg/kg/day for 54), Taxol-treated mice experi-
enced a maximal body weight loss of 20%, while no
weight loss was detected in mice treated with prodrug 54
Figure 1. Comparison of the abilities of Taxol–folate conjugates and
folate to inhibit 3H folic acid binding to cell-surface folate receptors.
Murine M109(A) or human KB (B) cells were incubated at 4 ꢀC for 1
h in FEMEM containing 10 Nm 3H folic acid and various concentra-
tions of 53 (!), 54 (&), 61 (&), 63 (!), or folic acid (*). After
thorough washing, cell-associated radioactive folates were stripped by
acid saline and measured with a liquid scintillation counter.
Figure 2. Cytotoxity of 54 and Taxol 14 in folate receptor positive KB
tumor cell line. Cells were incubated with Taxol (*), 54 (&), or 54
plus ꢁ500Â molar excess of free folic acid (~) for 18 h, washed, and
incubated further in fresh medium. The percentage of cell survival
compared to untreated control 24 h later was then determined. Data
are presented as the meanÆSD of three independent measurements.

J. W. Lee et al. / Bioorg. Med. Chem. 10 (2002) 2397–2414
2405
(Fig. 4). Therefore, 54 is far less toxic than the parent
drug Taxol.
Comparative Antitumor Activity In Vivo
While 54 was far from an ideal candidate, we decided to
proceed further in vivo to compare the antitumor activ-
ity of 54 with Taxol 14 in tumor-bearing mice, since in
vitro tests for comparing drug cytotoxicity may not
always predict in vivo therapeutic efficacy due to differ-
ences in drug stability, biodistribution and metabolism,
as well as tumor microenvironment. Mice implanted ip
with folate receptor positive M109tumors were treated
twice a day for 8 days with equimolar amounts of Taxol
14 and 54 starting on day 4 post tumor implantation.
While the control mice and the mice treated with the
injection vehicle survived approximately 26 days, Taxol
14 and prodrug 54 caused >177 and 73% increase in
life span and yielded cure rates of 73 and 18%, respec-
tively, at an equivalent Taxol 14 dose of 17.2 mg/kg/day
(ꢁ32 mg/kg/day for 54) (Fig. 5).
Figure 3. Comparison of the cytotoxicity of 54 and Taxol 14 towards
various tumor cell lines. Receptor-positive KB or M109, and receptor-
negative A549cells were incubated with Taxol and 54 at 2Â10À7 M for
18 h, washed, and incubated further in fresh medium. The percentage
of cell survival compared to untreated control cells at confluence was
then determined. Data are presented as the meanÆSD of three inde-
pendent measurements.
Conclusion
Our effort to enhance the pharmacologic efficacy of
Taxol through covalently linking the drug to a tumor-
targeting folate ligand was not successful in the form of
54. Although the Taxol–folate conjugate 54 retained
most of the receptor binding affinity of the folate ligand
and was far less toxic than Taxol in normal mice, it
failed to demonstrate selective killing of folate receptor-
expressing tumor cells in vitro or enhanced in vivo
antitumor activity over Taxol when administered in an
equimolar quantity formulated in the same injection
vehicle.
Figure 4. Comparison of the toxicity of 54 and Taxol 14 in nontumor
bearing mice. Female Balb/c mice were given four ip injections of 25
mg/kg/day Taxol 14 (*, !, &) or 46 mg/kg/day 54 (*, !, &) at 48-
h intervals. The mice were weighed every other day to record their
body weight changes.
There are a number of possible reasons to explain the
ineffectiveness of 54 in treating folate receptor-positive
tumors. First of all, 54 still has limited water solubility
despite a ꢁ20Â (molar ratio) improvement over Taxol.
The low water solubility prevented the administration
of a higher, more effective dose of 54 in tumor-bearing
mice. Secondly, both 54 and Taxol are large hydro-
phobic molecules that probably enter cells non-
specifically. Thus, the attachment of folate via a short
ester linker may not be enough to prevent nonspecific
binding and uptake in a receptor-independent manner.
Thirdly, the ester linkage was designed to hydrolyze in
vivo. Based on hydrolysis data at 37 ꢀC, 54 has a slower
hydrolysis rate (t1/2=197 h) at an acidic pH (pH 5)
than the hydrolysis rate (t1/2=109h) at neutral pH.
Upon binding of 54 to the cell surface receptor, it will
likely be internalized into endosomes where the pH
value is around 5.24 This acidic pH would certainly slow
down the hydrolysis of 54 to free Taxol inside cells.
However, once Taxol is released, the drug should have
no problem of diffusing out of endosomes and binding
to tubulin to exert its cytotoxic effect.
Figure 5. Antitumor effect of 54 and Taxol 14 in tumor-bearing mice.
Balb/c mice were inoculated ip with 5Â105 M109cells on day 0. Drugs
were administered from day 4 according to the schedule of q2dÂ8 for
untreated (*), injection vehicule (*), Taxol 14 at 17.2 mg/kg/day (!),
and 54 at 32.4 mg/kg/day (!). Each group consists of at least 10 mice.
The future of targeting Taxol to folate receptor-positive
tumor types via the folate ligand may require the design

2406
J. W. Lee et al. / Bioorg. Med. Chem. 10 (2002) 2397–2414
of new water-soluble linkers that are stable in the blood,
but hydrolyze quickly upon entering acidic endosomal
compartments to allow release of the cytotoxic parent
drug. The water solubility of any new folate–Taxol
derivative should also be high enough to avoid any need
for drug-solubilizing agents, so that an optimal drug
dosage may be achieved in vivo. Most importantly, in
order to achieve selective toxicity against receptor-positive
tumors, the mechanism of drug entry should be gov-
erned mainly by binding of the folate moiety to cell
surface receptors and not by simple membrane
diffusion.
Preparation of tetraalkylated ethylenediamine derivative
(3). A solution of N-Boc protected amine 2 (0.8 g, 2.25
mmol) in 20 mL of EtOAc was saturated with HCl (g)
and then stirred for 10 h at 25 ꢀC. The resulting solution
was concentrated to give the deprotected HCl salt. This
crude material was dissolved in 10 mL of DMF and
then tert-butyl bromoacetate (1.7 mL, 11.3 mmol) and
Hunig base (2 mL, 11.3 mmol) was added to the above
crude solution at 0 ꢀC. The reaction mixture was stirred
for 12 h at 25 ꢀC. EtOAc (100 mL) was added and then
the mixture was washed with water and brine. The
organic layer was dried and concentrated. The residue
was chromatographed with EtOAc/hexane (1:3) to
1
afford the desired product 3 (0.63 g, 58%). H NMR
(CDCl3) d 5.97–5.84 (m, 2H), 5.34–5.20 (m, 4H), 4.59
(d, J=5.7 Hz, 4H), 3.66 (s, 4H), 3.45 (s, 4H), 2.88 (t,
J=4.2 Hz, 2H), 1.43 (s, 18H); LRMS (CI) m/z 485
(M+H); HRMS (FAB) calcd for C24H41N2O8; (M+H)
485.2863, found 485.2868.
Experimental
General methods
Unless otherwise stated, reactions were carried out
under argon in flame-dried glassware. Flash chromato-
graphy on silica gel was carried out as described by Still
(230–400 mesh silica gel was used), and reversed-phase
LC was used for preparative purposes (LiChroprep
C-18, 310Â25 mm). 1H and 13C NMR spectra were
obtained using GE QE-300 NMR and Varian Gemini
200 NMR spectrometers at 300 or 200 MHz and 75 or
50 MHz respectively. Mass spectral data were obtained
on a Finnigan 4000 mass spectrometer (low resolution)
and a CEC 21 110 B high-resolution mass spectrometer,
with the molecular ion designated as M. All the chemi-
cals were supplied by Aldrich Chemical Co., Inc, Mil-
waukee, WI, USA, unless otherwise indicated.
Preparation of EDTA monoacid (4). A solution of com-
pound 3 (0.5 g, 1 mmol) in 6 mL of CH2Cl2 and 4 mL of
trifluoroacetic acid was stirred for 10 h at 25 ꢀC. The
resulting solution was concentrated and dried under
high pressure to give deprotected diacid as a quantita-
tive yield. 1H NMR (CDCl3) d 5.94–5.81 (m, 2H), 5.33–
5.21 (m, 4H), 4.58 (d, J=4.8 Hz, 4H), 4.21 (s, 4H), 3.64
(s, 4H), 3.47 (s, 2H), 3.15 (s, 2H); LRMS (FAB) m/z
372.8 (M+H); HRMS (FAB) calcd for C16H25N2O8;
(M+H) 373.1611, found 373.1610. A solution of DCC
(0.2 g, 1 mmol) in 2 mL of CH2Cl2 was added to a
solution of the deprotected diacid in 8 mL of CH2Cl2 at
0 ꢀC and stirred for 5 h at 25 ꢀC. The precipitate was
filtered and the filtrate was concentrated to give the
crude anhydride. 1H NMR (CDCl3) d 5.95–5.86 (m,
2H), 5.36–5.25 (m, 4H), 4.60 (d, J=6.0 Hz, 4H), 3.66 (s,
4H), 3.61 (s, 4H), 2.97 (t, J=6.0 Hz, 2H), 2.70 (t, J=6.0
Hz, 2H). This anhydride was dissolved with 8 mL of
allyl alcohol and one drop of triethyl amine and 5 mg of
DMAP was added to a solution. The resulting solution
was stirred for 15 h at 25 ꢀC and concentrated. The
residue was chromatographed using EtOAc/MeOH
Preparation of N-Boc-ethylenediamine allyl ester (2).
(Boc)2O (2.2 g, 10 mmol) in 20 mL of MeOH was
added dropwise to a solution of ethylenediamine 1 (6 g,
100 mmol) in 80 mL of MeOH at 0 ꢀC and the resulting
solution was stirred for 2 h at 0 ꢀC and then 2 h at 25 ꢀC.
The reaction mixture was concentrated. The residue was
dissolved with EtOAc/CH2Cl2/ether and insoluble
material was filtered and the filtrate was concentrated to
afford the pure crude N-Boc-ethylenediamine as a
quantitative yield which is used to the next reaction
1
(5:1) to afford the desired product 4 (0.24 g, 57%). H
1
without purification. H NMR (CDCl3) d 3.13 (td, 6.0,
NMR (CDCl3) d 5.97–5.84 (m, 3H), 5.35–5.23 (m, 6H),
4.60 (d, J=5.4 Hz, 6H), 3.61 (s, 4H), 3.55 (s, 2H), 3.50
(s, 2H), 2.88 (s, 4H); LRMS (CI) m/z 413 (M+H);
HRMS (FAB) calcd for C19H29N2O8; (M+H)
413.1924, found 413.1924.
J=5.7 Hz, 2H), 2.75 (t, J=6.0 Hz, 2H), 1.40 (s, 9H),
1.15 (br s, 2H); LRMS (CI) m/z 161 (M+H), 105;
HRMS (FAB) calcd for C7H17N2O2; (M+H)
161.1290, found 161.1288. Hunig base (4.4 mL, 25
mmol) and a solution of allyl bromoacetate (3.8 g, 22
mmol) in 5 mL of CH3CN were added to a solution of
N-Boc-ethylenediamine (1.6 g, 10 mmol) in 35 mL of
CH3CN at 0 ꢀC. The resulting solution was stirred for
20 h at 25 ꢀC and concentrated. EtOAc was added to
the residue and the organic phase was washed with
water and brine, dried, and concentrated to give a
crude product which was chromatographed from
EtOAc/hexane (1:3) to afford the desired product 2 (3.1 g,
Synthesis of di-allyl 3-((Allylcarbonyl)methyl)-6-(2-(tri-
fluoroacetyl)amino)ethyl-3,6-diazaoctanedioate (7). Ethyl
trifluoroacetate (1.46 g, 10.3 mmol) in 12 mL of CH2Cl2
was added dropwise to a solution of diethylenetriamine
6 (1.06 g, 10.3 mmol) in 8 mL of CH2Cl2 at 0 ꢀC. After
stirring 2 h at 0 ꢀC, the reaction mixture was stirred for
5 h at 25 ꢀC. The resulting solution was concentrated
and the residue was dissolved in 50mL of acetonitrile.
Hunig base (6.3 mL, 36 mmol) and allyl bromoacetate
(6.2 g, 36 mmol) added to the above solution at 0 ꢀC
and stirred for 20 h at 25 ꢀC. The resulting solution was
concentrated and the residue was dissolved in ethyl
acetate (100 mL). The organic solution was washed with
water and brine, dried, and concentrated. The crude was
1
86%). H NMR (CDCl3) d 5.97–5.84 (m, 2H), 5.49 (br
s, 1H), 5.34–5.21 (m, 4H), 4.60 (d, J=5.7 Hz, 4H), 3.57
(s, 4H), 3.15 (td, J=5.7, 5.4 Hz, 2H), 2.85 (t, J=5.7 Hz,
2H), 1.43 (s, 9H); LRMS (CI) m/z 357 (M+H); HRMS
(FAB) calcd for C17H29N2O6; (M+H) 357.2026, found
357.2024.

J. W. Lee et al. / Bioorg. Med. Chem. 10 (2002) 2397–2414
2407
chromatographed from EtOAc/hexane (1:2) to afford
1
5.84 (m, 3H), 5.34–5.19(m, 6H), 4.60–4.55 (m, 6H), 3.61
(s, 4H), 3.53 (s, 2H), 3.43 (s, 4H), 2.86–2.80 (m, 8H),
1.44 (s, 18H); LRMS (CI) m/z 626 (M+H); HRMS
(FAB) calcd for C31H52N3O10; (M+H) 626.3653, found
626.3656.
the desired product 7 (2.6 g, 52%). H NMR (CDCl3) d
8.37 (br s, 1H), 5.96–5.83 (m, 3H), 5.34–5.22 (m, 6H),
4.59(d, J=5.7 Hz, 2H), 4.58 (d, J=6.0 Hz, 4H), 3.61 (s,
4H), 3.43 (s, 2H), 3.35 (td, J=6.0, 4.2 Hz, 2H), 2.88 (t,
J=6.6 Hz, 2H), 2.86 (t, J=5.7 Hz, 2H), 2.78 (t, J=5.4
Hz, 2H); LRMS (CI) m/z 494 (M+H); HRMS (FAB)
calcd for C21H31F3N3O7; (M+H) 494.2114, found
494.2111.
Synthesis of DTPA monoacid (12). A solution of com-
pound 11 (0.5 g, 0.8 mmol) in 6 mL of CH2Cl2 and 4
mL of trifluoroacetic acid was stirred for 10 h at 25 ꢀC.
The solution was concentrated and dried under high
pressure to give deprotected diacid as a quantitative
yield. 1H NMR (CDCl3) d 5.95–5.81 (m, 3H), 5.40–5.21
(m, 6H), 4.61–4.57 (m, 6H), 4.07 (s, 4H), 3.72 (s, 6H),
3.50 (m, 2H), 3.15 (m, 2H), 2.96 (m, 4H); LRMS (FAB)
m/z 514 (M+H); HRMS (FAB) calcd for C23H36N3O10;
(M+H) 514.2401, found 514.2399. A solution of DCC
(0.17 g, 0.8 mmol) in 2 mL of CH2Cl2 was added to a
solution of the diacid in 6 mL of CH2Cl2 at 0 ꢀC and
stirred for 5 h at 25 ꢀC. The precipitate was filtered and
the filtrate concentrated to give the crude anhydride. 1H
NMR (CDCl3) d 5.97–5.84 (m, 3H), 5.35–5.22 (m, 6H),
4.60–4.57 (m, 6H), 3.65 (s, 4H), 3.61 (s, 4H), 3.54 (s,
2H), 2.92–2.86 (m, 6H), 2.70 (t, J=5.1 Hz, 2H). This
crude anhydride was dissolved in 8 mL of allyl alcohol
and one drop of triethylamine and 5 mg of DMAP was
added to a solution. The solution was stirred for 15 h at
25 ꢀC and concentrated. The residue was chromato-
graphed from EtOAc/MeOH (5:1) to afford the desired
Synthesis of DTTA monoacid (8). A solution of com-
pound 7 (0.25 g, 0.5 mmol) in 2 mL of DMF was added
dropwise to a suspension of 60% NaH (24 mg, 0.6
mmol) in 1 mL of DMF and stirred for 15 min followed
by adding trimethylsilyl bromoacetate (0.16 g, 0.75
mmol) to the above solution at 0 ꢀC. The resulting
solution was stirred for 12 h at 25 ꢀC and EtOAc (30
mL) was added. The solution was washed with satu-
rated sodium bicarbonate, brine, dried, concentrated,
and chromatographed from EtOAc/MeOH (10:1) to
afford the desired product 8 (0.2 g, 70%: 98% BRSM).
1H NMR (CDCl3) d 5.96–5.83 (m, 3H), 5.36–5.23 (m,
6H), 4.60 (d, J=5.7 Hz, 2H), 4.59(d, J=6.0 Hz, 4H),
4.25 (s, 2H), 3.71 (t, J=6.3 Hz, 2H), 3.59(s, 2H), 3.07–
2.97 (m, 4H), 2.95–2.84 (m, 2H); LRMS (CI) m/z 552
(M+H); HRMS (FAB) calcd for C23H33F3N3O9;
(M+H) 552.2169, found 552.2167.
1
Preparation of N-Boc-diethylenetriamine derivative (10).
(Boc)2O (1.6 g, 7.5 mmol) in 15 mL of MeOH was
added dropwise to a solution of diethylenetriamine 6
(0.7 g, 6.8 mmol) in 25 mL of MeOH at 0 ꢀC, stirred for
2 h at 0 ꢀC and then 10 h at 25 ꢀC. The reaction mixture
was concentrated, dissolved with CH3CN, insoluble
material filtered, and the filtrate was concentrated and
dissolved with 30 mL of CH3CN. Hunig base (4.8 mL,
27 mmol) and allyl bromoacetate (4.18 g, 25 mmol)
were added at 0 ꢀC. The resulting solution was stirred
for 12 h at 25 ꢀC and concentrated. EtOAc was added
and the organic phase was washed with water and brine,
dried, and concentrated to give a crude product which
was chromatographed using EtOAc/hexane (1:3) to
product 12 (0.3 g, 68%). H NMR (CDCl3) d 5.96–5.83
(m, 4H), 5.35–5.21 (m, 8H), 4.61–4.57 (m, 8H), 3.63 (s,
2H), 3.57 (s, 4H), 3.56 (s, 2H), 3.48 (s, 2H), 2.94–2.86
(m, 8H); LRMS (CI) m/z 554 (M+H); HRMS (FAB)
calcd for C26H40N3O10; (M+H) 554.2714, found
554.2711.
Preparation of 20-Alloc-Taxol (15). Hunig base (0.26
mL, 1.46 mmol) and allyl chloroformate (0.17 mL, 1.6
mmol) was added to a solution of Taxol (500 mg, 0.59
mmol) in 10 mL of methylene chloride at 0 ꢀC and stir-
red for 10 h. 20 mL of CH2Cl2 added and the mixture
washed with 0.1N HCl (20 mL), dried, and concentrated
to give solid which was recrystallized from CH2Cl2–
1
1
afford the desired product 10 (1.8 g, 52%). H NMR
ether to afford the pure product 15 (540 mg, 98%). H
(CDCl3) d 5.98–5.85 (m, 3H), 5.62 (br s, 1H), 5.35–5.22
(m, 6H), 4.61–4.58 (m, 6H), 3.61 (s, 4H), 3.44 (m, 2H),
2.86–2.73 (m, 6H), 1.44 (s, 9H); LRMS (CI) m/z 498
(M+H); HRMS (FAB) calcd for C24H40N3O8; (M+H)
498.2815, found 498.2818.
NMR (CDCl3) d 8.13 (dd, J=8.5, 1.2 Hz, 2H, Bz), 7.74
(dd, J=8.2, 1.2 Hz, 2H, Bz), 7.62 (t, J=7.5 Hz, 1H, Bz),
7.52–7.32 (m, 10H, Ar), 6.93 (d, J=9.3 Hz, 1H, NH),
6.29(s, 1H, 10-H), 6.28 (t, J=9.0 Hz, 1H, 13-H), 5.99
(dd, J=9.3, 2.4 Hz, 1H, 30-H), 5.92ꢁ5.86 (m, 1H,
¼CH), 5.69(d, J=6.9Hz, 1H, 2-H), 5.43 (d, J=2.7
Hz, 1H, 20-H), 5.39–5.28 (m, 2H, ¼CH2), 4.98 (b d,
J=8.0 Hz, 1H, 5-H), 4.64 (dd, J=3.9, 1.8 Hz, 2H,
allylic CH2), 4.44 (m, 1H, 7-H), 4.33 (A of AB, d,
J=8.4 Hz, 1H, 20-H), 4.21 (B of AB, d, J=8.4 Hz,
1H, 20-H), 3.82 (d, J=7.0 Hz, 1H, 3-H), 2.56 (m, 1H,
6-H), 2.47 (s, 3H, OAc), 2.40 (dd, 1H, 14-CH2), 2.23 (s,
Preparation of Pentaalkylated-diethylenetriamine deriva-
tive (11). A solution of N-Boc protected amine 10 (0.75
g, 1.5 mmol) in 15 mL of EtOAc was saturated with
HCl (g) and then stirred for 10 h at 25 ꢀC. The resulting
solution was concentrated to give deprotected product.
This crude material was dissolved with 7.5 mL of DMF
and then tert-butyl bromoacetate (1.1 mL, 7.5 mmol)
and Hunig base (1.6 mL, 9mmol) were added to the
solution at 0 ꢀC. The mixture was stirred for 12 h at
25 ꢀC. EtOAc (100 mL) was added and then the mixture
was washed with water and brine. The organic layer was
dried and concentrated. The residue was chromato-
graphed from EtOAc/hexane (1:3) to afford the desired
3H, OAc), 2.19(JJ, 1H, 14-CH ), 1.93 (s, 3H, 18-CH3),
2
1.89(m, 1H, 6-H), 1.69(s, 3H, 19-CH 3), 1.25 (s, 3H, 16-
CH3), 1.14 (s, 3H, 17-CH3); LRMS (FAB) m/z 937.8
(M).
General procedure for the preparation of fully protected
Taxol derivatives (16 and 17). A solution of DCC (0.2
mmol) and 5 mg of DMAP in 0.5 mL of CH2Cl2 was
1
product 11 (0.56 g, 60%). H NMR (CDCl3) d 5.97–

2408
J. W. Lee et al. / Bioorg. Med. Chem. 10 (2002) 2397–2414
added to a solution of monoacid 4 (or 12) (0.15 mmol)
and 20-alloc-Taxol 15 (0.1 mmol) in 1.5 mL of CH2Cl2
at 0 ꢀC. The resulting solution was allowed to warm to
25 ꢀC and stirred for 10 h. The reaction mixture was fil-
tered, concentrated, and chromatographed from
EtOAC/hexane (1:1 and/or 2:1) to afford the desired
product 16 or 18.
Preparation of azido-PEG Taxol (25 and 26). A solution
of DCC (78 mg, 0.38 mmol) and 5 mg of DMAP in 1
mL of CH2Cl2 was added to a solution of PEG-acid 23
(or 24) (0.4 mmol) and 20-alloc-Taxol 15 (250 mg, 0.27
mmol) in 3 mL of CH2Cl2 at 0 ꢀC. The resulting solu-
tion was allowed to warm to 25 ꢀC and stirred for 10 h.
The reaction mixture was filtered, concentrated, and
chromatographed from EtOAc to afford the desired
product (25 or 26).
1
16. 95% yield; H NMR (CDCl3) d 8.13 (d, J=7.2 Hz,
2H), 7.75 (d, J=7.2 Hz, 2H), 7.61 (t, J=7.2 Hz, 1H),
7.53–7.45 (m, 4H), 7.43–7.35 (m, 6H), 6.94 (d, J=9.3
Hz, 1H, NH), 6.25 (t, J=9.0 Hz, 1H), 6.23 (s, 1H), 5.99
(dd, J=9.3, 2.4 Hz, 1H), 5.97–5.84 (m, 4H), 5.68 (d,
J=6.9Hz, 1H), 5.60 (dd, J=7.2, 3.3 Hz, 1H), 5.43 (d,
J=2.4 Hz, 1H), 5.39–5.20 (m, 8H), 4.95 (d, J=9.3 Hz,
1H), 4.64 (dd, 4.5, 1.2 Hz, 2H), 4.58 (d, J=5.7 Hz, 6H),
4.33 (A of AB, d, J=8.4 Hz, 1H), 4.18 (B of AB, d,
J=8.4 Hz, 1H), 3.94 (d, J=6.6 Hz, 1H), 3.67–3.55 (m,
8H), 2.89(s, 4H), 2.61 (m, 1H), 2.45 (s, 3H), 2.39(m,
1H), 2.22 (m, 1H), 2.12 (s, 3H), 1.98 (s, 3H), 1.88 (m,
1H), 1.79(s, 3H), 1.20 (s, 3H), 1.14 (s, 3H); LRMS
(PDMS) m/z 1333 (M+H).
1
25. 94% yield. H NMR (CDCl3) d 8.13 (d, J=8.7 Hz,
2H), 7.74 (d, J=8.4 Hz, 2H), 7.62 (t, J=7.5 Hz, 1H),
7.54–7.34 (m, 10H), 7.22 (br s, 1H), 6.95 (d, J=9.6 Hz,
1H, NH), 6.26 (t, J=8.1 Hz, 1H), 6.19(s, 1H), 5.98 (dd,
J=9.3, 2.7 Hz, 1H), 5.94–5.83 (m, 1H), 5.68 (d, J=6.9
Hz, 1H), 5.66 (dd, J=5.5, 3.3 Hz, 1H), 5.43 (d, J=2.7
Hz, 1H), 5.39–5.27 (m, 2H), 4.97 (d, J=9.3 Hz, 1H),
4.64 (m, 2H), 4.35–4.03 (m, 6H), 3.95 (d, J=6.9Hz,
1H), 3.69–3.47 (m, 14H), 3.38 (t, J=5.4 Hz, 2H), 2.62
(m, 1H), 2.48 (s, 3H), 2.40 (m, 1H), 2.21 (m, 1H), 2.16
(s, 3H), 1.97 (s, 3H), 1.90 (m, 1H), 1.79 (s, 3H), 1.22 (s,
3H), 1.14 (s, 3H); LRMS (PDMS) m/z 1254 (M).
1
1
18. 94% yield; H NMR (CDCl3) d 8.13 (d, J=6.9Hz,
26. 95% yield. H NMR (CDCl3) d 8.12 (d, J=8.7 Hz,
2H), 7.75 (d, J=6.9Hz, 2H), 7.62 (t, J=6.6 Hz, 1H),
7.54–7.46 (m, 4H), 7.43–7.34 (m, 6H), 6.92 (d, J=9.3
Hz, 1H, NH), 6.26 (t, J=9.9 Hz, 1H), 6.23 (s, 1H),
5.99 (dd, J=9.3, 2.7 Hz, 1H), 5.97–5.84 (m, 5H),
5.68 (d, J=6.9Hz, 1H), 5.60 (dd, J=7.2, 3.3 Hz,
1H), 5.43 (d, J=2.4 Hz, 1H), 5.40–5.21 (m, 10H), 4.96
(d, J=8.1 Hz, 1H), 4.64 (dd, 4.5, 1.2 Hz, 2H), 4.60
(d, J=5.7 Hz, 8H), 4.33 (A of AB, d, J=8.4 Hz,
1H), 4.18 (B of AB, d, J=8.4 Hz, 1H), 3.94 (d,
J=6.9Hz, 1H), 3.64–3.52 (m, 10H), 2.84–2.80 (m,
8H), 2.63 (m, 1H), 2.46 (s, 3H), 2.40 (m, 1H), 2.23 (m,
1H), 2.12 (s, 3H), 1.98 (s, 3H), 1.84 (m, 1H), 1.79 (s,
3H), 1.21 (s, 3H), 1.14 (s, 3H); LRMS (PDMS) m/z 1476
(M+H).
2H), 7.73 (d, J=8.7 Hz, 2H), 7.61 (t, J=7.5 Hz, 1H),
7.58 (br s, 1H), 7.53–7.45 (m, 4H), 7.42–7.34 (m, 6H),
6.94 (d, J=9.3 Hz, 1H, NH), 6.24 (t, J=9.3 Hz, 1H),
6.21 (s, 1H), 5.97 (dd, J=9.3, 2.7 Hz, 1H), 5.92–5.83 (m,
1H), 5.68 (d, J=6.9Hz, 1H), 5.60 (dd, 6.9, 3.3 Hz, 1H),
5.42 (d, J=2.7 Hz, 1H), 5.38–5.26 (m, 2H), 4.95 (d,
J=8.4 Hz, 1H), 4.63 (m, 2H), 4.32 (A of AB, d, J=8.1
Hz, 1H), 4.17 (B of AB, d, J=8.7 Hz, 1H), 3.94 (d,
J=6.9Hz, 1H), 3.67–3.55 (m, 12H), 3.48 (t, 6.6 Hz,
2H), 3.45–3.24 (m, 4H), 3.21 (d, J=2.7 Hz, 2H), 2.62
(m, 1H), 2.45 (s, 3H), 2.41 (m, 1H), 2.38 (s, 3H), 2.22
(m, 1H), 2.12 (s, 3H), 1.97 (s, 3H), 1.89 (m, 1H), 1.79 (s,
3H), 1.20 (s, 3H), 1.13 (s, 3H); LRMS (PDMS) m/z 1268
(M+H).
General procedure for the preparation of Taxol EDTA/
DTPA (17 and 19). PhSiH3 (0.5 mmol) and Pd(PPh3)4
(0.004 mmol) was added to a solution of protected
Taxol 16 (or 18) (0.08 mmol) in 3 mL of methylene
chloride and stirred for 1 h at 25 ꢀC. 1 mL of MeOH
was added and the reaction solution stirred for 10 min.
The resulting solution was concentrated and the residue
was recrystallized from CH2Cl2/EtOAc/ether system to
afford the desired product 17 (or 19) in good yield. 17:
75% yield; LRMS (PDMS) m/z 1129(M+H). And 19.
70% yield; LRMS (PDMS) m/z 1230 (M+H).
Preparation of amino-PEG–Taxol (27 and 28). PhSH
(1.2 mmol) and Et3N (0.9mmol) were added to a solu-
tion of SnCl2 (0.3 mmol) in 4 mL of CH3CN. After 5
min, azido-compound 25 (or 26) (0.2 mmol) was added
and stirred for 1 h at 25 ꢀC. 30 mL of CH2Cl2 and 30
mL of saturated NaHCO3 solution was added. The
organic layer was separated and the separated aqueous
layer was extracted with methylene chloride twice. The
combined organic solution was washed with brine,
dried, and concentrated. The residue was dissolved in 5
mL of MeOH, stirred for 30 min at 25 ꢀC and con-
centrated. The purification by column chromatography
afforded the desired product amine.
Preparation of PEG-carboxylic acid (23 and 24). A
solution of amino-PEG-azide 20 (0.64 g, 2.90 mmol)
and anhydride 21 (or 22) (2.24 mmol) in 25 mL of
CH2Cl2 was stirred for 36 h at 25 ꢀC. Water (1 mL) was
added to destroy excess anhydride and left to stir over-
night. The reaction solution was dried and concentrated
to afford the desired product 23 or 24 (ꢁ98% yield). 23:
1H NMR (CDCl3) d 7.63 (brs, 1H), 4.18 (s, 2H), 4.14 (s,
2H), 3.74–3.62 (m, 10H), 3.56 (m, 4H), 3.40 (t, J=5.4
1
27. 90% yield; H NMR (CDCl3) d 8.13 (d, J=8.7 Hz,
2H), 7.74 (d, J=8.4 Hz, 2H), 7.62 (t, J=7.5 Hz, 1H),
7.54–7.34 (m, 10H), 7.22 (br s, 1H), 6.95 (d, J=9.6 Hz,
1H), 6.26 (t, J=8.1 Hz, 1H), 6.19(s, 1H), 5.98 (dd,
J=9.3, 2.7 Hz, 1H), 5.94–5.83 (m, 1H), 5.68 (d, J=6.9
Hz, 1H), 5.66 (dd, J=5.5, 3.3 Hz, 1H), 5.43 (d, 2.7 Hz,
1H), 5.39–5.27 (m, 2H), 4.97 (d, J=9.3 Hz, 1H), 4.64
(m, 2H, allylic CH2), 4.35–4.03 (m, 6H), 3.95 (d, J=6.9
Hz, 1H), 3.69–3.47 (m, 14H), 2.90 (t, J=5.1 Hz, 2H),
2.62 (m, 1H), 2.48 (s, 3H), 2.40 (m, 1H), 2.21 (m, 2H),
1
Hz, 2H). 24: H NMR (CDCl3) d 3.69–3.61 (m, 12H),
3.55 (t, J=4.5 Hz, 2H), 3.42 (s, 4H), 3.40 (t, J=5.4 Hz,
2H), 2.62 (s, 3H).

J. W. Lee et al. / Bioorg. Med. Chem. 10 (2002) 2397–2414
2409
2.16 (s, 3H), 1.97 (s, 3H), 1.90 (m, 1H), 1.79 (s, 3H), 1.22
(s, 3H), 1.14 (s, 3H); LRMS (PDMS) m/z 1228.4
(M+H).
Preparation of amino-PEG-glutamate (36). A solution
of azido-PEG-glutamate 35 (1.4 g, 3 mmol) and tri-
phenylphosphine (1 g, 3.6 mmol) in 20 mL of THF with
a drop of water was stirred for 24 h at 25 ꢀC. The reac-
tion solution was concentrated and chromatographed
with EtOAc and EtOAc/MeOH (10:1) to remove tri-
phenylphosphine derivatives and then MeOH/Et3N
1
28. 97% yield; H NMR (CDCl3) d 8.12 (d, J=8.7 Hz,
2H), 7.73 (d, J=8.4 Hz, 2H), 7.61 (t, J=7.5 Hz, 1H),
7.58 (br s, 1H), 7.53–7.45 (m, 4H), 7.42–7.34 (m, 6H),
6.99 (d, J=9.3 Hz, 1H), 6.24 (t, J=9.3 Hz, 1H), 6.21 (s,
1H), 5.97 (dd, J=9.3, 2.7 Hz, 1H), 5.92–5.83 (m, 1H),
5.68 (d, J=6.9Hz, 1H), 5.60 (dd, J=6.9, 3.3 Hz, 1H),
5.42 (d, 2.7 Hz, 1H), 5.38–5.26 (m, 2H), 4.95 (d, J=8.4
Hz, 1H), 4.63 (m, 2H, allylic CH2), 4.32 (A of AB, d,
J=8.1 Hz, 1H), 4.03 (B of AB, d, J=8.7 Hz, 1H), 3.94
(d, J=6.9Hz, 1H), 3.67–3.55 (m, 12H), 3.48 (t, J=6.6
Hz, 2H), 3.45–3.24 (m, 2H), 3.21 (d, J=2.7 Hz, 2H),
2.92 (t, J=5.3 Hz, 2H), 2.62 (m, 1H), 2.45 (s, 3H), 2.41
(m, 1H), 2.38 (s, 3H), 2.22 (m, 2H), 2.12 (s, 3H), 1.97 (s,
3H), 1.89(m, 1H), 1.79(s, 3H), 1.20 (s, 3H), 1.13 (s,
3H); LRMS (PDMS) m/z 1241.5 (M+H).
1
(98:2) to obtain the desired product in 99% yield. H
NMR (CDCl3) d 7.14 (br s, 1H, NH), 6.08 (d, J=8.4
Hz, 1H, NH), 6.00–5.84 (m, 2H), 5.34–5.19(m, 4H),
4.63 (d, J=5.7 Hz, 2H), 4.56 (d, J=5.4 Hz, 2H), 4.34
(m, 1H), 3.71–3.50 (m, 12H), 3.44 (t, J=4.8 Hz, 2H),
2.87 (t, J=5.1 Hz, 2H), 2.31 (br s, 2H, NH2), 2.28 (m,
2H), 2.22–2.17 (m, 1H), 2.07–1.97 (m, 1H); 13C NMR
(CDCl3, 75 MHz) d 172.2, 172.0, 156.3, 132.8, 131.7,
118.9, 117.8, 70.5, 70.2, 70.1, 70.0, 69.9, 69.8, 66.0, 65.8,
53.9, 41.4, 39.4, 32.3, 28.0; LRMS (CI) m/z 446(M+H),
383; HRMS (CI) calcd for C20H36N3O8; (M+H)
446.2502, found 446.2488.
Preparation of compounds 30 and 31. Taxol-amine 27
(or 28) (0.066 mmol) was added to a solution of DTPA-
dianhydride 29 (0.5 mmol) in 3 mL of DMSO and stir-
red for 3 h at 25 ꢀC. Water (2 mL) was added and the
solution stirred for several hours. The solution was lyo-
philized and the residue was dissolved with CH2Cl2 and
filtered. The filtrate was concentrated and recrystallized
from CH2Cl2/EtOAc/ether (1:2:3) to afford the desired
product 30 (or 31) as a good yield. 30: 80% yield.
LRMS (MALDI) m/z 1604 (M+H). 31: 85% yield.
LRMS (MALDI) m/z 1617 (M+H).
Preparation of Taxol-7-carbamoyl-PEG-glutamate (38).
A solution of 20-protected-Taxol–7-OCOIm 37 (40 mg,
0.04 mmol) and amino-PEG-glutamate 36 (175 mg, 0.4
mmol) in 3 mL of anhydrous isopropyl alcohol was
heated at reflux for 5 h. After cooling to the room tem-
perature, the reaction mixture was concentrated and the
residue was chromatographed from EtOAc to afford the
1
desired product (43 mg, 82%). H NMR (DMSO-d6) d
8.90 (d, J=7.6 Hz, 1H, NH), 7.95 (d, J=7.2 Hz, 2H),
7.87 (d, J=6.9Hz, 2H), 7.70 (t, J=6.4 Hz, 1H), 7.68–
7.37 (m, 10H), 7.20 (m, 1H), 6.30 (s, 1H), 6.18 (d, J=7.5
Hz, 1H), 5.87 (t, J=7.6 Hz, 1H), 5.91–5.82 (m, 2H),
5.40 (d, J=7.8 Hz, 1H), 5.37 (d, J=8.1 Hz, 1H), 5.30–
5.13 (m, 4H), 4.92 (d, J=9Hz, 1H), 4.75 (s, 1H), 4.57 (t,
J=7.8 Hz, 1H), 4.54 (d, J=5.1 Hz, 2H), 4.44 (d,
J=4.8HZ, 2H), 4.01 (m, 3H), 3.69(d, J=6.6 Hz, 1H),
3.47 (m, 4H), 3.40–3.24 (m, 8H), 3.14 (t, J=5.7 Hz, 2H),
3.04 (m, 2H), 2.37 (m, 1H), 2.20 (s, 3H), 2.14 (t, J=7.5
Hz, 2H), 2.05 (s, 3H), 1.96 (m, 2H), 1.79 (s, 3H), 1.76
(m, 3H), 1.61 (s, 3H), 1.00 (s, 3H), 0.98 (s, 3H); LRMS
(FAB) m/z 1325.0 (M).
Preparation of Taxol–PEG multiacid (32 and 33).
Et2NH (0.28 mmol) and Pd(PPh3)4 (2 mg) was added
to a solution of protected glutamate 30 (or 31) (0.028
mmol) in 4 mL of methylene chloride and stirred for 30
min at 25 ꢀC. The reaction mixture was concentrated
and the residue was recrystallized from CH2Cl2/EtOAc/
ether (1:1:3) system to afford the desired product 32 or
33 as a good yield. 32: 95% yield. LRMS (MALDI) m/z
1519(M+H). 33: 90% yield. LRMS (MALDI) m/z
1533 (M+H).
Preparation of Taxol–7-carbamoyl-PEG-glutamic acid
(39). A solution of protected glutamate 38 (95 mg, 0.7
mmol), Et2NH (100 mg, 1.4 mmol), and Pd(PPh3)4 (4
mg, 0.004 mmol) in 5 mL of methylene chloride was
stirred for 30 min at 25 ꢀC. The reaction mixture was
concentrated under reduced pressure. The residue was
recrystallized from CH2Cl2/ether system to afford the
Preparation of glutamate (35). A solution of DCC (3.15
mmol) and 5 mg of DMAP in 8 mL of CH2Cl2 was
added slowly to a solution of protected glutamic acid
34 (3 mmol) and N3-PEG-NH2 20 (3.3 mmol) in 16 mL
of THF at 0 ꢀC. The reaction mixture was stirred for
15 h at room temperature. After filtering to remove
solid, the filtrate was concentrated and chromato-
graphed with EtOAc/n-hexane to obtain the desired
1
desired product (72 mg, 87%). H NMR (DMSO-d6) d
9.36 (d, J=8.4 Hz, 1H, NH), 8.02 (br s, 1H), 7.93 (d,
J=7.8 Hz, 2H), 7.90 (d, J=10.8 Hz, 2H), 7.70 (t, J=7.2
Hz, 1H), 7.64–7.36 (m, 10H), 7.19(m, 1H), 7.19(m,
1H), 6.29(s, 1H), 5.85 (t, J=7.6 Hz, 1H), 5.39–5.28 (m,
4H), 4.90 (d, J=9Hz, 1H), 4.74 (s, 1H), 4.63 (t, J=8.1
Hz, 1H), 4.00 (m, 3H), 3.69(d, J=6.6 Hz, 1H), 3.46–
3.14 (m, 14H), 2.98 (m, 2H), 2.24 (m, 3H), 2.14 (s, 3H),
2.01 (s, 3H), 1.87 (m, 3H), 1.78 (s, 3H), 1.70 (m, 2H),
1.60 (s, 3H), 1.00 (s, 3H), 0.97 (s, 3H); LRMS (FAB)
m/z 1200.8 (M), 1223.8 (M+Na+).
1
product in 94% yield. H NMR (CDCl3) d 6.36 (br s,
1H, NH), 5.92–5.81 (m, 3H, involving NH), 5.33–5.16
(m, 4H), 4.60 (d, J=5.7 Hz, 2H), 4.53 (d, J=5.4 Hz,
2H), 4.32–4.30 (m, 1H), 3.66–3.57 (m, 10H), 3.52 (t,
J=5.1 Hz, 2H), 3.43–3.40 (m, 2H), 3.36 (t, J=4.8 Hz,
2H), 2.30–2.25 (m, 2H), 2.22–2.15 (m, 1H), 2.02–1.95
(m, 1H); 13C NMR (75 MHz, CDCl3) d 172.0, 171.8,
156.2, 132.7, 131.6, 119.0, 117.9, 70.8, 70.7, 70.6, 70.3,
70.1, 69.8, 66.1, 65.9, 53.8, 50.7, 39.4, 32.4, 28.2;
LRMS (CI) m/z 472 (M+H), 388, 219, 182, 170;
HRMS (CI) calcd for C20H34N5O8; (M+H) 472.2407,
found 472.2412.
Preparation of Taxol–7-carbamoyl-PEG-folate (40).
MTBD (10 mg, 0.063 mmol) was added to a solution

2410
J. W. Lee et al. / Bioorg. Med. Chem. 10 (2002) 2397–2414
of Taxol–7-carbamoyl-PEG-glutamic acid 39 (25 mg,
0.02 mmol) and pteroyl azide (8.5 mg, 0.025 mmol) in 1
mL of DMSO which the heterogeneous solution became
homogeneous within 30 seconds. The solution was left
to stir for 2 h at 25 ꢀC and was then loaded directly onto
a MPLC column (LiChroprop C18, 310Â25 mm) and
developed by 20 mL of 50 mM NH4HCO3 and then
73:27 50mM NH4HCO3/CH3CN, followed by 60:40
50 mM NH4HCO3/CH3CN for purification. After
lyopholization the desired product was obtained (17 mg,
reaction mixture was washed with 0.1 N HCl, dried,
concentrated, followed by column chromatography
(EtOA/MeOH=10:1) to afford the desired product.
1
43. 93% yield; H NMR (CDCl3) d 8.14 (d, J=8.5 Hz,
2H), 7.76 (d, J=7.5 Hz, 2H), 7.62 (t, J=7.5 Hz, 1H),
7.54–7.36 (m, 10H), 7.24 (br s, 1H), 7.02 (br s, 1H), 6.49
(br s, 1H), 6.30 (s, 1H), 6.23 (d, J=9.0 Hz, 1H), 6.02
(dd, J=9.0, 3.6 Hz, 1H), 5.95–5.81 (m, 3H), 5.68 (d,
J=6.9Hz, 1H), 5.59(d, J=3.6 Hz, 1H), 5.34–5.17 (m,
4H), 4.98 (d, J=7.8 Hz, 1H), 4.58 (d, J=6.0 Hz, 1H),
4.53 (d, J=5.4 Hz, 2H), 4.44 (m, 1H), 4.33–4.18 (m,
5H), 4.02 (s, 2H), 3.81 (d, J=6.9Hz, 1H), 3.65–3.56 (m,
10H), 3.50 (m, 4H), 3.38 (t, J=5.7 Hz, 2H), 2.56 (m,
1H), 2.46 (s, 3H), 2.34–2.21 (m, 4H), 2.23 (s, 3H), 2.14
(m, 1H), 1.98 (m, 1H), 1.93 (s, 3H), 1.89 (m, 1H), 1.69
(s, 3H), 1.23 (s, 3H), 1.14 (s, 3H); LRMS (FAB) m/z
1397.2 (M).
1
55%) as a yellow solid. H NMR (DMSO-d6) d 10.11
(d, J=7.5 Hz, 1H), 8.53 (s, 1H), 8.09(br s, 1H), 7.90 (d,
J=7.2 Hz, 4H), 7.71 (t, J=7.2 Hz, 1H), 7.64–7.34 (m,
10H), 7.58 (d, J=8.1 Hz, 2H), 7.14 (br s, 1H), 7.08 (br,
1H), 6.95 (br, 1H), 6.81 (br, 1H), 6.63 (d, 8.1 Hz, 2H),
6.28 (s, 1H), 5.81 (t, J=8.1 Hz, 1H), 5.35 (d, 5.7 Hz,
2H), 5.21 (t, J=4.5 Hz, 2H), 4.80 (m, 2H), 4.67 (s, 1H),
4.41 (t, J=4.8 Hz, 2H), 4.12 (m, 1H), 4.00 (s, 2H), 3.66
(d, J=6.6 Hz, 1H), 3.43–3.04 (m, 18H), 2.25 (m, 2H),
2.10 (s, 3H), 2.04 (s, 3H), 1.91–1.81 (m, 6H), 1.77 (s,
3H), 1.59(s, 3H), 1.00 (s, 3H), 0.98 (s, 3H); LRMS
(FAB) m/z 1494.5 (M), 1517.5 (M+Na+).
1
44. 95% yield; H NMR (CDCl3) d 8.21 (d, J=0.9Hz,
1H), 8.15 (d, J=7.2 Hz, 2H), 7.82 (d, J=7.2 Hz, 2H),
7.62 (t, J=7.2 Hz, 1H), 7.55–7.36 (m, 10H), 7.32 (m,
1H), 6.73 (d, J=7.2 Hz, 1H), 6.49(br s, 1H), 6.31 (s,
1H), 6.22 (d, J=9.0 Hz, 1H), 6.01 (dd, J=9.3, 3.9 Hz,
1H), 5.97–5.81 (m, 3H), 5.68 (d, J=7.5 Hz, 1H), 5.55 (d,
J=3.3 Hz, 1H), 5.36–5.17 (m, 4H), 4.97 (d, J=8.4 Hz,
1H), 4.58 (d, J=5.7 Hz, 1H), 4.53 (d, J=5.4 Hz, 2H),
4.42 (m, 1H), 4.35 (m, 1H), 4.32 (A of AB, d, J=8.4 Hz,
1H), 4.20 (B of AB, d, J=8.1 Hz, 1H), 3.81 (d, J=6.9
Hz, 1H), 3.62–3.43 (m, 16H), 3.37 (s, 2H), 3.25 (s, 2H),
2.56 (m, 1H), 2.48 (s, 3H), 2.38 (s, 3H), 2.34–2.21 (m,
4H), 2.23 (s, 3H), 2.14 (m, 1H), 1.98 (m, 1H), 1.94 (s,
3H), 1.89(m, 1H), 1.68 (s, 3H), 1.23 (s, 3H), 1.14 (s,
3H); LRMS (FAB) m/z 1410.0 (M).
Preparation of glutamyl-PEG-carboxylic acid (41 and
42). A solution of amino-PEG-glutamate 36 (0.9g, 2
mmol), diglycolic anhydride 21 (or 22) (3 mmol), and
DMAP (20 mg, 0.16 mmol) in 20 mL of CH2Cl2 was
stirred for 36 h at 25 ꢀC. H2O (0.5 mL) was added to the
solution and stirred for 24 h to destroy excess anhydride
and then dried, filtered, and concentrated to afford the
desired product.
41. 95% yield; 1H NMR (CDCl3) d 7.61 (br s, 1H, NH),
6.72 (br s, 1H, NH), 5.96–5.83 (m, 3H, including NH),
5.35–5.19(m, 4H), 4.62 (d, J=5.7 Hz, 2H), 4.55 (d,
J=5.4 Hz, 2H), 4.35 (m, 1H), 4.17 (s, 2H), 4.11 (s, 2H),
3.66–3.56 (m, 12H), 3.52 (t, J=4.8 Hz, 2H), 3.44 (br s,
2H), 2.32 (t, J=6.9Hz, 2H), 2.20 (m, 1H), 2.02 (m, 1H);
13C NMR (CDCl3, 75 MHz) d 172.8, 171.9, 171.7,
170.0, 156.4, 132.7, 131.6, 119.0, 118.0, 71.5, 70.6, 70.5,
70.3, 70.1, 69.9, 69.7, 69.2, 66.2, 66.0, 53.8, 39.4, 38.9,
32.4, 28.2; LRMS (FAB) m/z 562.2 (M+H); HRMS
(FAB) calcd for C24H40N3O12; (M+H) 562.2612, found
562.2594.
Preparation of Taxol–20-PEG-glutamic acid (45 and 46).
PhSiH3 (20 mg, 0.18 mmol) and Pd(PPh3)4 (4 mg,
0.004 mmol) was added to a solution of protected glu-
tamate 43 (or 44) (0.09mmol) in 5 mL of methylene
chloride and stirred for 1 h at 25 ꢀC. One drop of
MeOH was added and the reaction stirred for 5 min.
The reaction mixture was concentrated and the residue
was recrystallized from CH2Cl2/diethyl ether.
1
45. 92% yield; H NMR (DMSO-d6) d 9.34 (d, J=8.1
42. 94% yield; 1H NMR (CDCl3) d 7.98 (br s, 1H, NH),
6.90 (br s, 1H, NH), 6.05 (d, J=6.3 Hz, 1H, NH), 5.95–
5.82 (m, 2H), 5.34–5.18 (m, 4H), 4.62 (d, J=5.7 Hz,
2H), 4.55 (d, J=5.4 Hz, 2H), 4.33 (m, 1H), 3.64–3.56
(m, 12H), 3.48–3.39(m, 4H), 3.36 (s, 4H), 2.54 (s, 3H),
2.33 (t, J=6.9Hz, 2H), 2.19(m, 1H), 2.03 (m, 1H); 13C
NMR (CDCl3, 75 MHz) d 172.6, 172.0, 156.3, 132.7,
131.6, 119.0, 118.0, 70.6, 70.4, 70.3, 70.1, 69.8, 66.2,
66.1, 66.0, 60.5, 59.2, 53.9, 43.6, 39.3, 39.1, 32.4, 28.2;
LRMS (FAB) m/z 575.2(M+H); HRMS (FAB) calcd
for C25H43N4O11; (M+H) 575.2928, found 575.2903.
Hz, 1H, NH), 8.04 (br s, 1H), 7.95 (d, J=6.9Hz, 2H),
7.83 (d, J=7.2 Hz, 2H), 7.75 (t, J=7.2 Hz, 1H), 7.72–
7.37 (m, 10H), 7.15 (m, 1H), 6.27 (s, 1H), 5.79(t, J=7.6
Hz, 1H), 5.49(t, J=8.4 Hz, 1H), 5.40 (t, J=4.5 Hz,
1H), 5.38 (d, J=5.1 Hz, 1H), 4.88 (d, J=9.6 Hz, 1H),
4.62 (br s, 1H), 4.33 (s, 2H), 4.08 (m, 1H), 3.98 (d,
J=9.0 Hz, 2H), 3.94 (s, 2H), 3.55 (d, J=7.2 Hz, 1H),
3.48–3.15 (m, 16H), 2.25 (m, 1H), 2.22 (t, J=6.9Hz,
2H), 2.19(s, 3H), 2.08 (s, 3H), 1.88–1.80 (m, 3H), 1.77
(s, 3H), 1.56 (m, 2H), 1.46 (s, 3H), 1.00 (s, 3H), 0.97 (s,
3H); LRMS (FAB) m/z 1273.2 (M).
Preparation of Taxol–20-PEG-glutamate (43 and 44).
DIPC (16 mg, 0.12 mmol), Taxol 14 (104 mg, 0.12
mmol), and DMAP (10 mg) were added to a solution of
gulutamyl-PEG-carboxylic acid 41 (or 42) (0.1 mmol) in
5 mL of CH2Cl2 at 0 ꢀC. The resulting solution was
allowed to warm to 25 ꢀC and stirred for 12 h. The
46. 93% yield; H NMR (DMSO-d6) d 9.34 (d, J=8.1
1
Hz, 1H, NH), 8.02 (t, J=5.4 Hz, 1H), 7.94 (d, J=6.9
Hz, 2H), 7.83 (d, J=6.9Hz, 2H), 7.77–7.36 (m, 11H),
7.15 (m, 1H), 6.26 (s, 1H), 5.78 (t, J=9.0 Hz, 1H), 5.49
(t, J=8.4 Hz, 1H), 5.40 (t, J=5.1 Hz, 1H), 5.38 (d,
J=9.4 Hz, 1H), 4.88 (d, J=9.3 Hz, 1H), 4.61 (br s, 1H),

J. W. Lee et al. / Bioorg. Med. Chem. 10 (2002) 2397–2414
2411
4.08 (m, 1H), 3.99 (d, J=8.4 Hz, 2H), 3.96 (d, J=7.8
Hz, 2H), 3.55 (d, J=7.2 Hz, 1H), 3.51–3.12 (m, 18H),
3.11 (s, 2H), 2.26 (s, 3H), 2.25 (m, 1H), 2.22 (t, J=6.9
Hz, 2H), 2.20 (s, 3H), 2.07 (s, 3H), 1.88–1.80 (m, 3H),
1.75 (s, 3H), 1.56 (m, 2H), 1.46 (s, 3H), 0.99 (s, 3H), 0.96
(s, 3H); LRMS (FAB) m/z 1286.0 (M).
Preparation of Taxol–7-PEG-glutamic acid (51 and 52).
Et2NH (0.12 mL, 1.2 mmol) and Pd(PPh3)4 (4 mg,
0.004 mmol) was added to a solution of protected glu-
tamate 49 (or 50) (0.04 mmol) in 3 mL of methylene
chloride and stirred for 1 h at 25 ꢀC. The reaction mix-
ture was concentrated and the residue was recrystallized
from CH2Cl2/ether system to afford the desired product.
Preparation of Taxol–20-PEG-folate (47 and 48). Hunig
base (25 mg, 0.2 mmol) was added to a solution of
Taxol-PEG-glutamic acid 45 (or 46) (0.04 mmol) and
pteroyl azide (7 mg, 0.02 mmol) in 1.2 mL of DMSO
and stirred for 44 h at 25 ꢀC. After removing the volatile
components, the reaction mixture was loaded directly
onto a MPLC column (LiChroprop C18, 310Â25 mm)
and developed by 20 mL of 5mM NH4HCO3 and then
80:20 5mM NH4HCO3/CH3CN, followed by 73:27
5mM NH4HCO3/CH3CN for purification. After lyo-
pholization, the desired product was obtained in low
yield. 47: 25% yield as a yellow solid; LRMS (FAB) m/z
1566.8 (M). 48: 30% yield as a yellow solid; LRMS
(FAB) m/z 1579.8 (M).
1
51. 93% yield; H NMR (DMSO-d6) d 9.22 (d, J=7.9
Hz, 1H, NH), 8.02 (m, 1H), 7.93 (d, J=7.8 Hz, 2H),
7.89(d, J=7.2 Hz, 2H), 7.73–7.37 (m, 11H), 7.29(m,
1H), 7.20 (m, 1H), 6.02 (s, 1H), 5.87 (t, J=7.6 Hz, 1H),
5.45–5.35 (m, 4H), 4.93 (d, J=8.9Hz, 1H), 4.81 (s, 1H),
4.63 (d, J=8.1 Hz, 1H), 4.03 (m, 4H), 3.91 (s, 2H), 3.67
(d, J=7.2 Hz, 1H), 3.47–3.13 (m, 16H), 2.22 (m, 4H),
2.18 (s, 3H), 2.09(s, 3H), 1.81 (m, 4H), 1.73 (s, 3H), 1.62
(s, 3H), 1.01 (s, 3H), 0.95 (s, 3H); LRMS (FAB) m/z
1272.8 (M).
1
52. 96% yield; H NMR (DMSO-d6) d 9.22 (d, J=8.7
Hz, 1H, NH), 8.01 (t, J=5.4 Hz, 1H), 7.94 (d, J=7.2
Hz, 2H), 7.88 (d, J=6.6 Hz, 2H), 7.73–7.59(m, 4H),
7.55–7.41 (m, 3H), 7.37 (m, 4H), 7.19(m, 1H), 6.04 (s,
1H), 5.87 (t, J=7.6 Hz, 1H), 5.44–5.32 (m, 4H), 4.94 (d,
J=8.9Hz, 1H), 4.79(s, 1H), 4.63 (d, J=8.9Hz, 1H),
4.43 (m, 1H), 4.02 (br s, 2H), 3.70 (d, J=7.2 Hz, 1H),
3.47–3.15 (m, 18H), 3.03 (s, 2H), 2.25 (2, 3H), 2.20 (m,
4H), 2.17(s, 3H), 2.08 (s, 3H), 1.80 (m, 4H), 1.73 (s, 3H),
1.63 (s, 3H), 1.01 (s, 3H), 0.95 (s, 3H); LRMS (FAB)
m/z 1286.0 (M).
Preparation of Taxol–7-PEG-glutamate (49 and 50).
DIPC (11 mg, 0.08 mmol), 20-alloc-Taxol 15 (70 mg,
0.075 mmol), and DMAP (10 mg) were added to a
solution of gulutamyl-PEG-carboxylic acid 41 (or 42)
(0.082 mmol) in 4 mL of CH2Cl2 at 0 ꢀC. The resulting
solution was allowed to warm to 25 ꢀC and stirred for
12 h. The reaction mixture was washed with 0.1N HCl,
dried, and concentrated, followed by column chroma-
tography (EtOA/MeOH=10:1) to afford the desired
products.
Preparation of Taxol–7-PEG-folate (53 and 54). Hunig
base (25 mg, 0.2 mmol) was added to a solution of
Taxol–PEG-glutamic acid 51 (or 52) (0.023 mmol) and
Pteroyl azide (5 mg, 0.015 mmol) in 1.2 mL of DMSO
and stirred for 44 h at 25 ꢀC. After removing the volatile
components, the reaction mixture was loaded directly
onto a MPLC column (LiChroprop C18, 310Â25 mm)
and developed by 20 mL of 5 mM NH4HCO3 and then
80:20 5 mM NH4HCO3/CH3CN, followed by 70:30 5
mM NH4HCO3/CH3CN for purification. After lyopho-
lization obtained the desired product. 53: 39% yield as a
yellow solid; LRMS (FAB) m/z 1566.8 (M). 54: 35%
yield as a yellow solid; LRMS (FAB) m/z 1581.2
(M+H).
1
49. 52% yield (92% BRSM); H NMR (CDCl3) d 8.12
(d, J=8.1 Hz, 2H), 7.75 (d, J=8.4 Hz, 2H), 7.62 (t,
J=7.5 Hz, 1H), 7.54–7.35 (m, 10H), 7.23 (br s, 1H),
7.02 (d, J=9.0 Hz, 1H), 6.54 (br s, 1H), 6.22 (t, J=8.7
Hz, 1H), 6.19(s, 1H), 5.9(dd,
J=9.3, 2.4 Hz, 1H),
5.96–5.82 (m, 4H), 5.68 (d, J=6.9Hz, 1H), 5.65 (dd,
J=5.5, 3.3 Hz, 1H), 5.43 (d, J=2.4 Hz, 1H), 5.38–5.17
(m, 6H), 4.96 (d, J=8.7 Hz, 1H), 4.61 (m, 4H), 4.53 (d,
J=5.4 Hz, 2H), 4.35–4.07 (m, 7H), 3.94 (d, J=6.9Hz,
1H), 3.64–3.58 (m, 10H), 3.51 (m, 4H), 3.42 (t, J=5.7
Hz, 2H), 2.61 (m, 1H), 2.46 (s, 3H), 2.38 (m, 1H), 2.26
(t, J=6.6 Hz, 2H), 2.21 (m, 2H), 2.15 (s, 3H), 2.03 (m,
1H), 1.97 (s, 1H), 1.90 (m, 1H), 1.79 (s, 3H), 1.21 (s,
3H), 1.14 (s, 3H); LRMS (FAB) m/z 1481 (M+H), 1503
(M+Na).
Preparation of azido-PEG-iminodiacetic diallylester
(55). K2CO3 (1.42 g, 10.3 mmol) was added to a solu-
tion of azido-PEG-amine 20 (0.9g, 4 mmol) and allyl
bromoacetate (1.5 g, 8.7 mmol) in 40 mL of acetonitrile
and stirred for 30 h at room temperature. The reaction
mixture was filtered and evaporated. The residue was
dissolved in 100 mL of methylene chloride and then
washed with brine, dried, and concentrated to afford the
1
50. 57% yield (98% BRSM); H NMR (CDCl3) d 8.12
(d, J=8.1 Hz, 2H), 7.75 (d, J=8.4 Hz, 2H), 7.62 (t,
J=7.5 Hz, 1H), 7.55–7.38 (m, 10H), 7.32 (m, 1H), 7.00
(d, J=9.3 Hz, 1H), 6.50 (br s, 1H), 6.23 (t, J=8.7 Hz,
1H), 6.22 (s, 1H), 6.00–5.85 (m, 4H), 5.99 (dd, J=9.3,
2.4 Hz, 1H), 5.69(d, J=6.9Hz, 1H), 5.64 (dd, J=5.5,
3.3 Hz, 1H), 5.43 (d, J=2.7 Hz, 1H), 5.39–5.18 (m, 6H),
4.97 (d, J=8.7 Hz, 1H), 4.62 (m, 4H), 4.54 (d, J=5.4
Hz, 2H), 4.42 (m, 1H), 4.34 (d, J=8.1 Hz, 1H), 4.19(d,
J=8.4 Hz, 1H), 3.95 (d, J=6.9Hz, 1H), 3.64–3.42 (m,
16H), 3.44 (s, 2H), 3.23 (s, 2H), 2.61 (m, 1H), 2.46 (s,
3H), 2.41 (s, 3H), 2.37–2.04 (m, 5H), 2.14 (s, 3H), 1.98
(s, 3H), 1.93 (m, 2H), 1.80 (s, 3H), 1.21 (s, 3H), 1.15 (s,
3H); LRMS (FAB) m/z 1494.2 (M+H).
1
desired product 55 (1.65 g, 97%). H NMR (CDCl3) d
5.95–5.86 (m, 2H), 5.34–5.21 (m, 4H), 4.59 (d, J=4.8
Hz, 4H), 3.68–3.56 (m, 16H), 3.38 (t, J=5.1 Hz, 2H),
2.97 (t, J=5.7 Hz, 2H); 13C NMR (CDCl3, 75 MHz) d
171.1, 132.1, 118.6, 70.75, 70.73, 70.67, 70.50, 70.41,
70.11, 65.17, 55.85, 53.7, 50.7.
Preparation of azido-PEG-iminodiacetic acid (56).
PhSiH3 (0.92 mL, 7.4 mmol) and Pd(PPh3)4 (80 mg,

2412
J. W. Lee et al. / Bioorg. Med. Chem. 10 (2002) 2397–2414
0.068 mmol) was added to a solution of diallyl ester 55
(0.7 g, 1.7 mmol) in 17 mL of methylene chloride and
stirred for 1 h at 25 ꢀC. Then 0.5 mL of MeOH added
and stirred for 5 min. The reaction mixture was con-
centrated and the residue was washed with EtOAc twice
by decantation to afford the desired product 56 (0.52 g,
mg, 0.003 mmol) was added to a solution of protected
glutamate 59 (150 mg, 0.09mmol) in 5 mL of methylene
chloride and stirred for 1 h at 25 ꢀC. The reaction mix-
ture was concentrated and the residue was recrystallized
from CH2Cl2/ether to afford the desired product 60 (126
1
mg, 96%). H NMR (DMSO-d6) d 9.23 (d, J=8.7 Hz,
1
92%). H NMR (CDCl3) d 3.87 and 3.75 (br s, 4H),
3.65–3.60 (m, 12H), 3.36 (t, J=5.1 Hz, 4H).
1H, NH), 8.03 (t, J=5.4 Hz, 1H), 7.94 (d, J=7.2 Hz,
2H), 7.88 (d, J=6.6 Hz, 2H), 7.73–7.59(m, 4H), 7.55–
7.41 (m, 3H), 7.37 (m, 4H), 7.19(m, 1H), 6.04 (s, 1H),
5.87 (t, J=7.6 Hz, 1H), 5.44–5.32 (m, 4H), 4.94 (d,
J=8.9Hz, 1H), 4.79(s, 1H), 4.63 (d, J=8.9Hz, 1H),
4.43 (m, 1H), 4.02 (br s, 2H), 3.70 (d, J=7.2 Hz, 1H),
3.48–3.15 (m, 34H), 2.68 (br s, 2H), 2.20 (m, 4H), 2.17
(s, 3H), 2.08 (s, 3H), 1.80 (m, 4H), 1.73 (s, 3H), 1.63 (s,
3H), 1.01 (s, 3H), 0.95 (s, 3H); LRMS (FAB) m/z 1474.0
(M+H).
Preparation of azido-PEG-iminodiacetic anhydride (57).
A solution of DCC (0.28 g, 1.38 mmol) in 4 mL of
CH2Cl2 was added to a solution of diacid 56 (0.52 g,
1.56 mmol) in 10 mL of CH2Cl2 at 0 ꢀC and stirred for 5
h at 25 ꢀC. 10 mL of hexane was added and then stirred
for 5 min. The resulting solution was filtered and con-
centrated to give the desired product 57 as a crude anhy-
1
dride, which was used in the next reaction. H NMR
(CDCl3) d 3.69–3.59 (m, 16H, including a singlet for two
CH2), 3.39(t, J=5.1 Hz, 4H), 2.81 (t, J=5.1 Hz, 4H).
Preparation of Taxol–azido-PEG-Folate (61). MTBD
(24 mg, 0.157 mmol) was added to a solution of Taxol–
PEG-glutamic acid 60 (110 mg, 0.075 mmol) and pter-
oyl azide (28 mg, 0.082 mmol) in 1.5 mL of DMSO
which became homogeneous within 30 s. The solution
was left to stir for 2 h at 25 ꢀC and then loaded directly
onto a MPLC column (LiChroprop C18, 310Â25 mm)
and developed by 20 mL of 5mM NH4HCO3 and then
80:20 5mM NH4HCO3/CH3CN, followed by 70:30
5mM NH4HCO3/CH3CN for purification. After
lyopholization the desired product 61 was obtained
(57 mg, 43%) as a yellow solid. LRMS (FAB) m/z
1767.2 (M).
Preparation of glutamyl-PEG-carboxylic acid (58). A
solution of amino-PEG-glutamate 36 (0.45 g, 1 mmol)
and azido-PEG-iminodiacetic anhydride 57 (from above
reaction) in 10 mL of CH2Cl2 was stirred for 12 h at
25 ꢀC. The reaction solution was concentrated and
chromatographed with CH2Cl2/MeOH (5:1) or EtOAc/
MeOH (1:1) to afford the desired product 58 (0.59g,
1
78% for two steps). H NMR (CDCl3) d 7.92 (br s, 1H,
NH), 6.71 (br s, 1H, NH), 5.95–5.84 (m, 3H including
NH), 5.36–5.19(m, 4H), 4.63 (d, J=5.7 Hz, 2H), 4.56
(d, J=5.4 Hz, 2H), 4.34 (m, 1H), 3.70–3.54 (m, 24H),
3.51–3.34 (m, 10H, m which has two singlets and three
triplets), 2.91 (t, J=4.8 Hz, 2H), 2.33 (t, J=6.6 Hz, 2H),
2.21 (m, 1H), 2.04 (m, 1H).
Preparation of Taxol–amino-PEG-folate (62). PhSH
(0.12 mmol) and Et3N (0.09mmol) were added to a
solution of SnCl2 (0.03 mmol) in 4 mL of DMF. After
5 min, azide compound (0.02 mmol) was added and
stirred for 1 h at 25 ꢀC. MeOH (2 mL) was added and
the reaction was stirred for 1 h. The resulting solution
was concentrated to leave ca. 1 mL of volume and
then EtOAc/CH2Cl2 (5:1) added to effect precipita-
tion. The solid was taken after centrifugation by dis-
carding the supernatant and washed with CH2Cl2/
EtOAc (5:1). 95% yield; LRMS (PDMS) m/z 1742.6
(M+H).
Preparation of 7-glutamyl-PEG-aminoacetyl–Taxol (59).
A solution of DCC (49mg, 0.24 mmol) and 5 mg of
DMAP in 1 mL of CH2Cl2 was added to a solution of
acid 58 (200 mg, 0.26 mmol) and 20-alloc-Taxol 15 (154
mg, 0.16 mmol) in 4 mL of CH2Cl2 at 0 ꢀC. The result-
ing solution was allowed to warm to 25 ꢀC and stirred
for 12 h. The reaction mixture was filtered, con-
centrated, and chromatographed from EtOAc and then
EtOAc/MeOH (10:1) to afford the desired product 59
(256 mg, 93%). 1H NMR (CDCl3) d 8.13 (d, J=6.9Hz,
2H), 7.90 (br s, 1H), 7.75 (d, J=6.9Hz, 2H), 7.63 (t,
J=7.2 Hz, 1H), 7.55–7.48 (m, 4H), 7.46–7.36 (m, 6H),
6.96 (d, J=9.3 Hz, 1H, NH), 6.51 (br s, 1H, NH), 6.26 (t,
J=8.4 Hz, 1H), 6.21 (s, 1H), 5.98 (dd, J=9.3, 2.4 Hz, 1H),
5.95–5.83 (m, 4H), 5.69 (d, J=6.9Hz, 1H), 5.60 (dd,
J=6.9, 3.0 Hz, 1H), 5.43 (d, J=2.4 Hz, 1H), 5.39–5.18 (m,
6H), 4.96 (d, J=8.4 Hz, 1H), 4.66–4.58 (m, 4H, two
allylic CH2), 4.55 (d, J=5.4 Hz, 2H), 4.43 (m, 1H), 4.34
(A of AB, d, J=8.4 Hz, 1H), 4.18 (B of AB, d, J=8.4
Hz, 1H), 3.94 (d, J=6.6 Hz, 1H), 3.70–3.50 (m, 24H),
3.49–3.36 (m, 10H, m which has two singlets and three
triplet), 2.87 (m, 2H), 2.62 (m, 1H), 2.46 (s, 3H), 2.38 (m,
1H), 2.31 (t, J=6.6 Hz, 2H), 2.23 (m, 2H), 2.14 (s, 3H),
2.02 (m, 1H), 1.98 (s, 3H), 1.83 (m, 1H), 1.80 (s, 3H), 1.21
(s, 3H), 1.14 (s, 3H); LRMS (FAB) m/z 1682.2 (M+H).
Preparation of Y-shaped Taxol–folate-DTPA (63).
Amino-Taxol–folate 62 (8 mg, 0.0037 mmol) was
added to a solution of DTPA-dianhydride (5.2 mg,
0.015 mmol) in 2 mL of DMSO and then water (0.0026
mmol) was added after 4 min. The resulting solution
was stirred for 3 h at 25 ꢀC and was then loaded directly
onto a MPLC column (LiChroprop C18, 310Â25 mm)
and developed by 50 mL of 5 mM NH4HCO3, 150 mL
of 90:10 5 mM NH4HCO3/CH3CN, 200 mL of 85:15
5 mM NH4HCO3/CH3CN, and then 80:20 5 mM
NH4HCO3/CH3CN for purification. After lyopholiza-
tion the desired product 63 (80%) was obtained as a
yellow solid. LRMS (MALDI) m/z 2117 (M).
Biochemical materials. 3H folic acid was purchased from
Amersham (Arlington Heights, NY, USA). Cremophor
EL and 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetra-
zolium bromide (MTT) were purchased from Sigma
(St. Louis, MO, USA). KB (human nasopharyngeal
Preparation of Taxol–7-aminoacetyl-PEG-glutamic acid
(60). Et2NH (0.28 mL, 2.7 mmol) and Pd(PPh3)4 (3

J. W. Lee et al. / Bioorg. Med. Chem. 10 (2002) 2397–2414
2413
epidermal carcinoma) and A549(human lung carci-
noma) cell cultures were received as a gift from the
Purdue Cancer Center (West Lafayette, IN, USA).
M109(murine lung carcinoma) was provided by Dr.
Alberto Gabizon, Sharet Institute of Oncology,
Hadassah-Hebrew University Medical Center and
Medical School, Jerusalem, Israel25 All tissue culture
products were supplied by Gibco BRL (Grand Island,
NY, USA). All other chemicals were obtained from
major suppliers.
The cells were incubated for 24 h at 37 ꢀC before being
exposed to 150 mL of FDMEM containing (a) Taxol,
(b) 54, or (c) 54 plus >500Â molar excess of free folic
acid at drug concentrations ranging from 1.9Â10À12 to
5Â10À7 M for taxol 14 and 4.0Â10À12 to 1.05Â10À6
M
for 54. For drug solubility purposes, <0.085% of
cremophor/ethanol (1:1) was present in the growth
media including the control. After incubation for 18 h
at 37 ꢀC, the cells were washed 2Â with warm growth
medium to remove the unbound drug and incubated
further in fresh FDMEM until the control cells became
confluent (ꢁ24 h later). Thereafter, 15 mL of 5 mg/mL
MTT dissolved in PBS, pH 7.4, was added to each well
and the cells were stained for 2 h at 37 ꢀC. The formazan
crystals that formed were solubilized in 150 mL iso-
propanol containing 0.01N HCl, and the number of
viable cells in each well was determined from the
absorbance at 570 nm in a 96-well plate reader. For
analyses of cell cytotoxicity at a single-concentration,
KB, M109or A549cells were plated in 24-well plates at
ꢁ30% confluence and treated with 2Â10À7 M Taxol 14
or 54 for 18 h at 37 ꢀC [all growth media contained
<0.034% of cremophor/ethanol (1:1)]. Afterwards, the
cells were washed 2Â and incubated further until the
control cells were confluent. MTT (50 mL of 5 mg/mL)
was then added to each well containing 450 mL
FDMEM. MTT-stained viable cells were solubilized in
1 mL of isopropanol containing 0.01N HCl. The num-
ber of viable cells in each well was determined from the
absorbance at 570 nm.
Cell culture and tumor model
Monolayers of KB and A549cell lines were grown
continuously in folate-deficient Dulbecco’s modified
Eagles medium supplemented with 10% v/v heat-inac-
tivated fetal calf serum, 100 units/mL penicillin, 100 mg/
mL streptomycin and 2 mM l-glutamine (FDMEM) at
37 ꢀC in a humidified atmosphere containing 5% CO2.
M109cells were cultured in folate-deficient RPMI-1640
(Gibco BRL) supplemented with 10% v/v heat-inacti-
vated fetal calf serum, 100 units/mL penicillin, and 100
mg/mL streptomycin.
Female Balb/c mice were purchased from Harlan Labs,
Indianapolis, IN, USA and used for therapy studies as
soon as they reached 5–6 weeks of age. Mice were fed
on a folate-deficient diet starting 2 weeks prior to tumor
inoculation and ending 1 day after the last drug treat-
ment. All animal experiments were carried out in
accordance with procedures approved by the Purdue
Animal Care and Use Committee. M109tumors were
serially maintained as subcutaneous tumors in female
Balb/c mice at 2–3-week intervals. At appropriate times,
subcutaneous tumors were harvested and regenerated
according to established procedures.25 For intra-
peritoneal (ip) tumor implant, 5Â105 M109cells at an
early passage (P0 or P1) were implanted in 400 mL
volume into each animal.
Comparative toxicity and antitumor activity in vivo
Taxol 14 and 54 were first dissolved in 100% cremophor
followed by 1:1 dilution with ethanol and stored at 4 ꢀC.
The stock solution in 1:1 cremophor/ethanol was
further diluted 1:4 with PBS, pH 7.4, and used within 30
min. Thus, the injection vehicle consisted of cremophor/
ethanol/PBS at a volume ratio of 1:1:8. For general
toxicity studies, four groups of healthy tumor-free mice
at 3 mice/group were used. The mice (adapted to folate-
deficient diet) were ip injected with (a) PBS, (b) injection
vehicle, (c) Taxol 14 or (d) 54 at a Taxol equivalent dose
of 25 mg/kg/day (i.e., 46 mg/kg/day of 54). The dosing
schedule used consisted of a total of four injections at
2-day intervals. All mice were monitored and weighed
every other day to assess any general toxicity associated
with the drug treatment. The antitumor activity study
was conducted in four groups (a–d) of mice at 10–12
mice per group. The groups were designated as follows:
(a) untreated control; (b) injection vehicle only; (c)
Taxol 14 and (d) 54 formulated in the same injection
vehicle. Starting on day 4 after ip implantation of M109
tumor cells, mice were either left untreated (a) or treated
with 0.5 mL of the injection vehicle (b), or 17.2 mg/kg/
day of Taxol 14 (c), or 32.4 mg/kg/day of 54 (d). All
treatments were performed ip at a schedule of q2dÂ8,
that is, a total of eight ip injections at 2-day intervals.
Daily survival of the animals was recorded as a measure
of therapeutic efficacy due to treatment. Animals that
survived over 88 days were considered cured when no
macroscopic tumors were found in the peritoneal cavity
after euthanasia.
3H Folic acid competition study. The Taxol–folate con-
3
jugates’ competition with H folic acid for cell surface
receptor binding was studied at 4 ꢀC. Briefly, ꢁ90%
confluent cells in 24-well plates were incubated with
3
10 nM H folic acid plus various amounts of 53, 54,
61, 63 or non-radioactive folate at a concentration
range of 0.01–10 mM. After 1 h incubation and sub-
sequent washing with cold PBS, cell-surface bound H
3
folic acid was recovered in 0.5 mL acid saline (150 mM
NaCl, 10 mM sodium acetate, pH 3.5) and subjected
to liquid scintillation counting. The ligand-stripped
cells were lysed in 0.5 mL of 1% Triton X-100/PBS
solution and assayed for the protein content by the
standard BCA method. Each data point was obtained in
triplicate.
MTT assay
The cytotoxicity of 54 and free Taxol 14 towards cul-
tured tumor cells was determined using the MTT cell
viability assay.26 For concentration-dependent cyto-
toxicity studies, KB cells were plated into 96-well plates
at a density of 5Â104 cells/well in 150 mL of FDMEM.

2414
J. W. Lee et al. / Bioorg. Med. Chem. 10 (2002) 2397–2414
8. Sparano, J. A.; Wadler, S.; Liebes, L.; Robert, N. J.;
Acknowledgements
Schwartz, E. L.; Dutcher, J. P. Cancer Res. 1993, 53, 3509.
9. Gurney, H.; Dodwell, D.; Thatcher, N.; Tattersall, M. H. N.
Ann. Oncol. 1993, 4, 103.
10. Sutton, G. Gyn. Oncol. 1993, 51, 104. Buzzoni, R.; Bona-
donna, G.; Valagussa, P.; Zambetti, M. J. Clin. Oncol. 1991, 9,
2134.
We thank Endocyte, Inc. for financial support of this
research. Dr. Doug Lantrip is thanked for assistance in
preparation of this manuscript.
11. Dodwell, D. J. Lancet 1993, 341, 614. Davidson, N. E. J.
Clin. Oncol. 1992, 10, 517.
References and Notes
12. Yamaguchi, T.; Tsurumi, H.; Kotani, T.; Yamaoka, N.;
Otsuji, E.; Kitamura, K.; Takahashi, T. Jpn. J. Cancer Res.
1994, 85, 167. Pietersz, G. A.; Krauer, K. J. Drug Target 1994,
2, 183. Deonarain, M. P.; Epenetos, A. A. Br. J. Cancer 1994,
1. Folate mediated multiple drug delivery 5. For previous
papers in this series, see: Kim, H. Y.; Lantrip, D. A.; Fuchs, P.
L. Org Lett. 2001, 3, 2137 and ref 18.
2. Gehl, J.; Boesgaard, M.; Paaske, T.; Jensen, B. V.; Dom-
bernowsky, P. Seminars Oncol. 1996, 23, 35. Terwogt, J. M.
M.; Nuijen, B.; Ten Bokkel Huinink, W. W.; Beijnen, J. H.
Cancer Treat. Rev. 1997, 23, 87. Panchagnula, R. Int. J.
Pharm. 1998, 172, 1.
70, 786. Canevari, S.; Mezzanzanica, D.; Menard, S.; Ferrini,
´
S.; Moretta, L.; Colnaghi, M. I. Int. J. Cancer 1992, 7, 42.
13. Westerhof, G. R.; Jansen, G.; van Emmerik, N.; Kath-
mann, I.; Rijksen, G.; Jackman, A. L.; Schornagel, J. H.
Gancer Res. 1991, 51, 5507. Dixon, K. H.; Lanpher, B. C.;
Chiu, J.; Kelley, K.; Cowan, K. H. J. Biol. Chem. 1994, 269,
17. Mason, J. B.; Shoda, R.; Haskell, M.; Selhub, J.; Rosen-
berg, I. H. Biochim. Biophys. Acta 1990, 1024, 331.
14. Hayakawa, Y.; Uchiyama, M.; Kato, H.; Noyori, R.
Tetrahedron Lett. 1985, 26, 6505.
15. Dessolin, M.; Guillerez, M.-G.; Thieniet, N.; Guibe, F.;
Loffet, A. Tetrahedron Lett. 1995, 36, 5741.
16. Takahashi, T.; Tsukamoto, H.; Yarnada, H. Bioorg. Med.
Chem. Lett. 1998, 8, 113.
3. Khmelnitsky, Y. L.; Budcle, C.; Arnold, J. M.; Usyatinsky,
A.; Clark, D. S.; Dordick, J. S. J. Am. Chem. Soc. 1997, 119,
11554. Li, C.; Yu, D.-F.; Newman, R. A.; Cabral, F.;
Stephens, L. C.; Hunter, N.; Milas, L.; Wallace, S. Cancer
Res. 1998, 58, 2404. Golik, J.; Wong, H. S. L.; Chen, S. H.;
Doyle, T. W.; Wright, J. J. K.; Knipe, J.; Rose, W. C.;
Casazza, A. M.; Vyas, D. M. Bioorg. Med. Chem. Lett. 1996,
6, 1837. Nicolaou, K. C.; Riemer, C.; Kerr, M. A.; Rideout,
D.; Wrasidlo, W. Nature 1993, 364, 464. Greenwald, R. B.;
Pendri, A.; Bolikal, D. J. Org. Chem. 1995, 60, 331 and
references cited therein. Greenwald, R. B.; Gilbert, C. W.;
Pendri, A.; Conover, C. D.; Xia, J.; Martinez, A. J. Med.
Chem. 1996, 39, 424. DeBont, D. B. A.; Leenders, R. G. G.;
Haisma, H. J.; van der Meulen-Muileman, I.; Scheeren, H. W.
Bioorg. Med. Ghem. 1997, 5, 405.
4. Mathew, A. E.; Mejillano, M. R.; Nath, J. P.; Hirnes, R. H.;
Stella, V. J. J. Med. Chem. 1992, 35, 145. Au, S. M.; Hoe-
mann, M. Z.; Aube, J.; Mitscher, L. A.; Georg, G. I.; McCall,
R.; Jayasinghe, L. R. J. Med. Chem. 1995, 38, 3821. Kingston,
D. G. I. Pharm. Ther. 1991, 52, 1. Chen, S.-H.; Huang, S.;
Kant, J.; Fairchild, C.; Wei, J.; Farina, V. J. Org. Chem. 1993,
58, 5028.
5. Ford, L. G.; Brawley, O. W.; Penman, J. A.; Nayfield,
S. G.; Johnson, K. A.; Kramer, B. S. Cancer 1994, 74, 2726.
6. Chen, T.-L.; Passos-Coelho, J. L.; Noe, D. A.; Kennedy,
M. J.; Black, K. C.; Colvin, O. M.; Grochow, L. B. Cancer
Res. 1995, 55, 810. Motzer, R. J.; Gulati, S. C.; Tong, W. P.;
Menendez-Botet, C.; Lyn, P.; Mazumdar, M.; Viamis, V.; Lin,
S.; Bosl, G. J. Cancer Res. 1993, 53, 3730.
7. Peters, W. P.; Rosner, G.; Ross, M.; Vredenburgh, J.;
Meisenberg, B.; Kurtzberg, C.; Gilber, J. Blood 1993, 81, 1709.
Bitran, J. D.; Samuels, B. L.; Marsik, S.; Gambino, A.; White,
L. Gun. Cancer Res. 1995, 1, 185.
17. Schwabacher, A. W.; Lane, J. W.; Schiesher, M. W.;
Leigh, K. M.; Johnson, C. W. J. Org. Chem. 1998, 63, 1727.
18. Lee, J. W.; Fuchs, P. L. Org. Lett. 1999, 1, 179.
19. Greenwald, R. B.; Pendri, A.; Bolikal, D. J. Org. Chem.
1995, 60, 331.
20. Luo, J.; Smith, M. D.; Lantrip, D. A.; Wang, S.; Fuchs,
P. L. J. Am. Chem. Soc. 1997, 119, 10004.
21. Kingston, D. G. I. Pharm. Ther. 1991, 52, 1. Panchagnula,
R. Int. J. Pharm 1998, 172, 1. Golik, J.; Wong, H. S. L.; Chen,
S. H.; Doyle, T. W.; Wright, J. J. K.; Knipe, J.; Rose, W. C.;
Casazza, A. M.; Vyas, D. M. Bioorg. Med. Chem. Lett. 1996,
6, 1837.
22. McHugh, M.; Chang, Y. C. J. Biol. Chem. 1979, 254,
11312.
23. Wang, S.; Luo, J.; Lantrip, D. A.; Waters, D. J.; Mathias,
C. J.; Green, M. A.; Fuchs, P. L.; Low, P. S. Bioconjugate
Chem. 1997, 8, 673.
24. Lee, R. J.; Wang, S.; Low, P. S. Biochim. Biophys. Acta
1996, 1312, 237.
25. Gabizon, A.; Horowitz, A. T.; Goren, D.; Tzemach, D.;
Mandelbaum-Shavit, F.; Qazen, M. M.; Zalipsky, S. Bio-
conjugate Chem. 1999, 10, 289.
26. Milhaud, P. G.; Machy, P.; Lebleu, B.; Leserman, L. Bio-
chim. Biophys. Acta 1989, 987, 15.