Synthesis and immunological activity of water-soluble thalidomide prodrugs

Article abstract of DOI:10.1016/S0968-0896(00)00342-4

A series of new water-soluble thalidomide prodrugs was prepared. All compounds were derivatized on the nitrogen of the glutarimide ring. Esters of natural amino acids and succinic acid derivatives have been introduced by reaction with the hydroxymethyl thalidomide 2. Nicotinic acid derivatives were prepared from halomethyl derivatives. Additionally, a methoxymethyl derivative and a carboxymethyl derivative were prepared directly from thalidomide. Most compounds showed a very large increase in water solubility compared to thalidomide itself (0.012 mg/mL). The amorphous hydrochlorides of the N-methylalanine ester 8, valine ester 9, and glycylglycine ester 10, respectively, were the most soluble compounds showing solubility greater than 300 mg/mL, which equals an increase greater than 15,000-fold. The lipophilicity of the prodrugs has been determined by their HPLC capacity factors k′. The stability of selected compounds was determined. The hydrolysis rates follow pseudo-first order kinetics. In order to assess the immunological activity, the prodrugs were tested using tumor necrosis factor-α and interleukin-2 inhibition assays. Selected compounds were additionally investigated on their abililty to inhibit the local Shwartzman reaction, an assay to determine the vascular permeability. The prodrugs retained high effectiveness in the inhibition of TNF-α release. Our results indicated that the more stable prodrugs exhibited higher activity in the immunological assays. Some compounds showed higher activity than thalidomide itself, suggesting a high affine binding to the pharmacophore. In conclusion, the prodrugs exhibited high water solubility and high activity and might therefore be used in therapeutic applications.

Full text of DOI:10.1016/S0968-0896(00)00342-4

Bioorganic & Medicinal Chemistry 9 (2001) 1279±1291
Synthesis and Immunological Activity of
Water-Soluble Thalidomide Prodrugs
Sonja Hess,a,* Michaela A. Akermann,a Stephan Wnendt,b
Kai Zwingenbergerc,y and Kurt Egera
aPharmaceutical Chemistry, Institute of Pharmacy, University of Leipzig, Bruderstrasse 34, D-04103 Leipzig, Germany
È
bMolecular Pharmacology Department, Grunenthal, Zieglerstrasse 8, D-52068 Aachen, Germany
È
cMedical Scienti®c Division, Grunenthal, Zieglerstrasse 8, D-52068 Aachen, Germany
È
Received 6 November 2000; accepted 21 December 2000
AbstractÐA series of new water-soluble thalidomide prodrugs was prepared. All compounds were derivatized on the nitrogen of
the glutarimide ring. Esters of natural amino acids and succinic acid derivatives have been introduced by reaction with the
hydroxymethyl thalidomide 2. Nicotinic acid derivatives were prepared from halomethyl derivatives. Additionally, a methoxy-
methyl derivative and a carboxymethyl derivative were prepared directly from thalidomide. Most compounds showed a very large
increase in water solubility compared to thalidomide itself (0.012 mg/mL). The amorphous hydrochlorides of the N-methylalanine
ester 8, valine ester 9, and glycylglycine ester 10, respectively, were the most soluble compounds showing solubility greater than
300 mg/mL, which equals an increase greater than 15,000-fold. The lipophilicity of the prodrugs has been determined by their
HPLC capacity factors k0. The stability of selected compounds was determined. The hydrolysis rates follow pseudo-®rst order
kinetics. In order to assess the immunological activity, the prodrugs were tested using tumor necrosis factor-a and interleukin-2
inhibition assays. Selected compounds were additionally investigated on their abililty to inhibit the local Shwartzman reaction, an
assay to determine the vascular permeability. The prodrugs retained high e€ectiveness in the inhibition of TNF-a release. Our
results indicated that the more stable prodrugs exhibited higher activity in the immunological assays. Some compounds showed
higher activity than thalidomide itself, suggesting a high ane binding to the pharmacophore. In conclusion, the prodrugs exhibited
high water solubility and high activity and might therefore be used in therapeutic applications. # 2001 Elsevier Science Ltd. All
rights reserved.
Introduction
Due to its bad solubility, the therapeutic use of thali-
domide is limited. Attempts to solubilize thalidomide
with cyclodextrines did not lead to therapeutic applica-
tions.9,10 The development of water-soluble prodrugs
would be a great advantage in order to administer the
drug.
Thalidomide (2-(2,6-dioxo-3-piperidyl-isoindoline-1,3-
dione, Contergan1, 1) was developed as a sedative by
Grunenthal in the 1950's, but was shortly after its
introduction removed from the market when its terato-
genic side e€ects best described as heavy dysmelia
(stunted limb growth) were discovered.1À5 In recent
years, thalidomide has regained scienti®c interest because
of its e€ectiveness in various diseases including erythema
nodosum leprosum, graft-vs-host disease, arthritis, tuber-
culosis and conditions associated with human immuno-
de®ciency virus infections.6 In 1998, thalidomide has
been reintroduced to the market and is now under clin-
ical investigation for a series of immunomodulatory and
anti-in¯ammatory applications.7,8
Furthermore, water-soluble prodrugs might be valuable
tools to elucidate the mechanism of its bene®cial and
side e€ects. The immunological e€ects of thalidomide are
partly mediated by modulation of tumor necrosis factor-a
(TNF-a). For the rational development of thalidomide
derivatives without teratogenic side e€ects, it is abso-
lutely required to understand its molecular mechanism.
Biological experiments with poor soluble drugs such as
thalidomide are seldom reproducible and have often led
to misinterpretation. Precipitation leads to unreliable
biological results. The poor solubility of thalidomide
presumably hampered the discovery of its teratogenicity
earlier. Besides that, many of the investigated animal
*Corresponding author. Tel.: +49-341-97-36884; fax: +49-341-97-
36709; e-mail: hess@rz.uni-leipzig.de
yPresent address: Information Management, Grunenthal, Ziegle-
rstrasse 8, D-52068 Aachen, Germany.
0968-0896/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(00)00342-4

1280
S. Hess et al. / Bioorg. Med. Chem. 9 (2001) 1279±1291
species did not respond to its teratogenic side e€ects.
solubility should be greatly increased. The solubility of
the prodrugs has been determined as well as the pH values
of the aqueous solutions. In order to assess the lipophilicity
and hydrolysis kinetics of the prodrugs, an HPLC method
has been developed. Furthermore, we have investigated the
immunological e€ects of the resulting prodrugs with
respect to the introduced residues. Derivative 13 was syn-
thesized in order to serve for anity labeling of the poten-
tial biomolecule involved in the teratogenic mechanism.
Water-soluble prodrugs of thalidomide should be a
valuable tool in the ongoing search for the mechanism
of activity and teratogenicity.11À14 Up to now, more
than 30 hypotheses have been proposed. They include
hydroxyl radical mediated inhibition of angiogenesis in
embryoid bodies, inhibition of growth factors, inter-
calation with DNA synthesis as well as oxidative
damage of DNA. Recently, with the prevention of
homodimerization of N-cadherin monomers by interac-
tion of thalidomide, instead of the dimerization promot-
ing Trp2/Trp20 residues, a new concept has been
proposed.13
Results and Discussion
Chemistry
In the present study, we have prepared and investigated
a series of new water-soluble prodrugs. All variations
were introduced at the nitrogen of the glutarimide moiety,
enabling most prodrugs to be metabolized to thalidomide.
In general, Mannich bases (N-hydroxymethyl deriva-
tives) of thalidomide were prepared. Furthermore, ether
and carboxylic acid derivatives were synthesized. To
enable the identi®cation of the binding site of thalido-
mide to its orphan receptor(s), one compound for an
anity-labeling assay was additionally developed.
Amino acid derived prodrugs 7±10 were synthesized
starting from hydroxymethylthalidomide 2 and the
appropriate Boc-protected amino acids (Scheme 1)
using the dicyclohexylcarbodiimide method.16À18 To
improve yields, 4-pyrrolopyridine was added.19 Com-
pound 2 was prepared from racemic thalidomide reacted
with formaldehyde. The Boc-protection group of 3, 4, 5
and 6, respectively, was removed with tri¯uoroacetic
acid in dichloromethane followed by ion exchange with
20À22
Amberlite IR45.
Using the appropriate derivatization, prodrugs improve
the physicochemical properties of thalidomide and,
thus, bioavailability and membrane transport.15 This goal
was achieved by increasing the water solubility of thali-
domide by the introduction of both acidic and basic phy-
siologically derived moieties such as amino acids or
nicotinamide. Due to these ionizable groups, the water
Compound 13 was prepared from the Boc-protected
valine ester 5 whereby the protection group was also
removed with tri¯uoroacetic acid. The resulting residue
was then reacted with the N-hydroxysuccinimidyl deri-
vative 12 to a€ord 13.23,24 The succinic acid derivative
11 was synthesized from hydroxymethylthalidomide 2
and succinic anhydride.
Scheme 1. Synthetic route of the amino ester prodrugs 7±10, the succinyl derivative 11 and the anity-labeling compound 13. Reaction conditions:
(a) CH2O, Át; (b) N-Boc-aa, DCC, 4-PP, CH2Cl2, rt; (c) TFA (25%), ClÀ; (d) succinic anhydride, NEt3, 4-PP, CH2Cl2, rt; (e) TFA, NaHCO3, rt.

S. Hess et al. / Bioorg. Med. Chem. 9 (2001) 1279±1291
1281
In order to prepare the azido derivative 16, hydroxy-
methylthalidomide 2 was reacted with thionyl chloride
under ice cooling to give the chloromethylthalidomide
14 in good yields (Scheme 2). To increase reactivity, the
leaving group of chloromethylderivative 14 was exchanged.
Therefore, 14 was reacted with potassium iodide and
subsequently with sodium azide in acetone to a€ord 16.
compared to thalidomide. The best soluble compounds
were amorphous 8, 9 and 10 with solubilities greater
than 300 mg/mL, equivalent to more than 680 mmol/mL
thalidomide. In these cases, the solubility was increased
more than 15,000-fold. Solubility is not only dependent
on the hydrophilicity of the compounds but also
depends on the crystal structure. Due to their better
accessibility to solvents, amorphous powders are gen-
erally more soluble than highly ordered crystals. The
high increase in solubility of the amorphous amino acid
ester prodrugs should enable therapeutic use, for exam-
ple parenteral or topical application without addition of
solubilizers like cyclodextrines. The solubility of the
crystalline glycine ester 7 was increased 2780-fold, and the
solubility of the succinic acid derivative 11 was increased
920-fold. The nicotinamide derivative 17 and the nicotinic
acid derivative 19 showed lower increases of water solu-
bility, namely 280 and 170-fold, respectively.
The nicotinamide derivative 17 was obtained from
chloromethylthalidomide 14 reacted with nicotin-
amide.25À27 For the preparation of dihydropyridine
derivative 18, 17 was reduced with sodium dithionite.
Similarly, the nicotinic acid derivative 19 was prepared
from nicotinic acid and the iodomethyl derivative 15.
For the synthesis of the carboxymethyl derivative 21,
thalidomide was reacted with sodium hydride and tert-
butyl iodoacetate in DMF (Scheme 3). Again the pro-
tection group was removed with tri¯uoroacetic acid.
The methoxymethyl derivative 22 was prepared from
thalidomide and formaldehyde dimethylacetal using
phosphorus pentoxide in dichloromethane.28
Determination of the pH values of the aqueous solution
In addition to the water solubility, the pH values of the
aqueous solutions were determined. As shown in Table
1, the pH values ranged from 2.3 for the nicotinic acid
derivative 19 up to 5.7 for the nicotinamide derivative
17. The pH values of the ester prodrugs were between
4.6 and 5.0. Since thalidomide has been proven to be
very e€ective in skin lesions, some of the prepared
prodrugs may be established for topical use.
Reagents, yields and analytical data are given in the
Experimental section. All analytical data were in accor-
dance with the proposed structures.
Water solubility
Thalidomide exhibits very low water solubility
(0.012 mg/mL) which limits its therapeutic use. The
aqueous solubility of selected prodrugs has been deter-
mined using UV. As presented in Table 1, the water
solubility of the prodrugs has been greatly increased
Lipophilicity parameters
Permeation through membranes is essential for the
e€ectiveness of drugs. Therefore, not only aqueous
Scheme 2. Synthesis of the azido derivative 16 and the nicotinic acid derivative 17. Reaction conditions: (a) SOCl2, DMF; (b) KI, acetone; (c) NaN3,
acetone; (d) nicotinamide, DMF; (e) aq NaHCO3, Na2S2O3, 1 h, 0±5 ꢀC; (f) nicotinic acid, acetonitrile, 2.5 h.

1282
S. Hess et al. / Bioorg. Med. Chem. 9 (2001) 1279±1291
Scheme 3. Synthesis of the carboxymethyl derivative 21 and the methoxymethyl derivative 22. Reaction conditions: (a) NaH, DMF, tert-butyl
iodoacetate; (b) TFA, CH2Cl2, 1 h, rt; (c) P2O5, CH2Cl2, HC(OCH3)3, 72 h, rt.
Table 1. Physicochemical properties of selected compounds
Bodor reported a similar design with the successful
introduction of the nicotinamide/dihydropyridine redox
drug-delivery system to improve delivery through bio-
logical membranes.25 The lipophilic dihydropyridine
derivative 18 is supposed to penetrate easily through the
skin and after enzymatic oxidation to the quarterny 17,
the ionized form is captured in the stratum corneum.
Compound
Solubilitya
(mg/ml)
Solubility
(mmol/ml)
pH Value of the
aqueous solution
(c=5 mg/mL)
1
7
0.01236
49.7
0.00004736
0.13
n.d.
4.9
8
9
10
11
17
19
>300
>300
>300
16.7b
5.7
>0.73
>0.71
>0.68
0.043b
0.013
4.7
5.0
4.6
n.d.
5.7
2.3 (sat sol.)
Kinetics of hydrolysis and stability
The hydrolysis rates of the selected prodrugs at physio-
logical pH=7.4 were investigated in order to determine
their chemical stability. As presented in Chart 1, most pro-
drugs were hydrolyzed into hydroxymethylthalidomide (2)
and subsequently into thalidomide. The reactions were
monitored by HPLC whereby the decrease of prodrug
versus time was determined.
4.2
0.008
aDetermined in water.
bDetermined in phosphate bu€er, pH 7.4.
solubility but also lipophilicity are important physico-
chemical properties. The lipophilicity of the thalidomide
prodrugs was evaluated by means of reversed-phase
HPLC capacity factors (Table 2). Additionally, some
degradation compounds of thalidomide were investi-
gated using the same method in order to compare sta-
bility (see the following section). The log k0 values of the
prodrugs were correlated with log P values calculated by
the Broto's method using the computer program
ChemDraw Ultra (version 5).29,30 No correlation was
found for the log k0 and log P values for the investigated
ester prodrugs (Table 2).
Table 2. Retention times, capacity factors k0 and log P values of
investigated compounds
Compound
Retention time
(min)
Capacity
factors k0
Log Pa
1
2
7
8
17.9
17.2
9.2
13.3c
22.8, 24.0d
11.1
48.7
8.0
6.2
4.1
2.8
2.7
12.77
12.23
6.08
0.49
À0.42
À1.08
n.d.b
À0.51
À2.42
À0.24
n.d.b
À0.06
À0.06
À0.94
0.58
9.23c
16.54, 17.46d
7.54
Lipophilicity parameters can be used as indicators for
the permeability through membranes, suggesting the
acid 11 as possibly the drug with the highest membrane
permeability in this series. In order to achieve good
absorption of orally administered drugs, a log P value
greater than 2 is generally desired. The introduced pep-
tide bonds, however, do rather lower than increase the
log P values.31 Nevertheless, the amino acid ester
approach is successful in a variety of slightly soluble
drugs.17,32,33 This reported increase in e€ectiveness of
amino acid prodrugs is not only due to the increased
water solubility but also to the potential active transport
into cells via oligopeptide permeases.34,35
9
10
11
19
23
24
25
26
36.46
5.15
3.77
2.15
1.15
1.08
aFor the calculation of the log P values, the neutral compounds were
used.
bn.d., not determined; the log P value of this compound was not cal-
culable with the program.
cThe diastereomers were not separated with this method.
dTwo di€erent retention times were observed for the diastereomers.

S. Hess et al. / Bioorg. Med. Chem. 9 (2001) 1279±1291
1283
As shown in Chart 2, a variety of possible degradation
compounds can occur during hydrolysis.36 The main
hydrolytic compound is N-(2,6-dioxo-piperidin-3-yl)-
phthalamic acid 25 with a cleaved phthalimide ring,
whereas the glutarimide ring is more stable and, therefore,
the hydrolytic compounds 23 and 24 occur only in minor
percentages.36 The primary hydrolytic compounds 23, 24
and 25 are subjected to further hydrolysis.36 The reten-
tion times of the investigated prodrugs and degradation
compounds are given in Table 2, varying from 2.7 min
for 26 to 48.7 min for 11.
from 12 min for the N-methylalanine ester 8 up to 54
min for the relative stable valine ester 9. In the series of
the amino acid ester prodrugs 7, 8 and 9, the hydrolytic
half-lives increased with increasing steric hindrance of
the amino acid (Table 2). With the exception of 11 and
19, all investigated prodrugs were converted into thali-
domide, the parent compound. The hydrolytic products
of 11 and 19 remain to be identi®ed.
The determination of the hydrolytic products of the
prodrugs is complicated by the fact that thalidomide
undergoes hydrolytic cleavage itself. The half-life of
thalidomide, however, is longer than that of most pro-
drugs (approx. 300 min).36 Therefore, the formation of
hydrolytic compounds did not hamper the determina-
tion of the hydrolysis rates of the prodrugs with the
exception of the very stable 11.
Kinetics of the chemical hydrolysis of the prodrugs fol-
lowed a pseudo-®rst-order rate over several half-lives.
The rates of prodrug hydrolysis were within a broad
range (Table 3 and Fig. 1). On the one hand, the
hydrolytic rate constant of the glycylglycine derivative
10 was so high that it could not be determined with the
applied method. After 5 min, less than 10% of 10 was
determined. The succinic acid ester 11, on the other
hand, was the most stable prodrug. The hydrolytic rate
constant was less than 0.009 minÀ1, corresponding to a
half-life greater than 80 min. All other investigated pro-
drugs showed hydrolytic rate constants ranging from
0.013 to 0.060 minÀ1 resulting in half-lives (t1/2) ranging
Biological evaluation
The synthesized prodrugs were tested in TNF-a- and
interleukin-2 assays. The inhibition of the TNF-a release
was determined in the supernatant of human lipopoly-
saccharide-treated PBMCs.37,38 All compounds were dis-
solved in DMSO and diluted in aqueous bu€er solution.
Chart 1. Hydrolysis of the ester prodrugs.
Chart 2. Hydrolytic degradation of thalidomide.

1284
S. Hess et al. / Bioorg. Med. Chem. 9 (2001) 1279±1291
Table 3. Linearity, half-lives and hydrolytic rate constants of investigated prodrugs
Compound
Linearity of the calibration curve
(correlation coecient)
Hydrolytic rate constant
(minÀ1
Half-life
(min)
)
1
7
8
9
10
11
19
f(x)=542.44xÀ551.18 (r=0.9999)
f(x)=332.21x+3.3388 (r=0.9998)
f(x)=273.51xÀ30.805 (r=0.9998)
f(x)=272.22xÀ816.93 (r=0.9999)
n.d.b
0.0023a
0.034
0.060
0.013
300a
21
12
54
<3c
>80c
21
>0.23c
n.d.b
ꢁ0.009c
f(x)=235.32xÀ41.51 (r=1.0)
0.034
aData from ref 36
bn.d., not determined.
cEstimated from our own data.
Figure 1. Pseudo-®rst-order kinetic plots for hydrolysis of prodrugs 7±9 in physiological phosphate bu€er pH 7.4 at 37 ꢀC.
Most compounds were additionally investigated for
their ability to inhibit IL-2 in mice. The compounds
were administered intraperitoneally in mice that were
pretreated with staphylococcal super antigen to induce
IL-2 production.37,38 To investigate bioavailability,
compounds 7, 8 and 10 were additionally administered
perorally. Inhibition of IL-2 was determined from
heart-punctured blood using ELISA.
whereas the IC50 of thalidomide has been reported to be
200 mM.40 The glycine ester 7 showed 54% inhibition of
TNF-a followed by the N-methylalanine ester 8 exhi-
biting only weak inhibition of 12%. Interestingly, a
correlation between the stability of the prodrugs and the
in vitro activity has been observed: the more stable the
ester, the greater the inhibition of TNF-a. Additionally,
the very unstable glycylglycine ester 10 did not inhibit
but slightly increased the release of TNF-a. Further-
more, this observation was supported by the fact that
the Boc-protected glycine ester 3 exhibited stronger
inhibition of TNF-a than the unprotected 7.
The selected compounds thalidomide, 2 and 7, were also
tested for their ability to inhibit the local Shwartzman
reaction, an in¯ammatory lesion provoked by endo-
toxin exposition.39 Therefore, the dorsal skin of male
white mice was treated with lipopolysaccharide and
TNF-a. The inhibition of the extravasation of the dye
Evans Blue that was applied to the surrounding tissue
was determined using UV.
This ®nding was unexpected and showed that the pro-
drugs exhibited an e€ect of their own, rather than only
the e€ect of the released thalidomide. It can be assumed
that the ester prodrug 9 is binding itself with high anity
at the thalidomide recognition site of the pharmacophore.
Inhibition of TNF-ꢀ release
The highly e€ective valine ester 9 was used to develop
compound 13. After radioactive labeling with 125I, 13
may be used for anity labeling.23,24 The ability of 13 to
inhibit TNF-a is comparable to thalidomide. Therefore,
13 was developed as a pharmacological tool for the
investigation of the binding sites of thalidomide.
The amino acid ester prodrugs 3 and 7±10 in which the
Mannich base 2 is derivatized with a glycine, a valine, a
N-methylalanine or a glycylglycine were investigated
(Table 4). The valine ester 9 showed a greater inhibition
(88%) than thalidomide itself. The IC50 of 9 was 4.7 mM

S. Hess et al. / Bioorg. Med. Chem. 9 (2001) 1279±1291
1285
Table 4. Inhibition of TNF-a release and interleukin-2 of thalidomide
prodrugs
weight a slightly better e€ect than thalidomide. Simi-
larly, the nicotinic acid derivative 19 showed an ED40 of
767 mmol/kg body weight. All other derivatives showed
ED40 values greater than 890 mmol/kg body weight. The
test was performed under standard conditions within
2 h. Most prodrugs were not degraded completely at
that time. In contrast to the in vitro TNF-a assay, no
signi®cant synergistic e€ects of the prodrugs compared
to thalidomide were observed.
Compound
TNF-a
Interleukin-2 i.p.
Interleukin-2 p.o.
administered
b
% inhibitiona,b administered ED40
b
(concd (mM))
(mmol/kg)
% inhibition or ED40
(mmol/kg)
1
2
65 (194)
96 (174)
663
1010
Ester derivatives
3
78 (112)
898
602
976
944
912
In order to determine oral bioavailability, compounds 7,
8 and 10 have additionally been administered perorally.
Compound 7 showed an ED40 of 686 mmol/kg body
weight and therefore similar e€ectiveness compared to
intraperitoneal (ip) application. 8 showed an ED40 of
827 mmol/kg body weight, that is a small increase com-
pared to ip application. It is very likely that this e€ect is
mainly mediated by thalidomide. In contrast to the i.p
route, perorally administered drugs are subjected to
biotransformation via ®rst-pass e€ect in the liver. In the
case of prodrugs, this can lead to fast bioactivation into
thalidomide. The degradation of 10 was so fast that no
di€erence between intravenous and peroral application
could be observed (data not shown).
7
54 (131)
12 (122)
88 (118)
À10 (114)
63 (93)
ED40 686
ED40 827
55% (913)
8
9
10
13
Nicotinic acid derivatives
17
18
19
26 (117)
21 (127)
53 (96)
767
Other prodrugs
11
16
21
22
35 (129)
14 (160)
40 (150)
24 (165)
1323
aPercent inhibition at the indicated concentration.
bMedian of at least three experimental data; coecient of variation
was less than 20%.
Local Shwartzman reaction
The nicotinic acid derivative 19 exhibited 53% inhibi-
tion, whereas the nicotinamide derivative 17 and the
dihydropyridine derivative 18 showed only weak inhibi-
tion of 26 and 21%, respectively. As presented in Table
3, the stability of 9 and 19 has been found to be very
similar. Interestingly, their e€ectiveness to inhibit TNF-
a was comparable too. Compound 19, however, was not
degraded into thalidomide, suggesting 19 itself inhibited
the release of TNF-a. Again, this is in accordance with
the above-mentioned ®nding that more stable prodrugs
showed better activity.
The e€ects on the local Shwartzman reaction were
recorded in order to determine vascular permeability.
Thalidomide inhibited the extravasation of Evans Blue
by 48% after three dosages of 40 mg/mL with an ED50
of 178 mg/kg body weight. Hydroxymethyl derivative 2
and glycine ester 7 inhibited the extravasation in the
same range, namely 40%.
Conclusions
Using the prodrug approach, readily water-soluble
derivatives were prepared. The water solubility not only
depended on the introduced residue but also on the
crystal form. Amorphous 8, 9 and 10 showed 15,000-
fold increase of water solubility compared to thalido-
mide. The pH values of the aqueous solution varied
between 2.3 and 5.7, enabling therapeutic parenteral or
topical use.
Interestingly, the most e€ective compound in this series
has been proven to be hydroxymethylthalidomide (2),
inhibiting 96% of TNF-a-release, suggesting similar
binding with high anity.
Methoxymethylthalidomide 22 and the azido derivative
16 exerted only weak inhibition with 24 and 14%,
respectively, whereas the succinic acid derivative 11 and
the carboxymethyl derivative 21 have been determined
to show reasonable inhibition of TNF-a. Similar to 19,
the carboxymethyl derivative 21 cannot be metabolized
to thalidomide. This observation may also indicate that
stable prodrugs or prodrugs that delay degradation of
the intact glutarimide ring in thalidomide exhibited
more e€ective inhibition of TNF-a release.
Furthermore, the synthesis of the anity-labeling com-
pound 13 exhibiting similar TNF-a inhibition compared
to thalidomide provides a useful tool for further biolo-
gical studies to explain the mechanism of activity and
teratogenicity of thalidomide.
Most prodrugs retained activity in the immunological
assays performed. Interestingly, a positive correlation
between prodrug stability and activity was found, indicat-
ing a high ane binding of the prodrugs to the pharmaco-
phore. Initially, the development of short-lived prodrugs
was pursued in order to obtain prodrugs that release the
active compound fast. Obviously, bulky residues increased
stability by preventing degradation into thalidomide and
subsequent degradation of thalidomide. In future studies,
prodrugs with longer half-lives will be developed.
Inhibition of interleukin-2
The inhibition of interleukin-2 was determined in vivo
as described in the Experimental section. As reported in
Table 4, thalidomide showed an inhibition of IL-2 of
56% at a concentration of 194 mM. The ED40 was
determined to be 663 mmol/kg body weight. The glycine
ester 7 showed, with an ED40 of 602 mmol/kg body

1286
S. Hess et al. / Bioorg. Med. Chem. 9 (2001) 1279±1291
The majority of the new prodrugs might be useful
water-soluble tools in the ongoing search for the
mechanism of action of thalidomide. In particular, the
stable valine ester 9 with a TNF-a inhibition value of
4.7 mM (IC50) represents a new water-soluble lead for
further optimization studies for therapeutic applications.
and stability. Acetonitrile was HPLC gradient grade
quality and supplied by Merck (E. Merck, Darmstadt,
Germany). All other reagents and salts were supplied by
Aldrich (Steinheim, Germany). Ion-exchanged water
(UO 100, Letzner GmbH, Huckeswagen, Germany) was
further puri®ed by bidistillation. Before use, the aqueous
solution of sodium dihydrogen phosphate was ®ltered o€
through a cellulose acetate ®lter (0.45 mm).
Experimental
Chemistry
The hydrolytic compounds 23±26 were synthesized
according to Schumacher and Teubert.37,41
All chemicals and reagents were of the highest possible
purity. Thalidomide and N-phthaloylisoglutamine were
a generous gift from Grunenthal GmbH. NMRspectra
were measured on a Varian Gemini 300 spectrometer
(1H 300 MHz; 13C 75 MHz). The chemical shifts of the
remaining protons of the deuterated solvents served as
internal standard: d (1H, DMSO-d6=2.50, CDCl3=7.24,
13C, DMSO-d6=39.7, CDCl3=77.0). Unless otherwise
noted, DMSO was used as solvent. Infra-red spectra
were recorded on a Perkin-Elmer FT-IRPC 16 spectro-
photometer. All compounds were checked for purity by
TLC on 0.2 mm aluminium sheets with silica gel 60 F254
(Merck). The following eluents were used: dichloro-
methane:acetone (9:1) is eluent A, dichloromethane/
methanol (9:1) is eluent B, toluene/ethyl acetate (1:1) is
eluent C, toluene/ethanol/acetic acid (5:4:1) is eluent D
and toluene/ethylacetate/formic acid (10:9:1) is eluent E.
Melting points were taken on a Buchi 535 melting point
apparatus and are uncorrected. Thermospray (TSP,
positive ionization), electron and chemical ionization
mass spectra were recorded on a MS Engine HP 5989A
mass spectrometer (Hewlett Packard). The GC mass
spectrum was performed on HP5890/II-5971 mass
spectrometer (Hewlett Packard). Elemental analyses,
NMRspectra, and mass spectra were performed by the
Institute of Organic Chemistry, University of Leipzig
and University of Tubingen, respectively.
2-(1-Hydroxymethyl-2,6-dioxo-piperidine-3-yl)-1,3-dihydro-
2H-isoindole-1,3-dione (2).
A suspension of 12.9 g
(50 mmol) of rac-1 in 100 mL of a 35% aq formaldehyde
solution was re¯uxed until dissolved and then allowed
to cool down. After 24 h, the precipitate was collected
by ®ltration and washed with 3% aq formaldehyde
solution and dried. Yield: 10.1 g (70%); mp 165 ꢀC; IR
1
(KBR) n 3200, 2970, 1783, 1680; H NMR d 2.16±2.67
(m, 2H, H-40), 2.87±3.11 (m, 2H, H-50), 5.08 (d,
Jab=7.2 Hz, 2H, NCH2OH), 5.52 (m, 1H, CHCH2),
6.17 (t, Jab=7.2 Hz, Jbc=7.2 Hz, 1H, OH), 7.92 (s, 4H,
Harom).
General procedure for N-tert-butyloxycarbonyl-protected
ester 3±6. A solution of 4.32 g (15 mmol) 2, 15 mmol of
the appropriate N-Boc-protected amino acid, 3.09 g
(15 mmol) of dicyclohexylcarbodiimide and 0.22 g of
(1.5 mmol) 4-pyrrolidinopyridine in 75 mL of abs di-
chloromethane were reacted at rt for 24 h. Precipitated
dicylcohexylurea was ®ltered o€; the ®ltrate was washed
3Â with 75 mL of 5% acetic acid and 3Â with 75 mL of
water. The organic layer was dried over MgSO4 and
concentrated in vacuo. The residue was recrystallized
two times from ethanol.
N-tert-butyloxycarbonyl-2-aminoacetic acid-[3-(1,3-dihy-
dro-1,3-dioxo-2H-isoindole-2-yl)-2,6-dioxo-piperidine-1-
yl-methyl]ester (3). N-Boc-glycine (2.63 g) was used.
Yield: 3.88 g (58%); mp 108±111 ꢀC; DC (eluent A) Rf -
value=0.51; IR(KBr) n 3432, 3312, 2984, 2944, 1758,
Analytical methods
1
HPLC analyses were performed on a Sykam S 1000
pump, a 20 mL injection loop, a Linear UVis 205 detector
using Axxiom 747 software for the registration and data
evaluation. Chromatographic separation was achieved iso-
cratically at ambient temperature in Nucleosil 120-5, C18,
250Â4.6 mm I.D. with a pore diameter of 10 nm
(Macherey-Nagel, Duren, Germany) reversed-phase
columns.
1724, 1706, 1680, 1402, 1170, 722; H NMR d 1.37 (s,
9H, C(CH3)3), 2.11±2.15 (m, 1H, H-40), 2.56±2.71 (m,
1H, H-40), 2.81±2.88 (m, 1H, H-50), 3.01±3.15 (m, 1H,
H-50), 3.67 (d, J
=6.2 Hz, 2H, NHCH2CO2),
0
ꢂCH2ÀNH†
5.35 (dd, J3 aÀ4 e=5.3 Hz, J3 aÀ4 a=13.1 Hz, 1H, H-30),
5.66, 5.71 (AB, JAB=9.6 Hz, 2H, NCH2O), 7.23 (t,
0
0
0
J
=6.1 Hz, 1H, NH), 7.88±7.97 (m, 4H, Harom.);
ꢂNHÀCH2†
13C NMR d 20.80 (C-40), 28.14 (C(CH3)3), 31.07 (C-50),
41.57 (NHCH2COO), 49.47 (C-30), 63.40 (NCH2O),
78.33 (C(CH3)3), 123.52 (C-4, C-7), 131.20 (C-3a, C-7a),
134.98 (C-5, C-6), 155.78 (NHCOO), 167.04, 169.19,
169.48, 170.9 (C-1, C-3, C-20, C-60, CH2COO); CIMS
m/z 446 (MH+). Anal. calcd for (C21H23N3O8): C,
56.63; H, 5.20; N, 9.43; found: C, 56.85; H, 5.19; N,
9.67%.
The mobile phase consisted of two volumes of acetonitrile
and eight volumes of a 50mM sodium dihydrogenpho-
sphate bu€er. The bu€er was adjusted to pH 3.5 with
concentrated phosphoric acid and was degassed with
helium before and during use. Injection volume was
20 mL and ¯ow rate was 1.0 mL/min. The system was
equilibrated for approximately 30 min. Thalidomide, its
prodrugs and hydrolysis compounds were monitored at
300 and 265 nm. The calibration curves of the prodrugs
were linear over the range of 5±100 mg/mL. The method
was validated with the following parameters: speci®ty, lin-
earity, precision, limit of detection, limit of quanti®cation,
N-tert-butyloxycarbonyl-2-methylaminopropionic acid-[3-
(1,3-dihydro-1,3-dioxo-2H-isoindole-2-yl)-2,6-dioxo-piperi-
dine-1-yl-methyl]ester (4). N-Boc-l-N-methylalanine
4.32 g (15 mmol) was used. Crude 4 was chromato-
graphed on silica using dichloromethane/acetone (9:1).

S. Hess et al. / Bioorg. Med. Chem. 9 (2001) 1279±1291
1287
Yield: 4.44 g (63%); mp 123±125 ꢀC; DC (eluent A) Rf -
value=0.63; IR(KBR) n 2978, 1752, 1720, 1706, 1392,
720; 1H NMR d 1.29±1.38 (m, 12H, C(CH3)3, CHCH3),
2.12±2.17 (m, 1H, H-40), 2.58±2.64 (m, 1H, H-40), 2.73
(s, 3H, NCH3), 2.82±2.88 (m, 1H, H-50), 3.02±3.08 (m,
1H, H-50), 4.37, 4.60 (m, 1H, CHCH3), 5.34 (dd,
General procedure for prodrug esters 7±10. A solution of
4 mmol of the appropriate N-Boc-protected ester 3, 4, 5
or 6, respectively, in 25% of tri¯uoroacetic acid in di-
chloromethane was reacted at rt for 1 h. The solvent was
removed in vacuo and the residue was coevaporated
several times with dichloromethane until dry. The residue
was dissolved in 160 mL of methanol and ion-exchange
chromatographed using Amberlite IR45, pretreated
with 1 N hydrochloric acid. The solvent was concentrated
in vacuo and the residue was suspended in boiling
ethanol. Water was added dropwise until the solution
was clear. After cooling, the precipitate was collected by
®ltration and dried.
J3 aÀ4 e=5.3 Hz, J3 aÀ4 a=13.0 Hz, 1H, H-30), 5.71 (s,
1H, NCH2O), 7.88±7.96 (m, 4H, Harom); 13C NMR
d (CDCl3) 14.72, 15.09 (CHCH3), 21.76 (C-40), 28.33
(C(CH3)3), 30.49, 30.92 (NCH3), 31.84 (C-50), 49.91
(C-30), 53.36, 54.79 (CHCH3), 63.87 (NCH2O), 80.40
(C(CH3)3), 123.80 (C-4, C-7), 131.71 (C-3a, C-7a),
134.54 (C-5, C-6), 155.22, 155.83 (NCOO), 167.19,
167.93, 169.85, 169.93, 171.21 (C-1, C-3, C-20, C-60,
CHCOO). Anal. calcd for (C23H27N3O8): C, 58.35;
H, 5.75; N, 8.87; found: C, 58.31; H, 5.76; N,
8.80%.
0
0
0
0
2-Aminoacetic acid-[3-(1,3-dihydro-1,3-dioxo-2H-isoindole-
(7).
2-yl)-2,6-dioxo-piperidine-1-yl-methyl]esterÂHCl
Compound 3 (1.78 g, 4 mmol) was used. Yield: 0.81 g
(53 %); mp 227±232 ꢀC; DC (eluent C) Rf -value=0.6;
IR(KBR) n 2988, 2952, 2908, 2854, 2794, 2700, 2610,
1766, 1732, 1726, 1722, 1718, 1708, 1698, 1390, 1224,
N-tert-butyloxycarbonyl-2-amino-3-methylbutyric acid-[3-
(1,3-dihydro-1,3-dioxo-2H-isoindole-2-yl)-2,6-dioxo-piperi-
720; H NMR d 2.13±2.19 (m, 1H, H-40), 2.54±2.69 (m,
1
dine-1-yl-methyl]ester
(5).
N-Boc-l-valine
3.26 g
(15 mmol) was used. Crude 5 was chromatographed on
silica using dichloromethane/acetone (9/1). Yield: 3.85 g
(53%); mp 82±85 ꢀC; DC (eluent A) Rf -value=0.68; IR
1H, H-40), 2.84±2.89 (m, 1H, H-50), 3.04±3.16 (m, 1H,
H-50), 3.80 (s, 2H, CH2NH3), 5.35 (dd, J3 aÀ4 e=5.4 Hz,
0
0
J3 aÀ4 a=13.1 Hz, 1H, H-30), 5.77, 5.82 (AB,
JAB=9.6 Hz, 2H, NCH2O), 7.89±7.97 (m, 4H, Harom),
8.48 (s, 3H, NH3); 13C NMR d (D2O): 21.10 (C-40),
31.22 (C-50), 40.12 (CH2NH3), 49.94 (C-30), 64.55
(NCH2O), 123.95 (C-4, C-7), 130.96 (C-3a, C-7a),
135.34 (C-5, C-6), 167.08, 169.02, 170.68, 173.45 (C-1,
C-3, C-20, C-60, COO); TSPMS m/z 346 (M+). Anal.
calcd for (C16H16ClN3O6): C, 50.34; H, 4.22; N, 11.01;
found: C, 50.16; H, 4.30; N, 11.24%.
0
0
1
(KBr) n 3394, 2972, 2934, 1752, 1720, 1392, 720; H
NMR d 0.82±0.86 (m, 6H, CH(CH3)2), 1.37 (s, 9H,
C(CH3)3), 1.96±2.00 (m, 1H, CH(CH3)2), 2.12±2.17 (m,
1H, H-40), 2.60±2.66 (m, 1H, H-40), 2.83±2.88 (m, 1H, H-
50), 3.03±3.09 (m, 1H, H-50), 3.83 (t, J(CHÀNH)=6.1 Hz,
1H, CHNH), 5.30±5.38 (m, 1H, H-30), 5.61 (dd, 1H,
NCH2(A)O), 5.75 (t, 1H, NCH2(B)O), 7.13 (t, 1H, NH),
7.88±7.96 (m, 4H, Harom); 13C NMR d 18.04, 18.15,
18.75, 18.84 (CH(CH3)2), 20.80 (C-40), 28.13 (C(CH3)3),
29.49, 29.54 (CH(CH3)2), 31.03 (C-50), 49.46 (C-30),
59.07, 59.12 (CHNH), 63.19, 63.27 (NCH2O), 78.24
(C(CH3)3), 123.45 (C-4, C-7), 131.17 (C-3a, C-7a),
134.93 (C-5, C-6), 155.68 (NHCOO), 166.97, 169.06,
170.86, 170.94, 171.07, 171.12 (C-1, C-3, C-20, C-60,
CHCOO). Anal. calcd for (C24H29N3O8): C, 59.13; H,
6.00; N, 8.62; found: C, 58.74; H, 6.05; N, 8.64%.
2-Methylamino(propionic acid)-[3-(1,3-dihydro-1,3-dioxo-
2H-isoindole-2-yl)-2,6-di-oxo-piperidine-1-yl-methyl]esterÂ
HCl (8). Compound 4 (1.89 g, 4 mmol) was used.
Deviating from the general procedure, the resulting
residue was dissolved in a small amount of methanol
and diethyl ether was added dropwise to precipitate 8,
which was collected by ®ltration.
2-N-tert-butyloxycarbonyl-(aminoacetylamino)acetic acid-
[3-(1,3-dihydro-1,3-dioxo-2H-isoindole-2-yl)-2,6-dioxopiper-
idine-1-yl-methyl]ester (6). N-Boc-glycylglycine (3.48 g)
was used. Yield: 5.50 g (73%); mp 178±181 ꢀC; IR(KBr)
n 3442, 3404, 2978, 2932, 1760, 1716, 1392, 1162, 996,
Yield: 1.05 g (64 %); mp 147±151 ꢀC; DC (eluent C) Rf -
value 0.65; IR(KBR) n 2966, 2704, 2434, 1756, 1718,
1
1394, 1106, 720; H NMR d 1.42 (d, J=6.9 Hz, 3H,
CHCH3), 2.14±2.17 (m, 1H, H-40), 2.54 (s, 3H,
NH2CH3), 2.57±2.66 (m, 1H, H-40), 2.84±2.89 (m, 1H,
H-50), 3.04±3.16 (m, 1H, H-50), 4.10 (q, J=7.1 Hz, 1H,
1
720; H NMR d 1.38 (s, 9H, C(CH3)3), 2.12±2.17 (m,
1H, H-40), 2.61±2.66 (m, 1H, H-40), 2.83±2.89 (m, 1H,
CHCH3), 5.36 (dd, J3 aÀ4 e=4.5 Hz, J3 aÀ4 a=13.0 Hz,
1H, H-30), 5.76 (d, JAB=9.6 Hz, 1H, NCH2(A)O), 5.89
(dd, JAB=9.6 Hz, 1H, NCH2(B)O), 7.89±7.96 (m, 4H,
Harom), 9.57 (bs, 2H, NH2CH3); 13C NMR d 13.93
(CHCH3), 20.83 (C-40), 30.41, 30.46 (CH3NH2), 31.00
(C-50), 49.41 (C-30), 54.96 (CHCH3), 63.81 (NCH2O),
123.49 (C-4, C-7), 131.13 (C-3a, C-7a), 134.99 (C-5,
C-6), 166.99, 168.34, 169.21, 170.97, 171.04 (C-1, C-3,
C-20, C-60, CH2COO); TSPMS m/z 374 (M+). Anal.
calcd for (C18H20ClN3O6): C, 52.75; H, 4.92; N, 10.25;
found: C, 52.42; H, 4.82; N, 10.17%.
0
0
0
0
H-50), 3.03±3.13 (m, 1H, H-50), 3.59 (d, J
5.8 Hz, 2H, CH2NHCOO), 3.86 (d, J
=
=6.0Hz,
ꢂCH2ÀNH†
ꢂCH2ÀNH†
0
0
2H, NHCH2COO), 5.35 (dd, J3 aÀ4 e=5.3 Hz,
J3 aÀ4 a=13.0 Hz, 1H, H-30), 5.68, 5.73 (AB,
0
0
JAB=9.6 Hz, 2H, NCH2O), 6.97 (t, J =5.8 Hz,
1H, NHCOO), 7.89±7.96 (m, 4H, Harom), 8.20 (t,
=5.9 Hz, 1H, NHCOCH2); 13C NMR d
ꢂNHÀCH2†
J
ꢂNHÀCH2†
20.80 (C-40), 28.15 (C(CH3)3), 31.03 (C-50), 40.17
(CH2COO), 42.99 (CH2NHCOO), 49.44 (C-30), 63.31
(NCH2O), 78.05 (C(CH3)3), 123.48 (C-4, C-7), 131.16
(C-3a, C-7a), 134.93 (C-5, C-6), 155.72 (NHCOO),
167.0, 168.82, 169.15, 169.97, 170.95 (C-1, C-3, C-20, C-60,
CH2CONH, CH2COO). Anal. calcd for (C23H26N4O9):
C, 54.98; H, 5.22; N, 11.15; found: C, 55.18; H, 5.19; N,
10.89%.
2-Amino-3-methylbutyric acid-[3-(1,3-dihydro-1,3-dioxo-
2H-isoindole-2-yl)-2,6-dioxo-piperidine-1-yl-methyl]esterÂ
HCl (9). Compound 5 (1.95 g, 4 mmol) was used.
Deviating from the general procedure, the resulting

1288
S. Hess et al. / Bioorg. Med. Chem. 9 (2001) 1279±1291
residue was dissolved in a small amount of ethanol and
approximately a 10-fold excess of diethyl ether was
added dropwise to precipitate 9, which was collected by
®ltration. Yield: 1.01 g (60%); mp 145±150 ꢀC; DC
(C-1, C-3, C-20, C-60, COOCH2, COOH); TSPMS m/z
406 (M++18). Anal. calcd for (C18H16N2O8): C, 55.67;
H, 4.15; N, 7.21; found: C, 55.73; H, 4.01; N, 7.33%.
(eluent C) Rf -value=0.73; IR(KBR)
n 2970, 2902,
2-[3-(4-Hydroxyphenyl)propionylamino]-3-methylbutyric
acid-[3-(1,3-dihydro-1,3-dioxo-2H-isoindole-2-yl)-2,6-di-
1752, 1718, 1392, 720; 1H NMR d 0.94 (t, 6H,
CH(CH3)2), 2.14±2.18 (m, 2H, H-40, CH(CH)2), 2.59±
2.63 (m, 1H, H-40), 2.84±2.90 (m, 1H, H-50), 3.07±3.16
(m, 1H, H-50), 3.84 (bs, 1H, CHNH3), 5.34±5.39 (m, 1H,
H-30), 5.69 (dd, JAB=9.7 Hz, 1H, NCH2(A)O), 5.88 (dd,
JAB=9.5 Hz, 1H, NCH2(B)O), 7.88±7.95 (m, 4H, Harom),
8.62 (bs, 3H, NH3); 13C NMR d 17.33, 17.44, 18.11,
18.25 (CH(CH3)2), 20.77 (C-40), 29.33 (CH(CH3)2), 31.01
(C-50), 49.43 (C-30), 57.06 (CHNH3), 63.79 (NCH2O),
123.48 (C-4, C-7), 131.13 (C-3a, C-7a), 134.98 (C-5, C-
6), 166.95, 168.01, 168.10, 169.12, 169.20, 170.92, 171.00
(C-1, C-3, C-20, C-60, COO); CIMS m/z 388 (M+). Anal.
calcd for (C19H22ClN3O6): C, 53.84; H, 5.23; N, 9.91;
found: C, 54.06; H, 5.22; N, 9.62%.
oxo-piperidine-1-yl-methyl]ester (13).
A solution of
0.366 g (0.75 mmol) 5 in 10 mL of 25% tri¯uoroacetic
acid in dichloromethane was reacted at rt for 1 h. The
solvent was removed in vacuo and the residue was co-
evaporated several times with dichloromethane until
dry. The residue was taken up in 20 mL of aq phosphate
bu€er (0.067 M, pH 7.4) in an ice bath and treated with
0.395 g (1.5 mmol) 12 (3-(4-hydroxyphenyl)-propionic
acid-2,5-dioxo-pyrrolidine-1-yl-ester) dissolved in diox-
ane. The pH of the mixture was brought up to 8±9 using
an aq solution of sodium hydrogen carbonate and
stirred at rt for 5 h. Subsequently, the pH of the mixture
was adjusted to pH 2 using 1N hydrochloric acid.
Dioxane was concentrated in vacuo; the remaining aq
solution was 3Â extracted with ethyl acetate. The com-
bined organic layers were once washed with water and
dried over Na2SO4. The solvent was removed in vacuo
and the residue was chromatographed on silica
(30 cmÂ2 cm) using eluent D to yield 0.218 g (54%) of a
white foam.
2-(Aminoacetylamino)acetic acid-[3-(1,3-dihydro-1,3-dioxo-
2H-isoindole-2-yl-2,6-dioxopiperidine-1-yl-methyl]esterÂ
HCl (10). A solution of 2.01 g (4 mmol) 6 was reacted
as described for 7. The compound was recrystallized from
methanol:diethyl ether (1:1) to yield an amorphous foam.
Yield 1.35 g (77 %); DC (eluent B): Rf -value=0.15; IR
(KBr) n 3394, 3226, 3060, 1752, 1718, 1702, 1394, 720;
1H NMR d 2.12±2.16 (m, 1H, H-40), 2.59±2.68 (m, 1H,
H-40), 2.83±2.88 (m, 1H, H-50), 3.03±3.17 (m, 1H, H-50),
DC (eluent D) Rf -value=0.37; IR(KBR) n 3388, 2966,
2932, 1750, 1720, 1654, 1614, 1516, 1392, 720; 1H NMR
d 0.80±0.82 (m, 6H, CH(CH3)2), 1.95±1.99 (m, 1H,
CH(CH3)2), 2.12±3.08 (m, 8H, 2ÂH-40, 2ÂH-50,
NHCOCH2CH2), 4.16 (t, 1H, CHNH), 5.32±5.37 (m,
1H, H-30), 5.63 (dd, JAB=9.7 Hz, 1H, NCH2(A)O), 5.77
3.61 (s, 2H, CH2NH3), 3.96 (d, J
NHCH2CO2),
= 5.8 Hz, 2H,
0
ꢂCH2ÀNH†
0
5.35
(dd,
J3 aÀ4 e=5.3 Hz,
0
0
J3 aÀ4 a=12.9 Hz, 1H, H-3), 5.68, 5.75 (AB,
JAB=9.6 Hz, 2H, NCH2O), 7.90±7.97 (m, 4H, Harom),
00
00 00
00
=
8.15 (s, 3H, NH3), 8.87 (t, J
=5.8 Hz, 1H, NH);
(dd, JAB=9.6 Hz, 1H, NCH2(B)O), 6.50 (d, J3 À2 ,5 À6
ꢂNHÀCH2†
13C NMR d (D2O) 21.10 (C-40), 31.24 (C-50), 40.56,
42.12 (NHCH2CO2, CH2NH3), 49.96 (C-30), 64.20
(NCH2O), 123.95 (C-4, C-7), 131.03 (C-3a, C-7a),
135.33 (C-5, C-6), 167.79, 169.07, 169.99, 170.69, 173.50
(C-1, C-3, C-20, C-60, CONH, COO); TSPMS m/z 403
(M+). Anal. calcd for (C18H19ClN4O7): C, 49.27; H,
4.36; N, 12.77; found: C, 48.93; H, 4.26; N, 12.77%.
8.3 Hz, 2H, H-300, H-500), 6.98 (d, J2 À3 ,6 À5 =8.3 Hz,
2H, H-200, H-600), 7.88±7.96 (m, 4H, Harom), 8.06±8.10
(m, 1H, NH), 9.11 (s, 1H, OH); 13C NMR d 17.96,
18.03, 18.74, 18.83 (C(CH3)2), 20.81 (C-40), 29.72, 29.77
(CH(CH3)2), 30.27 (CH2CONH), 30.91 (C-50), 36.86
(CH2-phenyl), 49.46 (C-30), 57.02 (CHNH), 63.12
(NCH2O), 114.96 (C-300, C-500), 123.47 (C-4, C-7),
128.96 (C-200, C-600), 131.17, 131.24 (C-3a, C-7a, C-100),
134.93 (C-5, C-6), 155.37 (C-400), 166.99, 169.07, 169.11,
170.71, 170.74, 170.89, 170.97, 172.04 (C-1, C-3, C-20,C-
60, CONH, COO); TSPMS m/z 553 (M++18), 536
(M++1). Anal. calcd for (C28H29N3O8): C, 62.79; H,
5.46; N, 7.85; found: C, 62.71; H, 5.61; N, 7.74%.
00
00 00
00
Succinic acid mono[3-(1,3-dihydro-1,3-dioxo-2H-isoin-
dole-2-yl)-2,6-dioxo-piperidine-1-yl-methyl]ester (11). A
solution of 2.88 g (10 mmol) 2, 2 g (20 mmol) succinic
anhydride, 2.8 mL (20 mmol) of triethylamine and 0.15 g
(1 mmol) 4-pyrrolidinopyridine in 50 mL of abs di-
chloromethane were stirred at rt for 24 h. The reaction
mixture was extracted 3Â with 50mL of 5% hydrochloric
acid and 3Âtimes with 50mL of water. The organic layer
was dried over MgSO4 and the solvent was evaporated in
vacuo. The remaining residue was crystallized from ethyl
acetate and recrystallized from ethanol. Yield 2.66 g
(68%); DC (eluent E) Rf -value=0.25; IR(KBr) n 2990,
2960, 2904, 2784, 2684, 2574, 1740, 1716, 1704, 1392,
1208, 1162, 1100, 720; 1H NMR d 2.14±2.17 (m, 1H, H-
40), 2.62±2.68 (m, 4H, CH2CH2COOH), 2.82±3.07 (m,
3H, 2ÂH-50, H-40), 5.04±5.10 (m, 1H, H-30), 5.83, 5.88
(AB, JAB=9.5Hz, 2H, NCH2O), 7.75±7.78 (m, 2H,
Harom), 7.87±7.89 (m, 2H, Harom); 13C NMR d 20.81 (C-
40), 31.03 (C-50), 28.40 (CH2CH2COOH), 49.43 (C-30),
63.11 (NCH2O), 123.47 (C-4, C-7), 131.16 (C-3a, C-7a),
134.93 (C-5, C-6), 167.01, 169.15, 170.96, 171.23, 173.12
2-(1-Chloromethyl-2,6-dioxo-piperidine-3-yl)-1,3-dihydro-
2H-isoindole-1,3-dione (14). To a solution of 11.52 g
(40 mmol) of 2 in 100 mL of dimethylformamide, 8 mL
of thionyl chloride were added slowly under ice cooling.
After stirring for 1 h at 0 ꢀC, the mixture was poured
onto ice water. The precipitate was collected by ®ltration
and recrystallized from ethanol. Yield: 8.52 g (69 %);
mp 204±206 ꢀC; DC (eluent A) Rf -value=0.83; IR
(KBR) n 3074, 2908, 1786, 1772, 1716, 1702, 1392, 1156,
720; H NMR d 2.11±2.16 (m, 1H, H-40), 2.54±2.66 (m,
1
1H, H-40), 2.84±2.90 (m, 1H, H-50), 3.04±3.16 (m, 1H,
H-50), 5.38 (dd, J3 aÀ4 e=5.3 Hz, J3 aÀ4 a=13.0 Hz, 1H,
H-30), 5.52 (s, 2H, CH2Cl), 7.89±7.96 (m, 4H, Harom);
13C NMR d (CDCl3): 21.69 (C-40), 31.85 (C-50), 47.21
(CH2Cl), 49.94 (C-30), 123.88 (C-4, C-7), 131.70 (C-3a,
0
0
0
0

S. Hess et al. / Bioorg. Med. Chem. 9 (2001) 1279±1291
1289
00
00
00
00
00
C-7a), 134.59 (C-5, C-6), 167.23, 167.40, 169.40 (C-1,
C-3, C-20, C-60).
(d, J4 À5 =7.9 Hz, 1H, H-4 ), 9.22 (d, J6 À5 =6.1 Hz,
1Hz, 1H, H-600), 9.44 (s, 1H, H-200); 13C NMR(D O) d
2
20.98 (C-40), 30.98 (C-50), 49.81 (C-30), 62.62 (NCH2N),
124.00 (C-4, C-7), 128.64 (C-500), 131.05 (C-3a, C-7a),
134.06 (C-300), 135.40 (C-5, C-6), 145.21, 146.04, 147.45,
(C-400, C-200, C-600), 165.50, 169.08, 171.49, 173.61
(CONH2, C-1, C-3, C-20, C-60). Anal. calcd for
(C20H17ClN4O5): C, 56.02; H, 4.00; N, 13.07; found: C,
55.85; H, 4.04; N, 13.42%.
2-(1-Iodomethyl-2,6-dioxo-piperidine-3-yl)-1,3-dihydro-2H-
isoindole-1,3-dione (15). To a solution of 6.13 g (20 mmol)
14 in 200 mL of abs acetone, 16.60 g (100 mmol) potas-
sium iodide were added under argon atmosphere and
re¯uxed for 6 h. The solvent was evaporated in vacuo. The
residue was dissolved in 100 mL of a solution of 10% (m/
V) sodium thiosulfate and 3Â extracted with 50 mL of
dichloromethane. Then, the combined organic layers
were once extracted with 100 mL of a solution of 10%
(m/V) sodium thiosulfate and dried over sodium sulfate.
The solvent was evaporated in vacuo and the residue
was crystallized from acetonitrile. Yield 5.58 g (70%);
mp 207±209 ꢀC; DC (eluent A): Rf -value=0.86; IR
(KBr) n 3082, 2956, 2926, 2902, 1786, 1770, 1720, 1694,
1-[3-(1,3-Dihydro-1,3-dioxo-2H-isoindole-2-yl)-2,6-dioxo-
piperidine-1-yl-methyl]-1,4-dihydropyridine-3-carboxylic
acid amide (18). A solution of 0.43 g (1 mmol) 17 in
70 mL of water was degassed with nitrogen and cooled
to 0±5 ꢀC. A mixture of 0.5 g sodium hydrogencarbonate
and 0.7 g sodium dithionite were added and stirred for
1 h at 0±5 ꢀC. Then, the mixture was extracted 5±6Â
with 50 mL of degassed dichloromethane until the aq
phase was decolorized. The combined organic layers
were once washed with water and dried over Na2SO4.
The solvent was removed in vacuo and the yellow residue
was recrystallized from ethanol. Yield 0.126 (32%); mp
207±210 ꢀC; DC (eluent B) Rf -value=0.60; IR(KBr) n
3474, 3142, 1784, 1718, 1694, 1676, 1390, 716; 1H NMR
d 2.08±2.12 (m, 1H, H-40), 2.56±2.61 (m, 1H, H-40),
2.79±3.07 (m, 4H, 2ÂH-50, 2ÂH-400), 4.59±4.61 (m, 1H,
H-500) 4.94, 5.02 (AB, JAB=13.9 Hz, 2H, NHCH2N),
1
1390, 1138, 718; H NMR d (CDCl3) 2.12±2.14 (m, 1H,
H-40), 2.76±2.78 (m, 1H, H-40), 2.79±2.83 (m, 1H, H-50),
3.02±3.07 (m, 1H, H-50), 5.02±5.06 (m, 1H, H-30), 5.52
(s, 2H, CH2I), 7.77±7.79 (m, 2H, Harom), 7.88±7.91 (m,
2H, Harom); 13C NMR d (CDCl3) 5.22 (CH2I), 21.57
(C-40), 32.08 (C-50), 50.17 (C-30), 123.89 (C-4, C-7),
131.71 (C-3a, C-7a), 134.59 (C-5, C-6), 167.09, 167.21,
169.10 (C-1, C-3, C-20, C-60); TSPMS m/z 416
(M++18),
399
(MH+).
Anal.
calcd
for
(C14H11IN2O4): C, 42.23; H, 2.78; N, 7.04; found: C,
42.56; H, 2.84; N, 6.70%.
5.34 (dd, J3 aÀ4 e=5.0 Hz, J3 aÀ4 a=12.8 Hz, 1H, H-30),
0
0
0
0
00
00
00
5.97 (d, J6 À5 =7.7 Hz, 1H, H-6 ), 6.58 (s, 2H, CONH2),
7.02 (s, 1H, H-200), 7.89±7.96 (m, 4H, Harom); 13C NMR d
20.97 (C-40), 22.10 (C-400), 31.15 (C-50), 49.49 (C-30),
56.43 (NCH2N), 101.35 (C-300), 101.81 (C-500), 123.46
(C-4, C-7), 128.98 (C-600), 131.17 (C-3a, C-7a), 134.91
(C-5, C-6), 137.63 (C-200). 167.08, 168.77, 170.18, 172.18
(C-1, C-3, CONH2, C-20, C-60); TSPMS m/z 395 (M+).
Anal. calcd for (C20H18N4O5): C, 60.91; H, 4.60; N,
14.21; found: C, 61.19; H, 4.67; N, 14.38%.
2-(1-Azidomethyl-2,6-dioxo-piperidine-3-yl]-1,3-dihydro-2H-
isoindole-1,3-dione (16). A solution of 1.59 g (4 mmol) 15
in 25 mL of abs acetone was reacted with 0.52 g
(8mmol) sodium azide at rt for 24h. The reaction mixture
was ®ltered and the ®ltrate was poured into ice water.
After ®ltering o€ the precipitate, it was recrystallized
from ethanol. Yield 0.85 g (68%); mp 148±151 ꢀC; DC
(eluent A) Rf -value=0.79; IR(KBr) n 3056, 2966, 2920,
2126, 2100, 1786, 1772, 1716, 1692, 1392, 718; 1H NMR
d 2.10±2.18 (m, 1H, H-40), 2.55±2.66 (m, 1H, H-40), 2.81±
2.89 (m, 1H, H-50), 3.01±3.13 (m, 1H, H-50), 5.07, 5.12
3-Carboxy-1-[3-(1,3-dihydro-1,3-dioxo-2H-isoindoline-2-
yl)-2,6-dioxo-piperidine-1-ylmetyl]pyridinium iodide (19).
A solution of 1.59 g (4 mmol) 15 in 20 mL of abs aceto-
nitrile was re¯uxed with 0.49 g (4 mmol) nicotinic acid
for 2.5 h. After cooling the precipitate was collected by
®ltration and washed with ethanol. Yield 1.11 (53%);
mp 246±250 ꢀC; DC (eluent D) Rf -value=0.13; IR
0
0
(AB, JAB=12.1, 2H, CH2N3), 5.36 (dd, J3 aÀ4 e=5.3Hz,
0
0
J3 aÀ4 a=13.0 Hz, 1H, H-3), 7.88±7.96 (m, 4H, Harom);
13C NMR d 20.95 (C-40), 30.97 (C-50), 49.38 (C-30), 54.24
(CH2N3), 123.47 (C-4, C-7), 131.15 (C-3a, C-7a), 134.92
(C-5, C-6), 167.01, 169.69, 171.63 (C-1, C-3, C-20, C-60);
TSPMS m/z 331 (M++18). Anal. calcd for
(C14H11N5O4): C, 53.68; H, 3.54; N, 22.36; found: C,
53.77; H, 3.49; N, 22.19%.
1
(KBr) n 3058, 2904, 1784, 1772, 1712, 1392, 720; H
NMR d 2.15±2.18 (m, 1H, H-40), 2.68±2.72 (m, 1H, H-
40), 2.95±3.01 (m, 2H, 2ÂH-50), 5.32 (dd,
J3 aÀ4 e=4.8 Hz, J3 aÀ4 a=12.9 Hz, 1H, H-30), 6.53 (s,
0
0
0
0
3-Carbamoyl-1-[3-(1,3-dihydro-1,3-dioxo-2H-isoindoline-
2-yl)-2,6-dioxo-piperidine-1-yl-methyl]pyridinium chloride
(17). A solution of 1.23 g (4 mmol) 14 in 30 mL of abs
acetonitrile was re¯uxed with 0.98 g (8 mmol) nicotinamide
for 40 h. After cooling to 70 ꢀC, the precipitate was col-
lected by ®ltration.
2H, NCH2N), 7.93 (s, 4H, Harom), 8.28 (t, 1 H, H-500),
9.02 (d, J4 À5 =7.90H0 z, 1H, H-400), 9.11 (d,
00
00
00
J6 À5 =5.8 Hz, 1H, H-6 ), 9.40 (s, 1H, H-2 ); 13C NMR
00
00
d
20.69 (C-40), 31.33 (C-50), 49.50 (C-30), 61.77
(NCH2N), 123.53 (C-4, C-7), 127.80 (C-500), 131.09 (C-3a,
C-7a), 132.20 (C-300), 135.05 (C-5, C-6), 146.67 (C-200, C-
400), 147.59 (C-600), 162.84 (COOH), 166.95, 170.27,
172.04 (C-1, C-3, C-20, C-60); EIMS m/z 393 (M+).
Anal. calcd for (C20H16IN3O6): C, 46.08; H, 3.09; N,
8.06; found: C, 46.37; H, 3.14; N, 7.71%.
Yield 1.35 (79%); mp 277±280 ꢀC; DC (eluent D) Rf -
value=0.05; IR(KBr) n 3260, 3084, 1776, 1766, 1746,
1
1712, 1708, 1698, 1682, 1394, 1192, 726; H NMR d
2.29±2.35 (m, 1H, H-40), 2.63±2.73 (m, 1H, H-40), 3.10±3.15
(m, 2H, 2ÂH-50), 5.42 (dd, J3 aÀ4 e=5.5 Hz, J3 aÀ4 a
[3-(1,3-Dihydro-1,3-dioxo-2H-isoindole-2-yl)-2,6-dioxo-
piperidine-1-yl]-acetic acid tert-butyl ester (20). A solution
of 0.176 g (4.4 mmol) sodium hydride (60%) in 20 mL of
0
0
0
0
=
12.7Hz, 1H, H-30), 6.63, 6.69 (AB, JAB=13.3 Hz, 2H,
NCH2N), 7.91 (s, 4H, Harom), 8.30 (t, 1H, H-500), 9.04

1290
S. Hess et al. / Bioorg. Med. Chem. 9 (2001) 1279±1291
abs dimethylformamide was reacted with 1.03 g
(4 mmol) thalidomide while cooling. After the formation
of gas has ceased, 1.21 g (5 mmol) tert-butyl iodoacetate
were added slowly and stirred for further 24 h. The
mixture was poured into ice water; the pH was adjusted
to 2.0 with 5% hydrochloric acid. The precipitate was
®ltered o€ and recrystallized from ethanol. Yield 0.92
(62%); mp 143±144 ꢀC; DC (eluent A) Rf -value=0.83;
IR(KBr) n 2978, 1786, 1772, 1722, 1688, 1392, 1170,
134.89 (C-5, C-6), 167.08, 169.80, 171.68 (C-1, C-3,
C-20, C-60); GCMS m/z 302 (M+). Anal. calcd for
(C15H14N2O5): C, 59.60; H, 4.67; N, 9.27; found: C,
59.54; H, 4.72; N, 9.38%.
Solubility determination
The solubility of selected compounds was determined
using UV. In general, a saturated solution (200 mg/mL)
of the compounds in water was prepared and allowed to
reach equilibrium while stirring at rt for 1 h. Compound
11 was dissolved in bu€er, pH 7.4. After ®ltration
through cotton, the saturated stock solutions were dilu-
ted in order to enable UV determination. The solubility
of the compounds was then calculated using calibration
curves, previously carried out. Due to the good water
solubility of the compounds 8, 9 and 10, it was possible
to dissolve them in concentrations as high as 300 mg/mL.
No attempt was made to prepare saturated solutions in
these cases.
1
720; H NMR d 1.40 (s, 9H, C(CH3)3), 2.12±2.17 (m,
1H, H-40), 2.65±2.71 (m, 1H, H-40), 2.83±2.89 (m, 1H,
H-50), 3.04±3.10 (m, 1H, H-50) 4.26, 4.33 (AB,
0
0
JAB=16.7 Hz, 2H, CH2COOH), 5.29 (dd, J3 aÀ4 e=5.0 Hz,
J3 aÀ4 a=13.1 Hz, 1H, H-30), 7.88±7.95 (m, 4H, Harom);
13C NMR d 20.99 (C-40), 27.51(C(CH3)3), 30.95 (C-50),
41.70 (CH2COOH), 49.55 (C-30), 81.42 (C(CH3)3),
123.43 (C-4, C-7), 131.17 (C-3a, C-7a), 134.90 (C-5, C-
6), 166.32, 167.02, 169.07, 171.15 (C-1, C-3, C-20, C-60,
COOH); TSPMS m/z 390 (M++18). Anal. calcd for
(C19H20N2O6): C, 61.28; H, 5.41; N, 7.52; found: C,
61.56; H, 5.55; N, 7.84%.
0
0
Determination of the pH of the aqueous solution
[3-(1,3-Dihydro-1,3-dioxo-2H-isoindole-2-yl)-2,6-dioxo-
piperidine-1-yl]-acetic acid (21). A solution of 2.98 g
(8 mmol) 20 in a mixture of 40 mL of 25% tri¯uoro-
acetic acid in dichloromethane was reacted for 1 h at rt.
The solvent was evaporated in vacuo. The remaining
residue was coevaporated 3±4Â until dry. The amor-
phous powder was recrystallized from water. Yield
2.27 g (85%); mp 111±113 ꢀC; DC (eluent E) Rf -
value=0.24; IR(KBr) n 3004, 2964, 2632, 2548, 1772,
Five mg/mL of the compounds were dissolved in water
and the pH was measured at rt using a Metrohm 691
pH Meter (Metrohm, Herisau, Switzerland).
Lipophilicity parameters
In this method, the capacity factor (k0) of the solute was
taken as a measure of the relative lipophilicity and was
calculated as k0 ˆ ꢂtR À t0†=t0, where tR is the retention
time of the solute and t0 is the elution time of the
solvent. For analysis, a reversed phase Nucleosil
column equipped with a guard column was used. The k0
values were determined using acetonitrile/phosphate
bu€er (50 mM, pH 3.5) (20/80). The ¯ow rate was
1.0 mL/min, the UV detector was set at 300 and 265 nm,
respectively.
1
1720, 1690, 1398, 1172, 722; H NMR d 2.12±2.16 (m,
1H, H-40), 2.66±2.71 (m, 1H, H-40), 2.82±2.88 (m, 1H,
H-50), 3.04±3.10 (m, 1H, H-50) 4.29, 4.34 (AB,
0
0
=
JAB=16.9 Hz, 2H, CH2COOH), 5.29 (dd, J3 aÀ4 e
5.0 Hz, J3 aÀ4 a=13.0 Hz, 1H, H-30), 7.91 (m, 4H,
Harom); 13C NMR d 21.00 (C-40), 30.94 (C-50), 41.04
(CH2COOH), 49.53 (C-30), 123.42 (C-4, C-7), 131.16 (C-
3a, C-7a), 134.90 (C-5, C-6), 167.04, 168.72, 169.12,
171.17 (C-1, C-3, C-20, C-60, COOH); TSPMS m/z 334
(M++18), 317 (MH+). Anal. calcd for (C15H12N2O6Â
H2O): C, 53.90; H, 4.22; N, 8.38; found: C, 54.04; H,
4.28; N, 8.56%.
0
0
Hydrolysis kinetics
The hydrolysis of selected compounds was determined
using HPLC. Physiological phosphate bu€er pH 7.4 at
37 ꢀC was used. Aqueous solutions of the prodrugs were
added to the bu€ered solution to yield an initial con-
centration of 100 mg/mL. At selected time intervals,
samples were withdrawn and analyzed by HPLC.
2-(1-Methoxymethyl-2,6-dioxo-piperidine-3-yl)-1,3-dihydro-
1,3-dioxo-2H-isoindole-1,3-dione (22). A suspension of
1.03 g (4 mmol) 1 and 16 g phosphorus pentoxide in
30 mL of dichloromethane were reacted with 16 mL of
formaldehyde 1,1-dimethoxymethane at rt for 72 h.
Under cooling 20 mL of water was slowly added. The
mixture was extracted 3Â with 20 mL of diethyl ether.
The combined organic layer was dried over Na2SO4 and
the solvent was removed in vacuo. The residue was
crystallized from ethanol. Yield 0.80 (66%); mp 153±
155 ꢀC; DC (eluent A) Rf -value=0.64; IR(KBr) n 2958,
2830, 1784, 1772, 1720, 1686, 1394, 1114, 720; 1H NMR
d 2.08±2.14 (m, 1H, H-40), 2.54±2.67 (m, 1H, H-40),
2.79±2.84 (m, 1H, H-50), 2.99±3.11 (m, 1H, H-50), 3.25
(s, 3H, CH3), 5.04, 5.09 (AB, JAB=9.9 Hz, 2H,
Assays for inhibition of TNF-ꢀ synthesis by human
peripheral blood mononuclear cells (PBMCs)
PBMCs isolated from three healthy human blood
donors were cultured in RPMI 1640 (containing 10%
fetal calf serum, 100 mmol b-mercaptoethanol, 50 mg/mL
penicillin, 2 mM glutamine and 50 mg/mL streptomycin)
at a density of 1±3Â106 cells/mL at 37 ꢀC, 5% CO2 in
presence of test compounds in 24-well plates (Sigma,
Deisenhofen, Germany). Test compounds were routinely
dissolved in DMSO (dimethylsulfoxide, Merck,
Darmstadt, Germany) and applied 1 h before release of
TNF-a was stimulated by addition of lipopolysaccharide
from Escherichia coli serotype 0127: B8 (Sigma) at a
0
0
0
0
NCH2O), 5.32 (dd, J3 aÀ4 e=5.4 Hz, J3 aÀ4 a=13.1 Hz,
1H, H-30), 7.88±7.95 (m, 4H, Harom); 13C NMR d 20.98
(C-40), 31.15 (C-50), 49.52 (C-30), 56.53 (CH3O), 70.17
(NCH2O), 123.44 (C-4, C-7), 131.18 (C-3a, C-7a),

S. Hess et al. / Bioorg. Med. Chem. 9 (2001) 1279±1291
1291
®nal concentration of 0.1 mg/mL. The ®nal concentra-
References
tion of DMSO was adjusted to 0.1%, the ®nal com-
pound concentration was 50 mg/mL, corresponding to
93±194 mM, for each culture sample. After incubation
for 20 h at 37 ꢀC and 5% CO2, the TNF-a concentration
of all culture supernatants was determined using a
commercial TNF-a-ELISA (Boehringer Mannheim,
Mannheim, Germany). The percent inhibition of release
of TNF-a was calculated relative to cultures treated
with DMSO alone. The IC50 values were calculated by
linear regression.
1. Lenz, W. Dt. Med. Wochenschr. 1961, 86, 2555.
2. Lenz, W. Lancet 1962, 1, 271.
3. Lenz, W. Teratology 1988, 38, 203.
4. Ehninger, G.; Eger, K.; Stuhler, A.; Schuler, U. Bone
Marrow Transplant 1993, 12(Suppl.), 26.
5. Hashimoto, Y. Curr. Med. Chem. 1998, 5, 163.
6. Calabrese, L.; Fleischer, A. B., Jr. Am. J. Med. 2000, 108, 487.
7. Tseng, S.; Pak, G.; Washenik, K.; Pomeranz, M. K. J. Am.
Acad. Dermatol. 1996, 35, 969.
8. Spilker, B.; FritzSimmons, S.; Horan, M. Drug Develop.
Res. 1999, 48, 139.
9. Koch, H. P.; Steinacker, C. Arch. Pharm. Pharm. Med.
Chem. 1988, 321, 371.
Assay for inhibition of IL-2 synthesis in mice
10. Krenn, M.; Gamcsik, M. P.; Vogelsang, G. B.; Colvin,
O. M.; Leong, K. W. J. Pharm. Sci. 1992, 81, 685.
11. Sauer, H.; Guenther, J.; Hescheler, J.; Wartenberg, M.
Am. J. Pathol. 2000, 156, 151.
12. Thiel, R.; Kastner, U.; Neubert, R. Life Science 2000, 66,
133.
13. Thiele, A.; Thormann, M.; Hofmann, H.-J.; Naumann,
W. W.; Eger, K.; Hauschildt, S. Life Science 2000, 67, 457.
14. Stephens, T. D.; Fillmore, B. J. Teratology 2000, 61, 189.
15. Wermuth, C. G., Gaignault, J.-C., Marchandeau, C. In
The Practice of Medicinal Chemistry; Wermuth, C. G., ed.;
Academic: London, 1996, 671±696.
16. Williams, A.; Ibrahim, I. T. Chem. Rev. 1981, 81, 589.
17. Pop, E.; Liu, Z. Z.; Brewster, E. M.; Barenholz, Y.; Kor-
blyov, V.; Mechoulam, R.; Nadler, V.; Biegon, A. Pharm. Res.
1996, 13, 62.
18. Rautio, J.; Nevalainen, T.; Taipale, H.; Vepsalainen, J.;
Gynther, J.; Laine, K.; Jarvinen, T. J. Med. Chem. 2000, 43, 1489.
19. Hassner, A.; Alexanian, V. Tetrahedron Lett. 1978, 46, 4475.
20. Cho, M. J.; Haynes, L. C. J. Pharm. Sci. 1985, 74, 883.
21. Sieber, P.; Iselin, B. Helv. Chim. Acta 1968, 51, 522.
22. Ragnarsson, U.; Karlsson, S.; Lindeberg, G. Acta Chem.
Scand. 1970, 24, 2821.
23. Rudinger, J.; Ruegg, U. Biochem. J. 1973, 133, 538.
24. Bolton, A. E.; Hunter, W. M. Biochem. J. 1973, 133, 529.
25. Bodor, N.; Farrag, H. H. J. Med. Chem. 1983, 26, 313.
26. Pop, E.; Shek, E.; Murakami, T.; Bodor, N. S. J. Pharm.
Sci. 1989, 78, 609.
27. Phelan, M. J.; Bodor, N. Pharm. Res. 1989, 6, 667.
28. Gal, J. J. Pharm. Sci. 1979, 68, 1562.
29. Broto, P.; Moreau, G.; Vandyke, C. Eur. J. Med. Chem. Ð
Chim Ther. 1984, 19, 71.
30. ChemDraw Ultra Version 5.0; Cambridge Soft Corpora-
tion: Cambridge, MA, USA, 1999.
31. Buchwald, P.; Bodor, N. Curr. Med. Chem. 1998, 5, 353.
32. Trapani, G.; Latrofa, A.; Franco, A.; Lopedota, A.;
Maciocco, E.; Liso, G. Int. J. Pharm. 1998, 175, 195.
33. Beauchamp, L. M.; Krenitsky, T. A. Drugs Fut. 1993, 18, 619.
34. Payne, J. W. Drugs Exp. Clin. Res. 1986, 12, 585.
35. Ames, B. N.; Ames, G. G.-L.; Young, J. D.; Tsuchiya, D.;
Lecocq, J. Proc. Natl. Acad. Sci. U.S.A. 1973, 70, 456.
36. Schumacher, H.; Smith, R. L.; Williams, R. T. Br. J.
Pharmacol. 1965, 25, 324.
Male white mice (20±30 g body weight) were used. The
animals were kept under standardized housing condi-
tions. Test compounds were applied ip in a concentra-
tion of 400 mg/kg using carboxymethylcellulose as
vehicle and as control. After 30 min, staphylococcal-
superantigen was injected iv in a concentration of
200 mg/animal. After 2 h, the inhibition of the IL-2 con-
centration was determined collecting blood samples by
puncturing the heart. The IL-2 concentration of all
blood samples was determined using a commercial
ELISA (Boehringer Mannheim). The percent inhibition
of IL-2 synthesis was calculated relative to the blood
samples treated with carboxymethylcellulose alone. The
ED40 of compounds 7 and 8 was additionally deter-
mined after peroral application.
E€ect on local Shwartzman reaction in mice
The assay was carried out as previously described.39
Brie¯y, lipopolysacharide (LPS) from Salmonella typhi-
murium was injected intradermally at a dose of 100 mg in
a volume of 30 mL on one side of the shaved dorsum. On
the other side, the same volume of a sterile salt solution
was applied as a control. After 24 h, the Evans Blue
solution was injected in the ear vene. Immediately after
this injection, TNF-a (133 ng/0.4 mL) was injected sub-
cutaneously on the site of the intradermal injection.
Thalidomide derivatives were administered ip three
times in a concentration of 400 mg/kg, 10 min before
and 8 h after LPS injection as well as 10 min before the
TNF-a challenge injection. The dorsal skin was excised.
Evans Blue was extracted using formamide and its
extinction was determined at 623 nm.
37. Teubert, U.; Zwingenberger, K.; Wnendt, S.; Eger, K.
Arch. Pharm. Pharm. Med. Chem. 1998, 331, 7.
38. Zimmer, O.; Winter, W.; Wnendt, S.; Zwingenberger, K.;
Eger, K.; Teubert U., EP 856,513, August 15 1998.
39. Geist, C.; Woehrmann, T.; Schneider, J.; Zwingenberger,
K. FEMS Immunol. Med. Microbiol. 1995, 12, 165.
40. Muller, G. W.; Shire, M. G.; Wong, L. M.; Corral, L. G.;
Patterson, R. T.; Chen, Y.; Stirling, D. I. Bioorg. Med. Chem.
Lett. 1998, 8, 2669.
Acknowledgements
K.E. is grateful for support from the Fonds der Chem-
ischen Industrie and the Grunenthal GmbH, Aachen,
Germany. Our thanks are also expressed to S. Hees, Dr. J.
Schneider, J. Korioth, Dr. C. Geist, S. Krahe, all Gru-
nenthal, Stolberg, Germany for the determination of
some biological data and technical assistance.
41. Schumacher, H.; Smith, R. L.; Williams, R. T. Br. J.
Pharmacol. 1965, 25, 338.