Thalidomide metabolites and analogues. 3. Synthesis and antiangiogenic activity of the teratogenic and TNFα-modulatory thalidomide analogue 2-(2,6-dioxopiperidine-3-yl)phthalimidine

Article abstract of DOI:10.1021/jm020079d

Versatile synthesis of the teratogenic, TNFα-modulatory, and antiangiogenic thalidomide analogue 2-(2,6-dioxopiperidine-3-yl)phthalimidine (1) and its direct antiangiogenic properties are described. With thalidomide or thalidomide derivatives as precursors, the synthesis involved either carbonyl reduction/thiation-desulfurization or carbonyl reduction/acyliminium ion reduction protocols. Compared to earlier studies with thalidomide, which was only active with microsomal treatment, 1 exhibited marginal inhibitory activity in the rat aortic ring assay, thereby demonstrating the requirement for metabolic activation.

Full text of DOI:10.1021/jm020079d

J . Med. Chem. 2003, 46, 3793-3799
3793
Th a lid om id e Meta bolites a n d An a logu es. 3. Syn th esis a n d An tia n giogen ic
Activity of th e Ter a togen ic a n d TNF r-Mod u la tor y Th a lid om id e An a logu e
2-(2,6-Dioxop ip er id in e-3-yl)p h th a lim id in e1
Frederick A. Luzzio,*,† Alexander V. Mayorov,† Sylvia S. W. Ng,‡ Erwin A. Kruger,‡ and William D. Figg‡
Department of Chemistry, University of Louisville, 2320 South Brook Street, Louisville, Kentucky 40292, and
Medicine Branch, National Cancer Institute, Bethesda, Maryland 20892
Received February 18, 2002
Versatile synthesis of the teratogenic, TNFR-modulatory, and antiangiogenic thalidomide
analogue 2-(2,6-dioxopiperidine-3-yl)phthalimidine (1) and its direct antiangiogenic properties
are described. With thalidomide or thalidomide derivatives as precursors, the synthesis involved
either carbonyl reduction/thiation-desulfurization or carbonyl reduction/acyliminium ion
reduction protocols. Compared to earlier studies with thalidomide, which was only active with
microsomal treatment, 1 exhibited marginal inhibitory activity in the rat aortic ring assay,
thereby demonstrating the requirement for metabolic activation.
Ch a r t 1
In tr od u ction
Angiogenesis is the development of new blood vessels
from preestablished vasculature. The relevance of an-
giogenesis in tumor growth and metastasis has driven
the study and development of antiangiogenic agents as
cancer therapeutics.2 Thalidomide and its analogues
have attracted considerable attention as angiogenesis
inhibitors, and questions remain about the similarity
of the teratogenic and antiangiogenic mechanisms and
how these compounds are activated in each of the
biological responses. During early studies on the struc-
ture-activity relationships of thalidomide and recent
studies of its antiangiogenic activity, the deoxythalido-
mide analogue 2-(2,6-dioxopiperidin-3-yl)phthalimidine
(EM-12) or 2-deoxythalidomide (1, Chart 1) has emerged
as the most biologically significant congener.3 1 exhib-
ited a higher incidence of teratogenic effects than
thalidomide (2, Chart 1) when tested in white New
Zealand rabbits.4a Moreover, in the marmoset monkey,
which is the thalidomide-sensitive primate model, 1 has
been used as a probe to corroborate the activities of the
individual enantiomers of thalidomide with its terato-
genic effects.5-7 Consequently, 1 and its analogues have
seen increased use in bioassays where the mode of
action and metabolic fate of thalidomide are still under
investigation. The rationale for employment of 1 in
biological studies is connected with its hydrolytic stabil-
ity in the biological test system. The stability of the
phthalimidine ring of 1 compared to the phthalimide
ring of 2 contributes to a simpler hydrolytic profile so
that biological disposition studies are unencumbered by
a plethora of metabolites and their associated break-
down products.8 The tumor necrosis factor-R (TNFR)-
inhibitory activity of 1 has been demonstrated in
lipopolysaccharide-stimulated (LPS) human peripheral
blood mononuclear cell (PMBC) bioassays.9 The con-
figurationally fixed acyclic 3-methyl-N-phthalimidino
analogue 3 (Chart 1) was found to inhibit B16BL6
pulmonary experimental metathesis in mice.10 The
common structural feature of compounds 1 and 3 (Chart
1) is the phthalimidine ring, which, in contrast to
thalidomide, imparts increased hydrolytic resistance to
the molecule and presumably facilitates more efficient
transport to the molecular target.11 The antiangiogenic
activity of 1 was demonstrated in the basic fibroblast
growth factor (bFGF)-induced rabbit cornea micropocket
model and was found to be at least as potent as 2.12 The
antiangiogenic activities of 1 and 2 were also investi-
gated in the chicken chorioallantoic membrane (CAM)
bioassay, and both compounds were inactive in the CAM
model, thereby suggesting the requirement for metabolic
activation of both compounds.12 Finally, exposure of 2
to human or rabbit microsomes while incubating with
rat aortic ring and human aortic endothelial cell cul-
tures results in inhibition of angiogenesis. Control
cultures conducted during the same study revealed that
2 incubated without microsomes was not inhibitory.,4b,13
The increased biological activity associated with ana-
logues containing the phthalimidine ring together with
reports that 2 exhibits a range of activities without
exposure to an exogeneous metabolic activation system
(MAS)14 prompted us to explore new synthetic routes
to 1 and analogues as well as to probe the activity of 1
when administered directly to an ex vivo angiogenesis
assay system.
Resu lts a n d Discu ssion
* To whom correspondence should be addressed. Phone: 502 852-
7323. Fax: 502 852-8149. E-mail: faluzz01@athena.louisville.edu.
† University of Louisville.
The synthesis commences with the preparation of
N-paramethoxybenzyl (PMB) thalidomide from (()-3-
‡ National Cancer Institute.
10.1021/jm020079d CCC: $25.00 © 2003 American Chemical Society
Published on Web 07/30/2003

3794 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 18
Luzzio et al.
Sch em e 1. Synthetic Scheme for
(()-N-PMB-thalidomide 7a
Sch em e 2. Synthesis of 1 through N-PMB Derivativesa
a
Reaction conditions: (a) Al(Hg), THF, H2O, room temp, 3 h;
a
(b) thiophenol, p-toluenesulfonic acid, CH2Cl2, room temp, 15 h;
(c) Raney Ni, EtOH, room temp, 5 h, 80% (from 7); (d) CAN,
CH3CN, H2O, room temp, 2 h, 61%.
Reaction conditions: (a) 4-methoxybenzylamine, pyridine, 60
°C, 12 h; (b) Ac2O/(CF3CO)2O, 70 °C, 12 h (51% from 4); (c)
4-methoxybenzylamine, DCC, Et3N, HOBt, CH3CN, 50 °C, 15 h
(54%).
Ch a r t 2
phthalimidoglutaric anhydride 4 or the corresponding
glutaric acid 5 together with 4-methoxybenzylamine
(Scheme 1). 4-Methoxybenzylamine facilitates the in-
troduction of nitrogen as an “ammonia equivalent” in a
formal sense when considering that the product is
afforded at the end of the sequence by a deprotection
step.15,16 Two methods were used for N-(4-methoxyben-
zyl)glutarimide formation. Method A entailed the treat-
ment of 4 with 4-methoxybenzylamine followed by direct
treatment of intermediate 6 with acetic anhydride in
the presence of a catalytic amount of trifluoroacetic
anhydride to provide N-(4-methoxybenzyl)thalidomide
7 in 51% yield. Method B involved the exposure of (()-
2-phthalimidoglutaric acid 5 with triethylamine, 1-hy-
droxybenzotriazole (HOBt), 1, 3-dicyclohexylcarbodiim-
ide (DCC), and 4-methoxybenzylamine at 50 °C for 15
h. Purification of the crude product under the same
conditions as those of method A provided the N-PMB-
3-phthalimidoglutarimide 7 in 54% yield. The conver-
sion of the PMB-protected phthalimide 7 to the phthal-
imidine (isoindolinone) moiety of 10 was based on the
chemoselective aluminum-amalgam-mediated reduction
of the phthalimide group to the corresponding hydroxy-
lactam, followed by phenylthiation and desulfurization,
as briefly described in a preliminary communication
(Scheme 2).7 While the direct phthalimide to phthali-
midine conversion may sometimes be facilitated with
zinc/acetic acid,18 the overall procedure described herein
is milder and also avoids the employment of the less-
selective hydridic reducing agents. Exposure of the
N-PMB glutarimide 7 to freshly prepared aluminum
amalgam (Al/Hg) in tetrahydrofuran/water followed by
removal of the solid byproducts and reaction solvent
system provided the crude hydroxylactam-substituted
N-PMB-imide 8 without any concomitant reduction of
the glutarimide ring. Direct treatment of hydroxylactam
8 with thiophenol and p-toluenesulfonic acid in dichlo-
romethane resulted in quantitative phenylthiation,
thereby furnishing the phenylthiolactam 9. Reductive
desulfurization of the sensitive phenylthiolactam 9
utilized Raney nickel in ethanol over 5 h at room
temperature and provided the N-(4-methoxybenzyl)-
protected glutarimide (10) in 80% yield after purification
by column chromatography. The age of the Raney nickel
and the reaction time have a profound effect on the
outcome of the desulfurization. Older lots of Raney
nickel/ethanol combined with prolonged reaction times
promoted ring-opening of the sensitive glutarimide ring,
which led to the recovery of the N-PMB-glutamine and
isoglutamine ethyl esters 11a and 11b (Chart 2).
Treatment of the N-PMB-protected phthalimidino-
glutarimide (10) with ceric ammonium nitrate (CAN)
in wet acetonitrile resulted in removal of the 4-meth-
oxybenzyl group and furnished (()-1 as a white solid
in 61% yield after purification by silica gel column
chromatography. The employment of the benzyl group
for protection of the glutarimide nitrogen was evaluated
as a comparison to the PMB group. The N-benzyl
derivative (13) was prepared in 75% yield by the
aluminum amalgam/phenyl-thiation/desulfurization se-
quence starting with N-benzylthalidomide 12, which in
turn was prepared from benzylamine and 4, albeit in
low yield.19 However, catalytic hydrogenation of 13 in
tetrahydrofuran with 10% palladium on carbon under
hydrogen (6 atm, 144 h) failed to effect any removal of
the benzyl group as evidenced by 70% recovery of
unreacted 13.
The second route to 1 employed a variation of a
scheme reported by Galons for the asymmetric synthesis
of thalidomide via N-BOC-glutarimide derivatives
(Scheme 3).20 An adaptation of the Galons scheme was
of interest to us because its employment in conjunction
with our reduction sequence could, in principle, be used
in an asymmetric synthesis of 1 given that optically pure

Thalidomide Metabolites and Analogues
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 18 3795
Sch em e 3. Preparation of 1 from
N-Benzyloxythalidomide 14a
a
Reaction conditions: (a) Ac2O/(CF3CO)2O, then pyr/BnONH3Cl,
then Ac2O/(CF3CO)2O (97%); (b) Al(Hg), THF, H2O, room temp;
(c) Et3SiH, (CF3CO)2O, CH2Cl2, 0 °C, 2 h (41% from 14); (d) Pd-
C/MeOH /H2, 24 h; (e) BrCH2COPh, TEA, DMAP, CH3CN, 4 h,
room temp (21% from 15).
Sch em e 4. Direct Conversion of Thalidomide (2) to 1a
a
Reaction conditions: (a) Al(Hg), EtOH, H2O, AcOH; (b) Et3SiH,
(CF3CO)2O, AcOH, 70 °C, 3 h (66%).
starting materials were used. Conversion of 5 to 14
(97%), which was effected through anhydride 4, utilized
benzyloxyamine hydrochloride. Reduction of NBO tha-
lidomide 14 to the corresponding intermediate hydroxy-
lactam was effected with Al/Hg similar to the conversion
of 7 to 8. The crude NBO hydroxylactam was not
purified but directly treated with triethylsilane/trifluo-
roacetic acid21 in dichloromethane to afford 15 as an
amorphous white solid in 41% yield.
F igu r e 1. Top: inhibition of rat aortic ring microvessel
growth (daily treatment for 4 days). Assays were done with
medium (0.05% DMSO) and 30 µM carboxyamidotriazole
(CAI)/medium as controls and 50, 100, and 200 µM 1. Bot-
tom: inhibition of rat aortic ring microvessel growth (single
treatment over 4 days). Assays were done with medium and
30 µM carboxyamidotriazole (CAI)/medium as controls and 50,
100, and 200 µM 1 in medium (0.05% DMSO). The data
represent the mean ( SEM (n ) 4).
Removal of the N-benzyloxy group required a two-
step process.20 First, the benzyl group was cleaved with
palladium on activated carbon under an atmosphere of
hydrogen. Removal of the remaining N-hydroxyl group
was effected by phenacyl ether formation and cleavage
with phenacyl bromide/triethylamine/4-(dimethylami-
no)pyridine, thereby furnishing 1 in 21% yield after
flash chromatography and crystallization. A more direct
route to 1 was explored that did not involve various
modes of N-substitution or protection and employed
thalidomide itself as the starting material (Scheme 4).
(()-Thalidomide, prepared from 2-phthalimidoglutaric
acid 5 by a modified procedure described by Helm,3 was
treated with in situ prepared aluminum amalgam in a
mixture of ethanol/water/acetic acid (20:2:1) under
reflux. As in Scheme 3, the crude corresponding hy-
droxylactam again was not purified but directly treated
with a mixture of acetic/trifluoroacetic acid followed by
heating with triethylsilane to afford 1 in 66% yield after
recrystallization.
influx.23 The inhibition experiments utilized working
concentrations of 1 in endothelial basal media (EBM)
and 0.05% DMSO vehicle, with EBM/0.05% DMSO as
a control and CAI (30 µM) in EBM/0.05% DMSO.
The assays entailed both a single direct administra-
tion of the compounds to the rat aortic ring cultures on
day 1 together with a separate regimen that entailed
daily treatments of the cultures for 4 days. Digital
photographic images of the cultures were recorded each
day. The NIH Image program was used to quantify pixel
density in the digital images by using a gray-scale
analysis of background versus light pixels (outgrowths).24
In turn, peripheral microvessel growth is quantitated
and characterized. Moreover, nonspecific or inner-ring
growth can be omitted and the method of quantitation
allows for statistical analysis of the results. The title
compound 1 was increasingly inhibitory at 100 and 200
µM (30%) when compared to the positive control, car-
boxyamidotriazole (CAI), which inhibited outgrowths by
90% as indicated by the bar graph (Figure 1, top).
Microvessel outgrowth was not suppressed by a single
treatment of 1 at 100 or 200 µM (Figure 1, bottom).
The rat aortic ring microvessel growth assay was used
to evaluate the direct antiangiogenic properties of (()-
1.13 The bioassay is a standard laboratory ex vivo
preclinical screening experiment that is a modification
of the procedure developed by Nicosia.22 The technique
was employed to measure the inhibition of new mi-
crovessel growth by 1 versus a positive control, car-
boxyamidotriazole (CAI). CAI is an effective angiogen-
esis inhibitor in the rat aortic ring model, thereby
suppressing outgrowths through inhibition of calcium

3796 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 18
Luzzio et al.
The activity of thalidomide or its analogues when
bioassayed without in vivo or ex vivo metabolic activa-
tion is not without precedent. Direct administration of
2 significantly depletes glutathione and cysteine con-
centrations in New Zealand white rabbit whole embryo
(WEC) cultures. In turn, depleted glutathione and
cysteine levels interfere with normal redox cycling in
the developing embryo, thereby leading to developmen-
tal malformations.14
Con clu sion
We have detailed three serviceable synthetic routes
to the thalidomide analogue (1). The direct conversion
of thalidomide to 1 through the Al(Hg)/triethylsilane
reduction provided the best yields. The mild, selective
nature of all three routes makes them applicable to the
preparation of the isoindolinone core structure and
analogues of 1, in racemic and optically active form as
well as in suitable quantities for broad-range biological
evaluation. The direct antiangiogenic activity of 1 in the
rat aortic ring assay was found to be quite limited, and
although of minimal statistical significance along the
applied concentration range, an inhibitory trend was
apparent. Within the framework of previous in vivo and
ex vivo studies with thalidomide and in vivo studies
with 1, the lower range of inhibitory activity of 1 in the
present ex vivo study further demonstrates the require-
ment for prior metabolic activation.
Exp er im en ta l Section
Gen er a l P r oced u r es. Melting points are uncorrected. All
the starting materials and products described are racemic
mixtures. Dichloromethane and acetonitrile were distilled from
CaH2, and pyridine was distilled from barium oxide. Benzylic
amines were fractionally distilled prior to use as reactants.
All other reagents and solvents were ACS reagent grade and
used as commercially available. Flash column chromatography
was carried out using silica gel 60 (E. Merck 9385, 230-400
mesh).24 Silica gel filtrations employed silica gel 60 (E. Merck
7734, 70-230 mesh). Celite filtrations employed Celite 521.
Analytical thin-layer chromatographic (TLC) separations uti-
lized 0.25 mm glass-backed silica gel plates (E. Merck 5715,
silica gel 60 F254). The TLC chromatograms were developed
by dipping the plates in 2% anisaldehyde in ethanol or 2.5%
phosphomolybdic acid in ethanol followed by heating (hot
plate). Solutions, reaction mixtures, and chromatographic
fractions were concentrated under vacuum using a standard
rotary evaporator. Carbon signals marked with an asterisk
represent methyl and methyne carbons as determined by APT
experiments. High-resolution mass spectoscopic analyses
(HRMS) were performed by the Nebraska Center for Mass
Spectrometry, University of Nebraska, Lincoln and the Uni-
versity of CaliforniasRiverside Mass Spectrometry Facilty.
2-[1-(4-Meth oxyben zyl)-2,6-d ioxop ip er id in -3-yl]isoin -
d ole-1,3-d ion e (7). Meth od A. (()-3-Phthalimidoglutaric
anhydride 4 (1.49 g, 5.76 mmol) was placed in an oven-dried
round-bottom flask. Freshly distilled pyridine (8 mL) was
added and then followed by 4-methoxybenzylamine (750 µL,
5.74 mmol). The reaction mixture was stirred at room tem-
perature under an atmosphere of nitrogen (0.5 h) followed by
heating (60 °C, 12 h). Pyridine was then removed under
vacuum, which provided a syrupy residue. To the residue was
added acetic anhydride (15 mL) followed by a catalytic amount
of trifluoroacetic anhydride (100 µL), and the mixture was
heated at 70 °C (12 h). Upon completion of the reaction, as
determined by TLC, the volatile components were removed
under vacuum followed by addition of ethyl acetate and
filtration through a silica gel plug. The yellow filtrate was
concentrated under vacuum and flash-chromatographed (hex-
anes/ethyl acetate, 1:1) to provide 1.11 g (51%) of N-PMB-
F igu r e 2. Images of rat aortic ring assays: (top) treatment
of aortic ring culture with medium (0.05% DMSO); (middle)
treatment of aortic ring culture with 30 µM carboxyamido-
triazole (CAI)/medium; (bottom) treatment of aortic ring
culture with 200 µM 1 administered daily for 4 days in medium
(0.05% DMSO).
Photographic images of the aortic rings grown in
growth media (Figure 2, top), in the presence of CAI
(Figure 2, middle), and in the presence of 200 µM 1
administered daily (Figure 2, bottom) illustrate the
relative inhibitory effects of the compounds. Compared
to the direct daily administration of thalidomide in the
rat aortic ring assay in which no inhibition of microves-
sel growth over the 40-200 µM concentration range was
exhibited,13 the slightly increased inhibitory activity by
unactivated 1 is apparent. However, within the margin
of error, the activities of 1 and 2 are comparable and
the necessity for metabolic activation is demonstrated.

Thalidomide Metabolites and Analogues
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 18 3797
thalidomide 7 as an amorphous white solid. Rf ) ) 0.41
(hexane/ethyl acetate, 1:1). 1H NMR (500 MHz, CDCl3) δ: 7.87
(m, J ) 3.2, 5.5 Hz, 2H), 7.74 (m, J ) 5.2, 5.5 Hz, 2H), 7.31 (d,
J ) 8.6 Hz, 2H), 6.8 (d, J ) 8.6 Hz, 2H), 5.00 (d, J ) 14 Hz,
1H), 4.96 (m, 1H), 4.82 (d, J ) 14 Hz, 1H), 3.76 (s, 3H), 2.96
(m, 1H), 2.76 (m, 2H), 2.08 (m, 1H). 13C NMR (125 MHz,
CDCl3) δ: 170.7, 168.6, 167.4, 159.0, 134.4*, 131.8, 130.4*,
128.9, 123.8*, 113.8*, 55.2*, 50.2*, 43.3, 32.1, 22.0. HRMS m/z
(M+) calcd for C21H18N2O5 378.1216, found 378.1215. IR (KBr)
3-(1-Oxo-1,3-d ih yd r oisoin d ol-2-yl)p ip er id in e-2,6-d i-
on e (1). The N-PMB derivative 10 (66 mg, 0.181 mmol) was
dissolved in acetonitrile (2 mL). Water (400 µL) was added
followed by ceric ammonium nitrate (405 mg, 0.74 mmol). The
bright-orange solution was stirred (2 h) and then quenched
with aqueous sodium bicarbonate (1:1, w/w) and sodium
bisulfite (1:1, w/w) solutions. The volatile components were
removed under vacuum, and the solid residue was suspended
in acetone followed by filtration through a short silica gel (70-
230 mesh) column. Flash-column silica gel was then added to
the solution, the solvent was evaporated, and the silica gel
with the adsorbed compound was applied to a flash-chroma-
tography column and eluted with dichloromethane/acetone (10:
1). Upon elution of all the byproduct 4-methoxybenzaldehyde,
the polarity of the solvent system was increased to 2:1. The
homogeneous fractions were combined to provide 27 mg (61%)
of 1 as a white amorphous solid. Recrystallization of an
analytical sample from methyl cellosolve gave 1 as white
crystals. Mp 237-238 °C. Rf ) 0.22 (dichloromethane/acetone,
1785, 1771, 1726, 1682 cm-1
.
Meth od B. (()-2-Phthalimidoglutaric acid 5 (62 mg, 0.224
mmol) was dissolved in dry acetonitrile (2 mL) and cooled to
0 °C, and triethylamine (78 µL, 0.560 mmol) was added to the
mixture while stirring. To the resultant white suspension was
added 1-hydroxybenzotriazole (123 mg, 0.911 mmol), 4-meth-
oxybenzylamine (35 µL, 0.268 mmol), and 1,3-dicyclohexyl-
carbodiimide (236 mg, 1.14 mmol). The reaction mixture was
allowed to warm to room temperature over 1 h followed by
heating at 50 °C for 15 h. The reaction solvent was removed
under vacuum, and the yellow residue was dissolved in ethyl
acetate and filtered through a short pad of silica gel. Concen-
tration of the filtrate followed by flash-column chromatography
(hexane/ethyl acetate, 2:1) provided 46 mg (54%) of 7 by
trituration of the combined chromatographic fractions with
ethyl acetate. The spectral details were identical to the product
prepared above by method A.
1
5:1). H NMR (500 MHz, methanol-d4) δ: 7.82 (d, J ) 7.3 Hz,
1H), 7.65 (t, J ) 7.3 Hz, 1H), 7.60 (d, J ) 7.3 Hz, 1H), 7.54 (t,
J ) 7.3 Hz, 1H), 5.17 (dd, J ) 5.4 Hz, 13.7, 1H), 4.51 (m, J )
17.0, 27.3 Hz, 2H), 2.92 (ddd, J ) 5.4, 13.7, 18.1 Hz, 1H), 2.79
(ddd, J ) 2.4, 4.4, 13.7 Hz, 1H), 2.51 (ddd, J ) 4.9, 13.2, 26.3
Hz, 1H), 2.08 (m, J ) 2.4, 5.4, 18.1 Hz, 1H). 13C NMR (125
MHz, methanol-d4) δ: 173.5, 171.0, 170.3, 142.5, 132.2*, 131.5,
128.1*, 123.2*, 52.5*, 31.2, 22.9. HRMS m/z (M+) calcd for
1-(4-Met h oxyb en zyl)-3-(1-oxo-1,3-d ih yd r oisoin d ol-2-
yl)p ip er id in e-2,6-d ion e (10). Aluminum amalgam was pre-
pared by rinsing coils of food-grade aluminum foil (68 mg, 2.5
mmol) with diethyl ether. Each coil was immersed in 2%
aqueous mercuric chloride solution while agitating (20 s),
followed by immersing in distilled water (1-2 s) and then
adding to a solution of 7 (1.0 mmol) in THF (5 mL) and distilled
water (250 µL, 13.9 mmol). The reaction mixture was stirred
at room temperature (3 h), during which time evolution of
hydrogen occurred with the formation of a gray suspension.
The reaction mixture was then vacuum-filtered through Celite
while washing with methanol. The clear filtrate was concen-
trated to a syrup under aspirator vacuum followed by further
subjection to high vacuum (12 h) to remove residual water.
The syrupy crude lactam 8 was dissolved in dry dichlo-
romethane (3 mL) followed by addition of thiophenol (190 µL,
2.1 mmol) and p-toluenesulfonic acid (9.0 mg, 0.05 equiv). The
reaction mixture was then stirred under an atmosphere of
nitrogen at room temperature (15 h). Upon completion of the
reaction as determined by TLC analysis, the reaction mixture
was diluted with dichloromethane, washed with 10% aqueous
sodium bicarbonate solution, and dried over anhydrous sodium
sulfate. Removal of the drying agent followed by concentration
and flash-column chromatography (hexane/ethyl acetate, 1:1)
of the residue afforded the sensitive phenylthiolactam 9. The
semipure phenylthiolactam 9 was added to 95% ethanol (10
mL) followed by an aqueous suspension of Raney nickel (2 mL).
The black suspension was stirred at room temperature (5 h)
while monitoring by TLC. After completion of the desulfur-
ization, the reaction mixture was filtered through Celite under
gentle suction while washing the filter cake with methanol to
prevent the excess Raney nickel from drying. The filtrate was
concentrated and flash-column-chromatographed (toluene/
ethyl acetate, 2:1) to furnish 10 (291 mg, 80%) as white
crystals. Mp 68-71 °C. Rf ) 0.36 (hexane/ethyl acetate, 1:1).
1H NMR (500 MHz, CDCl3) δ: 7.88 (d, J ) 7.8 Hz, 1H), 7.51
(dt, J ) 0.98 Hz, 7.3, 1H), 7.47 (m, 2H), 7.32 (d, J ) 8.8 Hz,
2H), 6.80 (d, J ) 8.8 Hz, 2H), 5.20 (dd, J ) 5.4, 13.7 Hz, 1H),
4.88 (s, 2H,), 4.42 (d, J ) 16.1 Hz, 1H), 4.30 (d, J ) 16.1 Hz,
1H), 3.77 (s, 3H), 2.97 (ddd, J ) 2.4, 4.9, 18.1 Hz, 1H), 2.85
(ddd, J ) 5.4, 13.7, 18.1 Hz, 1H), 2.28 (ddd, J ) 4.9, 13.7, 26.3
Hz, 1H), 2.14 (m, J ) 2.4, 5.4, 12.7 Hz, 1H). 13C NMR (125
MHz, CDCl3) δ: 171.2, 170.2, 169.5, 141.7, 132.2*, 131.8,
130.8*, 129.3, 128.4*, 124.3*, 123.2, 114.0*, 55.5*, 52.8*, 47.3,
43.4, 32.4, 22.9. HRMS (FAB) m/z (M + H)+ calcd for
C
H
13H12N2O3 244.0848, found 244.0842. Anal. Calcd for C13
-
12N2O3: C, 63.92; H, 4.95; N, 11.45. Found: C, 63.73; H, 4.73;
N, 11.05. IR (KBr): 3204, 3099, 1732, 1705, 1680 cm-1
.
2-(1-Be n zyl-2,6-d ioxop ip e r id in -3-yl)isoin d ol-1,3-d i-
on e (12). N-Benzylthalidomide (12) was prepared (13%) from
(()-3-phthalimidophthalic anhydride 4 (871 mg, 3.36 mmol)
and benzylamine (370 mg, 3.39 mmol) in a manner analogous
to that of 7 by method A. Rf ) 0.42 (hexane/ethyl acetate, 1:1).
1H NMR (500 MHz, CDCl3) δ: 7.86 (m, J ) 2.9, 5.4 Hz, 2H),
7.75 (m, J ) 2.9, 5.4 Hz, 2H), 7.37 (d, J ) 8.3 Hz, 2H), 7.29 (t,
J ) 7.3 Hz, 2H), 7.23 (t, J ) 7.3 Hz, 1H), 5.08 (d, J ) 14.2 Hz,
1H), 5.03 (m, 1H), 4.91 (d, J ) 14.2 Hz, 1H), 2.97 (dt, J ) 29,
15.6 Hz, 1H), 2.80 (m, 2H), 2.10 (m, 1H). 13C NMR (125 Mz,
CDCl3) δ: 170.1, 168.9, 167.6, 136. 8, 134.6*, 132.0, 128.9*,
128.7*, 127.8*, 124.0*, 50.5*, 44.1, 32.3, 22.2. HRMS m/z (M+)
calcd for C20H16N2O4 348.1110, found 348.1121. IR (KBr):
1772, 1724, 1687, 1610 cm-1
.
1-Ben zyl-3-(1-oxo-1,3-d ih yd r oisoin d ol-2-yl)p ip er id in e-
2,6-d ion e (13). The N-benzyl derivative (13) was prepared in
the same manner as N-PMB derivative (10) starting from 2-(1-
benzyl-2,6-dioxopiperidin-3-yl)isoindol-1,3-dione (N-benzyltha-
lidomide) 12 (102 mg, 0.29 mmol). Flash column chromatog-
raphy (toluene/ethyl acetate, 2:1) provided 73 mg (75%) of 13
as a white amorphous solid. Rf ) 0.15 (hexane/EtOAc, 1:1).
1H NMR (CDCl3) δ: 7.89 (d, J ) 7.8 Hz, 1H), 7.57 (t, J ) 7.8
Hz, 1H), 7.47 (m, 2H), 7.37 (d, J ) 7.8 Hz, 2H), 7.27 (m, 3H),
5.28 (dd, J ) 5.4, 13.7 Hz, 1H), 4.96 (s, 2H), 4.43 (d, J ) 15.6
Hz, 1H), 4.32 (d, J ) 15.6 Hz, 1H), 2.99 (m, 1H), 2.87 (ddd, J
) 5.4, 13.7, 17.6 Hz, 1H), 2.31 (ddd, J ) 4.4, 13.2, 26.3 Hz,
1H), 2.16 (m, 1H). 13C NMR (125 MHz, CDCl3) δ: 171.2, 170.2,
169.5, 141.7, 137.0, 132.2*, 131.8, 129.2*, 128.7*, 128.4*,
127.9*, 124.4*, 123.2*, 52.7*, 47.4, 44.0, 32.4, 23.0. HRMS m/z
(M+) calcd for C20H18N2O3 334.1317, found 334.1322. IR
(KBr): 1770, 1721, 1708, 1662, 1621 cm-1
.
2-(1-Ben zyloxy-2,6-d ioxop ip er id in -3-yl)isoin d ol-1,3-d i-
on e (14). (()-2-Phthalimidoglutaric acid 5 (5.022 g, 18.1 mmol)
was heated with acetic anhydride (25 mL, mmol) and trifluo-
roacetic anhydride (0.5 mL, mmol) under an atmosphere of
nitrogen at 100° C for 13 h. The volatile components were
removed under high vacuum, which left a white residue of
crude anhydride 4. Pyridine (15 mL, mmol) was added, and
then benzyloxyamine hydrochloride (2.89 g, 18.1 mmol) was
added. The mixture was then heated under an atmosphere of
nitrogen at 60 °C (3 h) followed by removal of the pyridine
under high vacuum at room temperature to yield a dark syrupy
residue. To the residue was added acetic anhydride (25 mL,
mmol) and trifluoroacetic anhydride (0.5 mL, mmol) followed
C
21H21N2O4 365.1501, found 365.1495. IR (KBr): 1702, 1686,
1613 cm-1
.

3798 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 18
Luzzio et al.
by heating at 100 °C (4 h) under an atmosphere of nitrogen.
Concentration of the reaction mixture under high vacuum
followed by flash-column chromatography (hexane/ethyl ace-
tate, 2:1) afforded N-benzyloxythalidomide 14 (6.38 g, 97%)
as a white foam. Rf ) 0.32 (hexane/ethyl acetate, 1:1). 1H NMR
(500 MHz, CDCl3) δ: 7.90 (m, J ) 2.9, 5.4 Hz, 2H), 7.78 (m, J
) 2.9, 5.4 Hz, 2H), 7.55 (m, 2H), 7.37 (m, 3H), 5.06 (s, 2H),
5.09-5.04 (m, 1H), 3.00-2.9 (m, 1H), 2.83-2.74 (m, 2H), 2.16-
2.08 (m, 1H). 13C NMR (125 MHz, CDCl3) δ: 167.4, 167.1,
165.3, 134.8*, 134.0, 131.9, 130.3*, 129.4*, 128.7*, 124.1*, 78.7,
50.6*, 31.5, 22.2. HRMS m/z (M+H)+ calcd for C20H17N2O5
365.1137, found 365.1123. IR (KBr): 1784, 1771, 1751, 1719
glacial acetic acid (15 mL). Trifluoroacetic acid (9 mL, 121
mmol) was then added, and the mixture was heated at 70 °C
(10 min). Triethylsilane (6.6 mL, 41.3 mmol) was slowly added,
and heating was continued (3 h) at 70 °C. The resulting
heterogeneous mixture was allowed to cool to room tempera-
ture, diluted with methanol (50 mL), and then vacuum-filtered
through Celite. The pale-yellow filtrate was concentrated to a
solid and then traces of acetic and trifluoroacetic acid were
removed under vacuum (0.5 Torr) at 60 °C (14 h). The solid
residue was triturated with hot ethyl actetate and then cooled
to -10 °C for 2 h. The solid precipitate (2.41 g) was collected
by filtration, washed twice with ethyl acetate (40 mL), and
dried under vacuum. The filtrate was reduced to 10 mL and
provided another 0.96 g of crude product that was collected
by filtration. Combining both crops of crude product and
recrystallization from methyl cellosolve afforded 2.83 g (66%)
of pure 1 with melting points and spectral data that are
identical to those of the samples prepared above.
Ra t Aor tic Rin g Cu ltu r e Bioa ssa y. Female 6-8 week old
Sprague-Dawley rats were euthanized by asphyxiation with
CO2 followed by removal of the thoracic aorta. With a dissect-
ing microscope, 1 mm thick rings were then cut. Each ring
was rinsed eight times with culture medium, embedded in
Matrigel (Collaborative Biomedical, Bedford, MA), and incu-
bated in growth-factor-free endothelial cell basal medium
(EBM). The cell basal medium (Clonetics, Walkersville, MD)
contained 10 µg/mL gentiamycin, 100 units/mL penicillin, 100
mg/mL streptomycin, and 0.25 µg/mL amphotericin. At 24 h
intervals, the supernatant was replaced with 9 µg/mL 1
dissolved in DMSO. Incubation of the cultures was at 37 °C
with 5% CO2. The experiments utilized aortic rings taken from
four different rats. The aortic ring preparations were cultured
for 4 days, during which time microvessel growth was assessed
daily. To confirm the presence of endothelial cells in the new
microvessel growth, they were stained with factor VIII and
CD34 by previously detailed methods.13 Control cultures
consisted of EBM/0.05% DMSO and a positive control, car-
boxyamidotriazole (CAI, 30 µM) in EBM/0.05% DMSO. Digital
photographic images of both the control and test cultures were
recorded each day. The images of the rat aortic rings (Figure
2) were then scanned onto a photo CD, and the pixel density
in each image was quantified by employing a gray-scale
analysis of background versus light pixels (outgrowths) as
determined by the NIH Image program.24 The bar graphs in
Figure 1, which show relative percent microvessel growth, are
derived from the relative pixel density ((pixel)2 units) of the
images. The bar graphs were generated using KaleidaGraph
software. Comparisons were made using Student’s t-test with
P < 0.05 as the criterion for statistical significance. The data
were represented as the mean ( SEM (n ) 4).
cm-1
.
1-Ben zyloxy-3-(1-oxo-1,3-d ih yd r oisoin d ol-2-yl)p ip er i-
d in e-2,6-d ion e (15). N-Benzyloxythalidomide 14 (634 mg,
1.74 mmol) was dissolved in THF (6.0 mL) followed by the
addition of water (0.5 mL, 27.7 mmol). To the vigorously stirred
solution was added aluminum amalgam, which was prepared
from aluminum foil (141 mg, 5.22 mmol) as described above.
After the reduction was complete (3 h), the resulting gray
suspension was diluted with methanol (10 mL) and vacuum-
filtered through Celite. The clear filtrate was concentrated
under reduced pressure, and the residue was dissolved in dry
dichloromethane (10 mL). The solution was cooled to 0 °C.
Then trifluoroacetic acid (270 µL, 3.5 mmol) was added, and
the resulting yellow solution was stirred for 10 min. Trieth-
ylsilane (834 µL, 5.22 mmol) was then slowly added, and
stirring was continued for 2 h at 0 °C. The reaction mixture
was then concentrated under vacuum and purified by flash
chromatography (toluene/ethyl acetate, 3:1) to provide 15 (251
mg, 41%) as a white amorphous solid. Rf ) 0.23 (hexane/ethyl
1
acetate, 1:1). H NMR (500 MHz, CDCl3) δ: 7.88 (d, J ) 7.8
Hz, 1H), 7.58 (t, J ) 7.8 Hz, 1H), 7.50 (m, 3H), 7.44 (d, J )
7.8 Hz, 1H), 7.37 (m, 3H), 5.25 (m, 1H), 5.04 (dq, J ) 2.4, 9.3
Hz, 2H), 4.37-4.22 (dd, 2H), 2.94 (m, 1H), 2.84 (m, 1H), 2.24
(m, 1H), 2.10 (m, 1H). 13C NMR (125 MHz, CDCl3) δ: 169.4,
167.5, 166.7, 141.7, 133.9, 132.3*, 131.6, 130.4*, 129.5*, 128.7*,
128.5*, 123.2*, 78.4, 53.1*, 53.0*, 47.3, 31.6, 22.9. HRMS m/z
(M + H)+ calcd for C20H18N2O4 351.1346, found 351.1340. IR
(KBr): 1773, 1716, 1699, 1600, 1621 cm-1
.
P r ep a r a tion of 1 fr om th e N-Ben zyloxy Der iva tive
(15). N-Benzyloxy-protected 15 (4.69 g, 13.4 mmol)) was
dissolved in methanol (30 mL). Palladium on activated carbon
(10%, 354 mg) was carefully added, and the mixture was
stirred under an atmosphere of hydrogen (1 atm, 24 h). The
catalyst was removed by filtration through Celite, and the
filtrate was concentrated under reduced pressure. Anhydrous
acetonitrile (150 mL) was then added to the syrupy residue
followed by the addition of R-bromoacetophenone (2.66 g, 13.4
mmol), triethylamine (1.9 mL, 13.4 mmol), and 4-(dimethyl-
amino)pyridine (166 mg, 1.36 mmol). The resulting yellow
solution was stirred (4 h) and was then concentrated under
vacuum. The dark-orange syrupy resudue was flash-chromato-
graphed twice (toluene/ethyl acetate, 5:1; then toluene/ethyl
acetate, 3:1) to afford 700 mg (21%) of 1 as white crystals. The
compound exhibited the same melting point and spectral
characteristics as those of the compound obtained from the
N-deprotection of the N-PMB derivative (10).
P r ep a r a tion of 1 by Dir ect Red u ction -Deoxygen a tion
of (()-Th a lid om id e (2). A suspension of racemic thalidomide
2 (4.52 g, 1.75 mmol) in ethanol (100 mL) was heated to reflux
(5 min). Water (10 mL) waas added followed by glacial acetic
acid (5 mL). Aluminum foil (1.41 g, 5.26 mmol) was cut into 1
cm2 pieces, rinsed with diethyl ether, and added to the
thalidomide/acetic acid/water suspension. To the well-stirred
suspension, mercury(II) chloride (22 mg, 81 µmol) was added,
and the mixture was then refluxed.
Ack n ow led gm en t. The research was supported by
the Intramural Research Program of the National
Cancer Institute.
Refer en ces
(1) (a) Thalidomide Analogs and Metabolites, Part 3. For Part 2,
see the following. Luzzio, F. A.; Thomas, E. M.; Figg, W. D.
Tetrahedron Lett. 2000, 41, 7151-7155. (b) Presented at the
222nd National Meeting of the American Chemical Society,
Chicago, IL, August 29, 2001; Paper MEDI 0264.
(2) Rosen, L. Antiangiogenic Strategies and Agents in Clinical
Trials. Oncologist 2000, 5 (Suppl.), 20-27.
(3) Helm, F. C.; Frankus, E.; Friderichs, E.; Graudums, I.; Flohe´,
L. Comparative Teratological Investigation of Compounds Struc-
turally and Pharmacologically Related to Thalidomide Arzneim.-
Forsch. 1981, 31, 941-949.
(4) (a) Udagawa, T.; Verheul, H. M. W.; D’Amato, R. J . Thalidomide
and Analogs. In Antiangiogenic Agents in Cancer Therapy;
Teicher, B. A., Ed.; Humana Press: Totowa, NJ , 1999; pp 263-
274. (b) Figg, W. D.; Reed, E.; Green, S.; Pluda, J . Thalidomide
A Prodrug That Inhibits Angiogenesis. In Antiangiogenic Agents
in Cancer Therapy; Teicher, B. A., Ed.; Humana Press: Totowa,
NJ , 1999; pp 407-422.
(5) Schmahl, H.-J .; Nau, H.; Neubert, D. The Enantiomers of the
Teratogenic Thalidomide Analogue EM 12. Arch. Toxicol. 1988,
62, 200-204.
During the reaction, aliquots were analyzed by NMR, which
indicated that complete conversion to the corresponding hy-
droxylactam was complete in 3 h. The reaction mixture was
cooled, vacuum-filtered through a thin pad of Celite, and
concentrated under vacuum to give the crude hydroxylactam
as a white amorphous solid. The crude hydroxylactam was
dried under vacuum at room temperature and then mixed with

Thalidomide Metabolites and Analogues
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 18 3799
(6) Heger, W.; Klug, S.; Schmahl, H.-J .; Nau, H.; Merker, H.-J .;
Neubert, D. Embryotoxic Effects of Thalidomide Derivatives on
the Non-Human Primate Callithrix jacchus. Arch. Toxicol. 1988,
62, 205-208.
(7) Schmahl, H.-J .; Heger, W.; Nau, H. The Enantiomers of the
Teratogenic Thalidomide Analogue EM 12. Toxicol. Lett. 1989,
45, 23-33.
(8) Winckler, K.; Klinkmu¨ller, K. D.; Schmal, H. J . Determina-
tion of the Thalidomide Analogues 2-(2,6-Dioxopiperidine-3-yl)
Phthalimidine (EM 12), 2-(2,6-dioxopiperidine-4yl) Phthalimi-
dine (EM 16) and Their Metabolites in Biological Samples. J .
Chromatog. Biomed. Appl. 1989, 488, 417-425.
(9) Kaplan, G.; Sampaio, E. P. Method of Treating Abnor-
mal Concentrations of TNFR U.S. Patent 5,385,901, 1992 (to
Rockefeller University); Chem. Abstr. 1992, 117, 226313a.
(10) Shah, J . H.; Swartz, G. M.; Papathanassiu, A. E.; Treston, A.
M.; Fogler, W. E.; Madsen, J . W.; Green, S. J . Synthesis and
Enantiomeric Separation of 2-Phthalimidinoglutaric Acid Ana-
logues: Potent Inhibitors of Tumor Metastasis. J . Med. Chem.
1999, 42, 3014-3017.
(11) Neubert, R.; Nogueira, A. C.; Neubert, D. Thalidomide Deriva-
tives and the Immune System I. Changes in the Pattern of
Integrin Receptors and Other Surface Markers on T Lymphocyte
Subpopulations of Marmoset Blood. Arch. Toxicol. 1993, 67,
1-17.
(12) D’Amato, R. J .; Loughnan, M. S.; Flynn, E.; Folkman, J .
Thalidomide is an Inhibitor of Angiogenesis. Proc. Natl. Acad.
Sci. U.S.A. 1994, 91, 4082-4085.
(13) Bauer, K. S.; Dixon, S. C.; Figg, W. D. Inhibition of Angiogenesis
by Thalidomide Requires Metabolic Activation Which Is Species-
Dependent. Biochem. Pharmacol. 1998, 55, 1827-1834.
(14) Hansen, J . M.; Carney, E. W.; Harris, C. Differential Alteration
by Thalidomide of the Glutathione Content of Rat vs. Rabbit
Conceptuses in Vitro. Reprod. Toxicol. 1999, 13, 547-554.
(15) (a) Green, T. W.; Wuts, P. G. M. Protective Groups in Organic
Synthesis, 2nd ed.; Wiley: New York, 1991; p 400. (b) Although
the PMB route to 1 was described in an earlier publication (see
ref 17), no experimental details were provided. Full experimen-
tal details are provided in this report where the yields in some
steps differ because of scale-up.
(16) Luzzio, F. A.; Mayorov, A. V.; Figg, W. D. Thalidomide Metabo-
lites. Part I: Derivatives of (+)-2-N-Phthalimido-γ-hydroxy-
glutamic Acid Tetrahedron Lett. 2000, 41, 2275-228.
(17) Luzzio, F. A.; Piatt Zacherl, D. P.; Figg, W. D. A Facile Scheme
for Phthalimide/-Phthalimidine Conversion. Tetrahedron Lett.
1999, 40, 2087-2090.
(18) Guo, Z.; Schultz, A. G. Organic Synthesis Methodology. Prepara-
tion and Diastereoselective Birch Reduction-Alkylation of
3-Substituted 2-Methyl-2,3-Dihydroisoindol-1-ones. J . Org. Chem.
2001, 66, 2154-2157.
(19) (a) Benzylamines are not well-suited for glutarimide formation
because of their lower nucleophilicity (see ref 20). In contrast,
â-phenethylamine gave the corresponding glutarimide in 83%
yield using method A. (b) Fox, D. J .; Reckless, J .; Warren, S. J .;
Grainger, D. J . Design, Synthesis and Preliminary Pharmaco-
logical Evaluation of N-Acyl-3-aminoglutarimides as Broad-
Spectrum Chemokine Inhibitors in Vitro and Anti-inflammatory
Agents in Vivo. J . Med Chem. 2002, 45, 360-370.
(20) Robin, S.; Zhu, J .; Galons, H.; Pham-Huy, C.; Claude, J . R.;
Tomas, A.; Viossat, B. A Convenient Asymmetric Synthesis of
Thalidomide. Tetrahedron: Asymmetry 1995, 6, 1249-1252. The
electron-donating capacity of the NBO group presumably reduces
racemization of the product thalidomide.
(21) Hart, D. J .; Yang, T.-K. N-Acyliminium Ion Rearrangements:
Generalities and Application to the Synthesis of Pyrrolizidine
Alkaloids. J . Org. Chem. 1985, 50, 235-242.
(22) Nicosia, R. F.; Ottinetti, A. Growth of Microvessels in Serum-
Free Matrix Culture of Rat Aorta. Lab. Invest. 1990, 63, 115-
122.
(23) Hamby, J . M.; Showalter, H. D. H. Small Molecule Inhibitors of
Tumor-Promoted Angiogenesis, Including Protein Tyrosine Ki-
nase Inhibitors. Pharmacol. Ther. 1999, 82, 169-193.
(24) The NIH Image program was developed at the U.S. National
Institutes of Health and is available on the Internet at
http://rsb.info.nih.gov/nih-image/.
J M020079D