Neuroprotective effects of N-alkyl-1,2,4-oxadiazolidine-3,5-diones and their corresponding synthetic intermediates N-alkylhydroxylamines and N-1-alkyl-3-carbonyl-1-hydroxyureas against in vitro cerebral ischemia

Article abstract of DOI:10.1002/cmdc.200900418

Herein we report the synthesis and neuroprotective effects of new N-alkyl-1,2,4-oxadiazolidine-3,5-diones and their corresponding synthetic intermediates, N-alkylhydroxylamines and N-1-alkyl-3-carbonyl-1-hydroxyureas, in an in vitro model of ischemia. We found five analogues that protect HT22 cells from death in the concentration range of 1-5 μm. Because members of the MAP kinase family are known to be key players in nerve cell survival and death, we characterized the role of these kinases in the neuroprotective mechanisms of the newly synthesized analogues. The results indicate that these compounds provide neuroprotection through distinct mechanisms of action.

Full text of DOI:10.1002/cmdc.200900418

DOI: 10.1002/cmdc.200900418
Neuroprotective Effects of N-Alkyl-1,2,4-oxadiazolidine-3,5-
diones and Their Corresponding Synthetic Intermediates
N-Alkylhydroxylamines and N-1-Alkyl-3-carbonyl-1-
hydroxyureas against in vitro Cerebral Ischemia
[a]
[a]
[b]
[a]
Alain Cesar Biraboneye, Sebastien Madonna, Pamela Maher, and Jean-Louis Kraus*
Herein we report the synthesis and neuroprotective effects of
new N-alkyl-1,2,4-oxadiazolidine-3,5-diones and their corre-
sponding synthetic intermediates, N-alkylhydroxylamines and
N-1-alkyl-3-carbonyl-1-hydroxyureas, in an in vitro model of
ischemia. We found five analogues that protect HT22 cells
from death in the concentration range of 1–5 mm. Because
members of the MAP kinase family are known to be key play-
ers in nerve cell survival and death, we characterized the role
of these kinases in the neuroprotective mechanisms of the
newly synthesized analogues. The results indicate that these
compounds provide neuroprotection through distinct mecha-
nisms of action.
Introduction
Epidemiological, clinical, and biochemical studies have shown
that various types of dietary fatty acids can modify the risk of
mia assay, as well as a different assay of oxidative-stress-in-
duced nerve cell death using the same cell line. We also
looked at potential neuroprotective signaling pathways.
[
1,2]
stroke.
This has led to an increased focus on the potential
neuroprotective activities of free fatty acids in pharmacological
research. Besides being key biomolecules in metabolic process-
es, free fatty acids serve as substrates for cell membrane bio-
genesis (glyco- and phospholipids) and as precursors of intra-
cellular signaling molecules such as prostaglandins, leuko-
trienes, thromboxanes, and platelet-activating factor. Polyunsa-
turated fatty acids have been implicated in the prevention of
various human diseases including obesity, diabetes, coronary
artery disease, stroke, and inflammatory and neurological dis-
The bioisosterism approach
A lead compound with a desired pharmacological activity may
have associated with it undesirable side effects, characteristics
that limit its bioavailability, or structural features that adversely
influence its metabolism and excretion from the body. Bioiso-
sterism represents an approach used by medicinal chemists for
the rational modification of lead compounds into safer and
more clinically effective agents.
[
3]
orders. Saturated fats are usually regarded as unhealthy, but
nutritionists believe that the type of saturated fat is also im-
portant. Stearic acid is biochemically classified as a saturated
fatty acid, both for the purposes of food labeling and dietary
recommendations. Stearic acid, one of the most common fatty
acids in brain phospholipids, originates in the circulation and is
sequestered from the blood by the brain along with precursor
[
6]
Carboxylate bioisosteres have been extensively studied. Of
these, those most frequently used as biomimetics are phos-
phonic, sulfonic, and tetrazole groups. For example, at physio-
logical pH, the tetrazole group has a similar degree of ioniza-
tion (pK 4.9) as carboxylic acid but is ~10-fold more lipophilic
a
[
4]
fatty acids. The neuroprotective effects of stearic acid against
the toxicity of oxygen or glucose deprivation, or of glutamate
(Figure 1). Another carboxylate bioisostere that is less studied
is the 1,2,4-oxadiazolidine-3,5-dione bioisostere group
(Figure 1). This carboxylic bioisostere group (X=O) is found in
quisqualic acid, a natural compound isolated from the seeds of
[
5]
toward rat cortical or hippocampal slices, have been report-
ed. The behavior of stearic acid is especially unique with re-
spect to its effects on serum cholesterol levels. A beneficial
effect of stearic acid toward clotting factors can result in a di-
minished thrombogenic state. In the present study, we first as-
sayed the neuroprotective properties of various major fatty
acids in our in vitro ischemia assay using the mouse hippocam-
pal cell line, HT22. Several fatty acids were assayed including li-
noleic, abietic, docosahexaenoic, and stearic acids. Using the
bioisosterism approach, we then synthesized new fatty acid bio-
isostere derivatives bearing different N-alkyl chains and evalu-
ated the neuroprotective properties of this new series of stea-
ric acid analogues and their intermediates in the in vitro ische-
[7]
[8]
various varieties of Quisqualis chinesis,
Q. indica,
and
[9]
Q. fructus.
[
a] A. C. Biraboneye, S. Madonna, Prof. J.-L. Kraus
Laboratoire de Chimie Biomolꢀculaire, CNRS
IBDML-UMR 6216, Campus de Luminy Case 907
Universitꢀ de la Mediterranꢀe
1
3288 Marseille Cedex 090 (France)
Fax: (+33)04 91 82 94 16
E-mail: kraus@univmed.fr
[
b] Dr. P. Maher
The Salk Institute for Biological Studies
10010 North Torrey Pines Road, La Jolla, CA 92037 (USA)
ChemMedChem 2010, 5, 79 – 85
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
79

MED
J.-L. Kraus et al.
Figure 1. Three non-classical carboxylic acid bioisosteres.
It turns out that the biological properties of quisqualamine
are very similar to those of the neurotransmitter g-aminobuty-
[
10]
ric acid (GABA; Figure 2). Quisqualic acid is an agonist for
Scheme 1. a) PCC, CH
c) NaBH CN, MeOH, RT; d) OCNCO
dioxane, RT.
2
Cl
2
, RT, 2 h; b) NH
2
OH, HCl, H
2
O, NaOH, RT;
, CH Cl , RT; e) NaOH,
3
2
C
2
H
5
or SCNCO
2
C
2
H
5
2
2
Figure 2. Quisqualamine (left) and GABA (right) bioisosteres.
both metabotropic glutamate receptors coupled to phosphoi-
6a, 6b, 6c, and the corresponding thiooxo derivative 10a
were obtained directly by stirring the urea or thiourea inter-
mediates in 3.5n NaOH for a few hours at room temperature.
[
11]
nosotide hydrolysis and the non-NMDA ionotropic a-amino-
3
-hydroxy-5-methylisoxazole-4-propionic acid (AMPA) recep-
[
12]
tor.
Compounds containing 1,2,4-oxadiazolidine-3,5-dione
moieties were also found to be active as anti-hyperglycemic
[
13]
Comparison of physicochemical properties of the N-alkyl-
agents.
1,2,4-oxadiazolidine-3,5-dione moiety and carboxylic acid
Starting from the hypothesis that the 1,2,4-oxadiazolidine-
,5-dione group, through its acidic properties, is very similar to
3
We considered pK (ionization state) and ClogP (hydrophobic-
a
that of carboxylic acid and therefore could have similar neuro-
protective effects to those of stearic acid, the synthesis and
biological activity of new N-alkyl-1,2,4-oxadiazolidine-3,5-dione
derivatives and their synthetic intermediates were investigated.
Three alkyl chain lengths were selected: hexyl, decyl, and
stearyl.
ity) as the two most important parameters, as it is well known
that these molecular properties can influence passive absorp-
tion of which solubility and permeability are the key factors.
Table 1 summarizes the pK and ClogP values for the stearyl-,
a
decyl-, and hexyl-3,5-dioxo-1,2,4-oxadiazolidine bioisosteres
and their corresponding tetrazole and carboxylic acid ana-
logues.
As can be seen, replacement of the carboxylic acid group by
a 3,5-dioxo-1,2,4-oxadiazolidine or tetrazole moiety does not
significantly change the acidic properties of the resulting bioi-
sosteres, and only slight differences in hydrophobicity are ob-
Results
Chemistry
The new 1,2,4-oxadiazolidine-3,5-dione carboxylic bioisosteres
6
a, 6b, and 6c were synthesized as depicted in Scheme 1.
Table 1. Comparison of the physicochemical properties of 3,5-dioxo-
Starting from commercially available aldehydes 2b and 2c or
prepared from stearyl alcohol 1a in the case of stearyl alde-
hyde 2a, the corresponding aldoximes 3a, 3b, and 3c were
obtained in good yields. The reductive step leading to the cor-
responding hydroxylamines 4a, 4b, and 4c was achieved
using sodium cyanoborohydride in methanol as a reductive re-
agent. The compounds were purified by column chromatogra-
phy and immediately used for the next step, as these hydroxyl-
amine intermediates are rather unstable. The hydroxylamines
reacted quite easily with ethyl isocyanatoformate to give the
corresponding 3-ethoxycarbonyl-1-hydroxy-1-alkylureas 5a,
1
,2,4-oxadiazolidines and those of the corresponding carboxylic acids.
[
a]
[b]
Compound
R
ClogP
pK
a
Ref.
6a
6
6
À(CH
2
)
16CH
3
7.98
4.27
2.69
8.27
4.03
2.45
9.87
ND
2.26
10.5
4.90
2.16
b
c
À(CH
À(CH
2
)
)
8
4
CH
CH
3
2
3
stearic acid
decanoic acid
hexanoic acid
[15]
[16]
[14]
5
b, and 5c in acceptable yields. The stearylhydroxylamine in-
[
a] ClogP values were calculated using ChemDraw software (Cambridge-
termediate 4a was also treated with ethyl isothiocyanatofor-
mate to give the corresponding 3-ethoxycarbonyl-1-hydroxy-1-
stearylthiourea 9a. The N-alkyl-1,2,4-oxadiazolidine-3,5-diones
Soft, Cambridge, MA, USA). [b] Values were determined by UV/Vis spec-
troscopy (Sirius Analytical Instruments Ltd., “Applications and Theory
Guide to pH-Metric pK and logP Determination”, 1993.
a
8
0
www.chemmedchem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 79 – 85

Neuroprotective Effects of N-Alkyl-1,2,4-oxadiazolidine-3,5-diones
served. For example, in the case of the stearyl-3,5-dioxo-1,2,4-
Neuroprotective effects of the new fatty acid bioisosteres and
their synthetic intermediates
oxadiazolidine, the ClogP and pK values for stearic acid and
a
its bioisostere 6a are quite similar: 8.27 and 7.98 for the ClogP
values, respectively, and 10.5 and 9.87 for the respective pKa
values.
The new fatty acid bioisostere analogues and their synthetic
intermediates were tested for their neuroprotective effects in
two different models of nerve cell death. The IAA model was
already described. The second model, used to test the neuro-
protective effects of the new analogues, was oxidative gluta-
mate toxicity. In this widely used model of oxidative-stress-in-
duced death, treatment of HT22 cells with 5 mm glutamate in-
duces cell death within 24 h via a well-characterized pathway
involving glutathione depletion and the production of reactive
Biological results
Neuroprotective effects of fatty acids
Before testing the newly synthesized derivatives, we first as-
sayed the neuroprotective properties of stearic acid versus
other known unsaturated and cyclic fatty acids (linoleic, abietic,
and docosahexaenoic) as reference fatty acids, using a cell-
based assay that reproduces a number of the pathophysiologi-
cal changes associated with ischemia. Indeed, the changes ob-
served in nerve cells following treatment with iodoacetic acid
[19]
[20]
oxygen species. The flavonoid fisetin, which was previous-
[17]
ly reported to be neuroprotective in this assay, was used as
reference compound for comparison with the new analogues.
The results obtained with the HT22/IAA assay as well as the
glutamate toxicity assay are listed in Table 2.
Having identified new compounds, primarily 4a and 4b,
with neuroprotective effects, the question of their mechanism
of action became an important issue. We first investigated the
(
IAA) are very similar to the changes that have been observed
in the brain in animal models of ischemia. This primary assay
[
17]
was recently described by Maher et al. and uses the mouse
HT22 hippocampal nerve cell line in combination with IAA to
induce ischemia in vitro. IAA is a known irreversible glyceralde-
[21]
p38 MAPK and JNK pathways. Using specific antibodies to
the phosphorylated forms of JNK and p38 MAPK as well as
their specific substrates, the effects of some of the most active
compounds (4a, 4c, 4b, 5b, and 6a) on these pathways were
examined. We also looked at the effects of these compounds
on ERK activation, because this pathway is implicated in neuro-
protection. The results are listed in Table 3.
[
18]
hyde-3-phosphate dehydrogenase inhibitor
that induces
[
17]
chemical ischemia. Treatment of HT22 cells with IAA induced
a dose-dependent increase in cell death 20 h post-treatment
with <5% survival at 20 mm. Figure 3 shows the dose–re-
Discussion
From these results it is clear that the majority of the com-
pounds—bioisosteres 3,5-dioxo-1,2,4-oxadiazolidines 6a, 6b,
and 6c, the thiodiazoline 10a, as well as the synthetic inter-
mediates 5a, 5b, 5c, 4a, 4b, 4c, 3a, 3b, and 3c—show neu-
roprotective effects at concentrations ranging from 1.0 to
20 mm. The most active in our screening model are 4b and 4a,
with respective EC50 values of 1.0 and 2 mm. Both 4a and 4b
contain a hydroxylamine function linked to a linear stearyl or
decyl alkyl chain. Surprisingly, replacement of the carboxylic
acid function by a hydroxylamine group appears to greatly en-
hance the observed neuroprotective effect, as these analogues
have EC50 values better than that of stearic acid (EC =6.3 mm).
50
Figure 3. Dose–response curves for neuroprotection from IAA toxicity: repre-
sentative HT22 cell protection curves for fatty acids after IAA injury. HT22
cells were treated with 20 mm IAA for 2 h alone or in the presence of increas-
ing doses of linoleic, stearic, abietic, or docosahexaenoic acid. The same con-
centrations of fatty acids were also included in the fresh medium added
after the 2 h treatment with IAA. Percent survival was measured after 24 h
by MTT assay. Similar results were obtained in 3–5 independent experi-
ments.
We emphasize this point because, to our knowledge, the neu-
roprotective effects of N-alkylhydroxylamine compounds have
never been reported. Because N-hydroxylamine compounds
are known to be relatively unstable and sensitive to oxidation,
their potential use as chemotherapeutic agents could be limit-
ed. However, it is possible to stabilize such compounds
[22]
through the formation of their corresponding acid salts.
The data also indicate that the octadecyl (stearyl) alkyl chain
is not required for the neuroprotective effect. Interestingly,
compounds 4a and 4b have a potency similar to that of the
flavonoid fisetin, which was previously reported to be neuro-
sponse curves for four fatty acids: linoleic, abietic, docosahex-
aenoic, and stearic acids. As can be seen, the maximal neuro-
protective effects of the acids never exceeded 50% cell surviv-
al and were observed at concentrations between 1 and 2 mm.
In contrast to the other fatty acids, which all gave bell-shaped
curves, abietic acid gave a more typical neuroprotection curve
with a distinct plateau.
[17]
protective in the IAA toxicity assay; however, whereas fisetin
provided nearly 100% protection, the other analogues (with
the exception of compound 4b, which provided 83% protec-
tion) only protected the cells between 50 and 76%.
ChemMedChem 2010, 5, 79 – 85
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
81

MED
J.-L. Kraus et al.
that the electronic parameters
of the hydrophilic moiety (sulfur
being more nucleophilic and
bulkier than oxygen) greatly in-
fluence the neuroprotective
effect. The results of the reac-
tion mechanism studies report-
ed in Table 3 suggest that ana-
logues 4a, 4b, 4c, 5b, and 6a
exert their neuroprotective ef-
fects through distinct mecha-
nisms, as 4a, 4b, 4c, and 5b
induce ERK activation, while 6a
has no effect. Moreover, none
of the tested analogues affect
JNK or p38 MAP kinase activa-
tion; this indicates that all these
analogues induce neuroprotec-
tive effects through pathways
that are at least partially distinct
from the reference compound
fisetin and more similar to that
of the GM1-sialoganglioside an-
Table 2. Protection of HT22 cells from IAA and glutamate toxicity by fatty acid bioisosteres.
[
c]
Compound
EC50 [mm]
HT22/IAA
Max. Pro-
tection [%]
Toxicity [%]
HT22/Glu-
tamate [%]
JNK/p38 MAPK
Activation
ERK
Activation
[
a]
[b]
[d]
[e]
[f]
6
1
6
6
5
5
5
4
4
4
3
3
3
a
6.0
18.6
6.0
8.1
8.9
6.0
ND
2
1.0
2.7
8.9
5.8
ND
6.3
2.8
60
37
40
50
50
62
ND
62
83
76
49
62
ND
41
95
0
6
4
7
11
6
ND
50
3
9.9
7.7
8.7
9.5
12.8
9.6
ND
13.5
62.0
9.2
ND
9.4
ND
ND
95
No effect
ND
ND
No effect
ND
ND
0a (thio)
b
c
a
b
c
a
b
c
ND
ND
No effect
No effect
ND
No effect
No effect
No effect
ND
No effect
ND
ND
No effect
Increases
ND
Increases
Increases
Increases
ND
No effect
ND
ND
6
a
b
c
61
12
ND
16
15
Stearic acid
Fisetin
Decreases
Increases
[a] Half-maximal effective concentrations for protection from IAA toxicity were determined by exposing HT22
cells to various concentrations of each analogue in the presence of 20 mm IAA for 2 h; cell viability was deter-
mined after 24 h by the MTT assay. [b] The maximal percent survival at the most effective concentration (which
varied from 1 to 10 mm depending on the analogue). [c] HT22 cells were exposed to each analogue at 10 mm,
and cell viability was determined after 24 h by MTT assay. [d] HT22 cells were exposed to various concentra-
tions of each analogue in the presence of 5 mm glutamate for 24 h; cell viability was determined after 24 h by
MTT assay, and the maximal percent survival at the most effective dose (which varied from 1 to 10 mm depend-
ing on the analogue) was determined; the average percent survival for cells treated with glutamate alone was
1
5Æ5%. [e] The effect of the analogues on JNK and p38 MAPK activation by IAA treatment was determined by
alogues
scribed.
we
recently
de-
[
23]
SDS-PAGE and western blot with phospho-specific antibodies as described. [f] The effect of the analogues on
ERK activation in the presence of IAA was determined by SDS-PAGE and western blot with phospho-specific
antibodies as described.
[23]
[
23]
Together these data strongly
suggest that the new analogues
4
a, 4c, 4b, 5b, and 6a do not
act as antioxidants, but rather
as inducers of antioxidant and/
Table 3. Effect of analogues 4a, 4c, 4b, 5b, and 6a on JNK, p38 MAP
kinase, and ERK activation.
or neuroprotective protein synthesis. This conclusion is sup-
ported by their complete lack of activity in the Trolox equiva-
lent antioxidant capacity (TEAC) assay, an in vitro assay for anti-
[
a]
[b]
Compound
JNK/p38 MAPK Activation
ERK Activation
[24]
4
4
4
5
6
a
b
c
b
a
No effect
No effect
No effect
No effect
No effect
Decreases
Increases
Increases
Increases
Increases
No effect
Increases
oxidant activity (data not shown). Although their mechanism
of action has yet to be fully elucidated, our preliminary experi-
ments suggest that the whole group of newly designed com-
pounds does not provide neuroprotection through the same
mechanism as the reference compound fisetin, because they
have no effect on JNK or p38 MAP kinase activity. However, we
did observe that, similar to fisetin analogues, 4a, 4b, 4c, and
5b, but not analogue 6a activate ERK kinase. In summary, we
have successfully synthesized a series of new N-alkyl-1,2,4- oxa-
diazolidine-3,5-diones bearing alkyl chains of varying lengths
Fisetin
[a] Effects of new analogues on JNK and p38 MAP kinase activation: HT22
cells were untreated or treated with 10 mm 4a, 4b, 4c, 5b, or 6a in the
absence or presence of 20 mm IAA for 2 h. Following recovery for 2 h, cell
lysates were prepared, and equal amounts of protein were analyzed by
SDS-PAGE and immunoblotting with antibodies to phospho- and total
JNK and p38 MAP kinase. [b] Effects of new analogues on ERK phosphory-
lation: HT22 cells were untreated or treated with 10 mm 4a, 4b, 4c, 5b,
or 6a in the absence or presence of 20 mm IAA for 2 h for comparative
ERK activation estimation. Following recovery for 2 h, cell lysates were
prepared, and equal amounts of protein were analyzed by SDS-PAGE and
immunoblotting with antibodies to phospho- and total ERK.
(
hexyl, decyl, and stearyl), and, along with their synthetic inter-
mediates, we evaluated their neuroprotective effects. Some of
the new analogues exhibit potent neuroprotective effects
toward nerve cells exposed to ischemia-like insults. Analogues
4
a, 4c, 4b, and 5b are of particular interest, as they are very
efficacious, with EC50 values ranging between 1 and 2 mm.
Moreover, although the mechanism by which these new ana-
logues exert their neuroprotective effect remains unclear, it ap-
pears to be distinct from that of the reference compound fise-
tin. Given these potent effects in cell-based assays, further ex-
ploration of the neuroprotective effects of analogues in animal
models of ischemia is clearly warranted.
Notably, the acidic properties of the compounds (pK values
a
reported in Table 1) weakly influence the observed neuropro-
tective effect. The neuroprotective potency of compound 6a
(
pK 9.87), a true bioisostere of stearic acid (pK 10.5), is slightly
a a
higher than that of stearic acid. Replacement of the oxygen
atom in bioisostere 6a by a sulfur atom in 10a lowered the
observed neuroprotective effect by a factor of three, indicating
8
2
www.chemmedchem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 79 – 85

Neuroprotective Effects of N-Alkyl-1,2,4-oxadiazolidine-3,5-diones
organic phase gave the corresponding hydroxylamine, which was
Experimental Section
used either with or without purification: oily compound (62%);
1
Chemistry
Rf =0.3 (cHex/EtOAc 2:1); H NMR (250 MHz, CDCl ): d=2.87 (m,
3
2
H, -CH CHN (OH)), 1.50 (m, 2H, -CH CH CHN (OH)), 1.24 (m, 28H,
2 2 2
General methods: Starting materials and reagents were obtained
from commercial suppliers and were used without purification.
THF was distilled over sodium benzophenone ketyl immediately
prior to use; CH Cl was distilled over P O5 just prior to use.
13
-(CH ) -), 0.83 ppm (m, 3H, -CH ); C NMR (250 MHz, CDCl ): d=
2 14 3 3
4
2
4
4.5, 31.8, 29.7, 29.5 (10C), 29.3, 27.2, 22.7, 14.1 ppm; ESIMS: m/z
+
86 [M+H] , 100%; Anal. calcd for C H NO: C 75.72, H 13.70, N
18
39
2
2
2
.91, O 5.60.
1
H NMR spectra were recorded at 250 MHz on a Bruker AC-250
spectrometer. Elemental analyses were within 0.4% of theoretical
values for all compounds. Chemical shifts (d) are expressed ppm
downfield from (CH ) Si. ESIMS data were obtained on a Waters Mi-
N-decylhydroxylamine (4b): According to general procedure A,
the reaction of (E)-decanaloxime (200 mg, 1.1 mmol) with NaBH CN
3
3
4
(293 mg, 4.6 mmol) afforded the N-decylhydroxylamine as a white
1
cromass ZMD spectrometer by direct injection of the sample solu-
solid (180 mg, 94%): R =0.16 (cHex/EtOAc 3:1); H NMR (250 MHz,
f
bilized in CH CN. Analytical and preparative TLC was performed
CDCl ): d=2.92 (m, 2H, -CH CHNOH), 1.52 (m, 2H, -CH CH CHNOH),
1.48 (m, 12H, -(CH ) -), 0.87 ppm (m, 3H, -CH ); C NMR (250 MHz,
2 6 3
3
3
2
2
2
13
with silica gel plates 0.2 and 1 mm thick, respectively (60 F254
,
Merck). Preparative flash column chromatography was carried out
on silica gel (230–240 mesh, G60, Merck). Compound purity was
determined by elemental analysis.
CDCl ): d=44.6, 31.8, 29.3 (2C), 29.2 (2C), 27.0, 26.2, 22.6,
14.1 ppm; ESIMS: m/z 174 [M+H] , 100%; Anal. calcd for
3
+
C H NO: C 69.34, H 13.38, N 8.08, O 9.23.
10
23
Adehydes 2b and 2c are commercially available compounds,
whereas stearaldehyde 2a was prepared as follows: Pyridinium
chlorochromate (PCC; 1 g, 5.4 mmol) in anhydrous CH C1 (10 mL)
N-hexylhydroxylamine (4c): According to general procedure A,
the reaction of (E)-hexanaloxime (1.43 g, 12.1 mmol) with NaBH CN
3
2
2
(0.936 g, 14.8 mmol) afforded the N-hexylhydroxylamine as a white
1
was suspended in a 500-mL round-bottomed flask fitted with a
reflux condenser. 1-Stearol (1 g, 3.6 mmol) in CH Cl (20 mL) was
solid (828 mg, 55%): R =0.37 (cHex/EtOAc 2:1); H NMR (250 MHz,
f
2
2
CDCl3): d=2.87 (m, 2H, -CH CHN(OH)), 1.50 (m, 2H,
2
added in one portion to the magnetically stirred solution. After
h, dry Et O (15 mL) was added, and the supernatant was de-
canted. The insoluble residue was washed thoroughly with anhy-
-CH CH CHN(OH)), 1.24 (m, 6H, -(CH ) -), 0.83 ppm (m, 3H, -CH );
2
2
2 3
3
13
2
C NMR (250 MHz, CDCl ): d=44.5, 31.5, 26.5, 26.7, 22.7, 14.0 ppm;
3
2
+
ESIMS: m/z 118 [M+H] , 100%; Anal. calcd for C H N O : C 61.49,
14 28 2 4
drous Et O (3ꢁ10 mL), whereupon it became a black granular
H 12.90, N 11.95, O 13.65.
2
solid. The combined organic solution was dried, and the solvent
was evaporated. A solid white compound was obtained (808 mg,
General procedure B
1
8
3%): R =0.8 (cHex/EtOAc 9:1); H NMR (250 MHz, CDCl ): d=9.73
f
3
(
1
s, 1H, C(O)H), 2.40 (m, 2H, -CH C(O)), 1.57 (s, 2H, -CH CH C(O)),
.22 (m, 28H, -(CH ) -), 0.96 ppm (m, 3H, -CH ); C NMR (250 MHz,
2 14 3
N-octadecylhydroxylamine ethyl formylcarbamate (5a): Hydrox-
ylamine-N-octadecylhydroxylamine (0.24 g, 0.8 mmol) was dis-
2
2
2
13
CDCl ): d=203.6, 44.0, 32.0, 29.8 (2), 29.5, 22.8, 22.2, 21.4,
solved in anhydrous CH Cl (10 mL). Ethyl isocyanotoformate
2 2
3
+
1
4.3 ppm; ESIMS: m/z 269 [M+H] .
(0.098 g, 0.8 mmol) was added, and the reaction mixture was main-
tained at room temperature for 6 h. The solution was evaporated
under reduced pressure, EtOAc was added (10 mL), and the residue
Oximes: Oximes 3b and 3c are known compounds prepared from
the corresponding commercially available aldehydes according to
was washed with H O (3ꢁ10 mL). The organic phase was dried
[25]
2
published procedures. Stearaldehyde oxime 3a was prepared as
follows: solution of hydroxylamine hydrochloride (1 g,
4.5 mmol) in H O (10 mL) was added to a solution of stearalde-
over MgSO , and after filtration and evaporation under reduced
4
A
pressure, silica gel chromatography (cHex/EtOAc 2:1) of the residue
1
2
gave a transparent oil (62 mg, 18%): R =0.5 (cHex/EtOAc 2:1);
f
hyde (0.8 g, 2.9 mmol) in EtOH (10 mL), followed by the addition of
0% NaOH until a precipitate was formed. The reaction mixture
1
H NMR (250 MHz, CDCl ): d=9.36 (s, 1H, NH), 8.09 (s, 1H, NOH),
3
1
4
.25 (q, J=7.12 Hz, 2H, CH CH OC(O)), 3.68 (t, J=7.5 Hz, 2H,
3 2
was heated (T=708C) for 2 h, cooled to room temperature, and
-
CH N(OH)), 1.69 (m, 2H, -CH CH N(OH)), 1.24 (m, 31H, -(CH ) - and
2 2 2 2 14
the resulting white precipitate was filtered, washed with cold H O
13
2
CH CH OC(O)), 0.83 ppm (m, 3H, -CH ); C NMR (250 MHz, CDCl3):
3
2
3
and dried. A white solid was obtained in quantitative yield: R =
f
d=153.5, 151.5, 58.8, 52.5, 31.5, 30.2, 29.8 (13C), 26.6, 23.7, 22.4,
1
0
.87 (cHex/EtOAc 3:1); H NMR (250 MHz, CDCl ): d=7.43 (m, 1H,
+
3
1
4.5 ppm; ESIMS: m/z 401 [M+H] , 100%; Anal. calcd for
-CH=N(OH), trans), 6.93 (m, 1H, -CH=NOH, cis), 2.44 (m, 2H,
C H N O : C 65.96, H 11.14, N 6.99, O 15.98.
22
44
2
4
-
-
CH CH=N(OH)), 1.50 (m, 2H, CH CH CH=N(OH)), 1.29 (m, 28H,
2
2
1
2
3
(CH ) ), 0.87 ppm (m, 3H, -CH ); C NMR (250 MHz, CDCl ): d=
Ethyl 2-(3-hydroxy-3-decylureido)acetate (5b): According to gen-
eral procedure B, the reaction of N-decylhydroxylamine (100 mg,
0.5 mmol) with ethyl isocyanatoformate (66 mg, 0.5 mmol) afford-
2
14
3
3
1
2
4
52.5, 31.8, 30.2, 29.8 (12C), 25.2, 22.2, 21.4, 14.5 ppm; ESIMS: m/z
+
84 [M+H] , 100%; Anal. calcd for C H NO: C 76.32, H 13.14, N
18
37
.94, O 5.64.
ed 5b as a white solid in quantitative yield: R
f
=0.8 (EtOAc);
1
H NMR (250 MHz, CDCl ): d=8.71 (s, 1H, NH), 8.52 (s, 1H, NOH),
3
4
.18 (q, J=6.8 Hz, 2H, CH CH OC(O)), 3.49 (t, J=7.1 Hz, 2H,
3 2
Hydroxylamines 4a, 4b, and 4c
General procedure A
-CH NOH), 1.61 (m, 2H, -CH CH NOH), 1.25 (m, 17H, -(CH ) - and
2 2 2 2 7
13
CH CH OC(O)), 0.86 ppm (m, 3H, -CH ); C NMR (250 MHz, CDCl3):
3
2
3
d=153.5, 151.8, 61.6, 51.5, 48.5, 31.6, 29.3 (2), 29.1, 26.4, 22.4, 21.4,
N-octadecylhydroxylamine (4a): A solution of NaBH CN (133 mg,
+
3
1
3.9 ppm; ESIMS: m/z 289 [M+H] , 100%; Anal. calcd for
2
.2 mmol) in MeOH (2 mL) was added with concurrent dropwise
C H N O : C 58.31, H 9.79, N 9.71, O 22.19.
14
28
2
4
addition of aqueous 6n HCl/MeOH 1:1 to a stirred solution of the
oxime (500 mg, 1.7 mmol) in MeOH (10 mL) at À608C. The mixture
was then allowed to warm to À208C over 2 h while maintaining
pH 3. After evaporation the entire workup was carried out at 08C:
addition of saturated aqueous NaCl and basification with 6n KOH.
Extraction with Et O, drying over MgSO , and evaporation of the
Ethyl 2-(3-hexylureido)acetate (5c): According to general proce-
dure B, the reaction of N-hexylhydroxylamine (0.8 g, 0.5 mmol)
with ethyl isocyanatoacetate (0.87 g, 0.5 mmol) afforded 5c as a
1
white solid in quantitative yield: Rf =0.8 (EtOAc); H NMR
(250 MHz, CDCl ): d=7.94 (s, 1H, NH), 6.47 (s, 1H, NOH), 4.14 (m,
2
4
3
ChemMedChem 2010, 5, 79 – 85
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
83

MED
J.-L. Kraus et al.
2
H, CH CH OC(O)), 3.90 (m, 2H, -OC(O)CH NH), 3.44 (m, 2H,
2H, -CH CH N(OH)), 1.29 (m, 30H, -(CH ) -), 0.88 (m, 3H, -CH );
3
2
2
2
2
2 15
3
13
-CH NOH), 1.56 (m, 2H, -CH CH NOH), 1.25 (m, 9H, -(CH ) - and
C NMR (250 MHz, CDCl ): d=183.3, 159.2, 48.7, 31.5, 29.5 (12C),
3
2
2
2
2 3
13
+
CH CH OC(O)), 0.83 ppm (m, 3H, -CH ); C NMR (250 MHz, CDCl3):
26.3, 23.5, 22.7, 14.2 ppm; ESIMS: m/z 371 [M+H] , 100%; Anal.
3
2
3
d=169.5, 156.5, 61.0, 52.8, 44.5, 31.2, 26.3, 23.7, 22.7, 14.1 ppm;
calcd for C H N O S: C 64.82, H 10.30, N 7.56, O 8.54, S 8.65.
20
38
2
2
+
ESIMS: m/z 247 [M+H] , 100%; Anal. calcd for C H N O : C 53.31,
11
22
2
4
H 9.09, N 11.31, O 25.99.
Biology
General procedure C
Cell culture: Fetal calf serum (FCS) and dialyzed FCS (DFCS) were
obtained from Hyclone (Logan, UT, USA). Dulbecco’s modified
Eagle’s medium (DMEM) was purchased from Invitrogen (Carlsbad,
CA, USA). HT22 cells were grown in DMEM supplemented with
2
4
-octadecyl-1,2,4-oxadiazolidine-3,5-dione (6a): NaOH (3.5n,
mL) was added slowly under stirring to a solution of 5a (0.062 g,
0
0
.15 mmol) in dioxane (4 mL) at 08C. The mixture was stirred at
8C for 15 min, and then at room temperature for 7 h. The solvent
1
0% FCS and antibiotics.
was evaporated, the residue was taken up in H O, and the aqueous
solution was acidified to pH 5 with 1n HCl. The solution was ex-
2
Cytotoxicity assays: Cell viability was determined with a modified
version of the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium
bromide (MTT) assay based on the standard procedure. Cells were
tracted with EtOAc, the organic layers were dried over MgSO , and
4
the solvent was evaporated to yield a white solid (34 mg, 62%):
3
seeded onto 96-well microtiter plates at a density of 5ꢁ10 cells
1
Rf =0.5 (EtOAc); H NMR (250 MHz, CDCl ): d=3.68 (t, J=7.5 Hz,
3
per well. For the in vitro ischemia assay, the next day, the medium
was replaced with DMEM supplemented with 7.5% DFCS, and the
cells were treated with iodoacetic acid (IAA; 20 mm) alone or in the
presence of test compounds. After 2 h, the medium in each well
was aspirated and replaced with fresh medium without IAA, but
containing the compounds to be screened. After 20 h, the medium
in each well was aspirated and replaced with fresh medium con-
2
H, -CH N(OH)), 1.70 (m, 2H, -CH CH N(OH)), 1.24 (m, 30H, -(CH ) -
2 2 2 2 15
13
)
, 0.83 ppm (m, 3H, -CH3); C NMR (250 MHz, CDCl ): d=157.3,
3
1
3
7
53.2, 48.7, 31.5, 29.5 (12C), 26.3, 23.5, 22.7, 14.2 ppm; ESIMS: m/z
+
55 [M+H] , 100%; Anal. calcd for C H N O : C 67.72, H 10.80, N
20
38
2
3
.86, O 13.54.
2
-decyl-1,2,4-oxadiazolidine-3,5-dione (6b): According to general
À1
taining 2.5 mgmL MTT. After incubation for 4 h at 378C, the cells
procedure C, the reaction of 5b (100 mg, 0.34 mmol) with NaOH
10 mL) afforded the 2-decyl-1,2,4-oxadiazolidine-3,5-dione as a
were solubilized with a solution containing 50% N,N-dimethylfor-
mamide (DMF) and 20% SDS (100 mL, pH 4.7). For the oxidative
glutamate toxicity assay, the next day, the medium was replaced
with DMEM supplemented with 7.5% DFCS, and the cells were
treated with the test compounds alone or in the presence of 5 mm
glutamate. After 24 h, the medium in each well was aspirated and
(
1
white solid (48 mg, 58%): R =0.6 (EtOAc); H NMR (250 MHz,
f
CDCl ): d=9.58 (s, 1H, NH), 3.66 (t, J=7.1 Hz, 2H, -CH NO), 1.70 (m,
3
2
2
0
4
H, -CH CH NO), 1.25 (m, 14H, -(CH ) - and CH CH OC(O)),
2 2 2 7 3 2
13
.87 ppm (m, 3H, -CH3); C NMR (250 MHz, CDCl ): d=153.5, 151.8,
3
8.5, 31.6, 29.3 (2), 29.1, 26.4, 22.4, 21.4, 13.9 ppm; ESIMS: m/z 243
À1
+
replaced with fresh medium containing 2.5 mgmL MTT. After in-
[M+H] , 100%; Anal. calcd for C H N O : C 59.48, H 9.15, N 11.56,
12
22
2
3
cubation for 4 h at 37 C8, the cells were solubilized with a solution
containing 50% DMF and 20% SDS (100 mL, pH 4.7). For both
assays, the absorbance at l 570 nm was measured on the follow-
ing day with a microplate reader (Molecular Devices). Results were
confirmed by visual inspection of the wells. Controls included com-
pound alone to test for toxicity and compound with no cells to
test for interference with the assay chemistry.
O 19.81.
2
-hexyl-1,2,4-oxadiazolidine-3,5-dione (6c): According to general
procedure C, the reaction of 5c (1.8 g, 0.3 mmol) with 3.5m NaOH
10 mL) afforded the 2-hexyl-1,2,4-oxadiazolidine-3,5-dione as a
(
1
transparent oil in quantitative yield: R =0.6 (EtOAc); H NMR
f
(
250 MHz, CDCl ): d=9.77 (s, 1H, NH), 3.66 (t, J=6.9 Hz 2H, -CH N),
3 2
1
.69 (m, 2H, -CH CH N), 1.31 (m, 6H, -(CH ) -), 0.88 (m, 3H, -CH );
2 2 2 3 3
C NMR (250 MHz, CDCl ): d=157.5, 153.9, 48.9, 31.7, 26.2, 23.6,
1
3
SDS-PAGE and immunoblotting: HT22 cells from the same density
cultures as used for the cell death assays were untreated or treated
with the analogues alone or in the presence of 20 mm IAA for 2 h
followed by 2 h of recovery in the presence of the analogues. The
cells were washed twice in phosphate-buffered saline (PBS) and
solubilized in SDS sample buffer containing 0.1 mm Na VO and
3
+
2
2.7, 14.2 ppm; ESIMS: m/z 187 [M+H] , 100%; Anal. calcd for
C H N O : C 51.60, H 7.58, N 15.01, O 25.78.
8
14
2
3
Thio derivatives
3
4
1
mm phenylmethylsulfonyl fluoride (PMSF), boiled for 5 min, and
N-octadecylhydroxylamine ethyl thioformylcarbamate (9a): Ac-
cording to general procedure B, the reaction of N-octadecylhydrox-
ylamine (44 mg, 0.15 mmol) with ethoxycarbonyl isothiocyanate
either analyzed immediately or stored frozen at À708C. Proteins
were separated on 10% SDS-polyacrylamide gels and transferred
to nitrocellulose. Uniform loading and transfer of the samples were
confirmed by staining the nitrocellulose with Ponceau-S. Transfers
were blocked for 1 h at room temperature with 5% nonfat milk in
TBS/0.1% Tween-20 and then incubated overnight at 48C in the
primary antibody diluted in 5% bovine serum albumin (BSA) in
TBS/0.05% Tween-20. The primary antibodies used were phospho-
p44/42 MAP kinase antibody (no. 9101, 1:1000), phospho-p38 MAP
kinase (no. 9211, 1:1000), phospho-SAPK/JNK (no. 9255, 1:1000),
phospho c-Jun (no. 9261, 1:1000), phospho-MAPKAPK-2 antibody
(
30 mg, 0.23 mmol) afforded compound 9a as a yellow oil in quan-
1
titative yield: R =0.8 (EtOAc); H NMR (250 MHz, CDCl ): d=9.69 (s,
f
3
1
3
H, NH), 7.98 (s, 1H, NOH), 4.20 (q, J=7.12 Hz, 2H, CH CH OC(O)),
3 2
.61 (t, J=7.5 Hz, 2H, -CH N(OH)), 1.69 (m, 2H, -CH CH N(OH)), 1.24
2
2
2
(
m, 31H, -(CH ) - and CH CH OC(O)), 0.83 ppm (m, 3H, -CH );
2 14 3 2 3
1
3
C NMR (250 MHz, CDCl ): d=183.2, 153.4, 59.1, 57.5, 31.8, 30.2,
3
+
2
,
9.7 (12C), 27.6, 24.0, 22.7, 14.5, 13.8 ppm; ESIMS: m/z 417 [M+H]
100%; Anal. calcd for C H N O S: C 63.42, H 10.14, N 6.72, O
22
44
2
3
11.52, S 7.70.
(no. 3041, 1:1000), and total MAPKAPK-2 (no. 3042, 1:1000) from
2
-octadecyl-3-thioxo-1,2,4-oxadiazolidin-5-one (10a): According
Cell Signaling (Beverly, MA, USA); total p38 antibody (no. sc-728,
1:500) and c-Jun/AP-1 antibody (no. sc-44-G) from Santa Cruz Bio-
technology (Santa Cruz, CA, USA); JNK1 antibody (no. 15701A,
1:500) from Pharmingen; and pan-ERK antibody (no. E17120,
1:10000) from Transduction Laboratories (San Diego, CA, USA). The
to general procedure C, the reaction of 9a (0.67 g, 0.16 mmol) with
NaOH (10 mL) afforded the 2-octadecyl-3-thioxo-1,2,4-oxadiazoli-
din-5-one as a white solid (28 mg, 47%): R =0.6 (EtOAc); H NMR
1
f
(
250 MHz, CDCl ): d=3.70 (t, J=7.5 Hz, 2H, -CH N(OH)), 1.75 (m,
3
2
8
4
www.chemmedchem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 79 – 85

Neuroprotective Effects of N-Alkyl-1,2,4-oxadiazolidine-3,5-diones
transfers were rinsed with TBS/0.05% Tween-20 and incubated for
[7] Y. C. Tuan, Yao Hsueh Pao 1957, 5, 87–89.
[
[
8] S.-T. Fang, Hua Huseh Hsueh Pao 1964, 30, 226–229.
9] T. Takemoto, N. Takagi, T. Nakajima, K. Koiki, Yakugaku Zasshi 1975, 95,
1
h at room temperature in horseradish peroxidase (HRP)-conjugat-
ed goat anti-rabbit or goat anti-mouse (Biorad, Hercules, CA, USA)
diluted 1:5000 in 5% nonfat milk in TBS/0.1% Tween-20. The im-
munoblots were developed with the Super Signal reagent (Pierce,
Rockford, IL, USA).
176–179.
[
[
[
10] R. H. Evans, A. A. Francis, K. Hunt, M. R. Martin, J. C. Watkins, J. Pharm.
Pharmacol. 1978, 30, 364–367.
11] L. Littman, L. A. Chase, M. Renzi, A. B. Garlin, J. F. Koerner, R. L. Johnson,
M. B. Robinson, Neuropharmacology 1995, 34, 829–841.
12] W. A. Kinney, N. E. Lee, D. T. Garrison, E. J. Podlesny, J. T. Simmonds, Jr.,
D. Bramlett, R. R. Notvest, D. M. Kowal, R. P. Tasse, J. Med. Chem. 1992,
Acknowledgements
3
5, 4720–4726.
[
[
13] M. S. Malamas, J. Sredy, M. McCaleb, I. Gunawan, B. Mihan, D. Sullivan,
Eur. J. Med. Chem. 2001, 36, 31–42.
14] P. Pernice, M. Castaing, P. Mꢂnassa, J.-L. Kraus, Biophys. Chem. 1988,
IBDML-CNRS (Institut de Biologie du Developpement de Marseille
Luminy, France) is gratefully acknowledged for financial support.
This work was supported in part by an American Heart Associa-
tion grant to P.M. We thank Dr. Valerie Monnier (Mass Spectrom-
etry Unit, Spectropole Universitꢀ Paul Cezanne, Marseille, France)
for helpful discussions.
3
2,15–20.
[15] J. R. Kanicky, D. Shah, J. Coll. Interface Sci. 2002, 256, 201–207.
[
[
16] Food Preservatives (Eds.: N. J. Russell, G. W. Gould), Springer, Heidelberg,
003, pp. 1–403.
17] P. Maher, K. F. Salgado, J. A. Zivin, P. A. Lapchak, Brain Res. 2007, 1173,
17–125.
2
1
[
[
[
[
[
[
18] B. S. Winkler, M. W. Sauer, C. A. Starnes, Exp. Eye Res. 2003, 76, 715–723.
19] S. Tan, D. Schubert, P. Maher, Curr. Top. Med. Chem. 2001, 1, 497–506.
20] Y. Sagara, J. Vanhnasy, P. Maher, J. Neurochem. 2004, 90, 1144–1155.
21] S. L. Mehta, N. Manhas, R. Rahubir, Brain Res. Rev. 2007, 54, 34–66.
22] J. Blixt (Sçdertꢃlje, Sweden), Patent Application 09/380458, 1999.
23] A. C. Biraboneye, S. Madonna, Y. Laras, S. Krantic, P. Maher, J.-L. Kraus, J.
Med. Chem. 2009, 52, 4358–4369.
Keywords: bioisosteres · ischemia · N-Alkylhydroxylamines ·
neuroprotection · oxadiazolidine-3,5-diones
[
[
1] M. J. McArthur, B. P. Atshaves, A. Frolov, W. D. Foxworth, A. B. Kier, F.
Schroeder, J. Lipid Res. 1999, 40, 1371–1383.
2] P. Needleman, A. Raz, M. S. Minkes, J. A. Ferrendelli, H. Sprecher, Proc.
Natl. Acad. Sci. USA 1979, 76, 944–948.
[24] K. L. Wolfe, R. H. Liu, J. Agric. Food Chem. 2007, 52, 8890–8907.
25] Organic Synthesis Collective, Vol. II, Wiley, New York, 1943 p. 313.
[
[
3] N. G. Bazan, Mol. Neurobiol. 2005, 32, 89–103.
4] T. Tholstrup, P. Marckmann, J. Jespersen, B. Sandstrom, Am. J. Clin. Nutr.
[
1
994, 59, 371–377.
5] Z. J. Wang, G. M. Li, W. L. Tang, M. Yin, Acta Pharmacol. Sin. 2006, 27,
45–150.
6] G. A. Patani, E. J. LaVoie, Chem. Rev. 1996, 96, 3147–3176.
[
[
Received: October 7, 2009
Revised: November 2, 2009
Published online on November 26, 2009
1
ChemMedChem 2010, 5, 79 – 85
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
85