Synthesis and biological evaluation of ceramide analogues with substituted aromatic rings or an allylic fluoride in the sphingoid moiety

Article abstract of DOI:10.1021/jm000939v

The biological activity of synthetic ceramide analogues, having modified sphingoid and N-acyl chains, as well as fluorine substituents in the allylic position, was investigated in hippocampal neurons. Their influence on axonal growth was compared to that

Full text of DOI:10.1021/jm000939v

J . Med. Chem. 2000, 43, 4189-4199
4189
Syn th esis a n d Biologica l Eva lu a tion of Cer a m id e An a logu es w ith Su bstitu ted
Ar om a tic Rin gs or a n Allylic F lu or id e in th e Sp h in goid Moiety
Ilse Van Overmeire,§ Swetlana A. Boldin,# Krishnan Venkataraman,# Rivka Zisling,# Steven De J onghe,§
Serge Van Calenbergh,§ Denis De Keukeleire,§ Anthony H. Futerman,# and Piet Herdewijn*,§
Laboratory for Medicinal Chemistry (FFW), Ghent University, Harelbekestraat 72, B-9000 Ghent, Belgium, and
Department of Biological Chemistry, Weizmann Institute of Science, Rehovot 76100, Israel
Received March 14, 2000
The biological activity of synthetic ceramide analogues, having modified sphingoid and N-acyl
chains, as well as fluorine substituents in the allylic position, was investigated in hippocampal
neurons. Their influence on axonal growth was compared to that of C6-N-acyl analogues of
natural ceramides. D-erythro-Ceramides with a phenyl group in the sphingoid moiety and a
short N-acyl chain were able to reverse the inhibitory effect of fumonisin B1 (FB1), but not of
D-threo-1-phenyl-2-decanoylamino-3-morpholino-1-propanol (PDMP), on accelerated axonal
growth in hippocampal neurons. Moreover, we demonstrated that a ceramide analogue with
an aromatic ring in the sphingoid moiety is recognized as a substrate by glucosylceramide
synthase, which suggests that the observed biological effects are mediated by activation of the
ceramide analogue via glucosylation. Introduction of a methyl, pentyl, fluoro, or methoxy
substituent in the para position of the phenyl ring in the sphingoid moiety yielded partly active
compounds. Likewise, substitution of the benzene ring for a thienyl group did not abolish the
ability to reverse the inhibition of accelerated axonal growth by FB1. Both D-erythro- and L-threo-
ceramide analogues, having an allylic fluorine substituent, partly reversed the FB1 inhibition.
In tr od u ction
(2a ) must be metabolized to glucosylceramide (GlcCer)
to be able to reverse the inhibition of axonal growth
caused by FB1.8,9 This was demonstrated using different
stereoisomers of C6-ceramides. L-threo-C6-NBD-ceram-
ide (2b) is not a substrate for GlcCer synthase and does
not reverse the inhibition by FB1. Furthermore, neither
D-erythro- nor L-threo-C6-ceramide (or C6-NBD-ceram-
ide) can reverse the inhibitory effect of PDMP on
accelerated axonal growth.9
Ceramide, an intermediate in sphingolipid biosyn-
thesis and degradation, influences a variety of cellular
processes, e.g. cell differentiation and apoptosis.1,2 Al-
though the majority of studies on the biological activity
of ceramide make use of nonneuronal cell systems,
sphingolipids, and especially gangliosides, are at least
as important in the nervous system, where their abun-
dance is high.3
Inhibition of sphingolipid synthesis causes dramatic
effects on neuronal development. In a murine neuro-
blastoma cell line, neurite growth was inhibited by
D-threo-1-phenyl-2-decanoylamino-3-morpholino-1-pro-
panol (PDMP), an inhibitor of glucosylceramide synthe-
sis.4,5 Studies with fumonisin B1 (FB1), an inhibitor of
ceramide synthesis, demonstrated that ongoing synthe-
sis of ceramide is necessary to sustain the axonal growth
of hippocampal neurons.6 Likewise, in cerebellar Purkin-
je cells, the dendrite growth was inhibited by FB1.7
Later it was established that glucosylceramide synthesis
is required to maintain the axonal growth in hippoc-
ampal neurons.8,9
We previously described the synthesis of a new class
of short-chain ceramides, which possess a styryl group
instead of the alkenyl chain of naturally occurring
sphingolipids. D-erythro-Ceramide analogue 1 was able
to fully reverse the inhibition of FB1 on accelerated
axonal growth, when tested in hippocampal neurons.10
However we were unable to resolve the mechanism of
action of these modified ceramides. It is described that
D-erythro-C6-ceramide and D-erythro-C6-NBD-ceramide
We extended our studies by synthesizing new aro-
matic ceramide analogues to try to get insight in the
structure-activity relationship (SAR) of the lead com-
pound 1. These ceramide analogues possess either a
para-substituted benzene ring (8a -d , Scheme 1) or a
thienyl group (8e, Scheme 1) in the sphingoid part.
Regarding stereochemical implications, we focused only
on the D-erythro-ceramides, since the L-threo-epimer of
1 was previously shown to be unable to reverse the
inhibition of accelerated axonal development by FB1.10
In most of the new compounds a hexanoyl chain
represented the fatty acid residue, since this was the
optimum chain length for the lead compound 1. From
one of the congeners (8b) we also synthesized the
analogues with a shorter acid residue (8f,g), because
* Corresponding author. Tel and fax: +32 16 337387. E-mail:
piet.herdewijn@rega.kuleuven.ac.be.
§ Ghent University.
#
Weizmann Institute of Science.
10.1021/jm000939v CCC: $19.00 © 2000 American Chemical Society
Published on Web 10/04/2000

4190 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Van Overmeire et al.
Sch em e 1
Sch em e 2
the introduction of a pentyl substituent on the aromatic
ring increases the hydrophobicity of the molecule con-
siderably. We, likewise, analyzed the mechanism of
action of these aromatic short-chain ceramides.
Additionally, we investigated the biological activity
of fluorinated short-chain ceramides. Unlike the role of
the primary hydroxyl group of sphingolipids, little is
known about the function of the allylic alcohol group.
Whereas the primary hydroxyl group serves as a gly-
cosyl or phosphate acceptor site resulting in the forma-
tion of glycosphingolipids and phosphosphingolipids
(sphingomyelin, ceramide phosphate), no biochemical
modification of the secondary hydroxyl group is known.
To further explore the importance of the allylic hydroxyl
group of sphingolipids on axonal growth, we synthesized
fluorinated D-erythro- and L-threo-ceramide analogues
(18 and 19, Scheme 3) and investigated their activity
in hippocampal neurons.
Red-Al in Et2O, gave the N-t-Boc-protected sphingosine
analogues 6a -e. Hydrolysis of the carbamates by treat-
ment with 1 N HCl and subsequent N-acylation with
hexanoyl chloride under Schotten-Baumann condi-
tions13 afforded the ceramide analogues 8a -e. Ad-
ditionally, amine 7b was N-acylated using acetyl chlo-
ride and butanoyl chloride to afford ceramide analogues
8f,g, respectively. Compounds 8a -c, having a methyl,
pentyl, and fluoro substituent in the para position of
the phenyl ring, were readily obtained, as the required
phenylacetylenes are commercially available.
The alkynes 4-methoxyphenylacetylene (10) and
2-ethynylthiophene (12) (Scheme 2), required for the
synthesis of ceramides 8d ,e, respectively, were prepared
by reaction of 4-iodoanisole and 2-iodothiophene, re-
spectively, with trimethylsilylacetylene in the presence
of bis(triphenylphosphine)palladium(II) chloride and
cuprous(I) iodide and triethylamine.14 Desilylation of 9
and 11 was carried out by treatment with TBAF in THF
which afforded acetylenes 10 and 12 as volatile oils.15
The corresponding ceramides having a p-chloro or
p-bromo substituent could not be obtained in the same
way as the p-fluoro compound as reduction with Red-
Al resulted in dehalogenation of the phenyl ring.
B. F lu or in a ted Sh or t-Ch a in Cer a m id es. The sub-
stitution of the secondary hydroxyl group of ceramide
analogue 1310 (which has a short sphingoid and fatty
acid chain) by a fluorine atom was investigated using
diethylaminosulfur trifluoride (DAST) as fluorinating
reagens.16,17 Thus, the primary hydroxyl group of 13 was
Resu lts
Ch em istr y. A. Cer a m id es w ith a n Ar om a tic
Sp h in goid Moiety. These compounds were synthesized
according to the procedure described previously for the
synthesis of ceramide analogues with different sphin-
goid chains.10 Addition of the lithium salts of para-
substituted phenylacetylenes to the N-t-Boc-protected
oxazolidine aldehyde 311 in the presence of HMPA gave
erythro alkynols 4a -e (Scheme 1). Exclusive erythro
selectivity is explained by the Felkin-Ahn model for
additions to chiral aldehydes12 and confirmed by conver-
sion of 4a -e to the corresponding acetonides (data not
shown). As reported previously,10 determination of the
J 4,5 allowed unambiguous assignment of the erythro
configuration. Opening of the oxazolidine ring of 4a -e
by treatment with p-TsOH in CH3OH (5a -e), followed
by reduction of the triple bond to a trans alkene with

Synthesis and Evaluation of Ceramide Analogues
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4191
Sch em e 3
Ta ble 1. Ability of D-erythro Aromatic Short-Chain Ceramide Analogues To Reverse the Inhibitory Effects of FB1 on Accelerated
Axonal Growtha
statistical differences (p values) with
no. of axonal branch points/cell
treatment
at 51 h (normalized)b
bFGF-treated cells
(bFGF + FB1)-treated cells
none
bFGF
bFGF + FB1
1.03 ( 0.06
1.63 ( 0.10
0.89 ( 0.08
1.66 ( 0.08
1.01 ( 0.09
2.09 ( 0.14
1.46 ( 0.12
1.54 ( 0.10
1.50 ( 0.10
1.35 ( 0.06
1.38 ( 0.09
1.67 ( 0.09
1.83 ( 0.12
<0.005
<0.005
<0.01
<0.005
<0.005
bFGF + FB1 + D-erythro-C6-NBD-Cer
bFGF + FB1 + L-threo-C6-NBD-Cer
bFGF + FB1 + 1
<0.005
<0.05
<0.05
<0.05
<0.01
<0.05
<0.001
<0.005
bFGF + FB1 + 8a
bFGF + FB1 + 8b
bFGF + FB1 + 8c
bFGF + FB1 + 8d
bFGF + FB1 + 8e
bFGF + FB1 + 8f
bFGF + FB1 + 8g
a
bFGF (1 ng/mL) was added to hippocampal neurons at 48 h in culture together with FB1 (10 µM) and with or without short-acyl-
chain ceramide analogues (5 µM). Values are means ( SEM of measurements from three individual cultures in which 50 cells were
counted per cover slip, for two individual cover slips per treatment. The number of axonal branch points is proportional to the length of
the axon plexus.9 Statistical differences, calculated using the Students’ t-test, are shown as p values. When no p value is given, no statistical
b
differences (p > 0.05) were obtained. The numbers are normalized to 1 at 48 h for the particular set of data in the table.
protected as trityl ether 14 (Scheme 3). Reaction of 14
with 1.5 equiv of DAST in CH2Cl2 at -78 °C gave
It also provided us with a reference compound (21) to
investigate the biological activity of the fluorinated
ceramide 19.
1
compounds 15-17 in a ratio of 2:4:1. H NMR analysis
proved that 15 appeared as an epimeric mixture. Based
on the observation that D-erythro and L-threo epimers
can be distinguished by the upfield resonance of H-C-
OH in the 1H NMR spectrum of the erythro com-
pound,10,18 it was deduced from the δ values for H-C-F
that the L-threo fluoride of 15 was predominantly formed
(L-threo/D-erythro ≈ 2.5:1). The main compound 16 is
formed by a SN2′ type substitution of the hydroxyl group
in 14.19
After deprotection of 15 with Amberlyst 15 in CH3-
OH, both epimers 18 (D-erythro) and 19 (L-threo) could
be separated by preparative HPLC and were obtained
in pure form. In contrast, the mixture of epimers 20,
obtained from 16, could not be separated. From the 13C
NMR spectrum of 20 the ratio of epimers was estimated
to be ≈ 1:1. Prolonged exposure of the oxazoline deriva-
tive 17 to acid furnished the L-threo-ceramide 21,10
indicating that inversion of the configuration at the
allylic position had occurred during cyclization from 14.
Biologica l Eva lu a tion of New Com p ou n d s. The
ability of compounds 8a -g to reverse the inhibitory
effect of FB1 on accelerated axonal growth (by basic
fibroblast growth factor (bFGF)) of hippocampal neurons
was evaluated as previously described.10 D-erythro- and
L-threo-C6-NBD-ceramide (2a ,b) were used as reference
compounds as well as the lead compound 1. All new
compounds were able to partially or fully reverse the
FB1 inhibition (Table 1). The nature of the p-phenyl
substituent has only a minor influence on the activity
(8a -d ). Substitution of the benzene ring for a thienyl
group also yielded an active compound (8e). However,
the activity of the new compounds, having a methyl (8a ),
pentyl (8b), fluoro (8c), or methoxy (8d ) substituent, is
decreased compared to that of the unsubstituted cera-
mide analogue 1. When evaluating the activity of
analogues 8b,f,g, the N-acyl chain length was found to
have a more pronounced effect on the activity. The
aromatic ceramides having a p-pentyl substituent and

4192 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Van Overmeire et al.
Ta ble 2. Comparison of the Influence of Short-Chain Fluorinated and Nonfluorinated Ceramide Analogues on Axonal Growtha
statistical differences (p values) with
no. of axonal branch points/cell
treatment
at 51 h (normalized)b
bFGF-treated cells
(bFGF + FB1)-treated cells
none
bFGF
bFGF + FB1
1.08 ( 0.05
1.56 ( 0.07
1.07 ( 0.05
1.46 ( 0.06
1.06 ( 0.05
1.39 ( 0.07
1.29 ( 0.07
1.43 ( 0.06
1.11 ( 0.05
<0.001
<0.001
<0.005
<0.001
<0.001
<0.05
bFGF + FB1 + D-erythro-C6-NBD-Cer
bFGF + FB1 + L-threo-C6-NBD-Cer
bFGF + FB1 + 18
<0.05
<0.05
<0.01
bFGF + FB1 + 19
bFGF + FB1 + 13
bFGF + FB1 + 21
<0.005
a
Short-chain ceramide analogues were added as indicated in Table 1. Values are means ( SEM of measurements from three individual
b
cultures in which 50 cells were counted per cover slip, for two individual cover slips per treatment. The numbers are normalized to 1 at
48 h for the particular set of data in the table.
Ta ble 3. Inability of Ceramide Analogues To Reverse the Inhibitory Effects of PDMP on Accelerated Axonal Growtha
statistical differences (p values) with
no. of axonal branch points/cell
treatment
at 51 h (normalized)b
bFGF-treated cells
(bFGF + PDMP)-treated cells
none
bFGF
bFGF + PDMP
bFGF + PDMP+ D-erythro-C6-NBD-Cer
bFGF + PDMP + L-threo-C6-NBD-Cer
bFGF + PDMP + 1
1.03 ( 0.06
1.63 ( 0.10
1.06 ( 0.10
0.86 ( 0.13
0.94 ( 0.06
0.92 ( 0.10
<0.005
<0.005
<0.005
<0.005
<0.005
<0.005
a
bFGF (1 ng/mL) was added to hippocampal neurons at 48 h in culture together with PDMP (50 µM) and with or without short-acyl-
chain ceramide analogues (5 µM). Values are means ( SEM of measurements from three individual cultures in which 50 cells were
b
counted per cover slip, for two individual cover slips per treatment. The numbers are normalized to 1 at 48 h for the particular set of
data in the table.
Ta ble 4. Inability of Ceramide Analogues To Reverse the Inhibitory Effects of PDMP on Accelerated Axonal Growtha
statistical differences (p values) with
no. of axonal branch points/cell
treatment
at 51 h (normalized)b
bFGF-treated cells
(bFGF + PDMP)-treated cells
none
bFGF
bFGF + PDMP
bFGF + PDMP+ D-erythro-C6-NBD-Cer
bFGF + PDMP + L-threo-C6-NBD-Cer
bFGF + PDMP + 18
bFGF + PDMP + 19
bFGF + PDMP + 8e
1.08 ( 0.05
1.56 ( 0.07
1.02 ( 0.04
1.08 ( 0.06
1.06 ( 0.06
1.12 ( 0.06
1.16 ( 0.07
1.04 ( 0.06
<0.001
<0.001
<0.001
<0.005
<0.005
<0.005
<0.05
<0.005
a
Short-chain ceramide analogues were added as indicated in Table 3. Values are means ( SEM of measurements from three individual
b
cultures in which 50 cells were counted per cover slip, for two individual cover slips per treatment. The numbers are normalized to 1 at
48 h for the particular set of data in the Table.
a C2- or C4-acyl chain length are the most active new
compounds of the series and are able to fully reverse
the FB1 inhibition.
been shown before that accelerated axonal growth by
treatment with bFGF is inhibited by co-incubation with
PDMP.9 Neither D-erythro-C6-NBD-ceramide nor L-threo-
C6-NBD-ceramide can reverse the PDMP effect (Tables
3 and 4). If co-incubation of PDMP and 1 would lead to
a reversal of the PDMP inhibition, this should imply
that the activity of 1 is independent of its metabolism
to GlcCer. However, if 1 does not counteract the PDMP
effect, then glucosylation may be involved in its mode
of action, supposing that the compound is a substrate
for GlcCer synthase. Thus, hippocampal neurons were
incubated at 48 h in culture with bFGF, PDMP, and 1
and the number of axonal branch points was determined
at 51 h. It was clearly demonstrated that the inhibitory
effect of PDMP is not counteracted by co-incubation with
ceramide analogue 1 (Table 3).
In a second step, we investigated whether glucosyla-
tion of 1 by GlcCer synthase may occur. Therefore, we
examined the metabolism of its 14C-labeled analogue
(label in the carbonyl group) in a rat liver Golgi mem-
brane fraction, which is a rich source of GlcCer syn-
thase.20,21 [14C]-D-erythro-C6-ceramide was used as the
reference compound (Figure 1). The in vitro reactions
Likewise, the biological activity of the fluorinated
compounds 18 and 19 was investigated and compared
to that of the hydroxylated counterparts 13 and 21,
respectively (Table 2). D-erythro- and L-threo-C6-NBD-
ceramide served as reference compounds. Both fluori-
nated epimers (18 and 19) partly reversed the FB1
inhibition of axonal branching in a similar manner,
whereas among nonfluorinated compounds, only D-
erythro epimers (i.e. 13) were active.
Glu cosyla tion a s a Bioa ctivity Mech a n ism . The
ability of short N-acyl chain ceramides (D-erythro-C6-
NBD-ceramide and D-erythro-C6-ceramide) to affect ax-
onal branching of hippocampal neurons has been sug-
gested to involve the formation of GlcCer.9 Two obser-
vations have led to the conclusion that glucosylation is
important as activation step, leading to the biological
activity of aromatic ceramide analogues.
First, we investigated the ability of 1, the most active
of the ceramides having an aromatic residue, to reverse
the inhibition of the axonal growth by PDMP. It has

Synthesis and Evaluation of Ceramide Analogues
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4193
logues was performed as described for 1. D-erythro- and
L-threo-C6-NBD-ceramide were used in the same neu-
ronal cultures as reference compounds (Table 4). Nei-
ther 18 nor 19 caused reversal of the PDMP-inhibition.
It is known that GlcCer synthase does not recognize
L-threo-ceramides as substrates.20,22 In this regard, it
seemed interesting to investigate the metabolism of
D-erythro 18 and L-threo 19. To this end, the corre-
sponding 14C-labeled analogues were synthesized (label
at the carbonyl group). [14C]-18 and [14C]-19 were
separately incubated for 2 h at 37 °C with a rat liver
Golgi membrane fraction and UDP-glucose. After reac-
tion, the lipids were extracted and examined by TLC.
The observation of a number of unknown bands on the
TLCs (not shown) indicated that these 14C-labeled
fluorinated ceramides are unstable compounds. We also
noticed that they progressively degrade upon storage
in solution, even at -20 °C. However, for D-erythro
compound 18, comparison of incubations with and
without UDP-glucose clearly indicated that 18 might
be glucosylated (not shown). Thus, the allylic hydroxyl
group of D-erythro-ceramide does not seem to be critical
for recognition by GlcCer synthase. The complexity of
the thin layer chromatogram, obtained for L-threo
compound 19, prevented us to conclusively establish
whether this compound may be glucosylated or not by
GlcCer synthase. Therefore, the biological activities of
the fluorinated epimers 18 and 19 on axonal develop-
ment (Table 2) must be interpreted with caution, since
some of the effects could be due to degradation products.
F igu r e 1. Metabolism of [14C]-D-erythro-C6-ceramide and
[14C]-1 in a rat liver Golgi membrane fraction (GlcCer )
glucosylceramide, SM ) sphingomyelin). The in vitro reaction
mixtures contained rat liver Golgi membranes (127 µg of
protein/mL), UDP-glucose (5 mM), [14C]-D-erythro-C6-ceramide
or [14C]-1 (2 nmol/mL), MnCl2 (5 mM), and protease inhibitors
in a total volume of 1 mL of buffer (50 mM TRIS-HCl, 25 mM
KCl, pH 7.4), in the presence or absence of UDP-glucose (5
mM). The reactions were terminated after 2 h at 37 °C. Lipids
were extracted28 and separated on TLC using CHCl3/CH3OH/
9.8 mM CaCl2 (60:35:8, v/v/v) as the developing solvent.
of [14C]-D-erythro-C6-ceramide and [14C]-1 with GlcCer
synthase were performed during 2 h at 37 °C, in either
the presence or absence of UDP-glucose. TLC examina-
tion of the extracted lipids revealed formation of a
metabolite of [14C]-1, which, based on its Rf value, is
suggested to be the glucosylated form (Figure 1, lane 3:
incubation with UDP-glucose). This compound is not
formed in the absence of UDP-glucose (Figure 1, lane
4). Since [14C]-1 serves as a substrate of GlcCer syn-
thase, the presence of an alkyl chain in the sphingoid
moiety is not necessary to be recognized by this enzyme.
On the basis of the observations that ceramide analogue
1 reverses the inhibitory effect of FB1 on accelerated
axonal growth, but not the PDMP inhibition, and that
it may act as a substrate for GlcCer synthase, we
suggest that its glucosylated form is involved in the
mediation of the biological effects.
Con clu sion
New ceramide analogues, having either a para-
substituted phenyl ring in the sphingoid moiety or an
allylic fluoride, were synthesized, and their ability to
reverse the inhibited axonal growth of hippocampal
neurons, by FB1, was investigated. Introduction of a
para substituent on the phenyl ring of aromatic D-
erythro analogues 8a -d gave compounds of somewhat
lower activity than the unsubstituted lead compound
1. However, variation of the N-acyl chain length more
clearly influenced the activity of para-substituted aro-
matic ceramide analogues 8f,g. Substitution of the
benzene ring for a thiophene group gave a partially
active compound (8e). The activity of a D-erythro-
ceramide analogue having an allylic fluorine substituent
(18) was comparable to that of its hydroxylated coun-
terpart.
In analogy with the proposed mechanism of action for
1, it is tempting to speculate that glucosylation of the
ceramide analogues, having a substituted aromatic ring
in the sphingoid part (8a -g), is a prerequisite for the
observed biological activity in hippocampal neurons. The
inability of compound 8e to reverse the inhibition by
PDMP on accelerated axonal growth further supports
this hypothesis (Table 4). The lower activity of ceramide
analogues 8a -g with respect to 1 (Table 1) may be
related to the altered cellular uptake or recognition by
GlcCer synthase.
Due to the similar activity of both epimers of the
fluorinated ceramides (18 and 19, Table 2), their mode
of action was studied in more detail. First, the potential
ability to reverse the inhibitory actions of PDMP on
axonal growth was assayed. Co-incubation of bFGF with
PDMP and with or without fluorinated ceramide ana-
We provided indications for the involvement of a
glucosylation step in the mechanism of action of D-
erythro aromatic ceramide analogues. For the fluori-
nated ceramide analogues the mechanism of action was
not completely elucidated as the compounds were found
to be unstable on prolonged incubation. These findings
increase our understanding of the role of the sphingoid
alkyl chain and of the allylic hydroxyl group of ceram-
ides in the axonal growth in hippocampal neurons.
Exp er im en ta l Section
Hip p oca m p a l Cu ltu r es. Hippocampal neurons were cul-
tured at low density as previously described,23 however, with
some modifications.6,24 The dissected hippocampi of embryonic
day 18 rats (Wistar) were dissociated by trypsinization (0.25%
w/v, for 15 min at 37 °C). The tissue was washed in Mg2+
/
Ca2+-free Hank’s balanced salt solution (HBSS) (Gibco) and

4194 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Van Overmeire et al.
dissociated by repeated passage through a constricted Pasteur
pipet. Cells were plated in minimal essential medium (MEM)
with 10% horse serum, at a density of 12 × 103 cells/13-mm
glass cover slip (Assistent, Germany) that had been precoated
with poly-L-lysine (1 mg/mL). After allowing 3-4 h for the cells
to adhere to the substrate, cover slips were transferred into
24-well multidishes (Nunc) containing a monolayer of glial
cells. Cover slips were placed with the neurons facing down-
ward and separated from the glial cells by paraffin ‘feet’.
Cultures were maintained in serum-free medium (MEM),
which included N2 supplements,23 ovalbumin (0.1%, w/v), and
pyruvate (0.1 mM).
Ad d ition of Com p ou n d s. Stock solutions of FB1 and
PDMP were dissolved in Hepes buffer (20 mM, pH 7.4) and
added to cultures to give final concentrations of 10 and 50 µM,
respectively. Sphingolipid analogues were dissolved in ethanol
and added to the culture medium, whereby the final ethanol
concentration did not exceed 1%; control cultures were treated
with 1% ethanol. bFGF (1 ng/mL) was added to cultures after
48 h, together with FB1 (10 µM) or PDMP (50 µM), and
sphingolipid analogues (5 µM), as indicated, and the number
of axonal branch points was measured after 51 h.
An a lysis of Axon a l Gr ow th . After appropriate periods of
time, cover slips were removed from the 24-well multidishes;
neurons were fixed in 1% (v/v) glutaraldehyde in phosphate
buffered saline for 20 min at 37 °C, and mounted for micro-
scopic examination in 50% glycerol in phosphate buffered
saline. Neurons were examined by phase contrast microscopy
using an Achroplan 32×/0.4 n.a. phase 2 objective of a Zeiss
Axiovert 35 microscope. An axon was considered to branch
when the process that it gave rise to was more than 15 µm
long.9 Thin filipodia, which were occasionally observed along
the entire length of the axon, were not considered as branches.
Statistical analysis was performed using the Students’ t-test.
Syn th esis of 14C-La beled Cer a m id es. The 14C-labeled
analogue of ceramide 1 was synthesized by N-acylation of
(2S,3R,4E)-2-amino-5-phenyl-4-pentene-1,3-diol10 with [14C]-
hexanoic acid (55 mCi/mmol) (Biotrend, Ko¨ln, Germany), in
the presence of diethylphosphoryl cyanide and Et3N.25 Puri-
fication was performed by HPLC and UV detection using a
250 × 10 mm column (Biorad, Bio-Sil D90-10) filled with 10
µm silica, a Waters pump (model 510), a Philips Pye Unicam
PU 4025 UV-detector, and a Shimadzu C-R6A Chromatopac
recorder. The 14C-labeled analogues of ceramides 18 and 19
were synthesized by N-acylation of (2S,3R,4E)-2-amino-3-
fluoro-4-dodecen-1-ol and (2S,3S,4E)-2-amino-3-fluoro-4-dode-
cen-1-ol,17 respectively, using the N-hydroxysuccinimide ester
of 1-[14C]hexanoic acid (55 mCi/mmol) (Biotrend, Ko¨ln, Ger-
many).26 Attempts to purify the compounds by preparative
TLC failed due to decomposition. Therefore, the ceramides
were used without further purification after isolation (Et2O
extraction) from the reaction mixture.
Glu cosylcer a m id e Syn th a se Assa y in Ra t Liver Golgi
Mem br a n es w ith 14C-La beled Cer a m id es. A Golgi mem-
brane-enriched fraction was isolated from rat liver by the
method of Dominguez et al.,27 as follows. Rat liver was
homogenized in ice-cold 0.25 M sucrose, 50 mM Tris-HCl, pH
7.4, 25 mM KCl, 5 mM MgCl2, 4.5 mM CaCl2 (STKCM buffer)
using a motorized Potter-Elvehjem homogenizer. After cen-
trifugation at 400g for 10 min, supernatants were adjusted to
0.2 M sucrose in STKCM buffer and underlayed beneath a
discontinuous gradient of 0.90 and 0.4 M sucrose in STKCM.
After centrifugation at 25 000 rpm for 3 h in a SW 28 rotor at
4 °C, Golgi fractions were collected at the 0.4/0.9 M sucrose
interface.
The in vitro reaction mixtures contained a rat liver Golgi
membrane fraction20,21 (127 µg of protein/mL), UDP-glucose
(5 mM), [14C]-D-erythro-C6-ceramide or [14C]-1 (2 nmol/mL),
MnCl2 (5 mM), and protease inhibitors (Complete, EDTA-free
tablets, Boehringer Mannheim) in a total volume of 1 mL of
buffer (50 mM TRIS-HCl, 25 mM KCl, pH 7.4). Control
experiments were performed in the absence of UDP-glucose.
The reactions were terminated after 2 h at 37 °C by addition
of 3 mL of CHCl3/CH3OH (1:2 v/v). Lipids were extracted28 and
separated on TLC using CHCl3/CH3OH/9.8 mM CaCl2 (60:35:
8, v/v/v) as the developing solvent.
For 14C-labeled 18 and 19, the same procedure was followed
using a Golgi membrane fraction (100 µg of protein/mL), UDP-
glucose (5 mM), [14C]-D-erythro-C6-ceramide or 14C-labeled 18
or 19 (5 nmol/mL), MnCl2 (5 mM), and protease inhibitors
(Complete, EDTA-free) in a total volume of 1 mL of buffer (pH
7.4). Lipids were, after extraction, separated on TLC using
CHCl3/CH3OH/NH4OH/H2O (160:40:1:3, v/v/v/v) as the devel-
oping solvent.
Syn th esis. Gen er a l. Optical rotations were measured with
a Perkin-Elmer 241 polarimeter at 20 °C. IR spectra were
recorded with a Perkin-Elmer 1600 Series FTIR spectrometer.
1H NMR spectra were recorded at 500 MHz (Bru¨cker AM-500
or Varian Unity 500) or at 360 MHz (Bru¨cker WH-360). 13C
NMR spectra were recorded at 50 MHz (Varian Gemini-200),
90 MHz (Bru¨cker WH-360) or 125 MHz (Bru¨cker AM-500 or
Varian Unity 500), as indicated. Proton and carbon chemical
shifts are reported in parts per million (ppm) relative to CDCl3
(1H NMR, 7.23 ppm; 13C NMR, 77.00 ppm), CD3OD (1H NMR,
3.35 ppm) or DMSO-d6 (1H NMR, 2.50 ppm; 13C NMR, 39.70
ppm). Liquid secondary-ion mass spectra (LSIMS) were ob-
tained using a Kratos concept 1H mass spectrometer (Kratos,
Manchester, U.K.). Elemental analyses were performed at the
University of Konstanz, Germany, and are within (0.4% of
theoretical values unless otherwise specified.
Precoated Merck silica gel F254 plates were used for TLC,
and spots were examined with UV light at 254 nm and/or
ninhydrin (0.3% in EtOH) solution, phosphomolybdic acid
(0.5% in EtOH) solution and sulfuric acid-anisaldehyde spray.
Column chromatography was performed on SU¨ D-Chemie silica
gel (0.2-0.05 mm). Some compounds were purified by HPLC
using a RI-detector (Metz RI-detector LCD 312), a 250 × 10
mm column (Biorad, Bio-Sil D90-10) filled with 10 µm silica,
and a Kontron 422 pump. THF and Et2O were distilled with
sodium and benzophenone. CH2Cl2 was obtained by distillation
after reflux overnight with CaH2.
ter t-Bu tyl (4S)-4-[(1R)-1-Hyd r oxy-3-(4-m eth ylp h en yl)-
2-p r op yn yl]-2,2-d im eth yl-1,3-oxa zolid in e-3-ca r boxyla te
(4a ). To a stirred and cooled (-20 °C) solution of 4-ethynyl-
toluene (1.37 mL, 10.8 mmol) in dry THF (50 mL) was added
dropwise n-BuLi (1.6 M in hexane, 5 mL, 8.1 mmol). After
standing for 2 h, the solution was further cooled to -78 °C
and HMPA (1.89 mL, 10.8 mmol) was added, followed by a
cooled (-78 °C) solution of 311(1.46 g, 6.38 mmol) in dry THF
(4 mL). After standing for 3 h at -78 °C, the reaction mixture
was warmed to -20 °C and quenched by addition of a
saturated NH4Cl solution (75 mL). The aqueous layer was
extracted with Et2O (3×) and the combined organic layers were
washed with 0.5 N HCl and brine, dried over MgSO4, and
concentrated under reduced pressure. Flash chromatography
(pentane/EtOAc, 9:1) yielded 1.31 g (70%) of 4a as a colorless
oil: IR (cm-1, KBr) 3432, 2975, 2250, 1696; 1H NMR (500 MHz,
DMSO-d6) δ 1.35-1.50 (15 H, m, t-Bu, 2 CH3), 2.30 (3 H, s,
CH3-Ph), 3.90-4.12 (3 H, m, 2H-C(5), H-C(4)), 4.64 (1 H, t,
J ) 5.3 Hz, HCOH), 5.74-5.79 (1 H, m, OH), 7.15-7.20 (2 H,
m, arom), 7.30 (2 H, d, J ) 7.9 Hz, arom); HRMS (LSIMS)
calcd for C20H28NO4 (M + H)+ 346.2018, found 346.2001.
ter t-Bu tyl (1S,2R)-2-Hyd r oxy-1-(h yd r oxym eth yl)-4-(4-
m eth ylp h en yl)-3-bu tyn ylca r ba m a te (5a ). To a solution of
4a (1.06 g, 3.06 mmol) in CH3OH (20 mL) was added p-TsOH
(0.07 g, 0.37 mmol); the reaction mixture was stirred overnight
at room temperature and concentrated under reduced pres-
sure. The residue was dissolved in EtOAc, washed with
saturated NaHCO3 and H2O, and dried over MgSO4. Flash
chromatography (pentane/EtOAc, 6:4) afforded 0.6 g of 5a
(64%) as a colorless oil: IR (cm-1, KBr) 3412, 2975, 2250, 1690;
1H NMR (500 MHz, DMSO-d6) δ 1.30-1.50 (9 H, m, t-Bu), 2.30
(3 H, s, CH3-Ph), 3.45-3.65 (3 H, m, HOCH2, H-C(1)), 4.40
(1 H, t, J ) 6.5 Hz, H-C(2)), 4.58 (1 H, m, OH), 5.61 (1 H, d,
J ) 6.3 Hz, OH), 6.48 (1 H, d, J ) 8.8 Hz, NH), 7.18 (2 H, d,
J ) 7.7 Hz, arom), 7.30 (2 H, d, J ) 7.9 Hz, arom); HRMS
(LSIMS) calcd for C17H24NO4 (M + H)+ 306.1705, found
306.1729.

Synthesis and Evaluation of Ceramide Analogues
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4195
ter t-Bu tyl (1S,2R,3E)-2-Hyd r oxy-1-(h yd r oxym eth yl)-4-
(4-m eth ylp h en yl)-3-bu ten ylca r ba m a te (6a ). A solution of
5a (0.44 g, 1.44 mmol) in dry Et2O (2 mL) was added dropwise
to Red-Al (3.33 M in toluene, 5.2 mL, 17 mmol) and dry Et2O
(5 mL) at 0 °C. The mixture was stirred overnight at room
temperature and then quenched by adding CH3OH (2.5 mL)
at 0 °C. After dilution with Et2O (35 mL) and saturated
disodium tartrate (35 mL), the mixture was stirred for 2 h.
The aqueous layer was extracted with Et2O (3×) and the
organic layers were washed with saturated disodium tartrate
(2×) and brine and dried over MgSO4. Flash chromatography
(pentane/EtOAc, 6:4) yielded 0.35 g (79%) of 6a : IR (cm-1, KBr)
3432, 1685; 1H NMR (500 MHz, DMSO-d6) δ 1.32 (9 H, s, t-Bu),
2.28 (3 H, s, CH3-Ph), 3.40-3.56 (3 H, m, HOCH2, H-C(1)),
4.04-4.09 (1 H, m, H-C(2)), 4.48 (1 H, t, J ) 5.4 Hz, OH),
5.03 (1 H, d, J ) 5.0 Hz, OH), 6.19 (1 H, dd, J ) 15.9 and 6.8
Hz, H-C(3)), 6.36 (1 H, d, J ) 8.6 Hz, NH), 6.44 (1 H, d, J )
15.9 Hz, H-C(4)), 7.11 (2 H, d, J ) 7.8 Hz, arom), 7.27 (2 H,
d, J ) 7.9 Hz, arom); HRMS (LSIMS) calcd for C17H25NO4Na
(M + Na)+ 330.1681, found 330.1659.
(2S,3R,4E)-2-Am in o-5-(4-m eth ylp h en yl)-4-p en ten e-1,3-
d iol (7a ). To a solution of 6a (0.198 g, 0.64 mmol) in dioxane
(6.5 mL) was added 1 N HCl (3.25 mL). The solution was
stirred for 30 min at 100 °C and, after cooling to room
temperature, neutralized with 1 N NaOH (3.25 mL). The
aqueous layer was extracted several times with EtOAc and
the combined organic layers were concentrated under reduced
pressure. Flash chromatography (CH2Cl2/CH3OH/2 M NH3, 80:
20:2) gave 0.097 g (73%) of 7a (erythro/threo as estimated
from TLC: 95:5): 1H NMR (500 MHz, CD3OD) δ 2.30 (3 H, s,
CH3-Ph), 3.03-3.08 (1 H, m, H-C(2)), 3.61 (1 H, dd, J )
11.3 and 7.5 Hz, H-C(1)), 3.76 (1 H, dd, J ) 11.2 and 4.4
Hz, H-C(1)), 4.30 (1 H, t, J ) 6.2 Hz, H-C(3)), 6.22 (1 H, dd,
J ) 15.9 and 7.0 Hz, H-C(4)), 6.65 (1 H, d, J ) 15.9 Hz,
H-C(5)), 7.13 (2 H, d, J ) 8.0 Hz, arom), 7.32 (2 H, d, J ) 8.1
Hz, arom).
ter t-Bu tyl (1S,2R)-2-Hyd r oxy-1-(h yd r oxym eth yl)-4-(4-
p en tylp h en yl)-3-bu tyn ylca r ba m a te (5b). The same proce-
dure, as described for the synthesis of 5a , was followed using
2 g (5 mmol) of 4b and 0.19 g (1 mmol) of p-TsOH. Thus, 1.48
g (82%) of 5b was obtained: IR (cm-1, KBr) 3405, 2934, 2227,
1
1692; H NMR (500 MHz, DMSO-d6) δ 0.84 (3 H, t, J ) 7.1
Hz, CH3CH2), 1.20-1.33 (4 H, m, (CH2)2), 1.37 (9 H, s, t-Bu),
1.54 (2 H, quintet, J ) 7.5 Hz, CH2), 2.56 (2 H, t, J ) 7.6 Hz,
CH2-Ph), 3.45-3.51 (1 H, m, HOCHH), 3.55-3.65 (2 H, m,
HOCHH, H-C(1)), 4.40 (1 H, m, H-C(2)), 4.58 (1 H, t, J )
5.3 Hz, OH, exchangeable with D2O), 5.61 (1 H, d, J ) 5.0 Hz,
OH, exchangeable with D2O), 6.45 (1 H, d, J ) 8.9 Hz, NH,
exchangeable with D2O), 7.17 (2 H, d, J ) 8.0 Hz, arom), 7.30
(2 H, d, J ) 8.0 Hz, arom); HRMS (LSIMS) calcd for C21H32
-
NO4 (M + H)+ 362.2331, found 362.2384.
ter t-Bu tyl (1S,2R,3E)-2-Hyd r oxy-1-(h yd r oxym eth yl)-4-
(4-p en tylp h en yl)-3-bu ten ylca r ba m a te (6b). The same pro-
cedure, as described for the synthesis of 6a , was followed using
1.48 g (4.09 mmol) of 5b and 13.3 mL Red-Al (3.33 M in
toluene, 0.044 mol). After standing for 24 h, the reaction was
worked up in the usual way. Flash chromatography (pentane/
EtOAc, 6:4) yielded 1.21 g (90%) of 6b: IR (cm-1, KBr) 3426,
2933, 1687; 1H NMR (500 MHz, DMSO-d6) δ 0.85 (3 H, t, J )
7.1 Hz, CH3CH2), 1.20-1.40 (13 H, m, t-Bu, (CH2)2), 1.54 (2
H, t, J ) 6.7 Hz, CH2), 2.53 (2 H, t, J ) 7.0 Hz, CH2-Ph),
3.40-3.56 (3 H, m, HOCH2, H-C(1)), 4.08 (1 H, br s, H-C(2)),
4.47 (1 H, br s, OH), 5.00 (1 H, br s, OH), 6.20 (1 H, dd, J )
15.3 and 6.4 Hz, H-C(3)), 6.34 (1 H, d, J ) 8.0 Hz, NH), 6.45
(1 H, d, J ) 15.9 Hz, H-C(4)), 7.12 (2 H, d, J ) 7.0 Hz, arom),
7.27 (2 H, d, J ) 6.9 Hz, arom); HRMS (LSIMS) calcd for
C
21H34NO4 (M + H)+ 364.2488, found 364.2419.
(2S,3R,4E)-2-Am in o-5-(4-p en tylp h en yl)-4-p en ten e-1,3-
d iol (7b). The same procedure, as described for the synthesis
of 7a , was followed using 0.76 g (2.09 mmol) of 6b, 10.5 mL 1
N HCl and 21 mL dioxane. An analytical sample was obtained
by purification of the crude amine (0.95 g obtained) using flash
chromatography (CHCl3/CH3OH/2 M NH3, 90:10:1): 1H NMR
(500 MHz, CD3OD) δ 0.89 (3 H, t, J ) 7.1 Hz, CH3CH2), 1.26-
1.38 (4 H, m, (CH2)2), 1.60 (2 H, quintet, J ) 7.5 Hz, CH2),
2.57 (2 H, t, J ) 7.7 Hz, CH2-Ph), 2.87-2.92 (1 H, m, H-C(2)),
3.54 (1 H, dd, J ) 10.9 and 7.0 Hz, H-C(1)), 3.72 (1 H, dd, J
) 11.0 and 4.6 Hz, H-C(1)), 4.20 (1 H, dt, J ) 6.5 and 0.9 Hz,
H-C(3)), 6.25 (1 H, dd, J ) 15.9 and 7.2 Hz, H-C(4)), 6.62 (1
H, d, J ) 15.9 Hz, H-C(5)), 7.12 (2 H, d, J ) 8.1 Hz, arom),
7.34 (2 H, d, J ) 8.1 Hz, arom); HRMS (LSIMS) calcd for
N-[(1S,2R,3E)-2-Hydr oxy-1-(h ydr oxym eth yl)-4-(4-m eth -
ylp h en yl)-3-bu ten yl]h exa n a m id e (8a ). To a solution of
crude 7a (0.062 g, 0.3 mmol) in a mixture of THF (2.5 mL)
and a 33% aqueous NaOAc (2.5 mL) was added hexanoyl
chloride (34 µL, 0.24 mmol). The mixture was stirred at room
temperature overnight and then diluted with Et2O. The
organic layer was washed with brine and dried over MgSO4.
Flash chromatography (CHCl3/CH3OH, 97:3) and HPLC (CHCl3/
CH3OH, 98:2) afforded 0.06 g (66%) of 8a : IR (cm-1, KBr) 3426,
C
16H25NO2Na (M + Na)+ 286.1783, found 286.1817.
1
1641; H NMR (500 MHz, DMSO-d6) δ 0.76 (3 H, t, J ) 7.0
Hz, CH3CH2), 1.10-1.20 (4 H, m, (CH2)2), 1.36-1.44 (2 H, m,
CH2), 2.04 (2 H, t, J ) 7.5 Hz, COCH2), 2.28 (3 H, s, CH3-
Ph), 3.47-3.56 (2 H, m, HOCH2), 3.74-3.80 (1 H, m, H-C(1)),
4.10-4.15 (1 H, dd, J ) 6.3 and 5.8 Hz, H-C(2)), 4.56 (1 H, t,
J ) 5.3 Hz, OH), 5.08 (1 H, d, J ) 5.3 Hz, OH), 6.18 (1 H, dd,
J ) 15.9 and 6.6 Hz, H-C(3)), 6.45 (1 H, d, J ) 15.9 Hz,
H-C(4)), 7.12 (2 H, d, J ) 8.0 Hz, arom), 7.25 (2 H, d, J ) 8.1
Hz, arom), 7.53 (1 H, d, J ) 8.8 Hz, NH); 13C NMR (125 MHz,
DMSO-d6) δ 13.8, 20.8, 22.0, 25.2, 31.0, 35.6, 55.5 (C-1), 60.5
(HOCH2), 71.4 (C-2), 126.2 (arom), 129.2 (arom), 129.5 (C-3),
130.5 (C-4), 134.2 (Cipso), 136.5 (Cipso), 172.3 (CdO); HRMS
(LSIMS) calcd for C18H27NO3Na (M + Na)+ 328.1889, found
328.1862.
ter t-Bu tyl (4S)-4-[(1R)-1-Hyd r oxy-3-(4-p en tylp h en yl)-
2-p r op yn yl]-2,2-d im eth yl-1,3-oxa zolid in e-3-ca r boxyla te
(4b). The same procedure, as described for the synthesis of
4a , was followed using 1-ethynyl-4-pentylbenzene. Thus, 1.5
g (6.54 mmol) of 3,11 2 g (12 mmol) of 1-ethynyl-4-pentylben-
zene, 5.4 mL n-BuLi (1.6 M in hexane, 8.7 mmol), and 2 mL
(12 mmol) HMPA gave, after flash chromatography (pentane/
EtOAc, 9:1), 2 g (76%) of 4b: IR (cm-1, KBr) 3415, 3056, 2985,
2933, 2239, 1692, 1663; 1H NMR (360 MHz, DMSO-d6) δ 0.84
(3 H, t, J ) 7 Hz, CH3CH2), 1.20-1.33 (4 H, m, (CH2)2), 1.37-
1.60 (17 H, m, t-Bu, 2 CH3, CH2), 2.56 (2 H, t, J ) 7.6 Hz,
CH2-Ph), 3.90-4.11 (3 H, m, 2H-C(5), H-C(4)), 4.63-4.68
(1 H, m, HCOH), 5.79 (1 H, m, OH), 7.15-7.22 (2 H, m, arom),
7.32 (2 H, d, J ) 8.1 Hz, arom); HRMS (LSIMS) calcd for
C24H36NO4 (M + H)+ 402.2644, found 402.2616.
N-[(1S,2R,3E)-2-Hyd r oxy-1-(h yd r oxym eth yl)-4-(4-p en -
tylp h en yl)-3-bu ten yl]h exa n a m id e (8b). To a solution of
crude amine 7b (0.137 g, 0.52 mmol) in a mixture of THF (4
mL) and 33% aqueous NaOAc (4 mL) was added hexanoyl
chloride (72 µL, 0.51 mmol). The reaction mixture was stirred
at room temperature overnight and then diluted with Et2O.
The organic layer was washed with brine and separated. The
aqueous layer was extracted with Et2O (2×). The combined
organic layers were dried over MgSO4 and the solvent was
evaporated. HPLC purification (CH2Cl2/CH3OH, 97:3) yielded
0.092 g (49%) of 8b: IR (cm-1, KBr) 3426, 1641; 1H NMR (500
MHz, DMSO-d6) δ 0.80-0.88 (6 H, m, 2 CH3CH2), 1.10-1.15
(4 H, m, (CH2)2), 1.22-1.30 (4 H, m, (CH2)2), 1.35-1.43 (2 H,
m, COCH2CH2), 1.50-1.58 (2 H, m, CH2), 2.02-2.08 (2 H, m,
COCH2), 2.55-2.57 (2 H, m, CH2-Ph), 3.46-3.55 (2 H, m,
HOCH2), 3.75-3.80 (1 H, m, H-C(1)), 4.12 (1 H, t, J ) 6.5
Hz, H-C(2)), 4.55 (1 H, br s, OH), 5.10 (1 H, br s, OH), 6.17 (1
H, dd, J ) 15.9 and 6.6 Hz, H-C(3)), 6.44 (1 H, d, J ) 15.9
Hz, H-C(4)), 7.12 (2 H, d, J ) 8.0 Hz, arom), 7.26 (2 H, d, J
) 8.0 Hz, arom), 7.52 (1 H, d, J ) 8.8 Hz, NH); 13C NMR (125
MHz, DMSO-d6) δ 13.9, 22.0, 25.2, 30.9, 34.9, 35.6, 55.5 (C-1),
60.5 (HOCH2), 71.5 (C-2), 126.2 (arom), 128.5 (arom), 129.6
(C-3), 130.6 (C-4), 134.5 (Cipso), 141.5 (Cipso), 172.3 (CdO);
HRMS (LSIMS) calcd for C22H35NO3Na (M + Na)+ 384.2515,
found 384.2513. Anal. (C22H35NO3‚1/6CH3OH) C, N; H: calcd,
9.76; found, 9.26.
ter t-Bu tyl (4S)-4-[(1R)-3-(4-F lu or op h en yl)-1-h yd r oxy-
2-p r op yn yl]-2,2-d im eth yl-1,3-oxa zolid in e-3-ca r boxyla te

4196 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Van Overmeire et al.
(4c). The same procedure, as described for the synthesis of
4a , was followed using 1-ethynyl-4-fluorobenzene. Thus, 0.95
g (4.2 mmol) of 3,11 1 g (8.3 mmol) of 1-ethynyl-4-fluorobenzene,
4.55 mL n-BuLi (1.6 M in hexane, 7.3 mmol), and 1.46 mL
(8.3 mmol) HMPA gave 1.3 g (89%) of 4c. An analytical sample
was obtained by HPLC purification (hexane/EtOAc, 82.5:
lacetylene (10), 17.4 mL n-BuLi (1.6 M in hexane, 2.8 mmol),
and 1.4 mL (7.94 mmol) HMPA gave, after flash chromatog-
raphy (hexane/EtOAc, 82.5:17.5), 0.33 g (23%) of 4d : IR (cm-1
,
KBr) 3436, 2235, 1656; 1H NMR (500 MHz, DMSO-d6) δ 1.40-
1.48 (15 H, m, t-Bu, 2 CH3), 3.75 (3 H, s, CH3O), 3.95-4.10 (3
H, m, 2H-C(5), H-C(4)), 4.63 (1 H, br s, HCOH), 5.75 (1 H,
br s, OH), 6.92 (2 H, br s, Ph-3-H, Ph-5-H), 7.34 (2 H, d, J )
8.4 Hz, Ph-2-H, Ph-6-H); HRMS (LSIMS) calcd for C20H27NO5-
Na (M + Na)+ 384.1787, found 384.1795.
1
17.5): IR (cm-1, KBr) 3354, 2242, 1692, 1662; H NMR (500
MHz, DMSO-d6) δ 1.35-1.53 (15 H, m, t-Bu, 2 CH3), 3.90-
4.10 (3 H, m, 2H-C(5), H-C(4)), 4.60-4.67 (1 H, m, HCOH),
5.84 (1 H, t, J ) 7.1 Hz, OH), 7.19-7.25 (2 H, m, arom), 7.45
ter t-Bu tyl (1S,2R)-2-Hyd r oxy-1-(h yd r oxym eth yl)-4-(4-
m eth oxyp h en yl)-3-bu tyn ylca r ba m a te (5d ). The same pro-
cedure, as described for the synthesis of 5a , was followed using
0.52 g (1.44 mmol) of 4d and 0.055 g (0.29 mmol) of p-TsOH.
After standing for 3 h, the reaction was worked up in the usual
way. Flash chromatography (hexane/EtOAc, 6:4) gave 0.185
g (40%) of 5d . A small amount was purified by HPLC (hexane/
EtOAc, 1:1): IR (cm-1, KBr) 3427, 2965, 2233, 1693; 1H NMR
(500 MHz, DMSO-d6) δ 1.36 (9 H, s, t-Bu), 3.45-3.65 (3 H, m,
HOCH2, H-C(1)), 3.75 (3 H, s, CH3O), 4.35-4.42 (1 H, m,
H-C(2)), 4.59 (1 H, br s, OH), 5.59 (1 H, d, J ) 6.0 Hz, OH),
6.46 (1 H, d, J ) 8.7 Hz, NH), 6.92 (2 H, d, J ) 8.6 Hz, Ph-
3-H, Ph-5-H), 7.33 (2 H, d, J ) 8.6 Hz, Ph-2-H, Ph-6-H); HRMS
(LSIMS) calcd for C17H23NO5Na (M + Na)+ 344.1474, found
344.1451.
ter t-Bu tyl (1S,2R,3E)-2-Hyd r oxy-1-(h yd r oxym eth yl)-4-
(4-m eth oxyp h en yl)-3-bu ten ylca r ba m a te (6d ). The same
procedure, as described for the synthesis of 6a , was followed
using 0.15 g (0.47 mmol) of 5d and 0.7 mL Red-Al (3.33 M in
toluene, 2.34 mmol). After standing for 6 h, the reaction was
worked up in the usual way to yield 0.125 g (83%) of 6d . An
analytical sample was obtained by HPLC purification (hexane/
EtOAc, 45:55): 1H NMR (500 MHz, DMSO-d6) δ 1.28 (9 H, s,
t-Bu), 3.40-3.56 (3 H, m, HOCH2, H-C(1)), 3.75 (3 H, s,
CH3O), 4.00-4.08 (1 H, m, H-C(2)), 4.49 (1 H, br s, OH), 5.00
(1 H, d, J ) 4.7 Hz, OH), 6.09 (1 H, dd, J ) 15.8 and 6.8 Hz,
H-C(3)), 6.35 (1 H, d, J ) 8.4 Hz, NH), 6.41 (1 H, d, J ) 15.9
Hz, H-C(4)), 6.87 (2 H, d, J ) 8.3 Hz, Ph-3-H, Ph-5-H), 7.31
(2 H, d, J ) 8.4 Hz, Ph-2-H, Ph-6-H); HRMS (LSIMS) calcd
for C17H25NO5Na (M + Na)+ 346.1631, found 346.1626.
(2 H, t, J ) 6.3 Hz, arom); HRMS (LSIMS) calcd for C19H24
-
NO4FNa (M + Na)+ 372.1587, found 372.1583.
ter t-Bu tyl (1S,2R)-4-(4-F lu or op h en yl)-2-h yd r oxy-1-(h y-
d r oxym eth yl)-3-bu tyn ylca r ba m a te (5c). The same proce-
dure, as described for the synthesis of 5a , was followed using
1.3 g (3.74 mmol) of 4c and 0.142 g (0.75 mmol) of p-TsOH.
Thus, 0.83 g (72%) of 5c was obtained. A small amount was
purified by HPLC (hexane/EtOAc, 1:1): IR (cm-1, KBr) 3413,
3056, 2985, 2262, 1703; 1H NMR (500 MHz, DMSO-d6) δ 1.35
(9 H, s, t-Bu), 3.44-3.51 (1 H, m, HOCHH), 3.53-3.66 (2 H,
m, HOCHH, H-C(1)), 4.40 (1 H, t, J ) 6.7 Hz, H-C(2)), 4.60
(1 H, t, J ) 5.6 Hz, OH), 5.65 (1 H, d, J ) 6.4 Hz, OH), 6.50
(1 H, d, J ) 9.0 Hz, NH), 7.22 (2 H, t, J ) 8.8 Hz, arom), 7.45
(2 H, dd, J ) 8.5 Hz en 5.7 Hz, arom); HRMS (LSIMS) calcd
for C16H20NO4FNa (M + Na)+ 332.1274, found 332.1298.
ter t-Bu tyl (1S,2R,3E)-4-(4-F lu or op h en yl)-2-h yd r oxy-1-
(h yd r oxym eth yl)-3-bu ten ylca r ba m a te (6c). The same pro-
cedure, as described for the synthesis of 6a , was followed using
0.7 g (2.26 mmol) of 5c and 3.4 mL Red-Al (3.33 M in toluene,
11.3 mmol). After standing for 24 h, the reaction was worked
up in the usual way. Flash chromatography (hexane/EtOAc,
45:55) yielded 0.44 g (63%) of 6c: IR (cm-1, KBr) 3436, 1641;
1H NMR (500 MHz, DMSO-d6) δ 1.30 (9 H, s, t-Bu), 3.40-
3.57 (3 H, m, HOCH2, H-C(1)), 4.00-4.10 (1 H, m, H-C(2)),
4.50 (1 H, br s, OH), 5.08 (1 H, br s, OH), 6.20 (1 H, dd, J )
15.7 and 6.6 Hz, H-C(3)), 6.39 (1 H, d, J ) 8.0 Hz, NH), 6.47
(1 H, d, J ) 15.9 Hz, H-C(4)), 7.14 (2 H, t, J ) 8.4 Hz, arom),
7.38-7.48 (2 H, m, arom); HRMS (LSIMS) calcd for C16H22
-
NO4FNa (M + Na)+ 334.1431, found 334.1431.
N -[(1S ,2R ,3E )-4-(4-F lu or op h e n yl)-2-h yd r oxy-1-(h y-
d r oxym eth yl)-3-bu ten yl]h exa n a m id e (8c). To a solution
of 6c (0.215 g, 0.69 mmol) in dioxane (7 mL) was added 1 N
HCl (3.45 mL). The solution was stirred for 30 min at 100 °C
and, after cooling to room temperature, neutralized with 1 N
NaOH (3.45 mL). The aqueous layer was extracted with EtOAc
(3×) and the combined organic layers were concentrated under
reduced pressure. The crude amine (7c) was dissolved in a
mixture of THF (6 mL) and 33% aqueous NaOAc (6 mL), to
which hexanoyl chloride (78 µL, 0.56 mmol) was added. After
standing for 2 h, the reaction mixture was diluted with Et2O
and brine. The aqueous layer was extracted with Et2O (2×)
and the organic layers dried over MgSO4. Purification by flash
chromatography (CH2Cl2/CH3OH, 95:5) and HPLC (CH2Cl2/
CH3OH, 97:3) yielded 0.074 g (35%) of 8c: IR (cm-1, KBr) 3426,
N-[(1S,2R,3E)-2-Hydr oxy-1-(h ydr oxym eth yl)-4-(4-m eth -
oxyp h en yl)-3-bu ten yl]h exa n a m id e (8d ). A solution of 6d
(0.12 g, 0.37 mmol) in dioxane (4 mL) was treated with 1 N
HCl (1.9 mL) for 20 min at 100 °C. After cooling to room
temperature, the mixture was neutralized with 1 N NaOH (1.9
mL) and the aqueous layer was extracted with Et2O (5×) and
with EtOAc (6×). The combined organic layers were concen-
trated to yield the crude amine (7d ). The residue was dissolved
in a mixture of THF (4 mL), 33% aqueous NaOAc (4 mL), and
hexanoyl chloride (0.03 mL, 0.21 mmol). After standing for 2
h, the reaction mixture was diluted with Et2O and brine. The
aqueous layer was extracted with Et2O (2×) and the organic
layers dried over MgSO4. HPLC purification (CH2Cl2/CH3OH,
97:3) gave 0.046 g of 8d (39% from 6d ): 1H NMR (500 MHz,
DMSO-d6) δ 0.83 (3 H, t, J ) 7.0 Hz), 1.12-1.29 (4 H, m,
(CH2)2), 1.33-1.53 (2 H, m, CH2), 2.00-2.12 (2 H, m, CH2),
3.43-3.49 (2 H, m, HOCH2), 3.56-3.62 (1 H, m, H-C(1)), 3.72
(3 H, s, CH3O), 4.32-4.36 (1 H, m, H-C(2)), 4.61 (1 H, t, J )
5.5 Hz, OH), 5.04 (1 H, d, J ) 5.0 Hz, OH), 6.05 (1 H, dd, J )
15.9 and 5.9 Hz, H-C(3)), 6.44 (1 H, d, J ) 16.1 Hz, H-C(4)),
6.86 (2 H, d, J ) 8.5 Hz, Ph-3-H, Ph-5-H), 7.28 (1 H, d, J )
8.6 Hz, NH), 7.44 (2 H, d, J ) 8.7 Hz, Ph-2-H, Ph-6-H); 13C
NMR (50 MHz, DMSO-d6) δ 13.9, 21.9, 25.1,30.9, 35.5, 55.1,
60.9, 70.1, 114.1, 127.5, 129.3, 172.2; HRMS (LSIMS) calcd
for C18H27NO4Na (M + Na)+ 344.1838, found 344.1814.
ter t-Bu tyl (4S)-4-[(1R)-1-Hyd r oxy-3-(2-th ien yl)-2-p r o-
p yn yl]-2,2-d im et h yl-1,3-oxa zolid in e-3-ca r boxyla t e (4e).
The same procedure, as described for the synthesis of 4a , was
followed using 2-ethynylthiophene. Thus, 0.42 g (1.85 mmol)
of 3,11 0.4 g (3.7 mmol) of 2-ethynylthiophene (12), 1.7 mL
n-BuLi (1.6 M in hexane, 2.78 mmol), and 0.65 mL (3.7 mmol)
HMPA gave 0.57 g (91%) of 4e. An analytical sample was
obtained by HPLC purification (hexane/EtOAc, 85:15): 1H
NMR (500 MHz, DMSO-d6) δ 1.30-1.50 (15 H, m, t-Bu, 2 CH3),
3.87-4.08 (3 H, m, 2H-C(5), H-C(4)), 4.55-4.65 (1 H, m,
1
1641; H NMR (500 MHz, DMSO-d6) δ 0.74 (3 H, t, J ) 6.8
Hz, CH3CH2), 1.08-1.20 (4 H, m, (CH2)2), 1.33-1.45 (2 H, m,
COCH2CH2), 2.03 (2 H, dt, J ) 7.9 and 1.7 Hz, COCH2), 3.46-
3.58 (2 H, m, HOCH2), 3.75-3.83 (1 H, m, H-C(1)), 4.10-
4.15 (1 H, m, H-C(2)), 4.58 (1 H, br s, OH), 5.11 (1 H, d, J )
4.7 Hz, OH), 6.18 (1 H, dd, J ) 15.9 and 6.6 Hz, H-C(3)), 6.48
(1 H, d, J ) 15.9 Hz, H-C(4)), 7.14 (2 H, t, J ) 8.8 Hz, arom),
7.40 (2 H, dd, J ) 8.3 and 5.8 Hz, arom), 7.52 (1 H, d, J ) 8.8
Hz, NH); 13C NMR (50 MHz, DMSO-d6) δ 13.7, 21.9, 25.1, 30.8,
35.6, 55.4 (C-1), 60.5 (HOCH2), 71.4 (C-2), 115.2 (arom), 115.7
(arom), 128.0 (arom), 128.2 (arom), 128.5 (C-3), 131.7 (C-4),
133.7 (Cipso), 172.5 (CdO); HRMS (LSIMS) calcd for C17H24
-
-
NO3FNa (M + Na)+ 332.1638, found 332.1668. Anal. (C17H24
NO3F‚1/6CH3OH) C; H: calcd, 7.82; found, 7.64; N: calcd, 4.53;
found, 3.71.
ter t-Bu tyl (4S)-4-[(1R)-1-Hydr oxy-3-(4-m eth oxyph en yl)-
2-p r op yn yl]-2,2-d im eth yl-1,3-oxa zolid in e-3-ca r boxyla te
(4d ). The same procedure, as described for the synthesis of
4a , was followed using 4-methoxyphenylacetylene. Thus, 0.91
g (3.97 mmol) of 3,11 1.05 g (7.94 mmol) of 4-methoxypheny-

Synthesis and Evaluation of Ceramide Analogues
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4197
H-C-OH), 5.84-5.93 (1 H, m, OH), 7.08 (1 H, br s, H-C(4′)),
7.21-7.30 (1 H, m, H-C(3′)), 7.57 (1 H, br s, H-C(5′)); HRMS
(LSIMS) calcd for C17H23NO4SNa (M + Na)+ 360.1246, found
360.1236.
71.3 (C-2), 126.2 (arom), 128.6 (arom), 129.5 (C-3), 130.5 (C-
4), 134.5 (Cipso), 141.6 (Cipso), 169.4 (CdO); HRMS (LSIMS)
calcd for C18H27NO3Na (M + Na)+ 328.1889, found 328.1907.
Anal. (C18H27NO3‚1/12CH3OH) C, N; H: calcd, 8.91; found, 8.76.
ter t-Bu tyl (1S,2R)-2-Hyd r oxy-1-(h yd r oxym eth yl)-4-(2-
th ien yl)-3-bu tyn ylca r ba m a te (5e). The same procedure, as
described for the synthesis of 5a , was followed using 0.55 g
(1.63 mmol) of 4e and 0.062 g (0.33 mmol) of p-TsOH. Thus,
0.32 g (66%) of 5e was obtained. A small amount was purified
by HPLC (hexane/EtOAc, 6:4): IR (cm-1, KBr) 3427, 2233,
1698; 1H NMR (500 MHz, DMSO-d6) δ 1.37 (9 H, s, t-Bu),
3.43-3.64 (3 H, m, HOCH2, H-C(1)), 4.42 (1 H, t, J ) 6.5 Hz,
H-C(2)), 4.60 (1 H, br s, OH), 5.73 (1 H, d, OH), 6.53 (1 H, d,
J ) 9.0 Hz, NH), 7.05 (1 H, t, J ) 4.4 Hz, H-C(4′)), 7.23 (1 H,
d, J ) 3.3 Hz, H-C(3′)), 7.56 (1 H, d, J ) 5.1 Hz, H-C(5′));
HRMS (LSIMS) calcd for C14H19NO4SNa (M + Na)+ 320.0933,
found 320.0931.
N-[(1S,2R,3E)-2-Hyd r oxy-1-(h yd r oxym eth yl)-4-(4-p en -
tylp h en yl)-3-bu ten yl]bu ta n a m id e (8g). The same proce-
dure, as described for the synthesis of 8b, was followed, using
0.22 g (0.83 mmol) of crude amine 7b and 70 µL (0.68 mmol)
butanoyl chloride. Chromatographic purification (CHCl3/CH3-
OH, 95:5) and HPLC (CH2Cl2/CH3OH, 98:2) yielded 0.083 g
1
(30%) of 8g: IR (cm-1, KBr) 3415, 1636; H NMR (500 MHz,
DMSO-d6) δ 0.77 (3 H, t, J ) 7.4 Hz, CH3CH2), 0.85 (3 H, t, J
) 7.0 Hz, CH3CH2), 1.15-1.30 (4 H, m, (CH2)2), 1.38-1.48 (2
H, m, CH2), 1.49-1.58 (2 H, m, CH2), 2.03 (2 H, t, J ) 7.3 Hz,
COCH2), 2.53 (2 H, t, J ) 7.6 Hz, CH2-Ph), 3.43-3.56 (2 H,
m, HOCH2), 3.72-3.82 (1 H, m, H-C(1)), 4.10-4.17 (1 H, m,
H-C(2)), 4.56 (1 H, t, J ) 5.3 Hz, OH), 5.08 (1 H, d, J ) 5.2
Hz, OH), 6.19 (1 H, dd, J ) 15.8 and 6.6 Hz, H-C(3)), 6.46 (1
H, d, J ) 15.7 Hz, H-C(4)), 7.12 (2 H, d, J ) 7.8 Hz, arom),
7.26 (2 H, d, J ) 7.9 Hz, arom), 7.53 (1 H, d, J ) 8.7 Hz, NH);
13C NMR (125 MHz, DMSO-d6) δ 13.6, 14.0, 18.8, 22.0, 30.6,
30.9, 34.8, 37.5, 55.6 (C-1), 60.5 (HOCH2), 71.4 (C-2), 126.2
(arom), 128.5 (arom), 129.5 (C-3), 130.6 (C-4), 134.5 (Cipso),
ter t-Bu tyl (1S,2R,3E)-2-Hyd r oxy-1-(h yd r oxym eth yl)-4-
(2-th ien yl)-3-bu ten ylca r ba m a te (6e). The same procedure,
as described for the synthesis of 6a , was followed using 0.27 g
(0.9 mmol) of 5e and 2.75 mL Red-Al (3.33 M in toluene, 9
mmol). After standing for 4 h, the reaction was worked up in
the usual way. Thus, 0.218 g (81%) of 6e was obtained. A small
amount was purified by HPLC (hexane/EtOAc, 4:1): IR (cm-1
,
141.5 (Cipso), 172.2 (CdO); HRMS (LSIMS) calcd for C20H31
NO3Na (M + Na)+ 356.2202, found 356.2201. Anal. (C20H31
NO3‚1/8CH3OH) C, N; H: calcd, 9.37; found, 9.22.
-
-
1
KBr) 3415, 1687; H NMR (500 MHz, DMSO-d6) δ 1.33 (9 H,
s, t-Bu), 3.36-3.42 (1 H, m, H-C(1)), 3.43-3.49 (1 H, m,
HOCHH), 3.50-3.56 (1 H, m, HOCHH), 3.99-4.05 (1 H, m,
H-C(2)), 4.49 (1 H, t, J ) 5.5 Hz, OH), 5.10 (1 H, d, J ) 5.3
Hz, OH), 6.03 (1 H, dd, J ) 15.7 and 6.4 Hz, H-C(3)), 6.40 (1
H, d, J ) 9.0 Hz, NH), 6.65 (1 H, d, J ) 15.7 Hz, H-C(4)),
6.98 (1 H, dd, J ) 4.8 and 3.6 Hz, H-C(4′)), 7.01 (1 H, d, J )
2.6 Hz, H-C(3′)), 7.36 (1 H, d, J ) 4.9 Hz, H-C(5′)); HRMS
(LSIMS) calcd for C14H21NO4SNa (M + Na)+ 322.1089, found
322.1071.
Tr im eth yl[2-(4-m eth oxyp h en yl)eth yn yl]sila n e (9). To
a solution of 4-iodoanisole (4.97 g, 0.021 mol) in dry Et3N (55
mL) were added PdIICl2(PPh3)2 (0.147 g, 0.21 mmol), CuI (0.04
g, 0.21 mmol), and trimethylsilylacetylene (6 mL, 0.042 mol).
After standing for 1 h at room temperature, the volatiles were
removed in vacuum and the residue was dissolved in CH2Cl2
(100 mL). The organic layer was washed with 5% aqueous
disodium EDTA and H2O and dried over Na2SO4. Flash
chromatography (pentane/EtOAc, 99:1) yielded ≈4.3 g (100%)
of 9 as a colorless oil: IR (cm-1, KBr) 2955, 2155; 1H NMR
(500 MHz, CDCl3) δ 0.25 (9 H, s, Si-Me3), 3.80 (3 H, s, CH3O),
6.82 (2 H, d, J ) 7.9 Hz, Ph-3-H, Ph-5-H), 7.40 (2 H, d, J )
7.9 Hz, Ph-2-H, Ph-6-H).
N-[(1S,2R,3E)-2-Hyd r oxy-1-(h yd r oxym eth yl)-4-(2-th ie-
n yl)-3-bu ten yl]h exa n a m id e (8e). A solution of 6e (64 mg,
0.21 mmol) in dioxane (2 mL) was treated with 1 N HCl (1
mL) for 30 min at 100 °C. After cooling to room temperature,
the mixture was neutralized with 1 N NaOH (1 mL) and
extracted with EtOAc. The combined organic layers were
concentrated and the crude amine (7e) was dissolved in a
mixture of THF (1 mL) and 33% aqueous NaOAc (1 mL).
Hexanoyl chloride (22 µL, 0.16 mmol) was added and the
mixture was stirred overnight at room temperature and then
diluted with Et2O and brine. The aqueous layer was extracted
with Et2O (3×) and the organic layers were dried over MgSO4.
Purification by HPLC (CH2Cl2/CH3OH, 98:2) gave 0.03 g (48%)
of 8e: 1H NMR (500 MHz, CDCl3) δ 0.85-0.93 (3 H, m, CH3-
CH2), 1.25-1.40 (4 H, m, (CH2)2), 1.62-1.70 (2 H, m,
COCH2CH2), 2.24 (2 H, t, J ) 7.5 Hz, COCH2), 3.73-3.79 (1
H, m, HOCHH), 3.98-4.40 (2 H, m, HOCHH, H-C(1)), 4.55
(1 H, br s, H-C(2)), 6.10 (1 H, dd, J ) 15.7 and 5.7 Hz,
H-C(3)), 6.40 (1 H, br s, NH), 6.84 (1 H, d, J ) 15.8 Hz,
H-C(4)), 6.95-7.00 (2 H, m, H-C(3′), H-C(4′)), 7.17 (1 H, d,
J ) 4.4 Hz, H-C(5′)); 13C NMR (125 MHz, CDCl3) δ 13.8, 22.3,
25.3, 31.3, 36.6, 54.4 (C-1), 62.3 (HOCH2), 69.5 (C-2), 124.9,
4-Meth oxyp h en yla cetylen e (10). To a solution of 9 (5.1
g, 0.025 mol) in THF (300 mL) was added TBAF (1 M in THF,
27.5 mL, 0.0275 mol). After standing for 3 h at room temper-
ature, the mixture was diluted with Et2O and saturated
aqueous NaHCO3. The organic layer was removed and the
aqueous layer was extracted with Et2O (2×). The combined
organic layers were dried over MgSO4. Flash chromatography
(pentane/Et2O, 95:5) yielded 2.4 g (72%) of 10 as a colorless
1
oil: IR (cm-1, KBr) 3303, 2122; H NMR (500 MHz, CDCl3) δ
3.00 (1 H, s, ethynyl H), 3.75 (3 H, t, J ) 6.6 Hz, CH3O), 6.84
(2 H, d, J ) 8.8 Hz, Ph-3-H, Ph-5-H), 7.47 (2 H, d, J ) 8.8 Hz,
Ph-2-H, Ph-6-H).
Tr im eth yl[2-(2-th ien yl)eth yn yl]sila n e (11). To a solution
of 2-iodothiophene (2.8 mL, 0.025 mol) in dry Et3N (63 mL)
were added PdIICl2(PPh3)2 (0.175 g, 0.25 mmol), CuI (0.048 g,
0.25 mmol), and trimethylsilylacetylene (5 g, 0.051 mol). After
standing for 2 h at room temperature, the volatiles were
removed in vacuum and the residue was dissolved in CH2Cl2
(150 mL). The organic layer was washed with 5% aqueous
disodium EDTA and H2O and dried over Na2SO4. Flash
chromatography (pentane) yielded 4.67 g (100%) of 11 as a
126.2, 126.7, 127.3, 130.7; HRMS (LSIMS) calcd for C15H23
-
NO3SNa (M + Na)+ 320.1297, found 320.1277.
N-[(1S,2R,3E)-2-Hyd r oxy-1-(h yd r oxym eth yl)-4-(4-p en -
tylp h en yl)-3-bu ten yl]a ceta m id e (8f). The same procedure,
as described for the synthesis of 8b, was followed using 0.22
g (0.83 mmol) of crude amine 7b and 47 µL (0.67 mmol) acetyl
chloride. Flash chromatography (CHCl3/CH3OH, 95:5) and
HPLC (CH2Cl2/CH3OH, 95:5) yielded 0.154 g (61%) of 8f: IR
(cm-1, KBr) 3415, 1646; 1H NMR (500 MHz, DMSO-d6) δ 0.85
(3 H, t, J ) 7.0 Hz, CH3CH2), 1.20-1.33 (4 H, m, (CH2)2), 1.48-
1.58 (2 H, m, CH2), 1.74-1.84 (3 H, m, CH3CO), 2.53 (2 H, t,
J ) 7.6 Hz, CH2-Ph), 3.43-3.53 (2 H, m, HOCH2), 3.72-3.85
(1 H, m, H-C(1)), 4.12-4.18 (1 H, m, H-C(2)), 4.58 (1 H, br
s, OH), 5.08 (1 H, d, J ) 5.2 Hz, OH), 6.21 (1 H, dd, J ) 15.9
and 6.4 Hz, H-C(3)), 6.47 (1 H, d, J ) 15.7 Hz, H-C(4)), 7.13
(2 H, d, J ) 8.0 Hz, arom), 7.28 (2 H, d, J ) 8.0 Hz, arom),
7.63 (1 H, d, J ) 8.7 Hz, NH); 13C NMR (125 MHz, DMSO-d6)
δ 14.0, 22.0, 22.9, 30.7, 30.9, 34.9, 55.8 (C-1), 60.3 (HOCH2),
1
colorless oil: IR (cm-1, KBr) 2964, 2144; H NMR (500 MHz,
CDCl3) δ 0.25 (9 H, s, Si-Me3), 6.95 (1 H, dd, J ) 4.9 and 3.9
Hz, H-C(4′)), 7.24 (2 H, d, J ) 4.3 Hz, H-C(3′), H-C(5′)).
2-Eth yn ylth iop h en e (12). To a solution of 11 (4.67 g, 0.023
mol) in THF (295 mL) was added TBAF (1 M in THF, 28 mL,
0.028 mol). After standing for 2 h at room temperature, the
reaction mixture was diluted with Et2O (100 mL) and satu-
rated aqueous NaHCO3 (100 mL). The aqueous layer was
removed, and the organic layer was washed with H2O (2×)
and brine and dried over MgSO4. After concentration in
vacuum, the resulting oil (2.54 g) was dissolved in pentane
and filtered over silica to give 12 as a volatile oil: IR (cm-1
KBr) 3304, 2111; H NMR (500 MHz, CDCl3) δ 3.35 (1 H, s,
,
1

4198 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Van Overmeire et al.
ethynyl H), 6.98 (1 H, dd, J ) 5.1 and 3.6 Hz, H-C(4)), 7.20
(2 H, m, H-C(3), H-C(5)).
(1 H, dd, J ) 11.5 and 3.5 Hz, HOCHH), 3.92 (1 H, dd, J )
3
11.2 and 4.1 Hz, HOCHH), 4.07-4.17 (1 H, m, H-C(1), J H,F
) 21.0 Hz), 4.97 (0.5 H, t, J ) 5.9 Hz, H-C(2)), 5.06 (0.5 H, t,
N -{(1S ,2R ,3E )-2-H y d r o x y -1-[(t r it y lo x y )m e t h y l]-3-
u n d ecen yl}h exa n a m id e (14). A mixture of 139 (0.3 g, 0.96
mmol) and triphenylmethyl chloride (0.47 g, 1.67 mmol) in dry
pyridine (27 mL) was heated at 100 °C for 3 h. After cooling
to room temperature, it was poured into ice-cold 1 N HCl (370
mL) and stirred for 15 min. The aqueous layer was extracted
with Et2O (3×), and the combined organic layers were washed
with 2.5% NaHCO3 and H2O and dried over MgSO4. Flash
chromatography (hexane/EtOAc, 75:25) yielded 0.48 g of 14
2
J ) 6.0 Hz, H-C(2), J H,F ) 48.2 Hz), 5.51-5.60 (1 H, m,
H-C(3)), 5.84-5.92 (1 H, m, H-C(4)), 6.00 (1 H, d, J ) 8.0
Hz, NH); 13C NMR (50 MHz, CDCl3) δ 13.8, 14.0, 22.3, 22.6,
2
25.3, 28.7, 29.0, 31.3, 31.7, 32.2, 36.7, 53.6 (C-1, J C,F ) 22.9
1
Hz), 61.3 (HOCH2), 93.6 (C-2, J C,F ) 170.9 Hz), 124.5 (C-3,
3
2J C,F ) 18.3 Hz), 137.8 (C-4, J C,F ) 12.2 Hz), 173.7 (CdO);
HRMS (LSIMS) calcd for C18H34NO2FNa (M + Na)+ 338.2471,
found 338.2431.
20
N-[(1S,2S,3E)-2-Flu or o-1-(h ydr oxym eth yl)-3-u n decen yl]-
h exa n a m id e (19). Procedure: see 18; 1H NMR (500 MHz,
CDCl3) δ 0.87-0.97 (6 H, m, 2 CH3), 1.20-1.45 (14 H, m, 7
CH2), 1.60-1.67 (2 H, m, COCH2CH2), 2.02-2.11 (2 H, m, 2H-
C(5)), 2.23 (2 H, t, J ) 7.5 Hz, COCH2), 3.70-3.80 (2 H, m,
HOCH2), 4.04-4.09 (0.5 H, m, H-C(1)), 4.10-4.15 (0.5 H, m,
(90%) as a colorless oil: [R]D ) -1.98 (c 1.36, CHCl3); UV
(CH3OH) λmax 210 nm (log ꢀ 4.32); IR (cm-1, KBr) 3327, 2920,
1651; 1H NMR (360 MHz, CDCl3) δ 0.86-0.96 (6 H, m, 2 CH3),
1.20-1.40 (14 H, m, 7 CH2), 1.62-1.70 (2 H, m, COCH2CH2),
1.87-1.96 (2 H, m, 2H-C(5)), 2.20 (2 H, t, J ) 7.7 Hz, COCH2),
3.31 (1 H, dd, J ) 9.7 and 4.0 Hz, TrOCHH), 3.37-3.42 (2 H,
m, TrOCHH, OH), 4.03-4.10 (1 H, m, H-C(1)), 4.16-4.21 (1
H, m, H-C(2)), 5.26 (1 H, dd, J ) 15.4 and 6.2 Hz, H-C(3)),
5.64 (1 H, dt, J ) 15.4 and 6.6 Hz, H-C(4)), 6.08 (1 H, d, J )
7.9 Hz, NH), 7.20-7.45 (15 H, m, arom); HRMS (LSIMS) calcd
for C37H49NO3Na (M + Na)+ 578.3610, found 578.3602.
3
H-C(1), J H,F ) 23.8 Hz), 5.07 (0.5 H, dd, J ) 6.5 and 3.5 Hz,
2
H-C(2)), 5.16 (0.5 H, dd, J ) 6.5 and 3.5 Hz, H-C(2), J H,F
)
47.7 Hz), 5.46-5.55 (1 H, m, H-C(3)), 5.80-5.95 (2 H, m, NH
en H-C(4)); 13C NMR (50 MHz, CDCl3) δ 13.8, 14.0, 22.3, 22.6,
2
25.3, 28.7, 29.0, 31.3, 31.7, 32.2, 36.7, 54.3 (C-1, J C,F ) 19.8
1
Hz), 62.6 (HOCH2), 92.2 (C-2, J C,F ) 169.4 Hz), 124.5 (C-3,
N-{[1S,2(RS),3E]-2-F lu or o-1-[(t r it yloxy)m et h yl]-3-u n -
d ecen yl}h exa n a m id e (15). To a cooled (-78 °C) solution of
DAST (0.19 mL, 1.43 mmol) in dry CH2Cl2 (10 mL) was added
dropwise a solution of 14 (0.53 g, 0.95 mmol) in dry CH2Cl2
(10 mL). After standing for 45 min at -78 °C, examination by
TLC (hexane/EtOAc, 75:25) indicated completion of the reac-
tion. The reaction mixture was washed with 5% aqueous
NaHCO3 (3×) and the organic layer was dried over MgSO4.
Flash chromatography (hexane/EtOAc, 94:6) gave 0.07 g (14%)
of 17. Further elution (hexane/EtOAc, 9:1) afforded 0.15 g
(28%) of 15 and 0.3 g (57%) of 16. The pure threo compound
was obtained after HPLC purification of 15 (hexane/EtOAc,
9:1): 1H NMR (500 MHz, CDCl3) (data for threo) δ 0.85 (6 H,
m, 2 CH3), 1.20-1.40 (14 H, m, 7 CH2), 1.55-1.60 (2 H, m,
CH2), 1.98-2.08 (2 H, m, 2H-C(5)), 2.13 (2 H, t, J ) 7.6 Hz,
COCH2), 3.18-3.25 (2 H, m, TrOCH2), 4.26-4.37 (1 H, m,
H-C(1), 3J H,F ) 22.7 Hz), 5.14-5.18 (0.5 H, m, H-C(2)), 5.23-
2J C,F ) 18.7 Hz), 137.6 (C-4, J C,F ) 10.7 Hz), 174.2 (CdO);
3
HRMS (LSIMS) calcd for C18H35NO2F (M + H)+ 316.2652,
found 316.2696.
N -{[1R ,2E ,4(R S )]-4-F lu o r o -1-(h y d r o x y m e t h y l)-2-
u n d ecen yl}h exa n a m id e (20). A solution of 16 (0.174 g, 0.31
mmol) in CH3OH (3 mL) was treated with Amberlyst 15 (0.203
g) for 48 h at room temperature. The mixture was filtered over
Celite and concentrated. Flash chromatography (CH2Cl2/CH3-
OH, 98:2) yielded 0.06 g of 20 (61%): 1H NMR (360 MHz,
CDCl3) δ 0.83-0.97 (6 H, m, 2 CH3), 1.20-1.48 (14 H, m, 7
CH2), 1.60-1.82 (4 H, m, 2 CH2), 2.18-2.28 (2 H, m, COCH2),
3.64-3.77 (2 H, m, HOCH2), 4.51-4.63 (1 H, m, H-C(1)),
4.78-4.85 (0.5 H, m, H-C(4)), 4.92-4.98 (0.5 H, m, H-C(4),
J H,F ) 48.5 Hz), 5.66-5.80 (2 H, m, H-C(2), H-C(3)), 5.92-
6.00 (1 H, m, NH); 13C NMR (90 MHz, CDCl3) δ 13.8, 14.0,
22.3, 22.5, 24.6, 25.3, 29.0, 29.2, 31.3, 31.7, 35.2, 35.4, 36.7,
1
52.3 (C-1), 52.4 (C-1), 65.2 (CH2OH), 92.6 (C-4, J C,F ) 167.0
2
5.27 (0.5 H, m, H-C(2), J H,F ) 47.6 Hz), 5.44-5.53 (1 H, m,
1
3
Hz), 92.8 (C-4, J C,F ) 166.1 Hz), 129.3 (C-2, J C,F ) 11.1 Hz),
H-C(3)), 5.57 (1 H, d, J ) 9.4 Hz, NH), 5.79-5.88 (1 H, m,
H-C(4)), 7.20-7.45 (15 H, m, arom); HRMS (LSIMS) calcd
for C37H48NO2 (M - HF + H)+ 538.3685, found 538.3669.
129.7 (C-2, 3J C,F ) 11.0 Hz), 131.1 (C-3, 2J C,F ) 18.9 Hz), 131.3
2
(C-3, J C,F ) 18.9 Hz), 173.4 (CdO); HRMS (LSIMS) calcd for
C18H35NO2F (M + H)+ 316.2652, found 316.2746. Anal. (C18H34
NO2F‚1/10CH3OH) C, N; H: calcd, 10.86; found, 10.49.
-
N -{[1R ,2E ,4(R S )]-4-F lu or o-1-[(t r it yloxy)m e t h yl]-2-
1
u n d ecen yl}h exa n a m id e (16). Procedure: see 15; H NMR
(360 MHz, CDCl3) δ 0.85-0.95 (6 H, m, 2 CH3), 1.20-1.50 (14
H, m, 7 CH2), 1.50-1.80 (4 H, m, 2 CH2), 2.15 (2 H, t, J ) 7.5
Hz, COCH2), 3.19-3.29 (2 H, m, TrOCH2), 4.67-4.74 (1 H, m,
H-C(1)), 4.79-4.86 (0.5 H, m, H-C(4)), 4.93-4.99 (0.5 H, m,
Ack n ow led gm en t. We are grateful to Dr. C. Hen-
drix and Dr. K. Rothenbacher for recording the mass
spectra and thank Prof. R. Busson and Prof. A. De
Bruyn for help with the interpretation of NMR spectra.
We also acknowledge the help of Prof. G. Slegers and
Dr. F. Dumont with the radiolabeling of the compounds.
This work was financially supported by the Flemish
Institute for the Promotion of Scientific-Technological
Research in Industry, IWT, Brussels, Belgium (doctoral
fellowship to I.V.O.), the Special Research Fund of the
University of Ghent (BOZF Project No. 01101497), the
Israel Science Foundation, and the Buddy Taub Foun-
dation (research grant to A.H.F.).
2
H-C(4), J H,F ) 48.2 Hz), 5.61-5.87 (3 H, m, NH, H-C(2),
H-C(3)), 7.20-7.45 (15 H, m, arom); HRMS (LSIMS) calcd
for C37H48NO2FNa (M + Na)+ 580.3567, found 580.3586.
(4S,5S)-5-[(E)-1-Non en yl]-2-pen tyl-4-[(tr ityloxy)m eth yl]-
4,5-d ih yd r o-1,3-oxa zole (17). Procedure: see 15; 1H NMR
(500 MHz, CDCl3) δ 0.85-0.92 (6 H, m, 2 CH3), 1.20-1.42 (16
H, m, 8 CH2), 1.66-1.75 (2 H, m, CH2), 2.02-2.09 (2 H, m,
HCdCHCH2), 3.25-3.30 (2 H, m, TrOCH2), 3.94-3.98 (1 H,
m, H-C(4)), 4.80-4.85 (1 H, m, H-C(5)), 5.47 (1 H, dd, J )
15.3 and 8.1 Hz, HCdCHCH2), 5.72 (1 H, dt, J ) 15.3 and 6.7
Hz, HCdCHCH2), 7.20-7.45 (15 H, m, arom); HRMS (LSIMS)
calcd for C37H48NO2 (M + H)+ 538.3685, found 538.3711.
Su p p or tin g In for m a tion Ava ila ble: Analytical data for
compounds 8a -g and 20. This material is available free of
charge via the Internet at http://pubs.acs.org.
N-[(1S,2R,3E)-2-Flu or o-1-(h ydr oxym eth yl)-3-u n decen yl]-
h exa n a m id e (18). A solution of 15 (0.147 g, 0.26 mmol) in
CH3OH (5 mL) was treated with Amberlyst 15 (0.256 g) at
room temperature for 24 h. The heterogeneous mixture was
then filtered over Celite. Before concentration, the methanolic
solution was diluted with H2O and extracted with Et2O. The
Et2O layers were washed with saturated aqueous NaHCO3 and
dried (MgSO4). Flash chromatography (CH2Cl2/CH3OH, 98:2)
yielded 0.04 g (50%) of a mixture of 18 and 19 which was
separated by HPLC (CH2Cl2/CH3OH, 99:1): 1H NMR (500
MHz, CDCl3) δ 0.86-0.93 (6 H, m, 2 CH3), 1.20-1.43 (14 H,
m, 7 CH2), 1.60-1.67 (2 H, m, COCH2CH2), 2.04-2.11 (2 H,
m, 2H-C(5)), 2.21 (2 H, dt, J ) 7.7 and 2.0 Hz, COCH2), 3.73
Refer en ces
(1) Hannun, Y. A.; Obeid, L. M.; Wolff, R. A. The Novel Second
Messenger Ceramide: Identification, Mechanism of Action, and
Cellular Activity. Adv. Lipid Res. 1993, 25, 43-64.
(2) J arvis, W. D.; Grant, S.; Kolesnick, R. N. Ceramide and the
Induction of Apoptosis. Clin. Cancer Res. 1996, 2, 1-6.
(3) Schwarz, A.; Futerman A. H. The Localization of Gangliosides
in Neurons of the Central Nervous System: the Use of Anti-
ganglioside Antibodies. Biochim. Biophys. Acta 1996, 1286, 247-
267.

Synthesis and Evaluation of Ceramide Analogues
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4199
(4) Uemura, K.; Sugiyama, E.; Taketomi, T. Effects of an Inhibitor
of Glucosylceramide Synthase on Glycosphingolipid Synthesis
and Neurite Outgrowth in Murine Neuroblastoma Cell Lines.
J . Biochem. Tokyo 1991, 110, 96-102.
(5) Inokuchi, J .; Radin, N. S. Preparation of the Active Isomer of
1-Phenyl-2-decanoylamino-3-morpholino-1-propanol, Inhibitor of
Glucocerebroside Synthetase. J . Lipid Res. 1987, 28, 565-571.
(6) Harel, R.; Futerman, A. H. Inhibition of Sphingolipid Synthesis
Affects Axonal Outgrowth in Cultured Hippocampal Neurons.
J . Biol. Chem. 1993, 268, 14476-14481.
(17) De J onghe, S.; Van Overmeire, I.; Poulton, S.; Hendrix, C.;
Busson, R.; Van Calenbergh, S.; De Keukeleire, D.; Spiegel, S.;
Herdewijn, P. Structure-activity Relationships of Short-Chain
Sphingoid Bases as Inhibitors of Sphingosine Kinase. Bioorg.
Med. Chem. Lett. 1999, 9, 3175-3180.
(18) Herold, P. Synthesis of D-erythro- and D-threo-Sphingosine
Derivatives from L-Serine. Helv. Chim. Acta 1988, 71, 354-362.
(19) Magid, R. M. Nucleophilic and Organometallic Displacement
Reactions of Allylic Compounds: Stereo- and Regiochemistry.
Tetrahedron 1980, 36, 1901-1930.
(7) Furuya, S.; Ono, K.; Hirabayashi, Y. Sphingolipid Biosynthesis
is Necessary for Dendrite Growth and Survival of Cerebellar
Purkinje Cells in Culture. J . Neurochem. 1995, 65, 1551-1561.
(8) Schwarz, A.; Futerman, A. H. Distinct Roles for Ceramide and
Glucosylceramide at Different Stages of Neuronal Growth. J .
Neurosci. 1997, 17, 2929-2938.
(20) Paul, P.; Kamisaka, Y.; Marks, D. L.; Pagano, R. E. Purification
and Characterization of UDP-glucose:Ceramide Glucosyltrans-
ferase from Rat Liver Golgi Membranes. J . Biol. Chem. 1996,
271, 2287-2293.
(21) Futerman, A. H.; Pagano, R. E. Determination of the Intracel-
lular Sites and Topology of Glucosylceramide Synthesis in Rat
Liver. Biochem. J . 1991, 280, 295-302.
(9) Boldin, S.; Futerman, A. H. Glucosylceramide Synthesis is
Required for Basic Fibroblast Growth Factor and Laminin to
Stimulate Axonal Growth J . Neurochem. 1997, 68, 882-885.
(10) Van Overmeire, I.; Boldin, S.; Dumont, F.; Van Calenbergh, S.;
Slegers, G.; De Keukeleire, D.; Futerman, A. H.; Herdewijn, P.
Effect of Aromatic Short-Chain Analogues of Ceramide on
Axonal Growth in Hippocampal Neurons. J . Med. Chem. 1999,
42, 2697-2705.
(22) Pagano, R. E.; Martin, O. C. A Series of Fluorescent N-
acylsphingosines: Synthesis, Physical Properties, and Studies
in Cultured Cells. Biochemistry 1988, 27, 4439-4445.
(23) Goslin, K.; Asmussen, H.; Banker, G. Rat Hippocampal Neurons
in Low-Density Culture. In Culturing Nerve Cells; Banker, G.,
Goslin, K., Eds.; MIT Press: Cambridge, MA, 1998; pp 339-
370.
(11) Garner, P.; Park, J . M. The Synthesis and Configurational
Stability of Differentially Protected b-Hydroxy-a-Amino Alde-
hydes. J . Org. Chem. 1987, 52, 2361-2364.
(24) Schwarz, A.; Rapaport, E.; Hirschberg, K.; Futerman, A. H. A
Regulatory Role for Sphingolipids in Neuronal Growth. J . Biol.
Chem. 1995, 270, 10990-10998.
(12) Anh, N. T. Regio- and Stereo-Selectivities in Some Nucleophilic
Reactions. Top. Curr. Chem. 1980, 88, 144-162.
(25) Anand, J . K.; Sadozai, K. K.; Hakomori, S. A Simple Method for
the Synthesis of Ceramides and Radioloabeled Analogues. Lipids
1996, 31, 995-998.
(13) Wild, R.; Schmidt, R. R. Sphingosine and Phytosphingosine from
D-Threose, Synthesis of a 4-Keto-Ceramide. Tetrahedron: Asymm.
1994, 5, 2195-2208.
(26) Futerman, A. H.; Pagano, R. E. Use of N-([1-14C]Hexanoyl)-D-
erythro-sphingolipids to Assay Sphingolipid Metabolism. Meth-
ods Enzymol. 1992, 209, 437-446.
(14) Krause, M.; Ligneau, X.; Stark, H.; Garbarg, M.; Schwartz, J .
C.; Schunack, W. 4-Alkenylphenyl Imidazolylpropyl Ethers as
Selective Histamine H3-Receptor Antagonists with High Oral
Central Nervous System Activity. J . Med. Chem. 1998, 41,
4171-4176.
(27) Dominguez, M.; Fazel, A.; Dahan, S.; Lovell, J .; Hermo, L.;
Claude, A.; Melancon, P.; Bergeron, J . J . Fusogenic Domains of
Golgi Membranes are Sequestered into Specialized Regions of
the Stack That Can Be Released by Mechanical Fragmentation.
J . Cell Biol. 1999, 145, 673-688.
(28) Bligh, E. G.; Dyer, W. J . A Rapid Method of Total Lipid
Extraction and Purification. Can. J . Physiol. 1959, 37, 911-917.
(15) Merino, P.; Franco, S.; Merchan, F. L.; Tejero, T. Asymmetric
Addition Reactions of Lithium (Trimethylsilyl)acetylide with
Chiral R-Amino Nitrones. J . Org. Chem. 1998, 63, 5627-5630.
(16) Middleton, W. J . New Fluorinating Reagents. Dialkylaminosul-
fur Fluorides. J . Org. Chem. 1975, 40, 574-578.
J M000939V