Syntheses of 3-C-methylceramides

Article abstract of DOI:10.1002/ejoc.200400114

We report on the synthesis of a series of 3-C-methylceramide derivatives 2a,b, 12a,b, 13a,b, and 14a,b. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2004.

Full text of DOI:10.1002/ejoc.200400114

FULL PAPER
Syntheses of 3-C-Methylceramides
Peter Sawatzki[a] and Thomas Kolter*[a]
Keywords: Bioorganic chemistry / Cell surface engineering / Ceramide / Lipids / Sphingolipids
We report on the synthesis of a series of 3-C-methylceramide
derivatives 2a,b, 12a,b, 13a,b, and 14a,b.
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2004)
Introduction
5-methylsphingosine, (Z)-sphingosine and 1-deoxysphingo-
sine inhibited sphingolipid biosynthesis less efficiently than
(Z)-4-methylsphingosine and had no influence on cell mor-
phology. Some of the observed effects appear to be me-
diated by the 1-phosphorylated derivative of (Z)-4-methyl-
sphingosine, which might act as a metabolically stable ana-
log of sphingosine-1-phosphate.[7,8] 3-C-Methyl-substituted
sphinganine has been prepared as a mixture of stereoiso-
mers and its effect on the incorporation of a metabolic pre-
cursor, [3-14C]-serine, into sphingolipids of neuronal cells
has been investigated.[9]
Ceramide (N-acylsphingosine 1, Scheme 1) is the hydro-
phobic membrane anchor of mammalian glycosphingolip-
ids and a vital component of the human skin.[1] In addition,
ceramide attracted attention as a signaling substance that
has been implicated in a variety of biological processes.[2,3]
In living cells, it usually occurs in low concentrations as an
intermediate in sphingolipid- and glycosphingolipid metab-
olism, and, in higher concentrations, in the epidermis of
human skin.[1] A variety of ceramide analogs has been pre-
pared and investigated as potential enzyme inhibitors or as
ligands of putative ceramide binding proteins.[1,4] We report
on the synthesis of 3-C-methyl-substituted ceramide deriva-
tives like 2a (Scheme 3) as conceptionally novel tools for
cell surface engineering.
We decided to prepare 3-C-methyl-substituted ceramides
as cell-permeable prodrugs of neoglycosphingolipids with a
structurally modified membrane anchor. The stereospecific
substitution of the hydrogen atom in position 3 of the
sphingoid backbone by a methyl group is a small, but struc-
turally defined modification. After addition of the target
compounds to cultured cells, we expect intracellular meta-
bolic conversion of some of the target molecules into mem-
brane neoglycosphingolipids,[4] which, in turn, might alter
the properties of membrane-resident proteins (‘‘cell surface
engineering’’). It has been demonstrated that some cer-
amide analogs can be metabolized to neoglycolipids (review,
see ref.[4]). More and more evidence is accumulating that
suggests that the properties of membrane proteins can be
drastically influenced by the composition of their lipid sur-
roundings.[10] Physiologically relevant data on the inter-
ference of membrane proteins with glycosphingolipids are
rare, but several examples indicate that the activity of phar-
macologically relevant receptors can be modified by glyco-
sphingolipids.[1] This has been most convincingly demon-
strated for the down-regulation of the insulin receptor by
ganglioside GM3 in ganglioside GM3-synthase knockout-
mice.[11] Therefore, it would be interesting to know if ad-
dition of the target compounds to cultured cells leads to
the formation of membrane glycosphingolipids modified in
the lipid moiety, and if functional membrane properties can
be influenced by this modification. In addition to this
modulation of cellular plasma membranes, therapeutic ap-
plications of ceramide analogs are discussed for the treat-
ment of several disease states.[1,4]
Scheme 1
Compared with naturally occurring ceramide, the target
compounds are structurally modified in the sphingosine
backbone. C-Methyl-substituted sphingosine derivatives
have been prepared before.[5] Interest in such derivatives re-
sulted from the observation that the 4-methyl derivative of
(Z)-sphingosine[5] inhibits sphingolipid biosynthesis and in-
duces morphological changes in neuronal cells treated with
this compound.[6] The structurally related sphingosine anal-
ogs (E)-4-methylsphingosine, (Z)-5-methylsphingosine, (E)-
[a]
´
Kekule-Institut für Organische Chemie und Biochemie,
Universität Bonn,
Gerhard-Domagk-Straße 1, 53121 Bonn, Germany
Fax: (internat.) ϩ 49-228-737778
E-mail: tkolter@uni-bonn.de
Eur. J. Org. Chem. 2004, 3693Ϫ3700
DOI: 10.1002/ejoc.200400114
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3693

P. Sawatzki, T. Kolter
FULL PAPER
To investigate the influence of the additional methyl Results
group in position 3 of the sphingoid backbone, we decided
A key intermediate in the synthesis of the target com-
to synthesize stereoisomerically pure 3-C-methylceramide pounds 2a,b, 12a,b, 13a,b, and 14a,b is alkyne 6, which re-
2a in the -erythro configuration of the naturally occurring sulted from the two carbonϪcarbon bond-formation reac-
lipid. For structure-activity relationships, we prepared the tions of the chosen synthetic route (Scheme 2). Aldehyde 3
target compound in the -threo configuration also (2b, was synthesized most efficiently according to a procedure
Scheme 3). An important role has been attributed to the reported for its (R) enantiomer[19] starting from -serine.
(4E)-double bond of ceramide, since 4,5-dihydroceramide The synthesis of 6 from Garner’s aldehyde 3 can alterna-
derivatives are much less active as signaling substances.[12] tively be achieved by two reaction sequences, which differ
Therefore, we decided to vary the nature of the C-4ϪC-5 by the order of the alkylation/alkynylation steps. Addition
bond in addition to the replacement of the hydrogen atom of a methyl Grignard reagent to aldehyde 3 afforded amino
at C-3 by the methyl group. Two series of ceramide analogs alcohol 4 in 80% yield as a mixture of diastereomers. Sub-
in -threo and -erythro configuration resulted from this sequent Swern oxidation led to the amino ketone 5 in 58%
additional modification of the C-4ϪC-5 bond. In addition yield. Compound 5 has been prepared before by an inde-
to 2a,b we prepared the target compounds also with pendent method.[20] An X-ray structure of the racemic com-
C-4ϪC-5 single bonds, with (Z)-configured double bonds pound is available.[21] Alkynylation of 5 afforded key inter-
(12a,b), and with triple bonds (13a,b). For the purpose of mediate 6 in 50% yield. Compound 6 was obtained as a
cell culture application, the target compounds were synthe- mixture of diastereomers, which was only separable on a
sized with an acyl chain length of 12 instead of the 18 car- preparative scale after further transformation (Schemes 3
bon atoms predominantly found in the natural product. and 4). The diastereomeric ratio of 2:1 is in agreement with
This leads to higher solubility in water and easier appli- the assumption of nonchelating reaction conditions and
cation to cultured cells.[4] Compared with previously pre- with published data on related reactions.[20]
pared sphinganine analogs,[9] we expect a lower degree of
toxicity of the acylated compounds and different intra- turned out to be less suitable in terms of overall yield: Alky-
cellular distribution. nylation of Garner’s aldehyde with an alkynyllithium com-
The alternative order of alkylation/alkynylation reactions
Several preparative approaches to sphingolipids have pound to 7,[22Ϫ24] and the subsequent Swern oxidation to
been reported (reviews: refs.[13Ϫ16]), which can be modified amino ketone 8 proceeded in satisfactory yield. However,
for the synthesis of sphingolipid analogs and derivatives.[17] the Grignard reaction leading to 6 was only achieved in low
For the purpose of this study, we decided to start with the yields. The diastereoselectivity of the reaction as deter-
configuratively stable serine aldehyde 3,[18] since the stereo- mined by NMR spectroscopy favors the anti-Felkin product
chemistry of the desired modification at the 3-position in in a 2.5:1 ratio and can be explained by the assumption of
the sphingosine backbone can be controlled, while the (2S) chelate control of the reaction.[20] The subsequent synthetic
configuration is easily preserved during the subsequent re- steps utilize 6, which has been prepared via the intermedi-
actions.
ates 4 and 5.
Scheme 2
3694
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 3693Ϫ3700

Syntheses of 3-C-Methylceramides
FULL PAPER
Scheme 3
For the synthesis of the (Z)-3-C-methylceramides 12a
and 12b (Scheme 3), alkyne 9 was hydrogenated in the pres-
ence of Lindlar catalyst to afford (Z)-3-C-methylsphingo-
sine (11) in 79% yield as a mixture of diastereomers. Acyl-
ation led to the target compounds, which were readily sepa-
rated by column chromatography, in 74% yield.
The 3-C-methylceramides 13a and 13b with a triple bond
in both configurations at C-3 were prepared by acylation of
the alkyne analog of 3-C-methylsphingosine (9, Scheme 4).
The separated diastereomers were independently subjected
to catalytic hydrogenation with palladium on charcoal in
methanol to give the saturated target compounds 14a and
14b. An alternative route uses the reversed order of acyl-
ation and hydrogenation, and is also shown in Scheme 4.
The compound 3-C-methylsphinganine (15) that is derived
from 9 by hydrogenation is an intermediate in this strategy.
Acylation of 15 and subsequent separation of the stereoiso-
mers gives access to the target compounds 14a and 14b.
The assignment of the relative configuration to the target
compounds was made after transformation of 13a and 13b
into the corresponding 1,3-benzylidene acetals 16a and 16b,
respectively (Scheme 5).
Scheme 4
For the synthesis of the target compounds 2a and 2b in
-erythro and -threo configurations, respectively, alkyne 6
was deprotected to 9 with HCl in THF[20] in 86% yield
(Scheme 3). Reduction of the alkyne 9 to the (E)-alkene
with lithium in ethylamine[25] led to 3-C-methylsphingosine
(10) as a mixture of diastereomers. Attempts for the
separation of the diastereomers on a preparative scale were
not successful. However, separation of the target com-
pounds 2a and 2b, which were derived from 10 by acylation
with lauroyl chloride at Ϫ78 °C in 66% yield, was easily
achieved. To avoid the formation of 3-C-methylceramide
esters as side products, we used methanol as solvent; for
complete acylation, 2 equiv. of lauroyl chloride were neces-
sary.
Scheme 5
Eur. J. Org. Chem. 2004, 3693Ϫ3700
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3695

P. Sawatzki, T. Kolter
FULL PAPER
Contacts in the NOE differential spectra between the Conclusion
protons of the 3-methyl group and the amide proton in 16a
We present a straightforward preparative approach to a
series of 3-C-methylceramide derivatives.
were used for the assignment of the relative erythro configu-
ration to this compound, while 16b showed a contact be-
tween the 3-methyl protons and 2-H, which can be expected
for a substance with a threo configuration. The assignment
of the relative configuration to the other target molecules
was achieved after their respective hydrogenation and com-
parison of their TLC mobilities with those of the hydrogen-
ated compounds derived from 13a and 13b. Compound 13a
was used to demonstrate the configurative integrity of the
target compounds. Enantiomeric purity was determined by
1H NMR spectroscopy using Eu(hfc)3 as chiral shift re-
agent. The amount of the racemic by-product was less
than 1%.
Experimental Section
General Remarks: Melting points are uncorrected. NMR shifts (δ)
are given in ppm. 1H NMR spectra: Bruker AM-400 (400 MHz)
instrument, temperature of measurement 303 K, solvent as internal
standard (CDCl3: δH ϭ 7.24 ppm). 13C NMR spectra: Bruker AM-
400 (100 MHz), temperature of measurement 303 K, solvent as in-
ternal standard (CDCl3: δC ϭ 77.0 ppm). For the assignment of
NMR signals to individual atoms within the C18 alkyl chain, sphin-
gosine numbering is used; 3-H of 2-aminooctadec-4-ene-1,3-diol is
replaced by a methyl group. FAB-mass spectra (MS): Kratos Con-
cept 1 H, matrix ϭ m-nitrobenzoic acid. Optical rotations:
PerkinϪElmer 343 Polarimeter. Column chromatography: silica gel
60 (E. Merck, Darmstadt, Germany), thin-layer chromatography:
silica gel plates 60, thickness 0.25 mm (E. Merck, Darmstadt, Ger-
many). Elemental analyses were performed in the microanalytical
Discussion
´
department of the Kekule-Institut für Organische Chemie und Bio-
chemie, Bonn.
We report on a preparative approach to ceramide analogs
bearing an additional methyl group at the 3-position of the
sphingoid backbone. The key step is the CϪC bond-for-
ming reaction with the addition of alkylmagnesium and al-
kynyllithium reagents to N-protected α-amino carbonyl
compounds 3, 5, and 8, without loss of optical purity.[26,28]
The synthesis afforded the target compound 2a in an ef-
ficient and straightforward manner (Schemes 2 and 3). It
also gives access to diastereomer 2b with the -threo con-
figuration and permits investigation of the role of the
stereochemistry at this position. Further modifications re-
ported in this manuscript refer to the C-4ϪC-5 double
bond, which has been reported to drastically alter the sig-
naling properties of ceramide.[12] The synthesis also allows
access to the 3-methyl-modified sphingosine derivatives 9,
10, 11, and 15, which were, however, only obtained as mix-
tures of diastereomers. Access to these compounds as pure
stereoisomers would require deacylation of the correspond-
ing separated ceramide derivatives 13a,b, 2a,b, 12a,b, and
14a,b.
tert-Butyl (4S)-4-[(1R/S)1-Hydroxyethyl]-2,2-dimethyl-1,3-oxazolid-
ine-3-carboxylate (4): Compound 3 [7.99 g, 34.85 mmol; prep-
aration according to the (R) enantiomer as reported by Campbell
et al.[19]] was dissolved in diethyl ether (100 mL) and stirred under
reflux. Methylmagnesium bromide (3  in diethyl ether, 35 mL,
0.105 mol) was added slowly, and the solution was subsequently
stirred for 18 h at room temperature. The solution was diluted with
water and extracted with dichloromethane. The organic layer was
separated, dried with sodium sulfate, and concentrated under re-
duced pressure. The residue was purified by column chromatogra-
phy on silica gel [RF ϭ 0.27 (cyclohexane/ethyl acetate, 7:3)] to give
6.80 g of 4 as a colorless oil. 1H NMR (400 MHz, CDCl3, mixture
of rotamers): δ ϭ 1.18Ϫ1.34 [m, 3 H, CH(OH)CH3], 1.48Ϫ1.56
(m, 12 H, CH3), 1.65 (m, br., 1 H, OH), 3.45Ϫ4.25 (m, 4 H) ppm.
13C NMR (100 MHz, CDCl3, major rotamer):
δ ϭ 17.8
[CH(OH)CH3], 27.0, 28.3 and 29.6 (CH3), 47.8 (C-4), 65.0 (C-5),
67.2 [CH(OH)CH3], 79.2 [C(CH3)3], 98.8 (C-2), 155.9 (NCO) ppm.
C12H23NO4 (245.32): calcd. C 58.75, H 9.45, N 5.71; found C
58.93, H 9.51, N 5.63. FAB-MS (70 eV): m/z calcd. for C12H24NO4
[M ϩ H]ϩ 246.2; found 246.1.
This series of structural ceramide analogs enables struc-
ture-activity analyses with compounds of chemically de-
fined structure. At room temperature, the NMR spectra of
compounds 4, 5, 6, and 8 are complex due to the presence
of rotamers, or diastereomeric mixtures and rotamers. For
compound 5, a less complex spectrum was obtained in
tert-Butyl (4S)-4-Acetyl-2,2-dimethyl-1,3-oxazolidine-3-carboxylate
(5): The experimental conditions[26] were the same as those de-
scribed for the synthesis of compound 8. RF ϭ 0.15 (cylohexane/
ethyl acetate, 10:1). Preparation: 3 (6.80 g, 27.74 mmol) in dichloro-
methane (120 mL), oxalyl chloride (3.58 mL, 41.56 mmol) in di-
chloromethane (90 mL), DMSO (6.07 mL, 83.12 mmol), diisoprop-
[D6]DMSO at 120 °C due to the coalescence of signals.[20] ylethylamine (29.05 mL, 166.25 mmol in 120 mL dichlorometh-
ane), 0.5  HCl (250 mL). Yield: 3.90 g; colorless oil. [αD] ϭ Ϫ53.8
(c ϭ 1.66, CHCl3) {ref.[19] ϩ55.7 (c ϭ 2.34, CHCl3) for the (R)
enantiomer}. 1H NMR (400 MHz, CDCl3, major rotamer): δ ϭ
1.34Ϫ1.61 (m, 15 H, CH3), 2.10 (m, 3 H, COCH3), 3.85 (dd, J ϭ
3.0, J ϭ 9.4 Hz, 1 H, 5-Ha), 4.07 (dd, J ϭ 7.6, J ϭ 9.4 Hz, 1 H, 5-
Hb), 4.23 (dd, J ϭ 3.0, J ϭ 7.6 Hz, 1 H, 4-H) ppm. 13C NMR
(100 MHz, CDCl3, major rotamer): δ ϭ 23.5(COCH3), 25.8, 26.1,
26.2, 28.1 and 28.1 (CH3), 64.9 (C-4), 65.4 (C-5), 80.6 [C(CH3)3],
94.8 (C-2), 151.1 (NCO), 206.1 (COCH3) ppm. C12H21O4N
(243.28): calcd. C 59.24, H 8.70, N 5.76; found C 59.11, H 8.76, N
5.90. FAB-MS (70 eV): m/z calcd. for C12H22O4N [M ϩ H]ϩ
The optical rotation determined for compound 5 is in excel-
lent agreement with the published data; however, no litera-
ture data are available for substance 8. Remarkably, after
storage for some weeks at 4 °C, complete racemization of 5
was observed, such that X-ray data could be obtained only
for the racemic compound.[21] Therefore, in agreement with
reports on related derivatives,[28] the storage of the N-pro-
tected α-amino carbonyl compounds described here should
be avoided in order to preserve their configurational integ-
rity.
3696
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 3693Ϫ3700

Syntheses of 3-C-Methylceramides
FULL PAPER
244.15; found 244.1. 5 is further characterized by X-ray data of the
4), 78.7 [C(CH3)3], 80.7 and 95.3 (alkyne-C), 98.2 (C-2), 151.3
(NCO), 186.0 (CO) ppm. C26H45NO4 (435.65): calcd. C 71.68, H
10.41, N 3.22; found C 71.41, H 10.47, N 4.29. FAB-MS (70 eV):
m/z calcd. for C26H46NO4 [M ϩ H]ϩ 436.7; found 436.7.
racemic compound.[21]
tert-Butyl (4S)-4-[(1R/S)-1-Hydroxy-1-methylhexadec-2-ynyl]-2,2-
dimethyl-1,3-oxazolidine-3-carboxylate (6): The experimental con-
ditions for the preparation of 6 from 5 were the same as those
described for the synthesis of compound 7. Preparation: 5 (2.53 g,
10.36 mmol) in THF (50 mL), 1-pentadecyne (2.41 mL,
14.55 mmol) in THF (100 mL), n-butyllithium (5.41 mL,
12.48 mmol, 2.3  in n-hexane). RF ϭ 0.25 (petroleum ether/ethyl
acetate, 10:1). Yield: 2.36 g as a mixture of diastereomers (-er-
ythro/-threo ϭ 2:1, according to the analysis of the 13C NMR
spectrum); colorless oil. The spectroscopic data correspond to
those of the product obtained by alkylation of 8. C27H49NO4ϫH2O
(469.69): calcd. C 69.05, H 10.94, N 2.98; found C 69.60, H 10.65,
N 3.27. FAB-MS (70 eV): m/z calcd. for C27H50NO4 [M ϩ H]ϩ
452.4; found 452.3.
(2S,3R/S)-2-Amino-3-methyloctadec-4-yne-1,3-diol (9): 6 (2.27 g,
5.02 mmol) was dissolved in THF/1  HCl (1:1; 50 mL) and stirred
at 80 °C for 18 h. After cooling to room temperature, the reaction
mixture is treated with 1  NaOH until pH ϭ 8Ϫ9. The organic
solvent was removed under reduced pressure and the residue ex-
tracted with ethyl acetate. The organic layer was dried with sodium
sulfate, concentrated under reduced pressure and the residue was
purified by column chromatography on silica gel; RF ϭ 0.33 (-
erythro) and 0.38 (-threo) (chloroform/methanol, 1:1) to give
1
1.35 g of a white solid. H NMR (400 MHz, CDCl3): δ ϭ 0.86 (t,
J ϭ 7 Hz, 3 H, CH3), 1.24 (m, 22 H, CH2), 1.46 (s, 3 H, 3-CH3),
2.17 (m, 2 H, CH2), 2.83 (m, 1 H, 2-H), 2.97 (m, br., 4 H), 3.66
(m, 1 H, 1-Ha), 3.84 (m, 1 H, 1-Hb) ppm. 13C NMR (100 MHz,
CDCl3, mixture of -erythro and -threo diastereomers, ratio 2:1):
δ ϭ 14.1 (CH3), 18.6Ϫ31.9 (CH2), 26.7 (3-CH3, -t), 27.3 (3-CH3,
-e), 60.1 (C-2, -t), 60.8 (C-2, -e), 62.7 (C-1, -t), 63.5 (C-1, -
e), 69.3 (C-3, -e), 70.4 (C-3, -t), 81.4 (-e), 82.1 (-t), 85.4 (-t),
86.0 (-e) (C-4 and C-5) ppm. C19H37NO2 (311.51): calcd. C 73.26,
H 11.97, N 4.50; found C 72.61, H 11.87, N 4.04. FAB-MS (70 eV):
m/z calcd. for C19H38NO2 [M ϩ H]ϩ 312.3; found 312.2.
tert-Butyl (4S)-4-[(1R/S)-1-Hydroxy-1-methylhexadec-2-ynyl]-2,2-
dimethyl-1,3-oxazolidine-3-carboxylate (6) from 8: Compound 6
was synthesized according to the procedure described for 4. Prep-
aration: 8 (98.1 mg, 1.60 mmol) in diethyl ether (20 mL), methyl-
magnesium bromide (1.6 mL, 4.80 mmol; 3  in diethyl ether).
RF ϭ 0.25 (petroleum ether/ethyl acetate, 10:1). Yield: 139.7 mg as
colorless oil. 6 was obtained as a mixture of diastereomers (-er-
ythro/-threo ϭ 2.5:1). 1H NMR (400 MHz, CDCl3, mixture of rot-
amers): δ ϭ 0.81 (t, J ϭ 7 Hz, 3 H, CH3), 1.19 (m, 22 H, CH2),
1.30Ϫ1.66 (m, 18 H, CH3), 2.11 (m, 2 H, CH2), 3.57Ϫ4.18 (m, 2
H), 4.45 (m, br., 1 H, OH), 4.50Ϫ5.23 (m, 1 H) ppm. 13C NMR
(100 MHz, CDCl3, major rotamer): δ ϭ 14.1 (CH3), 18.7Ϫ31.9
(CH2), 21.86, 27.5, 27.8, 28.3, 29.7 and 31.2 (CH3), 51.3 (C-4),
62.5 (C-5), 67.0 (CϪOH), 79.5 [C(CH3)3], 81.5 and 95.2 (alkyne-
C), 100.0 (C-2), 155.6 (NCO) ppm. C27H49NO4 (451.69): calcd. C
71.80, H 10.93, N 3.10; found C 71.36, H 10.44, N 3.49. FAB-MS
(70 eV): m/z calcd. for C27H50NO4 [M ϩ H]ϩ 452.6; found 452.3.
(2S,3R/S,4E)-2-Amino-3-methyloctadec-4-ene-1,3-diol
(10):
9
(339.8 mg, 1.09 mmol) was dissolved in THF (25 mL) and added
slowly by syringe to a mixture of ethylamine (15 mL) and lithium
(117.3 mg, 16.90 mmol; deep blue solution) at Ϫ20 °C. Stirring was
continued overnight, and the solution was allowed to warm up to
room temperature. Solid ammonium chloride (3 g) was added, and
the reaction mixture was concentrated under reduced pressure. The
solid residue was extracted with a mixture of water and diethyl
ether, the organic phase was dried with sodium sulfate, filtered and
concentrated under reduced pressure. The residue was purified by
column chromatography on silica gel, RF ϭ 0.58 (-erythro) and
0.62 (-threo) (chloroform/methanol, 1:1), to afford 213.3 mg of a
tert-Butyl (4S)-4-[(1R/S)-1-Hydroxyhexadec-2-ynyl]-2,2-dimethyl-
1,3-oxazolidine-3-carboxylate (7):
7 was synthesized from 3
(593.5 mg, 2.59 mmol) according to ref.[23] The analytical data are
1
white solid. H NMR (400 MHz, CDCl3): δ ϭ 0.85 (t, J ϭ 7 Hz,
in agreement with those published. Yield: 862.2 mg of a colorless
oil.
3 H, CH3), 1.22 (m, 22 H, CH2), 1.44 (s, 3 H, 3-CH3), 2.01 (m, 2
H, CH2), 3.47 (m, 1 H, 2-H), 3.77 (md, J ϭ 11.5 Hz, 1 H, 1-Ha),
3.89 (md, J ϭ 11.5 Hz, 1 H, 1-Hb), 5.42 (d, J ϭ 15.3 Hz, 1 H, 4-
H), 5.79 (td, J ϭ 7, J ϭ 15.3 Hz, 1 H, 5-H) ppm. 13C NMR
(100 MHz, CDCl3, mixture of -erythro and -threo diastereomers;
only the chemical shifts for the -erythro diastereomer are given):
δ ϭ 14.1 (CH3), 22.7Ϫ32.4 (CH2), 26.8 (3-CH3), 59.6 (C-2), 60.9
(C-1), 72.3 (C-3), 130.5 and 132.1 (C-4 and C-5) ppm. C19H39NO2
(313.52): calcd. C 72.79, H 12.54 N, 4.47; found C 72.42, H 11.98,
N 4.18. FAB-MS (70 eV): m/z calcd. for C19H40NO2 [M ϩ H]ϩ
314.3; found 314.3.
tert-Butyl (4S)-2,2-Dimethyl-4-(pent-2-ynoyl)-1,3-oxazolidine-3-car-
boxylate (8): Procedure according to ref.[26] Dimethyl sulfoxide (260
µL, 3.94 mmol) was added to a solution of oxalyl chloride (170 µL,
1.97 mmol) in dichloromethane (3 mL) at Ϫ78 °C for 20 min. The
solution was allowed to warm up to Ϫ60 °C, and 7 (863.5 mg,
1.97 mmol), dissolved in dichloromethane (10 mL), was added
dropwise. After warming up to Ϫ45 °C, diisopropylethylamine
(1.37 mL, 7.88 mmol), dissolved in dichloromethane (6 mL), was
added. The cooling bath was removed, and the mixture was allowed
to warm up to 0 °C. After quenching with cooled (0 °C) 0.5  HCl
(15 mL), the solution was diluted with water and extracted with
dichloromethane. The organic layer was dried with sodium sulfate,
and the solvent was evaporated under reduced pressure. The crude
product was purified by column chromatography; RF ϭ 0.32
(2S,3R/S,4Z)-2-Amino-3-methyloctadec-4-ene-1,3-diol (11): The ex-
perimental conditions corresponded to those used in the synthesis
of 15. Product formation was carefully controlled by TLC to avoid
over-reduction to the saturated compound. Experimental details: 9
(217.4 mg, 697.9 µmol), catalytic amount of Lindlar catalyst (5%
Pd on CaCO3, poisoned with lead, Fluka), methanol (10 mL).
(petroleum ether/ethyl acetate, 10:1) to give 697.1 mg of 8 as a col-
1
orless oil. [αD] ϭ Ϫ52.0 (c ϭ 2.02, CHCl3). H NMR (400 MHz, RF ϭ 0.14 (chloroform/methanol, 1:1) (-erythro and -threo
CDCl3, major rotamer): δ ϭ 0.81 (t, J ϭ 7 Hz, 3 H, CH3), 1.19
form). Yield: 173.3 mg as a white solid. 1H NMR (400 MHz,
(m, 22 H, CH2), 1.35 (s, 3 H, CH3), 1.44 (s, 3 H, CH3), 1.48 (s, 3 CDCl3): δ ϭ 0.81 (t, J ϭ 7 Hz, 3 H, CH3), 1.19 (m, 22 H, CH2),
H, CH3), 1.63 (s, 3 H, CH3), 2.30 (m, 2 H, CH2), 4.00 (dd, J ϭ 1.27 (s, 3 H, 3-CH3), 2.22 (m, 2 H, CH2), 2.74 (dd J ϭ 3.5, J ϭ
3.7, J ϭ 9.4 Hz, 1 H, H-5a), 4.10 (dd, J ϭ 7.4, J ϭ 9.4 Hz, 1 H, 6.6 Hz, 1 H, 2-H), 2.88 (m, br., 4 H), 3.58 (dd, J ϭ 6.6, J ϭ 10.8 Hz,
H-5b), 4.29 (dd, J ϭ 3.7, J ϭ 7.4 Hz, 1 H, H-4) ppm. 13C NMR 1 H, 1-Ha), 3.70 (dd, J ϭ 3.5, J ϭ 10.8 Hz, 1 H, 1-Hb), 5.23 (d,
(100 MHz, CDCl3, major rotamer): δ ϭ 14.1 (CH3), 19.1Ϫ31.9 J ϭ 12.1 Hz, 1 H, 4-H), 5.35 (td, J ϭ 7.4, J ϭ 12.1 Hz, 1 H, 5-H)
(CH2), 24.3, 25.1, 25.2, 26.2 and 27.6 (CH3), 65.8 (C-5), 66.7 (C-
ppm. 13C NMR (100 MHz, CDCl3, mixture of -erythro and -
Eur. J. Org. Chem. 2004, 3693Ϫ3700
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3697

P. Sawatzki, T. Kolter
threo diastereomers, ratio 2:1): δ ϭ 14.1 (CH3), 22.7Ϫ31.9 (CH2), 174.1 (NCO) ppm. C31H59NO3 (493.81): calcd. C 75.40, H 12.04,
FULL PAPER
25.4 (3-CH3, -t), 26.7 (3-CH3, -e), 59.5 (C-2, -t), 59.9 (C-2, -
N 2.84; found C 74.67, H 11.96, N 2.57. FAB-MS (70 eV): m/z
e), 62.7 (C-1, -t), 63.5 (C-1, -e), 75.4 (C-3, -t), 75.8 (C-3, -e), calcd. for C31H59NO3 [Mϩ] 493.45; found 494.5 [M ϩ Hϩ].
132.5 (-e), 132.7 (-e), 133.2 (-t), 133.9 (-t) (C-4 and C-5) ppm.
N-[(1S,2S)-2-Hydroxy-1-(hydroxymethyl)-2-methylheptadec-3-
ynyl]dodecanamide (13b): RF ϭ 0.16 (chloroform/methanol, 30:1).
C19H39NO2 (313.52): calcd. C 72.79, H 12.54 N, 4.47; found C
71.99,
H 12.45, N 4.32. FAB-MS (70 eV): m/z calcd. for
[αD] ϭ Ϫ20.9 (c ϭ 1.02, CHCl3). M.p. 42 °C. 1H NMR (400 MHz,
CDCl3): δ ϭ 0.88 (t, J ϭ 7 Hz, 6 H, CH3), 1.25 (m, 38 H, CH2),
1.51 (s, 3 H, 3-CH3), 1.63 (m, 2 H, CH2), 2.19 (t, J ϭ 7 Hz, 2 H,
CH2), 2.24 (t, J ϭ 7.5 Hz, 2 H, CH2), 3.32 (m, br., 2 H, OH), 3.89
(m, 2 H, 1-HA and 1-HB), 4.06 (ddd, J ϭ 4.7, J ϭ 5, J ϭ 8.5 Hz, 1
H, 2-H), 6.00 (d, J ϭ 8.5 Hz, 1 H, NH) ppm. 13C NMR (100 MHz,
CDCl3): δ ϭ 14.1 (CH3), 18.6Ϫ37.0 (CH2), 30.3 (3-CH3), 57.3 (C-
2), 63.0 (C-1), 70.8 (C-3), 81.3 (C-5), 86.1 (C-4), 173.8 (NCO) ppm.
C31H59NO3 (493.81): calcd. C 75.40, H 12.04, N 2.84; found C
C19H40NO2 [M ϩ H]ϩ 314.3; found 314.4.
N-[(1S,2R,3Z)-2-Hydroxy-1-(hydroxymethyl)-2-methylheptadec-3-
enyl]dodecanamide (12a): The experimental conditions were the
same as those described for the synthesis of compound 13a. Prep-
aration: 11 (mixture of diastereomers, 159.0 mg, 507.1 µmol), tri-
ethylamine (210 µL, 1.52 mmol), lauroyl chloride (240 µL,
1.01 mmol), methanol (15 mL). RF ϭ 0.26 (chloroform/methanol,
30:1). The two diastereomers 12a and 12b were readily separated
in this workup procedure to give 122.5 mg of 12a and 65.3 mg of
12b as white solids. [αD] ϭ Ϫ10.2 (c ϭ 0.16, CHCl3). M.p. 44Ϫ45
°C. 1H NMR (400 MHz, CDCl3): δ ϭ 0.81 (t, J ϭ 7 Hz, 6 H, CH3),
1.20 (m, 38 H, CH2), 1.36 (s, 3 H, 3-CH3), 1.58 (m, 2 H, CH2),
2.18 (m, 4 H, CH2), 3.59 (m, br., 1 H, OH), 3.65 (dd, J ϭ 2.5, J ϭ
10.8 Hz, 1 H, 1-Ha), 3.70 (m, br., 1 H, OH), 3.86 (ddd, J ϭ 2.5,
J ϭ 3.0, J ϭ 9.1 Hz, 1 H, 2-H), 3.98 (dd, J ϭ 3.0, J ϭ 10.8 Hz, 1
H, 1-Hb), 5.35 (md, J ϭ 12.1 Hz, 2 H, 4-H and 5-H), 6.54 (d, J ϭ
9.1 Hz, 1 H, NH) ppm. 13C NMR (100 MHz, CDCl3): δ ϭ 14.1
(CH3), 22.7Ϫ37.0 (CH2), 27.8 (3-CH3), 55.8 (C-2), 63.7 (C-1), 77.4
(C-3), 132.4 (C-5), 133.4 (C-4), 174.1 (NCO) ppm. C31H61NO3
(495.83): calcd. C 75.09, H 12.40 N, 2.83; found C 74.96, H 12.42,
N 2.59. FAB-MS (70 eV): m/z calcd. for C31H62NO3 [M ϩ H]ϩ
496.5; found 496.4.
75.21,
H 11.91, N 2.57. FAB-MS (70 eV): m/z calcd. for
C31H60NO3 [M ϩ H]ϩ 494.45; found 494.4.
N-[(1S,2R)-2-Hydroxy-1-(hydroxymethyl)-2-methylheptadecyl]-
dodecanamide (14a): Route 1: 13a (154.3 mg, 312.4 µmol) was dis-
solved in methanol (10 mL) and degassed with stirring. A catalytic
amount of palladium on charcoal (10%) was added, and the solu-
tion was stirred under hydrogen for 18 h. The reaction mixture was
filtered, and the solvent was evaporated under reduced pressure.
The remaining residue was purified by column chromatography on
silica gel, RF ϭ 0.16 (chloroform/methanol, 30:1), to give 114.0 mg
of 14a as a white solid. Route 2: The synthesis was analogous to
the synthesis of 13a. Experimental details: 13a (mixture of dia-
stereomers, 97.1 mg, 307.7 µmol), triethylamine (130 µL, 923.2
µmol), lauroyl chloride (150 µL, 615.5 µmol), and methanol
(10 mL). The two diastereomers 14a and 14b were readily separated
in this workup procedure to give 97.1 mg of 14a and 35.4 mg of
14b as white solids. [αD] ϭ ϩ21.5 (c ϭ 0.38, CHCl3). M.p. 70Ϫ71
°C. 1H NMR (400 MHz, CDCl3): δ ϭ 0.81 (t, J ϭ 7 Hz, 6 H, CH3),
1.06 (s, 3 H, 3-CH3), 1.19 (m, 40 H, CH2), 1.56 (m, 2 H, CH2),
2.17 (t, J ϭ 7.5 Hz, 2 H, CH2), 3.36 (m, br., 2 H, OH), 3.68 (dd,
J ϭ 2, J ϭ 11.1 Hz, 1 H, 1-HA), 3.75 (td, J ϭ 2, J ϭ 8.9 Hz, 1 H,
2-H), 3.92 (dd, J ϭ 2.5, J ϭ 11.1 Hz, 1 H, 1-HB), 6.54 (d, J ϭ
8.4 Hz, 1 H, NH) ppm. 13C NMR (100 MHz, CDCl3): δ ϭ 14.1
(CH3), 22.7Ϫ40.4 (CH2), 24.2 (3-CH3), 55.0 (C-2), 62.9 (C-1), 75.5
(C-3), 174.0 (NCO) ppm. C31H63NO3 (497.85): calcd. C 74.79, H
12.76 N, 2.81; found C 74.75, H 12.65, N 2.48. FAB-MS (70 eV):
m/z calcd. for C31H64NO3 [M ϩ H]ϩ 498.5; found 498.4.
N-[(1S,2S,3Z)-2-Hydroxy-1-(hydroxymethyl)-2-methylheptadec-3-
enyl]dodecanamide (12b): RF ϭ 0.16 (chloroform/methanol, 30:1).
[αD] ϭ Ϫ13.8 (c ϭ 0.86, CHCl3). M.p. 36Ϫ38 °C. 1H NMR
(400 MHz, CDCl3): δ ϭ 0.81 (t, J ϭ 7 Hz, 6 H, CH3), 1.19 (m, 42
H, CH2), 1.39 (s, 3 H, 3-CH3), 2.14 (t, J ϭ 8 Hz, 2 H, CH2), 3.47
(m, br., 2 H, OH), 3.75 and 3.82 (m, 2 H, 1-Ha and 2-H), 3.94 (dd,
J ϭ 3.4, J ϭ 11.3 Hz, 1 H, 1-Hb), 5.20 (d, J ϭ 12.1 Hz, 2 H, 4-H),
5.31 (td, J ϭ 7, J ϭ 12.1 Hz, 1 H, 5-H), 6.33 (d, J ϭ 8.1 Hz, 1 H,
NH) ppm. 13C NMR (100 MHz, CDCl3): δ ϭ 14.1 (CH3),
22.7Ϫ37.0 (CH2), 26.9 (3-CH3), 57.3 (C-2), 62.9 (C-1), 76.8 (C-3),
132.9 (C-5), 133.1 (C-4), 174.0 (NCO) ppm. C31H61NO3 (495.83):
calcd. C 75.09, H 12.40 N, 2.83; found C 74.77, H 12.52, N 2.59.
FAB-MS (70 eV): m/z calcd. for C31H62NO3 [M ϩ H]ϩ 496.5;
found 496.4.
N-[(1S,2S)-2-Hydroxy-1-(hydroxymethyl)-2-methylheptadecyl]-
dodecanamide (14b): The experimental conditions were the same as
those described for the synthesis of 14a. Route 1: Experimental
details: 13b (99.2 mg, 200.9 µmol), catalytic amount of palladium
on charcoal (10%), and methanol (10 mL). RF ϭ 0.15 (chloroform/
N-[(1S,2R)-2-Hydroxy-1-(hydroxymethyl)-2-methylheptadec-3-
ynyl]dodecanamide (13a): 9 (470.3 mg, 1.51 mmol) and triethyl-
amine (630 µL, 4.53 mmol) were dissolved in methanol (15 mL)
and cooled to Ϫ78 °C. To this stirred solution, lauroyl chloride (720
µL, 3.02 mmol) was added dropwise and stirring was continued for
18 h, which allowed the temperature to raise to room temperature.
The solution was concentrated under reduced pressure, and the
residue was purified by column chromatography on silica gel; RF ϭ
0.26 (chloroform/methanol, 30:1). The two diastereomers 13a and
13b were readily separated in this workup procedure to give
372.6 mg of 13a and 185.3 mg of 13b as white solids. [αD] ϭ ϩ20.9
1
methanol, 30:1). [αD] ϭ ϩ13.8 (c ϭ 0.44, CHCl3). M.p. 67 °C. H
NMR (400 MHz, CDCl3): δ ϭ 0.87 (t, J ϭ 7 Hz, 6 H, CH3), 1.24
(m, 40 H, CH2), 1.29 (s, 3 H, 3-CH3), 1.64 (m, 2 H, CH2), 2.22 (t,
J ϭ 7.5 Hz, 2 H, CH2), 3.32 (m, br., 2 H, OH), 3.79 (m, 2 H, 1-
HA and 1-HB), 3.98 (md, J ϭ 8.6 Hz, 1 H, 2-H), 6.49 (d, J ϭ
8.1 Hz, 1 H, NH) ppm. 13C NMR (100 MHz, CDCl3): δ ϭ 14.1
(CH3), 22.7Ϫ40.0 (CH2), 24.5 (3-CH3), 54.7 (C-2), 63.5 (C-1), 75.3
(C-3), 173.8 (NCO) ppm. C31H63NO3 (497.85): calcd. C 74.79, H
12.76 N, 2.81; found C 74.98, H 12.70, N 2.54. FAB-MS (70 eV):
m/z calcd. for C31H64NO3 [M ϩ H]ϩ 498.5; found 498.4.
1
(c ϭ 0.57, CHCl3). M.p. 54Ϫ55 °C. H NMR (400 MHz, CDCl3):
δ ϭ 0.86 (t, J ϭ 7 Hz, 6 H, CH3), 1.24 (m, 36 H, CH2), 1.43 (s, 3
H, 3-CH3), 1.47 (m, 2 H, CH2), 1.63 (m, 2 H, CH2), 2.18 (t, J ϭ
7.5 Hz, 2 H, CH2), 2.24 (t, J ϭ 7.5 Hz, 2 H, CH2), 3.20 (m, br., 2
H, OH), 3.78 (dd, J ϭ 3.2, J ϭ 11.1 Hz, 1 H, 1-HA), 3.96 (ddd,
(2S,3R/S)-2-Amino-3-methyloctadecane-1,3-diol (15): The exper-
J ϭ 3.2, J ϭ 3.4, J ϭ 9.1 Hz, 1 H, 2-H), 4.29 (dd, J ϭ 3.4, J ϭ imental conditions were the same as those described for the syn-
11.1 Hz, 1 H, 1-HB), 6.41 (d, J ϭ 9.1 Hz, 1 H, NH) ppm. 13C NMR
(100 MHz, CDCl3): δ ϭ 14.1 (CH3), 18.7Ϫ37.0 (CH2), 28.6 (3-
CH3), 56.4 (C-2), 64.2 (C-1), 70.7 (C-3), 81.7 (C-5), 86.6 (C-4),
thesis of compound 14. Experimental details: 9 (115.4 mg, 370.4
µmol), catalytic amount of palladium on charcoal (10%), and
methanol (10 mL). RF ϭ 0.14 (chloroform/methanol, 1:1), -er-
3698
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 3693Ϫ3700

Syntheses of 3-C-Methylceramides
FULL PAPER
ythro and -threo form. Yield: 112.3 mg of a white solid. 1H NMR
1.36 mmol), and methanol (5 mL). RF: 0.26 (chloroform/methanol,
(400 MHz, CDCl3): δ ϭ 0.81 (t, J ϭ 7 Hz, 3 H, CH3), 1.19 (m, 28 30:1). The two diastereomers 2a and 2b were readily separated in
H, CH2), 1.36 (s, 3 H, 3-CH3), 2.73 (m, 1 H, 2-H), 2.95 (m, br., 4 this workup procedure to give 122.5 mg of 2a and 65.3 mg of 2b as
H), 3.57 (dd, J ϭ 7, J ϭ 10.5 Hz, 1 H, 1-HA), 3.70 (dd, J ϭ 3, J ϭ white solids. [αD] ϭ ϩ5.5 (c ϭ 1.13, CHCl3). M.p. 75 °C. 1H NMR
10.5 Hz, 1 H, 1-HB), 5.23 (d, J ϭ 12.1 Hz, 1 H, 4-H) ppm. 13C
NMR (100 MHz, CDCl3; mixture of -erythro and -threo dia-
stereomers; only the chemical shifts for the -erythro diastereomer
(400 MHz, CDCl3): δ ϭ 0.87 (t, J ϭ 7 Hz, 6 H, CH3), 1.25 (m, 38
H, CH2), 1.31 (s, 3 H, 3-CH3), 1.65 (m, 2 H, CH2), 2.06 (m, 2 H,
CH2), 2.26 (t, J ϭ 7.5 Hz, 2 H, CH2), 2.70 (m, br., 2 H, OH), 3.68
are given): δ ϭ 14.1 (CH3), 22.5Ϫ32.0 (CH2), 24.2 (3-CH3), 38.7 (dd, J ϭ 2.7, J ϭ 11.1 Hz, 1 H, 1-HA), 3.85 (ddd, J ϭ 2.7, J ϭ 3.0,
(C-4), 59.0 (C-2), 62.6 (C-1), 73.7 (C-3) ppm. C19H41NO2 (315.54): J ϭ 8.6 Hz, 1 H, 2-H), 3.98 (dd, J ϭ 3.0, J ϭ 11.1 Hz, 1 H, 1-HB),
calcd. C 72.32, H 13.10 N, 4.44; found C 72.23, H 12.95, N 4.07.
5.54 (d, J ϭ 15.5 Hz, 1 H, 4-H), 5.77 (td, J ϭ 7, J ϭ 15.5 Hz, 1 H,
FAB-MS (70 eV): m/z calcd. for C19H42NO2 [M ϩ H]ϩ 316.3; 5-H), 6.40 (d, J ϭ 8.6 Hz, 1 H, NH) ppm. 13C NMR (100 MHz,
found 316.3.
CDCl3): δ ϭ 14.1 (CH3), 22.7Ϫ37.0 (CH2), 26.8 (3-CH3), 55.6 (C-
2), 63.7 (C-1), 76.3 (C-3), 129.8 (C-5), 134.0 (C-4), 173.8 (NCO)
ppm. C31H61NO3 (495.83): calcd. C 75.10, H 12.40 N, 2.83; found
C 74.67, H 12.11, N 2.33. FAB-MS (70 eV): m/z calcd. for
C31H62NO3 [M ϩ H]ϩ 496.5; found 496.4.
N-[(4R,5S)-4-Methyl-2-phenyl-4-(tridec-1-ynyl)-1,3-dioxan-5-yl]-
dodecanamide (16a): The synthesis follows the procedure reported
by Brockway et al.[27] 13a (100.3 mg, 203.0 µmol), p-toluenesulfonic
acid monohydrate (16 mg, 83 µmol), and benzaldehyde dimethyl
acetal (60 µL, 368.6 µmol) were dissolved in DMF (5 mL) and
stirred at room temperature for 4d. Water (4 °C) was added, and
the mixture was extracted with ethyl acetate. The organic layer was
dried with sodium sulfate, filtered and concentrated under reduced
pressure. The remaining residue was purified by column chroma-
tography on silica gel, RF ϭ 0.33 (petroleum ether/ethyl acetate,
10:1), to give 11.7 mg of a white solid. [αD] ϭ ϩ36.4 (c ϭ 0.86,
N-[(1S,2S,3E)-2-Hydroxy-1-(hydroxymethyl)-2-methylheptadec-3-
enyl]dodecanamide (2b): RF ϭ 0.13 (chloroform/methanol, 30:1).
[αD] ϭ Ϫ4.6 (c ϭ 0.69, CHCl3). M.p. 74Ϫ75 °C. 1H NMR
(400 MHz, CDCl3): δ ϭ 0.88 (t, J ϭ 7 Hz, 6 H, CH3), 1.25 (m, 40
H, CH2, OH), 1.38 (s, 3 H, 3-CH3), 1.60 (m, 2 H, CH2), 2.01 (m,
2 H, CH2), 2.19 (td, J ϭ 2, J ϭ 7 Hz, 2 H, CH2), 3.82 (m, 1 H, 2-
H), 3.85 (dd, J ϭ 3.0, J ϭ 11.1 Hz, 1 H, 1-HA), 3.99 (dd, J ϭ 3.0,
J ϭ 11.1 Hz, 1 H, 1-HB), 5.44 (d, J ϭ 15.5 Hz, 1 H, 4-H), 5.67 (td,
J ϭ 7, J ϭ 15.5 Hz, 1 H, 5-H), 6.18 (d, J ϭ 7.4 Hz, 1 H, NH) ppm.
13C NMR (100 MHz, CDCl3): δ ϭ 14.1 (CH3), 22.7Ϫ37.0 (CH2),
26.6 (3-CH3), 56.4 (C-2), 63.1 (C-1), 75.3 (C-3), 129.8 (C-5), 134.0
(C-4), 173.7 (NCO) ppm. C31H61NO3ϫ1/2H2O (404.83): calcd. C
73.55, H 12.18 N, 2.78; found C 74.03, H 12.21, N 2.52. FAB-MS
(70 eV): m/z calcd. C31H62NO3 [M ϩ H]ϩ 496.5; found 496.4.
1
CHCl3). M.p. 52Ϫ53 °C. H NMR (400 MHz, CDCl3): δ ϭ 0.81
(t, J ϭ 7 Hz, 6 H, CH3), 1.18 (m, 34 H, CH2), 1.34 (m, 2 H, CH2),
1.41 (s, 3 H, 3-CH3), 1.48 (m, 2 H, CH2), 1.58 (m, 2 H, CH2), 2.17
(t, J ϭ 7.5 Hz, 2 H, CH2), 2.22 (t, J ϭ 7 Hz, 2 H, CH2), 3.58 (dd,
J ϭ 1.7, J ϭ 11.6 Hz, 1 H, 1-HA), 4.01 (td, J ϭ 2, J ϭ 10 Hz, 1
H, 2-H), 4.55 (dd, J ϭ 1.5, J ϭ 11.6 Hz, 1 H, 1-HB), 6.04 (s, 1 H,
ArCHO2); 6.18 (d, J ϭ 10 Hz, 1 H, NH), 7.31 (m, 3 H, Ar-H),
7.43 (m, 2 H, Ar-H) ppm. 13C NMR (100 MHz, CDCl3): δ ϭ 14.1
(CH3), 18.7Ϫ37.0 (CH2), 25.7 (3-CH3), 49.7 (C-2), 69.1 (C-1), 73.3
(C-3), 78.7 and 90.4 (C-4 and C-5), 97.0 (ArCHO2), 126.0, 128.3,
129.0 and 138.0 (Ar-C), 172.9 (NCO) ppm. C38H63NO3ϫ1/2H2O
(590.92): calcd. C 77.17, H 10.83 N, 2.37; found C 77.82, H 11.07,
N 2.37. FAB-MS (70 eV): m/z calcd. for C38H63NO3 [M ϩ H]ϩ
582.5; found 582.4.
Acknowledgments
We thank Prof. Dr. K. Sandhoff, Bonn, the Deutsche Forschungs-
gemeinschaft (SFB 284), and the Fonds der Chemischen Industrie
for the support of this work.
N-[(4S,5S)-4-Methyl-2-phenyl-4-(tridec-1-ynyl)-1,3-dioxan-5-yl]-
dodecanamide (16b): The experimental conditions were the same as
those described for the synthesis of 16a. Procedure: 13b (86.1 mg,
173.5 µmol), p-toluenesulfonic acid monohydrate (13.6 mg, 71.4
µmol), benzaldehyde dimethyl acetal (50 µL, 317.4 µmol), and
DMF (5 mL). RF ϭ 0.27 (petroleum ether/ethyl acetate, 10:1).
Yield: 52.2 mg as a white solid. [αD] ϭ Ϫ32.7 (c ϭ 2.48, CHCl3).
[1]
T. Kolter, K. Sandhoff, Angew. Chem. 1999, 111, 1633Ϫ1670,
Angew. Chem. Int. Ed. 1999, 38, 1532Ϫ1568.
[2]
Y. A. Hannun, L. M. Obeid, J. Biol. Chem. 2002, 277,
25847Ϫ25850.
W. J. van Blitterswijk, A. H. van der Luit, R. J. Veldman, M.
[3]
Verheij, J. Borst, Biochem. J. 2003, 369, 199Ϫ211.
1
M.p. 44 °C. H NMR (400 MHz, CDCl3): δ ϭ 0.81 (t, J ϭ 7 Hz,
[4]
S. Brodesser, P. Sawatzki, T. Kolter, Eur. J. Org. Chem. 2003,
6 H, CH3), 1.19 (m, 34 H, CH2), 1.37 (m, 2 H, CH2), 1.49 (s, 3 H,
3-CH3), 1.55 (m, 4 H, CH2), 2.14 (t, J ϭ 7.5 Hz, 2 H, CH2), 2.27
(t, J ϭ 7 Hz, 2 H, CH2), 3.63 (t, J ϭ 10.8 Hz, 1 H, 1-HA), 4.00
(dd, J ϭ 5.2, J ϭ 10.8 Hz, 1 H, 1-HB), 4.20 (ddd, J ϭ 5.2, J ϭ
10.6, J ϭ 10.8 Hz, 1 H, 2-H), 5.28 (d, J ϭ 10 Hz, 1 H, NH), 5.84
(s, 1 H, ArCHO2), 7.28 (m, 3 H, Ar-H), 7.42 (m, 2 H, Ar-H) ppm.
13C NMR (100 MHz, CDCl3): δ ϭ 14.1 (CH3), 18.7Ϫ37.0 (CH2),
27.00 (3-CH3), 49.2 (C-2), 66.8 (C-1), 74.6 (C-3), 77.1 and 90.8 (C-
4 and C-5), 96.9 (ArCHO2), 126.4, 128.3, 129.0 and 137.7 (Ar-C),
172.8 (NCO) ppm. C38H63NO3ϫ1/2H2O (590.92): calcd. C 77.17,
H 10.83 N, 2.37; found C 77.87, H 10.98, N 2.48. FAB-MS (70 eV):
m/z calcd. for C38H64NO3 [M ϩ H]ϩ 582.5; found 582.4.
2021Ϫ2034.
[5]
T. Bär, B. Kratzer, R. Wild, K. Sandhoff, R. R. Schmidt, Lie-
bigs Ann. Chem. 1993, 419Ϫ426.
G. van Echten-Deckert, A. Zschoche, T. Bär, R. R. Schmidt,
A. Raths, T. Heinemann, K. Sandhoff, J. Biol. Chem. 1997,
272, 15825Ϫ15833.
G. van Echten-Deckert, A. Schick, T. Heinemann, B. Schnied-
ers, J. Biol. Chem. 1998, 273, 23585Ϫ23589.
S. Nätzker, T. Heinemann, S. Figueroa-Perez, B. Schnieders, R.
R. Schmidt, K. Sandhoff, G. van Echten-Deckert, Biol. Chem.
2002, 383, 1885Ϫ1894.
T. Kolter, G. van Echten-Deckert, K. Sandhoff, Tetrahedron
[6]
[7]
[8]
[9]
1994, 50, 13425Ϫ13432.
[10]
´
M. Opekarova, W. Tanner, Biochim. Biophys. Acta 2003,
N-[(1S,2R,3E)-2-Hydroxy-1-(hydroxymethyl)-2-methylheptadec-3-
enyl]dodecanamide (2a): The experimental conditions are the same
as those described for the synthesis of compound 13a. Experimen-
tal details: 10 (mixture of diastereomers, 213.3 mg, 680.2 µmol),
triethylamine (280 µL, 2.04 mmol), lauroyl chloride (320 µL,
1610, 11Ϫ22.
[11]
T. Yamashita, A. Hashiramoto, M. Haluzik, H. Mizukami, S.
Beck, A. Norton, M. Kono, S. Tsuji, J. L. Daniotti, N. Werth,
R. Sandhoff, K. Sandhoff, R. L. Proia, Proc. Natl. Acad. Sci.
U. S. A. 2003, 100, 3445Ϫ3449.
Eur. J. Org. Chem. 2004, 3693Ϫ3700
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3699

P. Sawatzki, T. Kolter
FULL PAPER
[12]
[21]
A. Bielawska, H. M. Crane, D. Liotta, L. M. Obeid, Y. A.
P. Sawatzki, T. Mikeska, M. Nieger, H. Hupfer, T. Kolter, Acta
Crystallogr., Sect. E 2002, 58, o359Ϫo361.
P. Herold, Helv. Chim. Acta 1988, 71, 354Ϫ362.
P. Garner, J. M. Park, E. Malecki, J. Org. Chem. 1988, 53,
4395Ϫ4397.
H. Radunz, R. M. Devant, V. Eiermann, Liebigs Ann. Chem.
1988, 1103Ϫ1105.
R. A. Benkeser, G. Schroll, D. M. Sauve, J. Am. Chem. Soc.
1955, 77, 3378Ϫ3379.
A. Dondoni, D. Perrone, E. Turturici, J. Chem. Soc., Perkin
Trans. 1 1997, 16, 2389Ϫ2393.
Hannun, J. Biol. Chem. 1993, 268, 26226Ϫ26232.
R. M. Devant, Kontakte (Darmstadt) 1992, 3, 11Ϫ28.
P. M. Koskinen, A. M. P. Koskinen, Synthesis 1998,
1075Ϫ1091.
C. Curfman, D. Liotta, Methods Enzymol. 1999, 311, 391Ϫ440.
S. C. Bergmeier, Tetrahedron 2000, 56, 2561Ϫ2576.
R. Ghidoni, G. Sala, A. Giuliani, Biochim. Biophys. Acta 1999,
1439, 17Ϫ39.
[22]
[23]
[13]
[14]
[24]
[25]
[26]
[27]
[28]
[15]
[16]
[17]
[18]
P. Garner, J. M. Park, J. Org. Chem. 1987, 52, 2361Ϫ2364.
A. D. Campbell, R. M. Raynham, R. J. K. Taylor, Synthesis
[19]
C. Brockway, P. Kocienski, C. Pant, J. Chem. Soc., Perkin
Trans. 1 1984, 875Ϫ878.
J. Jurczak, A. Golebiowski, Chem. Rev. 1989, 89, 149Ϫ164.
Received February 19, 2004
1998, 1707Ϫ1709.
G. Ageno, L. Banfi, G. Cascio, G. Guanti, E. Manghisi, R.
[20]
Riva, V. Rocca, Tetrahedron 1995, 51, 8121Ϫ8134.
3700
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 3693Ϫ3700