A formal synthesis of 3-O-(4-methoxybenzyl)-azidosphingosine by a modified Julia olefination

Article abstract of DOI:10.1016/S0040-4020(02)00418-0

The stereoselective construction of the D-erythro-azidosphingosine characteristic trans double bond was accomplished by condensation between tetradecanal and a heterocyclic sulfone derived from diethyl-D-tartrate, following the Kocienski modification of the Julia olefination.

Full text of DOI:10.1016/S0040-4020(02)00418-0

TETRAHEDRON
Pergamon
Tetrahedron 58 12002) 4425±4428
A formal synthesis of 3-O-ꢀ4-methoxybenzyl)-azidosphingosine
by a modi®ed Julia ole®nation
Federica Compostella,a Laura Franchini,a Luigi Panza,b,p Davide Prosperib
and Fiamma Ronchettia
a
Á
Dipartimento di Chimica e Biochimica Medica, Universita di Milano, Via Saldini 50, 20133 Milano, Italy
b
Á
Dipartimento di Scienze Chimiche, Alimentari, Farmaceutiche e Farmacologiche, Universita del Piemonte Orientale, Viale Ferrucci 33,
28100 Novara, Italy
Received 14 February 2002;revised 18 March 2002;accepted 11 April 2002
AbstractÐThe stereoselective construction of the d-erythro-azidosphingosine characteristic trans double bond was accomplished by
condensation between tetradecanal and a heterocyclic sulfone derived from diethyl-d-tartrate, following the Kocienski modi®cation of
the Julia ole®nation. q 2002 Elsevier Science Ltd. All rights reserved.
1. Introduction
in the one-pot formation of alkenes by addition of
a-lithiated benzothiazol-2-yl sulfones to aldehydes, but
suffers from the fact that stereoselectivities are often
low.11±13 Kocienski found that the replacement of the
benzothiazol-2-yl group by a 1-phenyl-1H-tetrazol-5-yl
moiety leads, in the case of aliphatic aldehydes and
sulfones, to E alkenes with high selectivities. Furthermore,
the stereochemistry of the reaction is in¯uenced by the
choice of base and solvent, DME as solvent, and potassium
hexamethyldisilazide 1KHMDS) as base being the better
combination for high trans selectivity.4
Glycosphingolipids 1GSLs) are characteristic membrane
components of eukaryotic cells and they serve a variety of
functions either taking part in cell type-speci®c adhesion
processes or acting as binding sites for certain virus or
bacterial toxin.1 Each GLS carries a hydrophobic ceramide
moiety and a hydrophilic oligosaccharide chain. Among the
synthetic procedures for the coupling of the oligosaccharide
part of GSLs to the sphingosine moiety the `azidosphingo-
sine glycosylation procedure'2 has been by far preferred
over the direct attachment to the oligosaccharide of the
preformed ceramide unit or of the sphingosine moiety
followed by N-acetylation. This strategy, in fact, minimizes
the formation of by-products, moreover azidosphingosine is
an excellent precursor of sphingosine. A number of syn-
theses of d-erythro-azidosphingosine 11), mainly by the
asymmetric induction or the `ex chiral pool' approaches,
have been reported.3 When chiral substrates are used as
starting material, the characteristic trans double bond is
generally generated via Wittig related reactions;unfortu-
nately some preliminary attempts revealed that even the
simplest synthetic pathways are not free from problems. A
critical revision of the different approaches for the forma-
tion of E double bond, evidenced the Kocienski modi®-
cation of the ole®nation procedure developed by Sylvester
Julia as useful for this purpose.4 During the last years, this
methodology has been successfully applied to the synthesis
of natural product.5±10 The original Julia procedure consists
Herein the construction of the 4E double bond present in
3-O-14-methoxybenzyl) azidosphingosine 12), a coupling
partner in GSLs synthesis, using the Kocienski modi®cation
of the Julia protocol is reported.
2. Results and discussion
Known alcohol 3,14 derived from diethyl d-tartrate, was
chosen as starting material owing to its d-threo-diol
structure, suitable to obtain, after nitrogen introduction
with inversion of con®guration, the natural sphingosine
geometry.
Keywords: Julia ole®nation;azidosphingosine;sphingolipids;
alkenes;stereoselective synthesis.
Corresponding author. Tel.: 139-0321-657616;fax: 139-0321-657621;
e-mail: panza@pharm.unipmn.it
trans-
In the ®rst attempt, the two partners of the Julia coupling
were aldehyde 4, derived from alcohol 3 by Swern oxi-
dation,14 and sulfone 5. Compound 5 was obtained by
p
0040±4020/02/$ - see front matter q 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020102)00418-0

4426
F. Compostella et al. / Tetrahedron 58 02002) 4425±4428
Scheme 1. 1a) 1i) CH31CH2)12CH2Br, Et3N, THF, rt!608C. 1ii) MCPBA, CH2Cl2. 1b) KHMDS, DME, 2558C. 1c) 1i) 1-phenyl-1H-tetrazole-5-thiol, PPh3,
DEAD, THF, 408C. 1ii) MCPBA, CH2Cl2. 1d) CH31CH2)12CHO, KHMDS, DME, 2558C.
reaction of the commercially available 1-phenyl-1H-tetra-
zole-5-thiol with 1-bromotetradecane, Et3N in THF to afford
the thioether, which was in turn treated with MCPBA in
CH2Cl2 to give the corresponding sulfone 5 in 97% overall
yield 1Scheme 1). The reaction of metalated tetradecyl
sulfone 5 with aldehyde 4 1see Section 4, procedure A),
carried out in DME at 2558C by adding the base to a
mixture of sulfone and aldehyde, resulted in the stereoselec-
tive formation of the desired 4E ole®n 6 in a reasonable
yield 152%, E/Z, 4:1), but in the complete decomposition
of the unreacted starting materials;moreover, a serious
drawback of the reaction was the low solubility of sulfone
5 at the reaction temperature. An attempt to carry out the
condensation in THF instead of DME, lead on one side to a
slight increase of the sulfone solubility, but on the other to
lower yields and selectivities.
fashion, that consists in the addition of the base to a mixture
of the aldehyde and the sulfone. So, KHMDS 10.5 M in
toluene) was added to a solution of sulfone 7 in dry DME
at 2558C, followed by the addition of tetradecanal in DME
1see Section 4, procedure B). The coupling was carried out
different times by varying the equivalents of base 1from 1 to
1.5 equiv.) and/or aldehyde 1from 1 to 3 equiv.), and the
following considerations emerge. First of all, all the tests
con®rm the good stereoselection of the reaction with these
substrates: ole®n 6 was always recovered in a E/Z ratio of
4:1. As far as the reaction yields are concerned, compound 6
was obtained in only acceptable yields 145±53%), but it is
worth noting that it has always been possible to completely
recover the unreacted sulfone 7, resulting in nearly quanti-
tative yield based on conversion. The best result 153%) was
obtained when 1 equiv. of both the aldehyde and the base
were used;in any case the general trend is that it is better to
use a stoichiometric quantity of the base, since the only low
yield 136%) has been obtained when 1.5 base equivalents
were used. The reason why the reaction is not quantitative
might be due to a competitive a-deprotonation of the
aldehyde by the sulfone anion;in fact, an increase of the
aldehyde/sulfone ratio resulted in a decrease of the yield. In
order to augment the nucleophilicity of the anion, a more
coordinating solvent was tried, but the reaction carried
out in THF/DMPU at 2708C resulted in low yields and
decomposition of the starting material.
The good results in term of stereoselectivity stimulated the
planning of an alternative pathway ensuring a better solu-
bility of the reaction partners. The switch between sulfone
and aldehyde functional groups of the two synthons of the
above procedure was devised, leading to a new approach to
compound 6: the coupling between sulfone 7 and commer-
cially available tetradecanal. The synthesis of compound 7
was accomplished as shown in Scheme 1. Alcohol 3 was
®rst converted into the 1-phenyl-1H-tetrazole-5-yl thioether
under Mitsunobu conditions, and then oxidized to 7 in 75%
overall yields. A solution of sulfone 7 in DME at 2558C
was treated with KHMDS 11 equiv.) to prove both the
complete solubility of 7 and its stability despite the presence
of the p-methoxybenzyl group in b to the sulfone: the
anion of compound 7 was ef®ciently stabilized by the
sulfone group, so that no b-elimination product was
formed.15 Thus compound 7 is a good candidate for the
coupling.
While this work was in progress a relevant publication by
Jacobsen10 appeared;he attained, during a fragment
coupling by means of a Kocienski±Julia ole®nation with a
sulfone of the phenyltetrazole series, an high E-selectivity
with LiHMDS in DMF/DMPU mixture. However, when the
Jacobsen's conditions were applied to sulfone 7 and tetra-
decanal, great solubility problems arose;product 6 was
recovered in very low yield 15%), even if a slight increase
in the E/Z ratio was observed 1E/Z, 8.3:1.7). Once again this
result shows that is very dif®cult to generalize the conditions
After a preliminary screening, the standard Julia procedure
was in this second approach preferred over the Barbier-type

F. Compostella et al. / Tetrahedron 58 02002) 4425±4428
4427
of Julia ole®nation leading to high stereoselectivities, since
the outcome of the reaction is strictly dependent on the
features of the substrates.
3H, Jˆ7.5 Hz, CH3);1.20±1.40 1m, 20H, CH 2);1.51
1quint, 2H, Jˆ7.5 Hz, CH2);1.96 1m, 2H, CH );3.74 1m,
2
2H, CH2);7.56±7.74 1m, 5H, arom). 13C NMR: d 14.8;
22.6;23.4;28.8;29.5±30.3 18C);32.6;56.7;125.8 12C);
130.4 12C);132.1;133.7;154.2. MS: m/z 424 [M1NH4]1.
IR 1nujol) nmax 1160, 1340 cm21. C21H34N4O2S: calcd C
62.04, H 8.43, N 13.78;found C 62.22, H 8.31, N 13.65.
Lastly, compound 6 can be ef®ciently transformed into the
target 3-O-14-methoxybenzyl) azidosphingosine 12) with
the procedure described by Somfai and Olsson.8
4.1.2. ꢀ2R,3R)-1,2-O-Pentylidene-3-ꢀ4-methoxybenzyl)-
4-octadecen-1,2,3-triol ꢀ6). Procedure A: A 0.5 M solution
of KHMDS in toluene 11.55 mL, 0.77 mmol) was slowly
added to a mixture of aldehyde 4 10.13 g, 0.43 mmol) and
sulfone 5 10.23 g, 0.56 mmol) in dry DME 14 mL) at
2558C. The reaction mixture was stirred at 2558C for
30 min, then slowly warmed to room temperature. Water
15 mL) was added, and stirring continued for 30 min. The
resulting mixture was extracted with Et2O 13£5 mL). The
combined organic layers were washed with brine, dried over
sodium sulfate, and concentrated. The crude product was
puri®ed by ¯ash-chromatography 1hexane/AcOEt, 95:5) to
afford 6 10.11 g, 52%, E/Z, 8:2) as an oil.
3. Conclusion
A further example of the versatility of the Kocienski modi®-
cation of the Julia protocol applied, in this paper, to the
construction of the trans double bond of d-erythro-azido-
sphingosine skeleton has been described. This represents a
good alternative to other methods, since it allows the
generation of the double bond in high selectivities and
permits to recover and recycle all the expensive material
of the reaction.
4. Experimental
4.1. General
Procedure B: A 0.5 M solution of KHMDS in toluene
10.80 mL, 0.40 mmol) was slowly added to a solution of
sulfone 7 10.20 g, 0.40 mmol) in dry DME 12 mL) under
argon at 2558C. The solution was stirred for 30 min during
which time the color turned yellow. A solution of tetra-
decanal 10.086 g, 0.40 mmol) in DME 11.6 mL) was added
via cannula;the mixture was stirred at 2558C for 15 min,
and then quenched by the addition of water and warmed to
room temperature. After stirring for 30 min, the mixture was
extracted with Et2O. The combined organic layers were
washed with brine, dried, and concentrated. Puri®cation
by ¯ash chromatography 1hexane/AcOEt, 95:5, then 1:1)
gave 6 10.104 g, 53%, E/Z, 8:2), and sulfone 7 10.094 g).
Optical rotations were determined on a Perkin±Elmer 241
polarimeter in a 1 dm cell at 208C. Mass experiments were
performed through chemical ionization mass spectrometry
1CI-MS) as described in Ref. 16. All NMR spectra were
recorded at 303 K with a Bruker AM-500 spectrometer
equipped with an Aspect-3000 computer, a process control-
ler, and an array processor in CDCl3 solutions;chemical
shifts of NMR spectra are reported as d 1ppm) relative to
tetramethylsilane as internal standard. All reactions were
monitored by TLC on Silica Gel 60 F-254 plates 1Merck)
with detection by dipping in an ammonium molybdate
solution followed by heating. Flash column chromatography
was performed on Silica Gel 60 1230±400 mesh, Merck).
All evaporations were carried out under reduced pressure at
408C. Potassium bis1trimethylsilyl)amide 10.5 M in toluene)
and tetradecanal were purchased from Fluka;1-phenyl-1 H-
tetrazole-5-thiol was purchased from Aldrich. Compounds 3
and 4 were obtained as described in Ref. 14.
Compound 6 was recovered as E/Z mixture by ¯ash-
chromatography 1hexane/AcOEt, 95:5). The E/Z ratio was
established through H NMR analysis by integration of the
1
two doublets centered at 4.38 and 4.39 ppm due to the
OCHaHbPh±OMe proton. The two diastereoisomers were
separated by ¯ash-chromatography 1toluene/AcOEt,
40:0.5). Elution gave ®rst E-6: physical data were all in
agreement with those reported in Ref. 14, and then Z-6:
1
[a]Dˆ214 1c 0.3, CHCl3). H NMR 1CDCl3, 500 MHz):
4.1.1. 1-Phenyl-5-tetradecylsulfonyl-1H-tetrazole ꢀ5).
To a solution of 1-phenyl-1H-tetrazole-5-thiol 11.25 g,
7.01 mmol) in dry THF 125 mL), Et3N 11.17 mL,
8.41 mmol) was added, and the mixture stirred at room
temperature. After 40 min, 1-bromotetradecane 12.29 mL,
8.41 mmol) was added and the reaction re¯uxed for 6 h,
then diluted with water 140 mL), and extracted with Et2O
13£40 mL). The combined organic layers were dried and
evaporated at reduced pressure to give the crude thioether.
MCPBA 155%) 17.70 g, 24.53 mmol) was added in small
portions to a solution of the crude thioether in CH2Cl2
146 mL) at 08C, and the mixture was stirred at room
temperature for 24 h. The reaction mixture was washed
with NaHSO3 140 mL), and saturated NaHCO3 solution
13£30 mL). The organic layer was dried, and the solvent
removed by evaporation. The residue was submitted to
¯ash-chromatography 1hexane/AcOEt, 95:5) to afford
compound 5 12.76 g, 97%) as a sticky white solid. Mp
d 0.80±0.96 1m, 9H, 3CH3);1.15±1.42 1m, 22H, 11CH );
2
1.55±1.67 1m, 4H, 2CH2);1.96 1m, 2H, 2H 6);3.54 1dd, 1H,
J1a,1bˆ8.4 Hz, J1a,2ˆ7.0 Hz, H1a);3.78 1s, 3H, OCH );3.90
3
1dd, 1H, J1a,1bˆ8.4 Hz, J1b,2ˆ6.5 Hz, H1b);4.11±4.19 1m,
2H, H2 and H3);4.38 1d, 1H, Jˆ12.0 Hz, OCHaHbPh±
OMe);4.58 1d, 1H, Jˆ12.0 Hz, OCHaHbPh±OMe);5.22
1ddt, 1H, J4,5ˆ11.0 Hz, J3,4ˆ9.0 Hz, Jallˆ1.0 Hz, H4);5.68
1dt, J4,5ˆ11.0 Hz, J5,6ˆ7.5 Hz, H5);6.84 1d, 2H, Jˆ8.4 Hz,
arom.);7.25 1d, 2H, Jˆ8.4 Hz, arom.). 13C NMR: d 8.8;8.9;
14.8;23.4;28.9;30.0±30.4 111C);32.6;55.9;67.2;70.1;
75.5;79.3;114.3 13C);126.2;129.9 12C);131.6;137.4;
159.7. MS: m/z 506 [M1NH4]1. IR 1neat) nmax 3045,
2920, 1620 cm21. C31H52O4: calcd C 76.18, H 10.72;
found C 76.40 H 10.55.
4.1.3. ꢀ2R,3S)-3-O-ꢀ4-Methoxybenzyl)-1,2-O-ꢀ3-pentyl-
idene)-4-ꢀ1-phenyl-1H-tetrazol-5-sulfonyl)-1,2,3-butan-
triol ꢀ7). To a solution of 1-phenyl-1H-tetrazole-5-thiol
1
55.5±56.5 1from diisopropyl ether). H NMR: d 0.90 1t,

4428
F. Compostella et al. / Tetrahedron 58 02002) 4425±4428
10.69 g, 3.89 mmol), alcohol 3 11.10 g, 3.54 mmol), and
Ph3P 11.21 g, 4.60 mmol) in dry THF 150 mL), DEAD
10.66 mL, 4.24 mmol) was added dropwise at room
temperature. The reaction was stirred at 408C for 1 h, then
the solvent was evaporated at reduced pressure. Puri®cation
by ¯ash-chromatography 1hexane/AcOEt, 8:2) afforded
the thioether 11.57 g, 94%). MCPBA 155%) 13.67 g,
11.69 mmol) was added in small portions to a solution of
the thioether in CH2Cl2 143 mL) at 08C, and the mixture was
stirred at room temperature for 24 h. The reaction mixture
was washed with NaHSO3 140 mL), and saturated NaHCO3
solution 13£30 mL). The organic layer was dried, and the
solvent removed by evaporation. The residue was submitted
to ¯ash-chromatography 1hexane/EtOAc, 8:2) to afford
compound 7 11.35 g, 80%) as a sticky white solid. Mp
59.0±60.4 1from diisopropyl ether). [a]Dˆ114 1c 1 in
References
1. Huwiler, A.;Kolter, T.;Pfeilschifter, J.;Sandhoff, K.
Biochim. Biophys. Acta 2000, 1485, 63±99.
2. Vankar, Y. D.;Schmidt, R. R. Chem. Soc. Rev. 2000, 29, 201±
216.
3. Koskinen, P. M.;Koskinen, A. M. P. Synthesis 1998, 1075±
1091.
4. Blakemore, P. R.;Cole, W. J.;Kocienski, P. J.;Morley, A.
Synlett 1998, 26±28.
5. Williams, D. R.;Brooks, D. A.;Berliner, M. A. J. Am. Chem.
Soc. 1999, 121, 4924±4925.
È
6. Mulzer, J.;O hler, E. Angew. Chem., Int. Ed. Engl. 2001, 40,
3842±3846.
7. Metternich, R.;Denni, D.;Thai, B.;Sedrani, R. J. Org. Chem.
1999, 64, 9632±9639.
1
CHCl3). H NMR 1CDCl3, 500 MHz): d 0.84 1t, 3H, Jˆ
8. Smith, A. B.;Adams, C. M.;Kozmin, S. A.;Paone, D. V.
J. Am. Chem. Soc. 2001, 123, 5925±5937.
9. Smith, A. B.;Safonov, I. G.;Corbett, R. M. J. Am. Chem. Soc.
2001, 123, 12426±12427.
7.5 Hz, CH3);0.88 1t, 3H, Jˆ7.5 Hz, CH3);1.55 1m, 2H,
CH2);1.63 1m, 2H, CH );3.70 1dd, J4a,4bˆ8.5 Hz, J4a,3
ˆ
2
6.5 Hz, H4a);3.78 1s, 3H, OCH 3);3.85±3.89 1m, 2H,
2H1);3.93 1dd, J4a,4bˆ8.5 Hz, J4b,3ˆ6.5 Hz, H4b);4.22±
4.28 1m, 2H, H2 and H3);4.43 1d, 1H, Jˆ11.0 Hz,
OCHaHbPh±OMe);4.51 1d, 1H, Jˆ11.0 Hz, OCHaHbPh±
OMe);6.82 1m, 2H, arom.);7.05 1m, 2H, arom.);7.45±
7.62 1m, 5H, arom.). 13C NMR: d 8.9 12C);28.8;29.8;
56.0;57.5;65.8;73.6;74.1;76.1;114.5 12C), 114.8;
10. Liu, P.;Jacobsen, E. N. J. Am. Chem. Soc. 2001, 123, 10772±
10773.
11. Baudin, J. B.;Hareau, G.;Julia, S. A.;Ruel, O. Tetrahedron
Lett. 1991, 32, 1175±1178.
12. Baudin, J. B.;Hareau, G.;Julia, S. A.;Ruel, O.
Chim. Fr. 1993, 130, 336±357.
Bull. Soc.
126.2 12C);129.7;130.0 12C);130.3 12C);132.2;133.6;
13. Baudin, J. B.;Hareau, G.;Julia, S. A.;Lorne, R.;Ruel, O.
Bull. Soc. Chim. Fr. 1993, 130, 856±878.
154.9;160.2. MS: m/z 520 [M1NH4]1, 503 [M1H]1. IR
1nujol) nmax 1150, 1170, 1350, 1460, 1510, 1610 cm21
C24H30N4O6S: calcd C 57.36, H 6.02, N 11.15;found C
57.51, H 5.95, N 10.97.
.
14. Somfai, P.;Olsson, R. Tetrahedron 1993, 49, 6645±6650.
15. A prevoius example of a metallated b-alkoxy sulfone was
described: Pattenden, G.;Plowright, A. T.;Tornos, J. A.;
Ye, T. Tetrahedron Lett. 1998, 39, 6099±6102.
16. Colombo, D.;Marinone Albini, F.;Scala, A.;Taino, I. M.;
Toma, L. Tetrahedron: Asymmetry 1994, 5, 1993±1998.
Acknowledgements
We thank the University of Milano and the University of
Piemonte Orientale for ®nancial support.