Expedient synthesis of the α-C-glycoside analogue of the immunostimulant galactosylceramide (KRN7000)

Article abstract of DOI:10.1021/ol0613057

Key reactions in a concise synthesis of an α-C-galactosylceramide analogue of KRN7000 include a diastereoselective alkenylalane addition to an N-tert-butanesulfinyl imine and the use of an epoxidation/carbamate ring opening sequence to install the aminodi

Full text of DOI:10.1021/ol0613057

ORGANIC
LETTERS
2006
Vol. 8, No. 15
3375-3378
Expedient Synthesis of the
-C-Glycoside Analogue of the
Immunostimulant Galactosylceramide
(KRN7000)
r
Peter Wipf* and Joshua G. Pierce
Department of Chemistry, UniVersity of Pittsburgh, Pittsburgh, PennsylVania 15260
pwipf@pitt.edu
Received May 29, 2006 (Revised Manuscript Received June 17, 2006)
ABSTRACT
Key reactions in a concise synthesis of an
r-C-galactosylceramide analogue of KRN7000 include a diastereoselective alkenylalane addition
to an N-tert-butanesulfinyl imine and the use of an epoxidation/carbamate ring opening sequence to install the aminodiol stereotriad.
Glycolipids have been a target of increasing interest in
immunostimulant research since the discovery of the thera-
peutic potential of R-galactosylceramides, particularly
KRN7000 (1).1 Impressive activities have been recorded
against various disease models, including cancer,2a maleria,2b
and hepatitis B.2c The current model for the mode of action
of 1 involves sequential attachment to CD1d receptors on
antigen-presenting cells and natural killer T cells, resulting
in disease suppression.3 Of the various analogues of 1 that
have been prepared, the R-C-galactosylceramide analogue
2 developed by Franck et al. has shown a spectacular increase
in potency: a 1000-fold enhancement of 2 over 1 was found
in a mouse malaria assay, and a 100-fold activity increase
was detected in a mouse melanoma model.4 The initial
synthesis of 2 by Franck and co-workers involved the use
of the Ramberg-Baecklund reaction as the key step, and
this approach was subsequently improved through the use
of olefin cross-metathesis.4b Other groups have developed
alternative routes that allow for facile analogue preparation,
including the synthesis of the â-C-galactosylceramide;
however, these approaches are plagued by poor stereoselec-
tivity in installing the amino-diol stereotriad of 2.5
While investigating our cationic zirconocene addition to
(1) (a) Natori, T.; Morita, M.; Akimoto, K.; Koezuka, Y. Tetrahedron
1994, 50, 2771. (b) Morita, M.; Motoki, K.; Akimoto, K.; Natori, T.; Sakai,
T.; Sawa, E.; Yamaji, K.; Koezuka, Y.; Kobayashi, E.; Fukushima, H. J.
Med. Chem. 1995, 38, 2176. (c) Costantino, V.; Fattorusso, E.; Imperatore,
C.; Mangoni, A. J. Org. Chem. 2004, 69, 1174.
(2) (a) Hayakawa, Y.; Rovero, S.; Forni, G.; Smyth, M. J. Proc. Natl.
Acad. Sci. U.S.A. 2003, 100, 9464. (b) Gonzalez-Aseguinolaza, G.; de
Oliveira, C.; Tomaska, M.; Hong, S.; Bruna-Romero, O.; Nakayama, T.;
Taniguchi, M.; Bendelac, A.; Van Kaer, L.; Koezuka, Y.; Tsuji, M. Proc.
Natl. Acad. Sci. U.S.A. 2000, 97, 8461. (c) Kakimi, K.; Guidotti, L. G.;
Koezuka, Y.; Chisari, F. V. J. Exp. Med. 2000, 192, 921.
glycal epoxides,6 it became clear that the inclusion of
(4) (a) Yang, G.; Schmieg, J.; Tusji, M.; Franck, R. W. Angew. Chem.,
Int. Ed. 2004, 43, 3818. (b) Chen, G.; Schmieg, J.; Tsuji, M.; Frank, R. W.
Org. Lett. 2004, 6, 4077. (c) Franck, R. W.; Tsuji, M. Acc. Chem. Res.
2006, in press. (d) Chen, G.; Chien, M.; Tsuji, M.; Franck, R. W.
ChemBioChem 2006, 7, 1017.
(5) (a) Toba, T.; Murata, K.; Yamamura, T.; Miyake, S.; Annoura, H.
Tetrahedron Lett. 2005, 46, 5043. (b) Chaulagain, M. R.; Postema, M. H.
D.; Valeriote, F.; Pietraszkewicz, H. Tetrahedron Lett. 2004, 45, 7791.
(6) Wipf, P.; Pierce, J. G.; Zhuang, N. Org. Lett. 2005, 7, 483.
(3) Brigl, M.; Brenner, M. B. Annu. ReV. Immunol. 2004, 22, 817.
10.1021/ol0613057 CCC: $33.50
© 2006 American Chemical Society
Published on Web 06/29/2006

nitrogen in the glycoside side chain would provide molecules
with interesting biological properties.7 Accordingly, we
envisioned N-tert-butanesulfinyl imine 4 as a key intermedi-
ate that could undergo a diastereoselective alkenylmetal
addition followed by epoxidation and carbamate ring opening
of 5 to generate compounds such as 2 in a rapid fashion
(Scheme 1). From the outset, our requirements for this
this transformation failed, however. Conversion of aldehyde
7 to N-tert-butanesulfinylimine 4 was achieved in 94%
yield.10
With an efficient synthesis of imine 4 established, we
directed our efforts toward the stereoselective alkenylmetal
addition. Initial studies focused on the hydrozirconation11
of alkynes followed by transmetalation to dimethylzinc which
has proved effective for 1,2-addition to diphenylphosphi-
noylimines.12 Unfortunately, a variety of solvents, temper-
atures, and external Lewis acids failed to promote this
reaction.
Scheme 1. Retrosynthetic Approach toward the R-C-Glycoside
Analogue of KRN7000 (2)
Inspired by previous work in our group on the carboalu-
mination-sulfinimine addition of alkynes,13 we also inves-
tigated an alternative transmetalation of alkenyl zirconocenes
to trimethylaluminum14,15 and subsequent addition to 4. To
our delight, hydrozirconation of 1-hexyne (8) in CH2Cl2 at
room temperature, addition of Me3Al16 followed by 4 at 0
°C, and subsequent warming to room temperature for 3 h
generated the desired allylic amine 9 in 82% yield as a single
diastereomer by 1H NMR analysis (entry 1, Table 1).
Table 1. Hydrozirconation of Alkynes Followed by
Transmetalation to Trimethylaluminum and Addition to
N-tert-Butanesulfinylimine 4
undertaking were to synthesize 2 in the shortest possible
sequence and in a modular and stereoselective fashion.
Synthesis of aldehyde 7 was readily accomplished using
a literature procedure7 that entailed allylation of 3 with
allyltrimethylsilane, hydroboration/oxidation, and Swern
oxidation to generate the desired aldehyde in 71% overall
yield (Scheme 2). Later, we found that conditions developed
Scheme 2. Synthesis of Imine 4 via Allyltrimethylsilane
Addition
entry
alkyne
product
yield [%]
1
2
3
4
8
10
12
14
9
11
13
15
82a
81a
65b
85a
a Products were diastereomerically pure according to 1H NMR analysis
of the crude reaction mixtures. b A 93:7 mixture of diastereomers based on
HPLC analysis; all yields are based on isolated, pure material.
Furthermore, we were able to demonstrate that silyl ether,
carbamate, and sulfonamide functionalities were well toler-
(7) Palomo, C.; Oiarbide, M.; Landa, A.; Gonza´lez-Rego, M. C.; Garc´ıa,
J. M.; Gonza´lez, A.; Odriozola, J. M.; Martin-Pastor, M.; Linden, A. J.
Am. Chem. Soc. 2002, 124, 8637 and cited references.
(8) Panek, J. S.; Sparks, M. A. J. Org. Chem. 1989, 54, 2034. The
reaction is also effective using conditions described in this report. A 67%
yield was obtained on small scale; however, on larger scale, yields were
variable.
by Panek8 employing acetyloxyallylsilane 6 as a homoenolate
equivalent could be used to convert 3 via the resulting enol
acetate9 in situ to the desired aldehyde 7 in 67% yield
(Scheme 2). All attempts to employ silyloxyallylsilanes in
(9) Formed as a 10:1 mixture of anomers and a mixture of (E/Z)-isomers.
3376
Org. Lett., Vol. 8, No. 15, 2006

ated and provided allylic amides in high yield and excellent
diastereoselectivity (entries 2-4, Table 1). We propose the
four-membered chelate model 16 to account for the observed
selectivity, in analogy to additions of alkenylalanes derived
from alkyne carboaluminations.13 The mild and efficient
conditions for generating N-tert-butanesulfinyl imines coupled
with the rapid, stereoselective, and functional group tolerant
method of alkenylalane addition described herein provide
an attractive strategy for allylic amine synthesis.17
bearing the tert-butanesulfinyl protecting group; however,
epoxidation of this species also oxidized the sulfur to
generate the Bus protecting group18 that could not be
removed even under forcing conditions.
At this stage, conditions had to be identified to stereo-
selectively epoxidize the alkene 5 and effect the intra-
molecular cyclization of the tert-butylcarbamate to form
oxazolidinone 21.19 Prior studies by Roush20a and O’Brien20b
have demonstrated the feasibility of this sequence, although
Roush employed trichloroacetamides and O’Brien focused
on cyclic allylic amides.20c
An extension of this method toward the synthesis of
monoBoc-protected allylic amide 19 and bisBoc-protected
allylic amide 5 was straightforward. Hydrozirconation of
1-hexadecyne to generate alkenylzirconocene 17, followed
by the aluminum transmetalation/imine addition and conve-
nient in situ deprotection of the labile sulfinyl protecting
group with aqueous HCl afforded the desired allylic amine
in 72% yield (Scheme 3). A two step N-Boc-protection
Initial trials used MCPBA as the epoxidizing agent, under
variable temperature and solvent conditions. The best result
could be obtained at 0 °C in CH2Cl2 to yield 85% of 21 as
a 1.6:1 mixture of diastereomers (entry 1, Table 2). In situ
Table 2. Evaluation of Oxidation Conditions to Generate
Epoxide Intermediate 20 Followed by Intramolecular Opening of
the Epoxide to Form Carbamate 21
Scheme 3. Alkenylalane Addition/Deprotection/Boc-Protection
Sequence to Intermediate 5
entry
substrate
conditions
yield [%], dra
1
2
5
5
MCPBA, 0 °C
oxone, NaHCO3,
acetone, rt
85, 1.6:1
trace
3
4
5
TFAA, UHP, -20 °C,
Na2HPO4
TFAA, UHP, -20 °C,
Na2HPO4
93, 9:1
50, ND
19
aYields refer to isolated, pure products. Diastereoselectivity was deter-
mined by HPLC analysis of the crude reaction mixtures.
proved to be higher yielding than the one step approach. Our
original strategy involved epoxidation of the allylic amide
generated dimethyldioxirane (DMDO) was ineffective, pro-
ducing no conversion after 20 h at room temperature.
Increasing the electrophilicity of peracids has been shown
to increase selectivity in directed epoxidations.21 Trifluo-
roperacetic acid, generated in situ from trifluoroacetic acid
(10) For a review, see Ellman, J. A.; Owens, T. D.; Tang, T. P. Acc.
Chem. Res. 2002, 35, 984.
(11) For reviews, see (a) Wipf, P.; Jahn, H. Tetrahedron 1996, 52, 12853.
(b) Wipf, P.; Kendall, C. Top. Organomet. Chem. 2005, 8, 1.
(12) (a) Wipf, P.; Kendall, C.; Stephenson, C. R. J. J. Am. Chem. Soc.
2001, 123, 5122. (b) Wipf, P.; Kendall, C.; Stephenson, C. R. J. J. Am.
Chem. Soc. 2003, 125, 761.
(13) Wipf, P.; Nunes, R. L.; Ribe, S. HelV. Chim. Acta 2002, 85, 3478.
(14) Although first demonstrated by Schwartz and Carr in 1977 with
AlCl3,15 hydrozirconation/transmetalation to aluminum has not been explored
further for addition to electrophiles.
(15) Carr, D. B.; Schwartz, J. J. Am. Chem. Soc. 1977, 99, 638.
(16) The quality of Me3Al is of critical importance in this reaction.
Commercial solutions of Me3Al were ineffective, possibly because of
aggregate formation or traces of metal oxides; neat Me3Al that was freshly
diluted with CH2Cl2 was used in all cases.
(18) Sun, P.; Weinreb, S. M.; Shang, M. J. Org. Chem. 1997, 62, 8604.
(19) Wipf, P.; Lim, S. J. Am. Chem. Soc. 1995, 117, 558.
(20) (a) Roush, W. R.; Straub, J. A.; Brown, R. J. J. Org. Chem. 1987,
52, 5127. (b) O’Brien, P.; Childs, A. C.; Ensor, G. J.; Hill, C. L.; Kirby, J.
P.; Dearden, M. J.; Oxenford, S. J.; Rosser, C. M. Org. Lett. 2003, 5, 4955.
(c) Although frequently observed as a side reaction, this sequence is rarely
employed for anti-diol synthesis.
(21) (a) Jensen, A. J.; Luthman, K. Tetrahedron Lett. 1998, 39, 3213.
(b) Fehr, C. Angew. Chem., Int. Ed. Engl. 1998, 37, 2407. (c) Lee, K. W.;
Hwang, S. Y.; Kim, C. R.; Nam, D. H.; Chang, J. H.; Choi, S. C.; Choi, B.
S.; Choi, H.-W.; Lee, K. K.; So, B.; Cho, S. W.; Shin, H. Org. Process
Res. DeV. 2003, 7, 839.
(17) For a review, see Johannsen, M.; Jørgensen, K. A. Chem. ReV. 1998,
98, 1689.
Org. Lett., Vol. 8, No. 15, 2006
3377

anhydride (TFAA) and urea-hydrogen peroxide (UHP)
inclusion complex,22 improved the selectivity of the epoxi-
dation to 9:1 and increased the yield to 93% (entry 3, Table
2). MonoBoc protected amide 19 did not perform well in
this reaction, seemingly because of an increase in decom-
position. The assignment of the relative configuration in 20
was in accordance with previous reports;20a moreover, it was
confirmed by coupling constant analysis23 and by converting
both diastereomers to the desired C-glycosides 2 and 2′.24
Epoxidation of 5 under the TFAA/UHP conditions from
entry 3 in Table 2 was accompanied by epoxide ring opening
to give an N-Boc-protected carbamate (Scheme 4). Careful
Confirmation of the target structure was obtained by
comparison of the [R]D and 13C NMR data of previously
prepared 24 with 2 and 2′. As shown in Table 3, the 13C
Table 3. Comparison of Representative 13C NMR Chemical
Shifts and [R]D Valuesa
Scheme 4. Completion of the Synthesis of the
R-C-Galactosyl-ceramide Analogue of KRN700024
a
13C NMR data for 2 and 2′ were obtained in d5-pyridine at 126 MHz.
b Tentative assignments. [R]D values were obtained in pyridine (c 0.13)
and are reported as average values of three measurements.
c
NMR data for 2 compare well with the literature data, while
there are considerable differences with 2′. Further support
was achieved through comparison of [R]D values.
In conclusion, we have developed a short (10 steps for
the longest linear sequence, 12 overall transformations),
stereoselective synthesis of the R-C-glycoside analogue 2
of KRN7000 (1). This process allows for analogue prepara-
tions at multiple points along the route. The synthetic
sequence does not only showcase the utility of alkenylzir-
conocene-alkenylalane additions to N-tert-butanesulfinyl
imines for the synthesis of enantiomerically pure allylic
amines, but also the possibility to use a stereoselective trans-
alkene epoxidationsintramolecular carbamate cyclization
protocol as an attractive alternative to the dihydroxylation
of cis-alkenes for the synthesis of anti-diols.
reaction monitoring of a KOH/EtOH solution of this
intermediate allowed the selective cleavage of the cyclic
carbamate. N-Boc-protected amino diol 23 was thus obtained
in 89% overall yield as a 9:1 mixture of diastereomers that
were separated by column chromatography on SiO2. Removal
of the Boc group with HCl, coupling of the amine salt with
acid chloride 2425 and global deprotection of the galactosyl
benzyl ethers provided 2 in 71% yield over 3 steps.
Acknowledgment. This work has been supported by the
National Science Foundation (Grant CHE-0315205). We
would like to thank Mr. Juan Arredondo for help in obtaining
NMR data and Prof. R. W. Franck for helpful discussions
and copies of NMR spectra.
(22) Taliansky, S. Synlett 2005, 1962.
(23) Compound 21: major isomer, J9,10 ) 9.5 Hz; minor isomer, J9,10
) 5.2 Hz. These values are consistent with literature data wherein cis-
oxazolidinones have larger coupling constants than trans-oxazolidinones:
Bonini, B F.; Comes-Franchini, M.; Fochi, M.; Laboroi, F.; Mazzanti, G.;
Ricci. A.; Varchi, G. J. Org. Chem. 1999, 64, 8008.
(24) The 2 diastereomers of 23 (23 and 23′) were separated by column
chromatography, and each diastereomer was carried separately through the
remaining sequence with no notable differences. Yields for the sequence
shown in Scheme 4 represent averaged values.
Supporting Information Available: Experimental pro-
cedures and spectral data for 2, 2′, 4, 5, 7, 9, 11, 13, 15, 18,
and 23. This material is available free of charge via the
Internet at http://pubs.acs.org.
(25) Acyl chloride 24 was employed because of its greater solubility
and ease of preparation compared to activated esters used previously:
Heidelberg, T.; Martin, O. R. J. Org. Chem. 2004, 69, 2290.
OL0613057
3378
Org. Lett., Vol. 8, No. 15, 2006