Efficient synthesis of α-C-galactosyl ceramide immunostimulants: Use of ethylene-promoted olefin cross-metathesis

Article abstract of DOI:10.1021/ol0482137

(Chemical Equation Presented) Olefin cross-metathesis has been used to prepare α-C-galactosylceramide derivatives. The metathesis process merged vinyl and propenyl glycosides with vinyl derivatives of phytosphingosine. The use of ethylene enhanced the yie

Full text of DOI:10.1021/ol0482137

ORGANIC
LETTERS
2004
Vol. 6, No. 22
4077-4080
Efficient Synthesis of
r-C-Galactosyl
Ceramide Immunostimulants: Use of
Ethylene-Promoted Olefin
Cross-Metathesis
Guangwu Chen,† John Schmieg,‡ Moriya Tsuji,‡ and Richard W. Franck*,†
Department of Chemistry, Hunter College of CUNY, 695 Park AVenue,
New York, New York 10021, Department of Medical and Molecular Parasitology,
New York UniVersity School of Medicine, New York, New York 10016, and Aaron
Diamond AIDS Research Center, The Rockefeller UniVersity, New York, New York 10021
rfranck@hunter.cuny.edu
Received September 3, 2004
ABSTRACT
Olefin cross-metathesis has been used to prepare
r-C-galactosylceramide derivatives. The metathesis process merged vinyl and propenyl
glycosides with vinyl derivatives of phytosphingosine. The use of ethylene enhanced the yield of the methathesis step.
Agelasphins (R-galactosylceramides), a series of glyco-
sphingolipids isolated from a marine sponge of the genera
Agelas, showed potent effects on the immune system of
mice.1 This discovery led to the synthesis of a slightly
simplified analogue KRN7000 (1) as the lead compound for
further development.2 Detailed studies revealed that 1 is a
powerful immunostimulant which induces formation of both
interferon-γ (IFN-γ) and interleukins (IL)-12 and (IL)-4 by
first binding to antigen-presenting CD1d cells whereupon
the resulting complex then binds to natural killer T (NKT)
cells.3 This group of cytokines, which induce antagonistic
biological effects, i.e., Th1- and Th2-type responses, appar-
ently limit R-GalCer from eliciting a maximum of either
response.4 Interestingly, a recent report describes aza ana-
logues of 1 with good activity which apparently do not
interact via NKT cells.5
Our recent report of C-analogue 2 showed that it was 1000
times more active than 1 in a mouse malaria assay and 100-
fold more potent in a mouse melanoma model.6 Our first
synthesis employed the Ramberg-Backlund reaction as a
key step for linking a galactose derivative to a homophyto-
sphingosine which itself required a total synthesis.7 To make
(3) (a) Wilson, M. T.; Singh, A. K.; Van Kaer, L. Trends Mol. Med.
2002, 8 (5), 225-231. (b) Hayakawa, Y. Godfrey, D. I.; Smyth, M. J. Curr.
Med. Chem. 2004, 11 (2), 241-252.
† Hunter College of CUNY.
(4) Pal, E.; Tabira, T.; Kawano, T.; Taniguchi; Miyake, M. S.; Yama-
mura, T. J. Immunol. 2001, 166, 662-668.
‡ New York University School of Medicine and The Rockefeller
University.
(5) Mehta, A. S.; Gu, B.; Conyers, B.; Ouzounov, S.; Wang, L.; Moriarty,
R. M.; Dwek, R. A.; Block, T. M. Antimicrob. Agents Chemother. 2004,
48, 2085-2090.
(1) (a) Natori, T.; Morita, M.; Akimoto, K.; Koezuka, Y. Tetrahedron
1994, 50, 2771-2784. (b) Costantino, V.; Fattorusso, E.; Imperatore, C.;
Mangoni, A. J. Org. Chem. 2004, 69, 1174-1179 and references therein.
(2) 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, 2176-2187.
(6) Schmieg, J.; Yang, G.; Franck, R. W.; Tsuji, M. J. Exp. Med. 2003,
198, 1631-1641.
(7) Yang, G.; Schmieg, J.; Tsuji, M.; Franck, R. W. Angew. Chem. 2004,
116, 3906-3910; Angew. Chem., Int. Ed. 2004, 43, 3818-3822.
10.1021/ol0482137 CCC: $27.50
© 2004 American Chemical Society
Published on Web 10/06/2004

Scheme 1. Preparation of Sugar Olefins as CM Partners
(second generation) under an ethylene atmosphere afforded
C-vinyl galactoside 11 in an overall 50% yield after five
steps (Scheme 1). The R-C-vinyl galactoside with acetyl
protection 10 was also prepared in good yield. This CM
approach to C-vinyl glycosides has advantages over other
available methods8e-g,12 including the ethynyl sugar route
because of the simpler chemistry involved.
Figure 1. Cross-Metathesis Approach to C-Galactosyl Ceramide
Analogues.
The terminal olefin form of the sphingosine side chain
was prepared starting from the commercially available
phytosphingosine (Scheme 2). In the benzyl carbamate
larger quantities of our potent analogue available to the
immunology community, we have designed a shorter syn-
thesis which is based on olefin cross-metathesis.8 Here, we
report our second-generation convergent synthesis, with a
novel ethylene-promoted CM sequence as a key feature, of
the exact C-galactosyl ceramide analogue 2, its homologue
with one added methylene 4, and their unsaturated counter-
parts 3 and 5 (Figure 1).
Scheme 2. Preparation of Terminal Olefins of
Phytosphingosine
Our convergent approach begins with preparations of R-C-
vinyl and propenyl galactosides. The R-C-vinyl galactoside
11 was initially produced by controlled hydrogenation of the
correspondent ethynyl sugars 14 (Scheme 1). The latter was
made by using the protocol of Dondoni and Isobe.9 Although
this method worked smoothly in our hands, the overall yield
(30% after five steps from methyl galactoside) was limited
by the key C-glycosidation step. Thus, we investigated the
feasibility of a preparation of 11 via transformation of the
C-(1-propenyl) galactosides 9 to terminal olefins utilizing
cross metathesis with ethylene. Key material 9 was readily
produced from galactosyl pentaacetate using standard allyl
C-glycosidation,10 followed by palladium chloride mediated
isomerization.11 Subsequent treatment with Grubbs catalyst
sequence, after formation of the carbamate, the selectively
blocked primary alcohol 17 was obtained in four protecting
group manipulations in excellent yield. To obtain diol 15,
the primary alcohol was temporarily blocked by a silyl group,
which was then removed after the clean isopropylidenation
of the vicinal diol. Further transformation to the aldehyde
(8) (a) Scholl, M.; Ding, S.; Lee, C. W.; Grubbs, R. H. Org. Lett. 1999,
1, 953-956. (b) Zhu, L.; Mootoo, D. R. Org. Lett. 2003, 5, 3475-3478.
(c) Chen, G.; Kirschning, A. Chem. Eur. J. 2002, 8, 2717-2729. (d)
Blackwell, H. E.; O’Leary, D. J.; Chatterjee, A. K.; Washenfelder, R. A.;
Bussmann, D. A.; Grubbs, R. H. J. Am. Chem. Soc. 2000, 122, 58-71. (e)
Dondoni, A.; Giovannini, P. P.; Marra, A. J. Chem. Soc., Perkin Trans. 1
2001, 2380-2388. (f) Postema, M. H. D.; Piper, J. L. Tetrahedron Lett.
2002, 43, 7095-7099. (g) Nolen, E. G.; Kurish, A. J.; Wong, K. A.;
Orlando, M. D. Tetrahedron Lett. 2003, 44, 2449-2453. (h) Rai, A. N.;
Basu, A. Org. Lett. 2004, 6, 2861-2863.
(9) (a) Dondoni, A.; Mariotti, G.; Marra, A. J. Org. Chem. 2002, 67,
4475-4486. (b) Isobe, M. J. Synth. Org. Chem., Jpn. 1994, 52, 968-979.
(10) (a) Giannis, A.; Sandhoff, K. Tetrahedron Lett. 1985, 26, 1479-
1482. (b) Lewis, M. D.; Cha, J. K.; Kishi, Y. J. Am. Chem. Soc. 1982, 104,
4976-4978.
(11) (a) Brooks, G.; Edwards, P. D.; Hatto, J. D. I.; Smale, T. C.;
Southgate, R. Tetrahedron 1995, 51, 7999-8014. (b) Wilcox, C. S.; Long,
G. W.; Suh, H. Tetrahedron Lett. 1984, 25, 395-398.
(12) (a) Westermann, B.; Walter, A.; Flo¨rke, U.; Altenbach, H.-J. Org.
Lett. 2001, 3, 1375-1378. (b) Wang, W.; Zhang, Y.; Zhou, H.; Ble´riot,
Y.; Sinay, P. Eur. J. Org. Chem. 2001, 1053-1059.
4078
Org. Lett., Vol. 6, No. 22, 2004

19 was not totally problem-free. Standard Swern oxidation13
gave rise to epimerization at the R position, while a protocol
with sodium hypochlorite catalyzed by TEMPO14 led to the
acid as a product of overoxidation. Finally, we resorted to
the Kirschning solid phase oxidant in dilute neutral solution,
which successfully effected the transformation to 19 in 93%
yield.15 Terminal olefins 12 and 13 (t-Boc) were obtained
after treating the aldehydes with Tebbe reagent with overall
yields of 46% and 60%, respectively.
Table 2. Effects of Protecting Groups, Olefin Substitutions,
and Ethylene on CM Efficiency
The convergent olefin metathetic assembly was investi-
gated with the above suitably protected coupling partners.
C-Allyl glycosides (Table 1) participated well using Grubbs’
entry
sugar
lipid
solvent
product (R1,R2) yielda (%)
1
2
3
4
5
6
7
11
11
9
13
12
13
12
13
13
13
PhH
PhH
PhH
PhH
CH2Cl2
CH2Cl2
CH2Cl2
23 (Bn, Boc)
24 (Bn, Cbz)
23 (Bn, Boc)
25 (Ac, Cbz)
23 (Bn, Boc)
26 (Ac, Boc)
26 (Ac, Boc)
23 (10)
37 (9)
trace
trace
trace
8
Table 1. CM Formation of C-Glycolipids with C-Allylsugars
as Coupling Partnersa
9b
8b
8c
27 (56)
72 (46)
a All yields given are isolated ones, and those yields in parentheses refer
to isolated sugar homodimers. b Optimized conditions: degassing the
reaction mixture at the beginning; catalyst loading: 30 mol % in two portions
with 24 h interval; 0.1 M in dry CH2Cl2; reflux. c Condition b with presence
of ethylene and 10 mol % catalyst loading. All reactions were stirred under
reflux for 2 d with sugar in excess (2.5 equiv).
coupled product 26 with an isolated yield of 27% (entry 6).
Under the same conditions, there was no improvement with
benzylated propenyl sugar 9 as starting CM partner (entry
5).18 As a result of our success in using ethylene for the
conversion of C-(1-propenyl)glycosides to C-vinyl counter-
parts along with reports of ethylene promotion of enyne
cross-metathesis (“Mori conditions”),19 we tested ethylene
as a promoter for our cross-coupling process. In the event,
refluxing the side-chain olefin 13 with excess C-(1-propenyl)-
sugar 8 (2.5 equiv) in the presence of ethylene with a
cumulative addition of 10 mol % of second-generation
Grubbs’ catalyst in two portions led to greatly improved
formation of CM product 26 with more than 70% isolated
yield (entry 7). For this outcome, it must be the case that
CM of the product with ethylene is a slow step. The observed
enhancement of our CM by ethylene is probably the result
of improved ruthenocyclobutane formation of the phyto-
sphingosine partner. Hoye has recently reported an intramo-
lecular relay method to improve difficult RCM reactions.20
This also presumably promotes formation of slow-to-form
ruthenocyclobutane intermediates.
entry
sugar
solvent (T (° C))
product
yieldb (%)
1
2
6
7
CH2Cl2 (reflux)
CH2Cl2 (reflux)
21
22
61 (54)
61 (48)
a Both reactions were allowed to stir for 2 d with excess sugar partners
(2.5 equiv) and 30 mol % (in two portions) of Grubbs’ catalyst (second
generation). b All yields given are isolated ones, and those yields in
parentheses refer to isolated homodimers of sugars.
catalyst (second generation) in coupling with the tert-butyl
carbamate version of the lipid side chain in more than 60%
isolated yield under nonoptimized conditions. Considering
the bulky neighboring groups of both partner alkenes, we
believe this yield to be good. There was no significant
difference in reactivity between peracetyl- and perbenzyl-
protected sugar olefins 6 and 7. Vinyl homologue 11 (Table
2), comparatively, afforded significantly reduced cross-
coupling yields (entries 1 and 2), possibly attributed to
deactivating chelation between metal center, the multifunc-
tional groups around the rigid tetrahydropyran ring, and the
increased congestion at reaction sites.16 This result is quite
consistent with the reports from several other groups.8e,g,17
With substituted sugar olefins, either protected with acetyl
or benzyl groups (8 or 9), only traces of expected product
were detected in our initial tries (entries 3-5).8g After
optimization with respect to degassing, catalyst loading,
temperature, solvents, and concentration, the CM starting
with peracetyl-protected propenyl sugar 8 afforded cross-
Completion of the synthesis of our target molecules 2-5
requires amidation and deprotection of CM products 26 and
(16) (a) Jorgensen, M.; Hadwiger, P.; Madsen, R.; Stutz, A. E.; Wrodnigg,
T. M. Curr. Org. Chem. 2000, 4, 565-588. (b) Roy, R.; Das, S. K. Chem.
Commun. 2000, 519-529. (c) Trnka, T. M.; Grubbs, R. H. Acc. Chem.
Res. 2001, 34, 18-29.
(17) (a) Roy, R.; Das, S. K.; Dominique, R.; Trono, M. C.; Hernandez-
Mateo, F.; Santoyo-Gonzalez, F. Pure Appl. Chem. 1999, 71, 565-571.
(b) Godin, G.; Compain, P.; Martin, O. R. Org. Lett. 2003, 5, 3269-3272.
(18) McGarvey, G. J.; Benedum, T. E.; Schmidtmann, F. W. Org. Lett.
2002, 4, 3591-3594.
(19) (a) Kinoshita, A.; Sakakibara, N.; Mori, M. J. Am. Chem. Soc. 1997,
119, 12388-12389. (b) Diver, S. T.; Giessert, A. J. Chem. ReV. 2004, 104,
1317-1382 and references therein.
(20) Hoye, T. R.; Jeffrey, C. S.; Tennakoon, M. A.; Wang, J.; Zhao, H.
J. Am. Chem. Soc. 2004, 126, 10210-10211
(13) Mancuso, A. J.; Huang, S. L.; Swern, D. J. Org. Chem. 1978, 43,
2480-2482.
(14) Jurczak, J.; Gryko, D.; Kobrzycka, E.; Gruza, H.; Prokopowiczit,
P. Tetrahedron 1998, 54, 6051-6064.
(15) Monenschein, H.; Sourkouni-Argirusi, G.; Schubothe, K. M.;
O’Hare, T.; Kirschning, A. Org. Lett. 1999, 1, 2101-2104.
Org. Lett., Vol. 6, No. 22, 2004
4079

Scheme 3. Final Manipulation to Target Galactosylceramides
Table 3. Three Bioassays of Galactosyl Ceramide Ligands
time (h) T (h)
IL-4a
ligand × 10-3 (pg/mL) max IFNγ (pg/mL) max IL-4 IL-4
malariaa IFNγa
IFNγ/
none
250
25
<1
<1
0
1800
2100
600
0
1100
475
1
2
3
12
24
12
2
2
2
1.64
4.42
2.4
250
a 1 µg of ligand per mouse.
3-carbon (or 4-bond) linker. These materials are inactive.
This is the first demonstration in R-galactosylceramides that
a four-bond connector between galactose-C1 and ceramide
C-N precludes their recognition by the receptors of the
subject immune cascade.
In conclusion, we have developed a practical strategy for
our second-generation synthesis of the R-C-galactosylcera-
mide structure. This convergent construction exhibits very
high efficiency with an overall yield of 30% after 11 steps
for 2 (CRONY 101) starting from commercial phytosphin-
gosine. Compared with our previous synthesis, it is more
conveniently scaled up as a prelude for extensive bioassay
research. Beyond our specific application, CM synthesis of
C-vinyl glycosides and ethylene-promoted cross metathesis
have potential application for the construction of glyco-
conjugates and also may lead to improvements in the general
profile of cross metathesis.
21 (Scheme 3). Only one step requires comment. Addition
of triethylsilane in the acidic hydrolytic step21 blocked a
troublesome trifluoroacetylation.
We briefly describe three bioassays of alkene 3 with
C-glycoside 2 and O-glycoside 1 as positive controls. In the
mouse malaria assay, the animals were treated with glycolipid
(or no treatment as control) and then challenged with an
injection of sporozoites. After 48 h, the animals were
sacrificed and their livers were assayed for sporozoites.
Column 2 of Table 3 shows that both 2 and 3 are effective
at the 1 µg level in blocking sporozoite viability. The results
of two cytokine assays, IFNγ and IL-4, are shown in the
remaining columns. In the IL-4 assay, the O-glycoside 1 has
by far the most powerful effect. It is hypothesized that IL-4
and IFNγ are antagonistic.4 Thus, the relatively low levels
of IL-4 produced by the C-glycosides may permit a more
effective stimulation of the NKT cell cascade as compared
to the O-series. The time courses for IFNγ and IL-4
production are not correlated with each other or with the
malaria results. Thus, one can tentatively conclude that
variance of effects are not simply due to differences between
O- and C-glycolipid lifetimes (which was the fundamental
rationale for the study of C-analogues). Not shown is data
for the C-glycoside homologues 4 and 5 where we have a
Acknowledgment. The research was supported by grants
from NIH: to RWF for synthesis (GM 60271); to MT for
immunology (AI 45606, AI 01682, AI 47840) and to Hunter
College for infrastructure support (RCMI-RR 03037).
Supporting Information Available: Experimental pro-
cedures and characterization data. This material is available
free of charge via the Internet at http://pubs.acs.org.
(21) Pekari, K.; Schmidt, R. R. J. Org. Chem. 2003, 68, 1295-1308.
OL0482137
4080
Org. Lett., Vol. 6, No. 22, 2004