α-glycosphingolipids via chelation-induced anomerization of O- And s-glucuronic and galacturonic acid derivatives

Article abstract of DOI:10.1021/ol802915h

Bacterial glycolipids containing either α-glucuronic acid or α-galacturonic acid residues have an important role in the innate-type immune response to Gram-negative bacteria. Synthesis of closely related compounds, including a novel α-SO₂ glycolipid mimetic, is described from carbohydrate precursors where anomerization is a key step. Very high stereoselectivites (>97:3 in favor of α) were observed from O-glycoside precursors.

Full text of DOI:10.1021/ol802915h

ORGANIC
LETTERS
2009
Vol. 11, No. 4
939-942
r-Glycosphingolipids via
Chelation-Induced Anomerization of O-
and S-Glucuronic and Galacturonic Acid
Derivatives
Wayne Pilgrim† and Paul V. Murphy*,†,‡
UCD School of Chemistry and Chemical Biology and Centre for Synthesis and
Chemical Biology, UniVersity College Dublin, Belfield, Dublin 4, Ireland, and School
of Chemistry, National UniVersity of Ireland, Galway, UniVersity Road,
Galway, Ireland
paul.V.murphy@nuigalway.ie
Received December 18, 2008
ABSTRACT
Bacterial glycolipids containing either r-glucuronic acid or r-galacturonic acid residues have an important role in the innate-type immune
response to Gram-negative bacteria. Synthesis of closely related compounds, including a novel r-SO2 glycolipid mimetic, is described from
carbohydrate precursors where anomerization is a key step. Very high stereoselectivites (>97:3 in favor of r) were observed from O-glycoside
precursors.
Sphingomonous cell walls that do not contain lipopolysac-
charide (LPS) present glycosphingolipids such as GSL-1 1,1
PBS30 2, and PBS59 3.2 These contain glucuronic acid or
galacturonic acid residues R-linked to sphingosine derivatives
(Figure 1). These glycolipids are structurally related to the
antitumor agent KRN7000 4. The glycoprotein CD1d
presents 4 and stimulates natural killer T (NKT) cells to
release cytokines.3 Recently, 1-3 have been shown to be
targets for mice and human NKT cells.1,2 They can lead to
septic shock and bacterial clearance in infected mice,
demonstrating that glycolipids stimulate an innate-type
immune response to gram negative bacteria not presenting
LPS. Although ꢀ-glycosyl ceramides are more abundant in
† University College Dublin.
‡ National University of Ireland, Galway.
(2) Kinjo, Y.; Wu, D.; Kim, G.; Xing, G.-W.; Poles, M. A.; Ho, D. D.;
Tsuji, M.; Kawahara, K.; Wong, C.-H.; Kronenberg, M. Nature 2005, 434,
520–525.
(1) Mattner, J.; DeBord, K. L.; Ismail, N.; Goff, R. D.; Cantu, C.; Zhou,
D.; Saint-Mezard, P.; Wang, V.; Gao, Y.; Hoebe, K.; Schneewind, O.;
Walker, D.; Beutler, B.; Teyton, L.; Savage, P.; Bendelac, A. Nature 2005,
434, 525–529.
(3) Kobayashi, E.; Motoki, K.; Uchida, T.; Fukushima, H.; Koezuka,
Y. Oncol. Res. 1995, 7, 529–34.
10.1021/ol802915h CCC: $40.75
Published on Web 01/29/2009
2009 American Chemical Society

Scheme 2. Retrosynthesis of 5
Figure 1. Structures 1-4.
nature, the R-glycosidic linkage in 1-3 is essential for their
biological properties.
The development of novel strategies for the synthesis of
R-glycolipids and their mimetics is thus important in order
to generate and to evaluate the potency of new glycolipid
antigens. A number of syntheses of the glucuronic and
galacturonic acid based glycolipids have been reported,
and these have been based on the oxidation of a preformed
glycoside4 or glycoside bond formation using a uronic acid
based donor containing nonparticipating and “armed” pro-
tecting groups at C-2.5 The latter approach gave mixtures
of anomers in the glycoside bond forming reactions in some
cases. Herein, we describe the synthesis of the R-linked
uronic acid based glycolipids based on chelation-induced
anomerization (Scheme 1), which generates the R-linkage
enhancement was explained by chelation-induced anomer-
ization (Scheme 1) wherein a SnCl4 or TiCl4 catalyst gives
a 5-membered ring chelate after coordination to both the
pyranose ring oxygen atom and to the C-6 carbonyl group.
Consequently, anomerization would occur via cleavage of
the C1-O5 bond7 (endocyclic cleavage) as opposed to
cleavage of the glycosidic C1-O1 bond (exocyclic cleavage).
The synthesis of 6 was thus undertaken in order to determine
whether anomerization could be used to generate an inter-
mediate with the desired R-linkage that after subsequent
amide formation and deprotection would give 5. It was
envisaged that 6 would be obtained via glycoside coupling
of the acceptor 7 and donor 8, which would in turn be
prepared respectively from D-galactal and D-glucose.
Scheme 1. Chelation-Induced Anomerization
The dihydroceramide derivative 7 was first obtained from
3,4,6-tri-O-benzyl-D-galactal (Scheme 3).8 Initial reaction of
Scheme 3. Synthesis of 7
with very high stereoselectivity (>97:3) for both glucuronic
acid and galacturonic acid derivatives, even though partici-
pating and disarmed protecting groups are present. Chelation-
induced anomerization of S-glycosides is also demonstrated.
The retrosynthetic analysis of the initial target compound
5 is shown in Scheme 2. The rates of anomerization have
recently been observed to be faster for some glucuronic acid
derivatives than for their corresponding glucosides.6 This rate
(4) Okamoto, N.; Kanie, O.; Huang, Y.-H.; Fujii, R.; Watanabe, H.;
Shimamure, M. Chem. Biol. 2005, 12, 677–683.
(5) (a) Long, X.; Deng, S.; Mattner, J.; Zang, Z.; Zhou, D.; McNary,
N.; Goff, R. D.; Teyton, L.; Bendelac, A.; Savage, P. B. Nat. Chem. Biol.
2007, 3, 559–563. (b) Wu, D.; Xing, G.-W.; Poles, M. A.; Horowitz, A.;
Kinjo, Y.; Sullivan, B.; Bodmer-Narkevitch, V.; Plettenburg, O.; Kronen-
berg, M.; Tsuji, M.; Ho, D. D.; Wong, C.-.H Proc. Natl. Acad. Sci. U.S.A.
2005, 102, 1351–1356. (c) Stallforth, P.; Adibekian, A.; Seeberger, P. H.
Org. Lett. 2008, 10, 1573–1576.
the galactal under the Perlin conditions9 and subsequent
acetylation, as indicated also recently,10 gave the aldehyde
9.11 Reaction of 9with the Wittig reagent prepared by reaction
940
Org. Lett., Vol. 11, No. 4, 2009

prepared from D-glucose. Oxidation using TEMPO, subse-
quent base-promoted esterification of the acid with allyl
iodide, and reaction of the corresponding ester with
HBr-AcOH gave the R-glycosyl bromide 15. This bromide
was converted to the trichloroacetimidate donor 8 by
hydrolysis of 15 promoted by silver carbonate and reaction
of the resulting hemiacetal with DBU and trichloroacetoni-
trile.
Scheme 4. Synthesis of 8
The glycoside-coupling reaction of 7 and 8 was next
investigated (Scheme 5). This reaction gave the ꢀ-glycoside
of the phosphonium salt 1012 with n-butyllithium and
subsequent de-O-acetylation gave a diene as a mixture of
stereoisomers. Subsequent mesylation of this intermediate
diene gave 11. In contrast with the previous finding,10 we
believe the EZ isomer (EZ/EE, 12:1) is the major product
from the Wittig reaction in this sequence. Our assigment is
Scheme 5. Synthesis of 5
1
made on the basis of signals in the H NMR spectra (500
MHz) of the diene product mixture of 11. For H-6, a signal
for the EZ isomer is observed at δ 6.00 as an apparent triplet
with a coupling constant (J) of 11.0 Hz; the H-6 signal of
the EE isomer (minor product) appears at δ 6.06 (double
doublet overlapping) with a J of 15.1 Hz observed. These J
values are consistent with the EZ stereochemical assignment
for the major product from the Wittig reaction. Catalytic
hydrogenation of 11, followed by introduction of TES groups
and subsequent substitution of the mesyl group with azide,
gave 12. Removal of the TES groups from 12 using
TBAF-THF gave the diol 13. Efforts to carry out glyco-
sylation reactions with a variety of acylated glucuronic acid
donors and acceptors 1213 or 13 were unsuccessful. Thus,
13 was converted to the benzyl derivative 7 via the
regioselective introduction of a TIPS group at the primary
alcohol group, benzylation of the secondary hydroxyl group,
and subsequent removal of the TIPS group. The use of the
TIPS groups was found to be necessary in this sequence of
reactions, as it was more stable than a TBS protecting group,
for instance, which migrated during attempted benzylation
of the secondary hydroxyl group. Glycoside bond formation
from 7 was ultimately successful.
16 in excellent yield. The acetylated analogue of 8 was also
investigated as a donor, but in glycosylation reactions an
orthoester was obtained rather than the desired glycoside.
With the ꢀ-glycoside 16 in hand, its anomerization reactions
were investigated. Gratifyingly, anomerization with con-
comitant removal of the benzyl protecting group proceeded
efficiently using TiCl4 in dichloromethane to give 17 with
high stereoselectivity (R/ꢀ, 97:3) and yield. The Staudinger
reaction of azide 17 gave an amine that was converted to
amide 18 after treatment with nonadecanoyl chloride in the
presence of triethylamine. Removal of the protecting groups
from 18 gave target compound 5. The deprotection was
effected using hydroperoxide generated in n-propanol from
sodium propoxide and hydrogen peroxide. The use of
stronger and harder bases such as hydroxide and methoxide
led to the elimination of benzoic acid and formation of the
undesired unsaturated compound 19. The synthesis of the
The glycosyl donor 8 was obtained in seven steps from
D-glucose. Partially protected derivative 1414 was first
(6) O’Brien, C.; Polakova, M.; Pitt, N.; Tosin, M.; Murphy, P. V. Chem.
Eur. J. 2007, 13, 902–9.
(7) Lemieux, R. U. AdV. Carbohydr. Chem. 1954, 9, 1.
(8) For use of D-galactal derivatives in synthesis of sphingosine
derivatives by a different route to that described herein, see: Wild, R.;
Schmidt, R. R. Liebigs Ann. 1995, 75, 5–64.
(9) Gonzalez, F; Lesage, S.; Perlin, A. S Carbohydr. Res. 1975, 42,
267–74.
(10) For a closely related strategy to sphingosines and related compounds
recently published, see: Kokatla, H. P.; Sagar, R.; Vankar, Y. D. Tetrahedron
Lett. 2008, 49, 4728–30.
(11) For a recent application of this reaction in synthesis of dorrigocin
A analogues see Anquetin, G.; Rawe, S. L.; McMahon, K.; Murphy, E.;
Murphy, P. V. Chem. Eur. J. 2008, 14, 1592–1600.
(12) Dauben, W. G.; Gerdes, J. M.; Bunce, R. A. J. Org. Chem. 1984,
49, 4293–4295.
(13) For a synthesis of R-O-glucuronic acid derivatives from TMS or
TES ethers and a 6,1-lactone derivative, see: (a) Pola´kova´, M.; Pitt, N.;
Tosin, M.; Murphy, P. V. Angew. Chem., Int. Ed. 2004, 43, 2518–21. (b)
Tosin, M.; Murphy, P. V. Org. Lett. 2002, 4, 3675–78.
(14) Kovac, P.; Glaudemans, C. P. J. J. Carbohydr. Chem. 1988, 7, 317.
Org. Lett., Vol. 11, No. 4, 2009
941

bromide 23 via a mesylated intermediate. Next, coupling of
thiol 22 and alkyl bromide 23 using NaH (<1 equiv) gave
the ꢀ-S-thioglycoside 24. Gratifyingly, the TiCl4-catalyzed
anomerization gave the R-S-glycoside 25 in 55% yield. A
4:1 mixture of R- and ꢀ-thioglycolipids were obtained in
this latter reaction, and they were separated by silica gel
chromatography. Conversion of the azide group to an amine
and subsequent acylation followed by removal of the
protecting groups and concomitant oxidation of the sulfur
atom under the deprotection conditions gave 26, which can
be considered a novel glycolipid mimetic. The R-S-gly-
colipid16 would also be an interesting compound to obtain,
but conditions have not yet been established in our group
that preclude the use of hydrogen peroxide, which oxidizes
the thioglycoside.
Scheme 6. Synthesis of R-SO2 Containing Glycolipid
In summary, we have described the synthesis of glycolipids
and mimetics by a chelation-induced anomerization17 of
protected ꢀ-glycolipid precursors (Scheme 6). An advantage
of this strategy is that both ꢀ- and R-glycolipids can in
principle be generated. Deprotection conditions that will
preclude oxidation of sulfur need to be established with a
view to generating the S-glycolipid. The biological properties
of the new analogues will be described in due course.18
Acknowledgment. We are grateful to the UCD NMR and
MS Centres. The research was funded by the Programme
for Research in Third Level Institutions funded by the Higher
Education Authority of Ireland and the Astellas USA
Foundation.
Supporting Information Available: Experimental pro-
cedures and H and 13C NMR spectra. This material is
available free of charge via the Internet at http://pubs.acs.org.
1
galacturonic acid derivative 20 was also achieved using an
identical strategy from the corresponding galacturonic acid
donor. Within this sequence, the anomerization reaction gave
the desired R-O-galacturonic acid product in quantitative
yield and the ꢀ-anomer was not detected by NMR analysis,
indicating the stereoselectivity was >97:3.
Having achieved the synthesis of the target O-glycolipids,
we next investigated S-glycolipid anomerization. To the best
of our knowledge, the anomerization of S-glycosides have
rarely, if at all, been applied to complex neoglycoconjugate
synthesis.15 Thus, ꢀ-glycosyl thiol 22 was next prepared in
two steps from the galacturonosyl bromide derivative 21,
which had been obtained from a route similar to that
described for preparation of 15. Alcohol 7 was converted to
OL802915H
(15) For a review on synthesis of S-oligosaccharides and S-glycopeptides,
see: Pachamuthu, K.; Schmidt, R. R. Chem. ReV. 2006, 106, 160–87.
(16) The synthesis of the S-glycoside analogue of KRN7000 has recently
been reported. See: (a) Dere, R. T.; Zhu, X. Org. Lett. 2008, 10, 4641–4.
(b) Blauvelt, M. L.; Khalili, M.; Jaung, W.; Paulsen, J.; Anderson, A. C.;
Wilson, S. B.; Howell, A. R. Bioorg. Med. Chem. Lett. 2008, 18, 6374–76.
(17) For anomerization of mannopyranoside derivatives where chelation
is invoked, see: Wang, Y.; Cheon, H.-W.; Kishi, Y. Chem. Asian J. 2008,
3, 319–326.
(18) The results described herein were outlined in an oral communication
recently: Murphy, P. V.; Pilgrim, W. Novel synthesis of glycolipids
containing glucuronic acid. Abstract of Papers, 236th ACS National
Meeting, Philadelphia, PA, United States, August 17-21, 2008; American
Chemical Society: Washington, DC, 2008; CARB-092.
942
Org. Lett., Vol. 11, No. 4, 2009