Preparation of a lipid A derivative that contains a 27-hydroxyoctacosanoic acid moiety

Article abstract of DOI:10.1021/ol048746f

(Chemical Equation Presented) A general synthetic strategy for long-chain ω-1 hydroxy fatty acids has been developed, which employs as a key reaction step a cross metathesis between ω-unsaturated ester and 3-butene-2-ol. The resulting lipids were used for

Full text of DOI:10.1021/ol048746f

ORGANIC
LETTERS
2004
Vol. 6, No. 19
3333-3336
Preparation of a Lipid A Derivative That
Contains a 27-Hydroxyoctacosanoic
Acid Moiety
Balaji Santhanam and Geert-Jan Boons*
Complex Carbohydrate Research Center, The UniVersity of Georgia,
315 RiVerbend Road, Athens, Georgia 30602
gjboons@ccrc.uga.edu
Received July 1, 2004
ABSTRACT
A general synthetic strategy for long-chain ω-1 hydroxy fatty acids has been developed, which employs as a key reaction step a cross
metathesis between ω-unsaturated ester and 3-butene-2-ol. The resulting lipids were used for the preparation of lipid A derivatives of Rhizobium
sin-1, which have the ability to inhibit the E. coli LPS-dependent synthesis of tumor necrosis factor by human monocytes.
Lipopolysaccharides (LPS), which are important components
of the outer leaflet of the outer membrane of Gram-negative
bacteria,1 are major virulence factors of pathogenic bacteria2,3
as well as for symbionts such as the nitrogen-fixing Rhizobia
of legumes.4 LPS consist of an O-chain polysaccharide, a
core oligosaccharide, and an amphiphilic moiety referred to
as lipid A. The structure of lipid A is largely conserved
among most enteric bacteria, consisting of a â-(1-6)-linked
glucosamine disaccharide backbone with phosphate mo-
noesters at C-1 and C-4′ and â-hydroxyl fatty acyl groups
and acyloxyacyl residues at positions 2 and 3, and 2′ and 3′,
respectively (Figure 1). The nitrogen-fixing symbiont Rhizo-
(1) Alexander, C.; Zahringer, U. Trends Glycosci. Glycotechnol. 2002,
14, 69-86.
(2) Rietschel, E. T.; Brade, H.; Holst, O.; Brade, L.; Muller-Loennies,
S.; Mamat, U.; Zahringer, U.; Beckmann, F.; U., S.; Brandenburg, K.; Ulmer,
A. J.; Mattern, T.; Heine, H.; Schletter, J.; Loppnow, H.; Schonbeck, U.;
Flad, H. D.; Hauschildt, S.; Shade, F. U.; Di Padova, F.; Kusumoto, F.;
Figure 1. Structures of E. coli and R. sin-1 LPS.
Schumann, R. R. Curr. Top. Microbiol. Immunol. 1996, 216, 39-80.
(3) Raetz, C. R. H.; Whitfield, C. Annu. ReV. Biochem. 2002, 71, 635-
bium sin-1 has a structurally unusual lipid A (Figure 1),
differing in almost every aspect from those of enteric LPS.5
Its disaccharide backbone is devoid of phosphate, and the
700.
(4) Noel, K. D.; Vandenbosch, K. A.; Kulpaca, B. J. Bacteriol. 1986,
168, 1392-1401.
10.1021/ol048746f CCC: $27.50 © 2004 American Chemical Society
Published on Web 08/21/2004

glucosamine phosphate is replaced by 2-aminogluconolac-
tone. Another unique structural feature is the presence of an
unusual long-chain 27-hydroxyoctacosanoic acid (27OHC28:
0) moiety, which in turn can be esterified by â-hydroxy-
butyrate. It has been suggested that the presence of
27OHC28:0 fatty acid in the lipid A of many Rhizobial
species is required for maintaining the stability of the
bacterial membrane during endocytotic invasion and is
crucial for the survival of the bacterium within the plant-
derived symbiosome compartment.6-8 The 27OHC28:0 fatty
acid is also present among a number of facultative intra-
cellular pathogens that cause chronic infections such as
Brucella abortus, Bartonella henselae, and Legionella pneu-
mophila.6
We envisaged that olefin cross metathesis between an
ω-unsaturated ester (e.g., 6, Scheme 1) and 3-butene-2-ol
Scheme 1. Preparation of 27-Hydroxy-octacosanoic Acid
Recently, we demonstrated that LPS from R. sin-1 inhibits
the E. coli LPS-dependent synthesis of tumor necrosis factor
(TNF-R) by human monocytes.9,10 An LPS-mediated over-
production of host-derived inflammatory mediators such as
TNF-R may result in septicemia, which is a life-threatening
syndrome for which currently no treatment exists but
supportive therapy in an intensive care unit setting.11,12 Thus,
compounds such as R. sin-1 LPS may have the potential to
prevent the deleterious effects of enteric LPS. As a result of
the inherent molecular heterogeneity of R. sin-1 LPS, it
cannot be developed as a therapeutic agent for Gram-negative
sepsis. We have, however, demonstrated10 that a synthetic
analogue of the lipid A of R. sin-1 emulates the ability of
heterogeneous R. sin-1 LPS to antagonize enteric LPS, albeit
with somewhat higher IC50 values. The synthetic compound
contained an octacosanoic acid moiety rather than the natural
hydroxylated 27-hydroxyoctacosanoic acid.
To determine the contributions of the hydroxylation of the
long-chain 27-hydroxyoctacosanoic acid moiety for antago-
nistic properties, an efficient preparation of this fatty acid
was required. Furthermore, it was necessary that a synthetic
procedure was developed that allowed the 27-hydroxyocta-
cosanoic acid moiety to be introduced at a late stage of
synthesis. The new synthetic approach would also need to
allow an easy incorporation of other fatty acids for structure-
activity relationship studies.
(7) would give access to any ω-1 hydroxyl fatty ester after
the reduction of the double bond. In general, high selectivity
in olefin cross metathesis can be achieved when the two
olefins have significantly different reactivities. In this respect,
it has been shown that secondary allylic alcohols are of lower
reactivity than terminal olefins.14 Thus, it was expected that
a metathesis reaction of 6 with 7 would give a product in
good overall yield.
Methyl 25-hexacosenoiate (6), which is a starting material
for the cross metathesis reaction, was prepared from com-
mercially available bifunctional starting materials of ap-
propriate chain lengths (Scheme 1). Thus, reaction of
9-decen-1-ol (1) with CBr4 and PPh3 gave bromide 2 in a
yield of 92%, which was converted into a cuprate by reaction
with magnesium followed by transmetalation with dilithium
tetrachlorocuprate (Li2CuCl4). Condensation of the cuprate
with bromide 4 gave 25-hexacosenoic acid (5),15 which was
transformed into the corresponding methyl ester 6 by
treatment with freshly prepared diazomethane. A cross
metathesis reaction14 of ω-unsaturated ester 6 and 3-butene-
2-ol (7) using Grubbs first generation catalyst gave 8 in a
low yield of 15%. Fortunately, when Grubbs second genera-
tion catalyst was employed, compound 8 was isolated in a
much improved yield of 65% as mainly the trans-isomer (E/Z
) 20:1). A significantly lower yield of the cross metathesis
product was obtained when the hydroxyl of 7 was protected
as an acetyl ester or benzyl ether. Next, the double bond of
7 was reduced by hydrogenation over Rh/alumina to give 8.
Finally, benzylation of the ω-1 hydroxyl of 8 by treatment
with benzaldehyde, TMS2O, TMSOTf, and Et3SiH (f 9)
followed by saponification of the methyl ester with LiOH
gave 10 in a 73% overall yield (two steps).
An efficient approach for the chemical synthesis of long-
chain ω-1 hydroxy fatty acids (>C18) has not yet been
reported. These compounds have been obtained by an
enzymatic hydroxylation of inexpensive saturated fatty acid,13
a procedure that gave, however, mixtures of compounds in
which ω-2 and ω-3 were also hydroxylated.
(5) Jeyaretnam, B.; Glushka, J.; Kolli, V. S. K.; Carlson, R. W. J. Biol.
Chem. 2002, 277, 41802-41810.
(6) Bhat, U. R.; Carlson, R. W.; Busch, M.; Mayer, H. Int. J. Syst.
Bacteriol. 1991, 41, 213-217.
(7) Kannenberg, E. L.; Carlson, R. W. Mol. Microbiol. 2001, 39, 379-391.
(8) Vedam, V.; Kannenberg, E. L.; Haynes, J. G.; Sherrier, D. J.; Datta,
A.; Carlson, R. W. J. Bacteriol. 2003, 185, 1841-1850.
(9) Van den Plas, M. L.; Carlson, R. W.; Jeyaretnam, B.; McNeill, B.;
Barton, M. H.; Norton, N.; Murray, T. F.; Moore, J. N. J. Biol. Chem.
2002, 277, 41811-41816.
(10) Demchenko, A. V.; Wolfert, M. A.; Santhanam, B.; Moore, J. N.;
Boons, G. J. J. Am. Chem. Soc. 2003, 125, 6103-6112.
(11) Roberts, F. J.; Geere, I. W.; Coldman, A. ReV. Infect. Dis. 1991,
13, 34-46.
(12) Pittet, D.; Tarara, D.; Wenzel, R. P. J. Am. Med. Assoc. 1994, 271,
1598-1601.
(13) Schneider, S.; Wubbolts, M. G.; Sanglard, D.; Witholt, B. Tetra-
hedron: Asymmetry 1998, 9, 2833-2844.
(14) Chatterjee, A. K.; Choi, T. L.; Sanders, D. P.; Grubbs, R. H. J. Am.
Chem. Soc. 2003, 125, 11360-11370.
(15) Mirviss, S. B. J. Org. Chem. 1989, 54, 1948-1951.
3334
Org. Lett., Vol. 6, No. 19, 2004

Scheme 2. Synthesis of R. sin-1 Lipid A
Having successfully prepared 27-benzyloxy-octacosanoic
mediate can be acylated with other fatty acids (e.g., octa-
cosanic acid leading to compound 25), thus creating an
opportunity for the convenient preparation of a library of
compounds for structure-activity relationship studies.
Compound 20 could easily be prepared from the selec-
tively protected disaccharide 1110 and Lev ester or benzyl
ether protected â-hydroxy fatty acids 13,16 16, and 1817
acid 10, attention was focused on the preparation of lipid A
derivative 26 (Scheme 2), which contains this fatty acid
moiety. Lipid A derivatives are usually prepared by first
synthesizing â-hydroxy acyl- and acyloxoacyl acids, which
are then condensed with amines or alcohols of an ap-
propriately protected saccharide. We envisaged that an
acylation of the â-hydroxyl of the C-2′ myristate moiety of
20 with 27-hydroxyoctacosanoic acid 10 would offer a more
convergent approach and efficient use of expensive starting
material. Furthermore, the hydroxyl of this advanced inter-
(16) Duynstee, H. I.; van Vliet, M. J.; van der Marel, G. A.; van Boom,
J. H. Eur. J. Org. Chem. 1998, 303-307.
(17) Oikawa, M.; Kusumoto, S. Tetrahedron: Asymmetry 1995, 6, 961-
966.
Org. Lett., Vol. 6, No. 19, 2004
3335

(Scheme 2). Thus, removal of the phthalimido group and
acetyl esters of 11 by treatment with ethylenediamine in
refluxing n-butyl alcohol followed by selective N-acetylation
of 12 with 13 in the presence of DCC gave 14 in an overall
yield of 88%. Reduction of the azido moiety of 14 was easily
accomplished by reaction with 1,3-propanedithiol in a
mixture of pyridine, triethylamine, and water,18 and the amine
of the resulting compound 15 was acylated with 16 using
DCC as the activating reagent to give 17 in an overall yield
of 55%. The C-3 and C-3′ hydroxyls of 17 were acetylated
with 18 using DCC and DMAP to give 19 in 67% yield.
Thus, by employing the DDC-mediated acylation in the
presence or absence of DMAP, selectivity between O- and
N-acylation could be accomplished.
Selective removal of the Lev ester of 19 could easily be
achieved by treatment with hydrazine acetate19 to give the
advanced intermediate 20 in an excellent yield. The hydroxyl
of 20 was then acylated with octacosanic acid or 27-
benzyloxy-octacosanoic acid 10 using DCC and DMAP as
the activating reagent to give 21 and 22, respectively.
Hydrolysis of the thiophenyl moiety of 21 and 22 was
accomplished with NIS in wet dichloromethane in the
presence of TfOH.20 The corresponding lactols (R/â mix-
tures) were then oxidized with PCC to the lactones 23 and
24, respectively. Finally, the benzyl ethers and benzylidene
acetal of 23 and 24 were removed by catalytic hydrogenation
over Pd/C to afford the target compounds 25 and 26,
respectively.
In conclusion, an efficient synthetic approach for ω-un-
saturated ester has been developed by olefin cross metathesis
between an ω-unsaturated ester and 3-butene-2-ol followed
by a reduction of the double bond of the resulting product.
Furthermore, it has been demonstrated that an acylation of
the â-hydroxyl of the myristate moiety of a lipid A precursor
(e.g., 20) gives easy access to a range of compounds for
structure-activity relationship studies. In this study we have
employed two different fatty acids, namely, the octacosanoic
acid and 27-hydroxyoctacosanoic acid, to clarify the effect
of the ω-1 hydroxyl group to antagonistic properties of
natural Rhizobial lipid A. The results of these biological
studies will be reported elsewhere.
Acknowledgment. This research was supported by the
NIH Research Resource Center for Biomedical Complex
Carbohydrates (P41-RR-5351).
(18) Bayley, H.; Standring, D. N.; Knowles, J. R. Tetrahedron Lett. 1978,
Supporting Information Available: Experimental pro-
3633-3634.
1
cedures and H and 13C NMR spectra. This material is
(19) Van Boom, J. H.; Burgers, P. M. J. Tetrahedron Lett. 1976, 4875-
4878.
available free of charge via the Internet at http://pubs.acs.org.
(20) Veeneman, G. H.; van Leeuwen, S. H.; van Boom, J. H. Tetrahedron
Lett. 1990, 31, 1331-1334.
OL048746F
3336
Org. Lett., Vol. 6, No. 19, 2004