Chemical Synthesis of helicobacter pylori lipopolysaccharide partial structures and their selective proinflammatory responses

Article abstract of DOI:10.1002/chem.201003581

Helicobacter pylori is a common cause of gastroduodenal inflammatory diseases such as chronic gastritis and peptic ulcers and also an important factor in gastric carcinogenesis. Recent reports have demonstrated that bacterial inflammatory processes, such as stimulation with H. pylori lipopolysaccharide (LPS), initiate atherosclerosis. To establish the structures responsible for the inflammatory response of H. pylori LPS, we synthesized various kinds of lipid A structures (i.e., triacylated lipid A and Kdo-lipid A compounds), with or without the ethanolamine group at the 1-phosphate moiety, by a new divergent synthetic route. Stereoselective α-glycosylation of Kdo N-phenyltrifluoroacetimidate was achieved by use of microfluidic methods. None of the lipid A and Kdo-lipid A compounds were a strong inducer of IL-1β, IL-6, or IL-8, suggesting that H. pylori LPS is unable to induce acute inflammation. In fact, the lipid A and Kdo-lipid A compounds showed antagonistic activity against cytokine induction by E. coli LPS, except for the lipid A compound with the ethanolamine group, which showed very weak agonistic activity. On the other hand, these H. pylori LPS partial structures showed potent IL-18- and IL-12-inducing activities. IL-18 has been shown to correlate with chronic inflammation, so H. pylori LPS might be implicated in the chronic inflammatory responses induced by H. pylori. These results also indicated that H. pylori LPS can modulate the immune response: NF-κB activation through hTLR4/MD-2 was suppressed, whereas production of IL-18 and IL-12 was promoted.

Full text of DOI:10.1002/chem.201003581

DOI: 10.1002/chem.201003581
Chemical Synthesis of Helicobacter pylori Lipopolysaccharide Partial
Structures and their Selective Proinflammatory Responses
Atsushi Shimoyama,[a] Akinori Saeki,[a] Natsuko Tanimura,[b] Hiroko Tsutsui,[c]
Kensuke Miyake,[b] Yasuo Suda,[d] Yukari Fujimoto,*[a] and Koichi Fukase*[a]
Abstract: Helicobacter pylori is
a
thetic route. Stereoselective a-glycosy-
lation of Kdo N-phenyltrifluoroaceti-
midate was achieved by use of micro-
fluidic methods. None of the lipid A
and Kdo-lipid A compounds were a
strong inducer of IL-1b, IL-6, or IL-8,
suggesting that H. pylori LPS is unable
to induce acute inflammation. In fact,
the lipid A and Kdo-lipid A com-
pounds showed antagonistic activity
against cytokine induction by E. coli
LPS, except for the lipid A compound
with the ethanolamine group, which
showed very weak agonistic activity.
On the other hand, these H. pylori LPS
partial structures showed potent IL-18-
and IL-12-inducing activities. IL-18 has
been shown to correlate with chronic
inflammation, so H. pylori LPS might
be implicated in the chronic inflamma-
tory responses induced by H. pylori.
These results also indicated that
H. pylori LPS can modulate the
immune response: NF-kB activation
through hTLR4/MD-2 was suppressed,
whereas production of IL-18 and IL-12
was promoted.
common cause of gastroduodenal in-
flammatory diseases such as chronic
gastritis and peptic ulcers and also an
important factor in gastric carcinogene-
sis. Recent reports have demonstrated
that bacterial inflammatory processes,
such as stimulation with H. pylori lipo-
polysaccharide (LPS), initiate athero-
sclerosis. To establish the structures re-
sponsible for the inflammatory re-
sponse of H. pylori LPS, we synthe-
sized various kinds of lipid A structures
(i.e., triacylated lipid A and Kdo-
lipid A compounds), with or without
the ethanolamine group at the 1-phos-
phate moiety, by a new divergent syn-
Keywords: glycolipids · glycosyla-
tion · inflammation · immunochem-
istry · synthetic methods
Introduction
proximately 10–20% of infected people will develop peptic
ulcers, and 1% will develop gastric cancer.[3] Lipopolysac-
charide (LPS) of Gram-negative bacteria, also called endo-
toxin, is generally a potent immunostimulator and can cause
sepsis during severe infection. However, the H. pylori LPS
shows very low toxic activity in comparison with other en-
terobacterial preparations, such as Escherichia coli LPS, and
seems instead to be a key factor in virulence expression and
the innate immune response to infection.[4–7] In fact, the
plague bacillus Yersinia pestis produces LPS with poor stim-
ulating activity for Toll-like receptor 4 (TLR4) to evade
LPS-induced inflammation and to establish a deadly infec-
tion.[8]
It was recently reported that chronic infections are associ-
ated with atherosclerosis, which is a chronic inflammatory
vascular disease[9–11] and a major factor in cardiovascular dis-
ease, a primary cause of death worldwide. The inflammatory
process in atherosclerosis is thought to be initiated by infec-
tious agents such as H. pylori,[10,11] Chlamydia pneumo-
niae,[12–14] or Porphyromonas gingivalis.[15–18] In particular,
LPS has been implicated in atherosclerosis,[14–16,18] and Trian-
tafilou et al. showed that LPS from H. pylori and P. gingiva-
lis induces inflammatory responses in human vascular endo-
thelial cells.[19] Several recent reports also suggested a link-
age between neuropathological conditions in the central
nervous system (CNS), such as Alzheimerꢀs disease, and in-
fectious agents including some viruses and bacteria (e.g.,
Helicobacter pylori is one of the Gram-negative bacteria,
and its chronic colonization of the human stomach is the
major risk factor for gastroduodenal inflammatory diseases
such as chronic gastritis and peptic ulcers and also contrib-
utes to gastric carcinogenesis.[1,2] The prevalence of H. pylori
infections is estimated to be half of the world population.
Although the majority of infections are asymptomatic, ap-
[a] Dr. A. Shimoyama, A. Saeki, Prof. Y. Fujimoto, Prof. K. Fukase
Department of Chemistry, Graduate School of Science
Osaka University, Toyonaka, Osaka 560-0043 (Japan)
Fax : (+81)6-6850-5419
E-mail: yukarif@chem.sci.osaka-u.ac.jp
[b] Dr. N. Tanimura, Prof. K. Miyake
Division of Infectious Genetics
Institute of Medical Science, The University of Tokyo
4-6-1 Shirokanedai, Minatoku Tokyo 108-8639 (Japan)
[c] Prof. H. Tsutsui
Department of Microbiology, Hyogo College of Medicine
1-1, Mukogawa, Nishinomiya, Hyogo 663-8501 (Japan)
[d] Prof. Y. Suda
Department of Chemistry, Biotechnology and Chemical Engineering
Graduate School of Science and Engineering, Kagoshima University
Kagoshima 890-0065 (Japan)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201003581.
14464
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 14464 – 14474

FULL PAPER
H. pylori, C. pneumoniae, and spirochete).[20] Proinflamma-
tory cytokines, particularly interleukin 18 (IL-18),[21] are
thought to be related to chronic inflammatory diseases such
as atherosclerosis[22] and to neuroinflammatory/neurodege-
nerative processes in the CNS.[23] Although various studies
have shown a correlation between up-regulation of IL-18
and H. pylori infection,[24–26] the substances responsible for
IL-18 production have not been well elucidated, except in
Sudaꢀs report,[27] which indicated that H. pylori LPS induces
IL-18. We hence synthesized various lipid A partial struc-
tures in order to investigate the immunoregulatory effects of
H. pylori lipid A and LPS and to establish the structures re-
sponsible for the inflammatory response.
The endotoxic activity of LPS originates from lipid A, a
glycolipid terminal structure of LPS. In general, lipid A is
composed of two glucosamine (GlcN) units and fatty acids
linked to the 2- and 2’-amino and 3- and/or 3’-hydroxy
groups (Scheme 1, below). In the series of our synthetic
studies of lipid A,[28–30] we found that lipid A compounds
with smaller numbers of acyl groups (tri- and tetraacylated
with saturated fatty acids, for example) showed antagonistic
activity (against cytokine induction by E. coli LPS), whereas
the E. coli-type hexaacylated lipid A showed potent immu-
nostimulating activity. We observed that the presence of a
Kdo residue along with lipid A enhances the activities.[31]
The lipid A moieties of parasitic bacteria (e.g., H. pylori
and P. gingivalis) have characteristic structures, generally
containing fewer but longer fatty acids than those of entero-
bacteria such as E. coli and frequently lacking one of the
two phosphates. In fact, the characteristics of H. pylori
lipid A are as follows: 1) it has fewer (tri- and tetraacylation
are mostly found) but longer fatty acid residues (mainly
C18), 2) the 4’-phosphate group is often absent, and 3) an
ethanolamine group is often linked to the glycosyl phos-
phate moiety.[32,33] Detailed structural characterizations of
the lipid A components of H. pylori were performed inde-
pendently on the 206-1 strain by Suda et al.[32] and on the
NCTC 11637 strain by Moran et al.[33] The H. pylori 206-1
strain mainly produced a triacyl-type lipid A, whereas the
NCTC 11637 strain mainly produced the tetraacyl-type
lipid A in rough and smooth LPS (R- and S-LPS) and the
hexaacyl-type in the S-LPS as a minor component.
pure compounds is also the case for P. gingivalis lipid A, and
some synthetic works have also previously been done[35–37]
for analysis of the immunostimulatory activities.
In this study we also synthesized a compound of the
H. pylori Kdo-lipid A aminoethanol phosphate type
(HpKdoLAEA, 4) for the first time, as well as resynthesiz-
ing compounds 1–3 to elucidate their biological functions in
detail. We developed a new divergent strategy, in which the
key intermediate 9 (Scheme 1, below) was used as a
common precursor for the various lipid A and Kdo-lipid A
structures of H. pylori, P. gingivalis, and others. The diver-
gent strategy was also employed in our previous synthesis
and in Boonsꢀ synthesis.[38,39] Boons et al. synthesized Kdo-
lipid A of Neisseria meningitides by a divergent strategy
with use of a different synthetic intermediate.[40] Glycosyla-
tion of the lipid A part with the Kdo donor was difficult to
accomplish, because considerable amounts of the glycal
were formed as a by-product. We found that glycosylation
with the Kdo donor 6 under microfluidic conditions effec-
tively suppressed the glycal formation. The details of the
synthetic strategy and the biological activities of com-
pounds 1–4 are discussed below.
Results
The new diversity-oriented synthetic strategy for H. pylori
lipid A/Kdo-lipid A compounds is illustrated in Scheme 1.
The appropriately protected disaccharide 9 was synthesized
as a common key intermediate from the monosaccharide
donor 7 and the acceptor 8. Each protecting group in 9 (i.e.,
1-O-allyl, 2-N-allyloxycarbony (Alloc), 2’-N-2,2,2-trichloroe-
thyloxycarbonyl (Troc), 3-O-benzyloxycarbonyl (Z), 3’-O-p-
methoxybenzyl (MPM), and 4’,6’-benzylidene) could be se-
lectively removed to enable sequential introduction various-
ly of the Kdo moiety, acyl groups, or phosphate groups at
the appropriate position. We also developed the new Kdo
N-phenyltrifluoroacetimidate donor 6, which can be activat-
ed by catalytic quantities of a Lewis acid. The acid-labile
MPM group was hence utilizable as a temporary protective
group in the lipid A part.
As shown in Scheme 2, the lipid A moieties were synthe-
sized with and without the ethanolamine component at the
1-phosphate (compounds 1 and 2). The glycosyl donor 7 and
the acceptor 8 were synthesized independently from d-glu-
cosamine hydrochloride in six steps (Schemes S1 and S2 in
the Supporting Information). The donor 7 was synthesized
via allyl 4,6-O-benzylidene-GlcNTroc,[38] with introduction
of MPM and subsequent allyl group cleavage and introduc-
tion of trichloroacetoimidate at the anomeric position. The
The LPS and lipid A preparations from H. pylori are het-
erogeneous and could be contaminated by other immunosti-
mulants, so chemically synthesized pure compounds were re-
quired for precise biological studies. We have therefore pre-
viously synthesized the H. pylori lipid A structures triacyl-
type H. pylori lipid A (HpLA, 1, Scheme 1, below), triacyl-
type H. pylori lipid A aminoethanol phosphate (HpLAEA,
2), and H. pylori Kdo-lipid A (HpKdoLA, 3).[27,31,34]
HpLAEA (2), which contains an ethanolamine residue, in-
duced lower levels of cytokines, such as IL-8 and TNF-a,
upon activation of the LPS receptor, a Toll-like receptor 4
(TLR4)/MD-2 complex. In contrast, compounds 1 and 3
showed antagonistic activity against stimulation by E. coli
LPS. HpKdoLA (3) showed more potent inhibitory activity
than HpLA (1). A requirement for chemically synthesized
acceptor
8 was prepared via allyl 4,6-O-benzylidene-
GlcNAlloc,[41] with introduction of MPM and selective open-
ing of the benzylidene group to liberate the 6-hydroxy
group, followed by cleavage of the MPM group. Glycosyla-
tion between the donor 7 and the acceptor 8 afforded the
disaccharide with complete b-selectivity, owing to the neigh-
boring effect of the Troc group of 7, and the 3-position in
Chem. Eur. J. 2011, 17, 14464 – 14474
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
14465

Y. Fujimoto, K. Fukase et al.
Scheme 1. Synthetic strategy for the triacyl-type Helicobacter pylori lipid A and Kdo-lipid A.
Scheme 2. The synthesis of the Helicobacter pylori lipid A compounds 1 and 2.
the disaccharide was protected with a Z group to give the
key intermediate 9. The 2’-N-Troc group was then removed
through the action of a Zn/Cu couple and acetic acid and
the amino group was acylated with the acyloxycarboxylic
acid 10 (Scheme S3 in the Supporting Information) by Shii-
naꢀs method.[42] Other coupling methods gave considerable
amounts of the symmetrical acid anhydride of 10, which
showed low reactivity in acylation here. Subsequently, the 2-
N-Alloc group of 11 was removed by treatment with Pd-
AHCTUNGTRENNUNG
(PPh3)4 and dimethylaminotrimethylsilane (TMSDMA),[43]
and the free 2-amino group was smoothly acylated with the
benzyloxycarboxylic acid 12 (Scheme S3 in the Supporting
Information) with use of O-(7-azabenzotriazol-1-yl)-
N,N,N’,N’-tetramethyluronium
hexafluorophosphate
(HATU) and 4-(dimethylamino)pyridine (DMAP) to give
the protected lipid A backbone 13 in quantitative yield. The
14466
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 14464 – 14474

Synthesis and Proinflammatory Responses of H. pylori LPS Structures
FULL PAPER
allyl group was then isomerized to a prop-1-enyl group with
an Ir complex,[44] and the resulting prop-1-enyl group was
then removed with iodine and water to yield the disacchar-
ide 14, with a 1-hydroxy group. 1-O-a-Phosphorylation was
carried out by 1-O-lithiation with lithium hexamethyldisila-
zide (LHMDS), and subsequent treatment with tetrabenzyl
pyrophosphate in THF at ꢀ788C gave the desired 1-O-a-
phosphate 15 with high stereoselectivity.[45] After rapid but
careful purification, all protecting groups of 15 were re-
moved by hydrogenolysis with H2 (2.0 MPa) in the presence
of Pd-black to give HpLA (1) in quantitative yield.
For the synthesis of HpLAEA (2), phosphitylation of 14
with the phosphoramidite[46,47] in the presence of 1H-tetra-
zole and MS (4 ꢂ) gave the 1-O-phosphite, which was then
oxidized with DMDO[48] to afford the desired fully protected
1-O-a-phosphorylated 17. Hydrogenolysis of 17 with
Pd(OH)2/C under H2 (2.0 MPa) in THF/H2O/AcOH gave
the HpLAEA (2) in 89% yield.
glycosyl acceptors, but the reactions did not proceed. In-
stead, the glycals were formed as by-products with intramo-
lecular b-hydrogen elimination. We then examined glycosy-
lation with the disaccharide acceptor 19, which showed
better reactivity than the acylated disaccharide intermedi-
ates. The glycosylation was carried out at 08C with the iso-
propylidene Kdo donor 6 in excess (5 equiv) relative to 19
in the presence of TBSOTf (0.5 equiv) as a catalyst to give
the desired trisaccharide 20 in 70% yield with high a-selec-
tivity (Scheme 3B). The a-promoting solvent effect of cyclo-
pentyl methyl ether (CPME) was employed here.[51,52] The
trisaccharide 20 was thus obtained, although a large amount
of the glycal was formed. On the other hand, glycosylation
of 19 with the Kdo donor 18 did not occur at all. The lower
reactivity of 4,5-O-TBS Kdo fluoride relative to the corre-
sponding 4,5-O-isopropylidene was also observed in our pre-
vious study.[49]
Although the yield and stereoselectivity of the newly de-
veloped Kdo glycosylation were acceptable, a large excess
of the glycosyl donor 6 was required for completion of the
reaction. The low efficiency in the glycosylation might be
due to difficulties in temperature control under the conven-
tional batch reaction conditions, because the donor 6 was
highly reactive and would be expected to generate a signifi-
cant amount of heat during the reaction. We had previously
reported a-selective sialylation with N-phenyltrifluoroaceti-
midate donors by use of a continuous flow microreactor,
which enables efficient mixing and fast heat transfer, to in-
crease the stereoselectivity and to suppress glycal forma-
tion.[53] We thus applied microfluidic glycosylation of the
Kdo donor 6, expecting suppression of the glycal formation
owing to efficient removal of heat of reaction as well as the
promotion of intermolecular coupling reaction because of
efficient mixing (Scheme 3C). For this microfluidic Kdo gly-
cosylation, a CPME solution of the Kdo donor 6 and the ac-
ceptor 19 was mixed with TBSOTf solution in CPME at var-
ious concentrations in an IMM microreactor at a flow rate
of 0.5 mLminꢀ1. The reaction mixture was allowed to flow
for an additional 42 seconds through a tubular reactor (F=
1.0 mm) and then quenched by addition to a saturated aque-
ous solution of sodium bicarbonate. After the optimization
as shown in Scheme 3, the desired compound 20 was ob-
tained in 72% yield with high stereoselectivity through
mixing of a solution of the donor 6 (0.15m) and the acceptor
19 (0.10m) in CPME with a solution of TBSOTf (0.10m) in
CPME. Although the yield of microfluidic glycosylation was
almost the same as that of the batch glycosylation, the for-
mation of glycal was suppressed and only 1.5equivalents of
Kdo donor 6 relative to 19 were necessary for the microflui-
dic glycosylation. The rapid reaction achieved by the use of
relatively higher concentrations of the Lewis acid led to the
success of this microfluidic Kdo glycosylation without cleav-
age of the protecting groups.
The polysaccharide component of LPS is linked to lipid A
through a unique acidic octose: 3-deoxy-d-manno-2-octulo-
sonic acid (Kdo). In our previous synthetic study of E. coli
Re LPS compounds composed of lipid A and two Kdo resi-
dues, we showed that Kdo moieties enhanced the immunos-
timulating activity of lipid A.[49] On the other hand, HpKdo-
LA (3) showed stronger antagonistic activity than the
lipid A part (HpLA, 1).[31] Lipid A parts have previously
been considered to represent the activity of LPS, but our
studies showed that the effect of the Kdo varies depending
on the LPS species. We hence synthesized HpKdoLAEA (4)
to investigate the effect of the Kdo residue on the lipid A
part in this study. Our group has developed a-selective gly-
cosylation through the use of Kdo fluorides with bulky pro-
tecting groups such as isopropylidene and TBS groups at the
4- and 5-positions, and applied this to the synthesis of E. coli
Re-type LPS.[49] With these Kdo donors, the undesirable b-
side attack by a glycosyl acceptor is prevented by the pres-
ence of the bulky protective groups. Glycosylation with Kdo
fluorides, however, requires a large excess of Lewis acid for
the activation,[49] which has occasionally caused cleavage of
the acid-labile TBS or isopropylidene groups and hence de-
creased the stereoselectivity in the glycosylation. The syn-
thetic strategy was also constrained by the limited number
of protective groups resistant to strongly acidic conditions.
In this study, we therefore devised the new Kdo donors 6
and 18, possessing N-phenytrifluoroacetimidate moieties.
Glycosyl N-phenytrifluoroacetimidates have been shown to
be very efficient donors, which can be activated by catalytic
amounts of Lewis acid under mild conditions.[50] The Kdo
donors 6 and 18 were synthesized from the intermediate 5
(Scheme 1), which was prepared from d-mannose by our re-
ported methods (Scheme S4 in the Supporting Informa-
tion).[49] The glycosyl disaccharide acceptor 19 was prepared
by regioselective reductive ring-opening of the 4’,6’-O-ben-
zylidene group in
(Scheme 3A).
We first attempted the glycosylation of the Kdo donors 6
and 18 with the acylated disaccharide intermediates as the
9
with Me2NH·BH3 and BF3·OEt2
The synthesis of the desired Kdo-lipid A compounds 3
and 4 from the trisaccharide 20 is shown in Scheme 4. The
2’-N-Troc group of 20 was removed with a Zn/Cu couple
and the resulting amino group was acylated with the acyl-
Chem. Eur. J. 2011, 17, 14464 – 14474
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
14467

Y. Fujimoto, K. Fukase et al.
Scheme 3. The synthesis of Kdo-lipid A. A) The preparation of the disaccharide 19. B), C) The a-selective glycosylation of Kdo B) by the conventional
batch method, and C) by the microfluidic method.
oxycarboxylic acid 10 in the presence of 2-methyl-6-nitro-
benzoic anhydride (MNBA) and N,N-diisopropylethylamine
(DIPEA). For the acylation at this 2’-position, we also ex-
amined conditions with HATU, DMAP, and Et3N, but these
required more fatty acid and longer reaction times to finish
the reaction, although the yield was much the same. The 2-
N-Alloc group was then selectively cleaved by treatment
in 21 were removed to yield 22. The anomeric position in 22
was selectively phosphorylated with tetrabenzyl pyrophos-
phate and LHMDS, and all protecting groups were then re-
moved by hydrogenolysis (H2, 20 MPa) in the presence of
Pd-black to afford the desired HpKdoLA 3 in quantitative
yield.
HpKdoLAEA (4) was synthesized by use of the phos-
phoroamidite method for 1-O-phosphorylation, as in the
synthesis of 2. Because of the high reactivity of the phos-
phoroamidite reagent it was necessary to protect all of the
hydroxy groups except for the target 1-OH group for the
phsphorylation step. The isopropylidene group in 21 was
with PdACHTUNGTRENNUNG(PPh3)4 and TMSDMA, and the liberated amino
group was acylated with the benzyloxycarboxylic acid 12 in
the presence of HATU and Et3N to give the acylated trisac-
charide 21 in an excellent yield (92%). For the synthesis of
the HpKdoLA (3), the isopropylidene and 1-O-allyl groups
14468
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 14464 – 14474

Synthesis and Proinflammatory Responses of H. pylori LPS Structures
FULL PAPER
Scheme 4. The synthesis of the Helicobacter pylori Kdo-lipid A compounds 3 and 4.
converted into a benzylidene group for the final deprotec-
tion by catalytic hydrogenolysis. The allyl group of the ben-
zylidenated trisaccharide was cleaved to give the 1-O-free
compound 24. Phosphitylation of 24 was carried out with
the phosphoramidite, the resulting phosphite was oxidized
with dimethyldioxirane (DMDO), and all protecting groups
were then removed by hydrogenolysis with Pd(OH)2/C in
THF/H2O/AcOH under H2 (2.0 MPa) to give the HpKdo-
LAEA (4).
The immunomodulating activities of the synthesized
H. pylori lipid A and Kdo-lipid A compounds were evaluat-
ed by induction or suppression of cytokines such as IL-1b,
IL-6, IL-8, and IL-18 (Figure 1), as well as IL-12 and TNF-a
(Figure S1 in the Supporting Information), relative to the
E. coli LPS strain O111:B4. The cytokine-inducing activities
were tested in human peripheral whole blood cells.[54]
A
mixture of a test sample and heparinized human peripheral
whole blood was incubated at 378C for 20 h. The cytokine
levels in the supernatants of the incubated mixtures were
measured by an enzyme-linked immunosorbent assay
(ELISA). Inhibitory activities were measured according to
the abilities of the compounds to inhibit LPS-induced cyto-
kine production as follows: the prepared samples, LPS
(E. coli O111:B4, 500 pgmLꢀ1), and heparinized human pe-
ripheral whole blood were mixed and incubated, and the cy-
tokine levels were measured. Because of the lower solubili-
ties of the synthesized H. pylori lipid A and Kdo-lipid A
compounds, concentrations higher than those in Figure 1
were not tested. Only the HpLAEA showed weak inducing
activities of IL-1b, IL-6, IL-8, and TNF-a, whereas HpLA,
Chem. Eur. J. 2011, 17, 14464 – 14474
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
14469

Y. Fujimoto, K. Fukase et al.
Figure 1. Cytokine induction and inhibition by the synthesized H. pylori triacyl-type lipid A compounds 1 (HpLA) and 2 (HpLAEA) and by the Kdo-
lipid A compounds 3 (HpKdoLA) and 4 (HpKdoLAEA), relative to E. coli LPS O111:B4. Human peripheral whole-blood cells were assayed. A) Induc-
tion and B) inhibitory activities of IL-1b. C) Induction and D) inhibitory activities of IL-6. E) Induction and F) inhibitory activities of IL-8. G) Induction
and H) inhibitory activities of IL-18. In the inhibitory assay, the E. coli LPS coexisted in the samples during the incubation at a concentration of
500 pgmlꢀ1
.
HpKdoLA, and HpKdoLAEA showed not inducing activi-
ties but inhibitory activities toward cytokine induction by
E. coli LPS (Figure 1A–F, Figure S1 A–C in the Supporting
Information). HpKdoLA showed stronger inhibitory activi-
ties than HpLA, as we had previously reported. In the case
of the aminoethanol phosphate-type lipid A compounds, the
Kdo residue changed the activity of lipid A. Synergic en-
hancements of cytokine production with HpLAEA in com-
bination with E. coli LPS were also observed (Figure 1B,
1D, and 1F). In the case of HpLA, synergic enhancement
was also observed in the IL-6 inhibition assay between
HpLA concentrations of 40 ngmLꢀ1 and 1.25 mgmLꢀ1 in
HPWB (Figure S2 in the Supporting Information). The IC50
value of HpLA was not determined because of the synergic
effects, but the IC50 of HpKdoLA in IL-6 inhibition was de-
termined to be 340 ngmLꢀ1 (Figure S2 in the Supporting In-
formation).
On the other hand, all the synthesized H. pylori lipid A
and Kdo-lipid A compounds showed IL-18-inducing activity
(Figure 1G and 1H). HpLA showed the strongest inducing
activity for IL-18. In addition, all of the synthetic com-
pounds showed potent IL-12-inducing activity relative to the
E. coli LPS (Figure S1A in the Supporting Information).
The lipid A compounds, HpLA and HpLAEA, showed
stronger IL-12-inducing activity than the corresponding
Kdo-lipid A compounds HpKdoLA and HpKdoLAEA.
14470
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 14464 – 14474

Synthesis and Proinflammatory Responses of H. pylori LPS Structures
FULL PAPER
HpLA was shown to be the most potent IL-18 and IL-12 in-
ducer, whereas the same compound inhibited IL-1b, IL-6,
and IL-8 induction.
fects the binding of the lipid A part of the structure to the
MD-2/TLR4 complex and consequently elicits different re-
sponses. Another important finding is that all four synthe-
sized LPS partial structures, HpLA, HpKdoLA, HpLAEA,
and HpKdoLAEA, are potent inducers of IL-18 and IL-12
in human peripheral whole blood (HPWB). In this case, the
presence of Kdo decreased the potencies of the activities.
These results clearly indicated that IL-18 and IL-12 were in-
duced by H. pylori lipid A and Kdo-lipid A by pathways
that differed from the IL-1b, IL-6, and IL-8 pathways acti-
vated by E. coli LPS. E. coli LPS is a potent inducer of the
MyD88-dependent pathway factors, such as TNF-a, IL-1b,
IL-6, and IL-8, through activation of NF-kB. HpLA,
HpKdoLA, and HpKdoLAEA were antagonistic to the in-
duction of these factors and selectively activated the IL-18
and IL-12 pathways.
IL-18 was originally identified as a potent inducer of in-
terferon-g (IFN-g)[21] and has recently been shown to play a
role in the regulation of Th1 and Th2 responses,[58,59] as well
as providing a link between IL-12 and IL-18 in the regula-
tion of IFN-g production.[60] IL-18 is produced from pro-IL-
18 through activation with caspase-1. However, the activa-
tion pathway of caspase-1 from procaspase-1 remains a
matter of debate. Nakanishi et al. reported that caspase-1 is
activated by NALP3 through a TLR4- and TRIF-dependent
pathway and leads to IL-18 activation in Kupffer cells.[61]
NuÇez et al., on the other hand, described a TLR4-inde-
pendent caspase-1 activation pathway.[62] More precisely,
their research indicated that NALP3 is essential for caspase-
1 activation and that the NALP3-mediated caspase-1 activa-
tion proceeds independently of TLR signaling (TLR2 and
TLR4), the activation instead being mediated by pannexin-
1, a hemichannel protein that interacts with the P2X7 recep-
tor.[62,63] Our results and these reports indicated that IL-18
induction by LPS was independent of a NF-kB activation
pathway, although the pathways activated by H. pylori
lipid A and LPS have not been clarified. In contrast, IL-12
secretion was reported to be MyD88-dependent.[64] HpLA,
HpKdoLA, and HpKdoLAEA suppressed the induction of
MyD88-dependent cytokines, such as IL-1b, IL-6, and IL-8,
suggesting the presence of a MyD88-independent pathway
for the induction of IL-12.
Discussion
Recent studies have suggested several pathological implica-
tions of bacterial LPS due to their influence on the host
immune responses. Synthetic studies of lipid A and LPS par-
tial structures have contributed to investigations of the rela-
tionships of structures and functions of LPS, but these rela-
tionships are still not well understood. To address this issue,
we established an efficient synthetic strategy that is widely
applicable to the synthesis of various lipid A compounds, as
well as that of lipid A compounds containing Kdo moieties.
The strategy (Schemes 1 and 2) employs a key disaccharide
intermediate with selectively cleavable protecting groups at
the 1-O-, 2-N-, 3-O-, 2’-N-, 3’-O-, 4’-O-, and 6’-O-positions
for the synthesis of H. pylori lipid A compounds (HpLA
and HpLAEA) and Kdo-lipid A compounds (HpKdoLA
and HpKdoLAEA). Natural H. pylori LPS have shown only
very low endotoxic activity, so low activities or the absence
of immunostimulating activities would be expected for the
synthetic compounds. A bioassay clearly indicated that
HpLA, HpKdoLA, and HpKdoLAEA acted as antagonists
against LPS with respect to induction of IL-1b, IL-6, and IL-
8. On the other hand, the ethanolamine-form lipid A
HpLAEA showed weak activity in inducing IL-1b, IL-6, and
IL-8. In the previous reports, HpLAEA showed immunosti-
mulating activity through TLR4.[55] In these inhibition
assays, synergic enhancement effects were observed with
HpLAEA (Figure 1B, 1D, and 1F) and with HpLA (Fig-
ure S2 in the Supporting Information) in combination with
E. coli LPS. Phenomena of this kind have also previously
been observed with mixtures of the TLR4 agonist (synthetic
E. coli lipid A, compound 506) and the strong antagonist
(synthetic lipid A precursor IVa, compound 406), in which a
mixture of both compounds showed higher biological activi-
ties than the agonist itself in an inflammatory cytokine in-
duction assay.[56] We also observed similar effects with mix-
tures of weak antagonists (lipid A analogues) and a strong
agonist (E. coli LPS).[29] In natural LPS, one Kdo residue is
attached to lipid A, so a natural H. pylori LPS with a triacy-
lated lipid A should function as an antagonist in the induc-
tion of IL-1b, IL-6, and IL-8 through TLR4.
Cytokine induction by H. pylori LPS partial structures
demonstrates that these four types of structure either do not
induce or induce very low levels of the inflammatory cyto-
kines IL-1b, IL-6, and IL-8 through NF-kB stimulated by
hTLR4/MD-2. However, they simultaneously induce high
levels of IL-12 and IL-18. This phenomenon could give rise
to strong effects on chronic inflammatory reactions, provid-
ing a link to atherosclerosis and other inflammatory diseas-
es. The induction of both IL-18 and IL-12 would initiate Th1
cell development, and might be related to the finding that
H. pylori colonization is inversely associated with childhood
asthma.[65]
An important finding is that the presence of a Kdo resi-
due altered the activity of the lipid A compound HpLAEA.
Several reports, including our own studies, have indicated
that Kdo residues enhance the activities of lipid A com-
pounds.[57] However, no synthetic studies have previously re-
ported a transition from immunostimulating to antagonistic
activity (or vice versa) through the introduction of a Kdo
moiety. The molecular basis for the different activities is
presumably due to the differences in charge: HpLA,
HpKdoLA, and HpKdoLAEA are anionic, whereas
HpLAEA is neutral. The difference in charge probably af-
Chem. Eur. J. 2011, 17, 14464 – 14474
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
14471

Y. Fujimoto, K. Fukase et al.
(0.03%) in EtOH/conc. H2SO4/acetic acid buffer or ninhydrin (0.2%) in
EtOH/collidine/acetic acid buffer. MS (4 ꢂ and 3 ꢂ) were activated by
heating at 2508C in vacuo for 3 h before use. Unless otherwise stated, all
reactions were performed at room temperature. Non-aqueous reactions
were carried out under argon unless otherwise noted. Anhydrous CH2Cl2
was distilled from calcium hydride. Anhydrous THF and CPME were
purchased from Kanto Chemicals Japan. Anhydrous DMF was purchased
from NACALAI TESQUE, Japan. Distilled water was purchased from
Otsuka (Japan) or prepared by use of a combinations of Arium 611 UV
(Sautorius) or Toray Pure LV-308 (Toray) and GSL-200 (Advantec,
Japan). All other reagents and solvents used were also purchased from
commercial sources.
Conclusion
In this study we have established efficient methods for the
synthesis of various lipid A and Kdo/lipid A compounds and
have developed the a-selective glycosylation of Kdo and
lipid A backbone through application of a microfluidic reac-
tion. Using these newly developed methods, we succeeded
in the synthesis of the H. pylori LPS partial structures
HpLA (1), HpLAEA (2), HpKdoLA (3), and HpKdo-
LAEA(4), including the first total synthesis of HpKdo-
LAEA (4). The immunomodulatory activities of these syn-
thesized compounds revealed important links to the viru-
lence of H. pylori LPS. Namely, the results reported here
are fundamental to the understanding of the molecular basis
for the characteristic bioactivities of H. pylori LPS, especial-
ly with respect to cytokine induction by the lipid A-Kdo
structures (including higher induction of IL-18), which are
closely related to their virulence expression. In this study we
have focused on the synthesis of triacylated compounds.
Boons et al. synthesized tetraacylated lipid A compounds
derived from P. gingivalis and showed that the synthetic
lipid A compounds are antagonistic in humans.[37] They also
reported the antagonism of Rhizobium sin-1 lipid A, with a
2-aminogluconolactone and a very long chain fatty acid (27-
hydroxyoctacosanoic acid) lacking phosphate.[66,67] Piecing
together our results and theirs, the lipid A compounds from
these bacteria containing long-chain fatty acids but with
smaller overall numbers of acyl groups show antagonistic ac-
tivity in humans. Our synthetic methods in this report are
widely applicable for the synthesis of various kinds of
lipid A compounds, and we have recently applied them for
the synthesis of tetraacylated H. pylori lipid A (aminoetha-
nol phosphate-type) and Porphyromonas gingivalis lipid A
compounds (tri-, tetra-, and pentaacylated). The immuno-
modulatory activities, especially IL-12- and IL-18-inducing
activities, of these parasitic bacterial lipid A compounds will
also be revealed with the synthetic compounds.
Synthesis of H. pylori Kdo-lipid A (20)—batch reaction
Allyl 2-allyloxycarbonylamino-4-O-benzyl-6-O-{6-O-[(benzyl 7,8-di-O-
benzyl-3-deoxy-4,5-O-isopropylidene-a-d-manno-oct-2-ulopyranosid)o-
nate]-4-O-benzyl-2-deoxy-3-O-(4-methoxyphenylmethyl)-2-(2,2,2-trichlor-
oethoxycarbonylamino)-b-d-glucopyranosyl}-2-deoxy-a-d-glucopyrano-
side (20): A solution of TBSOTf (15.6 mL, 68.0 mmol) in dry CPME
(300 mL) was added at 08C to a mixture of the imidate donor 6 (489 mg,
680 mmol), the acceptor 19 (146 mg, 136 mmol), and MS (4 ꢂ) in dry
CPME (2.2 mL) and the mixture was stirred under Ar for 20 min. After
addition of saturated aqueous NaHCO3, the mixture was extracted with
CHCl3. The organic layer was washed with saturated aqueous NaHCO3
and brine, dried over anhydrous Na2SO4, filtered, and concentrated in
vacuo. The residue was purified by silica gel flash column chromatogra-
phy (toluene/EtOAc 20:1!10:1!3:1) to give 20 as
a white solid
(150 mg, 70%). 1H NMR (500 MHz, CDCl3): d=7.31–7.17 (m, 32H;
C6H5-CH2-, CH3O-C6H4-CH2-, C6H5-CH2-OCO-), 6.83 (d, J=8.5 Hz, 2H;
CH3O-C6H4-CH2-), 5.85 (ddd, J=22.6, 10.7, 5.6 Hz, 1H; -OCH2-CH=CH2
of Alloc group), 5.75 (ddd, J=22.5, 10.9, 5.7 Hz, 1H; -OCH2-CH=CH2 of
allyl group), 5.25 (dd, J=17.2, 1.3 Hz, 1H; -OCH2-CH=CH2 of Alloc
group), 5.20–5.11 (m, 6H; -OCH2-CH=CH2 of Alloc and allyl groups,
C6H5-CH2-OCO-), 5.05 (dd, J=10.5, 9.3 Hz, 1H; H-3), 5.01–4.98 (m, 2H;
2-NH, C6H5-CH2-OCO-), 4.89 (brs, 1H; 2’-NH), 4.73 (d, Jgem =11.2 Hz,
1H; C6H5-CH2-), 4.71 (d, J=2.7 Hz, 1H; H-1), 4.68 (d, Jgem =10.8 Hz,
1H; C6H5-CH2-), 4.66 (d, Jgem =9.0 Hz, 1H; CH3O-C6H4-CH2-), 4.63 (d,
J
gem =11.3 Hz, 2H; C6H5-CH2-), 4.58 (d, Jgem =11.0 Hz, 1H; CH3O-C6H4-
CH2-), 4.57 (d, Jgem =12.6 Hz, 2H; C6H5-CH2-), 4.51–4.44 (m, 6H; C6H5-
CH2-, -OCH2-CH=CH2 of Alloc group, -COO-CH2-CCl3), 4.41–4.36 (m,
2H; H-1’, H-4’’), 4.33 (dd, J=7.0, 1.3 Hz, 1H; H-5’’), 4.06 (d, J=10.5 Hz,
1H; H-7’’), 4.01 (dd, J=13.0, 5.2 Hz, 1H; -OCH2-CH=CH2 of allyl
group), 3.98 (dt, J=4.5, 2.0 Hz, 1H; H-5), 3.96–3.91 (m, 1H; H-2), 3.84–
3.81 (m, 2H; H-6a, H-6’’), 3.79–3.65 (m, 10H; H-3’, H-8’’a, CH3O-C6H4-
CH2-, H-6b, H-4’, H-6’a, H-4, -OCH2-CH=CH2 of allyl group), 3.59 (dd,
J=13.5, 3.2 Hz, 1H; H-8’’b), 3.37 (t, J=9.2 Hz, 1H; H-5’), 3.34–3.28 (m,
2H; H-6’b, H-2’), 2.57 (dd, J=14.9, 5.0 Hz, 1H; H-3’’a), 1.90 (dd, J=
14.9, 5.0 Hz, 1H; H-3’’b), 1.33 (s, 3H; methyl group of isopropylidene),
1.27 ppm (s, 3H; methyl group of isopropylidene); HRMS (ESI-Q-TOF,
positive): m/z: calcd for C84H93Cl3N2O23 [M+Na]+: 1625.5132; found:
1625.5127.
Experimental Section
1
General procedures: H NMR spectra were recorded in the solvents indi-
Representative microfluidic reaction procedure for the synthesis of 20:
Dry CPME was injected into the micromixer by syringe pump in advance
and saturated the microfluidic system. Subsequently, a solution of the
cated with JEOL JNM-LA 500, JEOL ECA 500, Varian INOVA 600, or
Bruker AVANCE III 400 spectrometers. The chemical shifts in CDCl3
are given in the form of d values from tetramethylsilane (TMS) as an in-
ternal standard. For measurement in D2O, the HDO signal (4.718 ppm at
308C) was used as a reference. ESI-TOF mass spectrometry was carried
out with an Applied Biosystems Mariner Biospectrometry Workstation.
ESI-IT mass spectrometry was carried out with a Bruker amaZon SL-k1
instrument. High-resolution mass spectrometry was performed with a
Waters Q-Tof micro (ESI-QTOF) instrument. Elemental analyses were
performed with a Yanaco CHNcorder MT-6. Silica gel column chroma-
tography was carried out with Kieselgel 60 (Merck, 0.040–0.063 mm) or
Silica Gel 60 N (Kanto Chemical, spherical, neutral, 0.040–0.050 mm) at
medium pressure (0.2–0.4 MPa). Gel permeation chromatography (GPC)
was carried out with Sephadex LH20 at atmospheric pressure. Precoated
Kieselgel 60 F 254 (Merck, 0.5 mm) was used for preparative thin layer
chromatography. TLC analysis was performed on silica gel 60 F254
(Merck) and compounds were visualized by use of UV (254 nm), phos-
phomolybdic acid solution (5.0% in EtOH), p-methoxybenzaldehyde
donor
6 (48.0 mg, 66.7 mmol, 0.15m) and the acceptor 19 (48.0 mg,
44.4 mmol, 0.10m) in CPME (500 mL) and a solution of TBSOTf (9.8 mL,
44.4 mmol, 0.10m) in CPME (500 mL) were injected to the IMM micro-
mixer, each by syringe pump at a flow rate of 0.5 mLminꢀ1, and mixed at
08C. After the reaction mixture had been allowed to flow through a
stainless reactor tube (F=1.0 mm, l=25 cm) at 08C for 42 s, the mixture
was added dropwise to saturated aqueous NaHCO3 at 08C. The mixture
was extracted with CHCl3, washed with saturated aqueous NaHCO3 and
brine, dried over Na2SO4, filtered, and concentrated in vacuo to give the
crude product. The residue was purified by silica gel flash column chro-
matography (toluene/EtOAc 20:1!10:1!3:1) to give the trisaccharide
20 (51 mg, 72%) and the stereoselectivity was analyzed by 1H NMR spec-
troscopy (a/b>95:5).
Other syntheses and spectroscopic data are shown in the Supporting In-
formation.
14472
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 14464 – 14474

Synthesis and Proinflammatory Responses of H. pylori LPS Structures
FULL PAPER
2-Deoxy-6-O-{2-deoxy-2-[(R)-3-(octadecanoyloxy)octadecanoylamino]-b-
d-glucopyranosyl}-2-[(R)-3-hydroxyoctadecanoylamino]-a-d-glucopyra-
nose 1-phosphate (1): Pd-black (10.0 mg) was added to a solution of 15
(11.0 mg, 5.63 mmol) in distilled THF (1.0 mL). The mixture was stirred
under hydrogen (2.0 MPa) at room temperature for 18 h and then neu-
tralized with Et3N in THF (10%, v/v, 20.0 mL), and the Pd catalyst was
removed by filtration. After removal of the solvent in vacuo, the residue
was lyophilized from tBuOH to afford 1 as a triethylammonium salt
(white solid, 7.0 mg, quant). 1H NMR (500 MHz, CDCl3/CD3OD): d=
5.47 (d, J=4.4 Hz, 1H; H-1), 5.27 (brs, 1H; b-CH of 2’-N-acyl), 4.62 (d,
J=8.8 Hz, 1H; H-1’), 4.08 (t, J=9.8 Hz, 1H; H-5), 4.02 (d, J=12.1 Hz,
1H; b-CH of 2-N-acyl), 4.02 (d, J=8.6 Hz, 1H; H-2), 3.89 (d, J=11.8 Hz,
1H; H-6a), 3.83 (dd, J=11.8, 6.3 Hz, 1H; H-6b), 3.78–3.65 (m, 2H; H-3,
H-2’), 3.45–3.24 (m, 4H; H-3’, H-4’, H-6’a, H-6’b), 3.23–3.16 (m, 2H; H-
4, H-5’), 2.58 (t, J=6.3 Hz, 2H; a-CH2 of 2’-N-acyl main chain), 2.39–
2.27 (m, 4H; a-CH2 of 2’-N-acyl side chain and 2-N-acyl), 1.76–1.15 (m,
84H; CH2 of 2-N-acyl and 2’-N-acyl), 0.89 ppm (t, J=6.2 Hz, 9H; -CH2-
CH3 of 2-N-acyl and 2’-N-acyl); HRMS (ESI-Q-TOF, negative): m/z:
calcd. for C66H127N2O17P [MꢀH]ꢀ: 1249.8794; found: 1249.8794.
hexane and water and then was lyophilized from tBuOH to afford 4 as a
white solid (110 mg, 15%). H NMR is shown in the Supporting Informa-
1
tion. HRMS (ESI-Q-TOF, negative): m/z: calcd for C76H144N3O24P
[MꢀH]ꢀ: 1512.9799; found: 1512.9792.
Cytokine assay: The cytokine-inducing activities of the synthetic lipid A
compounds and LPS (E. coli O111:B4, Sigma–Aldrich as a positive con-
trol) were tested in heparinized human peripheral whole blood (HPWB)
cells collected from the blood of an adult volunteer. The synthesized
samples (dissolved in 25 mL of 5% DMSO in saline) and HPWB (25 mL)
in RPMI1640 medium (75 mL) including meylon (3%), were incubated in
triplicate in a 96-well plastic plate at 378C under CO2 (5%). To maintain
the solubility of the synthetic compounds this culture included DMSO
(1%).[54] After 20 h of incubation and subsequent centrifugal separation
(at 300g for 2 min), the supernatant proteins were collected and the in-
duced quantities of cytokines were then measured by ELISA assay
(ELISA Ready-SET-Go! (eBioscience) for human IL-1b, IL-6, IL-12p70,
IL-8, and TNF-a; Human IL-18 ELISA Kit (MBL) for human IL-18).
For the inhibitory activities of the synthetic lipid A compounds and Kdo-
lipid A compounds against E. coli LPS, the synthetic compounds and ad-
ditional E. coli LPS (500 pgmLꢀ1) in RPMI 1640 medium including
meylon (3%) and a buffer solution were incubated and measured by a
method similar to that used for the inducing activities. Data represent
averages of three repeated assays with standard deviations from individu-
al experiments.
2-Deoxy-6-O-{2-deoxy-2-[(R)-3-(octadecanoyloxy)octadecanoylamino]-b-
d-glucopyranosyl}-2-[(R)-3-hydroxyoctadecanoylamino]-a-d-glucopyra-
nose 1-(aminoethyl)phosphate (2): Pd(OH)2/C (20 wt% on carbon,
20.0 mg) was added to a solution of 17 (20.0 mg, 9.79 mmol) in distilled
THF (1.0 mL), water (100 mL), and AcOH (50 mL). The mixture was
stirred under hydrogen (2.0 MPa) at room temperature for 2 d and then
neutralized with Et3N in THF (10%, v/v, 20.0 mL), and Pd catalyst was
removed by filtration. After removal of the solvent in vacuo, the residue
was lyophilized from tBuOH to afford 2 as a colorless solid (11.2 mg,
89%). 1H NMR (500 MHz, CDCl3/CD3OD): d=5.47 (d, J=8.3 Hz, 1H;
H-1), 5.26–5.22 (m, 1H; b-CH of 2’-N-acyl), 4.62 (d, J=8.5 Hz, 1H; H-
1’), 4.21 (dd, J=18.2, 8.5 Hz, 1H; H-5), 4.04 (d, J=12.1 Hz, 1H; H-6a),
3.95–3.90 (m, 2H; H-2, b-CH of 2-N-acyl), 3.88 (d, J=11.8 Hz, 1H;
-OPO-CH2-CH2-), 3.81 (t, J=10.9 Hz, 1H; H-6b), 3.72–3.65 (m, 2H; H-3,
-OPO-CH2-CH2-), 3.60–3.55 (m, 1H; H-2’), 3.48–3.20 (m, 6H; H-3’, H-4’,
H-6’a, H-6’b, -OPO-CH2-CH2-), 3.13–3.10 (m, 1H; H-5’), 2.95 (brs, 1H;
H-4), 2.56–2.42 (m, 2H; a-CH2 of 2’-N-acyl main chain), 2.41–2.22 (m,
4H; a-CH2 of 2’-N-acyl side chain and 2-N-acyl), 1.62–1.15 (m, 84H;
CH2 of 2-N-acyl and 2’-N-acyl), 0.88 ppm (brs, 9H; -CH2-CH3 of 2-N-
acyl and 2’-N-acyl); HRMS (ESI-Q-TOF, negative): m/z: calcd for
C68H132N3O17P [MꢀH]ꢀ: 1292.9216; found: 1292.9211.
Acknowledgements
This work was supported in part by Grants-in Aid for Scientific Research
(Nos. 22310136, 20241053, 19310144, 17310128, and 20–870) and the
NEXT program from the Japan Society for the Promotion of Science, as
well as by grants from Osaka University Global COE program (Frontier
Biomedical Science Underlying Organelle Network Biology), the Naito
Foundation, and the Takeda Science Foundation. The authors would also
like to acknowledge Prof. Yasuhiro Kajihara, Seiji Adachi, and Naoya In-
azumi of Osaka University for the NMR measurements and fruitful dis-
cussions.
2-Deoxy-6-O-{2-deoxy-6-O-(3-deoxy-a-d-manno-oct-2-ulopyranosid)onic
acid]-2-[(R)-3-(octadecanoyloxy)octadecanoylamino]-b-d-glucopyrano-
[1] P. B. Ernst, B. D. Gold, Annu. Rev. Microbiol. 2000, 54, 615–640.
[2] R. M. Peek, Jr., J. E. Crabtree, J. Pathol. 2006, 208, 233–248.
[3] P. B. Ernst, D. A. Peura, S. E. Crowe, Gastroenterology 2006, 130,
188–206; P. B. Ernst, D. A. Peura, S. E. Crowe, Gastroenterology
2006, 130, 212–213.
[4] S. O. Hynes, J. A. Ferris, B. Szponar, T. Wadstrom, J. G. Fox, J. OꢀR-
ourke, L. Larsson, E. Yaquian, A. Ljungh, M. Clyne, L. P. Andersen,
A. P. Moran, Helicobacter 2004, 9, 313–323.
[5] H. Nielsen, S. Birkholz, L. P. Andersen, A. P. Moran, J. Infect. Dis.
1994, 170, 135–139.
[6] G. I. Perez-Perez, V. L. Shepherd, J. D. Morrow, M. J. Blaser, Infect.
Immun. 1995, 63, 1183–1187.
[7] S. Yokota, H. Saito, T. Kubota, N. Yokosawa, K. Amano, N. Fujii,
Virology 2003, 306, 135–146.
[8] S. W. Montminy, N. Khan, S. McGrath, M. J. Walkowicz, F. Sharp,
J. E. Conlon, K. Fukase, S. Kusumoto, C. Sweet, K. Miyake, S.
Akira, R. J. Cotter, J. D. Goguen, E. Lien, Nat. Immunol. 2006, 7,
1066–1073.
syl}-2-[(R)-3-benzyloxyoctadecanoylamino]-a-d-glucopyranose
1-phos-
phate (3): Pd-black (6.0 mg) was added to a solution of 23 (6.3 mg,
2.70 mmol) in distilled THF (500 mL). The mixture was stirred under hy-
drogen (2.0 MPa) at room temperature for 5 d and then neutralized with
Et3N in THF (10%, v/v, 20.0 mL), and Pd catalyst was removed by filtra-
tion. After removal of the solvent in vacuo, the residue was lyophilized
from tBuOH to afford 3 as a triethylammonium salt (white solid, 4.3 mg,
quant). 1H NMR (600 MHz, CDCl3/CD3OD): d=5.48 (brs, 1H; H-1),
5.26–5.18 (m, 1H; b-CH of 2’-N-acyl), 4.46–4.40 (m, 1H; H-1’), 4.08–3.92
(m, 5H; H-2, H-5, H-4’’, H-7’’, b-CH of 2-N-acyl), 3.84–3.56 (m, 9H; H-3,
H-6a, H-6b, H-2’, H-5’, H-5’’, H-6’’, H-8’’a, H-8’’b), 3.51–3.42 (m, 3H; H-
3’, H-6’a, H-6’b), 3.38–3.26 (m, 2H; H-4, H-4’), 2.56–2.50 (m, 2H; a-CH2
of 2’-N-acyl main chain), 2.42–2.26 (m, 4H; a-CH2 of 2’-N-acyl side chain
and 2-N-acyl), 2.10–2.04 (m, 1H; H-3’’a), 1.92–1.84 (m, 1H; H-3’’b), 1.78–
1.14 (m, 84H; CH2 of 2-N-acyl and 2’-N-acyl), 0.89 ppm (t, J=7.2 Hz,
9H; -CH2-CH3 of 2-N-acyl and 2’-N-acyl); HRMS (ESI-Q-TOF, nega-
tive): m/z: calcd for C74H139N2O24P [MꢀH]ꢀ: 1469.9377; found: 1469.9371.
[9] R. Ross, N. Engl. J. Med. 1999, 340, 115–126.
2-Deoxy-6-O-[2-deoxy-6-O-(3-deoxy-a-d-manno-oct-2-ulopyranosid)onic
acid]-2-{[(R)-3-(octadecanoyloxy)octadecanoylamino]-b-d-glucopyrano-
syl}-2-[(R)-3-benzyloxyoctadecanoylamino]-a-d-glucopyranose 1-(amino-
ethyl)phosphate (4): Pd(OH)2/C (20 wt% on carbon, 1.0 mg) was added
to a solution of 25 (1.3 mg, 0.495 mmol) in distilled THF (435 mL), water
(43 mL), and AcOH (22 mL). The mixture was stirred under hydrogen
(2.0 MPa) at room temperature for 2 d and then neutralized with Et3N in
THF (10%, v/v, 20.0 mL), and Pd catalyst was removed by filtration.
After removal of the solvent in vacuo, the residue was washed with n-
[10] J. Danesh, R. Collins, R. Peto, Lancet 1997, 350, 430–436.
[11] J. Danesh, Y. Wong, M. Ward, J. Muir, Heart 1999, 81, 245–247.
[12] C. C. Kuo, J. T. Grayston, L. A. Campbell, Y. A. Goo, R. W. Wissler,
E. P. Benditt, Proc. Natl. Acad. Sci. USA 1995, 92, 6911–6914.
[13] J. T. Grayston, J. Infect. Dis. 2000, 181 Suppl 3, S402–410.
[14] A. Schumacher, I. Seljeflot, A. B. Lerkerod, L. Sommervoll, J. E.
Otterstad, H. Arnesen, Atherosclerosis 2002, 164, 153–160.
[15] G. Hajishengallis, A. Sharma, M. W. Russell, R. J. Genco, Ann. Pe-
riodontol. 2002, 7, 72–78.
Chem. Eur. J. 2011, 17, 14464 – 14474
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
14473

Y. Fujimoto, K. Fukase et al.
[16] F. C. Gibson III, H. Yumoto, Y. Takahashi, H. H. Chou, C. A.
Genco, J. Dent. Res. 2006, 85, 106–121.
[40] Y. Zhang, J. Gaekwad, M. A. Wolfert, G. J. Boons, Chem. Eur. J.
2008, 14, 558–569.
[41] S. Inamura, K. Fukase, S. Kusumoto, Tetrahedron Lett. 2001, 42,
7613–7616.
[42] I. Shiina, H. Ushiyama, Y.-k. Yamada, Y.-i. Kawakita, K. Nakata,
Chem. Asian J. 2008, 3, 454–461.
[43] A. Merzouk, F. Guibe, A. Loffet, Tetrahedron Lett. 1992, 33, 477–
480.
[44] D. Baudry, M. Ephritikhine, H. Felkin, J. Chem. Soc. Chem.
Commun. 1978, 694–695.
[17] G. A. Roth, B. Moser, F. Roth-Walter, M. B. Giacona, E. Harja,
P. N. Papapanou, A. M. Schmidt, E. Lalla, Atherosclerosis 2007, 190,
271–281.
[18] F. C. Gibson III, T. Ukai, C. A. Genco, Front. Biosci. 2008, 13,
2041–2059.
[19] M. Triantafilou, F. G. Gamper, P. M. Lepper, M. A. Mouratis, C.
Schumann, E. Harokopakis, R. E. Schifferle, G. Hajishengallis, K.
Triantafilou, Cell Microbiol. 2007, 9, 2030–2039.
[45] M. Inage, H. Chaki, S. Kusumoto, T. Shiba, Chem. Lett. 1982, 1281–
1284.
[46] A. Campbell, B. Fraser-Reid, Bioorg. Med. Chem. 1994, 2, 1209–
1219.
[47] G. D. Prestwich, J. F. Marecek, R. J. Mourey, A. B. Theibert, C. D.
Ferris, S. K. Danoff, S. H. Snyder, J. Am. Chem. Soc. 1991, 113,
1822–1825.
[20] K. Honjo, R. van Reekum, N. P. L. G. Verhoeff, Alzheimer’s Demen-
tia 2009, 5, 348–360.
[21] H. Okamura, H. Tsutsi, T. Komatsu, M. Yutsudo, A. Hakura, T. Ta-
nimoto, K. Torigoe, T. Okura, Y. Nukada, K. Hattori, Nature 1995,
378, 88–91.
[22] M. Trøseid, I. Seljeflot, H. Arnesen, Cardiovasc. Diabetol. 2010, 9,
11.
[48] R. W. Murray, R. Jeyaraman, J. Org. Chem. 1985, 50, 2847–2853.
[49] H. Yoshizaki, N. Fukuda, K. Sato, M. Oikawa, K. Fukase, Y. Suda,
S. Kusumoto, Angew. Chem. 2001, 113, 1523–1528; Angew. Chem.
Int. Ed. 2001, 40, 1475–1480.
[50] B. Yu, H. Tao, Tetrahedron Lett. 2001, 42, 2405–2407.
[51] H. Tokimoto, Y. Fujimoto, K. Fukase, S. Kusumoto, Tetrahedron:
Asymmetry 2005, 16, 441–447.
[52] K. Watanabe, N. Yamagiwa, Y. Torisawa, Org. Process Res. Dev.
2007, 11, 251–258.
[53] T. Goi, K. Tanaka, K. Fukase, J. Carbohydr. Chem. 2007, 26, 369–
394.
[54] Y. Suda, H. Tochio, K. Kawano, H. Takada, T. Yoshida, S. Kotani, S.
Kusumoto, FEMS Immunol. Med. Microbiol. 1995, 12, 97–112.
[55] T. Ogawa, Y. Asai, Y. Sakai, M. Oikawa, K. Fukase, Y. Suda, S. Ku-
sumoto, T. Tamura, FEMS Immunol. Med. Microbiol. 2003, 36, 1–7.
[56] M. Mueller, B. Lindner, S. Kusumoto, K. Fukase, A. B. Schromm, U.
Seydel, J. Biol. Chem. 2004, 279, 26307–26313.
[57] J. Gaekwad, Y. H. Zhang, W. Zhang, J. Reeves, M. A. Wolfert, G. J.
Boons, J. Biol. Chem. 2010, 285, 29375–29386.
[58] H. Tsutsui, T. Yoshimoto, N. Hayashi, H. Mizutani, K. Nakanishi,
Immunol. Rev. 2004, 202, 115–138.
[23] S. Alboni, D. Cervia, S. Sugama, B. Conti, J. Neuroinflammation
2010, 7, 9.
[24] T. Tomita, A. M. Jackson, N. Hida, M. Hayat, M. F. Dixon, T. Shi-
moyama, A. T. Axon, P. A. Robinson, J. E. Crabtree, J. Infect. Dis.
2001, 183, 620–627.
[25] M. T. Fera, M. Carbone, C. Buda, M. Aragona, S. Panetta, M. Gian-
none, F. La Torre, A. Giudice, E. Losi, Ann. N.Y. Acad. Sci. 2002,
963, 326–328.
[26] K. Yamauchi, I. J. Choi, H. Lu, H. Ogiwara, D. Y. Graham, Y. Ya-
maoka, J. Immunol. 2008, 180, 1207–1216.
[27] Y. Suda, Y. Kim, K. Nagata, Y. Hasegawa, A. Kawaguchi, M.
Oikawa, K. Nakajima, K. Aoyama, T. Ogawa, T. Shimoyama, T.
Tamura, S. Kusumoto, Symposium Papers: Symposium on the
Chemistry of Natural Products 2001, 43, 383–388.
[28] S. Kusumoto, K. Fukase, Y. Fujimoto in Microbial Glycobiology:
Structures, Relevance and Applications (Eds.: A. Moran, O. Holst,
P. J. Brennan, M. v. Itzstein), Elsevier, London, 2009, pp. 415–427.
[29] Y. Fukase, Y. Fujimoto, Y. Adachi, Y. Suda, S. Kusumoto, K.
Fukase, Bull. Chem. Soc. Japan 2008, 81, 796–819.
[30] Y. Fujimoto, Y. Adachi, M. Akamatsu, Y. Fukase, M. Kataoka, Y.
Suda, K. Fukase, S. Kusumoto, J. Endotoxin Res. 2005, 11, 341–347.
[31] Y. Fujimoto, M. Iwata, N. Imakita, A. Shimoyama, Y. Suda, S. Kusu-
moto, K. Fukase, Tetrahedron Lett. 2007, 48, 6577–6581.
[32] Y. Suda, T. Ogawa, W. Kashihara, M. Oikawa, T. Shimoyama, T.
Hayashi, T. Tamura, S. Kusumoto, J. Biochem. 1997, 121, 1129–
1133.
[59] K. Nakanishi, T. Yoshimoto, H. Tsutsui, H. Okamura, Annu. Rev.
Immunol. 2001, 19, 423–474.
[60] H. Okamura, S. Kashiwamura, H. Tsutsui, T. Yoshimoto, K. Naka-
nishi, Curr. Opin. Immunol. 1998, 10, 259–264.
[61] M. Imamura, H. Tsutsui, K. Yasuda, R. Uchiyama, S. Yumikura-Fu-
tatsugi, K. Mitani, S. Hayashi, S. Akira, S. Taniguchi, N. Van Rooi-
jen, J. Tschopp, T. Yamamoto, J. Fujimoto, K. Nakanishi, J. Hepatol.
2009, 51, 333–341.
[33] A. P. Moran, B. Lindner, E. J. Walsh, J. Bacteriol. 1997, 179, 6453–
6463.
[34] Y. Sakai, M. Oikawa, H. Yoshizaki, T. Ogawa, Y. Suda, K. Fukase,
S. Kusumoto, Tetrahedron Lett. 2000, 41, 6843–6847.
[35] N. Sawada, T. Ogawa, Y. Asai, Y. Makimura, A. Sugiyama, Clin.
Exp. Immunol. 2007, 148, 529–536.
[36] H. Kumada, Y. Haishima, K. Watanabe, C. Hasegawa, T. Tsuchiya,
K. Tanamoto, T. Umemoto, Oral Microbiol. Immunol. 2008, 23, 60–
69.
[37] Y. Zhang, J. Gaekwad, M. A. Wolfert, G. J. Boons, Org. Biomol.
Chem. 2008, 6, 3371–3381.
[38] K. Fukase, Y. Fukase, M. Oikawa, W. C. Liu, Y. Suda, S. Kusumoto,
Tetrahedron 1998, 54, 4033–4050.
[62] T. D. Kanneganti, M. Lamkanfi, Y. G. Kim, G. Chen, J. H. Park, L.
Franchi, P. Vandenabeele, G. NuÇez, Immunity 2007, 26, 433–443.
[63] T. Kawai, S. Akira, Int. Immunol. 2009, 21, 317–337.
[64] M. Krummen, S. Balkow, L. Shen, S. Heinz, C. Loquai, H. C. Probst,
S. Grabbe, J. Leukocyte Biol. 2010, 88, 189–199.
[65] Y. Chen, M. J. Blaser, J. Infect. Dis. 2008, 198, 553–560.
[66] M. Vasan, M. A. Wolfert, G. J. Boons, Org. Biomol. Chem. 2007, 5,
2087–2097.
[67] Y. H. Zhang, M. A. Wolfert, G. J. Boons, Bioorg. Med. Chem. 2007,
15, 4800–4812.
Received: December 11, 2010
[39] Y. H. Zhang, J. Gaekwad, M. A. Wolfert, G. J. Boons, J. Am. Chem.
Soc. 2007, 129, 5200–5216.
Revised: July 27, 2011
Published online: November 16, 2011
14474
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 14464 – 14474