Agonistic and antagonistic properties of a Rhizobium sin-1 lipid A modified by an ether-linked lipid

Article abstract of DOI:10.1039/b704427e

LPS from Rhizobium sin-1 (R. sin-1) can antagonize the production of tumor necrosis factor alpha (TNF-α) by E. coli LPS in human monocytic cells. Therefore these compounds provide interesting leads for the development of therapeutics for the prevention or

Full text of DOI:10.1039/b704427e

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Agonistic and antagonistic properties of a Rhizobium sin-1 lipid A modified by
an ether-linked lipid†
Mahalakshmi Vasan, Margreet A. Wolfert* and Geert-Jan Boons*
Received 23rd March 2007, Accepted 17th May 2007
First published as an Advance Article on the web 29th May 2007
DOI: 10.1039/b704427e
LPS from Rhizobium sin-1 (R. sin-1) can antagonize the production of tumor necrosis factor alpha
(
TNF-a) by E. coli LPS in human monocytic cells. Therefore these compounds provide interesting leads
for the development of therapeutics for the prevention or treatment of septic shock. Detailed structure
activity relationship studies have, however, been hampered by the propensity of these compounds to
undergo b-elimination to give biological inactive enone derivatives. To address this problem, we have
chemically synthesized in a convergent manner a R. sin-1 lipid A derivative in which the b-hydroxy ester
at C-3 of the proximal sugar unit has been replaced by an ether linked moiety. As expected, this
derivative exhibited a much-improved chemical stability. Furthermore, its ability to antagonize TNF-a
production induced by enteric LPS was only slightly smaller than that of the parent ester modified
derivative demonstrating that the ether-linked lipids affect biological activities only marginally.
Furthermore, it has been shown for the first time that R. sin-1 LPS and the ether modified lipid A are
also able to antagonize the production of the cytokine interferon-inducible protein 10, which arises
from the TRIF-dependent pathway. The latter pathway was somewhat more potently inhibited than the
MyD88-dependent pathway. Furthermore, it was observed that the natural LPS possesses much greater
activity than the synthetic and isolated lipid As, which indicates that di-KDO moiety is important for
optimal biological activity. It has also been found that isolated R. sin-1 LPS and lipid A agonize a
mouse macrophage cell line to induce the production of TNF-a and interferon beta in a Toll-like
receptor 4-dependent manner demonstrating species specific properties.
Introduction
to initiate specific down-stream processes, such as activation
of mitogen-activated protein (MAP) kinases and transcription
factors (NF-jB, AP-1 and interferon regulatory factor 3 (IRF-3)).
The end result is the up-regulation of hundreds of genes resulting
in the production of a multitude of cytokines and chemokines.
TLR4 initiates cell-signaling by two cascades that involve
recruitment of the intracellular TIR adaptor proteins MyD88 or
The innate immune system is an evolutionary ancient system
of defense against microbial infections. It responds rapidly to
1
highly conserved families of structural patterns, called pathogen
associated molecular patterns (PAMPs), which are integral parts
of pathogens and are perceived as danger signals by the host.
Recognition of PAMPs is mediated by sets of highly conserved
3,4
TRIF. Thus, the TIR domain of TLR4 can bind to the dimeric
2
receptors, each of which binds to a variety of PAMPs. Cellular
adapter protein MyD88, that then recruits and activates a number
of kinases, subsequently leading to activation of the MAP kinases,
such as p38, JNK and ERK1/2 and the transcription factor NF-
jB. This MyD88-dependent pathway results in the synthesis of pro-
inflammatory cytokines and chemokines including tumor necrosis
factor alpha (TNF-a), interleukin 1beta (IL-1b) and IL-6. Another
activation by these receptors results in acute inflammatory re-
sponses that include the production of a diverse set of cytokines
and chemokines, direct local attack against the invading pathogen
and initiation of responses that activate and regulate the adaptive
component of the immune response.
Lipopolysaccharides (LPSs) are structural components of the
outer surface membrane of Gram-negative bacteria that trigger
innate immune responses through Toll-like receptor 4 (TLR4),
a member of the TLR family that participates in pathogen
recognition. TLRs are transmembrane glycoproteins having an
extracellular domain that contains multiple leucine-rich repeating
motifs, a transmembrane domain and an intracellular signaling
6
adaptor protein, called TRIF, can also be recruited to the TIR
domain leading to activation of the transcription factor IRF-3,
NF-jB and the MAP kinase JNK. This TRIF-dependent pathway
results in the synthesis of important inflammatory mediators,
including interferon beta (IFN-b), interferon-inducible protein 10
(
IP-10) and nitric oxide.
Although the initiation of acute inflammatory responses is
3,4
domain. The intracellular domain serves as a docking site for
important for the prevention of infections, over-activation of this
response may lead to the clinical symptoms of septic shock. Sep-
ticemia is a serious world-wide health problem and is associated
5
a number of adaptor proteins, which in turn recruit kinases
Complex Carbohydrate Research Center, The University of Georgia, 315
Riverbend Road, Athens, GA, 30602, USA. E-mail: gjboons@ccrc.uga.edu,
mwolfert@ccrc.uga.edu; Fax: +1 706-542-4412
7,8
with mortality rates of 40–60%. It has been estimated that 1%
of hospital patients and 20–30% of ICU patients develop sepsis
and that septic shock results in 100000 deaths annually in the
†
Electronic supplementary information (ESI) available: Fig. S1 and S2 and
7,8
NMR spectra of synthesized compounds. See DOI: 10.1039/b704427e
United States. A number of strategies for the prevention and
This journal is © The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 2087–2097 | 2087

View Article Online
treatment of sepsis have been directed against the lipid A region
of LPS. For example, structural analogs of lipid A have been
chemical synthesis and biological evaluation of a compound
locked in the 2-aminogluconate established that this species
9
23
examined for their ability to antagonize the effects of LPS. These
possesses antagonistic properties. Detailedbiologicalevaluations
of the synthetic R. sin-1 lipid As have been hampered by their
chemical lability due to elimination to give the enone derivative 2.
To address this problem, we report here the chemical synthesis of
compound 3, which is derived from 1, however, the b-hydroxy ester
at C-3 of the proximal sugar unit has been replaced by an ether
derivative. It was anticipated that this compound would be less
prone to b-elimination due to the poor leaving group ability of the
10
antagonists include naturally occurring lipid A precursors, as
11,12
well as a number of synthetic analogs of these precursors.
The
most widely studied analog is a synthetic analog based on the
lipid A of Rhodobacter sphaeroides or R. capsulatus, two species
1
3–16
having very similar lipid A structures.
Although the lipid As
ꢀ
of R. sphaeroides/R. capsulatus and E. coli have the same bis-1,4 -
phosphorylated glucosamine disaccharide backbone their fatty
acyl patterns differ considerably. In this respect, R. sphaeroides/R.
capsulatus lipid A consists of two 3-oxomyristic acid, two b-
hydroxydecanoic acid and one dodecenoic acid residues.
24–26
ether.
Asaresult, wehavebeenabletoinvestigatetheabilityofa
R-sin 1 lipid A to antagonize cell-signaling events arising from the
MyD88- and TRIF-dependent pathways. Furthermore, species-
specific properties of 3 have been investigated by comparing
biological properties of the compound exposed to human and
mouse macrophages.
Recently, we reported that LPS from Rhizobium sin-1 (R. sin-1),
a nitrogen-fixing bacterial species, can prevent the induction of
17,18
TNF-aby E. coli LPS in human monocytic cells.
Furthermore,
another study showed that the biological properties of R. sin-1
LPS are species specific and most notably it was found that it can
agonize mouse macrophages in a TLR2-dependent manner.
19,20
Results and discussion
The lipid A of R. sin-1 LPS is a structurally unusual lipid A
Chemical synthesis
differing in almost every aspect from those known to contribute to
21
the toxicity of enteric LPS (Fig. 1). In particular, the disaccharide
moiety of Rhizobial lipid A is devoid of phosphate and the
glucosamine phosphate is replaced by 2-aminogluconolactone.
It contains a very long chain fatty acid, 27-hydroxyoctacosanoic
acid which, in turn, can be esterified by b-hydroxybutyrate. As
is the case with other naturally occurring LPSs, the inherent
microheterogeneity of the lipid A region of R. sin-1 makes it
impossible to be developed as a therapeutic agent for Gram-
negative septicemia. Furthermore, the inability to separate the
different species limits identification of specific structural features
that makes R. sin-1 lipid A an antagonist as opposed to an agonist.
To address these problems, we have synthesized a range of well-
defined lipid A derivatives based on the structure of R. sin-1
LPS for structure–activity relationship studies. We have already
It was envisaged that coupling of glycosyl donor 4 with glycosyl
acceptor 5 would give disaccharide 12 (Scheme 1), which is
appropriately protected for the selective introduction ofb-hydroxyl
fatty acids and oxidation of the C-1 position to lactone. Glycosyl
acceptor 5 is modified by an ether linked c-benzyloxy fatty acid,
because it was anticipated that the harsh conditions required
for its introduction would affect functionalities present in the
disaccharide. Another feature of 5 is that its anomeric center is
27,28
protected as a thioglycoside.
This functionality is stable under a
wide range of chemical conditions, however, it can be hydrolyzed at
a late stage of the chemical synthesis to give a lactol, which can then
be oxidized to a lactone. Furthermore, the selenoglycoside of 4 was
expected to be significantly more reactive towards activation with
NIS–TMSOTf than the thioglycoside of 5, and thus it was expected
that these compounds could be employed in a chemoselective
glycosylation to give 12. The phthalimido and azido functions
of 12 offer an attractive set of orthogonal protecting groups that
18
shown that the synthetic compound 1 is able to antagonize
E. coli LPS. In addition, we have demonstrated that the 2-
22
aminogluconolactones can exist as a 2-aminogluconate. The
Fig. 1 Structures of E. coli and R. sin-1 lipid A and synthetic R. sin-1 lipid A derivatives 1–3.
088 | Org. Biomol. Chem., 2007, 5, 2087–2097
2
This journal is © The Royal Society of Chemistry 2007

View Article Online
the b-hydroxyl fatty acids and oxidation of the anomeric center.
Thus, a NIS–TMSOTf mediated coupling of the glycosyl donor 4
◦
with acceptor 5 in dichloromethane at −35 C gave disaccharide
31,34,35
12 in a yield of 76%.
Only the b-anomer was formed due
to neighboring group participation of the phthalimido group.
Next, the phthalimido moiety and acetyl ester of compound 12
were removed by treatment with hydrazine hydrate in refluxing
36
ethanol and the amine of the resulting compound was selectively
acylated with alkanoyloxy fatty acid 7 in the presence DCC
to give 13. Reduction of the azido moiety of 13 was easily
accomplished by reaction with activated Zn in acetic acid and the
amine and hydroxyl of the resulting compound were immediately
acylated with 8 using 1,3-dicyclohexylcarbodiimide (DCC) and
4
-dimethylaminopyridine (DMAP) as the activation reagents to
afford 14. It is important to note that selective N-acylation could
be accomplished by performing the reaction in the absence of
DMAP, thereby making it possible to synthesize derivatives having
ꢀ
different substituents at C-2 amine and C-3 and C-3 hydroxyls.
The thioglycoside of compound 14 was hydrolyzed by treatment
37
with NIS–TMSOTf in wet dichloromethane and the resulting
lactol was oxidized with PCC in DCM to afford lactone 16.
Finally, the benzyl ethers and benzylidene acetal of 16 were
removed by catalytic hydrogenation over Pd/C to give the target
compound 3. As expected, this derivative had an excellent shelf-life
◦
and after storage for three months at −20 C no decomposition
was observed. Under similar storage conditions compound 1 was
decomposed.
It is important to note that alternative synthetic strategies,
which employed either an allyl or TBDMS ether for protection
of the anomeric center of the proximal sugar, led to failure.
Thus, the anomeric TBDMS function was not compatible with
the alkylation conditions required for the instalment of 6. Fur-
thermore, attempts to cleave an allyl glycoside at the final stage
of the synthesis led either to recovery of starting material or
decomposition. The use of a thioglycoside gave the best results
for the preparation of the target compound.
Scheme 1 Reagents and conditions: a) NaOMe, MeOH then C
6
H
5
CH-
·Et O,
NH
◦
(
OMe)
2
, CSA, CH
3
CN; b) 6, NaH, DMF, 0 C; c) BH
3
·NHMe
2
, BF
3
2
◦
◦
toluene, −30 C; d) NIS, TMSOTf, MS 4 A˚ , DCM, −35 C; e) NH
·
2
2
H O, EtOH, D, then 7, DCC, DCM; f) Zn, AcOH then 8, DCC, DMAP,
2
◦
DCM; g) NIS, TMSOTf, DCM, H
Pd/C, H , t-BuOH, THF.
2
O, 0 C; h) PCC, 3 A˚ MS, DCM; i)
2
allow selective derivatization of the two amino groups. Removal
of the phthalimido group will result in cleavage of the O-acetyl
ester. However, by exploiting the higher nucleophilicity of primary
amines compared to hydroxyls it is possible to selectively acylate
the amine.
Biological evaluation
1–5
Based on the results of recent studies, it is clear that enteric LPS-
induced cellular activation through TLR4 is complex as many
signaling elements are involved. However, it appears that there are
two distinct initiation points in the signaling process, one being a
specific intracellular adaptor protein called MyD88 and the other
an adaptor protein called TRIF, which operates independently
of MyD88. It is well established that TNF-a secretion is a
prototypical measure for activation of the MyD88-dependent
pathway, whereas secretion of IFN-b and IP-10 are commonly
used as an indicator of TRIF-dependent cellular activation.
Compound 3 was tested over a wide concentration range for
its ability to activate a human monocytic cell line (Mono Mac
6; MM6) to produce TNF-a and IP-10 protein and the resulting
values were compared with similar data obtained for E. coli LPS
and lipid A and R. sin-1 LPS and lipid A. Thus, MM6 cells
were exposed to the isolated and synthetic compounds and after
5.5 hours, the supernatants were harvested and examined for
Glycosyl acceptor 5 was readily obtained from known derivative
29
9.
Thus, the acetyl esters of 9 were cleaved by treatment with
sodium methoxide in methanol and the resulting triol was selec-
tively protected by reaction with benzylaldehyde dimethyl acetal
in the presence of camphorsulfonic acid (CSA) in acetonitrile to
give 10. The C-3 hydroxyl of 10 was alkylated with sulfonate 6 by
treatment with sodium hydride in DMF to give 11 in a good yield
26
of 79%. Next, the benzylidene acetal of 11 was regioselectively
opened by reaction with BH
3
·NHMe
2
and BF
3
·Et
2
O in toluene at
◦
−
30 C to give 5 in an excellent yield. A number of other reaction
conditions led to the formation of mixtures of regioisomeric benzyl
ethers. For example, the use of BH in DCM, which is the
·NHMe
conventional solvent for this reagent, gave a mixture of products.
3
2
3
0
31
26
32,33
Glycosyl donor 4 and fatty acids 6, 7 and 8
were prepared
by reported procedures.
human TNF-a and IP-10 using capture ELISAs. Potencies (EC50,
Having glycosyl donor 4 and acceptor 5 at hand, attention was
focused on the preparation of the disaccharide 12, instalment of
concentration producing 50% activity) and efficacies (maximal
level of production) were determined by fitting the dose–response
This journal is © The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 2087–2097 | 2089

View Article Online
Fig. 2 Concentration–response curves of E. coli LPS, E. coli lipid A, R. sin-1 LPS, R. sin-1 lipid A and synthetic compounds 1 and 3 in human monocytic
◦
cells. MM6 cells were incubated for 5.5 h at 37 C with increasing concentrations of E. coli LPS, E. coli lipid A, R. sin-1 LPS, R. sin-1 lipid A and synthetic
compounds 1 and 3 as indicated. TNF-a (a) and IP-10 (b) proteins in cell supernatants were measured using ELISAs. (Please note that R. sin-1 LPS,
R. sin-1 lipid A, 1 and 3 show background values and therefore overlap in the figure.) Treatment with E. coli LPS, E. coli lipid A, R. sin-1 LPS, R. sin-1
lipid A, 1 and 3 did not affect cell viability, as judged by cellular exclusion of trypan blue.
curves to a four parameter logistic equation using PRISM
software. As can be seen in Fig. 2, E. coli LPS and lipid A
yielded clear dose–response curves for TNF-a and IP-10, whereas
R. sin-1 LPS, R. sin-1 lipid A and synthetic compound 3 did not
induce significant production of the cytokines. The EC50 values
for E. coli 055:B5 LPS were significantly smaller than that of
E. coli lipid A (Table 1) which is probably due to the di-KDO
LPS or E. coli lipid A was employed as an activator. Thus,
it appears that the MyD88-dependent pathway is slightly more
responsive than TRIF-mediated cellular activation.
Based on its lack of proinflammatory effects, compound 3 was
tested over a wide concentration range for its ability to antagonize
TNF-a and IP-10 production by MM6 cells incubated with E. coli
−
1
LPS (1 ng mL ) (Fig. 3). An IC50 (concentration producing 50%
ꢀ
−1
moiety of LPS, which is attached to the C-6 position of lipid
inhibition) of 22 lM (38 lg mL ) was established for TNF-a
38
A. In this respect, recent studies have shown that meningococcal
lipid A expressed by a strain defect in KDO biosynthesis has
significantly reduced bioactivity compared to KDO containing
Meningococcal lipooligosaccharides. It has also been shown that
removal of the KDO moieties by mild acidic treatment reduces
inhibition by compound 3. Similar inhibition experiments with
R. sin-1 lipid A and compound 1 gave IC50 values of 2.0 lM (3.0 lg
−
1
−1
mL ) and 7.3 lM (13 lg mL ), respectively. As expected, R. sin-
1 LPS was a much more potent inhibitor of TNF-a production
than the corresponding lipid A and in this case an IC50 value of
18
−1
cellular responses.
6.5 nM (239 ng mL ) was determined. Thus, it is probable that the
It was observed that the EC50 values for TNF-a secretion were
approximately three times smaller than that of IP-10 when E. coli
KDO moiety of LPS accounts for the higher inhibitory activity.
Interestingly, for IP-10 secretion the IC50 values of R. sin-1 LPS and
compound 3 were smaller than the corresponding values for TNF-
a (1.4 nM; 51 ng mL ) and 3.7 lM (6.5 lg mL ), respectively)
and at the highest concentration tested compound 3 abolished
all IP-10 biosynthesis induced by enteric LPS. Similar inhibition
results were obtained when E. coli lipid A was employed as the
agonist (Supplementary data†).
The results of the cellular activation studies show that the
inhibitory activity of compound 3 is only slightly smaller than
that of the parent compound 1 demonstrating that the ether linked
fatty acid at C-3 of the proximal monosaccharide unit has only
marginal effect on the biological activity. However, compound 3
−
1
−1
a
Table 1 EC50 values (nM) of E. coli LPS and lipid A in MM6 cells
E. coli LPS
0.0016
E. coli lipid A
TNF-a
9.1
(7.3–11.4)
22.2
(0.0014–0.0019)
0.0042
IP-10
(
0.0032–0.055)
(18.5–26.8)
a
EC50 values are reported as best-fit values and as minimum–maximum
range (best-fit value ± std. error).
Fig. 3 Antagonism of E. coli LPS by R. sin-1 LPS, R. sin-1 lipid A and synthetic compounds 1 and 3 in human monocytic cells. TNF-a (a) and IP-10
b) concentrations were measured after preincubation of MM6 cells with increasing concentrations of R. sin-1 LPS, R. sin-1 lipid A, 1, or 3 as indicated
(
◦
−1
for 1 h at 37 C, followed by 5.5 h of incubation with E. coli LPS (1 ng mL ). Results are expressed as percentage of cytokine concentration of control
cells, which are incubated only with E. coli LPS.
2
090 | Org. Biomol. Chem., 2007, 5, 2087–2097
This journal is © The Royal Society of Chemistry 2007

View Article Online
has a much greater chemical stability than 1 making it a preferred
compound for biological studies. Furthermore, the KDO moiety
of E. coli and R. sin-1 LPS appears to significantly increase the
agonistic and antagonistic properties, respectively. The data also
reveal that the agonists and antagonists affect the MyD88 and
TRIF pathways slightly differently. Thus, the agonists (E. coli
LPS and lipid A) display somewhat higher potencies for TNF-a
compared to IP-10, whereas for the antagonists (R. sin-1 LPS and
lipid A), IP-10 was more potently inhibited at lower concentrations
compared to TNF-a.
was much larger between E. coli LPS and lipid A (10000-fold) than
between R. sin-1 LPS and lipid A (100-fold).
Recent reports indicate that LPS of non-enterobacterial species
such as Porphyromonas gingivalis, Leptospira interrogans and
R. sin-1 are capable of signaling independent of TLR4, instead
19,20,43,44
utilizing TLR2-mediated signal transduction.
However,
TLR2 can only recruit the adaptor protein MyD88 and as a result
can only initiate the production of MyD88-dependent cytokines
such as TNF-a, but not those TRIF-dependent cytokines such
3, 4
as IFN-b. The fact that our results show that R. sin-1 can
induce the production of IFN-b prompted us to investigate the
TLR utilization of these compounds. Thus, R. sin-1 LPS and
lipid A and E. coli LPS and lipid A were exposed at a range
of concentrations to HEK 293T cells stably transfected with
mouse TLR2/TLR6 or TLR4/MD2 and transiently transfected
with a plasmid containing the reporter gene pELAM-Luc (NF-
jB-dependent firefly luciferase reporter vector) and a plasmid
containing the control gene pRL-TK (Renilla luciferase control
reporter vector). As a negative control, wild type HEK 293T
cells transiently transfected with plasmids containing the reporter
gene pELAM-Luc and control gene pRL-TK were used. After
an incubation time of 4 h, the activity was measured using a
commercial dual-luciferase assay. E. coli LPS and the lipopeptide
There are several reports that indicate that structurally different
39–42
lipid As may differentially induce proinflammatory responses.
However the heterogeneity in the structure of lipid A within a
particular bacterial strain and possible contamination with other
inflammatory components of the bacterial cell-wall complicated
the dissecting of the biological responses to specific lipid As.
Synthetic compounds may address this important issue.
Next, attention was focused on cellular activation studies
using a mouse macrophage cell line (RAW 264.7 cNO(−) cells).
Thus, secretion of TNF-a and IFN-b protein was measured after
exposure of the cells for 5 h to a wide concentration range E. coli
LPS and lipid A, R. sin-1 LPS and lipid A and compound 3 (Fig. 4).
Interestingly, E. coli and R. sin-1 LPS and lipid As activated the
cells to produce TNF-a and IFN-b. No cytokine production was
measured for compound 3 even when a very high concentration
45
Pam
3
CysSK
4
were employed as positive controls for cellular
activation by TLR4 and TLR2/6, respectively. As can be seen
in Fig. 5, R. sin-1 LPS and lipid A can induce cellular activation
in a TLR4-dependent manner, whereas no activity was observed
in cells transfected with TLR2/6. The TLR4-dependent NF-jB
activation showed clear dose responses for E. coli LPS, R. sin-1
LPS and R. sin-1 lipid A (Supplementary data†).
−
1
was employed (57 lM; 100 lg mL ). Furthermore, compound 3
was not able to antagonize the production of TNF-a or IFN-b
induced by E. coli LPS.
For each agonist, the potency for TNF-asecretion was higher by
5
- to 7-fold compared to that of IFN-b. (Table 2). Furthermore, for
the E. coli derived compounds the EC50 values were significantly
smaller than those derived from R. sin-1. As expected, the lipid As
were less potent than their parent LPSs, however, the difference
Previously it was established that the lipid A region of
several Rhizobiaceae can stimulate bone marrow granulocytes
of TLR4-deficient mice to induce the expression of CD14.
19
Fig. 4 TNF-a and IFN-b production by murine macrophages after stimulation with E. coli LPS, E. coli lipid A, R. sin-1 LPS, R. sin-1 lipid A and
synthetic compound 3. Murine RAW cNO(−) cells were incubated for 5.5 h with increasing concentrations of E. coli LPS, E. coli lipid A, R. sin-1 LPS,
R. sin-1 lipid A and 3 as indicated. TNF-a (a) and IFN-b (b) in cell supernatants were measured using ELISAs. Treatment with E. coli LPS, E. coli lipid
A, R. sin-1 LPS, R. sin-1 lipid A and 3 did not affect cell viability, as judged by cellular exclusion of trypan blue.
a
Table 2 EC50 values (nM) of E. coli and R. sin-1 LPS and lipid A in RAW cells.
E. coli LPS
E. coli lipid A
R. sin-1 LPS
R. sin-1 lipid A
TNF-a
IFN-b
0.0028
21
(16–28)
124
2.5
(2.0–3.2)
19.3
171
(109–268)
932
(
0.0020–0.0041)
0.0118
0.0096–0.0145)
(
(105–147)
(11.3–32.8)
(816–1067)
a
EC50 values are reported as best-fit values and as minimum–maximum range (best-fit value ± std. error).
This journal is © The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 2087–2097 | 2091

View Article Online
differently activated or inhibited by the examined compounds.
Finally, the LPS agonist and antagonist were much more potent
indicating that the KDO moiety of LPS is important for optimal
biological properties.
Experimental
Chemical synthesis
General synthetic methods. Chemicals were purchased from
commercial suppliers and used without further purification, unless
otherwise noted. Dichloromethane (DCM) and toluene were
distilled from calcium hydride under Argon. Tetrahydrofuran
Fig. 5 Response of HEK 293T cells expressing murine TLRs to
R. sin-1 LPS and R. sin-1 lipid A. Induction of NF-jB activation was
determined in triplicate cultures of HEK 293T cells stably transfected
with murine TLR4/MD2 or TLR2/TLR6 and transiently transfected with
pELAM-Luc, pRL-TK and pcDNA3 plasmids. Forty-four h post-trans-
(
THF) was distilled under argon from sodium directly prior
to application. Dry N,N-dimethylamineformamide (DMF) was
˚
used without purification. Powdered molecular sieves (4 A) were
−
1
◦
fection, cells were treated with E. coli LPS (1 ng mL ), R. sin-1 LPS
activated in vacuo at 390 C for 8 h and cooled to room temperature
−
1
−1
−1
(
1 lg mL ), R. sin-1 lipid A (1 lg mL ) and Pam
3
CysSK
4
(1 lg mL
)
in vacuo prior to application. Column chromatography was per-
formed on silica gel 60 (EM Science, 70–230 mesh), size exclusion
was performed on Sephadex LH-20 and eluted with a mixture of
or were left untreated (control). Forty-eight h post-transfection, NF-jB
activation was determined by firefly luciferase activity relative to Renilla
luciferase activity. In the transfection experiments shown, human TNF-a
MeOH–CH
2
2
Cl , (1 : 1, v/v). Reactions were monitored by thin-
−
1
(
10 ng mL ) induced 24.5 ± 0.6 and 21.8 ± 0.3-fold activation of NF-jB
layer chromatography (TLC) on kieselgel 60 F254 (EM Science) and
compounds were visualized by examination under UV light and
by charring with cerium sulfate–ammonium molybdate solution.
in HEK 293T cells stably transfected with TLR4/MD2 and TLR2/6,
respectively.
◦
Organic solvents were removed under reduced pressure at <40 C.
Furthermore, no detectable levels of TNF-a were measured after
1
13
−
1
H NMR and C NMR spectra were recorded on a Merc 300,
Varian Inova 500 or Inova 600 equipped with Sun Workstations.
mouse peritoneal macrophages were exposed to 100 ng mL
R. sin-1 LPS. Surprisingly, HEK cells transfected with TLR2/6
with an ELAM luciferease reporter plasmid showed activity
at this concentration. Our results show clearly that at similar
concentrations, R. sin-1 LPS and lipid A can induce the production
of TNF-a and IFN-b in a TLR4-dependent manner.
The observation that synthetic compound 3 possessed no
activity in the mouse cell line was surprising. LPS and lipid
A isolated from R. sin-1 are composed of a complex mixture
of compounds differing in fatty acid substitution. Probably, a
compound with unique fatty acid composition is responsible for
the TLR4 agonistic properties. A larger range of derivatives will
need to be synthesized to establish which derivatives account
for this activity. The synthetic approach reported here provides
such an opportunity. Furthermore, the observation that TLR
ligands exhibit species-specific properties should be considered
when immuno modulators are being developed.
1
H NMR were recorded in CDCl
3
and referenced to residual
at 7.24 ppm. C NMR spectra were referenced to the
central peak of CDCl at 77.0 ppm. Assignments were made by
1
3
CHCl
3
3
standard gCOSY and gHSQC. High resolution mass spectra were
obtained on a Bruker model Ultraflex MALDI-TOF-TOF mass
spectrometer. Optical rotations were measured on a Jasco model
P-1020 polarimeter. Signals marked with a subscript L belong to
the ether-linked lipid at C-3, whereas signals marked with subscript
ꢀ
ꢀ
LL belong to the lipid at C-2 . Signals marked with subscript LL
refer to the C-28 side chain. Signals marked with subscript LA
ꢀ
belong to lipids at C-2, C-3 .
Phenyl 3-O-[(R)-3-benzyloxy-hexadecanoyl]-4,6-O-benzylidene-
2
-deoxy-2-azido-1-thio-b-D-glucopyranoside (11). NaH (55% oil
dispersion, 0.12 g, 5.0 mmol) was added to a solution of compound
◦
1
0 (0.34 g, 0.88 mmol) in DMF (6 mL) at 0 C followed by addition
of 6 (0.34 g, 0.79 mmol) dissolved in DMF (3 mL). The reaction
mixture was allowed to reach room temperature and stirring was
continued for 14 h. The reaction mixture was diluted with ethyl
acetate (20 mL), quenched with water (1 mL) and subsequently
Conclusion
It has been shown that a derivative of R. sin-1 lipid A in which
the C-3 fatty acid is replaced by an ether-linked moiety has a
much improved chemical stability. Furthermore, this compound
could antagonize cytokine production by a human monocytic cell
line induced by enteric LPS with a similar potency to the natural
ester-linked counter part. For the first time, it has been shown
that such an antagonist can inhibit both MyD88- and TRIF-
dependent cell signaling events. R. sin-1 LPS and lipid A agonized
mouse macrophages to produce TNF-aand IFN-b demonstrating
species specific properties. For the agonists examined, the potency
for TNF-a secretion was higher by 3–7 fold compared to that
of IFN-b or IP-10. For, the antagonists, the IC50 values for IP-
washed with saturated aqueous NaHCO
3
(2 × 10 mL) and brine
(2 × 10 mL). The combined organic layers were dried (MgSO
4
),
filtered and the filtrate was concentrated in vacuo. The residue was
purified by silica gel column chromatography (10% ethyl acetate in
hexane) to afford 9 as a colorless syrup (0.45 g, 79%, yield based
2
D
5
on mesylate): R
−55.94 (c = 1.0, CHCl
f
= 0.70 (20% ethyl acetate in hexane); [a]
=
1
3
). H NMR (300 MHz, CDCl
3
): d = 7.27–
7.59 (m, 15H, aromatic), 5.52 (s, 1H, >CHPh), 4.45–4.50 (m,
3H, H-1, H-3 CH
Ph), 4.38–4.03 (dd, 1H, J5,6a = 5.1 Hz, J6a,6b
10.2 Hz, H-6a), 4.02–3.98 (m, 1H, H-1 ), 3.84–3.73 (m, 2H, H-
, H-6b), 3.58–3.39 (m, 4H, H-3 , H-4, H-3, H-5), 3.30 (t, 1H,
J = 10.2 Hz, J = 8.7 Hz, H-2), 1.82–1.80 (m, 2H, H-2 ), 1.50–
1.46 (m, 2H, H-4 ), 1.31–1.24 [m, 22H, H-(5 –15 )], 0.89 (t, 3H,
L
2
=
L
1
L
L
1
0 were smaller than the corresponding values for TNF-a. These
L
data indicate that the MyD88 and TRIF pathways are somewhat
L
L
L
2
092 | Org. Biomol. Chem., 2007, 5, 2087–2097
This journal is © The Royal Society of Chemistry 2007

View Article Online
1
3
H-16
matic), 101.46 (>CHPh), 86.91 (C-1), 82.07 (C-4), 81.42 (C-3),
6.19 (C-3 ), 71.16 (C-3 CH Ph), 70.73 (C-5), 70.47 (C-1 ), 68.70
C-6), 65.08 (C-2), 35.02 (C-2 ), 34.24 (C-4 ), 32.1–22.93 [C-(5
)], 14.37 (C-16 ); HRMS (m/z) for C42
calcd 738.4019, found 738.4613.
L
); C NMR (300 MHz, CDCl
3
): d = 126.17–139.26 (aro-
(m, 1H, H-2), 1.91 (s, 3H, COCH
3
), 1.79–1.76 (m, 2H, H-2
L
),
1.51–1.42 (m, 2H, H-4 ), 1.32–1.24 [m, 22H, H-(4
L
L
–15 )], 0.86 (t,
L
1
3
7
(
L
L
2
L
3H, H-16
L
). C NMR (300 MHz, CDCl
3
): d = 170.41 (C=O),
ꢀ
L
L
L
+
–
123.80–139.11 (aromatic), 101.92 (CHPh), 98.79 (C-1 ), 85.79 (C-
ꢀ
15
L
L
H
57
N
3
O
5
S[M + Na] :
1), 85.39 (C-3), 79.58 (C-4 ), 78.44 (C-4, C-5), 76.39 (C-3
L
), 75.04
), 70.09 (C-3 ),
ꢀ
ꢀ
(C-3
L
CH
2
Ph, C-4 CH
2
Ph), 71.10 (C-6 ), 71.05 (C-1
L
ꢀ
ꢀ
6
3
1
8.94 (C-6), 68.56 (C-5 ), 66.56 (C-2), 55.49 (C-2 ), 35.06 (C-2
4.32 (C-4 ), 32.17–22.94 (C-5 –15 ), 20.83 (COCH ), 14.38 (C-
). HRMS (m/z) for C65
L
),
Phenyl 3-O-[(R)-3-benzyloxy-hexadecanoyl]-4-O-benzyl-2-deoxy-
-azido-1-thio-b-D-glucopyranoside (5). To a solution of com-
pound 11 (0.26 g, 0.36 mmol) in toluene (10 mL) was added
L
L
L
3
2
+
6
L
H
78
N
4
O12S[M + Na] : calcd 1161.5337,
found 1161.741.
BH
3
·NHMe
2
(0.11 g, 1.79 mmol). After cooling the reaction
◦
mixture (−30 C), BF
3
·OEt
2
(0.31 g, 2.15 mmol) was added
ꢀ
Phenyl 2-azido-4-O-benzyl-6-O-(4,6-O-benzylidene-2-deoxy-2 -
(R)-3-octacosanoyloxy-hexadecan]amido-b-D-glucopyranosyl)-3-
◦
dropwise. The temperature was allowed to reach 0 C over a
period of 1 h after which TLC analysis indicated completion of
the reaction. The reaction mixture was then quenched by the very
slow addition of methanol (3 mL) followed by evaporation of
the organic solvents in vacuo. The crude product was purified by
silica gel column chromatography (20% ethyl acetate in hexane) to
[
O-[(R)-3-benzyloxy-hexadecanoyl]-2-deoxy-1-thio-b-D-glucopyrano-
side (13). To compound 12 (0.73 g, 0.64 mmol) in ethanol
(
20 mL) was added hydrazine hydrate (1.5 mL). The reaction
◦
mixture was heated under reflux at 90 C for 5 h, after which
TLC analysis showed complete consumption of starting material.
The reaction mixture was cooled to room temperature and
concentrated in vacuo. The residue was purified by silica gel
column chromatography (5% MeOH in DCM) to afford free
obtain compound 5 (0.25 g, 95%) as a white solid. R
f
= 0.40 (20%
2
5
1
ethyl acetate in hexane); [a]D = −33.56 (c = 1, CHCl
300 MHz, CDCl
H, Ja,b = 10.8 Hz, H-3
H-3 CH Ph), 4.42 (bd, 2H, H-4 CH
.9 Hz, H-1), 3.91–3.76 (m, 3H, H-1 , H-6a), 3.61–3.57 (m, 1H,
H-6b), 3.47–3.43 (m, 1H, H-3 ), 3.39–3.33 (m, 1H, H-3), 3.28–3.13
m, 3H, H-4, H-5, H-2), 2.63 (s, 1H, H-6 OH), 1.82–1.76 (m, 2H,
3
). H NMR
(
2
3
): d = 7.19–7.46 (m, 15H, aromatic), 4.71 (d,
CH
Ph), 4.53 (d, 2H, Jb,a = 10.8 Hz,
Ph), 4.35 (d, 1H, J1,2
L
a
H
b
amine (0.59 g, 95% yield) as a white solid. R
f
= 0.25 (2% methanol
). H NMR (600 MHz,
): d = 7.57–7.30 (m, 20H, aromatic), 5.53 (s, 1H, >CHPh),
CH Ph), 4.56 (d, 1H, Ja,b
L
a
H
b
2
=
2
5
1
in DCM); [a]D = −23.76 (c = 1, CHCl
3
9
L
CDCl
3
L
4
.81 (d, 1H, Ja,b = 10.8 Hz, H-3
L
a
H
b
=
(
1
1.4 Hz, H-3
L
CH
a
H
b
Ph), 4.51–4.46 (dd, 2H, J = 11.4 Hz, H-4
H-2
L
), 1.30–1.48 (m, 2H, H-4
L
), 1.18 [m, 22H, H-(L
); C NMR (300 MHz, CDCl
): d = 127.67–139.13
aromatic), 86.28 (C-1), 85.56 (C-4), 79.76 (C-5), 77.50 (C-3), 76.46
C-3 ), 75.26 (C-3 CH Ph), 71.11 (C-4 CH Ph, C-1 , C-6), 65.48
C-2), 35.10 (C-2 ), 34.33 (C-4 ), 32.17–29.93 [C-(5 –15 )], 14.37
C-16 ); HRMS (m/z) for C42 S[M + Na] : calcd 740.4175,
5
–L15)], 0.81
ꢀ
CH
2
Ph), 4.44 (d, 1H, J1ꢀ,2ꢀ = 10.2 Hz, H-1 ), 4.32–4.30 (dd, 1H,
1
3
(
(
(
(
(
t, 3H, H-16
L
3
ꢀ
J
7
5ꢀ,6aꢀ = 4.8 Hz, J6aꢀ,6bꢀ = 10.2 Hz, H-6a ), 4.26 (d, 1H, J1,2
=
.8 Hz, H-1), 4.09 (bd, 1H, H-6a), 3.96–3.91 (m, 2H, H-1
L
), 3.76
L
L
2
2
L
ꢀ
(
5
(
t, 1H, J5ꢀ,6bꢀ = 10.2 Hz, J6bꢀ,6aꢀ = 10.2 Hz, H-6b ), 3.63–3.48 (m,
L
L
L
L
ꢀ
H, H-6b, H-3, H-4, H-3
L
, H-5), 3.39 (m, 1H, H-5 ), 3.32–3.24
+
L
H
59
N
3
O
5
ꢀ
ꢀ
ꢀ
m, 3H, H-4 , H-3 , H-2 ), 2.76 (t, 1H, J2,3 = J1,2 = 8.4 Hz, H-2),
.86 (m, 2H, H-2 ), 1.47–1.55 (m, 2H, H-4 ), 1.34–1.24 [m, 22H,
H-(5 –15 )], 0.88 (t, 3H, H-16
found 740.4748.
1
L
L
1
3
Phenyl 3-O-acetyl-6-O-(3-O-acetyl-4,6-O-benzylidene-2-deoxy-
-phthalimido-b-D-glucopyranosyl)-2-azido-3-O-[(R)-3-benzyloxy-
hexadecanoyl]-4-O-benzyl-2-deoxy-1-thio-b-D-glucopyranoside (12).
A suspension of 4 (0.59 g, 1.03 mmol) and 5 (0.61 g, 0.86 mmol)
and molecular sieves (4 A, 0.5 g) in DCM (10 mL) was stirred
under an argon atmosphere for 2 h. The mixture was cooled
L
L
L
); C NMR (300 MHz, CDCl
3
):
ꢀ
2
d = 126.49–139.10 (aromatic), 105.12 (C-1 ), 102.15 (>CHPh),
ꢀ
ꢀ
86.17 (C-1), 85.70 (C-4 ), 81.58 (C-3), 79.05 (C-3
76.43 (C-5), 75.19 (C-3 CH Ph), 73.49 (C-4), 71.20 (C-1
Ph), 69.09 (C-6), 68.95 (C-6 ), 66.66 (C-5 ), 65.42
), 34.31 (C-4 ), 32.15–22.92 (C
); HRMS (m/z) for C55
L
), 78.16 (C-3 ),
L
2
L
),
ꢀ
ꢀ
˚
71.09 (C-4 CH
(C-2 ), 58.22 (C-2), 35.09 (C-2
2
ꢀ
L
L
◦
+
(
−35 C) followed by the addition of NIS (0.23 g, 1.04 mmol) and
5
L
–15
L
), 14.36 (C-16
L
H
74
N
4
9
O S[M + Na] :
TMSOTf (0.01 g, 0.05 mmol). The reaction mixture was stirred
calcd 989.5177, found 989.6476. Lipid 7 (0.26 g, 0.39 mmol)
was dissolved in DCM (10 mL) and DCC (92.5 mg, 0.45 mmol)
was added. The mixture was stirred for 20 min and then the
above free amine was added (0.28 g, 0.29 mmol). The reaction
mixture was stirred for 16 h at room temperature after which
TLC analysis indicated completion of the reaction. The urea
was filtered off over a pad of Celite and the organic solvent was
concentrated in vacuo. The residue was purified by silica gel
column chromatography (15% ethyl acetate in toluene) to afford
◦
for 45 min allowing it to slowly reach −10 C after which TLC
analysis showed complete consumption of the starting materials.
The reaction mixture was quenched with pyridine (0.1 mL) and
diluted with DCM (10 mL). The molecular sieves were removed by
filtration through a pad of Celite. The filtrate was then washed with
aqueous Na
The organic phase was dried (MgSO
The crude mixture was purified by a silica gel chromatography
25% ethyl acetate in hexane) to afford 12 (0.74 g, 76%, yield
2
S
2
O
3
(2 × 20 mL, 15%) followed by water (20 mL).
4
) and concentrated in vacuo.
(
compound 13 (0.43 g, 89%) as a white solid. R
f
= 0.65 (30% ethyl
2
5
1
based on acceptor) as a white solid. R
f
= 0.55 (30% ethyl acetate
acetate in toluene); [a]D = −15.96 (c = 1, CHCl
3
). H NMR
2
5
1
in hexane); [a]D = −19.28 (c = 1, CHCl
3
). H NMR (600 MHz,
(500 MHz, CDCl
3
): d = 7.57–7.26 (20H, m, aromatic), 6.05 (d,
ꢀ
CDCl
3
): d = 7.70–7.07 (m, 24H, aromatic), 5.89 (t, 1H, J3ꢀ2ꢀ
=
1H, JNHꢀ,2ꢀ = 5.5 Hz, NH ), 5.57 (s, 1H, >CHPh), 5.07–5.05 (m,
ꢀ
9
.6 Hz, J3ꢀ4ꢀ = 10.2 Hz, H-3 ), 5.57–5.56 (d, 2H, J1ꢀ2ꢀ = 9.0 Hz,
1H, H-3LL), 4.79 (d, 1H, Ja,b = 11.0 Hz, H-3
L
CH
a
H
b
Ph), 4.75
ꢀ
ꢀ
ꢀ
H-1 , >CHPh), 4.49–4.39 (m, 5H, H-3
L
, H-4 CH
2
Ph, H-6 a), 4.38
(d, 1H, J1ꢀ,2ꢀ = 8.5 Hz, H-1 ), 4.54–4.46 (m, 4H, H-3
L
CH Ph,
a
H
b
ꢀ
(
t, 1H, J 2ꢀ3ꢀ = 9.6 Hz, J1ꢀ2ꢀ = 9.0 Hz, H-2 ), 4.27–4.24 (m, 2H,
H-1, C-4 CH
2
Ph), 4.36–4.33 (dd, 1H, J5ꢀ,6aꢀ = 5.0 Hz, J6aꢀ,6bꢀ
=
ꢀ
ꢀ
H-1, H-3
L
CH
b, H-4 , H-1
H, H-3
2
Ph), 4.06 (bd, 1H, H-6a), 3.87–3.79 (m, 4H, H-
10.5 Hz, H-6a ), 4.13–4.11 (bd, 1H, H-6a), 4.08 (t, 1H, H-3 ),
ꢀ
ꢀ
ꢀ
6
1
L
), 3.73–3.66 (m, 2H, H-5 , H-6b), 3.47–3.45 (m,
3.98–3.91 (m, 2H, H-1
10.5 Hz, H-6b ), 3.63–3.56 (m, 3H, H-6b, H-5, H-4 ), 3.53–3.44
L
), 3.79 (t, 1H, J5ꢀ,6aꢀ = 10.5 Hz, J6bꢀ,6aꢀ
=
ꢀ
ꢀ
L
), 3.33 (m, 1H, H-5), 3.19–3.18 (m, 2H, H-4, H-3), 3.12
This journal is © The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 2087–2097 | 2093

View Article Online
ꢀ
ꢀ
(
m, 3H, H-3
L
, H-2 , H-5 ), 3.36 (t, 1H, J2,3 = J3,4 = 9.0 Hz, H-3),
.26–3.21 (m, 2H, H-4, H-2), 2.25–2.31 (m, 4H, H-2LL, H-2LLꢀ ),
.89–1.85 (dd, H, H-2 ), 1.60–1.46 (m, 6H, H-4
, H-4LL, H-3LLꢀ ),
.41–1.15 [m, 92H, H-(5 –15
, 5LL–15LL, 4LLꢀ –27LLꢀ )], 0.89 (t, 9H,
, 16LL, 28LLꢀ ); C NMR (300 MHz, CDCl
toluene) to afford compound 14 (0.25 g, 86% yield) as a white
2
5
3
1
1
1
solid. R
f
= 0.35 (20% ethyl acetate in toluene); [a] = −34.48
D
1
L
L
(c = 1, CHCl
3
). H NMR (500 MHz, CDCl
3
): d = 7.23–7.49
ꢀ
(m, 30H, aromatic), 6.66 (d, 1H, JNHꢀ,2ꢀ = 9.0 Hz, NH ), 5.43 (s,
L
L
1
3
6
L
3
): d = 174.39, 172.03
1H, >CHPh), 5.37 (d, 1H, JNH,2 = 9.0 Hz, NH), 5.26 (t, 1H,
ꢀ
ꢀ
(
C=O), 126.60–139.08 (aromatic), 102.19 (CHPh), 101.41 (C-1 ),
J = 9.5 Hz, J = 10 Hz, H-3 ), 5.02–4.96 (m, 1H, H-3
L
), 4.75 (d,
ꢀ
8
7
7
6
(
5
5.63 (C-1), 85.51 (C-3), 81.72 (C-4 ), 79.09 (C-5), 78.45 (C-4),
1H, Ja,b = 11.0 Hz, H-3
L
CH
a
H
b
Ph), 4.71 (d, 1H, J1ꢀ,2ꢀ = 8.5 Hz,
CH Ph), 4.52–4.39
Ph), 4.34–4.32 (dd,
ꢀ
ꢀ
6.41 (C-3
1.24 (C-1
L
), 75.24 (C-3
), 71.08 (C-4 CH
6.67 (C-5), 65.09 (C-2), 59.48 (C-2 ), 42.58 (C-2LL, 2LLꢀ ), 35.08
), 34.30–34.79 (C-4
LL–15LL, 4LLꢀ –27LLꢀ ], 14.36 (16
L
CH
2
Ph), 71.97 (C-3 ), 71.62 (C-3LL),
H-1 ), 4.57 (d, 1H, Jb,a = 11.0 Hz, H-3
L
a
H
b
ꢀ
L
2
Ph), 69.14 (C-6), 68.83 (C-6 ),
(m, 7H, H-1, H-4 CH
2
Ph, H-3LA × 2 CH
2
ꢀ
ꢀ
1H, J5ꢀ,6aꢀ = 5.5 Hz, J6aꢀ,6bꢀ = 11.0 Hz, H-6a ), 4.01 (d, 1H, Ja,b
=
ꢀ
C-2
L
L
, C-4LL, C-3LLꢀ ), 32.16–22.92 [C-(5
L
–15
L
,
11.0 Hz, H-6a), 3.86–3.62 (m, 9H, H-2, H-6b , H-3
L
, H-3LA CH ×
ꢀ
ꢀ
, 16LL, 28LLꢀ ); HRMS (m/z) for
2, H-1
, H-6b, H-2 ), 3.54–3.40 (m, 5H, H-3, H-4, H-5, H-4 ,
L
L
+
ꢀ
C
99
H
158
N
4
O
12S[M + Na] : calcd 1650.1597, found 1650.2234.
H-5 ), 2.69–2.40 (m, 4H, H-2LA × 2), 2.31–2.07 (m, 4H, 2LL, 2LLꢀ ),
1
.76–1.79 (m, 2H, 2
1.33–1.27 (m, 136H, 5
0.90 (t, 15H, 16
CDCl
L
), 1.68–1.39 [m, 10H, H-(4
, 4LL, 3LLꢀ , 4LA],
L
ꢀ
Phenyl 4-O-benzyl-6-O-(4,6-O-benzylidene-3-O-[(R)-3 -benzyl-
oxy-hexadecanoyl]-2-deoxy-2 -[(R)-3-octacosanoyloxy-hexadecan]-
amido-b-D-glucopyranosyl)-2-[(R)-3-benzyloxy-hexadecan]-amido-
L
–15
L
, 5LL–15LL, 4LLꢀ –27LLꢀ , 5LA–15LA × 2),
ꢀ
13
, 16LL, 28Lꢀꢀ , 16LA × 2); C NMR (300 MHz,
L
): d = 174.00 (C=O), 171.44 (C=O), 171.32 (C=O), 169.70
3
3
-O-[(R)-3-benzyloxy-hexadecanoyl]-2-deoxy-1-thio-b-D-glucopyra-
(C=O), 139.22–126.33 (aromatic), 101.61 (>CHPh, C-1), 86.18
ꢀ ꢀ ꢀ
noside (14). Compound 13 (0.26 g, 0.16 mmol) was dissolved
in DCM (10 mL) and zinc dust (0.11 g, 1.67 mmol) was added
followed by acetic acid (100 lL, 1.75 mmol). The reaction mixture
was stirred for 2 h at room temperature after which TLC analysis
showed completion of the reaction. The reaction mixture was
washed with NaHCO
layer was dried (MgSO
product was purified by silica gel column chromatography (2%
MeOH in DCM) to afford a free amine (0.22 g, 84%) as a white
(C-1 ), 83.68 (C-5), 79.79 (C-4 ), 79.17 (C-4), 78.21 (C-5 ), 76.35
ꢀ
(C-1
L
), 75.82 (C-3), 74.84 (C-3
L
CH
2
Ph), 71.69 (C-3 ), 71.42
(C-4 CH
2
Ph), 71.32–70.65 (C-3
L
, C-3LA CH
2
Ph), 70.25 (C-3LL),
ꢀ
ꢀ
68.87 (C-6 ), 68.56 (C-6), 55.21 (C-2), 55.02 (C-2 ), 41.67, 41.53
(C-2LL, 2LLꢀ ), 39.81 (C-2LA), 35.09 (C-2
), 34.80–34.29 [C-(4LL
–15
,
,
L
(10 mL), water (10 mL) and the organic
) and concentrated in vacuo. The crude
3
LLꢀ , 4
L
, 4LA × 2)], 32.16–22.92 [C-(5
L
L
, 5LL–15LL, 4LLꢀ –27LLꢀ
3
5LA–15LA)], 14.34 (C-16
L
, 16LL, 28LLꢀ , 16LA); HRMS (m/z) for
4
+
C
99
H
158
N
4
O12S[M + Na] : calcd 2312.7020, found 2312.8816.
2
5
solid. R
CHCl
f
= 0.45 (2% MeOH in DCM); [a] = −9.24 (c = 1,
D
1
ꢀ
3
). H NMR (600 MHz, CDCl
3
): d = 7.56–7.28 (m, 20H,
4-O-Benzyl 6-O-(4,6-O-benzylidene-3-O-[(R)-3 -benzyloxy-hexa-
ꢀ
ꢀ
aromatic), 6.05 (d, 1H, JNHꢀ,2ꢀ = 6.6 Hz, NH ), 5.57 (s, 1H,
decanoyl]-2-deoxy-2 -[(R)-3-octacosanoyloxy-hexadecan]amido-b-
D-glucopyranosyl)-2-[(R)-3-benzyloxy-hexadecan]-amido-3-O-[(R)-
3-benzyloxy-hexadecanoyl]-2-deoxy-a-D-glucopyranose (15). Com-
pound 14 (0.05 g, 0.02 mmol) was dissolved in a mixture of
DCM and water (3.0 mL, 100 : 1 v/v) and the resulting solution
>
CHPh), 5.11–5.09 (m, 1H, H-3LL), 4.78 (d, 1H, Ja,b = 11.0 Hz,
ꢀ
H-3
(
J
L
-CH
a
H
b
Ph), 4.74 (d, 1H, J1ꢀ,2ꢀ = 8.5 Hz, H-1 ), 4.56–4.44
m, 4H, H-3
L
-CH
a
H
b
Ph, H-1, H-4 CH
2
Ph), 4.35–4.32 (dd, 1H,
ꢀ
6aꢀ,5ꢀ = 5.4 Hz, J6aꢀ,6bꢀ = 12.6 Hz, H-6a ), 4.12–4.14 (m, 1H,
ꢀ
◦
H-6a), 4.08 (t, 1H, J = 10.8 Hz, J = 11.4 Hz, H-3 ), 4.00–3.97
was cooled to 0 C. NIS (0.03 g, 0.13 mmol) and TMSOTf
(
(
(
m, 1H, H-1
L
-CH
1
H), 3.87–3.85 (m, 1H, H-1
L
-CHH
2
), 3.79
(0.5 lL, 0.28 lmol) were added and after stirring for 30 min
ꢀ
◦
t, 1H, J6bꢀ,5ꢀ = 12.0 Hz, J6bꢀ,6aꢀ = 12.6 Hz, H-6b ), 3.66–3.61
at 0 C, TLC analysis indicated completion of the reaction. It
ꢀ
ꢀ
ꢀ
m, 2H, H-4, H-3
L
), 3.60–3.43 (m, 4H, H-6b, H-4 , H-2 , H-5 ),
was then quenched with pyridine (0.1 mL) and washed with
3
1
.32–3.24 (m, 2H, H-3, H-5), 2.75 (t, 1H, J = 10.8 Hz, J =
1.4 Hz, H-2), 2.27–2.23 [m, 4H, H-(2LL, 2LLꢀ )], 1.86–1.83 (dd, 2H,
), 1.58–1.45 [m, 6H, H-(4
, 4LL, 3LLꢀ )], 1.39–1.14 [m, 92H,
–15
, 5LL–15LL, 4LLꢀ –27LLꢀ )], 0.896 (t, 9H, H-16 , 16LL, 28Lꢀꢀ ).
C NMR (300 MHz, CDCl
Na
2
S
2
O (8 mL, 15%) and water (8 mL). The organic layer was
3
dried (MgSO4) and concentrated in vacuo. The crude product
was purified by silica gel column chromatography (1% MeOH in
DCM) followed by size exclusion chromatography over LH-20
(MeOH–DCM, 1 : 1 v/v) to yield 15 (0.021 g, 44%) as a white
solid. R
CDCl
H-2
H-(5
L
L
L
L
L
1
3
3
): d = 139.02–126.61 (aromatic),
ꢀ
1
102.17 (>CHPh), 101.48 (C-1 ), 88.58 (C-1), 86.85 (C-4), 81.72
f
= 0.40 (1% methanol–DCM); H NMR (500 MHz,
ꢀ
(
C-4 ), 79.31 (C-3), 78.98 (C-2), 76.36 (C-3
L
), 75.07 (C-3
L
CH
2
Ph),
Ph),
3
): d = 7.48–7.24 (m, 25H, aromatic), 6.51 (d, 1H, JNH,2
=
ꢀ
ꢀ
7
6
2
2.07 (C-3 ), 71.49 (C-3LL), 70.88 (C-1
9.36 (C-6), 68.48 (C-6 ), 66.67 (C-5 ),59.52 (C-2 ), 42.42 (C-2LL
L
), 70.78 (C-4 CH
2
9.5 Hz, NH), 5.91 (d, 1H, JNHꢀ, 2ꢀ = 8.5 Hz, NH ), 5.45 (s, 1H,
ꢀ
ꢀ
ꢀ
ꢀ
,
>CHPh), 5.43–5.39 (m, 1H, H-3 ), 5.16 (d, 1H, J1ꢀ,2ꢀ = 8.5 Hz,
ꢀ
LLꢀ ), 35.04 (C-4
–15
, 5LL–15LL, 4LLꢀ –27LLꢀ ), 14.35 (16
m/z) for C99
624.3170. Lipid 8 (0.23 g, 0.62 mmol) was dissolved in DCM
13 mL) followed by the addition of DCC (0.16 g, 0.75 mmol) and
), 34.13–34.78 (C-4LL, 3LLꢀ , 2
), 22.92–32.16
H-1 ), 5.09 (bs, 1H, H-1), 4.99–4.97 (m, 1H, H-3LL), 4.81 (d, 1H,
L
L
(
(
C-5
L
L
L
, 16LL, 28LLꢀ ). HRMS
calcd 1624.1692, found
J
a,b = 11 Hz, H-3
L
CH
a
H
b
Ph), 4.54–4.38 (m, 7H, H-4 CH
2
Ph,
+
H
160
N
2
O12S[M + Na]
:
H-3 CH Ph, H-3LA CH
L
a
H
b
2
ꢀ
Ph × 2), 4.38–4.34 (dd, 1H, J5ꢀ,6ꢀ
=
1
(
5.5 Hz, J6aꢀ,6bꢀ = 11 Hz, H-6a ), 4.11–4.16 (m, 1H, H-2), 4.02–3.97
(m, 2H, H-1
L
), 3.86–3.74 (m, 5H, H-6a, H-3
L
, H-3LA CH × 2,
ꢀ
ꢀ
ꢀ
ꢀ
DMAP (0.046 g, 0.38 mmol). The reaction mixture was stirred
for 20 min followed by addition of the above amino compound
H-6b ), 3.68–3.62 (m, 5H, H-2 , H-6b, H-5 , H-4 , H-3), 3.58–3.42
(m, 3H, H-5, H-4, H-1 OH), 2.67–2.49 (m, 4H, H-2 LA CH
2.37–2.20 (m, 4H, H-2LL, 2LLꢀ ) 1.82–1.72 (m, 2H, H-2
), 1.60–1.48
(10H, 4
2
× 2),
(
0.20 g, 0.13 mmol). The reaction mixture was stirred for 16 h at
L
room temperature after which TLC analysis indicated completion
of the reaction. The urea was filtered off over a pad of Celite and
the organic solvent was concentrated in vacuo. The crude product
was purified by silica gel chromatography (20% ethyl acetate in
L
, 4LL, 3LLꢀ , 4LACH
2
× 2), 1.39–1.27 [m, 136H, H-(5
L
–15
, 16LL
L
,
,
5
LL–15LL, 4LLꢀ –27LLꢀ , 5LA–15LA × 2], 0.89 (t, 15H, H-16
L
+
28Lꢀꢀ , 16LA × 2); HRMS (m/z) for C139
H
N
O17[M + Na] : calcd
228
2
2220.6936, found 2220.9749.
2
094 | Org. Biomol. Chem., 2007, 5, 2087–2097
This journal is © The Royal Society of Chemistry 2007

View Article Online
ꢀ
46
4
-O-Benzyl 6-O-(4,6-O-benzylidene-3-O-[(R)-3 -benzyloxy-hexa-
reported elsewhere. The E. coli lipid A was reconstituted in PBS
ꢀ
◦
decanoyl]-2-deoxy-2 -[(R)-3-octacosanoyloxy-hexadecan]amido-
b-D-glucopyranosyl)-2-[(R)-3-benzyloxy-hexadecan]-amido-3-O-
(R)-3-benzyloxy-hexadecanoyl]-2-deoxy-a-D-glucono-1,5-lactone
16). Compound 15 (0.013 g, 5.92 lmol) was dissolved in DCM
˚
2 mL) and molecular sieves (3 A, 0.02 g) were added and, after
with DMSO (10%) and stored at −80 C. Synthetic compounds 1
◦
and 3 were stored lyophilized at −80 C and reconstituted in dry
[
(
(
THF on the day of the experiment; final concentrations of THF
in the biological experiments never exceeded 0.5% to avoid toxic
effects.
stirring the resulting suspension for 2 h under an atmosphere
of argon, PCC (0.063 g, 29.6 lmol) was added. The reaction
mixture was stirred for 2 h at room temperature after which TLC
analysis indicated completion of the reaction. After concentration
in vacuo, the crude product was purified by iatrobead column
chromatography (20% ethyl acetate in toluene) to afford 16
Cell maintenance. Mono Mac 6 (MM6) cells, provided by
Dr H.W.L. Ziegler-Heitbrock (Institute for Inhalation Biology,
Munich, Germany), were cultured in RPMI 1640 medium with
L-glutamine (BioWhittaker) supplemented with penicillin (100 u
−
1
−1
mL )–streptomycin (100 lg mL ; Mediatech, OPI supplement
1%; Sigma; containing oxaloacetate, pyruvate and bovine insulin)
(
(
0.008 g, 62%) as a white solid. R
f
= 0.60 (20% ethyl acetate
and fetal calf serum (FCS; 10%; HyClone). New batches of frozen
cell stock were grown up every 2 months and growth morphology
evaluated. Before each experiment, MM6 cells were incubated
1
in toluene); H NMR (500 MHz, CDCl
3
): d = 7.48–7.24 (m, 25H,
aromatic), 6.99 (d, 1H, JNH,2 = 8.5 Hz, NH), 6.66 (d, 1H, JNHꢀ,2ꢀ
=
ꢀ
ꢀ
7
1
1
.5 Hz, NH ), 5.67 (t, 1H, J = 9.5 Hz, J = 9.0 Hz, H-3 ), 5.39 (s,
−
1
with calcitriol (10 ng mL ; Sigma) for 2 days to differentiate
into macrophage like cells. RAW 264.7 cNO(−) cells, derived
from the RAW 264.7 mouse monocyte–macrophage cell line, were
obtained from ATCC. The cells were maintained in RPMI 1640
medium (ATCC) with L-glutamine (2 mM), adjusted to contain
H, >CHPh), 5.12 (m, 1H, H-3LL), 5.00 (d, 1H, J1ꢀ,2ꢀ = 8.0 Hz, H-
ꢀ
), 4.74 (d, 1H, Ja,b = 11.0 Hz, H-3
L
CH
Ph, H-3 LA CH
a
H Ph), 4.62–4.38 (m, 8H,
b
ꢀ
H-2, C-4 CH
4
(
(
5
2
3
1
2
Ph, H-3
L
CH
a
H
b
2
Ph, H-2LA CH
2
Ph),
ꢀ
.31–4.28 (dd, 1H, J5ꢀ,6aꢀ = 4.0 Hz, J6aꢀ,6bꢀ = 10.5 Hz, H-6a ), 4.08
d, 1H, J = 10.5 Hz, H-6a), 3.84–3.80 (m, 2H, H-1
L
), 3.74–3.65
−
1
−1
sodium bicarbonate (1.5 g L ), glucose (4.5 g L ), HEPES
10 mM) and sodium pyruvate (1.0 mM) and supplemented with
ꢀ
ꢀ
m, 4H, H-6b , H-3
L
, H-3LA CH ×2), 3.63–3.47 (m, 7H, H-4 , H-
(
ꢀ
ꢀ
ꢀ
, H-6b, H-2 , H-4, H-5, H-3), 2.69–2.46 (m, 4 H, H-2LA, H-3 LA),
−
1
−1
penicillin (100 u mL )–streptomycin (100 lg mL ) and FBS
10%). Human embryonic kidney (HEK) 293T cells were grown
in Dulbecco’s modified Eagle’s medium (ATCC) with L-glutamine
4 mM), glucose (4.5 g L ) and sodium bicarbonate (1.5 g L )
supplemented with penicillin (100 u mL )–streptomycin (100 lg
mL ), Normocin (100 lg mL ; InvivoGen) and FBS (10%).
Stably transfected HEK 293T cells with murine TLR4/MD2 and
murine TLR2/TLR6 were obtained from InvivoGen and grown
in the same growth medium as for HEK 293T cells supplemented
with the appropriate selective agents HygroGold (50 lg mL ;
InvivoGen) and blasticidin (10 lg mL ; InvivoGen). All cells
.42–2.24 (m, 4H, H-2LL, 2LLꢀ ), 2.07–1.44 (m, 12H, H-2
LLꢀ , H-4LA), 1.27–1.02 (m, 136H, 5
–15
, 4
L
, 4LL
,
L
(
, 5LL–15LL, 4LLꢀ –27LLꢀ , 5LA
–
L
L
ꢀꢀ
5
LA), 0.98–0.72 (t, 15H, 16
L
, 16LL, 28L , 16LA); HRMS (m/z) for
−
1
−1
(
+
C
139
H
226
N
2
O17[M + Na] : calcd 2218.6779, found 2218.8311.
−
1
ꢀ
ꢀ
−1
−1
2
-Deoxy-6-O-(2 -deoxy-3-O-[(R)-3 -benzyloxy-hexadecanoyl]-
ꢀ
2
-[(R)-3-octacosanoyloxy-hexadecan]amido-b-D-glucopyranosyl)-
2
-[(R)-3-benzyloxy-hexadecan]-amido-3-O-[(R)-3-benzyloxy-hexa-
decanoyl]-2-deoxy-a-D-glucono-1,5-lactone (3). Pd/C (10 mg,
0 wt%) was added to compound 16 (4.5 mg, 2.05 lmol) dissolved
in a mixture of THF–t-BuOH (2 mL, 1 : 1, v/v). The flask was
degassed and filled with H gas and then stirred for 12 h. The
−
1
1
−
1
◦
2
were maintained in a humid 5% CO atmosphere at 37 C.
2
reaction progress was monitored by MALDI. After completion
of the reaction, the catalyst was filtered off through a PTFE filter
and washed with THF (0.5 mL × 3) The combined filtrates were
concentrated in vacuo and the residue was purified by LH-20 size
exclusion column chromatography (i-PrOH–DCM, 1 : 1, v/v)
Cytokine induction and ELISAs. On the day of the exposure
assay differentiated MM6 cells were harvested by centrifugation
6
−1
◦
and gently suspended (10 cells mL ) in prewarmed (37 C)
5
medium and RAW 264.7 cNO(−) cells were plated as 2 × 10 cells
per well in 96-well tissue culture plates (Nunc). Cells were then
incubated with different combinations of stimuli for 5.5 hours.
1
to yield compound 3 (2.3 mg, 66%) as a white solid. H NMR
(
500 MHz, THF-D : (CD
3
)
2
CDOD 1 : 1): d = 5.12 (m, 1H, 3LL),
Culture supernatants were then collected and stored frozen
ꢀ
5
9
J
3
.02 (t, 1H, J = 10.0 Hz, J = 9.0 Hz, H-3 ), 4.67 (d, 1H, J 1ꢀ,2ꢀ
=
◦
(
−80 C) until assayed for cytokine production. All cytokine
ꢀ
.0 Hz, H-1 ), 4.21 (m, 1H, H-2), 4.16 (dd, 1H, J5ꢀ6ꢀ = 2.0 Hz,
ELISAs were performed in 96-well MaxiSorp plates (Nalge
Nunc International). Concentrations of human TNF-a protein in
culture supernatants were determined by a solid phase sandwich
ELISA. Plates were coated with purified mouse anti-human TNF-
a antibody (Pharmingen). TNF-a in standards and samples was
allowed to bind to the immobilized antibody. Biotinylated mouse
anti-human TNF-a antibody (Pharmingen) was then added, pro-
ducing an antibody–antigen–antibody “sandwich”. After addition
of avidin–horseradish peroxidase conjugate (Pharmingen) and
ABTS peroxidase substrate (Kirkegaard & Perry Laboratories),
a green color was produced in direct proportion to the amount
of TNF-a present in the sample. The reaction was stopped by
adding peroxidase stop solution (Kirkegaard & Perry Labora-
tories) and the absorbance was measured at 405 nm using a
microplate reader (BMG Labtech). Cytokine DuoSet ELISA
Development Kits (R & D Systems) were used for the cytokine
ꢀ
6aꢀ,6bꢀ = 11.5 Hz, H-6a ), 4.05 (m, 1H, H-6a), 3.94 (m, 2H, H-1L)
,
ꢀ
ꢀ
ꢀ
.86–3.45 (m, 11H, H-6b , H-2 , 3
L
, H-3LA × 2 CH, H-6b, H-5 ,
, 4 ),
–15
ꢀ
H-4 , H-3, H-4, H-5), 2.46–2.18 (m, 14H, H-2LA, 2LL, 2LLꢀ , 2
L
L
2
5
2
1
.06–1.94 (m, 6H, H-4LL, 3LLꢀ , 4LA), 1.84–1.27 (bm, 136H, 5
LL–15LL, 4LLꢀ –27LLꢀ , H-5LA–15LA), 1.65–0.67 (bm, 15H, 16
L
L
,
,
L
+
, 16LL
8
Lꢀꢀ , 16LA); HRMS (m/z) for C104
H
2
N O17[M + Na] : calcd
198
770.4588, found 1770.7673.
Biological experiments
Reagents for biological experiments. E. coli 055:B5 LPS was
obtained from List Biologicals, Pam CysSK was obtained from
3
4
Calbiochem and R. sin-1 LPS and lipid A were kindly provided
by Dr R. Carlson (CCRC, Athens, GA). All data presented in
this study were generated using the same batches of E. coli 055:B5
LPS and R. sin-1 LPS. The synthesis of E. coli lipid A has been
This journal is © The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 2087–2097 | 2095

View Article Online
quantification of human IP-10 and mouse TNF-a according to
the manufacturer’s instructions. The absorbance was measured at
References
1
2
B. Beutler, Mol. Immunol., 2004, 40, 845–859.
E. S. van Amersfoort, T. J. C. van Berkel and J. Kuiper, Clin. Microbiol.
Rev., 2003, 16, 379.
4
50 nm with wavelength correction set to 540 nm. Concentrations
of mouse IFN-b in culture supernatants were determined as
follows. Plates were coated with rabbit polyclonal antibody
against mouse IFN-b (PBL Biomedical Laboratories). IFN-b in
standards and samples was allowed to bind to the immobilized
antibody. Rat anti-mouse IFN-b antibody (USBiological) was
then added, producing an antibody-antigen-antibody “sandwich”.
Next, horseradish peroxidase (HRP) conjugated goat anti-rat IgG
3 S. Akira and K. Takeda, Nat. Rev. Immunol., 2004, 4, 499–511.
4
5
6
K. Takeda, T. Kaisho and S. Akira, Annu. Rev. Immunol., 2003, 21,
3
35–376.
A. F. McGettrick and L. A. J. O’Neill, Mol. Immunol., 2004, 41, 577–
582.
M. Yamamoto, S. Sato, H. Hemmi, K. Hoshino, T. Kaisho, H. Sanjo, O.
Takeuchi, M. Sugiyama, M. Okabe, K. Takeda and S. Akira, Science,
2
003, 301, 640–643.
(
3
H + L) antibody (Pierce) and a chromogenic substrate for HRP
,3 ,5,5 -tetramethylbenzidine (TMB; Pierce) were added. After
7
8
9
F. J. Roberts, I. W. Geere and A. Coldman, Rev. Infect. Dis., 1991, 13,
ꢀ
ꢀ
34–46.
D. Pittet, D. Tarara and R. P. Wenzel, JAMA, J. Am. Med. Assoc., 1994,
the reaction was stopped, the absorbance was measured at 450 nm
with wavelength correction set to 540 nm. All cytokine values are
presented as the means ± SD of triplicate measurements, with
each experiment being repeated three times.
2
71, 1598–1601.
M. Pollack and C. A. Ohl, Curr. Top. Microbiol. Immunol., 1996, 216,
275–297.
1
1
0 D. T. Golenbock, R. Y. Hampton, N. Qureshi, K. Takayama and C. R.
Raetz, J. Biol. Chem., 1991, 266, 19490–19498.
1 F. Peri, C. Marinzi, M. Barath, F. Granucci, M. Urbano and F. Nicotra,
Bioorg. Med. Chem., 2006, 14, 190–199.
Transfection and NF-jB activation assay. The day before
transfection, HEK 293T wild type cells and HEK 293T cells stably
transfected with murine TLR2/TLR6 or murine TLR4/MD2
were plated in 96-well tissue culture plates (16000 cells per well).
The next day, cells were transiently transfected using PolyFect
Transfection Reagent (Qiagen) with expression plasmids pELAM-
Luc (NF-jB-dependent firefly luciferase reporter plasmid, 50 ng
12 O. R. Martin, W. Zhou, X. F. Wu, S. Front-Deschamps, S. Moutel, K.
Schindl, P. Jeandet, C. Zbaeren and J. A. Bauer, J. Med. Chem., 2006,
4
9, 6000–6014.
1
3 J. H. Krauss, U. Seydel, J. Weckesser and H. Mayer, Eur. J. Biochem.,
989, 180, 519–526.
1
14 W. Christ, P. McGuinness, O. Asano, Y. Wang, M. Mullarkey, M. Perez,
L. Hawkins, T. Blythe, G. Dubuc and A. Robidoux, J. Am. Chem. Soc.,
1994, 116, 3637–3638.
47
per well) and pRL-TK (Renilla luciferase control reporter vector,
ng per well; Promega) as an internal control to normalize
1
5 M. Shiozaki, Y. Watanabe, Y. Iwano, T. Kaneko, H. Doi, D. Tanaka,
T. Shimozato and S. Kurakata, Tetrahedron, 2005, 61, 5101–5122.
1
experimental variations. The empty vector pcDNA3 (Invitrogen)
was used as a control and to normalize the DNA concentration for
all of the transfection reactions (total DNA 70 ng per well). Forty-
four h post-transfection, cells were exposed to the stimuli in the
presence of FCS to provide sCD14 at the indicated concentrations
for 4 h, after which cell extracts were prepared. The luciferase
activity was measured using the Dual-Luciferase Reporter Assay
System (Promega) according to the manufacturer’s instructions
and a combination luminometer–fluorometer microplate reader
16 D. P. Rossignol, L. D. Hawkins, W. J. Christ, S. Kobayashi, T. Kawata,
M. Lynn, I. Yamatsu and Y. Kishi, in Endotoxin in Health and Disease,
ed. B. Helmut, S. M. Opal, S. N. Vogel and D. C. Morrison, Marcel
Dekker, Inc., New York, 1999, vol. 1, pp. 699–717.
17 M. L. Vandenplas, R. W. Carlson, B. Jeyaretnam, B. McNeill, M. H.
Barton, N. Norton, T. F. Murray and J. N. Moore, J. Biol. Chem., 2002,
2
77, 41811–41816.
1
8 A. V. Demchenko, M. A. Wolfert, B. Santhanam, J. N. Moore and G. J.
Boons, J. Am. Chem. Soc., 2003, 125, 6103–6112.
19 R. Girard, T. Pedron, S. Uematsu, V. Balloy, M. Chignard, S. Akira
and R. Chaby, J. Cell Sci., 2003, 116, 293–302.
0 T. Pedron, R. Girard, B. Jeyaretnam, R. W. Carlson and R. Chaby,
2
(
BMG Labtech). Expression of the firefly luciferase reporter gene
Immunology, 2000, 101, 262–270.
was normalized for transfection efficiency with expression of
Renilla luciferase. The data are reported as the means ± SD of
triplicate treatments. The transfection experiments were repeated
at least twice.
21 B. Jeyaretnam, J. Glushka, V. S. K. Kolli and R. W. Carlson, J. Biol.
Chem., 2002, 277, 41802–41810.
2
2 B. Santhanam, M. A. Wolfert, J. N. Moore and G. J. Boons, Chem.–
Eur. J., 2004, 10, 4798–4807.
2
3 H. S. Lee, M. A. Wolfert, Y. H. Zhang and G. J. Boons, ChemBiochem,
2
006, 7, 140–148.
Data analysis. Concentration–response and inhibition data
were analyzed using nonlinear least-squares curve fitting in Prism
24 Y. Watanabe, K. Miura, M. Shiozaki, S. Kanai, S. Kurakata and M.
Nishijima, Carbohydr. Res., 2003, 338, 47–54.
2
5 M. Shiozaki, Y. Kobayashi, M. Arai, N. Ishida, T. Hiraoka, M.
Nishijima, S. Kuge, T. Otsuka and Y. Akamatsu, Carbohydr. Res., 1991,
222, 69–82.
(
GraphPad Software, Inc.). Concentration–response data were fit
with the following four parameter logistic equation: Y = Emax/(1 +
Hill slope
(
EC50/X)
), where Y is the cytokine response, X is logarithm
26 Y. Watanabe, K. Miura, M. Shiozaki, S. Kanai, S. Kurakata and M.
Nishijima, Carbohydr. Res., 2001, 332, 257–277.
of the concentration of the stimulus, Emax is the maximum response
and EC50 is the concentration of the stimulus producing 50%
stimulation. The Hillslope was set at 1 to be able to compare
the EC50 values of the different inducers. Inhibition data were
fit with the following logistic equation: Y = Bottom + (Top −
2
2
7 G. J. Boons, Tetrahedron, 1996, 52, 1095–1121.
8 P. J. Garegg, Adv. Carbohydr. Chem. Biochem., 1997, 52, 179–205.
29 H. Weidmann and H. K. Zimmerman, Justus Liebigs Ann. Chem., 1959,
6
28, 255–256.
3
0 K. Fukase, Y. Fukase, M. Oikawa, W. C. Liu, Y. Suda and S. Kusumoto,
Tetrahedron, 1998, 54, 4033–4050.
(X − log IC50
)
Bottom)/(1 + 10
), where Y is the cytokine response, X is
3
1 S. Mehta and B. M. Pinto, J. Org. Chem., 1993, 58, 3269–3276.
32 S. Hatakeyama, H. Mori, K. Kitano, H. Yamada and M. Nishizawa,
the logarithm of the concentration of the inhibitor and IC50 is the
concentration of the inhibitor that reduces the response by half.
Tetrahedron Lett., 1994, 35, 4367–4370.
3
3
3
3 D. S. Keegan, S. R. Hagen and D. A. Johnson, Tetrahedron: Asymmetry,
1
996, 7, 3559–3564.
4 H. M. Zuurmond, G. A. van der Marel and J. H. van Boom, Recl. Trav.
Chim. Pays-Bas, 1991, 110, 301–302.
Acknowledgements
5 S. Mehta and B. M. Pinto, Tetrahedron Lett., 1991, 32, 4435–4438.
This research was supported by the Institute of General Medicine
of the National Institutes of Health (GM061761).
36 S. Kurihara, T. Tsumuraya and I. Fujii, Bioorg. Med. Chem. Lett., 1999,
9, 1179–1184.
2
096 | Org. Biomol. Chem., 2007, 5, 2087–2097
This journal is © The Royal Society of Chemistry 2007

View Article Online
3
3
3
4
7 G. H. Veeneman, S. H. van Leeuwen and J. H. van Boom, Tetrahedron
Lett., 1990, 31, 1331–1334.
8 S. M. Zughaier, Y. L. Tzeng, S. M. Zimmer, A. Datta, R. W. Carlson
and D. S. Stephens, Infect. Immun., 2004, 72, 371–380.
9 S. M. Zughaier, S. M. Zimmer, A. Datta, R. W. Carlson and D. S.
Stephens, Infect. Immun., 2005, 73, 2940–2950.
0 M. G. Netea, B. J. Kullberg, L. A. Joosten, T. Sprong, I. Verschueren,
O. C. Boerman, F. Amiot, W. B. van den Berg and J. W. van der Meer,
Eur. J. Immunol., 2001, 31, 2529–2538.
43 C. Werts, R. I. Tapping, J. C. Mathison, T. H. Chuang, V. Kravchenko,
I. Saint Girons, D. A. Haake, P. J. Godowski, F. Hayashi, A. Ozinsky,
D. M. Underhill, C. J. Kirschning, H. Wagner, A. Aderem, P. S. Tobias
and R. J. Ulevitch, Nat. Rev. Immunol., 2001, 2, 346–352.
44 M. Hirschfeld, J. J. Weis, V. Toshchakov, C. A. Salkowski, M. J. Cody,
D. C. Ward, N. Qureshi, S. M. Michalek and S. N. Vogel, Infect. Immun.,
2001, 69, 1477–1482.
45 A. O. Aliprantis, R. B. Yang, M. R. Mark, S. Suggett, B. Devaux, J. D.
Radolf, G. R. Klimpel, P. Godowski and A. Zychlinsky, Science, 1999,
285, 736–739.
46 Y. Zhang, J. Gaekwad, M. A. Wolfert and G. J. Boons, J. Am. Chem.
Soc., 2007, 129, 5200–5216.
47 J. C. Chow, D. W. Young, D. T. Golenbock, W. J. Christ and F.
Gusovsky, J. Biol. Chem., 1999, 274, 10689–10692.
4
4
1 G. Mathiak, K. Kabir, G. Grass, H. Keller, E. Steinringer, T. Minor,
C. Rangger and L. F. Neville, Int. J. Mol. Med., 2003, 11, 41–
4
4.
2 P. von der Ley, L. Steeghs, H. J. Hamstra, J. ten Hove, B. Zomer and L.
van Alphen, Infect. Immun., 2001, 69, 5981–5990.
This journal is © The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 2087–2097 | 2097