Synthetic chemistry and function of bacterial cell surface glycoconjugates

Article abstract of DOI:10.1002/jccs.200200071

Typical bacterial glycoconjugates are known to stimulate immunological systems of higher animals and thereby play important roles in the primary defense of animals against bacterial infection. Lipopolysaccharide (LPS) of gram-negative bacteria is a representative of such glycoconjugates. LPS was first discovered as a potent bacterial toxin and named endotoxin but was soon found to exhibit immunostimulating activity. By the use of our synthetic pure preparations, the lipophilic partial structure of LPS, designated lipid A, proved to be the active entity responsible for both endotoxic and immunostimulating activities of LPS. This paper deals with our recent chemical synthesis and functional study of lipid A and related compounds. Synthesis is described of its various structural analogues, radio-labeled compound and Re-type LPS that contains two additional sugar moieties linked to lipid A.

Full text of DOI:10.1002/jccs.200200071

Journal of the Chinese Chemical Society, 2002, 49, 453-458
453
Invited Paper
Synthetic Chemistry and Function of Bacterial Cell Surface
Glycoconjugates+
Shoichi Kusumoto*, Koichi Fukase, Masato Oikawa and Yasuo Suda
Department of Chemistry, Graduate School of Science, Osaka University, Toyonaka, Osaka 560-0043, Japan
Typical bacterial glycoconjugates are known to stimulate immunological systems of higher animals and
thereby play important roles in the primary defense of animals against bacterial infection.Lipopolysaccharide
(LPS) of gram-negative bacteria is a representative of such glycoconjugates.LPS was first discovered as a po-
tent bacterial toxin and named endotoxin but was soon found to exhibit immunostimulating activity.By the
use of our synthetic pure preparations, the lipophilic partial structure of LPS, designated lipid A, proved to be
the active entity responsible for both endotoxic and immunostimulating activities of LPS.This paper deals
with our recent chemical synthesis and functional study of lipid A and related compounds.Synthesis is de-
scribed of its various structural analogues, radio-labeled compound and Re-type LPS that contains two addi-
tional sugar moieties linked to lipid A.
Bacterial cells are covered with a rigid cell envelope
which is mainly composed of complex glycoconjugates hav-
ing unique structures never found in higher animals. When
infection of bacteria occurs, defensive cells of animals such
as macrophages and lymphocytes start their action and pro-
duce protein mediators (cytokines) such as interleukins,
interferons, and tumor necrosis factors to enhance the entire
defense system. Interestingly, not whole living cells of bacte-
ria are required for such activation of defense mechanism but
certain typical cell surface glycoconjugates represent the role
of the living bacteria. It was further demonstrated that defi-
nite chemical entities corresponding to particular partial
structures of the glycoconjugates possess the activity.
ent highly challenging targets with complex chemical struc-
tures and amphiphilic natures difficult to handle. In this pre-
sentation we will focus on our synthetic work on lipopoly-
saccharide, a typical bioactive glycoconjugate of gram-nega-
tive bacteria.
Synthesis of lipid A, the lipid component of
lipopolysaccharide, and its analogues
Lipopolysaccharide (LPS) is the chemical entity of
bacterial endotoxin which was first described as early as
1892 as a heat-stable potent toxin bound to the cells of
Gram-negative organism. As the name endotoxin implies,
LPS exhibits various detrimental activities such as lethal tox-
icity, pyrogenicity, and tissue-damaging activity, which
sometimes cause serious clinical problems even today. Si-
multaneously, however, LPS also shows a wide range of ben-
eficial activities related to stimulation of immunological re-
sponses. LPS is composed of a hydrophilic polysaccharide
chain covalently bound to a glycolipid designated lipid A.
The latter was then found to be responsible for both benefi-
cial and toxic functions of LPS.2 The chemical structure of
lipid A from Escherichia coli Re mutant was deduced as 1 by
us.3 The proposed structure of lipid A (1) was soon confirmed
by our total synthesis.4 The synthetic lipid A showed identi-
cal biological activities to those of the natural counterpart
isolated from bacterial cells.5
The first example identified as an “active principle” of
cell surface glycoconjugates was N-acetylmuramyl-L-alanyl-
D-isoglutamine (so-called muramyl dipeptide, MDP) de-
scribed by two independent research groups including our-
selves.1 The muramyl peptide is a common building-block of
cell wall peptidoglycan. This particular molecular species
with molecular weight of only around 500 exhibits the defi-
nite immunostimulating activity similar to that of the cell
wall itself. This result encouraged us to continue our study on
other biologically important bacterial glycoconjugates in or-
der to identify the active structures responsible for their
immunostimulation and to disclose the mechanism of their
action. For synthetic organic chemists, these molecules pres-
+ Based on the lecture of Shoichi Kusumoto presented in “K.-T. Wang Bioorganic Chemistry Lectureship Symposium 2001”
held on September 28, 2001 in Taipei.

454 J. Chin. Chem. Soc., Vol. 49, No. 4, 2002
Kusumoto et al.
OH
O
OH
O
O
O
O
O
O
O
O
(HO)2P
(HO)2P O
HO
HO
O
O
O
O
O
O
NH
HN
O
O
O
O
OH
HN
O
OH
O P(OH)2
O
HN
O P(OH)2
O
O
O
O
OH
OH
O
OH
OH
O
: (R)-configuration
: (S)-configuration
E. coli lipid A
1
precursor Ia 2
: (R)-configuration
: (S)-configuration
4
3
OBn
NH
OBn
O
O
OH
OCOR1
O
P
O
CCl3
TMSOTf
O
O
O
O
O
OCOR1
OCOR1
O
P
O
OCOR1
O
O
+
NHTroc
O
HO
OAllyl
NHTroc
HO
OAllyl
NHTroc
NHTroc
74%
1) Zn - AcOH
2) R1CO2H
O
OBn
O
OBn
O
1) Ir comlex
2) I2 - aq. THF
DCC
O
O
O
OCOR1
OCOR1
OCOR1
OCOR1
O
P
O
OH
O
O
O
O
O
HO
OAllyl
NHCOR1
P
O
HO
NHCOR1
NHCOR1
NHCOR1
78%
82%
OBn
O
OH
O
1) n-BuLi
H2 - Pd
2) [(BzlO) 2 PO]2O
O
O
O
O
OCOR1
OCOR2
O
OCOR1
O
OCOR2
O
O
O
P
O
P(OBzl)2
O
HO
NHCOR1
(HO)2P
94%
O
HO
O
P(OH)2
NHCOR1
NHCOR2
NHCOR2
46%
O
4
5
(S)
(S)
R1 CO :CH3(CH2)10CHCH2CO-
OBn
CH (CH )
3
R2 CO
:
2 10 CHCH2CO-
OH
Various structural analogues of lipid A (1) were iso-
lated from cells of other Gram-negative bacteria. Some of
them were also synthesized and tested for their activities.6 A
novel efficient synthetic route to lipid A analogues was re-
cently established as exemplified in the scheme below.
The scheme illustrates preparation of an artificial ana-
logue 4 of a biosynthetic precursor of lipid A 2 which con-
tains only four moles of 3-hydroxytetradecanoic acid. Com-
pound 4 contains (S)-3-hydroxytetradecanoic acid instead of
the corresponding (R)-3-hydroxy acids present in the natural
counterpart 2 and lipid A (1).7 The (1-6)glycosidic linkage
of 3 was formed by the imidate method with the participation
of the neighboring N-Troc group. The use of this glycosyla-
tion method in place of the previously employed glycosyl
bromide4 enabled the protection of the distal phosphate group
as an o-xylidene ester. This cyclic benzyl type protecting
group was removed later smoothly by hydrogenolysis to-
gether with all the other benzyl groups in one step. After in-
troduction of the glycosyl phosphate group, the protected fi-
nal product 5 was carefully purified, which was essentially
important to obtain a high yield of the final product by hy-
drogenolysis. The free lipid A was purified by centrifugal
partition chromatography with a solvent system 1-butanol-
THF-H2O (9 : 7 : 20).7 The purification procedure is applica-
ble with some minor modifications to the synthesis of various
structural analogues of lipid A.8
Cytokine-inducing activity as a typical biological func-
tion of LPS was tested for the synthetic compounds. Figs. 1
and 2 show the interleukin-6 (IL-6)-inducing activity as
tested with human peripheral whole blood cells by the ELISA

Synthesis and Function of Bacterial Glycoconjugates
J. Chin. Chem. Soc., Vol. 49, No. 4, 2002 455
method. The data revealed the following interesting phenom-
ena. The number of acyl groups in the molecule is important
for the nature of the activity, whereas the configurations of
the 3-hydroxy acyl residues have no essential effect. Syn-
thetic E. coli-type lipid A (1) with 3-hydroxy- and 3-acyloxy-
acyl residues of natural (R)-configurations exhibits compara-
ble activity with that of a standard natural LPS (from E. coli
0111:B4 cells). An unnatural E. coli-type analog 3 with (S)-
3-hydroxy acids exhibited almost the same endotoxic activity
as the natural type 1. A biosynthetic precursor of lipid A des-
ignated precursor Ia (2) was found to act as an antagonist to 1
and LPS, suppressing the cytokine-inducing activity of the
latters.9 The precursor-type 4 with (S)-acids exhibited similar
but even stronger antagonistic activity than the natural-type
precursor Ia (2).7,8
Synthesis of radio-labeled lipid A analogues
Chemical synthesis has another obvious advantage that
one can prepare suitably labeled derivatives of lipid A which
are expected to be important tools for the investigation of
binding proteins and the mode of biological action. We next
attempted to incorporate radioactivity to the molecule of
phosphonooxyethyl analogues 6 of lipid A.10 In this artificial
compound one of the phosphate groups of lipid A is bound
not directly to the glycosidic hydroxy group but through an
-glycosidically linked ethylene glycol moiety. The phos-
phonooxyethyl analogue is stable enough so that its prepara-
tion was much easier than the corresponding natural lipid A.
In fact, 6 was obtained through short-step transformations
and purified readily by simple ion-exchange chromatogra-
phy.10 In spite of such modification, the phosphonooxyethyl
analogue 6 showed indistinguishable endotoxic activity from
that of 1.
8000
7000
6000
5000
4000
3000
2000
OH
O
O
O
O
O
(HO)2P
HO
O
O
O
HN
O P(OH)2
O
O
OH
HN
O
O
OH
O
O
O
O
O
1000
3
1
0
0.3
analogue3
0.1
lipid A1
concentration (ng/mL)
6
0.03
LPS
The phosphonooxyethyl analogue, therefore, seemed to
be an ideal target of the synthesis of radio labeled derivative.
We attempted to introduce radioactivity at the ethylene glycol
unit of these molecules. An entirely new route was exploited
for a tritium-labeled 6 as illustrated in the scheme below by
employing a similar strategy to that for the synthesis of 4
above.
Fig. 1. IL-6 induction by synthetic E. coli-type lipid A
1, its (S)-acyl analogue 3, and LPS (E. coli
0111:B4).
2500
2000
1500
1000
500
A disaccharide intermediate 7 was prepared which con-
tains all the acyl groups and 4 -phosphate as well as the
glycosidically bound ethylene glycol. A tritium was incorpo-
rated to the distal hydroxylmethylene group by NaB3H4-re-
duction of the corresponding aldehyde derivative. Phospho-
rylation followed by HPLC purification and hydrogenolytic
deprotection afforded highly pure tritium-labeled 6 with high
specific radioactivity.11,12 The labeled compound was used to
detect lipid A-binding proteins on the cell surface of macro-
phages.12
0
0
0.01
0.1
1
LPS (0.1 ng/mL) +
2
10
100
concentration (ng/mL)
LPS (0.1 ng/mL) +
4
Fig. 2. Inhibitory activity of synthetic lipid A ana-
logues 2 and 4 against IL-6 induction by LPS.

456 J. Chin. Chem. Soc., Vol. 49, No. 4, 2002
Kusumoto et al.
OH
O
OBn
NH
TMSOTf
MS4A
O
OH
OCOR2
+
CCl3
O
P
O
O
O
OMPM
ether
BnO
O
O
N3
NHFmoc
82%
1) DBU
2) R3CO2H, DCC
3) R1CO2H, DCC, DMAP
OBn
O
OBn
O
O
O
O
O
OCOR2
O
P
OCOR1
OCOR2
OH
O
P
O
OMPM
O
O
OMPM
O
BnO
NHCOR3
O
O
BnO
NHFmoc
O
O
N3
N3
43%
1) Zn, AcOH
O
O
DMSO
O
TMSOTf
2) R 1CO2H
DCC
(CF3CO) 2
O
CH3CN-CH2 Cl2
O
O
O
OCOR1
OCOR1
OCOR1
OH
OMPM
BnO
O
CHO
BnO
O
BnO
O
NHCOR1
NHCOR1
50%
(recovery 50%)
NHCOR1
50%
80%
1)
7
O
O
O
O
P
NEt2
O
OCOR1
NaB3H 4
O
O
P
O
O
OCOR1
1H-tetrazole
O
OH
BnO
R1COHN
O
17.1 MBq
(16.3 GB q/mmol)
BnO
R1COHN
O
2) mCPBA
3H
3H
32%
3.14 MBq
(3 .97 GBq/mmol)
58%
9.89 MBq
OH
O
H2 - Pd
O
O
O
OCOR2
OCOR4
O
MPM : CH3O
Fmoc :
CH2-
P(OH)2
O
(HO)2P
O
HO
O
3H
2.30 MBq
NHCOR3 R4COHN
73%
3H-labeled
6
CH2OCO-
R2 CO :CH3(CH2)10
R1 CO :
CH (CH )
2 10 CHCH2CO-
OBn
CHCH2CO- R3 CO : 2 10CHCH2CO- R4 CO :
CH (CH ) CH (CH )
3
2 10 CHCH2CO-
OH
3
3
CH3(CH2)12COO CH3(CH2)10COO
Synthesis of Re-type lipopolysaccharide (Re LPS)
The synthesis was completed by stepwise introduction
of two Kdo moieties and both phosphate followed by cata-
lytic hydrogenolysis. Purification was also effected by parti-
tion chromatogry to afford the highly pure first synthetic LPS
8.13 The intrinsic biological activities such as cytokine induc-
tion of Re LPS and its partial structure lacking one Kdo were
determined for the first time, the effects of contaminating
bacterial components being thereby completely excluded.
Recently, the basic framework of a new defense system
called “innate immunity” has been disclosed mainly by mo-
lecular biological approaches. This is a very basic mecha-
nism which animals are given by born to protect themselves
against a wide range of invading microorganisms. A family
of toll-like receptors (TLRs) were characterized and shown
to play thereby important roles to recognize various bacterial
glycoconjugates. By the use of our homogeneous synthetic
As described, lipid A has proved to be the endotoxic
principle of bacterial LPS. However, lipid A itself is an artifi-
cial molecule never exists in nature in the free form: it is ob-
tained only after mild acid hydrolysis of LPS. The most sim-
ple natural LPS ever found on living bacterial cells is one
called Re-type LPS produced by E. coli Re mutant. Re LPS
(8) contains only two additional acidic sugar moieties called
3-deoxy-D-manno-2-octulosonic acid (Kdo) linked to the
6 -position of lipid A. We also started a synthetic approach to
Re LPS which is a highly challenging target containing both
base-labile ester functionalities and the acid-labile glycosyl
phosphate and ketosidic linkages. After solving several basic
problems, such as stereoselective formation of -Kdo link-
ages, the synthesis was started with an acylated glucosamine
disaccharide devoid of the phosphate as shown in the scheme.

Synthesis and Function of Bacterial Glycoconjugates
J. Chin. Chem. Soc., Vol. 49, No. 4, 2002 457
BnO
BnO
TBSO
2 eq.
1)
O
O
BnO
BnO
TBSO
PNEt2
O CO2Bn
OTBS
O CO2Bn
OTBS
1H-tetrazole
MS4A, CH2Cl2
TESO
HO
O
O
O
F
O
O
OCOR1
OCOR3
O
O
BnO
NHCOR2
OAllyl
OCOR1
OCOR3
BF3 ether comlex
CH2Cl2
2) mCPBA
96%
NHCOR3
HO
BnO
OAllyl
NHCOR3
NHCOR2
89%
BnO
BnO
TBSO
BnO
BnO
HO
O CO2Bn
OTBS
O CO2Bn
OTES
TESCl
imidazole
47% HF : CH2Cl2
: MeCN
O
O
O
O
O
O
O
O
OCOR1
OCOR3
O
P
CH2Cl2
MS4A
OCOR1
OCOR3
1 : 5 : 5
96%
O
O
O
O
O
BnO
OAllyl
NHCOR3
P
O
BnO
NHCOR2
OAllyl
NHCOR3
NHCOR2
O
95%
BnO
BnO
HO
BnO
BnO
O
4 eq.
O CO2Bn
OH
BnO
BnO
HO
O CO2Bn
O
O
TFA
BnO
BnO
OBn
•`
O
(1.5 2.0%)
H2O (0.5%)
O CO2Bn
P
P O
OBn
F
Ir-complex
O
O
O
O
CH2Cl2
82%
O
O
then I2, H2O
81%
BF3 ether comlex
CH2Cl2
LiN(TMS)2 , THF
OCOR1
OCOR3
O
P
O
BnO
NHCOR2
OAllyl
O
NHCOR3
75%
OH
HO
HO
O
CO2 H
HO OH
HO
BnO
BnO
HO
HO
O CO2Bn
O
O
OH
BnO
BnO
HO
CO2 H
Pd(black)
H2 20 atm
O
O CO2Bn
O
O
O
P
O
O
HN
O
O
HO
(HO)2
O
O
O
THF
O
O
HN
O
O P(OH)2
O
O
O
O
O
O
OH
22%
(2 steps)
OCOR1
OCOR3
O
P
O
O
OH
O
O
O P(OBn)2
NHCOR3
O
BnO
NHCOR2
O
O
O
R1 CO=
R2 CO=
R3 CO=
O
O
O
O
O
8
p-CF3PhCH2O
Ir-complex = [Ir(H2)(cod)(MePh2 P)2]PF6
lipid A preparations, participation of TLR4 was unequivo-
cally proved in the process of cell activation by lipid A.14
Synthetic organic chemistry is a powerful tool particu-
larly in functional study of biological active complex
glycoconjugates as exemplified here. The major merit of
chemical synthesis in this field is its ability to produce homo-

458 J. Chin. Chem. Soc., Vol. 49, No. 4, 2002
Kusumoto et al.
Brade, H., Brade, L.; Freudenberg. M.; Schade, U.; Imoto,
M.; Yoshimura, H.; Kusumoto, S.; Shiba T. Eur. J. Biochem.
1985, 148, 1.
geneous and definite compounds including specifically la-
beled ones. Highly pure such preparations which are often
never available from natural sources are of crucial impor-
tance for determination of the active entities and also for the
study of action mechanisms. Synthetic chemists can thereby
take great pleasure in solving various problems and of course
in having the final success. In that way the frontier of chemis-
try may be ever expanding.
6. (a) Imoto, M.; Yoshimura, H.; Yamamoto, M.; Shimamoto,
T.; Kusumoto, S.; Shiba, T. Bull. Chem. Soc. Jpn. 1987, 60,
2197. (b) Kotani, S.; Takada, H.; Takahashi, I.; Tsujimoto,
M.; Ogawa, T.; Ikeda, T.; Harada, K.; Okamura, T.; Tamura,
S.; Tanaka, T.; Shiba, T.; Kusumoto, S.; Imoto, M.;
Yoshimura, H.; Kasai, N. Infect. Immun. 1986, 52, 872. (c)
Rietschel, E. Th.; Brade, L.; Schade, U.; Galanos, C.;
Freudenberg, M.; Lüderitz, O.; Kusumoto, S.; Shiba, T. Eur.
J. Biochem. 1987, 169, 27.
ACKNOWLEDGEMENT
7. Fukase, K.; Liu, W.-C.; Suda, Y.; Oikawa, M.; Wada, A.;
Mori, S.; Ulmer, A. J.; Rietschel, E. Th.; Kusumoto, S. Tet-
rahedron Lett. 1995, 36, 7455.
8. Liu, W.-C.; Oikawa. M.; Fukase, K.; Suda, Y.; Hendig, W.;
Mori, S.; Hashimoto, M.; Kusumoto, S. Bull. Chem. Soc.
Jpn. 1997, 70, 1441.
This work was partly supported by the Research for the
Future Program (No. 97L00502) from the Japan Society for
the Promotion of Science and by the Grand-in-Aid for Scien-
tific Research for Priority Areas (No. 06240105) from the
Ministry of Education, Science, Sports and Culture, Japan.
9. (a) Flad, H.-D.; Loppnow, H.; Feist, W.; Wang, M.-H.;
Brade, H.; Kusumoto, S.; Rietschel, E. Th.; Ulmer, A. J.
Lymphokine Res. 1989, 8, 235. (b) Wang, M.-H.; Feist, W.;
Herzbeck, H.; Brade, H.; Kusumoto, S.; Rietschel, E. Th.;
Flad, H.-D.; Ulmer, A. J. FEMS Microbiol. Immunol. 1990,
64, 179. (c) Wang, M.-H.; Flad, H.-D.; Feist, W.; Musehold,
J.; Kusumoto, S.; Brade, H.; Gerdesm, J.; Rietschel, E. Th.;
Ulmer, A. J. Lymphokine Cytok. Res. 1992, 11, 23.
10. (a) Kusama, T.; Soga, T.; Shioya E.; Nakayama, K.;
Nakajima, H.; Osada, Y.; Ono, Y.; Kusumoto, S.; Shiba, T.
Chem. Pharm. Bull. 1990, 38, 3366. (b) Kusama, T.; Soga,
T.; Ono, Y.; Kumazawa, E.; Shioya, E.; Osada, Y.;
Kusumoto, S.; Shiba, T. Chem. Pharm. Bull. 1991, 39, 1994.
(c) Liu, W.-C.; Oikawa, M.; Fukase, K.; Suda, Y.; Kusumoto,
S. Bull. Chem. Soc. Jpn. 1999, 72, 1377.
11. Fukase, K.; Kinoshita, I.; Suda, Y.; Aoki, Y.; Liu, W.-C.;
Oikawa, M.; Kurosawa, M.; Zähringer, U.; Seydel, U.;
Rietschel, E. Th.; Kusumoto, S. Synlett 1996, 252.
12. Fukase, K.; Kirikae, T.; Kirikae, F.; Liu, W.-C.; Oikawa, M.;
Suda, M.; Kurosawa, M.; Fukase, Y.; Yoshizaki, H.;
Kusumoto, S. Bull. Chem. Soc. Jpn. 2001, 74, 2189.
13. Yoshizaki, H.; Fukuda. N.; Sato, K.; Oikawa, M.; Fukase, K.;
Suda, Y.; Kusumoto, S. Angew. Chem. Int. Ed. 2001, 40,
1475.
Received February 22, 2002.
Key Words
Synthesis; Glycoconjugates; Lipid A;
Lipopolysaccharide; Biological activity;
Cytokines; Immunostimulation; Innate immunity.
REFERENCES
1. Kusumoto, S.; Tarumi, Y.; Ikenaka, K.; Shiba, T. Bull.
Chem. Soc. Jpn. 1976, 49, 533 and references therein.
2. (a) Rietschel, E. Th.; Brade, H. Scientific American 1992,
267, 54. (b) Rietschel, E. Th.; Brade, L.; Lindner, B.;
Zähringer, U. Bacterial Endotoxic Lipopolysaccharides;
Morrison, D. C.; Ryan, J. L., Eds.; CRC Press, 1992; Vol. 1, p
3.
3. Imoto, M.; Kusumoto, S.; Shiba, T.; Rietschel, E. Th.;
Galanos, C.; Lüderitz, O. Tetrahedron Lett. 1985, 26, 907.
4. (a) Imoto, M.; Yoshimura, H.; Sakaguchi, N.; Kusumoto, S.;
Shiba, T. Tetrahedron Lett. 1985, 26, 1545. (b) Imoto, M.;
Yoshimura, H.; Sakaguchi, N.; Kusumoto, S.; Shiba, T. Bull.
Chem. Soc. Jpn. 1987, 60, 2205.
14. Lien, E.; Means, K.; Heine, H.; Yoshimura, A.; Kusumoto,
S.; Fukase, K., Fenton, M. J.; Oikawa, M.; Qureshi, N.;
Monks, B.; Finberg, R. W.; Ingalls, R. R.; Golenbock, D. T.
J. Clinic. Invest. 2000, 105, 497.
5. Galanos, C.; Lüderitz, O.; Rietschel, E. Th.; Westphal, O.;