Lipid A structures containing novel lipid moieties: Synthesis and adjuvant properties

Article abstract of DOI:10.1016/S0960-894X(02)00362-1

Structurally well-defined immune stimulatory molecules are important components of new generation molecular vaccines. In this paper, the design and synthesis of two lipid A analogues containing an unnatural tri-lipid acyl group are described. In a totally synthetic liposomal vaccine system, these re-designed lipid A analogues demonstrate potent immune stimulatory properties including antigen specific T-cell activation.

Full text of DOI:10.1016/S0960-894X(02)00362-1

Bioorganic & Medicinal Chemistry Letters 12 (2002) 2193–2196
Lipid A Structures Containing Novel Lipid Moieties:
Synthesis and Adjuvant Properties
Zi-Hua Jiang, Mimi V. Bach, Wladyslaw A. Budzynski, Mark J. Krantz,
R. Rao Koganty*and B. Michael Longenecker
Biomira Inc., 2011-94 Street, Edmonton, Alberta, Canada T6N 1H1
Received 22 March 2002; accepted 24 April 2002
Abstract—Structurally well-defined immune stimulatory molecules are important components of new generation molecular vac-
cines. In this paper, the design and synthesis of two lipid A analogues containing an unnatural tri-lipid acyl group are described. In
a totally synthetic liposomal vaccine system, these re-designed lipid A analogues demonstrate potent immune stimulatory properties
including antigen specific T-cell activation. # 2002 Elsevier Science Ltd. All rights reserved.
With a steady growth in research towards molecular
vaccines for various applications, both in therapy and
prophylaxis, a significant attention is drawn towards
structurally well defined immune stimulatory adjuvants.1
However, the structural requirements for an adjuvant
and the molecular and cellular mechanisms underlying
their activities in vivo remain poorly understood.2
Nevertheless, the discovery of mammalian Toll-like
receptors (TLRs)3 and their essential role in innate and
adaptive immune responses4 has led to a better under-
standing of adjuvant actions. Recent identification of a
broad spectrum of specific TLR binding structural fea-
tures4 of pathogenic origin has expanded the repertoire
of adjuvants and their structural definition far beyond
the traditional secondary role of the crude adjuvant
preparations based on bacterial cell wall, such as
Freunds’ complete adjuvant (FCA).5
tion of lipid A, which in turn is defined mainly by the
number of charges, the number of acyl chains, and the
distribution and degree of un-saturation in the lipid
chains of the molecule. Furthermore, lipid A has been
suggested to be a ligand for TLR4,11,12 the Toll-like
receptor involved in the mediation of immune responses
to LPS/lipid A. Thus, structural mimics of natural lipid
A can function as potential ligands for TLR4 and display
agonistic or antagonistic activity to bacterial LPS. In this
paper we report the synthesis and preliminary biological
evaluations of two lipid A analogues, 1 and 2 (Fig. 1),
which are designed primarily as immune stimulatory
adjuvants.
Recently, Johnson et al.13 reported the syntheses of a
group of 3-O-desacyl monophosphoryl lipid A derivatives
with different acyl chain lengths. Their studies suggest that
fatty acid structure may be more important than the
presence or absence of the 1-phosphate group in deter-
mining the biological activity of lipid A molecules.
Compounds 1 and 2 are monophosphorylated analogues,
incorporating an unnatural tri-lipid group incorporating
an ether linkage.14 Although 3-(R)-hydroxy myristic
acid 3 (Fig. 2) is the most commonly found fatty acid in
natural lipid A, lipid acids such as 6, derived from two
molecules of 3, are not known in natural lipid A struc-
tures. It is generally believed that subtle modifications
to the lipid chains of lipid A may affect hydrophobic
balance and induce conformational changes that may
alter cellular uptake and expression of endotoxin activities
of lipid A.9,10,15 In an effort to determine whether the
immune system, or perhaps TLR4 itself, is tolerant
Lipopolysaccharide (LPS) preparations from Gram-
negative bacteria are effective in activating innate
immunity of the host following bacterial infection. Lipid
A, the lipid anchor of LPS, has been shown to be
responsible for the biological activities of LPS in most
in vitro and in vivo test systems. Structure–activity
relationship of lipid A has been a subject of wide
research interest over the last two decades.6,7 Recently,
Seydel et al.8ꢀ10 showed that the agonistic and antagonist
activity of lipid A are governed by the intrinsic conforma-
*Corresponding author. Tel.: +1-780-490-2816; fax: +1-780-988-
5963; e-mail: rkoganty@biomira.com
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00362-1

2194
Z.-H. Jiang et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2193–2196
Scheme 1. (a) Dodecyl triflate, K2CO3, ClCH2CH2Cl, 68%; (b) Zn,
HOAc, 86%; (c) DCC, DMAP, CH2Cl2, 81%; (d) KMnO4, hexane–
HOAc, 0 ꢁC, 61%.
Figure 1. Designed lipid A analogue 1 and 2 containing unnatural tri-
lipid moiety.
.
Scheme 2. (a) 5, EDCI, DMAP, CH2Cl2, 92%; (b) BH3 Me2NH,
.
BF3 OEt2, CH3CN, 72%; (c) (i) (BnO)2PN(iPr)2, tetrazole, CH2Cl2;
(ii) m-CPBA, CH2Cl2, 0 ꢁC, 72% (two steps); (d) (i) [bis(methyldiphe-
nylphosphine)](1,5-cyclooctadiene) iridium(I) hexafluorophosphate,
THF; (ii) NBS, THF–H2O, 79% (two steps); (e) CCl3CN, DBU,
CH2Cl2, 61%.
Figure 2. Structures of natural (3) and unnatural (5–7) lipid acids.
towards substantial structural modifications in the lipid
chains of lipid A, we designed an unnatural tri-lipid acid
7 (Fig. 2) as one lipid component of lipid A structures.
The synthesis of both lipid acids 5 and 7 is described in
Scheme 1. Reaction of the readily available 816 with
dodecyl triflate, followed by the removal of phenacyl
protecting group afforded 5. Further, 5 was coupled
with 917 to give the tri-lipid 10, which upon KMnO4
oxidation furnished the tri-lipid acid 7 in good overall
yield.
Well-established synthetic strategies for lipid
A
molecules13,18ꢀ21 mainly differ in the choice of protect-
ing group(s) and glycosylation donors in constructing
the b-1,6-di-glucosamine unit. Here, we choose benzyl
group as the global protecting group and the tri-
chloroacetimidate as the glycosylation donor. Scheme 2
describes the preparation of glycosyl imidate 14. The
introduction of lipid 5 to the 3-O-position of mono-
saccharide 1122 gave 12. Regio-selective reductive ring
opening of the benzylidene group in 12, followed by
dibenzylphosphite introduction and oxidation gave 13,
which was then converted to the imidate 14 after allyl
group removal. The glycosyl acceptor 18 was prepared
starting from the readily accessible 1523 (Scheme 3),
through the following sequence of reactions: benzyla-
tion (!16), removal of isopropylidene group (!17),
Scheme 3. (a) BnOC(NH)CCl3, CF3SO3H, hexane–CH2Cl2, 0 ꢁC,
56%; (b) HOAc–H2O, 45 ꢁC, 95%; (c) Zn, HOAc, 95%; (d) 7, DCC,
CH2Cl2, 38%.
removal of the Troc-group and introduction of the tri-
lipid acid (!18).
Glycosylation reaction between 14 and 18 afforded the
disaccharide 1924 in moderate yield (Scheme 4). After
the removal of the Troc-group, two different lipid acyl
groups (4 and 5, Fig. 2) were introduced at 20-amine to
give 20 and 21, which upon catalytic hydrogenolytic

Z.-H. Jiang et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2193–2196
2195
containing one phosphoryl group and six acyl chains
with a symmetric distribution (3/3), is expected to have
a small tilt angle and a cylindrical shape, the molecular
conformation that tends to exhibit low or no agonistic
activity, but probable antagonistic activity. Unexpectedly,
our result shows that 1 has strong agonistic activity of lipid
A. Whether the tri-lipid moiety in 1 differs structurally
from conventional three lipid chains (2+1) and has
different impact on the whole molecular shape of lipid A
need further investigation.
References and Notes
1. Singh, M.; O’Hagan, D. Nat. Biotech. 1999, 17, 1075.
2. Schijns, V. E. J. C. Crit. Rev. Immunol. 2001, 21, 75.
3. Heldwein, K. A.; Golenbock, D. T.; Fenton, M. J. Mod.
Asp. Immunobiol. 2001, 1, 249.
4. (a) Krutzik, S. R.; Sieling, P. A.; Modlin, R. L. Curr. Opin.
Immunol. 2001, 13, 104. (b) Akira, S.; Takeda, K.; Kaisho, T.
Nat. Immunol. 2001, 2, 675.
5. Freunds, J.; Casals, J.; Hosmer, E. P. Proc. Soc. Exp. Biol.
Med. 1937, 37, 509.
.
Scheme 4. (a) BF3 OEt2, CH2Cl2, 56%; (b) Zn, HOAc, 97%; (c) 4,
DCC, CH2Cl2, 76%; (d) 5, DCC, CH2Cl2, 75%; (e) Pd/C, H2, THF–
HOAc, 87% for 1 and 2.
6. Rietschel, E. T.; Brade, H.; Holst, O.; Brade, L.; Muller-
Loennies, S.; Mamat, U.; Zahringer, U.; Beckmann, F.; Sey-
del, U.; Brandenburg, K.; Ulmer, A. J.; Mattern, T.; Heine,
H.; Schletter, J.; Loppnow, H.; Schonbeck, U.; Flad, H.-D.;
Hauschildt, S.; Schade, U. F.; Di Padova, F.; Kusumoto, S.;
Schumann, R. R. Curr. Top. Microbiol. Immunol. 1996, 216, 39.
7. Takada, H.; Kotani, S. CRC Critic. Rev. Microbiol. 1989,
16, 477.
8. Schromm, A. B.; Brandenburg, K.; Loppnow, H.; Zahrin-
ger, U.; Rietschel, E. T.; Carroll, S. F.; Koch, M. H.; Kusu-
moto, S.; Seydel, U. J. Immunol. 1998, 161, 5464.
9. Seydel, U.; Oikawa, M.; Fukase, K.; Kusumoto, S.; Bren-
denburg, K. Eur. J. Biochem. 2000, 267, 3032.
de-benzylation afforded the designed lipid A analogue 1
and 2,25 respectively.
Compounds 1 and 2 were tested for immune stimulatory
(adjuvant) properties in terms of their ability to induce
antigen specific T-cell proliferation (blastogenesis) and
secretion of cytokine interferon-gamma (IFN-g) (Fig. 3).
Liposomal vaccine formulation containing MUC1-
derived peptide BLP2526 as an antigen and lipid A as an
adjuvant was used to immunize mice.27 Preliminary test
results show that both 1 and 2 induce the same magni-
tude of IFN-g level (rg/mL) and T-cell proliferation
(CPM) as the natural Lipid A28 preparation, indicating
that 1 and 2 are potent immune stimulatory agents for
antigen specific T-cell activation. These data show that
murine immune system, probably mediated by TLR4,
exhibits greater tolerance to structural modifications of
lipid chains in lipid A than previously thought.
10. Schromm, A. B.; Brandenburg, K.; Loppnow, H.; Moran,
A. P.; Koch, M. H. J.; Rietschel, E. T.; Seydel, U. Eur. J.
Biochem. 2000, 267, 2008.
11. Yoshida, K.; Ono, M.; Sawada, H. J. Endotoxin Res.
1999, 5, 127.
12. Kutuzova, G. D.; Albrecht, R. M.; Erickson, C. M.; Qur-
eshi, N. J. Immunol. 2001, 167, 482.
13. Johnson, D. A.; Keegan, D. S.; Sowell, C. G.; Livesay,
M. T.; Johnson, C. L.; Taubner, L. M.; Harris, A.; Myers,
K. R.; Thompson, J. D.; Gustafson, G. L.; Rhodes, M. J.;
Ulrich, J. T.; Ward, J. R.; Yorgensen, Y. M.; Cantrell, J. L.;
Brookshire, V. G. J. Med. Chem. 1999, 42, 4640.
The fact that 1 is strongly active as an immunostimulant
is somewhat surprising in view of the observation by
Seydel et al.8ꢀ10 that biological activities of lipid A are
determined by the shape of the molecule. Analogue 1,
14. (a) Installing an ether linkage instead of natural ester
linkage in lipid chains of lipid A is to improve the stability of
the molecule. For examples of synthetic lipid A analogues with
lipid chains containing an ether linkage, please see: (a) Christ,
W. J.; Asano, O.; Robidoux, A. L. C.; Perez, M.; Wang, Y.;
Dubuc, G. R.; Gavin, W. E.; Hawkins, L. D.; McGuinness,
P. D.; Mullarkey, M.; Lewis, M. D.; Kishi, Y.; Kawata, T.;
Bristol, J. R.; Rose, J. R.; Rossignol, D. P.; Kobayashi, S.;
Hishinuma, I.; Kimura, A.; Asakawa, N.; Karayama, K.;
Yamatsu, I. Science 1995, 268, 80. (b) Watanabe, Y.; Mochi-
zuki, T.; Shiozaki, M.; Kanai, S.; Kurakata, S.-I.; Nishijima,
M. Carbohydr. Res. 2001, 333, 203.
15. Rietschel, E. T.; Kirikae, T.; Schade, F. U.; Ulmer, A. J.;
Holst, O.; Brade, H.; Schmidt, G.; Mamat, U.; Grimmecke, H.-D.;
Kusumoto, S.; Zahringer, U. Immunobiology 1993, 187, 169.
16. Kiso, M.; Tanaka, S.; Fujita, M.; Fujishima, Y.; Ogawa,
Y.; Hasegawa, A. Carbohydr. Res. 1987, 162, 247.
17. Hanessian, S.; Tehim, A.; Chen, P. J. Org. Chem. 1995,
58, 7768.
Figure 3. Adjuvant properties of synthetic lipid A analogue 1 and 2:
antigen specific T-cell proliferation (blastogenesis) (CPM, counts per
min) and interferon-gamma (IFN-g, pg/mL) production in mice
immunized with MUC1-based liposomal vaccine adjuvanted with lipid
A analogue. Lipid A: natural lipid A product obtained from Salmo-
nella minnesota, R595.

2196
Z.-H. Jiang et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2193–2196
18. Christ, W. J.; McGuinness, P. D.; Asano, O.; Wang, Y.;
Mullarkey, M. A.; Perez, M.; Hawkins, L. D.; Blythe, T. A.;
Dubuc, G. R.; Robidoux, A. L. J. Am. Chem. Soc. 1994, 116,
3637.
19. Yoshizaki, H.; Fukuda, N.; Sato, K.; Oikawa, M.;
Fukase, K.; Suda, Y.; Kusumoto, S. Angew. Chem., Int. Ed.
2001, 40, 1475.
confirmed by MS. 1: Rf 0.20 (chloroform/methanol/water/
ammonium hydroxide, 7:3:0.4:0.2). [a]2D0=ꢀ3.5 (c 0.2, chloro-
form/methanol, 4:1). ESI-MS calcd for C92H177N2O20P:
1661.3; found (negative mode): 1660.3 (87, MꢀH), 1661.3
(100, MꢀH, 13C isotope peak). 2: Rf 0.38 (chloroform/metha-
nol/water/ammonium hydroxide, 7:3:0.4:0.2). [a]2D0=+2.0 (c
0.2, chloroform/methanol, 4:1). ESI-MS calcd for
20. Fukase, K.; Fukase, Y.; Oikawa, M.; Liu, W.-C.; Suda,
Y.; Kusumoto, S. Tetrahedron 1998, 54, 4033.
21. Griffiths, S. L.; Madsen, R.; Fraser-Reid, B. J. Org. Chem.
1997, 62, 3654.
C104H201N2O20P: 1829.4; found (negative mode): 1828.5 (80,
MꢀH), 1829.5 (100, MꢀH, 13C isotope peak).
26. Palmer, M.; Parker, J.; Modi, S.; Butts, C.; Smylie, M.;
Meikle, A.; Kehoe, M.; MacLean, G.; Longenecker, B. M.
Clin. Lung Cancer 2001, 3, 49.
22. Schule, G.; Ziegler, T. Liebigs Ann. 1996, 1599.
23. Schule, G.; Ziegler, T. Tetrahedron 1996, 52, 2925.
24. The regio-selectivity of the glycosylation at 6-O-position
and its b-linkage are asserted based on the result of other
similar glycosylation reaction carried out in this laboratory.
The anomeric protons of the disaccharide are recognizable
and its acetylated derivative confirms the 6-O-linkage. 19: Rf
0.33 (hexane/ethyl acetate, 2:1). [a]2D0=+24.0 (c 0.3, chloro-
form). 1H NMR (500 MHz, CDCl3) d 0.88 (t, J=6.5 Hz, 15H,
5 CH3), 1.25 (m, 90H), 1.40–1.58 (m, 10H), 2.25–2.45 (m, 6H),
2.95 (br s, 1H, OH), 3.21 (m, 1H), 3.36 (m, 3H), 3.47–3.53 (m,
2H), 3.57–3.67 (m, 4H), 3.72 (m, 2H), 3.80 (m, 2H), 4.03 (br d,
J=11.0 Hz, 1H), 4.25 (ddd, J=10.0, 9.0, 3 5 Hz, 1H, H-2),
4.44 (d, J=11.5 Hz, 1H), 4.48 (m, 3H), 4.59 (m, 2H), 4.67 (d,
J=11.5 Hz, 1H), 4.69 (d, J=11.5 Hz, 1H), 4.75 (d, J=11.5
Hz, 1H), 4.85–4.92 (m, 6H), 5.04 (m, 1H), 5.36 (m, 2H), 5.80
(d, J=9.0 Hz, 1H, NH), 7.30 (m, 25H, Ar–H).
27. Mice were immunized subcutaneously with a single injec-
tion of liposomal vaccine formulation containing MUC1-
derivatived lipopeptide antigen BLP25 (40 mg per dose),
H2N–STAPPAHGVTSAPDTRPAPGSTAPPK(palmitoyl)G–
OH, and a lipid A analogue (20 mg per dose), either synthetic
or natural, as an adjuvant. After 9 days, the lymphocytes were
taken from the draining lymph nodes of immunized mice and
incubated in in vitro culture in the presence of the same
boosting antigen BLP25 and antigen presenting cells (APC).
3
T-cell proliferation was evaluated using a standard H thymi-
dine incorporation assay and IFN-g level was determined in
the cell culture supernatants using enzyme-linked immu-
noabsorbent assay (ELISA). Both CPM and IFN-g data
reported in Figure 3 were averaged from sextet experimental
measurements.
28. The natural Lipid A product isolated from Salmonella
minnesota, R595, was purchased from Avanti Polar Lipids,
Inc. For structures of S. minnesota R595 lipid A, please see:
Qureshi, N.; Mascagni, P.; Ribi, E.; Takayama, K. J. Biol.
Chem. 1985, 260, 5271.
1
25. The H NMR spectra of fully de-protected lipid A analo-
gue 1 and 2 are poorly resolved in various solvents tested,
indicating the aggregation of these molecules in solution. The
purity of 1 and 2 were checked by HP-TLC and their identity