In vivo protection provided by a synthetic new alpha-galactosyl ceramide analog against bacterial and viral infections in murine models

Article abstract of DOI:10.1128/AAC.00368-10

Alpha-galactosyl ceramide (α-GalCer) has been known to bind to the CD1d receptor on dendritic cells and activate invariant natural killer T (iNKT) cells, which subsequently secrete T-helper-cell 1 (Th1) and Th2 cytokines, which correlate with anti-infecti

Full text of DOI:10.1128/AAC.00368-10

ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Oct. 2010, p. 4129–4136
0066-4804/10/$12.00 doi:10.1128/AAC.00368-10
Copyright © 2010, American Society for Microbiology. All Rights Reserved.
Vol. 54, No. 10
In Vivo Protection Provided by a Synthetic New Alpha-Galactosyl
Ceramide Analog against Bacterial and Viral Infections
in Murine Modelsᰔ
Kun-Hsien Lin,1,2† Jian-Jong Liang,3† Wen-I Huang,1 Shao-Ying Lin-Chu,1 Ching-Yao Su,1,4
Yi-Ling Lee,3 Jia-Tsong Jan,1 Yi-Ling Lin,1,3* Yih-Shyun E. Cheng,1
and Chi-Huey Wong1,2,4
*
Genomics Research Center1 and Institute of Biomedical Sciences,3 Academia Sinica, Taipei 115, Taiwan, and Department of
Chemistry2 and Institute of Biochemical Sciences,4 National Taiwan University, Taipei 106, Taiwan
Received 18 March 2010/Returned for modification 10 May 2010/Accepted 14 July 2010
Alpha-galactosyl ceramide (␣-GalCer) has been known to bind to the CD1d receptor on dendritic cells and
activate invariant natural killer T (iNKT) cells, which subsequently secrete T-helper-cell 1 (Th1) and Th2
cytokines, which correlate with anti-infection activity and the prevention of autoimmune diseases, respectively.
␣-GalCer elicits the secretion of these two cytokines nonselectively, and thus, its effectiveness is limited by the
opposing effects of the Th1 and Th2 cytokines. Reported here is the synthesis of a new ␣-GalCer analog
(compound C34), based on the structure of CD1d, with a 4-(4-fluorophenoxy) phenyl undecanoyl modification
of the N-acyl moiety of ␣-GalCer. Using several murine bacterial and viral infection models, we demonstrated
that C34 has superior antibacterial and antiviral activities in comparison with those of several other Th1-
selective glycolipids and that it is most effective by administering it to mice in a prophylactic manner before
or shortly after infection.
Natural killer T (NKT) cells contribute to a variety of im-
munological processes through the recognition of NKT cell
receptors by lipid and glycolipid antigens presented by CD1d
molecules (2, 35). CD1d molecules are major histocompatibil-
ity complex class I-like proteins and are expressed by most
monocytes, macrophages, dendritic cells, and B cells as well as
by nonlymphoid cells. CD1d presents glycolipid antigens to
CD1d-restricted T cells (or NKT cells), which are implicated in
the host innate defense system through the production of T-
helper-cell 1 (Th1) and Th2 types of cytokines, such as gamma
interferon (IFN-␥) and interleukin-4 (IL-4), respectively (32,
39). The production of Th1 cytokines is shown to correlate with
the antitumor, antibacterial, and antiviral effects of glycolipids,
while the ability to induce the Th2 cytokines is thought to
correlate with the amelioration of certain autoimmune dis-
eases, such as type 1 diabetes (9, 10, 28).
Among the variety of ligands that bind to CD1d, the most
well studied one is alpha-galactosyl ceramide (␣-GalCer) (7,
17, 30), which is a synthetic, structurally optimized com-
pound of the natural product lead identified from the ma-
rine sponge (Agelas mauritianus) as agelasphin-9b (24). The
synthetic analog (2S,3S,4R)-1-O-(␣-D-galactopyranosyl)-2-
(N-hexacosanoylamino)-1,3,4-octadecanetriol was made as
KRN7000 at Kirin Brewery, Japan (23), and has generally been
referred to as ␣-GalCer. Mice treated with ␣-GalCer were
shown to be protected against a variety of infections (13, 15, 16,
22, 33). However, the effectiveness of ␣-GalCer is limited by
the opposing effects of the Th1 and Th2 cytokines induced by
this glycolipid. In order to develop selective Th1 or Th2 acti-
vators, many ␣-GalCer analogs were synthesized (1, 3, 5, 8, 12,
14, 21, 34, 36), and their immune-modulating activities were
shown to be related to the affinity of binding to CD1d (8).
Recent studies using a CD1d array established that the secre-
tion of IFN-␥ and IL-4 by NKT cells is determined by the
binding of ␣-GalCer analogs to CD1d (21), and a higher af-
finity for CD1d would shift the cytokine release profile toward
a stronger Th1 response (3, 21).
This report described the synthesis of the 4-(4-fluorophe-
noxy)phenyl undecanoyl derivative of ␣-GalCer as a selective
Th1 activator. This analog was designed on the basis of the
structure of CD1d, where the N-acyl group of the glycolipid
interacts most favorably with the lipid binding moiety of CD1d
(36). Its efficacies for protection against both bacterial and
viral infections were compared with those of ␣-GalCer and
several other analogs using murine models of infection. Our
study found that among the synthetic ␣-GalCer analogs tested,
compound C34 is the most effective in offering protection
against all of the bacterial and viral infection models tested,
especially when it is given in a prophylactic manner or shortly
after infection.
MATERIALS AND METHODS
* Corresponding author. Mailing address: Genomics Research Cen-
ter, Academia Sinica, 128 Academia Road, Section 2, Nankang Dis-
trict, Taipei 115, Taiwan. Phone for Chi-Huey Wong: 886-2-2789-9400.
Fax: 886-2-2785-3852. E-mail: chwong@gate.sinica.edu.tw. Phone for
Yi-Ling Lin: 886-2-2652-3902. Fax: 886-2-2785-8847. E-mail: yll@ibms
.sinica.edu.tw.
Reagents and glycolipid testing agents. All the reagents were commercial
reagent grade and were used without further purification, unless otherwise spec-
ified. All solvents were anhydrous grade. CH2Cl2 and methanol (MeOH) were
purchased from Acros. Anhydrous CHCl3 was purchased from Merck. Molecular
sieves (4 Å) for glycosylation were purchased from Acros and activated by flame.
Reactions were monitored by analytical thin-layer chromatography (TLC) on
Merck silica gel 60 F254 plates and were visualized under UV light (254 nm)
† These authors contributed equally to this work.
ᰔ Published ahead of print on 26 July 2010.
4129

4130
LIN ET AL.
ANTIMICROB. AGENTS CHEMOTHER.
used in this study were previously reported (3, 8, 21, 36). Compound C34 is a new
␣-GalCer analog; its synthesis is described in this report (Fig. 1B). Glycolipid
stock solutions were prepared as 1-mg/ml dimethyl sulfoxide (DMSO) solutions.
The test samples were diluted with phosphate-buffered saline (PBS) to 10 to 30
␮g/ml and were used for the animal studies.
Synthetic procedures for compound 3. To a solution of galactosyl donor 1 (5.8
g, 11.8 mmol) (18), dimethyl sulfide (1.1 ml, 15.6 mmol), 4-Å molecular sieve (1
g), and 2-chloropyridine (3.6 ml, 39 mmol) in anhydrous CH2Cl2 (30 ml) under
Ar at Ϫ45°C was added trifluoromethanesulfonic anhydride (2 ml, 11.9 mmol).
The reaction mixture was stirred for 20 min at 0°C and 20 min at room temper-
ature. Galactosyl acceptor 2 (6) in CH2Cl2 (10 ml) was slowly added, and the
reaction mixture was stirred at room temperature for 16 h and then filtered. The
organic layer was washed with water and brine, dried over MgSO4, and concen-
trated under reduced pressure. The residue was chromatographed on silica gel
(ethyl acetate-hexanes [EA-Hex], 1:20 to 1:15) to give the product (compound 3)
as a yellow oil (4 g, 60%). 1H NMR (600 MHz, CDCl3) ␦ 7.37 to 7.24 (m, 15H),
4.95 (d, J ϭ 11.5 Hz, 1H), 4.91 (d, J ϭ 3.6 Hz, 1H), 4.86 (d, J ϭ 11.5 Hz, 1H), 4.77
(d, J ϭ 11.5 Hz, 1H), 4.71 (d, J ϭ 11.5 Hz, 1H), 4.67 (d, J ϭ 11.5 Hz, 1H), 4.59
(d, J ϭ 11.5 Hz, 1H), 4.11 (dd, J ϭ 10.7 Hz, 7.6 Hz, 1H), 4.08 (m, J ϭ 4.4 Hz, 1H),
4.05 (dd, J ϭ 10.1 Hz, 3.6 Hz, 1H), 4.02 (dd, J ϭ 10.7 Hz, 3.5 Hz, 1H), 4.01(dd,
J ϭ 10.5 Hz, 2.4 Hz, 1H), 4.00 (dd, J ϭ 9.2 Hz, 4.4 Hz, 1H), 3.95 (dd, J ϭ 10.1
Hz, 2.7 Hz, 1H), 3.92 (dd, J ϭ 7.6 Hz, 3.5 Hz, 1H), 3.83 (d, J ϭ 2.7 Hz, 1H), 3.69
(dd, J ϭ 10.5 Hz, 6.6 Hz, 1H), 3.40 (ddd, J ϭ 9.2 Hz, 6.6 Hz, 2.4 Hz, 1H), 1.95
(s, 3H), 1.60 (m, 1H), 1.50 (m, 1H), 1.35 (s, 3H), 1.34 to 1.20 (m, 27H), 0.85 (t,
J ϭ 6.8 Hz, 3H). 13C NMR (150 MHz, CDCl3) ␦ 170.79, 138.95 ϫ 2, 138.37,
128.67 ϫ 2, 128.58 ϫ 2, 128.55 ϫ 2, 128.47 ϫ 2, 128.00, 127.83 ϫ 2, 127.80 ϫ 2,
127.74, 127.70, 108.39, 98.96, 78.77, 77.93, 76.71, 75.46, 75.02, 74.73, 73.83, 73.15,
69.83, 69.13, 63.91, 59.93, 32.13, 29.90 ϫ 3, 29.87 ϫ 2, 29.83, 29.81, 29.77, 29.57,
29.52, 28.40, 26.73, 25.91, 22.90, 21.03, 14.34. High-resolution mass spectrometry
(HRMS; electrospray ionization–time of flight [ESI-TOF]) for C50H71N3O9Naϩ
[M ϩ Na]ϩ: calculated, 880.5083; found, 880.5064.
Synthesis of compound 4. (10-Carboxynonyl)triphenylphosphonium bromide
(4.8 g, 19.1 mmol) (29) was dissolved in 186 ml of tetrahydrofuran (THF). The
solution was cooled to 0 to 5°C and was added dropwise to a solution of lithium
bis(trimethylsilyl)amide (LHMDS; 1 M in THF-ethylbenzene, 38 mmol) to pro-
duce an orange ylide within 1 h at 0 to 5°C. 4-(4-Fluorophenoxy)benzaldehyde (2
g, 9.25 mmol), dissolved in 18 ml of THF, was added dropwise, and the solution
was stirred for 4 h at room temperature. Afterwards, the reaction was quenched
with methanol and concentrated. The residue was extracted with diethyl ether,
water, and brine and then dried over MgSO4. After removal of the solvent, the
mixture was chromatographed on silica gel (EA-Hex, 1:5 to 1:2) to give the
unsaturated fatty acid.
The saturated fatty acid was prepared by catalytic hydrogenation in 50 ml of
methanol containing 10 mol% of 10% palladium on charcoal (Pd/C). The reac-
tion mixture was stirred under H2 (1 atm) at room temperature for 1 day. The
hydrogenated product was filtered through Celite 545 to remove the catalyst. The
resulting solution was concentrated and chromatographed on silica gel (EA-Hex,
1:4 to 1:2) to give the product (compound 4) as a yellow solid (2 g, 58%, two
steps). 1H NMR (600 MHz, MeOD) ␦ 7.16 (m, 2H), 7.06 (m, 2H), 6.96 (m, 2H),
6.87 (m, 2H), 2.59 (t, J ϭ 7.6 Hz, 2H), 2.27 (m, J ϭ 7.6 Hz, 2H), 1.59 (m, 4H),
1.32 (m, 12H). 13C NMR (150 MHz, MeOD) ␦ 177.97, 160.92, 159.33, 156.94,
155.19, 139.34, 130.88 ϫ 2, 121.27, 121.22, 119.61 ϫ 2, 117.34, 117.19, 36.24,
35.11, 32.93, 30.72, 30.66 ϫ 2, 30.50, 30.36, 26.21. HRMS (ESI-TOF) for
FIG. 1. ␣-GalCer analogs used in the present study. (A) Names
and structures of ␣-GalCer analogs used. (B) Reagents and conditions
used for the synthesis of C34 were as follows: a, Me2S, 2-Cl-pyridine,
Tf2O, CH2Cl2, 4-Å molecular sieve, Ϫ45°C, 16 h, 60%; b, tri-
phenylphosphine, pyridine, H2O, 50°C, 9 h; c, 4, EDC, HBTu, Et3N,
CH2Cl2, 88%, two steps; d, NaOMe, MeOH-CH2Cl2 ϭ 1:4, 9 h; e,
AcOH-H2O ϭ 4:1, 70°C, 16 h; f, Pd(OH)2, H2, MeOH-CHCl3 ϭ 4:1,
5 h, 40%, three steps. BnO, benzyl ether; OAc, acetate.
C
23H29FO3Naϩ [M ϩ Na]ϩ: calculated, 395.1993; found, 395.1998.
Synthesis of compound 5. To a solution of compound 3 (4 g, 4.7 mmol) in
pyridine-H2O (10:1, 187 ml) was added triphenylphosphine (2.5 g, 9.4 mmol).
The mixture was stirred for 16 h at 45°C, concentrated, dissolved with ethyl
acetate, washed with brine, and concentrated again to dryness. The residue was
used for the next step without prior purification.
To a stirred solution of this compound, fatty acid 4 (2.1 g, 5.6 mmol) in 36 ml
of anhydrous CH2Cl2 (36 ml) was added, along with triethylamine (Et3N; 1.3 ml),
1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDC; 1.3 g, 7.1
mmol), and O-benzotriazole-N,N,NЈ,NЈ-tetramethyl-uronium-hexafluoro-phos-
phate (HBTu; 2.7 g, 7.1 mmol). The solution was stirred for 16 h at ambient
temperature and was extracted sequentially with CH2Cl2, water, and brine and
then dried over MgSO4. The mixture was concentrated and chromatographed on
silica gel (EA-Hex, 1:10 to 1:5) to give the product (compound 5) as a white solid
(4.8 g, 88%, two steps). 1H NMR (600 MHz, CDCl3) ␦ 7.40 to 7.23 (m, 15H), 7.10
(d, J ϭ 8.4 Hz, 2H), 6.99 (t, J ϭ 8.8 Hz, 2H), 6.92 (m, 2H), 6.85 (d, J ϭ 8.4 Hz,
2H), 6.01 (d, J ϭ 9.2 Hz, 1H), 4.99 (d, J ϭ 3.6 Hz, 1H), 4.94 (d, J ϭ 11.5 Hz, 1H),
4.81 (d, J ϭ 11.5 Hz, 1H), 4.79 (d, J ϭ 11.5 Hz, 1H), 4.76 (d, J ϭ 11.5 Hz, 1H),
4.66 (d, J ϭ 11.5 Hz, 1H), 4.60 (d, J ϭ 11.5 Hz, 1H), 4.14 to 4.02 (m, J ϭ 3.6 Hz,
5H), 3.97 (m, 1H), 3.90 to 3.86 (m, J ϭ 2.6 Hz, 2.5 Hz, 3H), 3.84 (dd, J ϭ 11.1
and/or by staining with acidic ceric ammonium molybdate or ninhydrin. Flash
column chromatography was performed on a silica gel 60 Geduran column
(particle size, 40 to 63 ␮m; Merck). 1H nuclear magnetic resonance (NMR)
spectra were recorded on a Bruker AV-600 spectrometer (600 MHz) at 20°C.
Chemical shifts (␦ ppm) were assigned according to the standard signals of
CDCl3 (␦ ϭ 7.24 ppm), MeOD (␦ ϭ 3.31 ppm), and pyridine-d5 (␦ ϭ 7.58 ppm).
13C NMR spectra were obtained on a Bruker AV-600 spectrometer (150 MHz)
and were reported on the ␦ ppm scale using the signals of CDCl3 (␦ ϭ 77.23
ppm), MeOD (␦ ϭ 49.15 ppm), and pyridine-d5 (␦ ϭ 150.35 ppm) for calibration.
The structures of all glycolipids and their code names used in this study are
shown in Fig. 1A. The synthesis and some of the properties of the glycolipids

VOL. 54, 2010
ANTI-INFECTION ACTIVITIES OF ␣-GalCer ANALOG
4131
Hz, 2.6 Hz, 1H), 3.62 (dd, J ϭ 11.1 Hz, 2.5 Hz, 1H), 2.54 (t, J ϭ 7.7 Hz, 2H), 2.08
to 1.98 (m, 2H), 1.97 (s, 3H), 1.60 to 1.50 (m, 1H), 1.47 (m, 1H), 1.40 (s, 3H), 1.36
(m, 1H), 1.30 (s, 3H), 1.29 to 1.19 (m, 39H), 0.85 (t, J ϭ 6.7 Hz, 3H). 13C NMR
(150 MHz, CDCl3) ␦ 172.54, 170.83, 159.60, 158.00, 158.58, 153.57, 138.67,
138.43, 138.26, 138.06, 129.77 ϫ 2, 128.67 ϫ 2, 128.65 ϫ 2, 128.60 ϫ 2, 128.58 ϫ
2, 128.19 ϫ 2, 128.12, 128.09, 127.92, 127.72 ϫ 2, 120.30, 120.25, 118.54 ϫ 2,
116.44, 116.29, 108.12, 98.85, 79.08, 77.98, 76.86, 75.86, 74.79, 74.57, 73.80, 73.22,
69.07, 68.80, 64.03, 48.72, 37.04, 35.41, 32.14, 31.86, 29.93 ϫ 4, 29.88 ϫ 2, 29.85 ϫ
2, 29.79, 29.76, 29.67, 29.58 ϫ 2, 29.54, 29.07, 28.34, 26.80, 26.12, 25.90, 22.90,
21.03, 14.34. HRMS (matrix-assisted laser desorption ionization–time of flight)
for C73H100FNO11Naϩ [M ϩ Na]ϩ: calculated, 1,208.7173; found, 1,208.7135.
Synthesis of compound C34. To a solution of 5 (4.8 g, 4 mmol) in cosolvent
MeOH-CH2Cl2 (4:1, 50 ml) was added sodium methoxide (NaOMe; 0.4 mmol).
The solution was stirred overnight at room temperature. The mixture was neu-
tralized with Amberlite IR-120 resin and then filtered. The solution was concen-
trated and used for the next step without prior purification.
The solution in acetyl hydroxide (AcOH)-H2O (4:1, 50 ml) was stirred for 16 h
at 70°C and was concentrated to dryness. The residue was purified by flash
column chromatography on silica gel (EA-Hex-MeOH, 1:1:0.1).
The deacetonide derivative was dissolved in 50 ml of cosolvent MeOH-CHCl3
(4:1, 50 ml) containing palladium hydroxide on carbon (20% Pd, trace), and the
mixture was stirred at room temperature for 1 day in an H2 atmosphere. The
hydrogenated product was filtered through Celite 545 to remove the catalyst, and
the filtrate was concentrated to dryness. The residue was eluted from LH20
(MeOH-CHCl3, 1:1) to give compound C34 (1.3 g, 40%, three steps). 1H NMR
(600 MHz, pyridine-d5) ␦ 8.53 (d, J ϭ 8.6 Hz, 1H), 7.27 (d, J ϭ 8.4 Hz, 2H), 7.14
(t, J ϭ 8.4 Hz, 2H), 7.05 to 7.09 (m, 4H), 5.59 (d, J ϭ 3.8 Hz, 1H), 5.27 (m, 1H),
4.70 to 4.65 (m, 2H), 4.55 (d, J ϭ 2.6 Hz, 1H), 4.52 (t, J ϭ 6.0 Hz, 1H), 4.45 to
4.39 (m, 4H), 4.35 to 4.30 (m, 1H), 2.59 (t, J ϭ 7.6 Hz, 2H), 2.45 (t, 2H), 2.30
(m, 1H), 1.91 (m, 2H), 1.81 (m, 2H), 1.68 (m, 2H), 1.59 (m, 2H), 1.47 to 1.15
(m, 40H), 0.87 (t, J ϭ 6.9 Hz, 3H). 13C NMR (150 MHz, pyridine-d5) ␦ 173.79,
160.19, 158.60, 156.32, 154.53, 138.89, 130.73 ϫ 2, 121.06, 121.01, 119.46 ϫ 2,
117.25, 117.09, 102.00, 77.17, 73.53, 72.96, 72.09, 71.46, 71.29, 70.79, 69.07, 63.12,
51.95, 37.27, 35.86, 34.82, 32.60, 32.47, 30.84, 30.63, 30.48 ϫ 3, 30.41, 30.35, 30.28,
30.25 ϫ 2, 30.20, 30.10, 30.03, 26.99, 26.87, 23.42, 14.77. HRMS (ESI-TOF) for
FIG. 2. Antibacterial efficacies of glycolipids in mice infected with
Sphingomonas capsulata. (A) Infected mice were grouped and received
100 ␮g/kg ␣-GalCer analogs or vehicle control. They were killed at
24 h postinfection to determine the liver bacterial loads, which are
represented by dots. The mean numbers of CFU of each treatment
group are indicated with short horizontal lines. (B) An experiment
similar to that described for panel A was conducted by administering
50 ␮g/kg of selected glycolipids to different groups of mice. Asterisks
indicate groups with results significantly different from those for the
PBS-treated control group: , P Ͻ 0.05; , P Ͻ 0.01.
C
47H76FNO10Naϩ [M ϩ Na]ϩ: calculated, 856.5345; found, 856.5362.
Mice, viruses, and bacteria used. C57BL/6 and BALB/c mice at 6 to 8 weeks
of age were used for the studies. Mice were housed in plastic cages with free
access to food and water and were allowed to acclimate at least 1 week prior to
the start of the experiments. The viruses used in this study are influenza virus
strain WSN (A/WSN/1933/H1N1) and Japanese encephalitis virus (JEV; strain
RP-9), as described previously (20, 27). The bacterial strain Sphingomonas cap-
sulata (ATCC 14666) was obtained from BCRC, Taiwan. Luminescent strain
Staphylococcus aureus Xen29 (19) was purchased from Xenogen Corporation
(CA). The mouse experiments were approved and performed in accordance with
the guidelines of the Academia Sinica Institutional Animal Care and Utilization
Committee. The immunodeficient mice used in this study are as described pre-
viously (20).
Antibacterial efficacy study using Sphingomonas capsulata-infected mice. Six-
to-eight-week old female C57BL/6 mice were injected intraperitoneally (i.p.)
with 5 ϫ 108 Sphingomonas capsulata organisms. The mice were grouped into
treatment and control groups at 5 to 10 mice per group. Three hours after the
infection, the mice in the treatment group were injected intraperitoneally with
the test glycolipids at 50 or 100 ␮g/kg of body weight, and the mice in the control
group were injected with the same volume of PBS. Twenty-four hours after
bacterial infection, mice from all groups were killed. The livers were removed
and homogenized in 0.9% NaCl with 0.02% Tween 80 using tissue homogenizers.
The numbers of CFU of Sphingomonas capsulata in the liver homogenates were
determined by plating diluted samples on nutrient plates. Colonies were counted
after incubation for 48 h at 37°C. The antibacterial results were analyzed using
a one-way analysis of variance (ANOVA) to assess the significance of the dif-
ference between treatment groups, followed by the Tukey posttest.
*
**
The infection was done by i.p. injection of 5 ϫ 105 PFU of the RP-9 strain of JEV
in 500 ␮l PBS, and 30 ␮l PBS was simultaneously injected intracerebrally, as
described previously (4, 20, 38). The survival of the mice was monitored daily.
The survival curves for the glycolipid-treated groups were compared to those for
the control group using Prism software (GraphPad Software, San Diego, CA).
Anti-influenza animal model study. For the evaluation of the in vivo anti-
influenza efficacies of the ␣-GalCer analogs, BALB/c mice were treated with C34
at different times after nasal infection using 10 50% lethal doses (LD50s) of strain
WSN influenza virus. The survival of the mice in all groups was examined daily
for 14 days postinfection. The survival curves for the glycolipid-treated groups
were compared to those for the control group using Prism software (GraphPad
Software).
RESULTS AND DISCUSSION
Antibacterial efficacy study using luminescent Staphylococcus aureus (Xen29)
by image analyses. BALB/c female mice were weighed, grouped, and injected in
the posterior quadricep muscle of the left thigh with 200 ␮l PBS containing 3 ϫ
108 CFU of S. aureus Xen29. At different times after Xen29 infection, the mice
were injected i.p. with vehicle or glycolipid, as indicated. Mice were periodically
imaged using an IVIS in vivo imaging system (Xenogen Corporation), as de-
scribed previously (19). The imaging results, in relative luminescence units, were
analyzed using one-way ANOVA to assess the significance of the difference
between treatment groups, followed by the Tukey posttest.
Synthesis of the new ␣-GalCer analog C34. The structures of
all glycolipids and their code names used in this study are
shown in Fig. 1A. Compounds C1, C3, C9, C11, C14, C16, C17,
C23, and 7DW8-5 were prepared as described previously (3, 8,
21, 36). As shown in Fig. 1B, the synthesis of C34 involved the
coupling of the galactosyl donor (compound 1) with phyto-
sphingosine (compound 2) using trifluoromethanesulfonic an-
hydride (Tf2O) and dimethyl sulfide (Me2S) as the promoter
afforded compound 3 at a 60% yield. The ␣/␤ selectivity was
JEV infection animal model. Seven-week-old C57BL/6 mice were grouped,
and at various times postinfection the JEV-infected mice received glycolipids.

4132
LIN ET AL.
ANTIMICROB. AGENTS CHEMOTHER.
mice were used as the infection model to evaluate the antibac-
terial efficacies of some of the synthetic glycolipids. S. capsulata
is a common environmental bacterial strain that is found in
many places, such as air and water. It can easily be identified
on nutrient agar plates because of its yellow colony color.
Unlike most Gram-negative bacteria, S. capsulata does not
contain lipopolysaccharide (LPS), which activates the host an-
tibacterial activities. The antibacterial activities of glycolipids
in S. capsulata are mediated through the activation of NKT
cells by glycolipid-bound CD1d molecules; thus, the antibac-
terial efficacies reflect the impact of the NKT cell-mediated
pathway. Figure 2A shows that mice in the treatment groups
receiving 100 ␮g/kg of compound C1, C11, C14, or C16 had
significantly lower S. capsulata CFU numbers in the livers than
the control group at 24 h after bacterial infection. In contrast,
the reduction in the numbers of CFU in the groups treated
with compounds C3, C9, and C17 were not significantly differ-
ent from the numbers of CFU found in the control group. The
result of this study suggested that the glycolipids ␣-GalCer
(compound C1) and C16 appeared to be more potent than
compounds C11 and C14 at suppressing S. capsulata in vivo,
though the differences are not significant.
The S. capsulata-infected mice were then used to compare
the antibacterial potency of ␣-GalCer with the potencies of
compound C34 and other analogs, such as compounds C23 and
7DW8-5, that induce high levels of Th1 cytokine secretion
(21). These glycolipids were evaluated at 50 ␮g/kg using the S.
capsulata-infected C57BL/6 mouse model. Figure 2B shows
that treatments with any one of the three glycolipids (com-
pound C34, C23, or 7DW8-5) reduced the bacterial loads sig-
nificantly compared with the loads in the vehicle-treated mice,
with P values being less than 0.01. The mean number of CFU
for the C34-treated mice appeared to be lower than that for the
␣-GalCer-treated group. The present study further confirmed
that glycolipids biased toward the Th1 pathway are more ef-
fective as antibacterial agents than compound C1. However,
differences between these glycolipid-treated groups were not
significant in the present study.
Antibacterial activities of ␣-GalCer analogs in the thigh
wound mouse model using luminescent Staphylococcus aureus
(Xen29). As shown above, intraperitoneal injection of
C57BL/6 mice with S. capsulata results in transient infections,
and treatment with some glycolipids hastens the clearance in
the 24 h after infection. However, in the absence of any treat-
ment, the infections are often cleared in 2 to 3 days. To eval-
uate the antibacterial efficacies of the ␣-GalCer analogs, we
then used a more relevant disease model, the mouse model of
deep thigh wound infection, by infection of mice in the hind
thigh muscle with luminescent S. aureus Xen29. The S. aureus
Xen29 infection can be monitored by image analyses of live
mice, which show the reductions in the numbers of CFU in
the experimental mice and allows repeated examination of the
infections without the need to sacrifice the mice prior to the
end of the study. Figure 3 shows that treatment of the thigh
muscle-infected mice with compounds ␣-GalCer and 7DW8-5
3 h after infection resulted in increased clearance for many
treated mice; however, a statistically significant difference (P Ͻ
0.05) in comparison to the clearance for the vehicle-treated
group was noted only for the group receiving C34. This result
FIG. 3. Antibacterial efficacies of selected glycolipids in the Staph-
ylococcus aureus thigh infection model. The antibacterial efficacies of
glycolipids (150 ␮g/kg) that were administered by i.p. injection at 3 h
postinfection were evaluated by image analyses of luminescent S. au-
reus Xen29. (A) Images of mice in different treatment groups taken
48 h after infection. (B) The relative luminescence unit (RLU) values
of each mouse measured 48 h after infection were plotted as individual
symbols, and the mean values for the treatment groups are shown with
horizontal lines. , P Ͻ 0.05.
*
increased when the acetyl protecting group was at the 6 posi-
tion of the galactosyl donor (compound 1). The azide (com-
pound 3) was reduced using the Staudinger reaction, and the
resulting amine was coupled with fatty acid (compound 4) to
give compound 5 at an 88% yield. Finally, the protecting
groups of compound 5 were removed by several reactions,
resulting in C34 at a 40% yield in three steps.
Antibacterial efficacy of ␣-GalCer analogs in Sphingomonas
capsulata-infected mice. The Sphingomonas capsulata-infected

VOL. 54, 2010
ANTI-INFECTION ACTIVITIES OF ␣-GalCer ANALOG
4133
FIG. 4. Glycolipids enhance host protection against JEV infection.
Seven-week-old C57BL/6 mice were injected i.p. with 1 ␮g of com-
pound C1, 7DW8-5, C23, or C34 or with solvent 1 day before virus
challenge. The mice were infected i.p. with 5 ϫ 105 PFU of JEV (RP-9
strain) and simultaneously injected intracerebrally with 30 ␮l PBS.
One day after viral infection, the mice were boosted i.p. with 1 ␮g of
compound C1, 7DW8-5, C23, or C34 or with solvent. The mice were
monitored daily for 25 days, and the percent of survival is shown.
Fourteen mice were included in each experimental group. Survival
curve comparisons were performed by statistical analysis by the log
rank test using Prism software, and a P value of Ͻ0.05 for the result for
the treated mice compared to the result for the solvent-treated control
mice is indicated ( ).
*
suggests that C34 treatment may have a higher probability of
being beneficial for bacterial clearance in this disease model.
Antiviral evaluation of ␣-GalCer analogs using the JEV
infection animal model. To determine whether the glycolipids
elicit protective immunity against viral infection, we used a
two-dose protocol in which glycolipid was given 1 day before
and 1 day after JEV challenge of mice. As shown in Fig. 4,
compounds C1 and C23 increased the rate of mouse survival
from 21% to 57% by comparison with the rate for the solvent-
treated control group. Furthermore, the protective effects were
even more profound in mice receiving compounds C34 and
7DW8-5, with the survival rates being 64% and 71%, respec-
tively (Fig. 4). These results suggest that glycolipid-stimulated
NKT cells likely contribute to the host defense against JEV
infection. Similarly, C34, which is effective in protecting against
JEV infections, was also found to prolong the survival of in-
fluenza virus-infected BALB/c mice (see below). To further
dissect the requirement for the glycolipid-mediated immune
protection, we tested whether the beneficial effects of C34 and
7DW8-5 were still noted in mice deficient in the gene for signal
transducer and activator of transcription 1 (Stat-1), Ig ␮ chain,
or CD8 ␣ chain. Our studies showed that the protective activity
offered by either C34 or 7DW8-5 in wild-type C57BL/6 mice
was not demonstrated in immune-deficient mice. Apparently,
both the innate and adaptive immune components are re-
quired to trigger protective immunity against JEV infection by
these glycolipids, as no beneficial effect was noted in mice
lacking the gene for Stat-1, the immunoglobulin ␮ chain, or the
CD8 ␣ chain (Fig. 5).
FIG. 5. Glycolipids require host innate and adapted immunity com-
ponents to trigger protection against JEV infection. Seven-week-old
immunodeficient mice lacking Stat-1 (A), immunoglobulin ␮ chain
(B), or CD8 ␣ chain (C) were administered compound C34 or 7DW8-5
or solvent, as described in the legend to Fig. 4, using a two-dose
protocol. The Stat-1-knockout mice were intraperitoneally challenged
with 0.1 PFU of JEV (RP-9 strain). The Ig ␮ chain-knockout and CD8
␣ chain-knockout mice were challenged as described in the legend to
Fig. 4. The mice were monitored daily for 25 days. The percent survival
is shown, and the number of mice in each experimental group is shown
in the key to each panel.

4134
LIN ET AL.
ANTIMICROB. AGENTS CHEMOTHER.
FIG. 6. A two-dose protocol, with doses being administered 1 day before and 1 day after JEV infection, gives the best protective effect.
Glycolipid C34 (A) or 7DW8-5 (B) was administered to 7-week-old C57BL/6 mice by a two-dose protocol, as described in the legend to Fig. 4 [day
(Ϫ1) & day (ϩ1)] or only once 1 day before [day (Ϫ1)], on the same day [day (0)], or 1 day after [day (ϩ1)] viral infection. The mice were
challenged and monitored as described in the legend to Fig. 4. The percent survival and the number of mice in each experimental group are shown
in the key above each panel.
The protective activities of ␣-GalCer analogs are most ef-
fective when it is administered in a prophylactic manner. We
assessed whether these glycolipids could generate protective
immunity if they were administered after viral infection. Be-
sides the two-dose protocol described above, several one-dose
protocols were also tested. In comparison to the two-dose
protocol, C34 administered only once on the day before infec-
tion elicited a weaker protective effect against JEV. However,
if the glycolipids were given once either at the time of infection
or 1 day after viral infection, no protection against JEV chal-
lenge could be noted (Fig. 6A), suggesting that glycolipid-
mediated immune modulation against infection is a time-de-
pendent event. Reduced protective effects were also noted
when one dose of compound 7DW8-5 was administered to
mice 1 day before JEV infection. When the one dose was given
1 day after infection, no protective effects were found (Fig.
6B). The effects of C34 administration times versus viral infec-
tion time were also evaluated in influenza virus (strain WSN)-
infected BALB/c mice. Administration of C34 1 day before
WSN infection was the most beneficial to mouse survival (P Ͻ
0.01). Treatment with two doses of C34 administered both 1
day before and 1 day after WSN infection could also signifi-
cantly prolong survival (P Ͻ 0.05) so that it was longer than
that for the vehicle-treated group. Similar to JEV infection, a
single dose of C34 given 1 day after influenza virus infection
was not beneficial for survival (Fig. 7). We also studied the
FIG. 7. Survival profile of influenza virus-infected mice receiving
effects of the C34 administration times on bacterial clearance
compound C34 at different times and different doses. BALB/c mice
using the S. aureus thigh infection model. When mouse infec-
tions were examined 48 h after infection, bacterial clearance
was most profound in the group of mice administered C34
immediately after the thigh infections were introduced. For the
group receiving C34 6 h after thigh infection, an improvement
in bacterial clearance was noted, but to a lesser extent than that
were infected with 10 LD50s influenza virus WSN by the intranasal
(i.n.) route. Mice were grouped into four groups receiving two i.p.
injections on both day Ϫ1 and day ϩ1 relative to the time of influenza
virus infection. (A) Study protocol describing the contents of the i.p.
injections (1 ␮g/mouse C34 or PBS vehicle) and the injection times of
the different treatment groups are shown. (B) Survival curves for the
different treatment groups.

VOL. 54, 2010
ANTI-INFECTION ACTIVITIES OF ␣-GalCer ANALOG
4135
REFERENCES
1. Blauvelt, M. L., M. Khalili, W. Jaung, J. Paulsen, A. C. Anderson, S. B.
Wilson, and A. R. Howell. 2008. Alpha-S-GalCer: synthesis and evaluation
for iNKT cell stimulation. Bioorg. Med. Chem. Lett. 18:6374–6376.
2. Brigl, M., P. van den Elzen, X. Chen, J. H. Meyers, D. Wu, C. H. Wong, F.
Reddington, P. A. Illarianov, G. S. Besra, M. B. Brenner, and J. E. Gumperz.
2006. Conserved and heterogeneous lipid antigen specificities of CD1d-
restricted NKT cell receptors. J. Immunol. 176:3625–3634.
3. Chang, Y. J., J. R. Huang, Y. C. Tsai, J. T. Hung, D. Wu, M. Fujio, C. H.
Wong, and A. L. Yu. 2007. Potent immune-modulating and anticancer effects
of NKT cell stimulatory glycolipids. Proc. Natl. Acad. Sci. U. S. A. 104:
10299–10304.
4. Chen, L. K., Y. L. Lin, C. L. Liao, C. G. Lin, Y. L. Huang, C. T. Yeh, S. C.
Lai, J. T. Jan, and C. Chin. 1996. Generation and characterization of organ-
tropism mutants of Japanese encephalitis virus in vivo and in vitro. Virology
223:79–88.
5. Chiba, A., S. Oki, K. Miyamoto, H. Hashimoto, T. Yamamura, and S.
Miyake. 2004. Suppression of collagen-induced arthritis by natural killer T
cell activation with OCH, a sphingosine-truncated analog of alpha-galacto-
sylceramide. Arthritis Rheum. 50:305–313.
6. Fan, G., Y. S. Pan, K. C. Lu, Y. P. Cheng, W. C. Lin, S. Lin, C. H. Lin, C. H.
Wong, J. M. Fang, and C. C. Lin. 2005. Synthesis of ␣-galactosyl ceramide
and the related glycolipids for evaluation of their activities on mouse spleno-
cytes. Tetrahedron 61:1855–1862.
7. Franck, R. W., and M. Tsuji. 2006. Alpha-c-galactosylceramides: synthesis
and immunology. Acc. Chem. Res. 39:692–701.
8. Fujio, M., D. Wu, R. Garcia-Navarro, D. D. Ho, M. Tsuji, and C. H. Wong.
2006. Structure-based discovery of glycolipids for CD1d-mediated NKT cell
activation: tuning the adjuvant versus immunosuppression activity. J. Am.
Chem. Soc. 128:9022–9023.
FIG. 8. Effects of time of C34 administration on clearance of in-
fection in the murine thigh wound infection model. Mice infected in
the left thigh with luminescent S. aureus were grouped into vehicle
control group and two treatment groups, which received 150 ␮g/kg C34
at 0 and 6 h after infection, respectively. Mice were imaged at 48 h
postinfection. The images of the luminescent bacterial infection and
the relative luminescence units (RLU) are shown as in Fig. 3.
,
**
significant, with P being Ͻ0.01.
9. Godfrey, D. I., H. R. MacDonald, M. Kronenberg, M. J. Smyth, and L. Van
Kaer. 2004. NKT cells: what’s in a name? Nat. Rev. Immunol. 4:231–237.
10. Gonzalez-Aseguinolaza, G., L. Van Kaer, C. C. Bergmann, J. M. Wilson, J.
Schmieg, M. Kronenberg, T. Nakayama, M. Taniguchi, Y. Koezuka, and M.
Tsuji. 2002. Natural killer T cell ligand alpha-galactosylceramide enhances
protective immunity induced by malaria vaccines. J. Exp. Med. 195:617–624.
11. Ho, L. P., L. Denney, K. Luhn, D. Teoh, C. Clelland, and A. J. McMichael.
2008. Activation of invariant NKT cells enhances the innate immune re-
sponse and improves the disease course in influenza A virus infection. Eur.
J. Immunol. 38:1913–1922.
12. Hung, L. C., C. C. Lin, S. K. Hung, B. C. Wu, M. D. Jan, S. H. Liou, and S. L.
Fu. 2007. A synthetic analog of alpha-galactosylceramide induces macro-
phage activation via the TLR4-signaling pathways. Biochem. Pharmacol.
73:1957–1970.
13. Johnson, T. R., S. Hong, L. Van Kaer, Y. Koezuka, and B. S. Graham. 2002.
NK T cells contribute to expansion of CD8(ϩ) T cells and amplification of
antiviral immune responses to respiratory syncytial virus. J. Virol. 76:4294–
4303.
14. Kaieda, S., C. Tomi, S. Oki, T. Yamamura, and S. Miyake. 2007. Activation
of invariant natural killer T cells by synthetic glycolipid ligands suppresses
autoantibody-induced arthritis. Arthritis Rheum. 56:1836–1845.
15. Kakimi, K., L. G. Guidotti, Y. Koezuka, and F. V. Chisari. 2000. Natural
killer T cell activation inhibits hepatitis B virus replication in vivo. J. Exp.
Med. 192:921–930.
for mice receiving C34 immediately after the thigh infection
(Fig. 8).
In the present study, we used murine models of both bacte-
rial and viral infections to evaluate the infection suppression
efficacies of several ␣-GalCer analogs and found that com-
pound C34 is the most efficacious in all models. Previous stud-
ies showed that ␣-GalCer treatment of mice was complicated
by its association with detrimental side effects and that treat-
ment efficacy was influenced by a variety of parameters (37).
NKT cells respond to infection-associated glycolipids and are
stimulated by cytokines produced in dendritic cells upon infec-
tion (31). Thus, encountered with glycolipid administration
and infection challenges, the immune systems of the experi-
mental mice may experience complicated stimulations. Differ-
ent cytokine responses were documented in the influenza vi-
rus-infected mice, ␣-GalCer-treated mice, or both influenza
virus-infected and ␣-GalCer-treated mice. Greater serum cy-
tokine levels were often found in the doubly stimulated mice.
However, the profiles are complicated and varied at different
observation times (11).
16. Kinjo, Y., D. Wu, G. Kim, G. W. Xing, M. A. Poles, D. D. Ho, M. Tsuji, K.
Kawahara, C. H. Wong, and M. Kronenberg. 2005. Recognition of bacterial
glycosphingolipids by natural killer T cells. Nature 434:520–525.
17. Kobayashi, E., K. Motoki, T. Uchida, H. Fukushima, and Y. Koezuka. 1995.
KRN7000, a novel immunomodulator, and its antitumor activities. Oncol.
Res. 7:529–534.
The efficacy of ␣-GalCer for offering protection against LPS-
induced shock was found to be dependent on the time of
administration, and the glycolipid needs to be administered
before or within 2 h after LPS challenge (25, 26). Our in vivo
studies using bacterial and viral infection models also suggest
that compound C34 is the most effective at suppressing infec-
tions when it is administered before or shortly after infection.
Infections by different microorganisms could result in varied
immune stimulations with different kinetics. Thus, for effective
protection against infections, both the doses and the numbers
of administrations need to be fine-tuned to be beneficial.
18. Koto, S., N. Morishima, K. Takenaka, K. Kanemitsu, N. Shimoura, M. Kase,
S. Kojiro, T. Nakamura, T. Kawase, and S. Zen. 1989. 2-Methoxyethyl group
for protection of reducing hydroxyl group of aldose. Bull. Chem. Soc. Jpn.
62:3549–3566.
19. Kuklin, N. A., G. D. Pancari, T. W. Tobery, L. Cope, J. Jackson, C. Gill, K.
Overbye, K. P. Francis, J. Yu, D. Montgomery, A. S. Anderson, W. Mc-
Clements, and K. U. Jansen. 2003. Real-time monitoring of bacterial infec-
tion in vivo: development of bioluminescent staphylococcal foreign-body and
deep-thigh-wound mouse infection models. Antimicrob. Agents Chemother.
47:2740–2748.
20. Liang, J. J., C. L. Liao, J. T. Liao, Y. L. Lee, and Y. L. Lin. 2009. A Japanese
encephalitis virus vaccine candidate strain is attenuated by decreasing its
interferon antagonistic ability. Vaccine 27:2746–2754.
21. Liang, P. H., M. Imamura, X. Li, D. Wu, M. Fujio, R. T. Guy, B. C. Wu, M.
Tsuji, and C. H. Wong. 2008. Quantitative microarray analysis of intact
glycolipid-CD1d interaction and correlation with cell-based cytokine produc-
tion. J. Am. Chem. Soc. 130:12348–12354.
ACKNOWLEDGMENTS
22. Minagawa, S., C. Ohyama, S. Hatakeyama, N. Tsuchiya, T. Kato, and T.
Habuchi. 2005. Activation of natural killer T cells by alpha-galactosylcer-
amide mediates clearance of bacteria in murine urinary tract infection.
J. Urol. 173:2171–2174.
We thank the National Science Council, Taiwan, and Academia
Sinica, Taiwan, for financial support.
We declare that we have no conflicts of interest.

4136
LIN ET AL.
ANTIMICROB. AGENTS CHEMOTHER.
23. Morita, M., K. Motoki, K. Akimoto, T. Natori, T. Sakai, E. Sawa, K. Yamaji,
Y. Koezuka, E. Kobayashi, and H. Fukushima. 1995. Structure-activity re-
lationship of alpha-galactosylceramides against B16-bearing mice. J. Med.
Chem. 38:2176–2187.
24. Natori, T., M. Morita, K. Akimoto, and Y. Koezuka. 1994. Agelasphins, novel
antitumor and immunostimulatory cerebrosides from the marine sponge
Agelas mauritianus. Tetrahedron 50:2771–2784.
25. Sireci, G., M. P. La Manna, D. Di Liberto, M. Lo Dico, M. Taniguchi, F.
Dieli, and A. Salerno. 2008. Prophylaxis of lipopolysaccharide-induced shock
by alpha-galactosylceramide. J. Leukoc. Biol. 84:550–560.
26. Sireci, G., M. P. La Manna, C. Di Sano, D. Di Liberto, S. A. Porcelli, M.
Kronenberg, F. Dieli, and A. Salerno. 2007. Pivotal advance: alpha-galacto-
sylceramide induces protection against lipopolysaccharide-induced shock.
J. Leukoc. Biol. 81:607–622.
27. Su, C. Y., S. Y. Wang, J. J. Shie, K. S. Jeng, N. J. Temperton, J. M. Fang,
C. H. Wong, and Y. S. Cheng. 2008. In vitro evaluation of neuraminidase
inhibitors using the neuraminidase-dependent release assay of hemaggluti-
nin-pseudotyped viruses. Antiviral Res. 79:199–205.
28. Taniguchi, M., M. Harada, S. Kojo, T. Nakayama, and H. Wakao. 2003. The
regulatory role of Valpha14 NKT cells in innate and acquired immune
response. Annu. Rev. Immunol. 21:483–513.
29. Thurnhofer, S., and W. Vetter. 2007. Synthesis of (S)-(ϩ)-enantiomers of
food-relevant (n-5)-monoenoic and saturated anteiso-fatty acids by a Wittig
reaction. Tetrahedron 63:1140–1145.
killer T cells in the response to microorganisms. Nat. Rev. Microbiol. 5:405–
417.
32. van der Vliet, H. J., J. W. Molling, B. M. von Blomberg, N. Nishi, W. Kolgen,
A. J. van den Eertwegh, H. M. Pinedo, G. Giaccone, and R. J. Scheper. 2004.
The immunoregulatory role of CD1d-restricted natural killer T cells in
disease. Clin. Immunol. 112:8–23.
33. van Dommelen, S. L., H. A. Tabarias, M. J. Smyth, and M. A. Degli-Esposti.
2003. Activation of natural killer (NK) T cells during murine cytomegalovi-
rus infection enhances the antiviral response mediated by NK cells. J. Virol.
77:1877–1884.
34. Velmourougane, G., R. Raju, G. Bricard, J. S. Im, G. S. Besra, S. A. Porcelli,
and A. R. Howell. 2009. Synthesis and evaluation of an acyl-chain unsatur-
ated analog of the Th2 biasing, immunostimulatory glycolipid, OCH. Bioorg.
Med. Chem. Lett. 19:3386–3388.
35. Wu, D., M. Fujio, and C. H. Wong. 2008. Glycolipids as immunostimulating
agents. Bioorg. Med. Chem. 16:1073–1083.
36. Wu, D., D. M. Zajonc, M. Fujio, B. A. Sullivan, Y. Kinjo, M. Kronenberg,
I. A. Wilson, and C. H. Wong. 2006. Design of natural killer T cell activators:
structure and function of a microbial glycosphingolipid bound to mouse
CD1d. Proc. Natl. Acad. Sci. U. S. A. 103:3972–3977.
37. Wu, L., and L. Van Kaer. 2009. Natural killer T cells and autoimmune
disease. Curr. Mol. Med. 9:4–14.
38. Wu, S. F., C. J. Lee, C. L. Liao, R. A. Dwek, N. Zitzmann, and Y. L. Lin. 2002.
Antiviral effects of an iminosugar derivative on flavivirus infections. J. Virol.
76:3596–3604.
39. Yoshimoto, T., A. Bendelac, C. Watson, J. Hu-Li, and W. E. Paul. 1995. Role
of NK1.1ϩ T cells in a TH2 response and in immunoglobulin E production.
Science 270:1845–1847.
30. Trappeniers, M., S. Goormans, K. Van Beneden, T. Decruy, B. Linclau, A.
Al-Shamkhani, T. Elliott, C. Ottensmeier, J. M. Werner, D. Elewaut, and S.
Van Calenbergh. 2008. Synthesis and in vitro evaluation of alpha-GalCer
epimers. Chem. Med. Chem. 3:1061–1070.
31. Tupin, E., Y. Kinjo, and M. Kronenberg. 2007. The unique role of natural