Synthesis and Evaluation of Acyl-Chain- and Galactose-6′′-Modified Analogues of α-GalCer for NKT Cell Activation

Article abstract of DOI:10.1002/cbic.201200004

α-GalCer is an immunostimulating glycolipid that binds to CD1d molecules and activates invariant natural killer T (iNKT) cells. Here we report a scaled-up synthesis of α-GalCer analogues with modifications in the acyl side chain and/or at the galactose 6′′-position, together with their evaluation in vitro and in vivo. Analogues containing 11-phenylundecanoyl acyl side chains with aromatic substitutions (14, 16-21) and Gal-6′′-phenylacetamide-substituted α-GalCer analogues bearing p-nitro- (32), p-tert-butyl (34), or o-, m-, or p-methyl groups (40-42) displayed higher IFN-γ/IL-4 secretion ratios than α-GalCer in vitro. In mice, compound 16, with an 11-(3,4-difluorophenyl)undecanoyl acyl chain, induced significant proliferation of NK and DC cells, which should be beneficial in killing tumors and priming the immune response. These new glycolipids might prove useful as adjuvants or anticancer agents.

Full text of DOI:10.1002/cbic.201200004

DOI: 10.1002/cbic.201200004
Synthesis and Evaluation of Acyl-Chain- and Galactose-6’’-
Modified Analogues of a-GalCer for NKT Cell Activation
Ming-Han Hsieh,[a] Jung-Tung Hung,[b] Ya-Wen Liw,[a] Yin-Jen Lu,[a] Chi-Huey Wong,*[b]
Alice L. Yu,*[b] and Pi-Hui Liang*[a, b]
a-GalCer is an immunostimulating glycolipid that binds to
CD1d molecules and activates invariant natural killer T (iNKT)
cells. Here we report a scaled-up synthesis of a-GalCer ana-
logues with modifications in the acyl side chain and/or at the
galactose 6’’-position, together with their evaluation in vitro
and in vivo. Analogues containing 11-phenylundecanoyl acyl
side chains with aromatic substitutions (14, 16–21) and Gal-6’’-
phenylacetamide-substituted a-GalCer analogues bearing p-
nitro- (32), p-tert-butyl (34), or o-, m-, or p-methyl groups (40–
42) displayed higher IFN-g/IL-4 secretion ratios than a-GalCer
in vitro. In mice, compound 16, with an 11-(3,4-difluoropheny-
l)undecanoyl acyl chain, induced significant proliferation of NK
and DC cells, which should be beneficial in killing tumors and
priming the immune response. These new glycolipids might
prove useful as adjuvants or anticancer agents.
Introduction
a-GalCer (1, a-galactosylceramide, also known as KRN7000,
Scheme 1), a glycolipid composed of a-linked galactose, sphin-
gosine, and an acyl tail, was obtained by structure optimization
of agelasphin,[1] a natural product isolated from the marine
sponge Agelas mauritianus.[2] It was found to exhibit both in
vitro and in vivo antitumor activity and shown to be the most
potent ligand yet known for both mouse and human invariant
natural killer T cells (iNKT cells).[3] a-GalCer can bind with
CD1d—a molecule similar to the major histocompatibility com-
plex (MHC) class I—on antigen-presenting cells to generate
CD1d/a-GalCer biomolecular complexes, which are recognized
by the semi-invariant Tcell receptor (TCR, Va24-Ja18 in
humans and Va14-Ja18 in mice) in iNKT cells, resulting in the
secretion of T helper 1 (TH1; IFN-g, for example) and T helper 2
(TH2; IL-4, IL-10, for example) cytokines.[3]
biased cytokine profiles are highly desirable for future clinical
applications.
Miyamoto et al. reported the truncated (in the sphingosine
tail) analogue OCH[10] (2, Scheme 1), which was found to
induce IL-4-biased secretion and to protect mice from the
development of experimental autoimmune encephalomyelitis
(EAE), a mouse model for multiple sclerosis.[11] Porcelli et al.
demonstrated that C20:2 (3, Scheme 1) was able to induce
TH2-biased cytokines.[12] Conversely, another analogue, a-C-
GalCer (4, Scheme 1), showed enhanced TH1 response and dis-
played 100- to 1000-fold improved activity against melanoma
metastases relative to a-GalCer.[13]
In earlier studies we modified the acyl chain of a-GalCer
with aromatic substituents at the end and found that these
compounds increased the activity for TH1 response.[14–15] The
binding affinities of acyl-chain-modified analogues of a-GalCer
with CD1d were evaluated in a quantitative glycan microar-
ray,[16] and the results showed that compounds bearing an 11-
phenylundecanoyl moiety, such as compound 5 (Scheme 1),
exhibited TH1 cytokine selectivity twice as high as that of a-
GalCer, as well as a strong binding affinity with CD1d. The in-
Both TH1 and TH2 cytokines have important biological func-
tions. In general, TH2 cytokines are associated with immuno-
suppression and immunomodulation, whereas TH1 cytokines
are correlated with anti-infection and antitumor activity.[4–5]
Type I diabetes mellitus, for example, is characterized by over-
activation of TH1 cells and hypo-responsiveness to TH2. Skew-
ing this cytokine induction profile towards TH2 would be bene-
ficial for the treatment of this disease, where enhancement of
TH1 response will boost tumor immunity. Indeed, a-GalCer has
been evaluated in two anticancer Phase I clinical trials. Patients
with advanced cancers were injected with a-GalCer or a-
GalCer-loaded immature dendritic cells (iDCs), and although no
durable tumor regression was noted, transient reduction of
tumor markers and size was observed in a few patients.[6–7] The
lack of significant anticancer efficacy of a-GalCer might be the
result of the massive induction of a cytokine storm with con-
current and reciprocal functionalities from the activation of
iNKT cells.[8–9] Although the mechanism that skews the immune
response towards TH1 or TH2 is not clear, compounds with
[a] M.-H. Hsieh,+ Y.-W. Liw, Y.-J. Lu, Prof. Dr. P.-H. Liang
School of Pharmacy, College of Medicine, National Taiwan University
1, Jen-Ai Road, Section 1, Taipei 100 (Taiwan)
E-mail: phliang@ntu.edu.tw
[b] Dr. J.-T. Hung,+ Prof. Dr. C.-H. Wong, Prof. Dr. A. L. Yu, Prof. Dr. P.-H. Liang
Genomics Research Center, Academia Sinica
128, Academia Road, Section 2, Nankang, Taipei 115 (Taiwan)
chwong@sinica.edu.tw
E-mail: ayu@sinica.edu.tw
[+] These authors contributed equally to this work.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cbic.201200004: experimental procedures and
characterization of compounds 7–43 as well as in vivo cytokine secretions
of compounds 1, 16, and 42.
ChemBioChem 0000, 00, 1 – 10
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
1
&
ÞÞ
These are not the final page numbers!

C.-H. Wong, A. L. Yu, P.-H. Liang, et al.
Scheme 1. Structures of a-GalCer derivatives.
creased ligand affinity of compound 5 was thought to be the
result of favorable p–p interactions between the phenyl moiet-
ies and the aromatic side chains in the A’-tunnel of CD1d.[14,17]
The crystal structures of human iNKT TCR–a-GalCer–CD1d
complexes indicate that the 2’’-OH, 3’’-OH, and 4’’-OH groups
of galactose (for numbering see compound 1 in Scheme 1)
participate in hydrogen bonds to Gly96a, Ser30a, and Phe29a,
respectively, located on the invariant TCR a-chain in the iNKT
cell.[18] A nonglycosidic lipid ligand—threitolceramide, in which
galactose C5’’-OH and C6’’-OH of a-GalCer had been re-
moved—was demonstrated to activate iNKT cells effectively
through TCR with a weaker affinity for CD1d/threitolcreamide
than CD1d/a-GalCer complexes.[19] Another new series of non-
and at the Gal 6’’-position (e.g., 6)[22] were reported to skew
toward TH1 cytokine secretions, it seemed logical to design
molecules modified at the two sites for further biological eval-
uation. Furthermore, the substitution effects of these aromatic
rings remained unclear. To study their structure–activity rela-
tionships, a library of various a-GalCer derivatives bearing dif-
ferently substituted aromatic rings on the acyl chain and Gal
6’’-positions were synthesized. The new compounds in this
study were evaluated for their in vitro and in vivo immune
modulating activities by measurement of cytokine releasing
profiles and immune effector cell proliferation.
glycosidic agonists, aminocyclitol phytoceramides, were also Result and Discussion
found to be weak agonists for NKT TCR.[20,21] It was interesting
Chemistry
to note that these nonglycosidic ligands induced TH1-biased
secretion of cytokines in vivo, which were thought to be good
immune modulators because they avoided the cytokine storm
induced by a-GalCer.[19–21] These studies of nonglycosidic li-
gands also indicate that the galactose 6’’-OH group in a-
GalCer is not essential for activation of iNKT TCR. Furthermore,
the crystal structure of human iNKT TCR–a-GalCer–CD1d com-
plex also showed that the galactose 6’’-OH group in a-GalCer
was not involved in any H-bond interaction with CD1d or TCR.
Recently, Van Calenbergh et al. synthesized Gal-6’’ derivatives
such as the compound NU-a-GalCer (6, Scheme 1), which was
found to reduce IL-4 production significantly with retention of
TH1 activity.[22–23] X-ray crystallography studies of TCR–NU-a-
GalCer–mCD1d revealed that the aromatic group at Gal 6’’
formed a third anchor with CD1d, which pulled the galactose
slightly down into the groove, increasing the van der Waals
interactions between the ligand and CD1d.[23]
A scalable synthesis of glycolipid analogues was first devel-
oped (Scheme 2).The key phytosphingosine acceptor for glyco-
sylation was prepared by use of method similar to
a
Schmidt’s,[35] with modifications. The amino group of commer-
cial phytosphingosine hydrochloride[36] was converted into an
azido group in a diazo transfer reaction with freshly prepared
TfN3.[37] Selective tritylation of the primary alcohol, followed by
dibenzylation of the two secondary alcohols (BnCl/NaH) and
acidic removal of the trityl group, afforded the desired com-
pound 7. More than 100 grams of 7 were obtained by this
method.
Various glycosylation methods, involving donors with fluo-
ride, trichloroacetoimidate, bromide, and iodide as leaving
groups, have been used in glycolipid syntheses.[38–41] The glyco-
syl imidate used initially in our synthesis was obtained with ex-
cellent yield (89%) and anomeric selectivity (a/b 9:1). The imi-
date group is easily hydrolyzed, however, and donors bearing
it must be freshly prepared and purified by column chroma-
tography, so this compound is not suitable for large-scale syn-
thesis. An alternative method based on 4-methylphenyl 2,3-O-
dibenzyl-4,6-O-benzylidene-1-thiol-d-galactopyranoside (8)[40]
The majority of previous studies have focused only on single
site modifications of a-GalCer: variously in the acyl
chain,[12,14–16,24,25] in the sphingosine chain,[10,15,26,27] or in the
sugar moiety.[15,19,20,28–34] Because analogues of a-GalCer modi-
fied both in the phenylalkyl-containing acyl chain (e.g., 5)[14,15]
&2
&
www.chembiochem.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 0000, 00, 1 – 10
ÝÝ
These are not the final page numbers!

a-GalCer Analogues for NKT Cell Activation
as the donor was developed; this was activated with Me2S2/
Tf2O at À108C[42] to afford 9 in a 64% yield (7.1% b form).
Azide reduction was accomplished with LiAlH4 in THF to give
the corresponding amine intermediate, which was coupled
with various prepared carboxylic acids (Scheme S1 in the Sup-
porting Information) to give compounds 10–13. Global depro-
tection of compounds 10–13 was achieved under hydrogenol-
ysis conditions in the presence of catalytic Pd(OH)2 and H2 in
a mixture of MeOH and CH2Cl2 to yield analogues 14–17 in
72–94% yields.
(Scheme 2), which was coupled with acids in the presence of
HBTU and NMM to give analogues 19–21.
The synthesis of 6’’-modified Gal analogues began with com-
pound 22 (Scheme 3).[22] Reduction of the 6’’-azide gave an
amine intermediate, which was condensed with different com-
mercial aromatic acids to give 23–25 and then 26–28 after
deprotection.
For the synthesis of the doubly modified compounds in
which the acyl chain was an 11-(2,4-difluorophenyl)undecanoyl
group, a different strategy that avoided tedious protective
group interconversions was adopted (Scheme 4). The Gal 6’’-
OH group of 16 was tosylated (TsCl/pyridine), and the tosyloxy
group was substituted with azide to give 29. Staudinger reduc-
tion of this azide gave amine 31. Amidation with a variety of
carboxylic acids yielded 32–43. The poor solubility of fully de-
Unfortunately, though, these reactions proceeded with con-
comitant reductive dehalogenation of the aryl halides of 19–
21, and so it was decided to remove the benzyl groups before
the amide bond formation. Compound 9 was subjected to
hydrogenolysis conditions to give the fully deprotected 18
Scheme 2. Reagents and conditions: a) i: TfN3, K2CO3, CuSO4, CH2Cl2, MeOH, H2O, 08C to RT, 12 h; ii: trityl chloride, TEA, toluene, 50–558C, 21 h; iii: BnCl, NaH,
DMF, toluene, 50–608C, 18 h; iv: HCl, toluene, MeOH, 50–608C, 20 h, 36% (over 4 steps); b) 8,[42] Me2S2/Tf2O, THF, 4 ꢁ MS, À108C, 20 min, 64%; c) i: LiAlH4,
THF, 08C to RT, 70 min; ii: RCO2H, HBTU, NMM, CH2Cl2, RT, 12 h, 38–63% (over 2 steps); d) Pd(OH)2, H2 (550 kPa), MeOH, CH2Cl2, AcOH, 16 h, RT, quant.;
e) Pd(OH)2, H2, MeOH, CH2Cl2, 12 h, RT, 72–94%; f) RCO2H, HBTU, NMM, CH2Cl2, MeOH, RT, 12 h, 19–25%.
Scheme 3. Reagents and conditions: a) PPh3, THF, H2O, RT, 2 days; b) Ph(CH2)nCO2H, HBTU, NMM, CH2Cl2, RT, 12 h; c) Pd(OH)2, H2, CH2Cl2, MeOH, RT, 15 h, 22–
56% (over 3 steps).
ChemBioChem 0000, 00, 1 – 10
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
&
3
&
ÞÞ
These are not the final page numbers!

C.-H. Wong, A. L. Yu, P.-H. Liang, et al.
not found to have a significant
effect. This observation is consis-
tent with a recent X-ray crystalli-
zation study,[17] which indicated
that minor conformational differ-
ences of residues in the A’-
tunnel of CD1d did not induce
any major changes that could
readily be discriminated by the
iNKT TCR. With regard to the gal-
actose-modified analogues, a-
GalCer bearing a phenylaceta-
mide group at the Gal 6’’-posi-
tion
observed to induce higher IL-2
cytokine secretion
(compound
26)
was
(47.4Æ
2.07 ngmLÀ1) than the other Gal
6’’-phenylalkanamide derivatives
(29.2Æ4.72 ngmLÀ1 and 11.6Æ
0.43 ngmLÀ1 for 27 and 28;
Figure 1). Interestingly, in com-
pounds 27 and 28 (with de-
creased IL-2 levels in relation to
26) the phenyl groups are held
further away from the rest of the
Scheme 4. Reagents and conditions: a) TsCl, pyridine, 08C to RT, 16 h, 25% (100% based on recovery of starting
material); b) NaN3, DMF, 1008C, 2 days, 82%; c) PPh3, THF, H2O, 2 days, 53%; d) RCH2CO2H, HBTU, NMM, MeOH,
CH2Cl2, RT, 16 h, 9–42%.
molecules by progressively longer chains of methylene groups.
Recent findings indicated that functional groups at Gal-6’’
(such as in compound 6) interacted with the backbone resi-
dues Met69 (a1 helix) and Thr159 (a2 helix) of mCD1d.[22] The
phenylacetamide component of compound 26 seems to fit
better in this pocket. Compounds 26–28 were found to be
poorly soluble in DMSO in comparison with a-GalCer
(ꢀ1 mgmLÀ1), however, which might make them unsuitable
for further development.
Because different aromatic substitutions on the phenylunde-
canoyl analogues (5, 14–17, 19–21) were not found to affect
IL-2 secretion significantly, doubly modified compounds in
which the 11-(3,4-difluorophenyl)undecanoyl acyl chain moiety
was conserved and the Gal 6’’-phenylacetamido substituents
were varied were studied next. The 11-(3,4-difluorophenyl)un-
decanoyl moiety on the acyl chain was chosen because of its
good solubility characteristics (>10 mgmLÀ1 of DMSO) and
high total preparative yield (>10% based on the synthesis of
compound 16 from phytosphingosine). Again, compounds 32–
43, bearing various aromatic substituents at the Gal 6’’-phenyl-
acetamido position, were screened for their abilities to induce
IL-2 secretions in the A20-CD1d and mNK1.2 cell systems
(Figure 1). Decreases in IL-2 level were observed both for the
two compounds substituted with the electron-withdrawing
nitro group (32 and 33; 19.9 and 15.1 ngmLÀ1, respectively)
and for the halogenated compounds 37–39 (range: 19.5–
28.1 ngmLÀ1). In contrast, electron-donating substituents (36,
40–43; range: 39–59.2 ngmLÀ1) were shown to result in en-
hanced IL-2 cytokine production, suggesting superior abilities
to activate NKT cells.
Figure 1. Effects of a-GalCer analogues on IL-2 production by mNK1.2 cells.
A20-mCD1d cells were loaded with indicated a-GalCer analogues (0.2 mm)
and co-cultured with mNK1.2 cells. The supernatants were harvested and
the production of IL-2 was determined by ELISA. Data are given as means
ÆSDs, D: DMSO; - indicates no glycolipid.
protected 31 might account for the low yield of this amidation
step.
Biology
In order to ascertain the abilities of these compounds to acti-
vate NKT cells, a mouse B lymphoma cell line (A20-CD1d) was
selected as a source of antigen-presenting cells for loading
with the various glycolipids. After co-incubation with Va14-
TCR-expressing mNK1.2 cells as effector cells, IL-2 secretion
was measured by ELISA. As shown in Figure 1, all of the syn-
thesized acyl-chain-modified analogues (5, 14–17, 19–21) in-
duced higher levels of IL-2 cytokine secretion (range: 49.2–
62.6 ngmLÀ1) than a-GalCer (44.1Æ0.14 ngmLÀ1). Com-
pounds 14 and 21, with 4-phenoxy and 4-bromo substitution,
respectively, were found to produce slightly higher IL-2 cyto-
kine levels, but aromatic substitution in these molecules was
To investigate further whether these analogues could elicit
TH1- or TH2-biased cytokine responses, splenocytes from
&4
&
www.chembiochem.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 0000, 00, 1 – 10
ÝÝ
These are not the final page numbers!

a-GalCer Analogues for NKT Cell Activation
(Table 1). Modification only of the acyl chain of a-GalCer gave
a twofold improvement in TH1-biased selectivity, which was
similar to our previous finding with human Va24NKT cells.[16] To
our delight, however, Gal 6’’-phenylacetamides substituted
with p-nitro- (32), p-tert-butyl (34), or o-, m-, or p-methyl (40,
41, 42) groups induced three- to sixfold improvements in
selectivity over a-GalCer, although the levels of cytokines in-
duced by 40 and 41 were lower than those induced by 42 and
a-GalCer.
Table 1. Induction of cytokine production by a-GalCer analogues.[a]
Compound
IFN-g secretion
IL-4 secretion
IFN-g/IL-4 ratio[b]
DMSO
1
0.07
0
–
1.00
8.33Æ1.86
11.5Æ0.20
8.26Æ0.68
6.96Æ0.94
6.95Æ0.07
8.21Æ2.60
3.95Æ0.85
12.0Æ1.50
2.42Æ0.90
0
0.15Æ0.02
0.12Æ0.02
0.09Æ0.02
0.10Æ0.01
0.08Æ0.01
0.08Æ0.007
0.04Æ0.01
0.14Æ0.02
0.01Æ0.007
0
14
16
17
20
21
32
33
34
35
36
37
38
39
40
41
42
43
1.96Æ0.24
2.02Æ0.23
2.00Æ0.47
1.90Æ0.68
1.91Æ0.13
3.51Æ0.82*
1.78Æ0.74
4.34Æ0.69*
–
Numerous factors have been reported to cause cytokine
profile shifts; contributing causes might be the stabilities of
the CD1d/glycolipid binary complex[43,44] or the CD1d/glycoli-
pid/TCR ternary complex.[45] One possibility is that an addition-
al aromatic ring in the Gal 6’’-position, such as in the doubly
modified analogues (32, 34, 40–42), might provide additional
p–p interactions with Trp-153 of CD1d and Phe-29a of
CDR1a,[18] and enhance the stability of the CD1d/glycolipid/
TCR complex.[45] The pharmacokinetic stabilities of glycolipids
are an important issue for cytokine polarization, because TH1-
biasing glycolipids require longer stimulus periods.[12,45,46]
Modifications at the Gal 6’’-position are less susceptible to hy-
drolysis by galactosidase, so these compounds (32, 34, 40–42)
might be more stable and might polarize stronger TH1 re-
sponses in vivo. It is noteworthy that all of the these doubly
modified compounds were more soluble than a-GalCer (1–
10 mgmLÀ1 vs. ꢀ1 mgmLÀ1 of DMSO), suggesting they might
be well suited for future development.
4.18Æ1.07
0
0.09Æ0.02
0
0
1.15Æ0.16
–
–
0.01Æ0.01
0.08Æ0.04
3.28Æ1.30
2.38Æ0.96
7.73Æ0.50
4.85Æ2.33
0
–
0.02Æ0.005
0.01Æ0.004
0.05Æ0.01
0.05Æ0.006
5.80Æ0.76*
4.57Æ1.13*
3.45Æ0.18*
2.45Æ0.77
[a] Splenocytes from C57BL/6 mice were cultured with the indicated gly-
colipids (0.1 mm). After 60 h, IFN-g and IL-4 production (ngmLÀ1) was
measured by ELISA. Assays were performed in triplicate and data are
given as means ÆSDs. [b] The ratios of IFN-g to IL-4 were obtained after
normalization of the levels of IFN-g and IL-4 from the indicated glyco-
lipids with the levels of IFN-g and IL-4 from a-GalCer. * p<0.05 vs. a-
GalCer.
C57BL/6 female mice were cultured in vitro with the glycoli-
pids (0.1 mm) for 60 h and the proliferation of splenocytes was
determined by the alarmarBlue assay. DMSO and a-GalCer (1)
were used in this assay as negative and positive controls, re-
spectively. All of these compounds were found to be capable
of inducing splenocyte proliferation, and no cytotoxicity was
observed (Figure S1). The secretion profiles of IFN-g and IL-4
from the supernatant were examined by ELISA. Similarly to the
IL-2 secretion results, considerable levels of secretion of IFN-g
and IL-4 were observed for compounds 14 (11.5Æ0.2 and
0.12Æ0.02 ngmLÀ1), 16 (8.26Æ0.68 and 0.09Æ0.02 ngmLÀ1),
17 (6.96Æ0.94 and 0.1Æ0.01 ngmLÀ1), and 20 (6.95Æ0.07 and
0.08Æ0.01 ngmLÀ1), in which the only modification was in the
acyl chain of a-GalCer (Table 1). Comparison of the cytokine se-
cretion profiles for different substitution groups on the phenyl
rings in the acyl chains showed no significant observable sub-
stitution effects for these analogues. Significantly, various sub-
stituted groups on the Gal 6’’-phenylacetamido component
did indeed affect the cytokine secretions. In contrast with a pre-
vious report, in which the Gal 6’’-(4-Cl, 3-CF3)-benzamide of a-
GalCer was found to induce high IFN-g secretion,[22] we found
that halogen-containing Gal 6’’-phenylacetamide compounds
(35, 37–39) failed to induce cytokine secretions. On the other
hand, compounds with o,p-dinitro groups (33; 12.0Æ
1.5 ngmLÀ1) or a p-methyl group (42; 7.73Æ0.5 ngmLÀ1) were
able to induce significant IFN-g secretion.
To investigate the induction of cytokine/chemokine in vivo
further, several analogues with different cell-based activity pro-
files were selected. Compound 21 with higher IL-2 secretion,
compound 33 with higher TH1 and TH2 secretion, and com-
pounds 40–42 with better TH1/TH2 ratios were compared with
a-GalCer (positive control), compound 16 (parent compound
of the doubly modified analogues), and DMSO (negative con-
trol, 1%, 100 mL). Compounds were i.v. (intravenously) injected
into female C57BL/6 mice and serum was harvested immedi-
ately before injection as well as at 2 and 18 h after injection,
coincidental with the times after which maximum production
of IFN-g and of IL-4 are reached. The serum levels of TH1 cyto-
kines (IFN-g, IL-2, and IL-12), TH2 cytokines (IL-4, IL-5, IL-6, IL-10,
and IL-13), inflammatory cytokines (IL-1a, IL-1b, IL-8, IL-17,
TNF-a, and GM-CSF), growth-promoting cytokines (IL-7, IL-9, IL-
15, and G-CSF), and chemokines (RANTES, MIP-1a, MCP-1, IP-
10, and KC) were determined with the aid of the Luminex
system with multiplex kit (see Figure 2 and Figures S2 and S3).
As predicted, a-GalCer induced most of the cytokine/chemo-
kines massively and concurrently. Similarly to the responses of
a-GalCer, compounds 16 and 21 induced significant secretion
of cytokines/chemokines tested, except for IL-7 and IL-17 at
2 h and IL-2, IL-4, IL-6, IL-7, IL-13, IL-1b, IL-17, TNF-a and GM-
CSF at 18 h. Compound 16 induced a potent production of
IFN-g at 18 h, which is about half the amount induced by a-
GalCer at the same dose (Figure 2). Compound 21 induced less
IFN-g secretion than 16 at 18 h. In contrast with the compara-
tively strong production found in the in vitro stimulation assay
(Table 1), compound 33 was essentially unable to induce cyto-
kine secretion in vivo except for marginal levels of IL-5, MIP1a,
To compare the TH1/TH2 selectivities of these compounds
with that of a-GalCer, the cytokine (IFN-g and IL-4) production
profiles of glycolipids normalized to that of a-GalCer were ob-
tained, together with the selectivity ratios of these compounds
ChemBioChem 0000, 00, 1 – 10
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
&
5
&
ÞÞ
These are not the final page numbers!

C.-H. Wong, A. L. Yu, P.-H. Liang, et al.
knock-out mice (data not shown), indicating CD1d-dependent
activation of NKT cells.
Previous studies have shown that the glycolipids could stim-
ulate NKT cells to secrete TH1/TH2 cytokine/chemokines to acti-
vate various types of immune cells.[5,47] Because immune cell
numbers are known to be lower in patients with malignan-
cies,[48–50] proliferation of splenocytes and immune effector
cells with these glycolipids might have therapeutic benefits.
We analyzed the populations of immune cells in the spleens of
female C57BL/6 mice at 72 h after injection (2 mg/100 mL) of a-
GalCer, 16, 21, 33, 40, 41, and 42, together with DMSO (1%,
100 mL). As shown in Figure 3, after i.v. administration, no sig-
nificant proliferation of splenocytes was observed in compari-
son with DMSO. The subpopulations of immune cells in spleen
were determined by flow cytometry. In general, a-GalCer was
superior to compound 16, which was in turn superior to other
glycolipids in increasing most of the subpopulations of
immune cells. In a-GalCer-treated mice there were significant
increases in the percentages and absolute cell numbers of NKT
(0.6Æ0.1%, 0.7Æ0.2ꢂ106), CD3 T (29.6Æ3.5%, 36.4Æ2.3ꢂ106),
CD4 T (17.8Æ2.5%, 21.8Æ1.9ꢂ106), and DC (4.0Æ0.3%, 4.9Æ
0.6ꢂ106) cells (Figure 3). Mice treated with 16 showed higher
percentages of NK (3.4Æ0.1%) and DC (3.6Æ0.5%) cells rela-
tive to the DMSO group (2.2Æ0.4% and 2.4Æ0.2%, respective-
ly). The ability of compound 16 to increase populations of NK
and DC cells is important for killing tumors and priming
immune response. With regard to doubly modified compounds
(33, 40–42), compound 42 induced significant proliferation of
DC cells (3.2Æ0.09%) and slight proliferation of CD3 T, CD8 T
and B cells. The mechanism of activation of immune cells by
42 remains enigmatic. It is possible that 42 might interact with
other receptors (such as Toll-like receptor or lectins on DCs), or
it might directly activate CD1d-expressing immune cells.[51]
Figure 2. Effects of a-GalCer analogues on IFN-g and IL-4 production in vivo.
a-GalCer (1), 16, 21, 33, and 40–42 (2 mg/100 mL) and DMSO (1%, 100 mL)
were injected i.v. into C67BL/6 mice. The serum was harvested at 2 h and
18 h after injection and the levels of cytokines were determined using a
luminex multiplex assay kit. Data indicate the means for three independent
mice.
and IP-10 secretions (Figure 2 and Figures S2 and S3). Com-
pounds 40, 41, and 42, which are Gal 6’’-phenylacetamides
substituted with a methyl group at the ortho-, meta-, and
para-position, respectively, induced the secretion of only a few
cytokines/chemokines. The secretion of IFN-g can only be de-
tected in mice treated with 42 (Figure 2), consistently with the
observation of IFN-g induced in vitro (Table 1), in which 42
indeed induced higher production of IFN-g than 40 and 41.
IL-4 secretion was poorly induced by the new compounds
tested (Figure 2). Compound 16 induced only a quarter of the
amount of IL-4 secretion induced by a-GalCer and com-
pound 21 induced 2% of the amount induced by a-GalCer at
the same dose. Calculation of the IFN-g/IL-4 ratios for these
compounds shows 16, 21, and 42 to elicit more TH1-biased re-
sponses than a-GalCer at 18 h. The higher TH1-biased respons-
es of these compounds suggest that they might be better in
polarizing the immune response toward TH1 cells, which could
prime the cytotoxic Tcells to kill tumor cells. Moreover, com-
pound 16 also induced the secretion of stimulatory cytokines
(G-CSF, IL-9, TNF-a, and GM-CSF) and chemokines (IP-10, KC,
MCP-1, and RNATES), which could activate and recruit immune
cells. The doubly modified compounds (33, 40–42) displayed
weak secretion of cytokines/chemokines in vivo. They might
be considered weak agonists of NKT cells or need to be inject-
ed into mice in higher doses, if such biological activity as
occurs merits further investigation. No cytokine secretion was
observed when these glycolipids were injected into CD1d
Conclusions
a-GalCer analogues bearing aromatic substituents at their acyl
and/or galactose 6’’-positions were synthesized and their ca-
pacities to activate immune cells were investigated. The results
indicated that aromatic substituents on the phenylundecanoyl
acyl side chain of a-GalCer induced a significant release of cy-
tokines, but the nature of substitutions at this position seems
to be less important. However, substitutions in the Gal 6’’-phe-
nylacetamide component have significant effects on NKT TCR
binding. Compounds bearing halogen substituents in the Gal
6’’-phenylacetamide component were unable to induce cyto-
kine secretion. Compounds bearing a nitro group (e.g., 32),
a tBu group (e.g., 34), or electron-donating groups (e.g., 40–
42 with methyl groups) on the Gal 6’’-phenylacetamide com-
ponent were able to polarize the in vitro production of TH1 cy-
tokine significantly. In in vivo experiments, compounds 16, 21,
and 42 elicited more TH1-biased responses than a-GalCer. The
stronger TH1-biased responses of these compounds suggest
that they might be better than a-GalCer in polarizing the
immune response toward TH1 cells, which could prime the cy-
totoxic Tcells to kill tumor cells.
&6
&
www.chembiochem.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 0000, 00, 1 – 10
ÝÝ
These are not the final page numbers!

a-GalCer Analogues for NKT Cell Activation
Figure 3. Proliferation of murine splenocytes after injection of a-GalCer (1), 16, 21, 33, and 40–42. Female C57BL/6 mice (n=3 per group) were injected with
1, 16, 21, 33, and 40–42 (2 mg/100 mL), together with DMSO (D, 1%, 100 mL), the spleens were harvested at 72 h after injection, and the populations of
immune cells were then analyzed by FACS. Total numbers of splenocytes and percentage increases in subpopulations of immune cells including NK, NKT,
CD3 T, CD4 T, CD8 T, DC, B, and MDSC cells were determined. Assays were performed in triplicate and data are given as means ÆSDs. * P<0.05, relative to
DMSO.
quantified with a DuoSet ELISA Development System (R&D System,
MN, USA). Briefly, the capture antibody was coated to the plate
Experimental Section
overnight at 48C. The plate were washed and blocked with block-
ing buffer (BSA in PBS, 1%). The samples were then added, and
the plates were incubated at room temperature for 2 h. After three
washes with wash buffer (Tween 20 in PBS, 0.05%), the detection
antibody was added to the plates, which were incubated at room
temperature for 2 h. After three washes with wash buffer, streptavi-
din-HRP was added and the plates were incubated at room tem-
perature for 20 min. The plates were then washed four times with
wash buffer, substrate solution was added to each well, and incu-
bation was performed for 20 min at room temperature. Finally,
stop solution (H2SO4, 2.0n) was added and the optical density
(OD450) was determined with a microplate reader (SpectraMax M2,
Molecular Devices, Sunnyvale, CA, USA).
Synthesis: See Schemes 2–4 and the Supporting Information.
Animal experiments: C57BL/6 female mice were purchased from
the National Laboratory Animal Center (Taipei, Taiwan). Mice were
maintained under specific pathogen-free conditions in an animal
facility and treated in accordance with the institutional guidelines.
All animal experiments were approved by the Institutional Animal
Care and Ethics Committee.
Preparation of a-GalCer and a-GalCer analogues: All glycolipids
were dissolved in DMSO (100%) at concentrations of 1–10 mm or
1–10 mgmLÀ1. Before use, the solutions were diluted with phos-
phate-buffered saline (PBS, pH 7.4) to obtain final concentrations
of 0.01 mm, 0.1 mm, 0.2 mm, or 20 mgmLÀ1
.
In vitro cell proliferation assay: Splenocytes of C57BL/6 female
mice (16 w) were stimulated with the indicated glycolipids at
Cytokine secretion assay: Supernatant or serum was collected at
the indicated time and the production of IL-2, IL-4, and IFN-g was
ChemBioChem 0000, 00, 1 – 10
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
&
7
&
ÞÞ
These are not the final page numbers!

C.-H. Wong, A. L. Yu, P.-H. Liang, et al.
[12] K. O. A. Yu, J. S. Im, A. Molano, Y. Dutronc, P. A. Illarionov, C. Forestier, N.
Fujiwara, I. Arias, S. Miyake, T. Yamamura, Y.-T. Chang, G. S. Besra, S. A.
Porcelli, Proc. Natl. Acad. Sci. USA 2005, 102, 3383–3388.
[13] G. Yang, J. Schmieg, M. Tsuji, R. W. Franck, Angew. Chem. 2004, 116,
3906–3910; Angew. Chem. Int. Ed. 2004, 43, 3818–3822.
[14] M. Fujio, D. Wu, R. Garcia-Navarro, D. D. Ho, M. Tsuji, C.-H. Wong, J. Am.
Chem. Soc. 2006, 128, 9022–9023.
[15] Y.-J. Chang, J.-R. Huang, Y.-C. Tsai, J.-T. Hung, D. Wu, M. Fujio, C.-H.
Wong, A. L. Yu, Proc. Natl. Acad. Sci. USA 2007, 104, 10299–10304.
[16] P.-H. Liang, M. Imamura, X. Li, D. Wu, M. Fujio, R. T. Guy, B.-C. Wu, M.
Tsuji, C.-H. Wong, J. Am. Chem. Soc. 2008, 130, 12348–12354.
[17] A. Schiefner, M. Fujio, D. Wu, C.-H. Wong, I. A. Wilson, J. Mol. Biol. 2009,
394, 71–82.
[18] N. A. Borg, K. S. Wun, L. Kjer-Nielsen, M. C. Wilce, D. G. Pellicci, R. Koh,
G. S. Besra, M. Bharadwaj, D. I. Godfrey, J. McCluskey, J. Rossjohn, Nature
2007, 448, 44–49.
[19] J. D. Silk, M. Salio, B. G. Reddy, D. Shepherd, U. Gileadi, J. Brown, S. H.
Masri, P. Polzella, G. Ritter, G. S. Besra, E. Y. Jones, R. R. Schmidt, V. Cerun-
dolo, J. Immunol. 2008, 180, 6452–6456.
[20] J. Kerzerho, E. D. Yu, C. M. Barra, E. Alari-Pahisa, E. Girardi, Y. Harrak, P.
Lauzurica, A. Llebari, D. M. Zajonc, O. Akbari, A. R. CastaÇo, J. Immunol.
2012, 188, 2254–2265.
[21] Y. Harrak, C. M. Barra, A. Delgado, A. R. CastaÇo, A. Llebaria, J. Am.
Chem. Soc. 2011, 133, 12079–12084.
[22] M. Trappeniers, K. Van Beneden, T. Decruy, U. Hillaert, B. Linclau, D. Ele-
waut, S. Van Calenbergh, J. Am. Chem. Soc. 2008, 130, 16468–16469.
[23] S. Aspeslagh, Y. Li, E. D. Yu, N. Pauwels, M. Trappeniers, E. Girardi, T.
Decruy, K. V. Beneden, K. Venken, M. Drennan, L. Leybaert, J. Wang,
R. W. Franck, S. V. Calenbergh, D. M. Zajonc, D. Elewaut, EMBO J. 2011,
30, 2294–2305.
0.01 mm and 0.1 mm. Cell culture was carried out for three days, ala-
marBlue (AbD serotec, Oxford, UK) was added, and cells were cul-
tured for another 7 h. The cell proliferation was determined by
using SpetraMax M2 (Molecular Devices, CA, USA) of fluorescence
emission at 590 nm.
In vivo cytokine secretion and cell expansion assay: Wild-type
and CD1d knockout female C67BL/6 mice were i.v. injected with
one dose (2 mg/100 mL) of glycolipid. Serum was collected immedi-
ately before injection and at 2 h and 18 h afterwards, and the pro-
files of cytokines/chemokines were detected with a MILLIPLEX
Mouse Cytokine/Chemokine kit (Millipore, MA), with measurement
with a Luminex 200 (Luminex, TX) machine. Data were analyzed
with Luminex xPONENT software. Three days after glycolipid injec-
tions, the mice were euthanized and spleens were harvested. The
weight of each spleen was measured with a balance (Mettler Tole-
do AG, Switzerland) and viable cells were counted with a Vi-cell XR
instrument (Beckman Coulter, CA). Subpopulations of immune cells
were determined by FACS analysis. Briefly, splenocytes were
stained with fluorescence-conjugated antibody (Biolegend, CA),
including NK1.1, CD3, CD4, CD8, CD11c, CD19, CD11b, Gr1, and
CD25, for 30 min on ice. After incubation, cells in the plates were
washed and resuspended with FACS buffer [BSA (1%) and NaN3
(0.1%) in PBS]; fluorescence intensities were then determined by
FACSCanto (BD Biosciences, CA) and analyzed with FACSDiva soft-
ware (BD Bioscience, CA).
[24] T. Toba, K. Murata, K. Nakanishi, B. Takahashi, N. Takemoto, M. Akabane,
T. Nakatsuka, S. Imajo, T. Yamamura, S. Miyake, H. Annoura, Bioorg. Med.
Chem. Lett. 2007, 17, 2781–2784.
Acknowledgements
[25] M. Michieletti, A. Bracci, F. Compostella, G. De Libero, L. Mori, S. Fallarini,
G. Lombardi, L. Panza, J. Org. Chem. 2008, 73, 9192–9195.
[26] K. Fuhshuku, N. Hongo, T. Tashiro, Y. Masudo, R. Nakagawa, K. Seino, M.
Taniguchi, K. Mori, Bioorg. Med. Chem. 2008, 16, 950–964.
[27] M. Trappeniers, S. Goormans, K, Van Beneden, T. Decruy, B. Linclau, A.
Al-Shamkhani, T. Elliott, C. Ottensmeier, J. M. Werner, D. Elewaut, S. Van
Calenbergh, ChemMedChem 2008, 3, 1061–1070.
[28] D. M. Zajonc, C. Cantu, 3rd., J. Mattner, D. Zhou, P. B. Savage, A. Bende-
lac, I. A. Wilson, L. Teyton, Nat. Immunol. 2005, 6, 810–818.
[29] V. Costantino, E. Fattorusso, C. Imperatore, A. Mangoni, Tetrahedron
2002, 58, 369–375.
We thank the National Science Council of Taiwan, ROC (NSC 98-
2320-B-002-017-MY3), the National Taiwan University, and the
Academia Sinica for financial support. We also acknowledge the
excellent technical assistance, service, and advice provided by the
cell imaging/flow cytometry facility at the Division of Medical
Biology, Genomics Research Center, Academia Sinica.
Keywords: alpha-galactosylceramide · CD1d · cytokines · NKT
cells · substituent effects
[30] L. Barbieri, V. Costantino, E. Fattorusso, A. Mangoni, N. Basilico, M. Mon-
dani, D. Taramelli, Eur. J. Org. Chem. 2005, 3279–3285.
[31] G.-W. Xing, D. Wu, M. A. Poles, A. Horowitz, M. Tsuji, D. D. Ho, C.-H.
Wong, Bioorg. Med. Chem. 2005, 13, 2907–2916.
[32] X. T. Zhou, C. Forestier, R. D. Goff, C. Li, L. Teyton, A. Bendelac, P. B.
Savage, Org. Lett. 2002, 4, 1267–1270.
[1] M. Morita, K. Motoki, K. Akimoto, T. Natori, T. Sakai, E. Sawa, K. Yamaji, Y.
Koezuka, E. Kobayashi, H. Fukushima, J. Med. Chem. 1995, 38, 2176–
2187.
[33] R. Raju, B. F. Castillo, S. K. Richardson, M. Thakur, R. Severins, M. Kronen-
berg, A. R. Howell, Bioorg. Med. Chem. Lett. 2009, 19, 4122–4125.
[34] T. Tashiro, R. Nakagawa, S. Inoue, M. Shiozaki, H. Watarai, M. Taniguchi,
K. Mori, Tetrahedron Lett. 2008, 49, 6827–6830.
[2] T. Natori, Y. Koezuka, T. Higa, Tetrahedron Lett. 1993, 34, 5591–5592.
[3] T. T. Kawano, J. Cui, Y. Koezuka, I. Toura, Y. Kaneko, K. Motoki, H. Ueno,
R. Nakagawa, H. Sato, E. Kondo, H. Koseki, M. Taniguchi, Science 1997,
278, 1626–1629.
[35] B. Kratzer, T. G. Mayer, R. R. Schmidt, Eur. J. Org. Chem. 1998, 291–298.
[36] A highly pure form of phytosphingosine hydrochloride [(2S,3S,4R)-2-
aminooctadecane-1,3,4-triol] from the appropriate yeast fermentation
broth is commercially available in large quantities for a reasonable
price.
[4] M. Kronenberg, A. Rudensky, Nature 2005, 435, 598–604.
[5] M. Taniguchi, M. Harada, S. Kojo, T. Nakayama, H. Wakao, Annu. Rev. Im-
munol. 2003, 21, 483–513.
[6] G. Giaccone, C. J. Punt, Y. Ando, R. Ruijter, N. Nishi, M. Peters, B. M. von
Blomberg, R. J. Scheper, H. J. van der Vliet, A. J. van den Eertwegh, M.
Roelvink, J. Beijnen, H. Zwierzina, H. M. Pinedo, Clin. Cancer Res. 2002, 8,
3702–3709.
[37] P. B. Alper, S.-C. Hung, C.-H. Wong, Tetrahedron Lett. 1996, 37, 6029–
6032.
[7] A. Ishikawa, S. Motohashi, E. Ishikawa, H. Fuchida, K. Higashino, M.
Otsuji, T. Iizasa, T. Nakayama, M. Taniguchi, T. Fujisawa, Clin. Cancer Res.
2005, 11, 1910–1917.
[8] M. J. Smyth, D. I. Godfrey, Nat. Immunol. 2000, 1, 459–460.
[9] M. A. Exley, L. Lynch, B. Varghese, M. Nowak, N. Alatrakchi, S. P. Balk,
Clin. Immunol. 2011, 140, 184–195.
[38] H. Takikawa, S. E. Muto, K. Mori, Tetrahedron 1998, 54, 3141–3150.
[39] R. M. Ndonye, D. P. Izmirian, M. F. Dunn, K. O. A. Yu, S. A. Porcelli, A.
Khurana, M. Kronenberg, S. K. Richardson, A. R. Howell, J. Org. Chem.
2005, 70, 10260–10270.
[40] O. Plettenburg, V. Bodmer-Narkevitch, C.-H. Wong, J. Org. Chem. 2002,
67, 4559–4564.
[10] K. Miyamoto, S. Miyake, T. Yamamura, Nature 2001, 413, 531–534.
[11] A. K. Singh, M. T. Wilson, S. Hong, D. Olivares-Villagꢃmez, C. Du, A. K.
Stanic, S. Joyce, S. Sriram, Y. Koezuka, L. Van Kaer, J. Exp. Med. 2001, 194,
1801–1811.
[41] W. Du, S. S. Kulkarni, J. Gervay-Hague, Chem. Commun. 2007, 2336–
2338.
[42] J. Tatai, P. Fugedi, Org. Lett. 2007, 9, 4647–4650.
[43] C. R. Berkers, H. Ovaa, Trends Pharmacol. Sci. 2005, 26, 252–257.
&8
&
www.chembiochem.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 0000, 00, 1 – 10
ÝÝ
These are not the final page numbers!

a-GalCer Analogues for NKT Cell Activation
[44] A. Banchet-Cadeddu, E. Henon, M. Dauchez, J.-H. Renault, F. Monneaux,
A. Haudrechy, Org. Biomol. Chem. 2011, 9, 3080–3104.
B. M. von Blomberg, R. J. Scheper, A. J. van den Eertwegh, Int. J. Cancer
2005, 116, 87–93.
[45] T.-N. Wu, K.-H. Lin, Y.-J. Chang, J.-R. Huang, J.-Y. Cheng, A. L. Yu, C.-H.
Wong, Proc. Natl. Acad. Sci. USA 2011, 108, 17275–17280.
[50] S. Fujii, K. Shimizu, V. Klimek, M. D. Geller, S. D. Nimer, M. V. Dhodapkar,
Br. J. Haematol. 2003, 122, 617–622.
[46] B. A. Sullivan, N. A. Nagarajan, G. Wingender, J. Wang, I. Scott, M. Tsuji,
R. W. Franck, S. A. Porcelli, D. M. Zajonc, M. Kronenberg, J. Immunol.
2010, 184, 141–153.
[51] Y. J. Kim, H. J. Ko, Y. S. Kim, D. H. Kim, S. Kang, J. M. Kim, Y. Chung, C. Y.
Kang, Int. J. Cancer 2008, 122, 2774–2783.
[47] Y. Hayakawa, K. Takeda, H. Yagita, M. J. Smyth, L. Van Kaer, K. Okumura,
I. Saiki, Blood 2002, 100, 1728–1733.
[48] S. Motohashi, T. Nakayama, Immunotherapy 2009, 1, 73–82.
[49] J. W. Molling, W. Kçlgen, H. J. van der Vliet, M. F. Boomsma, H. Kruizen-
ga, C. H. Smorenburg, B. G. Molenkamp, J. A. Langendijk, C. R. Leemans,
Received: January 4, 2012
Published online on && &&, 0000
ChemBioChem 0000, 00, 1 – 10
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
&
9
&
ÞÞ
These are not the final page numbers!

FULL PAPERS
M.-H. Hsieh, J.-T. Hung, Y.-W. Liw, Y.-J. Lu,
C.-H. Wong,* A. L. Yu,* P.-H. Liang*
Compounds with biased cytokine pro-
files: A selection of a-GalCer derivatives
with modifications in the acyl side chain
and/or at the galactose 6’-position were
synthesized and assayed for NKT cell ac-
tivation. Gal-6’-phenylacetamide-based
derivatives substituted with p-nitro (32),
p-tert-butyl (34), or o-, m-, or p-methyl
groups (40–42) were found to promote
polarization biased towards T helper
1 (TH1) in vitro.
&& – &&
Synthesis and Evaluation of Acyl-
Chain- and Galactose-6’’-Modified
Analogues of a-GalCer for NKT Cell
Activation
&10
&
www.chembiochem.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 0000, 00, 1 – 10
ÝÝ
These are not the final page numbers!