Galacto-configured aminocyclitol phytoceramides are potent in vivo invariant natural killer T cell stimulators

Article abstract of DOI:10.1021/ja202610x

A new class of α-galactosylceramide (αGC) nonglycosidic analogues bearing galacto-configured aminocyclitols as sugar surrogates have been obtained. The aminocyclohexane having a hydroxyl substitution pattern similar to an α-galactoside is efficiently obtained by a sequence involving Evans aldol reaction and ring-closing metathesis with a Grubbs catalyst to give a key intermediate cyclohexene, which has been converted in galacto-aminocyclohexanes that are linked through a secondary amine to a phytoceramide lipid having a cerotyl N-acyl group. Natural Killer T (NKT) cellular assays have resulted in the identification of an active compound, HS161, which has been found to promote NKT cell expansion in vitro in a similar fashion but more weakly than αGC. This compound stimulates the release of Interferon-γ (IFNγ) and Interleukin-4 (IL-4) in iNKT cell culture but with lower potency than αGC. The activation of Invariant Natural Killer T (iNKT) cells by this compound has been confirmed in flow cytometry experiments. Remarkably, when tested in mice, HS161 selectively induces a very strong production of IFN-γ indicative of a potent Th1 cytokine profile. Overall, these data confirm the agonist activity of αGC lipid analogues having charged amino-substituted polar heads and their capacity to modulate the response arising from iNKT cell activation in vivo.

Full text of DOI:10.1021/ja202610x

ARTICLE
pubs.acs.org/JACS
Galacto-Configured Aminocyclitol Phytoceramides Are Potent in Vivo
Invariant Natural Killer T Cell Stimulators
Youssef Harrak,† Carolina M. Barra,§ Antonio Delgado,†,‡ A. Raꢀul Casta~no,*,§ and Amadeu Llebaria*,†
†Research Unit on BioActive Molecules (RUBAM), Departament de Química Biomꢁedica, Institut de Química Avancada de Catalunya
)
(IQAC-CSIC), Jordi Girona 18-26, 08034 Barcelona, Spain
‡Unitat de Química Farmacꢁeutica (Unitat Associada al CSIC), Facultat de Farmꢁacia, Universitat de Barcelona (UB), Avgda. Joan XXIII,
s/n, 08028 Barcelona, Spain
§Grupo de Inmunología Molecular, Institut de Biotecnologia i Biomedicina, Universitat Autꢁonoma de Barcelona, Bellaterra 08193
Cerdanyola del Vallꢁes, Barcelona, Spain
S Supporting Information
b
ABSTRACT: A new class of R-galactosylceramide (RGC)
nonglycosidic analogues bearing galacto-configured aminocy-
clitols as sugar surrogates have been obtained. The aminocy-
clohexane having a hydroxyl substitution pattern similar to an
R-galactoside is efficiently obtained by a sequence involving
Evans aldol reaction and ring-closing metathesis with a Grubbs
catalyst to give a key intermediate cyclohexene, which has been
converted in galacto-aminocyclohexanes that are linked through
a secondary amine to a phytoceramide lipid having a cerotyl N-acyl group. Natural Killer T (NKT) cellular assays have resulted in the
identification of an active compound, HS161, which has been found to promote NKT cell expansion in vitro in a similar fashion but
more weakly than RGC. This compound stimulates the release of Interferon-γ (IFNγ) and Interleukin-4 (IL-4) in iNKT cell culture
but with lower potency than RGC. The activation of Invariant Natural Killer T (iNKT) cells by this compound has been confirmed
in flow cytometry experiments. Remarkably, when tested in mice, HS161 selectively induces a very strong production of IFN-γ
indicative of a potent Th1 cytokine profile. Overall, these data confirm the agonist activity of RGC lipid analogues having charged
amino-substituted polar heads and their capacity to modulate the response arising from iNKT cell activation in vivo.
nvariant Natural Killer T (iNKT) cells are a distinct lympho-
cyte class that regulates a broad range of immune responses.1
relevant for TCR recognition. Moreover, the galactose hydroxy-
methyl group, not involved in RGC CD1d binding and TCR
recognition, is replaced in HS44 by a hydroxyl group. The in vitro
activity found for HS44 is in accordance with the iNKT activity of
the structurally related R-glucosylphytoceramide,6 which is an
agonist weaker than RGC. This interaction mode suggested the
design of the new aminocyclitol derivatives 1and 2(alsodesignedas
HS161 and HS138, respectively, see Chart 1) having substituents to
better mimic galactose and specially address the important role of
40OH in RGC recognition by a TCR protein and therefore with
potentially higher activity on NKT stimulation. Other related
compounds resulting from the substitution of glycosidic oxygen
atoms by carbon are known, such as R-carbaGal-GC11 or R-C-GC12
and found to strongly stimulate iNKT cells exhibiting biological
profiles different from RGC. In this article, we describe the synthesis
of 1 and 2 and the results obtained in their biological assays.
I
They recognize glycolipid antigens presented by the MHC
class I-like protein CD1d.2 iNKT cells are involved in the regula-
tion of autoimmunity and 3 the response to tumors4 or microbial
infections.5 The glycolipid R-galactosylceramide (RGC) is the
prototypical lipid antigen presented by CD1d that stimulates the
invariant T cell receptor (TCR) expressed by iNKT cells.6 How-
ever, the exceptional potency of RGC on iNKT cell stimulation is
associated to different side effects, arising from the simultaneous
production of Th1 and Th2 type cytokines, which have opposite cell
polarizing functions, followed by an unresponsive phase before
recovering homeostatic equilibirum.7 This has prompted the design
and synthesis of RGC analogues to improve its biological properties
and modulate the strong induced response.8,9 In this context, we
have previously described the in vitro activity of aminocyclitol
phytoceramides as iNKT cell activators.10 These nonglycosidic
RGC mimetics bear a polyhydroxylated aminocyclohexane as a
galactose surrogate, which is linked through a secondary amine to a
phytoceramide lipid (Chart 1).
’ RESULTS
Synthesis of Aminocyclitol Compounds. The cyclitol phy-
toceramides 1 and 2 were obtained by the regioselective opening
of aziridine 3 with cyclitol amines 4 or 5, using a method
One of these aminocyclitol phytoceramides, HS44, activated
iNKT cells despite that it better matches an R-gluco than R-galacto
configuration because of the equatorial orientation of 40OH, a
change that would not influence ligand binding to CD1d but is
Received: March 23, 2011
Published: July 05, 2011
r
2011 American Chemical Society
12079
dx.doi.org/10.1021/ja202610x J. Am. Chem. Soc. 2011, 133, 12079–12084
|

Journal of the American Chemical Society
ARTICLE
Chart 1. Aminocyclitol and Related Glycolipid Mimetics of
R-Galactosylceramide
Scheme 2. Synthesis of Key Cyclohexene Intermediates for
Aminocyclitolsa
a Reagents and conditions: (a) Bu2BOTf, Et3N, 83%; (b) 2 M LiBH4,
Et2O, 95%; (c) Grubbs’second-generation catalyst, CH2Cl2, 91%; (d)
NaH, BnBr, DMF, 94%; (e) Pyr-trityl BF4, CH3CN; (f) NaH, BnBr,
DMF; (g) p-TsOH, MeOH (89% combined eꢀg).
The plan for the synthesis of the necessary aminocyclitols 4
and 5 was to obtain them from key intermediates cyclohexene 12
and its related epoxide 13.
Scheme 1. Synthesis of rGC Analogues 1 and 2a
As depicted in Scheme 2, cyclohexene 12 was efficiently
obtained in four steps involving an aldol reaction and ring-
closing metathesis (RCM) as strategic steps. We started from
aldehyde 14, obtained from D-xylose,14 which was treated with
Evans’ oxazolidinone 15,15 following conditions similar to those
reported in the literature,16,17 to give the anti-aldol adduct 16 as
the only isolated compound in high yield. Reduction of the chiral
auxiliary amide gave the corresponding hydroxymethyl derivative
17, having all the substituents with the required stereochemistry
for the galacto carbasugar synthesis. RCM with second-gene-
ration Grubbs catalyst18,19 gave cyclohexene 18a, which was
benzylated to give the key intermediate 12. The epoxidation of
compound 12 under several conditions (m-CPBA/CH2Cl2;
m-CPBA/NaH2PO4/THF; or Caro’s salt in K2CO3/acetone)
was unsuccessful, recovering the starting material unaltered.
When these conditions were applied to diol 18a, starting material
was consumed, but only trace amounts of the desired epoxide
were observed, together with unidentified degradation products.
On the other hand, m-CPBA epoxidation of tri-O-benzyl deri-
vative 18d, arising from transient trityl protection of the primary
hydroxyl group in 18a, benzylation of the secondary hydroxyl
group of 18b, and final trityl removal from 18c (Scheme 2), was
also fruitless since the starting material was recovered in all cases.
Apparently, the presence of an axial substituent at the 40 position
(see Scheme 2 or Chart 1 for numbering) affects the reactivity of
cyclohexenes 18d and 12 toward epoxidation. Moreover, the
stability of the expected epoxides formed from 18a under the
above reaction conditions differs also from that observed in
gluco-related systems (epimeric with 18a at the 40 position),
whose epoxidation has been reported to take place without
incident.14
a Reagents and conditions: (a) MeCN, reflux; 6, 81%; 7, 83%; (b) PhSH,
Cs2CO3, MeCN, 25 °C; 8, 74%; 9, 83%; (c) EDC, cerotic acid, THF,
reflux; 10, 71%; 11, 79%; (d) H2, Pd/C, MeOH, HCl (cat.) 25 °C; 1,
81%; 2, 85%.
developed to obtain 1-amino-1-deoxyphytosphingosine deriva-
tives,13 and employed previously in the preparation of HS44.10
The reaction of 3 with aminocyclitols 4 and 5 was straightfor-
ward with exclusive attack at the less substituted aziridine carbon
atom to give 6 and 7, respectively, in good yields. Desulfonyla-
tion to diamines 8 and 9 and selective N-acylation at the primary
amine with cerotic acid gave the hydroxyl-protected products 10
and 11 (Scheme 1), which were transformed in the final
aminocyclitol lipids 1 and 2 after the simultaneous removal of
acetal and benzyl protecting groups by catalytic hydrogenolysis
in acidic media.
12080
dx.doi.org/10.1021/ja202610x |J. Am. Chem. Soc. 2011, 133, 12079–12084

Journal of the American Chemical Society
ARTICLE
Scheme 3. Synthesis of Aminocyclitol 4 from 12a
a Reagents and conditions: (a) (1) BH3,THF; then H2O2, aq. NaOH, 0
to 25 °C; (2) IBX, DMSO, 25 °C, 75% (2 steps). (b) CeCl3 7H2O,
3
NaBH4, MeOH, ꢀ50 °C, chromatography; 21, 53%; 22, 27%. (c) From
21, (1) MsCl, Et3N, CH2Cl2, 0 to 25 °C; (2) NaN3, DMF, 90 °C; 75%
(2 steps). (d) LiAlH4, THF 0 to 25 °C, 93%.
Scheme 4. Synthesis of Epoxide 13 and Aminocyclitol 5 from
12a
Figure 1. In vitro cytokine induction in cultured splenocytes. IFNγ and
IL-4 present in the culture supernatants at day 4 after compound
incubation was measured by ELISA. * Denotes statistical significance
of low cytokine levels.
The major cis-diol 25 was converted to epoxide 13 by a two-step
sequence involving the reaction with excess 2-acetoxyisobutyryl
bromide (MattocksꢀMoffat reagent)20ꢀ22 to give a mixture of
regioisomeric trans-cyclohexane bromoacetates, which were not
isolated but directly converted into tetrabenzylated epoxide 13 after
K2CO3/MeOH treatment. This stable epoxide was shown to be a
convenient precursor for aminocyclitols 4and 5(Scheme4). Onone
hand, it was transformed to cyclohexanol 21, a precursor of 4 by
reductive opening. Thus, treatment of epoxide 13 with LiAlH4
afforded a 1:2 mixture of the previously obtained alcohols 21 and
22, confirming the same stereochemistry on the new hydroxyl
groups introduced in compound 12 by the hydroboration and
hydroxylation routes. When LiHEt3B was used as the hydride source,
a 3:2 mixture of 21 and 22 was obtained favoring the desired
regioisomer. On the other hand, the galacto-configurated epoxide 13
was transformed in aminocyclitol 5 by nucleophilic opening with
sodium azide. Unexpectedly, the epoxide 13 was rather unreactive.
After trying different conditions, it was found that NH4Cl activation
with heating at reflux in ethoxyethanol gave excellent yields of a 1:1
mixture of azidoalcohols 26 and 27, which could be separated by
chromatography. We later discovered that the use of LiClO4 in
acetonitrile at reflux in this reaction favored the formation of the
desired regioisomer 27 over 26 (4:1) in good yields. Finally, the
reduction of the azido group in 27 afforded aminocyclitol 5.
Biological Activity. The biological activity of the aminocycli-
tols 1 (HS161) and 2 (HS138) as iNKT agonists was assayed in
vitro by measuring the production of Th1 and Th2 prototypic
cytokines, Interferon-γ (IFNγ) and Interleukin- 4 (IL-4) of
spleen cultures. As depicted in Figure 1, HS161 induced IFNγ
production, as mesaured at day 4 of culture. The amount of IFNγ
averaged 1/3 of the amount produced by RGC-stimulated
cultures, at concentrations in which it reaches the plateau of
maximum IFNγ production (low micromolar range) and closer
to OCH, a Th2 agonist analogue of RGC with a shorter
sphingosine chain (see Chart 1).23 We used RGC at 100 ng/
mL because at this concentration it reaches the maximum of
IFNγ production, and the use of higher concentrations has
a Reagents and conditions: (a) OsO4, NMNO, 25 °C, acetone/H2O
95%; (b) (1) BrCOC(CH3)2OAc; (2) K2CO3, MeOH, 0 to 25 °C 91%;
(c) NaN3, EtOCH2CH2OH, 150 °C, 93%, 26/27 1:1; or NaN3, LiClO4
MeCN, reflux, 89%, 26/27 1:4; (d) from 27, LiAlH4, THF 0 to 25 °C,
93%; (e) LiAlH4, THF, 50 °C, 21, 20% 22, 43%; or LiHBEt3, THF ꢀ15
°C; 21, 55% 22, 37%.
For the synthesis of aminocylitol 4, we tried a route not
requiring epoxide 13 (Scheme 3). The hydroborationꢀoxidation
of 12 gave a mixture of four alcohols that were directly oxidized to
an inseparable mixture of cyclohexanones 19 and 20, which were
reduced with Luche reagent to a 2:1 mixture of β-alcohols 21 and
22 that were isolated after chromatography. The hydroboration
of cyclohexenes 18aꢀd did not improve the results obtained
with 12. Cyclohexanol 21 was transformed after conventional
mesylation and azidation in the epimeric azide 23, which was
converted to the desired amine 4 by LiAlH4 reduction.
For the synthesis of aminocyclitol 5, we needed epoxide 13 or
a cyclohexane compound with an equivalent substitution pattern.
Due to the difficulties found in the direct formation of epoxide 13
from alkenes 12 or 18a,d, we resorted to the osmium-promoted
dihydroxylation of cyclohexene 12, which proceeded in good
yield to give a 4:1 mixture of diastereomeric tetra-O-benzyl-cis-1,
2-diols 24 and 25 (Scheme 4) that were separated by chromatography.
12081
dx.doi.org/10.1021/ja202610x |J. Am. Chem. Soc. 2011, 133, 12079–12084

Journal of the American Chemical Society
ARTICLE
Figure 3. In vivo cytokine induction in mice. Amounts of 1 μg of
aminocyclitols and 100 ng of RGC and OCH were i.p. injected into
mice, and IFNγ and IL-4 in serum were measured after 2 and 24 h.
Figure 2. Proliferation of NKT cells after compound activation. Flow
cytometry analysis of splenocyte cultures incubated for 6 days with the
different compounds at indicated concentrations is shown. NKT
percentages (double positive cells) were calculated among electronically
gated life lymphocytes.
sensitive techniques, HS138 induced iNKT specific proliferation
with a much lower potency than RGC and IFNγ, and IL-4
induction was some 30-fold lower. Overall, these analyses con-
clusively show that HS161 is recognized by iNKT cells, inducing
their activation, proliferation, and production of cytokines in a
way similar to the prototypic CD1d ligand RGC, although with
less efficiency and lower potency, while HS138 induces a much
weaker response.
To more meaningfully address the biological activity of these
analogues and their physiological consequences, we resorted to
in vivo studies. We intraperitoneally administered 1 μg of the
aminocyclitol analogues and measured the levels of IFNγ and IL-
4 in serum at two time points, 2 and 24 h, the peaks of maximum
production of these cytokines. As controls, we injected mice with
a dose of 100 ng of RGC or OCH, an amount that gives a
response close to the maximum attainable (less than 50%
difference) but is more useful for comparison of the in vivo
potential of the compounds.
detrimental effects in cell cultures, including high cell mortality
and consequent variability in results. In contrast to HS161, the
structurally related aminocyclitol phytoceramide HS138 did not
induce a significant production of IFNγ.
On the contrary, production of prototypic Th2 cytokine IL-4
was extremely weak, although statistically significant, after
HS161 stimulation (Figure 1). OCH induced more than 5-fold
higher levels of IL-4 than HS161 at equal concentrations. Stimu-
lation with a suboptimal concentration of RGC also induced
higher levels of IL-4, corroborating the weaker recognition of
HS161 by iNKT cells. Again, as was the case for IFNγ, HS138
did not induce significant production of IL-4 above background
levels.
The in vitro data strongly suggest that the HS161 compound
was recognized by iNKT cells inducing their activation in a way
that seems to be more biased toward the Th1 pathway, while
HS138 was not recognized and could not activate iNKT re-
sponses in a meaningful manner. To confirm that the observed
production of cytokines is due to the recognition of analogues by
NKT cells, the splenocyte cultures were analyzed by flow
cytometry to detect the presence of NKT cells. As shown in
Figure 2, HS161 induced the expansion of double positive NKT
cells, up to three times relative to negative controls, at the highest
concentrations tested. This activation is similar to the weaker
proliferation induced by OCH and weaker than that induced by
RGC at concentrations at the plateau level. As in the cytokine
assays, HS138 increased the number of iNKT cells, but so weakly
that it is not statistically significant. This result is in accordance
with a complementary study performed with this compound,
which showed that it is a relatively weak iNKT agonist when
compared to RGC and R-C-GC.24 In this study, using more
Surprisingly, HS161 induced a very strong production of
IFNγ upon in vivo administration, as detected in the serum of
treated animals 24 h later (Figure 3). This is up to two times
higher than the amount induced by the suboptimal dose of 100 ng
of RGC. This stimulatory potential is higher than anticipated from
the in vitro assays, and it may be related to the expected prolonged
in vivo half-life of the compound. Similarly to splenocyte cultures,
HS138 was essentially unable to induce IFNγ production, except
for marginal levels just above background.
In opposition, IL-4 induction in vivo was weaker than expected
(Figure 3). Basically, a very little amount of IL-4 was detected in
the serum after HS161 administration. In this case, OCH was as
strong an inducer of cytokine production as RGC at a suboptimal
dose, as expected from its known propensity to bias Th2
response by iNKT cells.23 Again, treatment with HS138 pro-
duced amounts of IL-4 barely over background levels with no
statistical relevance at 2 h, and essentially no IL-4 in serum was
found at 24 h in any treatment.
12082
dx.doi.org/10.1021/ja202610x |J. Am. Chem. Soc. 2011, 133, 12079–12084

Journal of the American Chemical Society
ARTICLE
’ DISCUSSION
these important regulatory T cells by aminocyclitol analogues of
RGC and that becomes a potent Th1 activator when used in vivo,
despite its lower intrinsic capacity to be recognized by NKT cells.
A more complete assessment of the biological properties of
HS161 is guaranteed.
In the search for new agonists of iNKT cells that may be useful
for immunotherapeutical approaches, including the possibility of
obtaining new vaccine adjuvants, we resorted to modifying the
polar head structure of the prototypical agonist RGC by using
aminocyclitols as galactose replacements. As pointed out before,
the structural modifications introduced in 1 and 2 are expected to
be differentially recognized by iNKT cells, to induce their
activation in a more controllable and specific way, and to increase
the relatively weak activation of HS44. These new sugar mimetics
with R-galacto configuration could be obtained by an interesting
synthetic sequence involving Evans aldol chemistry coupled with
ring-closing metathesis to give cyclohexene 12 that was trans-
formed into the final amines 4 and 5 by conventional reactions.
The coupling of these aminocyclitols to the phytosphingosine
aziridine 3, deprotection reactions, and N-acylation allowed us to
obtain the desired RGC glycolipid mimetics.
’ ASSOCIATED CONTENT
S
Supporting Information. Experimental procedures for
b
the synthesis and biological testing of the compounds, character-
ization data, and NMR spectra of molecules. This material is
available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
amadeu.llebaria@iqac.csic.es; raul.castano@uab.cat.
The experiments here described show that HS161 is a very
efficient activator of Th1 response, as deduced from the high
level of IFNγ in the serum of treated animals, compared with the
very weak IL-4 induction. Although it could not be deduced from
its in vitro results, the strong HS161 in vivo activity was actually
awaited in the design of the aminocyclitol phytoceramides, which
pursued ligands capable of activating iNKT cells but resistant to
the degradative action of glycosidases and, hence, with increased
bioavailability in vivo. HS161 is recognized by iNKT, and its
presumed stability would allow it to activate iNKT cells in a more
sustained manner, thus becoming a strong Th1 inducer, in a
mechanism similar to that proposed for the analogue R-C-GC.25
Interestingly, the ether analogue R-carbaGal-GC is also a Th1
biased iNKT cell agonist with increased metabolic stability over
RGC .26 Thus, HS161 has an efficacy as high as RGC as an inducer
of Th1 response but does not activate Th2 cytokines, so avoiding
one of the main problems in the therapeutical application of RGC,
the simultaneous and potent induction of contradictory responses.
In addition, the lower potency of HS161 suggests that induction of
anergythat followsexcessive NKT activation, the other mainpitfall
of RGC biological functionality, could be much lower, if not totally
avoided, adding to the advantageous possibilities of HS161 as an
immunotherapeutic tool.
In contrast, HS138 is weakly recognized by iNKT, either in
vitro or in vivo. This compound incorporates an additional
hydroxyl group (60OH) in the cyclitol ring with respect to
HS161, in place of the pyranose oxygen in the original galactose.
We previously showed that this extra substitution in the weak
activator HS44 significantly diminished the recognition by
iNKT, probably due to a distorting interaction with VRPro28
of the TCR. Despite having increased the capacity of the polar
head to be recognized by using a galactocyclitol, the introduction
of a 60OH still seems to preclude TCR interaction and therefore
the activation of iNKT cells in a very relevant manner. So,
relatively minor structural alterations at this position are not well
tolerated. The weak activity of HS138 is in accordance with
Tashiro et al.,11 who reported that a similar modification in R-
carbaGal-GC analogues abolished its agonist iNKT activity.
’ ACKNOWLEDGMENT
This work was supported by MICINN (Project CTQ2008ꢀ
01426/BQU), Fondos Feder (EU), Generalitat de Catalunya
(2005SGR01063), UAB (PRP2007ꢀ06), and CSIC (200480E561).
The authors thank E. Dalmau for HRMS analysis and Dr, M. Egido-
Gabas and Dr. Amaya Castro for analytical support. Y.H. thanks
MICINN for a Juan de la Cierva fellowship.
’ REFERENCES
(1) Bendelac, A.; Savage, P. B.; Teyton, L. Annu. Rev. Immunol. 2007,
25, 297–336.
(2) Matsuda, J. L.; Mallevaey, T.; Scott-Browne, J.; Gapin, L. Curr.
Opin. Immunol. 2008, 20, 358–368.
(3) Ronchi, F.; Falcone, M. Front. Biosci. 2008, 13, 4827–4837.
(4) Molling, J. W.; Moreno, M.; van der Vliet, H. J.; van den
Eertwegh, A. J.; Scheper, R. J.; von Blomberg, B. M.; Bontkes, H. J.
Clin. Immunol. 2008, 129, 182–194.
(5) Brigl, M.; Brenner, M. B. Semin. Immunol. 2010, 22, 79–86.
(6) Kawano, T.; Cui, J.; Koezuka, Y.; Toura, I.; Kaneko, Y.; Motoki,
K.; Ueno, H.; Nakagawa, R.; Sato, H.; Kondo, E.; Koseki, H.; Taniguchi,
M. Science 1997, 278, 1626–1629.
(7) Parekh, V. V.; Wilson, M. T.; Olivares-Villagomez, D.; Singh,
A. K.; Wu, L.; Wang, C. R.; Joyce, S.; Van Kaer, L. J. Clin. Invest. 2005,
115, 2572–2583.
(8) Wu, D.; Fujio, M.; Wong, C. H. Bioorg. Med. Chem. 2008,
16, 1073–1083.
(9) Wun, K. S.; Borg, N. A.; Kjer-Nielsen, L.; Beddoe, T.; Koh, R.;
Richardson, S. K.; Thakur, M.; Howell, A. R.; Scott-Browne, J. P.; Gapin,
L.; Godfrey, D. I.; McCluskey, J.; Rossjohn, J. J. Exp. Med. 2008,
205, 939–949.
(10) Harrak, Y.; Barra, C. M.; Bedia, C.; Delgado, A.; Castano, A. R.;
Llebaria, A. ChemMedChem 2009, 4, 1608–1613.
(11) Tashiro, T.; Nakagawa, R.; Hirokawa, T.; Inoue, S.; Watarai, H.;
Taniguchi, M.; Mori, K. Bioorg. Med. Chem. 2009, 17, 6360–6373.
(12) Yang, G.; Schmieg, J.; Tsuji, M.; Franck, R. W. Angew. Chem.,
Int. Ed. 2004, 43, 3818–3822.
(13) Harrak, Y.; Llebaria, A.; Delgado, A. Eur. J. Org. Chem.
2008, 4647–4654.
(14) Hansen, F. G.; Bundgaard, E.; Madsen, R. J. Org. Chem. 2005,
70, 10139–10142.
(15) Evans, D. A.; Sjogren, E. B.; Bartroli, J.; Dow, R. L. Tetrahedron
Lett. 1986, 27, 4957–4960.
(16) Fleming, K. N.; Taylor, R. E. Angew. Chem., Int. Ed. 2004,
43, 1728–1730.
(17) Taylor, R. E.; Hearn, B. R.; Ciavarri, J. P. Org. Lett. 2002,
4, 2953–2955.
’ CONCLUSIONS
In summary, the synthesis of phytoceramide compounds
structurally related to RGC having aminocyclitols as galactose
mimetics resulted in the identification of HS161, a new non-
glycosidic NKT agonist that improves the capacity to activate
12083
dx.doi.org/10.1021/ja202610x |J. Am. Chem. Soc. 2011, 133, 12079–12084

Journal of the American Chemical Society
ARTICLE
(18) Scholl, M.; Ding, S.; Lee, C. W.; Grubbs, R. H. Org. Lett. 1999,
1, 953–956.
(19) Plumet, J. Mini Rev. Org.Chem. 2007, 201–216.
(20) Mattocks, A. R. J. Chem. Soc. 1964, 4840–4845.
(21) Mattocks, A. R. J. Chem. Soc. 1964, 1918–1930.
(22) Russell, A. F.; Greenberg, S.; Moffatt, J. G. J. Am. Chem. Soc.
1973, 95, 4025–4030.
(23) Miyamoto, K.; Miyake, S.; Yamamura, T. Nature 2001, 413,
531–534.
(24) Patel, O.; Cameron, G.; Pellicci, D. G.; Liu, Z.; Byun, H.-S.;
Franck, R. W.; Casta~no, A. R.; Llebaria, A.; Bittman, R.; Porcelli, S. A.;
Godfrey, D. I.; Rossjohn, J., submitted.
(25) Schmieg, J.; Yang, G.; Franck, R. W.; Tsuji, M. J. Exp. Med.
2003, 198, 1631–1641.
(26) Tashiro, T.; Sekine-Kondo, E.; Shigeura, T.; Nakagawa, R.;
Inoue, S.; Omori-Miyake, M.; Chiba, T.; Hongo, N.; Fujii, S.; Shimizu,
K.; Yoshiga, Y.; Sumida, T.; Mori, K.; Watarai, H.; Taniguchi, M. Int.
Immunol. 2010, 22, 319–328.
12084
dx.doi.org/10.1021/ja202610x |J. Am. Chem. Soc. 2011, 133, 12079–12084