Structure-based discovery of glycolipids for CD1d-mediated NKT cell activation: Tuning the adjuvant versus immunosuppression activity

Article abstract of DOI:10.1021/ja062740z

Introduction of an aromatic group into the fatty acyl chain of α-GalCer modulates the activity and selectivity of IFN-γ/IL-4 secretion through CD1d-mediated activation of NKT cells. Compound 14-16 are more potent than α-Galcer and biased for IFN-γ than fo

Full text of DOI:10.1021/ja062740z

Published on Web 06/23/2006
Structure-Based Discovery of Glycolipids for CD1d-Mediated NKT Cell
Activation: Tuning the Adjuvant versus Immunosuppression Activity
Masakazu Fujio,§ Douglass Wu,§ Raquel Garcia-Navarro,† David D. Ho,† Moriya Tsuji,*,† and
Chi-Huey Wong*,§
The Scripps Research Institute, La Jolla, California 92037, and Aaron Diamond AIDS Research Center, The
Rockefeller UniVersity, New York, New York 10016
Received April 19, 2006; E-mail: wong@scripps.edu; mtsuji@adarc.org
R-Galactosylceramide (R-GalCer; 1)1 is a derivative of a marine
sponge natural product and has been shown to possess antitumor
activity.2 Further studies revealed that R-GalCer activates natural
killer T (NKT) cells through CD1d-mediated antigen presentation.
Unlike other T cells, NKT cells are restricted to a nonmajor
histocompatibility complex (MHC) molecule, CD1d, which binds
lipids and glycolipids instead of peptides.3,4 R-GalCer, when bound
to CD1d, activates NKT cells by means of T cell receptor (TCR)
recognition to produce T helper 1 (Th1) and T helper 2 (Th2)
cytokines, such as interferon-γ (IFN) and interleukin 4 (IL-4),
respectively.5
Figure 1. Structure of R-GalCer 1 and fatty acyl chain analogues 2-16.
The production of Th1 cytokines is thought to correlate with
the antitumor, antiviral/bacterial and adjuvant effects of R-GalCer,
while Th2 cytokine production is thought to correlate with the
amelioration of certain autoimmune diseases (e.g., type 1 diabetes
and multiple sclerosis).6-8 However, the efficacy of R-GalCer has
been limited because of the reciprocal inhibition exhibited by Th1
and Th2 cytokines. In a Phase I study, R-GalCer was ineffective
in the treatment of solid tumors possibly because the therapeutic
effects of IFN-γ were hindered by IL-4 and thus gave no net
benefit.9 Therefore, compounds which increase the selectivity
toward either Th1 or Th2 cytokines responses may be more
advantageous.10,11
Structure-activity relationship studies of R-GalCer analogues
revealed that truncation of the fatty acyl or the phytosphingosine
group allows for more selective Th2 response.12,13 OCH, a
phytosphingosine truncated analogue of R-GalCer, selectively
induces Th2 cytokines from NKT cells.14 Also, introduction of
double bonds into the fatty acyl chain of R-GalCer, C20:1 cis/trans
and C20:2 analogues, seemed to bias toward Th2 responses.15 The
C-glycoside analogue of R-GalCer seems to bias toward the Th1
response but is less potent than R-GalCer during short incubation
times.16,17
Although numerous factors likely play a role in shifting the
cytokine profile, stability of the CD1d/glycolipid complex may be
a contributing cause. Time-course cytokine release profiles show
that, upon VR14 NKT-cell activation, IL-4 levels peak within 2 h
while IFN-γ levels peak 12 h after treatment with R-GalCer in
mice.15,16 IFN-γ production requires longer TCR stimulation than
IL-4.13 Full activation of NKT cells seems to require a high-affinity
TCR interaction with a long half-life.18,19 Thus, a model to bias
the cytokine profile toward a Th2 response would be to generate
less stable glycolipid/CD1d complexes to shorten NKT cell
stimulation times. Conversely, a biased Th1 response could be
created by prolonged stimulation of TCRs on NKT cells through
increasingly stable glycolipid/CD1d complexes.
Figure 2. IFN-γ and IL-4 secretion by human NKT cell line when
stimulated by 10 ng/mL of indicated glycolipids. IFN-γ and IL-4 release
was measured after 16 h of culture. Results are expressed as relative
activities as mean of duplicate assays ( standard deviation. Representative
data from one of three experiments are shown.
We have found that the bacterial glycolipid, GalA-GSL, activates
NKT cells through CD1d-mediated antigen presentation.20 The
crystal structure of mouse CD1d/GalA-GSL21 and activity study
revealed that the 4′-OH of phytosphingosine chain and the galactose
6-OH are important in addition to the 3′-OH of phytosphingosine
chain and galactose 2-OH to activate NKT cells by comparison
with human CD1d/R-GalCer22 and mCD1d/PBS-25.23 These crystal
structures confirmed that the lipid chains of R-linked glycosphin-
golipids were accommodated in two hydrophobic pockets. Numer-
ous aromatic side-chain residues line the binding groove and could
be utilized to make specific aromatic interactions. Specifically,
Tyr73, Phe114, Phe70, and Trp114 of the A′ pocket seemed most
accessible. This observation prompted us to introduce an aromatic
group into the fatty acyl chain of R-GalCer with the hope of making
a tighter binding glycolipid to CD1d.
In this study, we designed and synthesized a variety of fatty acyl
chain analogues of R-GalCer (Figure 1) based on the available
CD1d/glycolipid structures and evaluated NKT cell activation by
measuring cytokine release profiles. We synthesized various fatty
acyl chain analogues (Supporting Information Scheme S1 and
Figure S1) and evaluated their ability to activate human VR24NKT
cells as measured by their IFN-γ and IL-4 production relative to
R-GalCer20 (Figure 2). Shorter chain analogues 2-4 and 7-9
showed only moderate activity. However, 2-thienylacetyl analogue
5 and 2-naphthylacetyl analogue 6 demonstrated almost equal or
better cytokine secretion than R-GalCer. These results suggest that
the spacer chain length was too short to interact with the aromatic
residues in the CD1d hydrophobic groove. These results prompted
§ The Scripps Research Institute.
† The Rockefeller University.
9
9022
J. AM. CHEM. SOC. 2006, 128, 9022-9023
10.1021/ja062740z CCC: $33.50 © 2006 American Chemical Society

C O M M U N I C A T I O N S
more potent than R-GalCer and also exhibit a stronger Th1 cytokine
response, probably due to enhanced binding to CD1d or selective
interaction with CD8+ versus CD4+ NKT cells,26 although this issue
has been debated.27 The origin of the enhanced potency and Th1
selectivity remains to be fully addressed. This study provides a new
direction for the development of novel glycolipid-based immuno-
therapeutic agents which are more potent than R-GalCer and are
able to exhibit greater Th1-type cytokine profiles.
Figure 3. IFN-γ and IL-4 secretion by human NKT cell line when
stimulated by 10 ng/mL of indicated glycolipids. IFN-γ and IL-4 release
was measured after 16 h of culture. Results are expressed as relative
activities as mean of duplicate assays ( standard deviation. Representative
data from one of three experiments are shown.
Acknowledgment. We would like to thank Prof. I. A. Wilson
for providing hCD1d protein. We also thank the Skaggs Institute
and the NIH for research support.
Supporting Information Available: Synthesis of fatty acyl chain
analogues 2-35; amounts of IFN-γ and IL-4 secretion in response to
these glycolipids; results from competitive binding study of hCD1d
with glycolipids; complete refs 9, 15, 18, 20b, and 22. This material is
available free of charge via the Internet at http://pubs.acs.org.
References
(1) Morita, M.; Motoki, K.; Akimoto, K.; Natori, T.; Sakai, T.; Sawa, E.;
Yamaji, K.; Koezuka, Y.; Kobayashi, E.; Fukushima, H. J. Med. Chem.
1995, 38, 2176-2187.
(2) Hayakawa, Y.; Godfrey, D. I.; Smyth, M. J. Curr. Med. Chem. 2004, 11,
241-252.
(3) Joyce, S.; Woods, A. S.; Yewdell, J. W.; Bennink, J. R.; De Silva, A. D.;
Boesteanu, A.; Balk, S. P.; Cotter, R. J.; Brutkiewicz, R. R. Science 1998,
279, 1541-1544.
(4) Moody, D. B.; Porcelli, S. A. Nat. ReV. Immunol. 2003, 3, 11-22.
(5) Kronenberg, M. Annu. ReV. Immunol. 2005, 23, 877-900.
(6) Godfrey, D. I.; MacDonald, H. R.; Kronenberg, M.; Smyth, M. J.; Van
Kaer, L. Nat. ReV. Immunol. 2004, 4, 231-237.
(7) Taniguchi, M.; Harada, M.; Kojo, S.; Nakayama, T.; Wakao, H. Annu.
ReV. Immunol. 2003, 21, 483-513.
(8) Gonzalez-Aseguinolaza, G.; Van Kaer, L.; Bergmann, C. C.; Wilson, J.
M.; Schmieg, J.; Kronenberg, M.; Nakayama, T.; Taniguchi, M.; Koezuka,
Y.; Tsuji, M. J. Exp. Med. 2002, 195, 617-624.
(9) Giaccone, G.; et al. Clin. Cancer Res. 2002, 8, 3702-3709.
(10) Smyth, M. J.; Godfrey, D. I. Nat. Immunol. 2000, 1, 459-460.
(11) Berkers, C. R.; Ovaa, H. Trends Pharmacol. Sci. 2005, 26, 252-257.
(12) Goff, R. D.; Gao, Y.; Mattner, J.; Zhou, D.; Yin, N.; Cantu, C., III; Teyton,
L.; Bendelac, A.; Savage. P. B. J. Am. Chem. Soc. 2004, 126, 13602-
13603.
Figure 4. Superimposition of the docking results of fatty acyl chain
analogues 14-16 with hCD1d. The R2 helix is removed for clarity. 14
(green); 15 (yellow); 16 (orange).
us to evaluate analogues with a longer spacer chain (Figure 3).
Elongation of the spacer chain length drastically enhanced overall
cytokine production, while piperidine analogue 12 diminished their
activity. 4-Fluorophenylpropionyl analogue 11 demonstrated more
potent cytokine production than 4-CF3 analogue 10 and 4-OMe
analogue 8. Of these compounds, the longer alkyl chain analogues
14-16 are 4 times more potent and biased for IFN-γ secretion.
These results show that it is possible to potentiate and tune the
Th1/Th2 cytokine profile by the introduction of a terminal aromatic
group with an appropriate length of spacer chain. To visualize the
interactions between the designed analogues and human CD1d,
Autodock 3.024 was utilized to model the binding of selected
compounds in the hCD1d hydrophobic groove (Figure 4). 13-16
were individually docked, and their results did not vary significantly
from the crystal structure of R-GalCer bound to hCD1d.22 In each
case, the phytosphingosine tail extended into the F′ pocket and the
A′ pocket was occupied by the fatty acyl chain with the galactose
headgroup presented in nearly the same configuration. Introduction
of a terminal phenyl group in the R-GalCer analogues seemed to
promote additional specific interactions between 14, 15, and the
phenol ring of Tyr73 and between 16 and Trp40. Docking of
compounds with a shorter spacer chain on the fatty acyl tail showed
limited interaction between terminal functional groups and aromatic
residues in the A′ pocket.
(13) Oki, S.; Chiba, A.; Yamamura, T.; Miyake, S. J. Clin. InVest. 2004, 113,
1631-1640.
(14) Miyamoto, K.; Miyake, S.; Yamamura, T. Nature 2001, 413, 531-534.
(15) Porcelli, S. A.; et al. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 3383-
3388.
(16) Schmieg, J.; Yang, G.; Franck, R. W.; Tsuji, M. J. Exp. Med. 2003, 198,
1631-1641.
(17) Yang, G.; Schmieg, J.; Tsuji, M.; Franck, R. W. Angew. Chem., Int. Ed.
2004, 43, 3818-3822.
(18) Kronenberg, M.; et al. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 12254-
12259.
(19) Cantu, C., III; Benlagha, K.; Savage, P. B.; Bendelac, A.; Teyton, L. J.
Immunol. 2003, 170, 4673-4682.
(20) (a) Kinjo, Y.; Wu, D.; Kim, G.; Xing, G. W.; Poles, M. A.; Ho, D. D.;
Tsuji, M.; Kawahara, K.; Wong, C. H.; Kronenberg, M. Nature 2005,
434, 520-525. (b) Wong, C. H.; et al. Proc. Natl. Acad. Sci. U.S.A. 2005,
102, 1531-1536. (c) Xing, G. W.; Wu, D.; Poles, M. A.; Horowitz, A.;
Tsuji, M.; Ho, D. D.; Wong, C. H. Bioorg. Med. Chem. 2005, 13, 2907-
2916.
(21) Wu, D.; Zajonc, D. M.; Fujio, M.; Sullivan, B. A.; Kinjo, Y.; Kronenberg,
M.; Wilson, I. A.; Wong, C. H. Proc. Natl. Acad. Sci. U.S.A. 2006, 103,
3972-3977.
We also examined hCD1d binding of compounds 14-16 using
isoelectric focusing (IEF) electrophoresis25 (see Supporting Infor-
mation). Compounds 14-16 demonstrated more potent inhibition
of GT1b-hCD1d binding than R-GalCer and a less potent analogue
4, supporting our hypothesis.
In conclusion, we have found that introduction of an aromatic
group to the fatty acyl chain greatly enhances IFN-γ/IL-4 secretion
and enables the tuning of Th1/Th2 cytokine profile, possibly through
alteration of glycolipid/CD1d complex stability. Compounds 14-
16 represent the first examples of NKT cell agonists which are
(22) Cerundolo, V.; et al. Nat. Immunol. 2005, 6, 819-826.
(23) Zajonc, D. M.; Cantu, C., III; Mattner, J.; Zhou, D.; Savage, P. B.;
Bendelac, A.; Wilson, I. A.; Teyton, L. Nat. Immunol. 2005, 6, 810-
818.
(24) Morris, G. M.; Goodsell, D. S.; Halliday, R. S.; Huey, R.; Hart, W. E.;
Belew, R. K.; Olson, A. J. J. Comput. Chem. 1998, 19, 1639-1662.
(25) Cantu, C., III; Benlagha, K.; Savage. P. B.; Bendelac, A.; Teyton, L. J.
Immunol. 2003, 170, 4673-4682.
(26) Takahashi, T.; Chiba, S.; Nieda, M.; Azuma, T.; Ichihara, S.; Shibata,
Y.; Juji, T.; Hirai, H. J. Immunol. 2002, 168, 3140-3144.
(27) Ho, L. P.; Urban, B. C.; Jones, L.; Ogg, G. S.; McMichael, A. J. J Immunol.
2004, 172, 7350-7358.
JA062740Z
9
J. AM. CHEM. SOC. VOL. 128, NO. 28, 2006 9023