Efforts toward rational design of Th2-bias immune stimulator through modification on D-Gal-C-4 of α-GalCer derivative

Article abstract of DOI:10.1016/j.tet.2021.132168

α-Galactosylceramide (α-GalCer) initiates immune response by activating iNKT cells in a CD1d restricted way. However, its clinical application is quite limited mainly because of the simultaneous secretion of antagonizing Th1- and Th2-type cytokines. With the help of molecular computation, we conducted a rational design of glycolipid to expect a Th2-bias immune response. Accordingly, some of the designed D-galactopyranosyl lipids were synthesized, and their abilities to induce Th2 cytokines through stimulating iNKT cells were investigated both in vitro and in vivo. We found that modification on Gal-C-4 of glycolipids could possibly increase Th2-selecivity, although the cytokine productivity was suppressed ascribing to the weaken interaction of CD1d and TCR in CD1d/glycolipid/TCR triplex.

Full text of DOI:10.1016/j.tet.2021.132168

Tetrahedron 89 (2021) 132168
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Efforts toward rational design of Th2-bias immune stimulator through
modification on D-Gal-C-4 of a-GalCer derivative
,
, b, c, *
Chuanfang Zhao a b, Liyan Hu c, Juan Ma a, Wei Cui b, Yang Jing c, Yuguo Du a
a State Key Laboratory of Environmental Chemistry and Eco-toxicology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences,
Beijing, 100085, China
b School of Chemical Sciences, University of Chinese Academy of Sciences, Beijing, 100049, China
c National Engineering Research Center for Carbohydrate Synthesis, Jiangxi Normal University, Nanchang, 330022, Jiangxi, China
a r t i c l e i n f o
a b s t r a c t
Article history:
a
-Galactosylceramide (a-GalCer) initiates immune response by activating iNKT cells in a CD1d restricted
Received 18 February 2021
Received in revised form
8 April 2021
Accepted 13 April 2021
Available online 30 April 2021
way. However, its clinical application is quite limited mainly because of the simultaneous secretion of
antagonizing Th1- and Th2-type cytokines. With the help of molecular computation, we conducted a
rational design of glycolipid to expect a Th2-bias immune response. Accordingly, some of the designed D-
galactopyranosyl lipids were synthesized, and their abilities to induce Th2 cytokines through stimulating
iNKT cells were investigated both in vitro and in vivo. We found that modification on Gal-C-4 of glyco-
lipids could possibly increase Th2-selecivity, although the cytokine productivity was suppressed
ascribing to the weaken interaction of CD1d and TCR in CD1d/glycolipid/TCR triplex.
© 2021 Elsevier Ltd. All rights reserved.
Keywords:
KRN7000
a
-GalCer
Th2 cytokines
iNKT cells
Autoimmune disease
1. Introduction
stimulation facilitating anticancer activities [14,15]. However, the
key factor which selectively induces autoimmune Th2-type cyto-
D
-Galactopyranosyl lipid molecules, as represented by the well-
kines is quite confusing in view of the facts that either shortening
sphingosine chain of -GalCer or modifying 4-/6-OH group of D-Gal
known leading compound KRN7000, could stimulate natural killer
T cells to produce Th1 and Th2 cytokines [1e4]. An acceptable
mechanism of this process has been suggested based on X-ray
crystal structures of both the binary complex of KRN7000/CD1d
and the ternary complex of TCR/KRN7000/CD1d (Fig. 1A) [5e7].
Disappointingly, KRN7000 did not present a satisfactory thera-
peutic activity as it evokes simultaneous secretion of antagonizing
Th1 and Th2 cytokines [8e10]. Since then, many glycolipid mole-
cules have been designed through modification on each part of
KRN7000 skeleton (glycosyl, sphingosine chain, acyl chain, glyco-
sidic bond) in order to bias either Th1 or Th2 production of iNKT
cells [11e13].
a
residue could possibly increase Th2 bias [16e19]. Very few Gal-4-
OH modified substrates have been designed so far, mainly due to
the worries of vanishing crucial hydrogen bond between Gal-4-OH
and ASN-30 of TCR shown in the crystallographic study [7].
Inconsistent results were also obtained from those reports of
attempted modification on Gal-4-OH with alkyl or aryl derivatives
[18,20e22]. For example, it was reported that a-GalCer derivatives
with axial/equatorial/eliminated 4-OH group presented the similar
stimulation pattern on iNKTcells, implying that some modifications
on Gal-4-OH might be tolerated with respected to the immunos-
timulating activity [4,20e23]. We have successfully designed and
Previous reports suggested that a strength binding between
a
-
synthesized some glycolipids with modified 6-OH on D-gal-
Galactosylceramide -GalCer) and CD1d is in favor of Th1
(
a
actopyranosyl residue [19], and the subsequent in vivo tests
exhibited better Th2 selectivity than that of the well-known Th2-
type stimulator OCH [16,24,25]. Taking advantages of this prom-
ising experience, we here would like to disclose our efforts on
* Corresponding author. State Key Laboratory of Environmental Chemistry and
Eco-toxicology, Research Center for Eco-Environmental Sciences, Chinese Academy
of Sciences, Beijing, 100085, China.
rational trimming of
D-galactopyranosyl C-4 to improve the selec-
tivity of Th2-type cytokine production.
E-mail address: duyuguo@rcees.ac.cn (Y. Du).
https://doi.org/10.1016/j.tet.2021.132168
0040-4020/© 2021 Elsevier Ltd. All rights reserved.

C. Zhao, L. Hu, J. Ma et al.
Tetrahedron 89 (2021) 132168
Fig. 1. Crystal structure of the ternary complex among CD1d (labeled in green), TCR (labeled in orange), and glycolipid (labeled in cyan) for KRN7000 (A) or OCH (B), and the
potential binding sites of TCR toward OCH were labeled in magenta (B).
2. Results and discussion
propyl at the outer end of C-4-amide structure initiated a hydro-
phobic interaction with TCR VAL-50 (7). Theoretically, all these
additional interactions could enhance the substrate affinity toward
TCR and thus in favor of Th2-type selectivity. To check the validity
of this prediction, we synthesized some of the compounds and
tested their stimulation abilities in vivo.
With careful analysis of the published molecules showing Th2
activity [16,18,19,26,27], we hypothesized that glycolipid present-
ing a relatively strong affinity with TCR or a relatively weak inter-
action with CD1d is beneficial to generate Th2-type cytokines on
condition of a stable TCR/glycolipid/CD1d ternary complex. A
The synthetic route was initially explored with the preparation
of compound 2, as shown in Scheme 1. To introduce a C-4-amine
shorten chain on either acyl or sphingosine part of a-GalCer could
slightly reduce its binding ability with CD1d and thus lead to an
improved Th2-type cytokines production, as represented by the
molecule OCH [26,27]. Further structure analyses of the crystal TCR/
OCH/CD1d (Fig. 1B) revealed the existence of anion residues (ASP-
29, ASP-69), cation residues (LYS-68, LYS-71, ARG-95), polar resi-
dues (THR-27, ASN-30, GLN-52, SER-97), and hydrophobic residue
(VAL-50) around the binding sites of TCR. These facts encouraged us
group on
2,3,6-tri-O-benzyl-1-thio-
starting material. The hydroxyl group of 8 underwent mitsunobu
reaction with diphenylphosphoryl azide (DPPA) to give C-4-
inverted azido compound 9 in 80% yield. Convergently, the cer-
D
-galactopyranosyl residue, we chose 4-methylphenyl
-glucopyranoside (8) [28] as the
b-D
amide analogue 10 was synthesized from D-mannose following our
previous work [19] in 12 steps with 15% overall yield. Coupling of
thioglycoside donor 9 and ceramide acceptor 10 under standard
glycosylation conditions [29] of NIS/TMSOTf obtained a low yield of
the desired compound 11, ascribing to the poor solubility of cer-
amide derivative 10 in CH2Cl2 and quick decomposition of donor 9
under the reaction conditions. Fortunately, the same glycosylation
was successfully preceded under “inverse procedure” [30] through
a slow addition of donor 9 into a pre-mixed solution of 10/NIS/
to put aryl, hydrophobic or hydrophilic groups on Gal-C-4 of
a-
GalCer and to predict the binding activities with the aid of Surflex-
Dock in SYBYL-X software (Tripos Inc., St. Louis, MO, USA). In our
preliminary modeling study, we also observed that introducing an
amide group on Gal-C-4 could provide extra H-bond and polar
interaction relative to its ether counterpart, offering
D-galactopyr-
anosyl 4-OH a potential modifiable site in -GalCer molecule. Based
a
on all these information, we proposed some glycolipid structures
(Fig. 2) with higher TotalScore in docking computation (Table 1) to
expect an improved TCR-binding ability in TCR/glycolipid/CD1d
triplex.
TMSOTf in CH2Cl2 at 0 ꢀC. The expecting
from a separable -mixture of 3:1 in a yield of 48%. The azide
group of 11 was reduced with Me3P in pyridine/H2O at 50 ꢀC to give
the key building block 4-amino- -galactopyranosyl derivative 12 in
a-glycoside 11 was isolated
a,b
D
Besides hydrophobic interactions of ceramide residue with A/F
pockets of CD1d, the precise structure analysis (Fig. 3) of TCR/OCH/
CD1d triplex proclaimed the existence of H-bond interactions be-
an isolated yield of 85%. Attempted azido-reduction of 11 with PPh3
generated only 15% yield of the product under the same reaction
conditions owing to the bulk environment around Gal-C-4.
Condensation of 12 with indolyl-3-propionic acid in the presence
of EDCI/HOBt/TEA in DMF at rt obtained amide derivative 13 in 73%
yield over two steps from 11. Treatment of 13 with HCl (3 M) in THF/
MeOH at rt successfully removed isopropylidene group, however,
the subsequent Pd(OH)2/C catalyzed hydrogenolysis of 13 with H2
gave an unexpected mixture without separable compound 2. We
believed that the bulk groups on both C-1 and C-4 warped the
tween CD1d ASP-80 and sphingosine 4-OH, ASP-153 and
D-gal-
actopyranosyl 3-OH, as well as the H-bond affinities from
D
-
galactopyranosyl OHs toTCR residues ASN-30 and GLY-96. Different
from the known OCH, our designed glycolipids presented more
potential binding sites toward TCR. For examples, the C-4-amide
could form H-bond with ASN-30 of TCR (3, 7), and the hydroxyl
group of phenol residue offered extra H-bond with ASP-51 (6). The
appended anion -COO- group at the end of C-4 substitution
generated electrostatic interaction with LYS-68 of TCR (1). More
conformation of D-galactose residue and thus prevented the normal
reduction of 2,3,6-tri-O-benzyl groups. Alternatively, cleavage of
acetal group from 11 was performed smoothly with aqueous 70%
AcOH in THF (v/v 2:1), generated diol 14 in 92% yield. Compound 14
was subjected to hydrogenation with H2 under the conditions of
importantly, the newly introduced aryl group could generate
HeN interaction with ASN-30 (5), … cation interaction with LYS-
68 (3, 6), or interaction with HIS-31 (5). Introducing iso-
p …
p
p
… p
2

C. Zhao, L. Hu, J. Ma et al.
Tetrahedron 89 (2021) 132168
Fig. 2. Structures of the designed glycolipids by modifying Gal-C-4 of OCH.
Table 1
in vitro results also indicated that glycolipids 3, 4, or 7 exhibited
lower stimulating activity on iNKT cells than OCH.
The TotalScore of TCR/glycolipid/CD1d triplex.
To test their selectivity in cytokine production, OCH, 3, 4, 5 and 7
(8 nmol/mice) were intravenous (i.v.) injected into C57BL/6 (B6)
mice (6 mice per group), and the mice blood were collected at 0,
3 h, 6 h, 9 h, 12 h, 24 h, 36 h, and 48 h after injection. The quantity of
Ligand
OCH
1
2
3
TotalScore
Ligand
TotalScore
15.45
4
19.44
18.95
5
20.37
19.77
6
19.22
20.77
7
18.52
IFN-
Fig. 6A and Fig. 6B, the glycolipids OCH, 3, 4, 5, and 7 could stim-
ulate iNKT cells to produce IFN- and IL-4 in vivo, reaching the
highest serum concentration of IFN- and IL-4 at 12 h and 3 h after
injection, respectively. The peak value ratio and AUC ratio for the
production of IFN- and IL-4 (Fig. 6C and D) indicated that com-
pounds 3, 4, 5, or 7 provided comparable or better Th2 selectivity,
but lower productivity for both IFN- and IL-4 than OCH itself. The
relatively lower percentage of IFN- expressing iNKTcells, 0.6e0.7%
g and IL-4 in serum was characterized by ELISA. As shown in
g
Pd(OH)2/C in aqueous HCl (3 M), and the affording key intermediate
15 was used for the next reaction without further purification.
Condensation of 15 with indolyl-3-propionic acid in the presence of
EDCI and HOBt furnished compound 2 in a yield of 73% over two
steps.
g
g
g
The synthetic new compound 2 was subjected to cytokine
g
production test in vivo by measuring the serum IL-4 and IFN-
g
for the designed compounds versus 1.0% for the reference OCH
(Fig. 5), may be responsible for the diminished cytokine secretion.
This phenomenon has also been observed from other Gal-C-4
modified glycolipids, e.g. trimming C-4 with AcHN-, HOCH2CH2O-
levels in C57BL/6 mice via intraperitoneal injection. Compared to
KRN7000 control, compound 2 induced a significantly compressed
production of IFN-g and a moderately decreased IL-4 productivity,
thus exhibiting an increased selectivity of Th2-type cytokine
secretion (Fig. 4).
Encouraged by this exciting result, we next completed the
syntheses of other designed molecules (Scheme 2). Based on the
successful synthesis of 2, we followed the same synthetic strategy
to prepare the designed compounds 3e7 using 14 as the core
building block. Thus, hydrogenation of 14 (/ 15), followed by a
direct condensation of 15 with 3-phenyl propionic acid, 4-phenyl
butyric acid, 3-(4-benzyloxyphenyl) propionic acid, or 5-
methylhexanoic acid in the presence of co-catalyst EDCI/HOBt
generated the expecting compounds 3, 4, 5 and 7 in the yield of
79%, 78%, 82%, or 78%, respectively, over two steps. Pd/C catalyzed
H2 reduction of compound 5 generated phenol 6, which could be
oxidized in partial during the purification process.
, or aryl groups led to a reduced production of IFN-
their C-4-OH analogues [18,21,22].
g and IL-4 than
Curious about the weak stimulation activity on iNKT cells, we
mimicked TCR/glycolipid/CD1d interaction with molecular dy-
namic computation using AMBER14 software [33]. In order to
compare the CD1d and TCR affinity to the newly designed glyco-
lipid, respectively, we also calculated the binding energy of glyco-
lipid/CD1d and TCR/glycolipid with the help of MMPBSA method in
AMBER software [34].
Molecular dynamic computational results (Table 2) indicated
that introducing phenyl or iso-propyl residue on Gal-C-4 of OCH
(forming 3, 4, 5 and 7) increased their TCR affinity (shown lower Ec
value). However, these compounds weakened their affinities to
CD1d in the ternary model in contrast to the binary complex,
showing bigger Ea values than the corresponding GBa values in
Table 2. In particular, the higher Eac values of Gal-C-4 modified
glycolipids suggest the weakened interaction in CD1d/glycolipid/
TCR ternary complex by comparison with CD1d/OCH/TCR triplex.
This may disfavor the iNKT stimulation and decrease the produc-
The synthesized new compounds 3, 4 and 7 were subjected to
the biological evaluation of stimulating cytokine production both
in vitro and in vivo. In the in vitro experiments, glycolipids OCH, 3, 4
and 7 presented higher percentage of IFN-g expressing iNKT cells
than the blank significantly, indicating that the designed glyco-
lipids could activate iNKT cells in a CD1d-restricted way (Fig. 5A).
This has also been confirmed by IL-2 production in the same
experimental system as shown in Fig. 5B, while IL-2 is believed to
be an indicator of iNKT stimulation activity [31,32]. Besides, the
tion of IFN-
g and IL-4 [22]. To further investigate the binding
interaction of TCR/glycolipd/CD1d, we performed structure ana-
lyses and energy decomposition analyses for the ternary complex.
Energy decomposition of Ea in the triplex system displayed a
3

C. Zhao, L. Hu, J. Ma et al.
Tetrahedron 89 (2021) 132168
Fig. 3. Structures of the designed glycolipids (labeled out in cyan) with CD1d (green cartoon) and TCR (orange cartoon) in the ternary complex.
4

C. Zhao, L. Hu, J. Ma et al.
Tetrahedron 89 (2021) 132168
Scheme 1. Reagents and conditions: (a) PPh3, DIAD, DPPA, THF, 0 ꢀC - rt, 80%; (b) 10, NIS, TMSOTf, 4 Å molecular sieves, DCM, 0 ꢀC, inverse procedure, 48%; (c) Me3P, pyridine/H2O,
50 ꢀC; (d) 3-Indolepropionic acid, EDCI, HOBt, TEA, DMF, rt, 73% for 13 from 11, 73% for 2 from 14; (e) HCl (3 M), THF/MeOH, rt, 95%; (f) H2 (3 atm), Pd(OH)2/C, HCl (3 M, 2 eq.), THF/
MeOH (v/v, 1/1), rt; (g) 70% AcOH/THF (v/v 2/1), reflux, 12 h, 92%.
similar interaction pattern between the substrate (OCH, 3, 4, 5 or 7)
and CD1d, indicating the dominant roles of PHE-70, TYR-73, PHE-
77, ASP-80, ASP-153, GLY-155, THR-156, THR-159, VAL-160, and
LEU-163 in the binding process (Fig. 7). Further structural analyses
revealed that glycolipids bind with CD1d by hydrogen bond inter-
action with residues ASP-80, ASP-153, GLN-154, THR-156, THR-159
3. Conclusions
In summary, we have designed some new Th2-biasing glyco-
lipids through modification on D-galactopyranosyl C-4 of a-GalCer
derivatives under the guidance of TCR structure analysis and
docking computation. Both in vitro and in vivo experiments indi-
cated that the designed compounds could trigger Th2-type im-
(Fig. 8),
p … HeC interaction with residues PHE-70, TYR-73, PHE-
77, PHE-171, and the hydrophobic interaction with residues LEU-
13, VAL-30, MET-69, VAL-72, ILE-81, LEU-150, VAL-160, LEU-163,
LEU-164.
Energy decomposition of Ec in the ternary system (Fig. 9) indi-
cated that the residues PRO-28, ASP-29, ASN-30, VAL-50, ARG-95,
GLY-96, and LEU-99 of TCR contributed fundamental affinities to-
ward glycolipid in ternary complex. As also shown in Fig. 8,
hydrogen bond was the major binding format among TCR residues
of ASP-29, ASN-30, ASP-94, ARG-95, GLY-96, and glycolipids, while
mune response by inducing less IFN-g production and a persistent
low level secretion of IL-4 as well. Binding energy analyses under
dynamic calculation suggested that the C-4 modified compounds
could enhance TCR affinity to the binary, but probably sacrifice the
stability of TCR/glycolipid/CD1d triplex at the same time. This, in
turn, would weaken their abilities to stimulate iNKT cells and result
in a relatively low productivity of cytokines. Our results suggested
that the TCR affinity to the glycolipid, as well as the ternary stability
of TCR/glycolipid/CD1d, should be considered in designing of Th2-
type glycolipid. Our findings would be helpful to the molecular
a potential
p … HeC interaction between Gal-C-4 phenyl and VAL-
50 alkyl groups, a hydrophobic interaction between LEU-99 and
sphingosine residue were also observed. Surprisingly, dynamic
calculation suggested that LYS-68 did not strengthen TCR affinity
design through modification of a-GalCer in the related research
works.
via an expected cation …
p interaction with 3 or 4, instead, it
4. Experimental section
decreased TCR affinity to the Gal-C-4 modified glycolipids as shown
a positive energy in dynamic analysis. This is further confirmed by a
repelling interaction between LYS-68 and C-4 substituent in the
ternary structure after dynamic simulation in comparison with
docking results (Fig. 3). Our designed glycolipids (3, 4, 5, 7) pre-
sented extra binding sites with TCR residues (ASP-29, VAL-50, LYS-
68, and ASP-94) indicating a closer tie of TCR/glycolipid in the
ternary complex, and this would be more favorable to initiate a
Th2-bias immune response. The bigger Eac value of TCR and CD1d
also suggested that groups introduced onto Gal-C-4 may weaken
the interaction between TCR and CD1d, and thus formed less stable
ternary (CD1d/glycolipid/TCR) relative to TCR/OCH/CD1d, resulting
in a weak iNKT stimulating activity.
4.1. General experimental
Commercially available materials were used without further
purification if not noted. Solvents were dried with established
procedures before use. Flash chromatography (FC) was performed
on silica gel (200e300 meshes) column and the reactions were
monitored by thin-layer chromatography (TLC) on gel F254 plates.
Optical rotations were measured using a polarimeter with a ther-
mally jacketed 5 cm cell at approximately 20 ꢀC. Melting points
were recorded in open capillary tubes with a melting point appa-
ratus (Buchi Melting Point M ꢁ 560). IR (FT-IR) spectra were
recorded on a PerkinElmer BXII spectrometer with KBr pellets.
High-resolution mass spectrometry data (HRMS) were acquired
using a Q-TOF analyzer in acetone or methanol as solvent. 1H NMR,
5

C. Zhao, L. Hu, J. Ma et al.
Tetrahedron 89 (2021) 132168
Fig. 4. Bioactivity of compound 2 showing good Th2 selectivity.
Scheme 2. (a) H2 (3 atm), Pd(OH)2/C, HCl (3 M, 2 eq.), THF/MeOH (v/v, 1/1), rt; then 3-phenyl propionic acid, EDCI, HOBt, TEA, DMF, rt, 79% over two steps; (b) H2 (3 atm), Pd(OH)2/C,
HCl (3 M, 2 eq.), THF/MeOH (v/v, 1/1), rt; then 4-phenyl butyric acid, EDCI, HOBt, TEA, DMF, rt, 78% over two steps; (c) H2 (3 atm), Pd(OH)2/C, HCl (3 M, 2 eq.), THF/MeOH (v/v, 1/1),
rt; then 3-(4-benzyloxyphenyl) propionic acid, EDCI, HOBt, TEA, DMF, rt, 82% over two steps; (d) H2, Pd/C, MeOH; (e) H2 (3 atm), Pd(OH)2/C, HCl (3 M, 2 eq.), THF/MeOH (v/v, 1/1), rt;
then 5-methylhexanoic acid, EDCI, HOBt, TEA, DMF, rt, 78% over two steps.
13C NMR were measured on 400 MHz or 100 MHz spectrometers
(NMR in CDCl3 with TMS as an internal standard). Chemical shifts
4.1.1. Synthesis of 9
To a solution of 8 (356 mg，0.64 mmol) in THF (10 mL) was
added Ph3P (0.37 g, 1.41 mmol), DIAD (0.28 mL, 1.41 mmol) and
DPPA (0.30 mL, 1.39 mmol) at 0 ꢀC under nitrogen. The mixture was
stirred at this condition for another 2 h, then concentrated, and the
residue was extracted between ethyl acetate and water. The organic
(d) are given in ppm relative to residual solvent (usually chloro-
form; d 7.26 for 1H NMR or 77.0 for decoupled 13C NMR), and
coupling constants (J) in Hz. Multiplicity is tabulated as s for singlet,
d for doublet, t for triplet, q for quadruplet, and m for multiplet.
6

C. Zhao, L. Hu, J. Ma et al.
Tetrahedron 89 (2021) 132168
Fig. 5. The percentage of IFN-gþ iNKT in the total iNKT cells (A) and the content of IL-2 (B) in cultured splenocytes on treatment with compounds OCH, 3, 4 and 7, respectively.
Fig. 6. Mice serum cytokines IFN-
(n ¼ 6). The relative activity with respect to IFN-
g
(A) and IL-4 (B) on treatment with OCH, 3, 4, 5, or 7 at designed time post injection through i. v. Administration. There were 6 mice at each group
and IL-4 production was determined based on peak value ratio (C) and AUC ratio (D), taking OCH as the reference.
g
layer was further washed with saturated NaCl, then dried over
anhydrous Na2SO4, concentrated, and the resulting residue was
purified by silica gel column chromatography to get 9 (298 mg, 80%)
d
7.36e7.25 (m, 17H), 7.07 (d, J ¼ 8.0 Hz, 2H), 4.79 (d, J ¼ 12.0 Hz,
1H), 4.77e4.72 (m, 2H), 4.69 (d, J ¼ 12.0 Hz, 1H), 4.54e4.50 (m, 3H),
4.04 (d, J ¼ 1.2 Hz, 1H), 3.71e3.69 (m, 2H), 3.67e3.65 (m, 2H),
as a white solid. [
a]
20 ꢁ 9.4 (c 1, CHCl3); 1H NMR (400 MHz, CDCl3)
3.60e3.57 (m, 1H), 2.31 (s, 1H); 13C NMR (100 MHz, CDCl3)
d 138.3,
D
7

C. Zhao, L. Hu, J. Ma et al.
Tetrahedron 89 (2021) 132168
Table 2
137.9, 137.8, 137.7, 132.8, 129.8, 129.7, 128.72, 128.67, 128.6, 128.5,
128.2, 128.13, 128.08, 128.03, 88.4, 83.0, 77.3,76.0, 75.7, 73.9, 72.9,
68.8, 60.2, 21.3; HRMS (ESI): [C34H35N3O4S þ Na]þ calc. 604.2240;
found 604.2245.
The relative binding energy (kcal/mol) of glycolipids with CD1d and TCR. GBa stands
for the binding energy of CD1d and glycolipid in the binary complex, while Ea, Ec,
and Eac represents the binding energy between CD1d and glycolipid, glycolipid and
TCR, CD1d and TCR in the ternary complex, respectively.
GBa
Ea
Ec
Eac
4.1.2. Synthesis of 11
OCH
0
0
0
0
3
4
5
7
ꢁ7.71
ꢁ2.54
ꢁ8.44
ꢁ8.17
ꢁ2.37
0.13
2.30
ꢁ5.54
ꢁ11.71
ꢁ6.94
ꢁ12.11
ꢁ1.11
0.23
3.95
0.34
1.49
To
0.86 mmol) and 4 Å molecular sieves in CH2Cl2 (30 mL) was added
NIS (305 mg, 1.29 mmol) and TMSOTf (78
L, 0.43 mmol) at 0 ꢀC
under nitrogen. The solution of galactosyl donor 9 (750 mg,
a mixture of phytosphingosin derivative 10 (501 mg,
m
Fig. 7. Key CD1d residues contributed to glycolipid and CD1d interaction in the ternary complex.
Fig. 8. Structures of ternary complex at the lowest binding energy, showing glycolipids in cyan, CD1d in green, TCR in orange, and the H-bond interaction in red dash line.
8

C. Zhao, L. Hu, J. Ma et al.
Tetrahedron 89 (2021) 132168
d
172.87, 172.40, 138.33, 138.31, 137.62, 136.19, 128.56, 128.49,
128.35, 128.24, 127.95, 127.91, 127.73, 127.68, 127.01, 122.11, 122.05,
119.33, 118.59, 114.40, 111.16, 107.86, 99.40, 77.82, 76.71, 76.40,
75.89, 75.26, 73.45, 73.20, 71.40, 70.60, 69.76, 68.50, 48.63, 47.61,
37.19, 36.75, 31.94, 31.77, 29.73, 29.68, 29.61, 29.51, 29.43, 29.38,
28.79, 28.27, 26.28, 26.02, 25.64, 22.71, 22.64, 21.12, 14.15, 14.07;
HRMS (ESI): [C74H109N3O9
1206.8067.
þ
Na]þ calc. 1206.8056; found
4.1.4. Synthesis of 14
Compound 11 (126 mg, 0.121 mmol) in 70% AcOH (10 mL) and
THF (5 mL) was refluxed at 80 ꢀC for 18 h, then concentrated to a
syrup which was purified on silica gel column (MeOH/DCM: 1/20)
Fig. 9. The key TCR residues contributed to the binding interaction between glycolipid
and TCR in the ternary complex.
to give the product 14 (112 mg, 92%) as white solid. [
a
]
20 þ29.2 (c 1,
D
CHCl3); 1H NMR (400 MHz, CDCl3)
d
7.38e7.30 (m, 15H), 6.35 (d,
J ¼ 8.0 Hz, 1H), 4.88 (d, J ¼ 11.6 Hz, 1H), 4.79e4.72 (m, 3H), 4.64 (d,
J ¼ 11.6 Hz, 1H), 4.56 (d, J ¼ 11.6 Hz, 1H), 4.50 (d, J ¼ 11.6 Hz, 1H),
4.20e4.18 (m, 1H), 4.06 (d, J ¼ 3.2 Hz, 1H), 3.98e3.91 (m, 2H),
3.87e3.83 (m, 3H), 3.59e3.52 (m, 2H), 3.47 (m, 2H), 2.13 (t,
J ¼ 7.6 Hz, 1H), 1.60e1.54 (m, 1H), 1.47e1.45 (m, 1H), 1.25 (m, 51H),
1.29 mmol) in dried CH2Cl2 (10 mL) was added to the system
dropwisely through a drop funnel with constant pressure. The
resulting mixture was stirred for 2 h at 0 ꢀC and then at rt for 1 h.
The reaction was quenched with saturated Na2S2O3, and extracted
with saturated NaHCO3 (3 ꢂ 5 mL). The organic layer was filtered
over a pad of Celite, and the filter-cake was washed with CH2Cl2
(3 ꢂ 15 mL). The filtrate was concentrated under reduced pressure,
and the resulting residue was purified by silica gel column chro-
matography to give the product 11 (430 mg, 48%) as white solid.
0.88 (t, J ¼ 6.4 Hz, 6H); 13C NMR (100 MHz, CDCl3)
d 173.30, 137.75,
137.72, 137.52, 128.74, 128.72, 128.32, 128.17, 127.91, 99.14, 78.33,
76.30, 75.79, 74.72, 73.93, 73.43, 72.81, 70.25, 68.91, 67.87, 60.84,
49.54, 36.94, 33.37, 32.11, 31.99, 29.91, 29.89, 29.84, 29.73, 29.58,
29.55, 29.51, 25.91, 25.72, 22.88, 22.82, 14.32, 14.25; HRMS (ESI):
[C60H94N4O8 þ Na]þ calc. 1021.6964; found 1021.6961.
[
a
]
20 þ17.6 (c 0.5, CHCl3); 1H NMR (400 MHz, CDCl3)
d 7.39e7.30 (m,
D
15H), 6.05 (d, J ¼ 4.5 Hz, 1H), 4.86 (d, J ¼ 3.6 Hz, 1H), 4.81e4.77 (m,
3H), 4.64 (d, J ¼ 11.2 Hz, 1H), 4.56 (d, J ¼ 12.0 Hz, 1H), 4.50 (d,
J ¼ 12.0 Hz, 1H), 4.09e3.98 (m, 6H), 3.95e3.84 (m, 2H), 3.63e3.55
(m, 3H), 2.08e1.95 (m, 2H), 1.69 (s, 1H), 1.54 (t, J ¼ 6.2 Hz, 2H), 1.46
(d, J ¼ 8.8 Hz, 1H), 1.40 (s, 3H), 1.32 (s, 3H), 1.30e1.24 (m, 46H),
4.1.5. Synthesis of 2
To a solution of compound 14 (100 mg, 0.10 mmol) in THF/MeOH
(5 mL/5 mL) was added Pd(OH)2/C (80 mg) and aqueous HCl (3 M,
65 mL, 0.2 mmol). The reaction was shaken under hydrogen atmo-
sphere for 6 h, then neutralized with anion exchange resin to pH 7.
The reaction mixture was filtered over Celite, and the filter-cake
was washed with MeOH (2 ꢂ 2 mL). The filtrate was concen-
trated under reduced pressure. To the above crude product in dry
DMF (6 mL) was added Indole-3-propionic acid (18 mg,
0.90e0.85 (m, 6H); 13C NMR (100 MHz, CDCl3)
d 172.44, 138.10,
137.93, 137.40, 128.57, 128.50, 128.07, 128.00, 127.89, 127.74, 107.90,
99.34, 77.82, 76.46, 75.53, 73.80, 73.74, 72.86, 70.15, 69.26, 67.71,
61.09, 48.66, 31.95, 31.78, 29.73, 29.68, 29.62, 29.51, 29.44, 29.39,
28.81, 28.18, 26.32, 25.96, 25.70, 22.72, 22.65, 14.16, 14.08; HRMS
(ESI): [C63H98N4O8 þ Na]þ calc. 1061.7283; found 1061.7276.
0.094 mmol), triethylamine (32
mL, 0.24 mmol), EDCI (23 mg,
0.12 mmol) and HOBt (25 mg, 0.12 mmol) at 0 ꢀC. The reaction
mixture was stirred rt for 9 h, then concentrated to a syrup which
was partitioned between EtOAc (3 ꢂ 6 mL) and water. The com-
bined organic layer was washed with brine, dried over anhydrous
MgSO4, and concentrated under reduced pressure. The obtained
syrup was subjected to silica gel column chromatography (MeOH/
4.1.3. Synthesis of 13
To a solution of 11 (125 mg, 0.12 mmol) in pyridine (5 mL) and
water (0.5 mL) was added Me3P (11.5 mg, 0.15 mmol). The reaction
mixture was heated to 50 ꢀC for 12 h, concentrated, and the residue
was partitioned between water (20 mL) and EtOAc (3 ꢂ 10 mL). The
combined organic layer was washed with brine, dried over anhy-
drous Na2SO4, and concentrated to a syrup. To the mixture of this
syrup and 3-(3-indolyl) propionic acid (24.1 mg, 0.13 mmol) in
CH2Cl2 1:15) to give 2 as white solid (64 mg, 73% over two steps).
20
[
a
]
þ17.8 (c 0.5, CHCl3); 1H NMR (400 MHz, MeOD)
d 7.57 (d,
D
J ¼ 7.6, 1H), 7.31 (d, J ¼ 8.0, 1H), 7.08e7.04 (m, 2H), 7.01e6.97 (m,
1H), 4.83 (d, J ¼ 4.0 Hz, 1H), 4.37 (dd, J ¼ 4.8, 1.6 Hz, 1H), 4.20e4.61
(m,1H), 3.94e3.90 (m, 2H), 3.84 (dd, J ¼ 10.8, 4.8 Hz,1H), 3.68e3.63
(m, 2H), 3.59 (t, J ¼ 6.4 Hz, 1H), 3.53 (dd, J ¼ 8.7, 6.1, 2.1 Hz, 1H),
3.29e3.23 (m, 2H), 3.09e3.05 (m, 2H), 2.75e2.62 (m, 2H), 2.21 (t,
J ¼ 7.6 Hz, 2H), 1.64e1.54 (m, 4H), 1.39e1.22 (m, 54H), 0.92e0.87
dried DMF (6 mL) was added Et3N (32
mL, 0.24 mmol), EDCI
(22.9 mg, 0.12 mmol) and HOBt (16.2 mg, 0.12 mmol) at 0 ꢀC under
nitrogen. The reaction mixture was stirred at rt for 9 h, and then
concentrated. The residue was partitioned between water (12 mL)
and EtOAc (3 ꢂ 8 mL), and the organic layer was combined, washed
with brine, dried over MgSO4, and concentrated under reduced
pressure. The desired product 13 was obtained by silica gel column
(m, 6H); 13C NMR (100 MHz, MeOD)
d 177.6, 175.9, 138.2, 128.5,
123.1, 122.3, 119.5, 119.3, 115.1, 112.2, 107.5, 100.9, 75.5, 72.9, 71.5,
70.4, 70.2, 68.4, 62.5, 52.4, 51.7, 37.8, 37.2, 33.12, 33.08, 33.0, 30.80,
30.78, 30.76, 30.7, 30.6, 30.5, 30.4, 27.1, 26.7, 23.8, 23.7, 22.6, 14.49,
14.45; HRMS (ESI): [C50H87N3O9 þ Na]þ calc. 896.6335; found
896.6322.
chromatography (MeOH/DCM: 1/18) as an amorphous solid
20
(104 mg, 73% based on 11). [
a
]
þ22.1 (c 1, CHCl3); 1H NMR
D
(400 MHz, CDCl3)
d
7.57 (d, J ¼ 7.6, 1H), 7.41e7.20 (m, 16H), 7.16 (t,
J ¼ 8.0, 1H), 7.10 (t, J ¼ 8.0, 1H), 6.81 (d, J ¼ 1.6 Hz, 1H), 6.14 (d,
J ¼ 8.0 Hz, 1H), 5.44 (d, J ¼ 8.5, 1H), 4.83 (d, J ¼ 10.8 Hz, 1H),
4.76e4.71 (m, 2H), 4.68 (d, J ¼ 11.6 Hz, 1H), 4.56 (d, J ¼ 12.0 Hz, 1H),
4.49 (d, J ¼ 10.8 Hz, 1H), 4.45, 4.36 (2 d, J ¼ 11.6 Hz, 2H), 4.12e4.00
(m, 4H), 3.95e3.84 (m, 2H), 3.62 (d, J ¼ 8.4, 1H), 3.35 (dd, J ¼ 8.6,
2.1 Hz, 1H), 3.26e3.15 (m, 2H), 3.10e3.00 (m, 2H), 2.65e2.55 (m,
2H), 2.08e1.90 (m, 2H), 1.54e1.40 (m, 4H), 1.40 (s, 3H), 1.32 (s, 3H),
1.30e1.24 (m, 46H), 0.90e0.80 (m, 6H); 13C NMR (100 MHz, CDCl3)
4.1.6. Synthesis of 3
Compound 3 (64 mg, 79%) was obtained from compound 14
(100 mg, 0.10 mmol) using the same procedure as described in the
preparation of 2, applying 3-phenyl propionic acid (14.1 mg,
20
0.094 mmol) instead of Indole-3-propionic acid. [
a
]
þ41.1 (c 0.5,
D
CHCl3); 1H NMR (400 MHz, MeOD/CDCl3 ¼ 1/15)
d 7.29 (d,
J ¼ 7.2 Hz, 2H), 7.22e7.18 (m, 3H), 4.88 (d, J ¼ 4.0 Hz, 1H), 4.35 (d,
9

C. Zhao, L. Hu, J. Ma et al.
Tetrahedron 89 (2021) 132168
J ¼ 4.4 Hz,1H), 4.17 (d, J ¼ 4.8 Hz, 1H), 4.00e3.84 (m, 3H), 3.73e3.62
(m, 4H), 3.38e3.34 (m, 1H), 3.27e3.23 (m, 1H), 2.96 (t, J ¼ 7.6 Hz,
2H), 2.65e2.57 (m, 7H), 2.21 (t, J ¼ 7.6 Hz, 2H), 1.61e1.40 (m, 4H),
1.32e1.25 (m, 46H), 0.90e0.86 (m, 6H). 13C NMR (100 MHz, MeOD/
4.2. Computational details
4.2.1. Surflex-Dock
To predict their CD1d, TCR affinity, we mimic their binding
interaction with CD1d and TCR with Surflex-Dock in SYBYL-X
software (Tripos Inc., St. Louis, MO, USA) [35]. The crystal struc-
CDCl3 ¼ 1/15)
d 175.89, 174.28, 140.50, 128.60, 128.36, 126.40, 99.06,
74.59, 72.38, 69.53, 69.17, 69.10, 67.99, 60.85, 51.08, 37.82, 36.54,
32.30e31.68 (m), 29.78e29.40 (m), 25.86, 25.55, 22.73, 22.67, 14.14,
14.07; HRMS (ESI): [C48H86N2O9 þ Na]þ calc. 857.6226; found
857.6215.
ture of ternary complex of a-GalCer and CD1d, TCR were down-
loaded from Protein Data Bank (http://www.rcsb.org/, PDB code,
3HE6) and used in the computation [7]. In the tetramer complex
only A chain (CD1d) and C chain (TCR) interact with glycolipid
directly, so they were used to study the interaction of glycolipids
with CD1d and TCR. The designed glycolipids were derived from
OCH, which was extracted from the crystal structure. Gasteiger-
Huckel charge was added for ligand atoms. Before docking, water
and other small molecules in the protein were deleted, all hydrogen
atoms were added, and MMFF94 charge was added for atoms in
protein. Besides, the ligand was extracted, and was used as refer-
ence molecules in the docking computation. The protocol was
generated with threshold 0.5 and Bloat 10.0 Å. In the computation,
we found that the start conformation affected the docking results,
so three conformations for each glycolipid were considered, these
conformations was generated by minimizing energy with the limit
of Max Interactions 0, 100 and 1000 cycles or Gradient 0.05 kcal/
(mol*Å). In the Surflex-Dock, most parameters used default value
except using the extracted ligand as reference molecule. The
highest TotalScore conformation with reasonable binding mode
with both CD1d and TCR, was used to further structure analysis.
4.1.7. Synthesis of 4
Compound 4 (63.8 mg, 78%) was obtained from compound 14
(100 mg, 0.10 mmol) using the same procedure as described in the
preparation of 2, applying 4-phenyl butyric acid (15.4 mg,
0.094 mmol) instead of Indole-3-propionic acid. [
a
]
20 þ114.4 (c 0.5,
D
CHCl3); 1H NMR (400 MHz, MeOD/CDCl3 ¼ 1/15)
d 7.30e7.26 (m,
3H), 7.21e7.17 (m, 2H), 7.10 (d, J ¼ 8.4 Hz, 1H), 6.81 (d, J ¼ 8.4 Hz,
1H), 4.90 (d, J ¼ 4.0 Hz, 1H), 4.38e4.35 (m, 1H), 4.21e4.16 (m, 1H),
3.97e3.94 (m, 2H), 3.87 (dd, J ¼ 10.4, 4.4 Hz, 1H), 3.69 (dd, J ¼ 10.8,
4.0 Hz, 1H), 3.64 (dd, J ¼ 10.4, 4.0 Hz, 1H), 3.59e3.57 (m, 2H),
3.53e3.39 (m, 2H), 2.67e2.64 (m, 7H), 2.30 (t, J ¼ 7.6 Hz, 2H), 2.20
(t, J ¼ 6.8 Hz, 2H), 1.96 (t, J ¼ 7.4 Hz, 2H), 1.64e1.54 (m, 4H),
1.40e1.25 (m, 46H), 0.91e0.86 (m, 6H). 13C NMR (100 MHz, MeOD/
CDCl3 ¼ 1/15)
d 176.64, 174.36, 141.48, 128.54, 126.12, 99.37, 74.89,
72.29, 69.71, 69.32, 69.04, 68.07, 61.10, 51.26, 50.19, 36.70, 36.65,
35.65, 35.60, 35.27, 35.20, 32.61, 32.03, 31.99, 29.84e29.47 (m),
27.44, 25.92, 25.59, 22.79, 22.75, 14.20, 14.14; HRMS (ESI):
[C49H88N2O9 þ Na]þ calc. 871.6382; found 871.6372.
4.2.2. Molecular dynamic analysis
To find key residues that contribute to increase CD1d and TCR
affinity of glycolipids, molecular dynamics were performed to
mimic interaction of glycolipids with CD1d and TCR with AMBER12
software [33]. The glycolipids were optimized with HF/6-31G(d)
method [36] and single-point computation of electrostatic poten-
tial were carried out at the same level of theory with Gaussian 09
software. Based on these computations, the force field parameters,
structural parameters and atomic RESP charge of glycolipids were
generated with Antechamber in Amber12 [37]. GAFF force field was
used for glycolipids and FF03 force field was used for amino acid.
Hydrogen atoms of protein structures were added with tleap pro-
gram [38]. To neutralize the unit system, counter ions (Naþ) were
added. The binding complex was solvated with explicitly solvation
model and filled with TIP3P water molecules [39]. In the cuboid
water box, the solvent depth is at least 10 Å from any protein or
ligand atom. Before molecular dynamic simulation, the system was
minimized (8000 cycles of steepest descent and 4000 cycles of
conjugate gradient minimizations) in the order of restrain all pro-
tein atoms, restrain the backbone atoms and without restrain. Then
the system was gradually heated to 300 K in the speed of 50 K per
5 ps? To get a stable sampling system, the system equilibrated at
300 K for 600 ps? The production simulations of all TCR/ligand/
CD1d and ligands/CD1d systems lasted for 10 ns, the snapshots
were collected at an interval of 10 ps, 100 frames for each system,
which used to calculate binding energy with MMGBSA method
[40]. All the equilibration and production simulations were per-
formed at 300 K. The minimization, heating and equilibrium mo-
lecular dynamics simulations were carried out with sander
program, and the production simulation with pmemd program in
AMBER12. SHAKE was used to constrain covalent bonds involving
hydrogen atom with a 2.0 fs time step [41]. And the particle mesh
Ewald (PME) algorithm was used to calculate the long-range elec-
trostatics, and the non-bond cutoff is 10 Å [42].
4.1.8. Synthesis of 5
Compound 5 (75.2 mg, 82%) was obtained from compound 14
(100 mg, 0.10 mmol) using the same procedure as described in the
preparation of 2, applying 3-(4-benzyloxyphenyl) propionic acid
(24.2 mg, 0.095 mmol) instead of Indole-3-propionic acid.
[
d
a
]
20 þ54.1 (c 0.5, CHCl3); 1H NMR (400 MHz, MeOD/CDCl3 ¼ 1/15)
D
7.44e7.34 (m, 5H), 7.13 (d, J ¼ 8.4 Hz, 2H), 6.90 (d, J ¼ 8.0 Hz, 2H),
5.03 (s, 1H), 4.87 (d, J ¼ 4.0 Hz, 1H), 4.35e4.32 (m, 1H), 4.19e4.16
(m, 1H), 3.94e3.89 (m, 2H), 3.85 (dd, J ¼ 10.4, 4.8 Hz, 1H), 3.66 (dd,
J ¼ 10.8, 4.0 Hz, 1H), 3.59e3.52 (m, 3H), 3.40e3.34 (m, 1H),
3.30e3.25 (m, 1H), 3.23e3.00 (m, 6H), 2.90 (t, J ¼ 8.8 Hz, 2H),
2.59e2.55 (m, 2H), 2.19 (t, J ¼ 8.7 Hz, 2H), 1.68e1.52 (m, 4H),
1.40e1.25 (m, 48H), 0.90e0.86 (m, 6H); 13C NMR (100 MHz, MeOD/
CDCl3 ¼ 1/15)
d 175.8, 174.2, 157.3, 137.0, 132.9, 129.3, 128.5, 127.9,
127.4, 114.9, 99.2, 74.8, 72.1, 70.1, 69.5, 69.1, 68.9, 67.8, 60.8, 51.0,
50.0, 38.0, 36.5, 32.6, 31.9, 31.8, 30.8, 29.68, 29.65, 29.6, 29.5, 29.4,
29.33, 29.31, 25.8, 25.4, 22.64, 22.6, 14.0, 13.98. HRMS (ESI):
[C55H92N2O10 þ Na]þ calc. 963.6644; found 963.6625.
4.1.9. Synthesis of 7
Compound 7 (60.1 mg, 78%) was obtained from compound 14
(100 mg, 0.10 mmol) and 5-methylhexanoic acid (12.1 mg,
0.092 mmol) using the same procedure as described in the prepa-
20
ration of 2. [
CDCl3 ¼ 1/12)
a]
þ11.9 (c 0.5, CHCl3); 1H NMR (400 MHz, MeOD/
D
d
4.92 (d, J ¼ 3.6 Hz, 1H), 4.35 (d, J ¼ 4.4 Hz, 1H), 4.19
(q, J ¼ 4.3 Hz, 1H), 3.99e3.94 (m, 2H), 3.89 (dd, J ¼ 10.8, 4.4 Hz, 1H),
3.71 (dd, J ¼ 10.8, 4.00 Hz, 1H), 3.66 (dd, J ¼ 10.4, 3.6 Hz, 1H),
3.61e3.56 (m, 2H), 3.55e3.44 (m, 2H), 2.40e2.25 (m, 6H),
1.68e1.52 (m, 8H), 1.41e1.18 (m, 52H), 0.91e0.86 (m, 12H). 13C NMR
(100 MHz, MeOD/CDCl3 ¼ 1/12)
d 177.03, 174.16, 99.29, 74.94, 72.31,
69.54, 69.33, 69.07, 68.17, 61.02, 51.19, 38.48, 36.61, 31.96, 31.91,
29.69 (m), 29.60, 29.46, 29.40, 27.83, 25.83, 25.52, 23.71, 22.72,
22.68, 22.50, 22.48, 14.14, 14.07; HRMS (ESI): [C46H90N2O9 þ Na]þ
calc. 837.6539; found 837.6530.
4.2.3. MM/GBSA calculations
The binding free energy for each system was calculated using
10

C. Zhao, L. Hu, J. Ma et al.
Tetrahedron 89 (2021) 132168
the MM/GBSA technique [43e45] according to:
Declaration of competing interest
D
E ¼ Ecomplex - Eprotein - Eligand
¼
D
EMM
þ
D
EGB
þ
D
ESA - T
D
S
The authors declare that they have no known competing
financial interests or personal relationships that could have
appeared to influence the work reported in this paper.
Where EMM is the molecular mechanics interaction energy
D
between the protein and the inhibitor; EGB and ESA are the
electrostatic and nonpolar contributions to dissolvation upon in-
hibitor binding, respectively; and the T S is the conformational
D
D
Acknowledgments
D
entropy change. In this paper, we compute all the binding energies
of GCs with CD1d and TCR with Molecular Mechanics Generalized
Born Surface Area (MMGBSA) method. Because the newly designed
glycolipids had similar structures with OCH, their conformational
entropy change is similar, and the entropy effect is not considered
here. GBa means the binding energy of CD1d and glycolipids in
binary complex CD1d/GCs. In the ternary complex CD1d/GCs/TCR,
Ea, Ec and Eac means the binding energy of CD1d with glycolipids,
glycolipids with TCR, and CD1d with TCR, respectively.
All the structures of CD1d/GCs/TCR used to geometry structure
analyses are lowest energy structure which is extracted from pro-
duction of molecular dynamics. PyMol software (www.pymol.org)
is used to perform the binding interaction analysis of glycolipids
with CD1d and TCR and draw pictures.
This work was supported by National Key Research and Devel-
opment Program of China (no. 2019YFC1605802) and National
Natural Science Foundation of China (Projects 21672255). We thank
Xuefeng Sun and Chengxiang Shang for high resolution mass
determination.
Appendix A. Supplementary data
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.tet.2021.132168.
References
[1] T. Natori, Y. Koezuka, T. Higa, Tetrahedron Lett. 34 (1993) 5591e5592.
[2] T. Natori, M. Morita, K. Akimoto, Y. Koezuka, Tetrahedron 50 (1994)
2771e2784.
[3] M. Morita, K. Motoki, K. Akimoto, T. Natori, T. Sakai, E. Sawa, K. Yamaji,
Y. Koezuka, E. Kobayashi, H. Fukushima, J. Med. Chem. 38 (1995) 2176e2187.
[4] 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 278 (1997)
1626e1629.
[5] M. Koch, V.S. Stronge, D. Shepherd, S.D. Gadola, B. Mathew, G. Ritter,
A.R. Fersht, G.S. Besra, R.R. Schmidt, E.Y. Jones, V. Cerundolo, Nat. Immunol. 6
(2005) 819e826.
[6] 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 448 (2007)
44e49.
[7] D.G. Pellicci, O. Patel, L. Kjer-Nielsen, S.S. Pang, L.C. Sullivan, K. Kyparissoudis,
A.G. Brooks, H.H. Reid, S. Gras, I.S. Lucet, R. Koh, M.J. Smyth, T. Mallevaey,
J.L. Matsuda, L. Gapin, J. McCluskey, D.I. Godfrey, J. Rossjohn, Immunity 31
(2009) 47e59.
[8] M.J. Smyth, D.I. Godfrey, Nat. Immunol. 1 (2000) 459e460.
[9] C.R. Berkers, H. Ovaa, Trends Pharmacol. Sci. 26 (2005) 252e257.
[10] K.O. Yu, S.A. Porcelli, Immunol. Lett. 100 (2005) 42e55.
[11] H. Iijima, K. Kimura, T. Sakai, A. Uchimura, T. Shimizu, H. Ueno, T. Natori,
Y. Koezuka, Bioorg. Med. Chem. 6 (1998) 1905e1910.
4.3. In vitro and in vivo activities
4.3.1. Intracellular staining of IFN-g in iNKT cells
The in vitro NKT cell activation was carried out with previously
reported method [16,46]. C57BL/6 mice (obtained from Vital River,
China) spleen was collected and single cell suspension was pre-
pared gently. After erythrolysis was performed with 0.84% NH4Cl,
splenocytes were washed completely with cold PBS twice. 2 ꢂ 106
per well splenocytes in 48-well plate were cultured with 1000 ng/
mL of KRN7000, OCH and compounds 3, 4, 7 for 16 h, respectively.
In the blank control, splenocytes were incubated in complete me-
dium containing DMSO equal to the exposure group. Brefeldin A
(eBioscience) was added in the last 2 h of the incubation. The
collected cells were incubated with 10 mL mCD1d/a-GalCer Dex-
trame conjugated with R-phycoerythrin (PE) (PBS44 variant,
YD8002-PE, Immudex Fruebjergvej 3, Denmark) at room temper-
[12] X. Laurent, B. Bertin, N. Renault, A. Farce, S. Speca, O. Milhomme, R. Millet,
P. Desreumaux, E. Henon, P. Chavatte, J. Med. Chem. 57 (2014) 5489e5508.
[13] A. Banchet-Cadeddu, E. Henon, M. Dauchez, J.H. Renault, F. Monneaux,
A. Haudrechy, Org. Biomol. Chem. 9 (2011) 3080e3104.
[14] 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. 130 (2008) 12348e12354.
[15] X. Li, M. Fujio, M. Imamura, D. Wu, S. Vasan, C.H. Wong, D.D. Ho, M. Tsuji, Proc.
Natl. Acad. Sci. U. S. A. 107 (2010) 13010e13015.
[16] K. Miyamoto, S. Miyake, T. Yamamura, Nature 413 (2001) 531e534.
[17] K.O. 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. U. S. A. 102 (2005) 3383e3388.
ature for 20 min in 100 mL dilution. The following antibodies were
used for gating CD3þNK1.1 þ CD1d tetramer þ NKT cells: APC
labeled anti-mouse CD3 mAb, FITC labeled anti-mouse NK1.1 mAb
(PK136). And then intracellular staining using Percp-cy5.5 labeled
anti-mouse IFN-g was performed. FACS analysis was done on a
FACS analyzer LSRII (BD Biosciences) according to standard
instructions.
4.3.2. Cytokines profile detection
The concentration of cytokines including IL-4, IFN-
g
was
[18] C. Xia, W. Zhang, Y. Zhang, W. Chen, J. Nadas, R. Severin, R. Woodward,
B. Wang, X. Wang, M. Kronenberg, P.G. Wang, ChemMedChem 4 (2009)
1810e1815.
[19] L. Hu, C. Zhao, J. Ma, Y. Jing, Y. Du, Bioorg. Med. Chem. Lett 29 (2019)
1357e1362.
determined in culture media collected from splenocytes post in-
cubation with the designed compounds for 24 h (n ¼ 5). For the
in vivo cytokines profile analysis, mice (n ¼ 5) were injected with
8 nmol of the substrates or buffer only (control) through tail
intravenous injection (I$V.). At each designed exposure times, blood
[20] W. Zhang, C. Xia, J. Nadas, W. Chen, L. Gu, P.G. Wang, Bioorg. Med. Chem. 19
(2011) 2767e2776.
[21] J. Janssens, T. Decruy, K. Venken, T. Seki, S. Krols, J. Van der Eycken, M. Tsuji,
D. Elewaut, S. Van Calenbergh, ACS Med. Chem. Lett. 8 (2017) 642e647.
[22] J. Janssens, A. Bitra, J. Wang, T. Decruy, K. Venken, J. van der Eycken,
D. Elewaut, D.M. Zajonc, S. van Calenbergh, ChemMedChem 14 (2019)
147e168.
[23] K.S. Wun, N.A. Borg, L. Kjer-Nielsen, T. Beddoe, R. Koh, S.K. Richardson,
M. Thakur, A.R. Howell, J.P. Scott-Browne, L. Gapin, D.I. Godfrey, J. McCluskey,
J. Rossjohn, J. Exp. Med. 205 (2008) 939e949.
[24] C. Forestier, T. Takaki, A. Molano, J.S. Im, I. Baine, E.S. Jerud, P. Illarionov,
R. Ndonye, A.R. Howell, P. Santamaria, G.S. Besra, T.P. DiLorenzo, S.A. Porcelli,
J. Immunol. 178 (2007) 1415e1425.
[25] M. Mizuno, M. Masumura, C. Tomi, A. Chiba, S. Oki, T. Yamamura, S. Miyake,
J. Autoimmun. 23 (2004) 293e300.
was drawn and the amount of IL-4 and IFN-
ELISA kit purchased from NeoBioscience (QuantiCyto® Mouse IFN-
ELISA kit, EMC101.96, QuantiCyto® Mouse IL-4 ELISA kit,
g were quantified by
g
EMC003.96).
All the bioassay experiments were performed in accordance
with ‘the Regulations for the Administration of Affairs Concerning
Experimental Animals Approved by the State Council of People's
Republic of China’. And the care and experimental protocols were
approved by ‘the regulation of the Institute of Microbiology, Chi-
nese Academy of Sciences of Research Ethics Committee’
(PZIMCA2012008).
[26] R.D. Goff, Y. Gao, J. Mattner, D. Zhou, N. Yin, C. Cantu 3rd, L. Teyton,
A. Bendelac, P.B. Savage, J. Am. Chem. Soc. 126 (2004) 13602e13603.
11

C. Zhao, L. Hu, J. Ma et al.
Tetrahedron 89 (2021) 132168
[27] 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 17 (2007) 2781e2784.
[28] R.J. Williams, J. Iglesias-Fernandez, J. Stepper, A. Jackson, A.J. Thompson,
E.C. Lowe, J.M. White, H.J. Gilbert, C. Rovira, G.J. Davies, S.J. Williams, Angew
Chem. Int. Ed. Engl. 53 (2014) 1087e1091.
[29] D. Crich, M. Smith, J. Am. Chem. Soc. 123 (2001) 9015e9020.
[30] F. Faltin, V. Fehring, R. Miethchen, Syn. Stuttgart (2002) 1851e1856.
[31] N. Pauwels, S. Aspeslagh, D. Elewaut, S. Van Calenbergh, Bioorg. Med. Chem.
20 (2012) 7149e7154.
J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev, A.J. Austin, R. Cammi,
C. Pomelli, J.W. Ochterski, R.L. Martin, K. Morokuma, V.G. Zakrzewski,
G.A. Voth, P. Salvador, J.J. Dannenberg, S. Dapprich, A.D. Daniels, O. Farkas,
J.B. Foresman, J.V. Ortiz, J. Cioslowski, D.J. Fox, Gaussian 09, Revision B.01, ed.,
Gaussian, Inc., Wallingford, CT, 2010.
[35] A.N. Jain, J. Comput. Aided Mol. Des. 10 (1996) 427e440.
[36] M.M. Francl, W.J. Pietro, W.J. Hehre, J.S. Binkley, M.S. Gordon, D.J. DeFrees,
J.A. Pople, J. Chem. Phys. 77 (1982) 3654e3665.
[37] C.I. Bayly, P. Cieplak, W.D. Cornell, P.A. Kollman, J. Phys. Chem. 97 (1993)
10269e10280.
[32] M. Michieletti, A. Bracci, F. Compostella, G. De Libero, L. Mori, S. Fallarini,
G. Lombardi, L. Panza, J. Org. Chem. 73 (2008) 9192e9195.
[38] J. Wang, R.M. Wolf, J.W. Caldwell, P.A. Kollman, D.A. Case, J. Comput. Chem. 25
(2004) 1157e1174.
[33] D.A. Case, T.E. Cheatham 3rd, T. Darden, H. Gohlke, R. Luo, K.M. Merz Jr.,
A. Onufriev, C. Simmerling, B. Wang, R.J. Woods, J. Comput. Chem. 26 (2005)
1668e1688.
[39] W.L. Jorgensen, J. Chandrasekhar, J.D. Madura, R.W. Impey, M.L. Klein, J. Chem.
Phys. 79 (1983) 926e935.
[40] H. Gohlke, D.A. Case, J. Comput. Chem. 25 (2004) 238e250.
[41] J.-P. Ryckaert, G. Ciccotti, H.J.C. Berendsen, J. Comput. Phys. 23 (1977)
327e341.
[42] T. Darden, D. York, L. Pedersen, J. Chem. Phys. 98 (1993) 10089e10092.
[43] H. Sun, Y. Li, S. Tian, L. Xu, T. Hou, Phys. Chem. Chem. Phys. 16 (2014)
16719e16729.
[44] T. Hou, J. Wang, Y. Li, W. Wang, J. Comput. Chem. 32 (2011) 866e877.
[45] T. Hou, J. Wang, Y. Li, W. Wang, J. Chem. Inf. Model. 51 (2011) 69e82.
[46] M. Fang, M.T. Orr, P. Spee, T. Egebjerg, L.L. Lanier, L.J. Sigal, Immunity 34
(2011) 579e589.
[34] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb,
J.R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G.A. Petersson,
H. Nakatsuji, M. Caricato, X. Li, H.P. Hratchian, A.F. Izmaylov, J. Bloino,
G. Zheng, J.L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda,
J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven,
J.A. Montgomery Jr., J.E. Peralta, F. Ogliaro, M. Bearpark, J.J. Heyd, E. Brothers,
K.N. Kudin, V.N. Staroverov, T. Keith, R. Kobayashi, J. Normand,
K. Raghavachari, A. Rendell, J.C. Burant, S.S. Iyengar, J. Tomasi, M. Cossi,
N. Rega, J.M. Millam, M. Klene, J.E. Knox, J.B. Cross, V. Bakken, C. Adamo,
12