Site-specific effect of polar functional group-modification in lipids of TLR2 ligands for modulating the ligand immunostimulatory activity

Article abstract of DOI:10.1016/j.bmcl.2018.03.042

Toll-like receptor 2 (TLR2), a member of the TLR innate immune receptor family, recognizes lipoproteins from bacteria and modulates the immune response by inducing the expression of various cytokines. TLR2 has a large hydrophobic pocket that recognizes long fatty acyl groups on TLR2 ligands. However, few studies have focused on the property of the hydrophobic TLR2 pocket. Based on the X-ray crystal structure of TLR2, small polar regions were found in the hydrophobic TLR2 pocket. Interactions between the polar residues and ligands were explored here by designing and synthesizing a Pam₂CSK₄ derivative of the TLR2 ligands, containing an amide group within the lipid moiety. We evaluated the binding affinities and immunomodulatory activities of these ligands. Results suggested that the amide groups in the lipid chain interacted with the polar residues in the hydrophobic lipid-binding pocket of TLR2.

Full text of DOI:10.1016/j.bmcl.2018.03.042

Bioorganic & Medicinal Chemistry Letters xxx (2018) xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Site-specific effect of polar functional group-modification in lipids of
TLR2 ligands for modulating the ligand immunostimulatory activity
Yohei Arai , Shinsuke Inuki a,b, Yukari Fujimoto a,
a
⇑
a
Faculty of Science and Technology, Keio University, 3-14-1 Hiyoshi, Yokohama, Kanagawa 223-8522, Japan
Graduate School of Pharmaceutical Sciences, Kyoto University, 46-29 Yoshida-Shimo-Adachi-cho, Sakyo-ku, Kyoto 606-8501, Japan
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Toll-like receptor 2 (TLR2), a member of the TLR innate immune receptor family, recognizes lipoproteins
from bacteria and modulates the immune response by inducing the expression of various cytokines. TLR2
has a large hydrophobic pocket that recognizes long fatty acyl groups on TLR2 ligands. However, few
studies have focused on the property of the hydrophobic TLR2 pocket. Based on the X-ray crystal struc-
ture of TLR2, small polar regions were found in the hydrophobic TLR2 pocket. Interactions between the
Received 3 January 2018
Revised 15 March 2018
Accepted 16 March 2018
Available online xxxx
2 4
polar residues and ligands were explored here by designing and synthesizing a Pam CSK derivative of
Keywords:
Toll-like receptor 2
Pam CSK
2 4
Structure–activity relationship
Lipopeptide
the TLR2 ligands, containing an amide group within the lipid moiety. We evaluated the binding affinities
and immunomodulatory activities of these ligands. Results suggested that the amide groups in the lipid
chain interacted with the polar residues in the hydrophobic lipid-binding pocket of TLR2.
Ó 2018 Elsevier Ltd. All rights reserved.
Innate immune receptors play a critical role in the recognition
of microbial components and the activation of an immune
response in a host cell.1 Toll-like receptor 2 (TLR2) is one such
receptor that recognizes lipoproteins from bacterial cell mem-
branes and induces heterodimerization with TLR1 or TLR6, leading
to the activation of an immune response.3 The dimerization of
TLR1/2 or TLR2/6 leads to release of various proinflammatory
cytokines, such as interleukin-1b (IL-1b), IL-6, and tumor necrosis
structure–activity relationship (SAR) studies of lipid moieties on
TLR2 ligands have been conducted. Ulmer and co-workers focused
on the length of the acyl chain in the TLR2 ligand, and they
,2
3 4
designed Pam CSK
containing various acyl chain lengths;13 David
and co-workers designed and synthesized monoacyl lipopeptides
that exhibited hTLR2-specific agonist activities greater than the
,4
1
4
2 4
Pam CSK activity. The molecular recognition mechanism under-
lying the lipid binding site in TLR2, especially the interactions
among polar amino acid residues in the pocket, remains unclear.
We recently focused on small confined hydrophilic regions in
the large hydrophobic pocket of the lipid-binding protein CD1d.
We revealed that these regions play an important role in ligand
factor alpha (TNF-a). These proinflammatory cytokines activate
5
adaptive immunity to affect other immune cells. TLR2 ligands
are, therefore, promising candidates for vaccine adjuvants and
6
–11
pharmaceuticals that support cancer immunotherapies.
1
5
The TLR1/2 heterodimer recognizes triacyl lipoproteins from
Gram-negative bacteria and some Gram-positive bacteria. On the
other hand, the TLR2/6 heterodimer recognizes diacyl lipoproteins
from Gram-positive bacteria and mycoplasma. The triacyl lipopep-
recognition. The designed CD1d ligands containing polar func-
tional groups in the lipid chain, which were expected to interact
with the hydrophilic regions of the binding pocket, displayed
higher binding affinities and cytokine production activities com-
pared to the CD1d ligand containing an unmodified lipid chain.
Molecular dynamics simulations suggested that the formation of
hydrogen bonds between the polar functional groups and the con-
fined hydrophilic residues of CD1d modulated ligand recognition.
These results were consistent with the following results; Several
groups have investigated the general water-shielding properties
of hydrogen bonds located in a hydrophobic binding pocket,
revealing that the microenvironment increases the kinetic stability
3 4 2 4
tide Pam CSK and the diacyl lipopeptides Pam CSK are known
12
potent TLR1/2 and TLR2/6 ligands, respectively (Fig. 1). The lipid
component of these lipoproteins is located at the N-terminal
2 4 3 4
cysteine of the proteins, and both Pam CSK and Pam CSK have
a diacylglycerol that connects to the cysteine thiol. X-ray crystal-
lography indicated that TLR2 includes extensive hydrophobic
pockets that can recognize the lipid moieties of bacterial lipopro-
teins via hydrophobic interactions;3 however, only limited
,4
1
6,17
of the hydrogen bonds.
In the present research, we designed
4
derivatives containing amide groups
and synthesized Pam CSK
2
⇑
Corresponding author.
E-mail address: fujimotoy@chem.keio.ac.jp (Y. Fujimoto).
in the lipid chain in an effort to investigate the properties of the
https://doi.org/10.1016/j.bmcl.2018.03.042
960-894X/Ó 2018 Elsevier Ltd. All rights reserved.
0
Please cite this article in press as: Arai Y., et al. Bioorg. Med. Chem. Lett. (2018), https://doi.org/10.1016/j.bmcl.2018.03.042

2
Y. Arai et al. / Bioorganic & Medicinal Chemistry Letters xxx (2018) xxx–xxx
2 4 3 4
Fig. 1. Structures of the TLR2 ligands, Pam CSK and Pam CSK .
hydrophilic amino acid residues in the TLR2 lipid binding pocket
and to develop novel TLR2 ligands.
We first obtained detailed information about the hydrophilic
region in the TLR2 lipid-binding pocket. Using a computational
1
8
tool, SiteMap, we analyzed two crystal structures of TLR2–ligand
3
complexes, the hTLR1-hTLR2-Pam
3 4
CSK complex (PDB: 2Z7X),
4
and the mTLR2-mTLR6-Pam CSK complex (PDB: 3A79). Two spa-
2
4
tially confined hydrophilic regions, including hydrogen bond
acceptors, were identified in the large hydrophobic pocket of
TLR2 (Fig. 2). One region was involved in the backbone amide
group of Leu334. The other region contained the side chain of the
Ser345 residue. The former residue (Leu334) was present only in
hTLR2. By contrast, the latter (Ser345) was present in both hTLR2
and mTLR2.
2 4
Fig. 3. Structures of the Pam CSK derivatives containing an amide group in the
lipid chain.
2 4
The synthesis of the Pam CSK derivatives containing amide
Focusing on these confined hydrophilic regions, we designed
two types of Pam CSK derivatives (1a–d and 2a–e) containing a
2 4
groups in the lipid chain is shown in Scheme 1. The amide
group-containing fatty acids 5a–i were synthesized through a
Schotten–Baumann reaction of the commercially available car-
boxylic acid chlorides 3a–i and the corresponding amino acids
polar group in the lipid moiety, allowing the interaction between
the polar residue and the ligand (Fig. 3). The ligands 1a–d included
polar group-containing lipid acyl chains at the primary alcohol
position of the glycerol moieties. By contrast, a polar group was
introduced into the lipid acyl chains attached to the second alcohol
positions of the glycerol moieties in the ligands 2a–e. We expected
the ligands 1a–d to form hydrogen bonding interactions with the
Leu334 residue and the ligands 2a–e to interact with the Ser346
residue in TLR2. In view of previous reports, we selected the amide
as the polar functional group because amide-containing lipids
could be easily accessed from commercially available compounds.
4
a–i. With all carboxylic acid derivatives 5a–i in hand, we next
sought to convert 5a–i to the Pam CSK derivatives starting from
the common precursor 6. The key precursor 6 was prepared effi-
2
4
ciently from
the literature.
L-cystine according to the procedure published in
12,19,20
The introduction of the carboxylic acid deriva-
tives 5a–d to the primary hydroxyl group was conducted using
WSC and DMAP to smoothly give the corresponding acyl chain-
containing compounds 7a–d. Acylation of the second hydroxyl
group with palmitic acid, cleavage of the tert-butyl ester, and a
solid-phase synthesis using Rink Amide Resin as a solid phase sup-
port led to the desired Pam
a–e were synthesized from the common precursor 6 in the same
manner as described above.
2 4
CSK derivatives 1a–d. The compounds
2
We next evaluated the immunostimulatory activities of our
designed/synthesized ligands using human macrophage cell lines
(
phorbol 12-myristate 13-acetate (PMA)-differentiated THP-1
human monocytic cells) and mouse macrophage RAW 264.7 cells,
which are antigen-presenting cells and express immune receptors,
including TLR2. In these assays, human or mouse macrophage cell
lines were incubated in the presence of various concentrations of
the synthesized ligands, and the induction levels of inflammatory
cytokines IL-6 and TNF-
a were measured by enzyme-linked
immunosorbent assay (ELISA) (Fig. 4 and Supplementary Fig. S1).
Initially, we investigated the IL-6 induction activities of the com-
pounds 1a–d, the amide groups of which were expected to interact
with the Leu334 residue in human TLR2 (Fig. 4A). All ligands dis-
played cytokine production in a dose-dependent manner. The
designed compounds 1a–d were less active than the unmodified
2 4
ligand Pam CSK . Interestingly, the levels of cytokine production
depended on the position of the amide group in the lipid chain.
The ligands 1c and 1d exhibited slightly higher levels of cytokine
production compared with the corresponding amide-containing
ligands 1a and 1b (10 and 100 nM). Next, we examined the
stimulation of cytokine release by compounds 2a–e, which could
Fig. 2. Hydrophilic and hydrophobic map of the hTLR2 pocket (PDB: 2Z7X),3
generated by SiteMap. The color of the surface represents the nature of the pocket:
yellow indicates hydrophobic regions, blue and red indicate hydrophilic regions.
3 4
The Pam CSK structure is shown in the green model.
Please cite this article in press as: Arai Y., et al. Bioorg. Med. Chem. Lett. (2018), https://doi.org/10.1016/j.bmcl.2018.03.042

Y. Arai et al. / Bioorganic & Medicinal Chemistry Letters xxx (2018) xxx–xxx
3
2 4
Scheme 1. Synthesis of the Pam CSK derivatives 1a–d and 2a–e.
Fig. 4. Proinflammatory cytokine induction by the synthesized compounds 1a–d and 2a–e. (A) and (B) IL-6 induction in PMA-differentiated THP-1 human monocytic cells. (C)
IL-6 induction in murine macrophage RAW264.7 cells. The graphs show mean ± standard error for triplicate values. Similar results were obtained from two or three
independent experiments.
interact with the Ser346 residue present in both human and mouse
TLR2. Human macrophage cell lines and mouse macrophage RAW
the cytokine production level measurement in macrophage cell
lines revealed that the position of the polar functional group in
the lipid chain was associated with their cytokine induction levels.
The position of the amide group in a lipid chain affected
the immunostimulatory activities; therefore, we investigated the
binding potential of the synthesized ligands to TLR2. We
performed a competitive binding assay based on the Alpha (Ampli-
2
64.7 cells were used for this study (Fig. 4B and C), and IL-6 cyto-
kine secretion levels were quantified. Although compounds 2a–e
induced lower levels of cytokine production than Pam CSK , the
2
4
amide position in the acyl chain influenced the ligand activities.
The ligand 2b, in particular, induced relatively high levels of cyto-
kine production among the lipid-modified ligands. Taken together,
Ò
fied Luminescent Proximity Homogeneous Assay) Screen system
Please cite this article in press as: Arai Y., et al. Bioorg. Med. Chem. Lett. (2018), https://doi.org/10.1016/j.bmcl.2018.03.042

4
Y. Arai et al. / Bioorganic & Medicinal Chemistry Letters xxx (2018) xxx–xxx
using TLR2-conjugated acceptor beads, streptavidin-coated donor
2 4
Pam CSK , which displayed an immunostimulatory activity similar
-biotin as a tracer.2
1,22
Without competitive
to that of these ligands (Fig. 4). These results suggested that site-
specific interactions between these amide groups and the polar
residues affected the ligand binding interactions and could be
involved in the immunological activities of the compounds. Dock-
ing studies of the synthesized ligands based on X-ray crystal struc-
tures (Fig. 2) suggested that the amide groups in compounds 1c,
1d, and 2b were positioned closer to the hydrophilic regions in
TLR2. These experimental results suggested that the amide groups
could interact with polar amino acid residues in the lipid-binding
pocket of TLR2. On the other hand, the introduction of amide
groups into the lipid chain decreased the binding affinities and
immunostimulatory activities, which would be affected by the
lipid chain flexibility.
beads, and Pam
ligands (synthesized ligands), TLR2-conjugated acceptor beads
interacted with Pam CSK -biotin recognized by the streptavidin-
2
CSK
4
2
4
coated donor beads, bringing the donor and acceptor beads in close
proximity and permitting excitation of the acceptor beads by sin-
glet oxygen produced after illumination of the donor beads. By
2 4
contrast, competitive ligands competed with Pam CSK -biotin,
which decreased the signal. Initially, we assessed the binding
potential of the designed ligands 1a–d using an hTLR2-His tagged
protein (Fig. 5A).
All ligands containing amide groups in the lipid chain displayed
2 4
lower binding affinities compared with the unmodified Pam CSK ,
consistent with the cytokine induction assay results. 1c and 1d dis-
played higher binding affinities than the other amide-containing
synthesized ligands. Next, we examined the interactions between
the designed ligands 2a–e and the hTLR2-His-tagged protein, and
mTLR2-Fc fusion protein (Fig. 5B and C). As shown in Fig. 5B and
C, the synthesized compounds 2a and 2b exhibited higher affinities
toward human and mouse TLR2. The ligand binding affinities to
TLR2 depended on the amide position in the lipid chain of
2 4
In this study, we synthesized novel Pam CSK analogues con-
taining an amide group in the lipid moiety to act as TLR2 ligands.
We evaluated the binding potentials of the synthesized ligands
using a competitive binding assay based on the Alpha Screen.
The immunomodulatory activities were measured using human
and mouse macrophage cell lines. Both assays revealed that
although the introduction of the amide group did not improve
the ligand affinities and immunological activities, the amide posi-
tion in the lipid did affect the affinities and activities. These results
suggest that the amide groups may specifically interact with polar
residues in the lipid binding pocket, and these residues would be a
target for the regulation of TLR2 lipid recognition.
Acknowledgments
This research was supported by grants from the Grant-in-Aid
for Scientific Research (Nos. JP17H02207, JP17H05800,
JP16H01162, and JP16K16638), the Mizutani Foundation for
Glycoscience, Mishima Kaiun Memorial Foundation, Takeda
Science Foundation, Keio University Doctorate Student Grant-in-
Aid Program.
A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at https://doi.org/10.1016/j.bmcl.2018.03.042.
References
1.
2.
3.
4.
5.
6.
7.
Akira S, Uematsu S, Takeuchi O. Cell. 2006;124:783.
Creagh EM, O’Neill LA. Trends Immunol. 2006;27:352.
Jin MS, Kim SE, Heo JY, et al. Cell. 2007;130:1071.
Kang JY, Nan X, Jin MS, et al. Immunity. 2009;31:873.
O’Neill LA, Golenbock D, Bowie AG. Nat Rev Immunol. 2013;13:453.
Kovacs-Simon A, Titball RW, Michell SL. Infect Immun. 2011;79:548.
Hutchings MI, Palmer T, Harrington DJ, Sutcliffe IC. Trends Microbiol.
2009;17:13.
8
9
.
.
Kanzler H, Barrat FJ, Hessel EM, Coffman RL. Nat Med. 2007;13:552.
Jarnicki AG, Conroy H, Brereton C, et al. J Immunol. 2008;180:3797.
10. Marshall NA, Galvin KC, Corcoran AM, Boon L, Higgs R, Mills KH. Cancer Res.
2012;72:581.
1
1
1
1. Adams S. Immunotherapy. 2009;1:949.
2. Fujimoto Y, Hashimoto M, Furuyashiki M, et al. ChemBioChem. 2009;10:2311.
3. Buwitt-Beckmann U, Heine H, Wiesmüller KH, et al. J Biol Chem. 2006;281:
9049.
14. Salunke DB, Connelly SW, Shukla NM, Hermanson AR, Fox LM, David SA. J Med
Chem. 2013;56:5885.
15. Inuki S, Aiba T, Hirata N, et al. ACS Chem Biol. 2016;11:3132.
1
6. Schmidtke P, Javier Luque F, Murray JB, Barril X.
011;133:18903.
J Am Chem Soc.
2
1
1
1
2
7. Gao J, Bosco DA, Powers ET, Kelly JW. Nat Struct Mol Biol. 2009;16:684.
8. Halgren TA. J Chem Inf Model. 2009;49:377.
9. Wiesmuller K, Bessler W, Jung G. Hoppe-Seyler’s Z Physiol Chem. 1983;364:593.
0. Reichel F, Roelofsen AM, Geurts HP, Hämäläinen TI, Feiters MC, Boons GJ. J Am
Chem Soc. 1999;121:7989.
TM
Fig. 5. Binding assay based on AlphaScreen , applied to the synthesized lipopep-
tides. Competitive inhibition assay between Pam CSK -biotin and (A) 1a–d using
2 4
2
1. Bosse R, Illy C, Chelsky D. Application Note. Principles of AlphaScreen
(PerkinElmer Life Science, Montreal), 2002.
hTLR2-His tag, (B) 2a–e using hTLR2-His tag, (C) 2a–e using mTLR2-Fc. Similar
22. Koymans KJ, Feitsma LJ, Brondijk TH, et al. Proc Natl Acad Sci U S A.
results were obtained from two or three independent experiments.
2015;112:11018.
Please cite this article in press as: Arai Y., et al. Bioorg. Med. Chem. Lett. (2018), https://doi.org/10.1016/j.bmcl.2018.03.042