Synthesis of characteristic Mycobacterium peptidoglycan (PGN) fragments utilizing with chemoenzymatic preparation of meso-diaminopimelic acid (DAP), and their modulation of innate immune responses

Article abstract of DOI:10.1039/c5ob02145f

Peptidoglycan (PGN) is a major component of bacterial cell wall and is recognized as a potent immunostimulant. The PGN in the cell envelope of Mycobacterium Tuberculosis has been shown to possess several unique characteristics including the presence of N-glycolyl groups (in addition to N-acetyl groups) in the muramic acid residues, and amidation of the free carboxylic acid of d-Glu or of meso-DAP in the peptide chains. Using a newly developed, highly stereoselective, chemoenzymatic approach for the synthesis of meso-DAP in peptide stems, we successfully synthesized for the first time, a series of Mycobacterium PGN fragments that include both mono- and disaccharides of MurNGlyc or 1,6-anhydro-MurNGlyc, as well as peptide-amidated variants. The ability of these PGN fragments to stimulate the immune system through activation of human Nod1 and Nod2 was examined. The PGN fragments were found to modulate immune stimulation, specifically, amidation at the d-Glu and meso-DAP in the peptide stem strongly reduced hNod1 activation. This effect was dependent on modification position. Additionally, N-glycolyl (instead of acetyl) of muramic acid was associated with slightly reduced human Nod1 and Nod2 stimulatory capabilities.

Full text of DOI:10.1039/c5ob02145f

Organic &
Biomolecular Chemistry
View Article Online
View Journal
PAPER
Synthesis of characteristic Mycobacterium
peptidoglycan (PGN) fragments utilizing
with chemoenzymatic preparation of
meso-diaminopimelic acid (DAP), and their
modulation of innate immune responses†
Cite this: DOI: 10.1039/c5ob02145f
Qianqian Wang,a,b Yusuke Matsuo,a Ambara R. Pradipta,a Naohiro Inohara,c
Yukari Fujimoto*b and Koichi Fukase*a
Peptidoglycan (PGN) is a major component of bacterial cell wall and is recognized as a potent immuno-
stimulant. The PGN in the cell envelope of Mycobacterium Tuberculosis has been shown to possess
several unique characteristics including the presence of N-glycolyl groups (in addition to N-acetyl groups)
in the muramic acid residues, and amidation of the free carboxylic acid of D-Glu or of meso-DAP in the
peptide chains. Using a newly developed, highly stereoselective, chemoenzymatic approach for the syn-
thesis of meso-DAP in peptide stems, we successfully synthesized for the first time, a series of Myco-
bacterium PGN fragments that include both mono- and disaccharides of MurNGlyc or 1,6-anhydro-
MurNGlyc, as well as peptide-amidated variants. The ability of these PGN fragments to stimulate the
immune system through activation of human Nod1 and Nod2 was examined. The PGN fragments were
found to modulate immune stimulation, specifically, amidation at the D-Glu and meso-DAP in the peptide
stem strongly reduced hNod1 activation. This effect was dependent on modification position. Additionally,
N-glycolyl (instead of acetyl) of muramic acid was associated with slightly reduced human Nod1 and
Nod2 stimulatory capabilities.
Received 17th October 2015,
Accepted 24th November 2015
DOI: 10.1039/c5ob02145f
www.rsc.org/obc
rily conserved microbial components called pathogen-associ-
ated molecular patterns (PAMPs). PRRs include membrane-
Introduction
Mycobacterium tuberculosis (TB) is a major global health issue, bound Toll-like receptors (TLRs) and cytosolic Nod-like recep-
ranked as the second leading cause of death from an infec- tors (NLRs).3,4 Peptidoglycan (PGN), as a major component of
tious disease worldwide, after the human immunodeficiency bacterial cell walls in both Gram-positive and Gram-negative
virus (HIV).1 Mycobacteria, including Mycobacterium tuber- bacteria,5 is known as a potent immunostimulator. Nod1 and
culosis (M. tuberculosis), are intracellular pathogens that evade Nod2, founding members of the NLR protein family, have
both innate and adaptive host immunity.2 There is a clear and been shown to recognize bacterial PGN fragments (muro-
urgent need for a deeper understanding of Mycobacteria peptides) and regulate innate and adaptive immune responses
biology and the epidemiology of infection.
by activating transcription factors including NF-κB.6 Nod1
Innate immune responses are the first line of defense recognizes meso-DAP-containing muropeptides, with the basic
against foreign invaders, and host immune cells employ motif required for ligand binding being iE-DAP (γ-D-glutamyl
pattern recognition receptors (PRRs) to recognize evolutiona- diaminopimelic acid).7 The minimum ligand for Nod2, on the
other hand, is MDP (muramyl dipeptide: MurNAc-L-Ala-
D-isoGln).8,9
aDepartment of Chemistry, Graduate School of Science, Osaka University,
PGN is comprised of long glycan chains with repeating
Machikaneyama 1-1, Toyonaka, Osaka 560-0043, Japan
bDepartment of Chemistry, Faculty of Science and Technology, Keio University,
Hiyoshi 3-14-1, Yokohama, Kanagawa 223-8522, Japan.
E-mail: fujimotoy@chem.keio.ac.jp
β(1–4) disaccharide, generally N-acetylglucosamine (GlcNAc)
and N-acetylmuramic acid (MurNAc), which are linked via
peptide bridges. The branched position of the peptide is
usually a diaminocarboxylic acid, such as L-Lys (in many
Gram-positive bacteria) or meso-diaminopimelic acid (meso-
DAP, in Gram-negative bacteria and some Gram-positive
cDepartment of Pathology, University of Michigan Medical School, Ann Arbor,
Michigan 48109, USA
†Electronic supplementary information (ESI) available. See DOI: 10.1039/
c5ob02145f
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
bacteria, e.g., Mycobacteria and Bacillus species). The highly sequences with various amidation patterns, in order to investi-
cross-linked PGN layer in M. tuberculosis displays unique gate their Nod1 and Nod2 stimulatory potentials. We have
chemical features:10 amidated carboxylic acid of D-Glu and shown that smaller PGN fragments, such as monosaccharide
meso-DAP in the peptide stem, and N-glycolylmuramic acid and disaccharide fragments, exhibited higher Nod2-stimulat-
(MurNGlyc), in addition to the conventional N-acetylmuramic ing activity than the corresponding longer glycan frag-
acid (MurNAc), in the glycan chain (Fig. 1A). This amidation ments.16,17 It was also demonstrated that monosaccharide and
may prevent the bacteria from being recognized by host Nod1 disaccharide fragments with DAP-containing tripeptides
—a finding also observed in Bacillus subtilis.11 The glycolylated exhibited potent Nod1-stimulating activity.18 Hence, we chose
PGN exhibited increased stability in the presence of β-lactam to focus on monosaccharide and disaccharide fragments as
antibiotics and lysozyme and was also hypothesized to stabil- synthetic targets in this study.
ize the cell wall through hydrogen bonding.12 These modifi-
The (2S,6R)-meso-DAP has been identified as the most
cations were thought to be closely related to the intracellular potent Nod1 stimulator compared with other three isomers in
survival and consequent pathogenesis of M. tuberculosis, but our previous study.19 Although several synthetic methods for
no direct evidence had been obtained so far.
generation of meso-DAP, the key moiety for Nod1 recognition,
With the exception of N-glycolyl MDP,13–15 the immuno- have been reported,18 we developed a chemoenzymatic method
stimulatory activity of PGN partial structures specific to for the synthesis of the orthogonally protected meso-DAP. This
M. tuberculosis has not been well characterized. Especially, the newly developed method was advantageous for large-scale syn-
immunostimulatory capability of meso-DAP-containing Myco- thesis over the reported methods due to short reaction steps,
bacterium PGN fragments has been rarely studied. In this mild conditions, and high enantioselectivity. Tripeptides
study, we synthesized various PGN partial structures, including (L-Ala-D-isoGln-meso-DAP) and tetrapeptides (L-Ala-D-isoGln-
MurNGlyc and meso-DAP containing tri- or tetrapeptide meso-DAP-D-Ala) were prepared and then coupled to mono-
and disaccharides to construct a series of Mycobacterium PGN
fragments with glycolyl- and amide-modifications (Fig. 1B). We
then examined their modulation of human innate immune
system via the interaction with cytosolic receptor proteins
Nod1 and Nod2. This is the first demonstration that Myco-
bacterium PGN fragments exhibited very weak hNod1 and
hNod2 activation, which provides direct proof for the cytosolic
immune evasion of M. tuberculosis.
Results and discussion
An efficient chemoenzymatic method for the large-scale
synthesis of meso-DAP derivatives
We initially developed a chemoenzymatic method for the
large-scale synthesis of meso-DAP (Scheme 1). L-Aminoacylase
and D-aminoacylase can specifically hydrolyze N-acetyl groups
with L- and D-configurations, respectively, under mild con-
ditions (40 °C in 0.05 M phosphate buffer). The commercially
available racemic 2,6-diaminopimelic acid was first acetylated.
Then, the L-N-acetyl group of 3 was hydrolyzed by L-amino-
acylase (Acylase Amano) and the free amino group was pro-
tected with a tert-butoxycarbonyl (Boc) group to generate 4
with S- and R-chiral centers. Herein, a mixture of three com-
pounds: (2R,6R)-di-N-acetyl DAP, (2S,6S)-di-N-Boc DAP, and
compound 4, was obtained, which was completely separated
by extraction with ethyl acetate at different pH values. At a pH
range of 5–6, all of the (2S,6S)-di-N-Boc DAP and a small
amount of compound 4 were extracted to the organic layer.
After acidification at a pH range of 1–2, compound 4 was trans-
ferred to the organic layer while (2R,6R)-di-N-acetyl DAP
remained in the water layer. Compound 4 displayed high
purity as revealed by its 1H-NMR spectrum.
In order to facilitate the deprotection of acetyl group and
also improve optical purity, N-acetyl in D-configuration of 4 was
Fig. 1 (A) Schematic structure of M. Tuberculosis peptidoglycan. (B)
Chemically synthesized M. Tuberculosis peptidoglycan fragments.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
The liberated amino group of compound 5 was protected
with a benzyloxycarbonyl (Z) group to obtain compound 6. The
two carboxylic acids were protected with benzyl groups yielding
the orthogonally protected meso-DAP 7. The Boc group of com-
pound 7 was deprotected with trifluoroacetic acid (TFA) and
coupled to the side-chain carboxyl group of Boc-L-Ala-D-isoGln
using 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydro-
chloride (WSCD·HCl), 1-hydroxybenzotriazole (HOBt), and tri-
ethylamine in THF, to give fully protected tripeptide 12. The
optical purity of 12 was confirmed by chiral HPLC analysis
(CHIRALPAK AD-H, Daicel Chemical), showing none of the
other enantiomer and diastereomers were contained. The Boc
group of 12 was then removed to obtain the key intermediate,
tripeptide 14 (H-L-Ala-D-isoGln-meso-DAP).
To prepare tetrapeptide, the Boc group of compound 6 was
cleaved with TFA, and deprotected product 8 was reacted with
triphosgene to form a N-carboxyanhydride (NCA) ring in com-
pound 9.20 The ring-opening polymerization of α-amino acid
N-carboxyanhydrides (NCAs) has been reported to be the most
economical and expedient process for the synthesis of poly-
peptide-containing block copolymers.21 Herein, the ring-
opening reaction was conducted by coupling with HCl·H-D-Ala-
OBn to generate the dipeptide with a free amino group, which
was protected with Boc group in 10. Tetrapeptide 15 (H-L-Ala-
D-isoGln-meso-DAP-D-Ala) was synthesized from 10 in a similar
manner to the synthesis of 14. With the peptides in hand,
we continued the synthesis of meso-DAP containing PGN
fragments.
Scheme 1 Synthesis of meso-DAP derivatives 14 and 15: (a) Ac2O, pyri-
dine, H2O, overnight; (b) L-aminoacylase, CoCl2·6H2O, phosphate buffer,
40 °C, 5 d; (c) Boc2O, TEA, H2O/1,4-dioxane = 2 : 1, overnight; (d) D-ami-
noacylase, CoCl2·6H2O, phosphate buffer, 40 °C, 7 d; (e) ZCl, Na2CO3,
H2O/1,4-dioxane = 2 : 1, 6 h; (f) BnBr, Na2CO3, DMF, 7 h; (g) 20% TFA/
DCM, 30 min; (h) 1 M HCl in Et2O; (i) triphosgene, THF, 50 °C, 6 h; ( j)
HCl·NH2-D-Ala-OBn, Et3N, DMF, 0 °C – r.t., 1 h; (k) Boc2O, TEA, H2O/
1,4-dioxane = 1 : 1, overnight; l) BnBr, Cs2CO3, DMF, 6 h; (m) Boc-L-Ala-
D-isoGln-COOH, WSCD·HCl, HOBt, Et3N, THF, overnight; (n) TFA,
30 min.
Synthesis of meso-DAP containing Mycobacterium PGN
fragments
The synthetic procedures for the generation of mono-
saccharide compounds containing N-glycolyl are depicted in
Schemes 2 and S1.† (The synthesis of MurNAc-containing
structures is described in Scheme S2.†) For MurNGlyc-contain-
ing fragments, the Alloc protection group in compound 16 16
was cleaved and then glycolylated using 4-(4,6-Di-
methoxy-1,3,5-triazin-2-yl)-4-methylmorpholinium
Chloride
(DMT-MM)22 to prevent esterification of the liberated hydroxy
group of glycolic acid. The resulting compound 17 was
obtained in quantitative yield. The allyl group in compound 17
was isomerized using H2-activated [Ir(cod)(MePh2P)2]PF6
(cod = 1,5-cyclooctadiene), followed by the deprotection of
ethyl ester with LiOH to give the product 18. Compound 18
was subsequently condensed with tripeptide 14, tetrapeptide
cleaved by D-aminoacylase to yield chiral compound 5 in a
manner similar to L-aminoacylase.
In the enzyme reactions, the addition of proper concen-
tration of cobalt(II) ions effectively accelerated the reaction. We
optimized the concentration of cobalt(II) (Tables S1 and S2†).
For the L-aminoacylase reaction generating compound S1 from
compound 3, addition of 1 mM CoCl2 led to higher yield
(60% meso-DAP in 4 d) than reaction conducted in the absence
of Co(II) ions (32% in 4 d). However, 10 mM CoCl2 produced
almost the same yield as 1 mM. For D-aminoacylase reaction
generating compound 5 from compound 4, yield of compound
18
15, or HCl·NH2-L-Ala-D-isoGlu-OBn-meso-DAP(OBn)2 to yield
compounds 19, 20, or 21, respectively. The removal of vinyl
group with iodine and water and final hydrogenation afforded
the final compounds 1b, 1d, and 1i in high yields. The
1,6-anhydro-muramic acid-containing compounds were syn-
thesized from an azide sugar 25.18 The azide group of com-
pound 25 was reduced to a free amino group which was
glycolylated to generate compound S4. Compound S4 under-
went ethyl-ester removal, peptide-coupling, and deprotection
to generate the final compounds, 1f and 1h.
5
was increased when the concentration of Co(II) was
increased; Specifically, 10% (no CoCl2 in 4 d), 32% (1 mM
CoCl2 in 4 d), 67% (10 mM CoCl2 in 4 d), and 83% (10 mM
CoCl2 in 8 d).
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
Scheme 2 Synthesis of Mycobacterium PGN fragments with monosac-
charides. (a) Pd(PPh3)4, TMSDMA, CHCl3, 2 h; (b) glycolic acid, DMT-MM,
THF, 2.5 h; (c) [Ir(cod)(H)(MePh2P)2]PF6, THF, 2 h; (d) LiOH, THF/1,4-
dioxane/H2O = 4 : 2 : 1, 1 h; (e) 14, 15 or HCl·NH2-L-Ala-D-isoGlu-OBn-
meso-DAP(OBn)2, DMT-MM, Et3N, CH3OH/H2O = 2 : 1, overnight; (f) I2,
THF/H2O = 1 : 1, 2 h; (g) H2 (2.0 MPa), Pd(OH)2, AcOH; (h) Zn, THF/AcOH
= 1 : 1, 30 min; (i) H2 (2.0 MPa), Pd(OH)2, THF/H2O/AcOH = 1 : 1 : 0.1.
The synthesis of Mycobacterium PGN fragments with disac-
charides was also succeeded following the routes in Schemes 3
and S3† (MurNGlyc-containing structure) and Scheme S4†
(MurNAc-containing structure). For GlcNAc-MurNGlyc-contain-
ing structures (Scheme 3A), the glycosyl acceptor 29 was pre-
pared from the regioselective ring opening of 4,6-O-
benzylidene of 16 to form 6-O-benzyl derivative by using
BH3·Me3N/BF3·Et2O. Glycosylation of compound 29 with
N-Troc glucosaminyl trichloroacetimidate 30 in the presence of
TMSOTf afforded the desired disaccharide 31. Compound 30
was chosen as the glycosyl donor because of its excellent
β-selective glycosylation ability due to the neighboring partici-
pation of Troc group. The Troc group in 31 was cleaved with
Scheme 3 Synthesis of Mycobacterium PGN fragments with disacchar-
ides. (A) having a MurNGlyc moiety and (B) a MurNGlyc(anh) moiety. (a)
Me3N·BH3/BF3·Et2O, CH3CN, 1 h; (b) TMSOTf, MS(4 Å), CH2Cl2, −15 °C,
15 min; (c) Zn/Cu, AcOH/THF = 1 : 1, 1 h; (d) Ac2O, pyridine, overnight;
(e) Pd(PPh3)4, TMSDMA, CHCl3, 2 h; (f) glycolic acid, DMT-MM, THF, 2 h;
(g) [Ir(cod)(H)(MePh2P)2]PF6, THF, 2 h; (h) LiOH, THF/1,4-dioxane/H2O =
4 : 2 : 1, 1 h; (i) DMT-MM, Et3N, CH3OH/H2O = 2 : 1, overnight; ( j) I2, THF/
H2O = 1 : 1, 2 h; (k) H2 (2.0 MPa), Pd(OH)2, AcOH, 24 h; (l) H2 (1.5 MPa),
Pd(OH)2, THF, 36 h; (m) AllocCl, Et3N, CH2Cl2, 1.5 h; (n) H2 (2.0 MPa),
Pd(OH)2, THF/H2O/AcOH = 1 : 1 : 0.1, 24 h.
Zn/Cu and the liberated amino group was acetylated to give Coupling of compound 33 with peptides and subsequent de-
compound S14. The Alloc group in compound S14 was then protection of the vinyl group (with I2) and benzyl-type groups
replaced either by a glycolyl group (as shown in Scheme 3A) to (via hydrogenation) generated the N-glycolyl disaccharide frag-
generate the key intermediate 32 for the synthesis of glycolyl- ments 2b and 2d. The GlcNAc-MurNGlyc(anh)-containing
containing compounds, or an acetyl group to generate the structures 2f and 2h (Scheme 3B) were synthesized similarly to
intermediate for synthesis of acetyl-containing compounds the synthesis of compounds 2b and 2d from compound 38.
(Scheme S4†). After isomerization of the allyl group of com- Compound 38 was prepared in 57% yield from azide sugar 25
pound 32 to a vinyl group with an Ir-complex, the cleavage of via hydrogenation and the introduction of an Alloc group in 2
ethyl ester generated compound 33 in 73% yield in 2 steps. steps. The synthesis of Mycobacterium PGN fragments with
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
Scheme 4 Synthesis of amidated derivatives of meso-DAP-containing
peptides: (a) triphosgene, THF, 50 °C, 5 h; (b) 0.5 M NH3 in 1,4-dioxane,
0 °C – r.t., 1.5 h; (c) Boc2O, TEA, H2O/DMF = 1 : 2, overnight; (d) BnBr,
Cs2CO3, DMF, overnight; (e) ZCl, Na2CO3, H2O/1,4-dioxane = 1 : 1, 2 h;
(f) 28% ammonia, WSCD·HCl, HOBt, Et3N, THF, overnight; (g) TFA,
30 min; (h) 1 M HCl in Et2O; (i) Boc-D-Glu-OBn, WSCD·HCl, HOBt, Et3N,
THF, overnight; ( j) Boc-L-Ala-OH, WSCD·HCl, HOBt, Et3N, THF,
overnight.
monosaccharides (compounds 1a–1h) and disaccharides (com-
pounds 2a–2h) were thus achieved.
Fig. 2 Stimulation of human Nod1 with synthetic PGN fragments. (A)
Monosaccharide-containing fragments (1a–1h) of M. tuberculosis PGN,
and iE-DAP and 1n (MS-3PA-iE-DAP; non-amidated compound) were used
as references. (B) Monosaccharide-containing fragments with or
without amidation. (C) Disaccharide-containing fragments (2a–2h) of
M. tuberculosis PGN.
In order to comprehensively compare the effect of amid-
ation modification on the host immune stimulation, PGN frag-
ments 1j, 1k and 1m with peptides, 50j, 50k and 50m, were
also synthesized (Schemes 4 and S5†). For the synthesis of
L-amide-meso-DAP 45, NCA in 9 was reacted with anhydrous
NH3 (5 equiv.) in 1,4-dioxane to give the L-amidated 44. After
protection of the free amino and carboxyl groups, 45 was
obtained successfully. The D-amide-meso-DAP 48 was also pre-
pared from 5 in a similar way. The two –COOH groups in 6
were amidated by 28% ammonia aqueous solution with
WSCD·HCl, HOBt and triethylamine to give L,D-di-amide-meso-
DAP 49. The amidated meso-DAPs were then reacted with the
side-chain carboxyl group of Boc-D-Glu-OBn and Boc-L-Ala con-
secutively to afford the tripeptides. After Boc-deprotection, the
appropriately protected key intermediates, tripeptides 50j, 50k
and 50m, were successfully synthesized, which were used for
the synthesis of compounds 1j, 1k, and 1m in Scheme S5.†
In the human Nod1 assay (Fig. 2), iE-DAP and 1n (MS-
3PA-iE-DAP), were used as the references. Compound 1n had
been shown to have more potent activity than iE-DAP in our
previous study.18 Analysis of structure–activity relationships
systematically revealed the effects of amidation patterns at
distinct peptides (Glu or meso-DAP). Amidation at the α-carb-
oxylic acid of D-Glu dramatically reduced hNod1 activation
(1a–1h) compared to 1n (Fig. 2A), which was consistant with
our finding from the corresponding peptides: tripeptide
A-iE-DAP (L-Ala-γ-D-glutamyl-meso-DAP) showed stronger
hNod1 activity than A-iQ-DAP (L-Ala-γ-D-glutaminyl-meso-
DAP).23 In Fig. 2B, different amidation patterns were compared
based on their hNod1 activation potentials. Amidation of one
of the two carboxyl groups of meso-DAP (1j and 1k) displayed a
milder influence on hNod1 activation in comparison with the
D-Glu modification (1a). Amidation of the carboxylic acid at
the D-position of meso-DAP (1k) led to a greater reduction in
hNod1 activation potential than that at the L-position (1j),
implying that the D-COOH was more important than the
L-COOH for hNod1 sensing. The glycolyl modification had
little effect on hNod1 activation as demonstrated by 1n and 1i.
Innate immunomodulation of Mycobacterium PGN fragments
The immunomodulating abilities of the synthetic Mycobacter-
ium PGN fragments were evaluated via hNod1 and hNod2 acti-
vation with cell assays utilizing HEK293T based conditions;
hNod1 assay was done with stably human NOD1 transfected
cells (HEG1G) (Fig. 2), whereas hNod2 assay was conducted
with transiently human NOD2 transfected cells (Fig. 3), to find
out the structural motifs that affect the innate immune
modulation.
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
As expected, fragments with (anh)MurN (1e–1h and 2e–2h)
were unable to activate Nod2 due to modification of the MurN
backbone.
Conclusions
The effective chemoenzymatic synthesis of meso-DAP deriva-
tives greatly facilitated the preparation of various meso-DAP-
containing PGN fragments whose innate immunity modulat-
ing efficacy was comprehensively explored. Glycolylation and
amidation are the major structural modification characteristics
of M. tuberculosis peptidoglycans. This study initially manifests
that Mycobacterium PGN fragments exhibit very weak hNod1
and hNod2 activation, providing a potential mechanism for
M. tuberculosis cytosolic immune evasion. Amidation is respon-
sible for the reduced hNod1 responses, which is consistent
with previously reported studies using peptides and regular
monosaccharide-containing structures.7,10 In addition, this
study systematically evaluated the effect of the meso-DAP
residue on hNod2 activation for the first time. We previously
reported that MDP, the minimal ligand required for hNod2
activation, elicited the strongest hNod2 activation and Lys-con-
taining tri- or tetrapeptides also possessed significant Nod2-
stimulatory activity, although their potency decreased as the
length of the peptide chain increased.16,17 We demonstrate
here that the meso-DAP residue considerably reduces the
hNod2-stimulatory activity of PGN fragments.
Fig. 3 Stimulation of human Nod2 with synthetic PGN fragments. (A)
Monosaccharide-containing fragments (1a–1h) of M. tuberculosis PGN,
and MDP and 1p (MurNAc-L-Ala-D-isoGln-L-Lys) were used as refer-
ences. (B) Disaccharide-containing fragments (2a–2h) of M. tuberculosis
PGN.
We also found that anhydro-MurN-containing structures were
slightly more potent activators of hNod1 than regular MurN
structures (Fig. 2A and C). Disaccharide fragments (2a–2h)
showed similar hNod1 activation tendencies as their corres-
ponding monosaccharides varieties. The above results demon-
strate that amidation at D-Glu as well as meso-DAP in
Mycobacterium PGN fragments reduce their recognition by
hNod1, and this may provide advantage to the bacteria by
allowing them to escape from the detection by the host’s
innate immune system.
On the other hand, the glycolylation modification had no
significant effect on either hNod1 or hNod2 stimulation;
N-glycolylated compounds showed slightly weaker activities
than the corresponding acetylated ones. The glycolyl group
may, rather, contribute to maintenance of cell wall integrity
and improve resistance to lysozyme—another important
element for bacterial survival.
Experimental section
In the human Nod2 assay (Fig. 3), MDP and 1p (MurNAc-L-
Ala-D-isoGln-L-Lys)16 were used as the references. The N-glycolyl
MDP [MDP(Glyc)] displayed strong hNod2 activation; however,
it was less potent than MDP. Comparing N-acetylated tri-
peptide monosaccharide 1a and disaccharide 2a with N-glyco-
lylated tripeptide monosaccharide 1b and disaccharide 2b,
N-acetylated 1a and 2a induced slightly greater activation of
hNod2 than the corresponding 1b and 2b, respectively. Thus,
N-glycolyl modification led to reduced hNod2 activation,
although the effect was not so significant. The effect of DAP
residues on activation of hNod2 was also investigated for the
first time in detail. Tripeptide-containing fragments, 1a, 1b,
2a, and 2b, displayed very weak hNod2 stimulatory capabili-
ties, while Lys-containing tripeptide, 1p (MurNAc-L-Ala-D-
isoGln-L-Lys), showed modest activity. These results clearly
demonstrate that the meso-DAP moiety dramatically decreases
hNod2 activation. Almost no hNod2-activating function was
observed for DAP-containing tetrapeptides, 1c, 1d, 2c, and 2d.
General
1H NMR and 13C spectra were recorded in indicated solvents
by using a JEOL ECS 400, or a JEOL ECA 500, or a Varian
INOVA 700 spectrometers. The chemical shifts in CDCl3 are
given in δ values from tetramethylsilane as an internal stan-
dard. HRMS measurement was carried out by using Thermo
Fisher Scientific ESI-LIT-orbitrap mass spectrometry. Silica-gel
column chromatography was carried out by using Kieselgel 60
(Merck, 0.040–0.063 mm) at medium pressure (2–4 kg cm−2).
Reverse phase chromatography was carried out using Sephadex
HP20 at atmospheric pressure. TLC analysis was performed on
Silica-gel 60 F254 (Merck) and compounds were visualized by
UV (254 nm), phosphomolybdic acid solution (5.0% in EtOH),
0.03% p-methoxybenzaldehyde in EtOH-conc. H2SO4-acetic
acid buffer or 0.2% ninhydrin in EtOH-collidine-acetic acid
buffer. Anhydrous CH2Cl2 was prepared by distillation from
calcium hydride. Non-aqueous reactions were carried out
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
under argon atmosphere unless otherwise noted. All other washed with diethylether, and the water layer was acidified to
reagents and solvents used were purchased from commercial pH 2–3 and extracted with ethyl acetate three times. The com-
sources. Molecular sieves 4 Å were activated with microwave bined organic layers were dried over Na2SO4 and concentrated
1 min × 3 times before use.
(2S,6R)-6-acetamido-2-((tert-butyloxycarbonyl)amino)hept- column
anedioic acid (4). Compound 3 (27 g, 98 mmol) was dissolved 12 : 1 : 0.1) to give compound 6 as a white solid (9.18 g, 90%).
in phosphate buffer (pH = 8.2, 0.05 M) to a final concentration ), 5.08
1H NMR (400 MHz, CD3OD): δ 7.36–7.25 (5H, m; ArH
as 0.05 M, heated to 40 °C in water bath, then L-aminoacylase (2H, s; –CH2Ph–), 4.15 (1H, dd, 4.6 Hz, 8.4 Hz;
(5.4 g) and CoCl2·6H2O (466 mg, 10−3 M) were added. The solu- –CH
tion was stirred at 40 °C for 5 d. The solution was cooled to 1.93–1.60 (4H, m; 3-CH
room temperature and acidified to pH 2–3 before being added (CH3)3, 4-CH2
in vacuo. The crude compound was purified by silica-gel flash
chromatography (CHCl3:CH3OH : CH3COOH
=
̲
̲
J =
(NHBoc)–), 4.07 (1H, dd, J = 4.6 Hz, 8.2 Hz; –CH
(NHZ)–),
), 1.54–1.42 (11H, m; –OC-
̲
2, 5-CH2
̲
̲
̲
); 13C NMR (100 MHz, CD3OD):
δ
176.1,
activated carbon, then warmed to 70 °C for 10 min. The black 175.8, 158.6, 158.1, 138.2, 129.4, 128.9, 128.7, 80.5, 67.6, 55.2,
suspension was filtrated, and the filtrate was evaporated to 54.7, 32.4, 32.3, 28.7, 23.3.; HRMS (ESI-LIT-orbitrap):
give a crude compound.
Without any purification, the crude compound was dis- 447.1744.
solved in H2O (200 mL) and triethylamine (60 mL). The solu- Dibenzyl
tion of Boc2O (70 mL) in 1,4-dioxane (100 mL) was added to butyloxycarbonyl) amino}heptanedioate (7). Compound
m/z: calcd for C20H28N2O8Na: 447.1738 [M + Na]+; found:
(2S,6R)-6-[{(benzyloxy)carbonyl}amino]-2-{(tert-
6
the water solution dropwise at 0 °C. The reaction was stirred at (4.0 g, 9.42 mmol) and Na2CO3 (2.0 g, 18.8 mmol) were dis-
room temperature and monitored by TLC until complete reac- solved into dry DMF (25 mL). The resulting suspension was
tion. The mixture was adjusted to pH 5–6 with 2 M aq. HCl stirred at room temperature for 30 min before BnBr (4.5 mL,
and extracted with ethyl acetate to remove di-butoxycarbonated 37.7 mmol) were added dropwise. The reaction was reacted for
compound. Further adjusted to pH 2–3 and extracted with 7 h before quenching with H2O and extracted with ethyl
ethyl acetate several times to get pure 4. The ethyl acetate layer acetate twice. The combined organic layers were washed with
was dried over Na2SO4 and concentrated in vacuo. The oily brine, dried over Na2SO4 and concentrated in vacuo. The crude
compound was resuspended to 50 mL water and neutralized compound was purified by silica-gel flash column chromato-
with 2 M aq. NH3, then the solution was lyophilized to give graphy (toluene : AcOEt = 8 : 1) to give compound 7 as a white
compound 4 as a white solid (11.40 g, 35%, 2 steps).
solid (5.18 g, 91%).
(2R,6S)-7-[{(R)-1-(benzyloxy)-1-oxopropan-2-yl}amino]-2-
1H NMR (400 MHz, D2O): δ 4.06 (1H, dd, J = 4.8 Hz, 8.7 Hz;
–CH
–NHCOCH
m; –OC(CH
̲
̲
̲
̲
̲
̲
̲
179.5, 174.3, 158.3, 81.7, 56.2, 55.5, 32.0, 31.7, 28.4, 22.6.; stirred for 30 min at room temperature. The solution was then
HRMS (ESI-LIT-orbitrap): m/z: calcd for C14H24N2O7Na: concentrated and co-evaporated with toluene to remove TFA.
355.1476 [M + Na]+; found: 355.1480.
Addition of 1 M HCl diethylether solution gave a white precipi-
(2S,6R)-6-amino-2-{(tert-butyloxycarbonyl)amino}heptane- tate. The supernatant was removed to give Boc-deprotected
dioic acid (5). D-aminoacylase (5.65 g) was dissolved in phos- intermediate 8, which was lyophilized in 1,4-dioxane. The
phate buffer (pH = 8.2, 0.05 M), heated to 40 °C. CoCl2·6H2O crude 8 was dissolved in dry THF at 50 °C, followed by addition
(1.62 g, 6.81 mmol) was added to the solution followed by the of triphosgene (1.54 g, 5.18 mmol) in dry THF. The solution
addition of 4 (11.3 g). Without stirring, the mixture was was stirred at 50 °C for 6 h. After cooling to room temperature,
reacted at 40 °C for 7 d. The solution was cooled to room temp- the solvent was evaporated quickly, and the oily solid 9 was
erature and acidified to pH 2–3 before being added activated washed by dry hexane twice. Without purification, the next
carbon, then warmed to 70 °C for 10 min. The black suspen- step was proceeded immediately. To the solution of the crude
sion was filtrated, and the filtrate was evaporated to give a compound 9 in dry DMF (60 mL) was added HCl·H-D-Ala-OBn
crude compound. The residue was purified with HP20 to give (3.05 g, 14.13 mmol) and triethylamine (2.6 mL) in dry DMF
compound 5 as a purple solid (7.90 g, 80%).
1H NMR (500 MHz, D2O): δ 4.18 (0.2H, br s; –CH
3.93 (1H, br s; –CH
2.02–1.67 (4H, m; 3-CH
(CH3)3, 4-CH2
C12H22N2O6Na: 313.1370 [M + Na]+; found: 313.1372.
(30 mL) at 0 °C. The mixture was stirred at 0 °C for 0.5 h, and
̲
̲
(NH2)–), ing DMF, the next step was carried out directly. The crude com-
), 1.52–1.49 (11H, m, –OC- pound was dissolved into H2O/1,4-dioxane (1 : 1, 80 mL)
); HRMS (ESI-LIT-orbitrap): m/z: calcd for followed by addition of triethylamine (4.8 mL) at 0 °C. Boc2O
(13 mL) was then added dropwise to the solution, and the
̲
̲
(2S,6R)-6-[{(benzyloxy)carbonyl}amino]-2-{(tert-butyloxycarb- solution was stirred at room temperature overnight. The solu-
onyl)amino}heptanedioic acid (6). To the solution of com- tion was acidified to pH 2–3 and extracted with ethyl acetate
pound 5 (7.5 g, 26 mmol) and Na2CO3 (10.95 g) in water five times. The combined organic layers were dried over
(250 mL) was added benzyl chloroformate (5.57 mL, 39 mmol) Na2SO4 and concentrated in vacuo. The crude compound was
in 1,4-dioxane (125 mL) dropwise at 0 °C, stirred at 0 °C for purified twice by silica-gel flash chromatography, first eluted
1 h, then stirred at room temperature for 5 h. The solution was with CHCl3:MeOH : CH3COOH = 15 : 1 : 0.1, then eluted with
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
CHCl3:MeOH = 10 : 1 to give the pure compound 10 (606.8 mg,
Protected H-L-Ala-D-isoGln-meso-DAP (14). Compound 12
22% for 5 steps).
(800 mg, 1 mmol) was dissolved in TFA (3 mL) and the result-
1H NMR (400 MHz, CD3OD): δ 7.36–7.24 (10H, m; ArH
̲
5.17–5.08 (4H, m; –O–CH
̲
D-Ala-αH
̲
̲
̲
̲ ,
̲
̲
̲
̲
157.7, 138.2, 137.3, 129.6, 129.5, 129.3, 129.2, 129.0, 128.8,
Protected L-Ala-D-isoGln-meso-DAP-D-Ala (13). Compound 13
80.7, 67.9, 67.6, 55.8, 55.3, 33.2, 32.4, 28.7, 23.2, 17.4; HRMS was synthesized from compound 11 with similar method to
(ESI-LIT-orbitrap): m/z: calcd for C30H40N3O9: 586.2759 the synthesis of compound 12. The crude compound was puri-
[M + H]+; found: 586.2764.
fied by silica-gel flash column chromatography (CHCl3/MeOH
Protected meso-DAP-D-Ala (11). Compound 10 (600 mg, 30 : 1) to give 13 as a white solid (450 mg, 65%).
1.02 mmol) and Cs2CO3 (199 mg, 0.61 mmol) were dissolved
1H NMR (500 MHz, CD3OD): δ 7.34–7.26 (15H, m; ArH
into dry DMF (5 mL). The resulting suspension was stirred at 5.18–5.04 (6H, m; –CH2-Ph × 3), 4.43 (1H, q, J = 7.3 Hz; D-Ala-
room temperature for 30 min before BnBr (242 μL, 2.4 mmol) αH), 4.33 (1H, dd, J = 3.2 Hz, 10.9 Hz; Gln-αH), 4.26 (1H, dd,
were added dropwise. The reaction was reacted for 6 h before J = 5.5 Hz, 7.6 Hz; DAP 2-H), 4.17 (1H, m; DAP 6-H), 4.00 (1H,
quenching with H2O and extracted with ethyl acetate three q, J = 7.2 Hz; L-Ala-αH), 2.32–2.22 (3H, m; Gln-βCH, γCH2),
times. The combined organic layers were washed with brine, 1.84–1.42 (16H, m; Gln-βCH, DAP 3-CH2, 5-CH2, 4-CH2, –OC-
dried over Na2SO4, and concentrated in vacuo. The crude com- (CH3)3), 1.37 (3H, d, J = 7.3 Hz; D-Ala-βCH3), 1.28 (3H, d, J =
pound was purified by silica-gel flash column chromatography 7.2 Hz; L-Ala-βCH3
); 13C NMR (CDCl3, 98.5 MHz): δ 174.2,
(CHCl3 : acetone = 20 : 1) to give compound 11 as a white solid 173.9, 173.7, 173.4, 172.6, 172.3, 156.5, 156.2, 137.9, 135.5,
(5.18 g, 75%). 135.4, 128.7, 128.6, 128.6, 128.5, 128.4, 128.2, 128.1, 128.1,
̲),
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
Fully-protected L-Ala-D-isoGln-meso-DAP (12). Compound 7 80.7, 67.2, 67.1, 67.0, 54.0, 53.5, 51.2, 50.7, 48.4, 31.6, 31.2,
(1.0 g, 1.65 mmol) was dissolved in 50% TFA in CH2Cl2 31.1, 29.9, 28.3, 21.7, 17.7, 16.9; HRMS (ESI-LIT-orbitrap): m/z:
(20 mL) and the mixture was stirred for 30 min at room temp- calcd for C45H58N6O12Na: 897.4005 [M + Na]+; found: 897.4016.
erature. The reaction mixture was then concentrated and co-
evaporated with toluene. Addition of 1 M HCl diethylether 15 was synthesized from compound 13 with similar method to
solution gave white precipitate. The supernatant was the synthesis of compound 14. The compound was used
removed to give Boc-deprotected intermediate. To the residue without further purification.
Protected H-L-Ala-D-isoGln-meso-DAP-D-Ala (15). Compound
a
was added anhydrous THF, WSCD·HCl (477 mg, 2.47 mmol),
The synthetic procedures of 1b, 1d and 1i in Scheme 2 are
HOBt (334 mg, 2.47 mmol), and Boc-L-Ala-D-isoGln-COOH shown in ESI.†
(785 mg, 2.47 mmol) were added. Triethylamine (689 μL,
Compound 1b. 1H NMR (500 MHz, D2O): δ 5.13 (1H, d, J =
4.95 mmol) was then added to the solution at 0 °C. The 3.5 Hz; H-1β), 4.30–4.15 (4H, m; Ala-αH
̲
̲, Lac-αH̲, DAP
mixture was stirred overnight at room temperature. 10% 2-H
citric acid solution was added to the mixture and m; DAP 6-H
̲), 4.04–3.93 (3H, m; –NHCOCH̲
̲
̲
̲
̲
̲ ,
̲
̲
̲
̲
compound was purified by silica-gel flash column chromato- NMR (125 MHz, D2O): δ 176.6, 176.4, 176.0, 175.8, 175.5,
graphy (CHCl3/MeOH 20 : 1) to give 12 as a white solid (1.11 g, 175.3, 175.1, 91.5, 79.6, 76.4, 72.3, 69.8, 61.6, 61.4, 55.2, 54.7,
84%).
1H NMR (500 MHz, CDCl3): δ 7.72 (1H, s; NH
(16H, m; ArH, NH), 6.66 (1H, m; NH
), 5.77 (1H, d, J = 7.0 Hz; [M + Na]+; found: 703.2767.
NH), 5.55 (1H, m; NH), 5.16–5.06 (7H, m; –CH2-Ph × 3, NH),
Compound 1d. 1H NMR (500 MHz, D2O): δ 5.12 (1H, d, J =
4.83 (1H, s; Gln-αH), 4.51 (1H, dd, J = 7.4 Hz, 12.5 Hz; DAP 3.3 Hz; H-1β), 4.29–4.14 (5H, m; D-Ala-αH, Ala-αH, Gln-αH, Lac-
2-H), 4.40 (1H, dd, J = 7.2 Hz, 12.2 Hz; DAP 6-H), 4.11 (1H, s; αH DAP 2-H), 4.03–3.92 (3H, m; –NHCOCH2OH, H-2),
Ala-αH), 2.35–2.24 (2H, m; Gln-γCH2), 2.10–2.05 (1H, m; Gln- 3.85–3.65 (4H, m; DAP 6-H, H-3, H-5, H-6), 3.57–3.42 (2H, m;
βCH), 1.92 (1H, br s; Gln-βCH), 1.87–1.66 (4H, m; DAP 3-CH2 H-4, H-6′), 2.32 (2H,br s; Gln-γCH2), 2.14–2.06 (1H, m; Gln-
5-CH2), 1.52–1.43 (11H, m; DAP 4-CH2, –OC(CH3)3), 1.23 (3H, βCH), 1.92–1.66 (5H, m; Gln-βCH
d, J = 6.9 Hz; Ala-βCH 2, Lac-βCH3, Ala-βCH3
3); 13C NMR (CDCl3, 125 MHz): δ 173.5, 1.41–1.28 (11H, DAP 4-CH
53.9, 53.7, 50.4, 32.4, 31.4, 30.7, 27.4, 21.8, 19.3, 17.2; HRMS
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
,
̲
̲
̲
̲
̲
̲
DAP 3-CH2
̲
,
5-CH2
̲
̲
),
);
̲
̲
̲
̲
, D-Ala-βCH3
173.1, 172.1, 156.1, 155.5, 136.2, 135.3, 135.2, 128.7, 128.6, 13C NMR (125 MHz, D2O): δ 176.5, 176.4, 176.3, 176.0, 175.7,
128.5, 128.5, 128.3, 128.2, 80.3, 67.4, 67.1, 67.1, 53.7, 52.6, 175.5, 175.1, 173.7, 91.5, 79.6, 76.4, 72.3, 69.8, 61.6, 61.4, 55.2,
51.4, 50.8, 32.6, 31.7, 31.0, 30.5, 28.3, 21.2, 18.0. HRMS 54.5, 54.0, 53.5, 51.2, 50.4, 32.2, 31.5, 30.7, 27.5, 21.7, 19.3,
(ESI-LIT-orbitrap): m/z: calcd for C42H53N5O11Na: 826.3634 18.0, 17.2; HRMS (ESI-LIT-orbitrap): m/z: calcd for
[M + Na]+; found: 826.3646.
C29H49N7O16Na: 774.3128 [M + Na]+; found: 774.3136.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
Compound 1i. 1H NMR (400 MHz, D2O): δ 5.07 (1H, d, J = 0.116 mmol) was added. The mixture was stirred at room
2.5 Hz; H-1β), 4.23–4.16 (4H, m; Ala-αH, Glu-αH, Lac-αH, DAP temperature for 10 min before adding DMT-MM, then allowed
2-H), 3.99–3.87 (3H, m; –NHCOCH2OH, H-2), 3.77–3.63 (4H, to react overnight. After dissolving the white solid to H2O, the
m; DAP 6-H, H-3, H-5, H-6), 3.58–3.39 (2H, m; H-4, H-6′), mixture was extracted with CHCl3 three times. The combined
2.27–2.23 (2H, m; Glu-γCH2), 2.11–2.02 (1H, m; Glu-βCH), organic layers were washed with 10% citric acid solution, satu-
1.91–1.62 (5H, m; Glu-βCH, DAP 3-CH2, 5-CH2), 1.37–1.25 (8H, rated aq. NaHCO3, brine and dried over Na2SO4, and then con-
DAP 4-CH2, Lac-βCH3, Ala-βCH3
). 13C NMR (100 MHz, D2O): δ centrated in vacuo. The crude compound was purified by silica-
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
176.2, 176.0, 176.0, 175.7, 175.4, 175.3, 174.8, 91.5, 79.6, 76.4, gel flash column chromatography (CHCl3/MeOH 20 : 1) to give
72.2, 69.7, 61.6, 61.4, 56.5, 54.9, 53.9, 53.4, 50.2, 32.2, 30.9, compound 27 as a white solid (60 mg, 97%).
30.5, 27.5, 21.6, 19.2, 17.4. HRMS (ESI-LIT-orbitrap): m/z: calcd
1H NMR (500 MHz, CDCl3): δ 7.84 (1H, d, J = 6.6 Hz; NH
̲
for C26H44N5O16: 682.2778 [M + H]+; found: 682.2780.
7.55 (1H, d, J = 7.6 Hz; NH
̲), 7.39–7.30 (21H, m; ArH̲, N̲
2-Glycolylamino-1,6-anhydro-2-deoxy-4-O-benzyl-3-O-[(1R)-1- (1H, d, J = 9.5 Hz; NH
(carbonyl)ethyl]-β-D-glucopyranose (26). To the solution of Hz; NH
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
to give the reduced intermediate as a white solid (83.8 mg, (1H, s; H-3), 3.47 (1H, s; H-4), 2.46 (1H, s; –NHCOCH2OH
̲
90%). Glycolic acid (27.2 mg, 0.358 mmol) and the above inter- 2.35–2.26 (2H, m; Gln-γCH
mediate were mixed in THF for 10 min before being added (1H, br s; Gln-βCH), 1.87–1.69 (4H, m; DAP 3-CH
DMT-MM (132 mg, 0.477 mmol). The mixture was stirred at 1.47–1.33 (8H, DAP 4-CH2, Lac-βCH3, Ala-βCH3
̲
2), 2.08 (1H, br s; Gln-βCH
̲
̲
2̲ , 5-CH2
̲
),
̲
̲
̲
); 13C NMR
room temperature for further 2.5 h. 10% citric acid solution was (125 MHz, CDCl3): δ 173.5, 173.5, 173.2, 173.0, 172.1, 171.2,
added to the mixture and extracted with CHCl3 twice. The com- 156.1, 137.1, 136.2, 135.3, 135.2, 128.7, 128.7, 128.6, 128.6,
bined organic layers were washed with saturated aq. NaHCO3, 128.5, 128.5, 128.4, 128.3, 128.2, 100.5, 76.5, 76.0, 75.8, 74.0,
brine and dried over Na2SO4, and then concentrated in vacuo. 71.7, 67.3, 67.2, 67.1, 65.1, 62.2, 53.7, 52.7, 51.8, 49.6, 47.1,
The crude compound was purified by silica-gel flash column 32.5, 31.2, 30.6, 30.2, 21.3, 17.8, 17.1; HRMS (ESI-LIT-orbitrap):
chromatography (CHCl3/MeOH 25 : 1) to give compound, 1,6- m/z: calcd for C55H66N6O16Na: 1089.4428 [M + Na]+; found:
anhydro-2-deoxy-2-glycolylamino-4-O-benzyl-3-O-[(1R)-1-(carbo-nyl)
1089.4413.
ethyl]-β-D-glucopyranose as a white solid (81 mg, 84%).
Protected monosaccharide (anh) containing N-glycolylmur-
1H NMR (500 MHz, CDCl3): δ 7.39–7.32 (5H, m; ArH
̲
(1H, d, J = 9.6 Hz; NH
Hz; PhCH
̲
̲
̲
̲
̲
̲
̲),
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
,
5-CH2
̲
),
);
̲
̲
̲
̲
column, first washed by H2O, then by methanol to give com- 13C NMR (125 MHz, CDCl3): δ 173.6, 173.5, 173.1, 172.8, 172.3,
pound 26 as a white solid (62.3 mg, 87%). HRMS (ESI-LIT- 171.9, 171.5, 156.2, 137.1, 135.3, 128.7, 128.6, 128.5, 128.4,
orbitrap): m/z: calcd for C18H23NO8Na: 404.1316 [M + Na]+; 128.2, 128.1, 127.9, 127.8, 100.5, 75.9, 75.7, 73.9, 71.7, 67.1,
found: 404.1328.
65.1, 62.2, 53.8, 53.2, 51.6, 49.7, 48.3, 47.3, 32.1, 30.8, 29.9,
Protected monosaccharide (anh) containing N-glycolylmur- 21.6, 17.8, 17.5, 17.3; HRMS (ESI-LIT-orbitrap): m/z: calcd for
amyl group and tripeptide (27). Compound 26 (22 mg, C58H71N7O17Na: 1160.4799 [M + Na]+; found: 1160.4813.
0.058 mmol) and 14 (64 mg, 0.087 mmol) were dissolved to
Monosaccharide (anh) containing N-glycolylmuramyl group
MeOH/H2O 2 : 1 (12 mL), then triethylamine (16 μL, and tripeptide (1f). Compound 27 (50 mg, 0.0469) and
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
Pd(OH)2 (60 mg, 0.427 mmol) were mixed well in THF/H2O/ 4.09 (1H, q,
J = 7.6 Hz; D-Ala-αH̲), 4.01–3.97 (2H, m;
AcOH 1 : 1 : 0.1 (2.1 mL), and reacted under H2 atmosphere –COCH
(2.0 MPa) for 24 h. The Pd(OH)2 was removed by membrane fil- m; H-2′, H-5, H-5′, H-6′, H-3, H-6, DAP 6-H
tration, the filtrate was concentrated in vacuo. The residue was H-3′), 3.36–3.33 (2H, m; H-6′, H-4′), 2.36–2.33 (2H, m; Gln-
resuspended with ultrapure H2O and lyophilized to give 1f as a γCH2), 2.12–2.06 (1H, m; Gln-βCH), 1.97 (3H, s; –NHCOCH3),
white solid (20 mg, 65%). 1.94–1.64 (5H, m; Gln-βCH, DAP 3-CH2, 5-CH2), 1.41–1.24
1H NMR (500 MHz, D2O): δ 5.42 (1H, s; H-1), 4.63 (1H, d, J = (11H, m; DAP 4-CH ); 13C
2, Ala-βCH3, Lac-βCH3 D-Ala-βCH3
4.6 Hz; H-5), 4.30–4.20 (4H, m; Ala-αH, Gln-αH, H-6, DAP 2-H), NMR (150 MHz, D2O): δ 179.5, 175.9, 175.7, 175.2, 175.0,
4.11 (1H, q, J = 6.8 Hz; Lac-αH), 4.01 (2H, d, J = 2.9 Hz; 175.0, 174.7, 174.5, 172.9, 100.4, 90.2, 79.2, 77.1, 76.1, 75.6,
–NHCOCH2OH), 3.95 (1H, s; H-2), 3.84 (1H, s; H-4), 3.77–3.74 73.6, 71.2, 70.3, 61.2, 61.0, 59.8, 56.1, 54.5, 53.8, 53.5, 52.9,
(2H, m; DAP 6-H, H-6′), 3.35 (1H, s; H-3), 2.35–2.32 (2H, m; 50.8, 49.9, 31.6, 30.9, 30.0, 26.9, 22.1, 21.0, 18.4, 17.5, 16.6;
Gln-γCH2), 2.14–2.07 (1H, m; Gln-βCH), 1.95–1.66 (5H, m; Gln- HRMS (ESI-LIT-orbitrap): m/z: calcd for C37H62N8O21Na:
βCH, DAP 3-CH2, 5-CH2), 1.45–1.40 (2H, m; DAP 4-CH2
), 1.36 977.3922 [M + Na]+; found: 977.3939.
(3H, d, J = 7.2 Hz; Ala-βCH3), 1.30 (3H, d, J = 6.8 Hz; Lac-βCH3);
Compound 2f. 1H NMR (500 MHz, D2O): δ 5.41 (1H, s; Hanh
13C NMR (125 MHz, D2O): δ 176.3, 175.5, 175.4, 175.2, 175.1, 1), 4.75–4.65 (2H, m; Hanh-5, H-1), 4.30 (1H, q, J = 7.5 Hz; Ala-
174.5, 101.7, 80.7, 77.5, 77.1, 70.3, 66.3, 62.5, 54.7, 54.1, 53.5, αH), 4.25–4.22 (2H, m; Gln-αH, DAP 2-H), 4.19 (1H, dd, J = 5.0
50.7, 33.0, 32.1, 31.3, 30.1, 28.8, 22.6, 18.5, 17.9; HRMS Hz, 8.5 Hz; Hanh-6), 4.13 (1H, q, J = 6.5 Hz; Lac-αH), 4.06 (2H,
̲
2OH), 3.88–3.78 (3H, m; H-2, H-6, H-4), 3.75–3.60 (7H,
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
,
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
-
̲
̲
̲
(ESI-LIT-orbitrap): m/z: calcd for C26H42N6O14Na: 685.2651 s; –NHCOCH
̲
2OH), 3.98 (1H, s; Hanh-2), 3.92 (1H, s; Hanh-4),
[M + Na]+; found: 685.2661.
3.83–3.80 (1H, m; H-6), 3.76 (1H, dd, J = 6.0 Hz, 8.5 Hz; Hanh
-
Monosaccharide (anh) containing N-glycolylmuramyl group 6′), 3.71–3.64 (3H, m; DAP 6-H̲, H-5, H-2), 3.59 (1H, s; Hanh-3),
and tetrapeptide (1h). Compound 1h was synthesized from 28 3.54–3.49 (1H, m; H-3), 3.40–3.39 (2H, m; H-6, H-4), 2.35–2.32
with similar method to the synthesis of 1f (quant.). (2H, m; Gln-γCH2), 2.15–2.08 (1H, m; Gln-βCH), 1.99 (3H, s;
1H NMR (500 MHz, D2O): δ 5.56 (1H, s; H-1), 4.77 (1H, br s; –NHCOCH
3), 1.96–1.64 (5H, m; Gln-βCH, DAP 3-CH2, 5-CH2),
H-5), 4.47–4.34 (5H, m; D-Ala-αH, Ala-αH, Gln-αH, H-6, DAP 1.44–1.36 (5H, m; DAP 4-CH2, Ala-βCH3), 1.31 (3H, d, J = 6.5
2-H), 4.25 (1H, q, J = 6.8 Hz; Lac-αH), 4.15 (2H, d, J = 3.3 Hz; Hz; Lac-βCH
3); 13C NMR (125 MHz, D2O): δ 178.1, 176.9, 176.5,
–NHCOCH2OH), 4.09 (1H, s; H-2), 4.01–3.97 (2H, m; DAP 6-H 176.0, 175.9, 175.7, 175.4, 175.0, 101.9, 100.7, 77.7, 77.1, 77.0,
H-4), 3.89 (1H, dd, J = 5.7 Hz, 7.9 Hz; H-6′), 3.49 (1H, dd, J = 75.9, 75.1, 74.5, 70.7, 65.9, 61.9, 61.6, 56.7, 55.4, 54.7, 54.0,
1.6 Hz, 1.6 Hz; H-3), 2.48–2.44 (2H, m; Gln-γCH2), 2.28–2.21 50.6, 49.5, 32.6, 31.6, 31.0, 27.8, 23.2, 22.1, 19.0, 17.7; HRMS
(1H, m; Gln-βCH), 2.06–1.79 (5H, m; Gln-βCH, DAP 3-CH2 (LTQ-orbitrap MS): m/z: calcd for C34H56N7O19: 866.3625
5-CH2), 1.59–1.52 (2H, m; DAP 4-CH2
), 1.51–1.43 (9H, m; Ala- [M + H]+; found: 866.3633.
βCH3, Lac-βCH3 D-Ala-βCH3
); 13C NMR (125 MHz, D2O): δ
177.6, 176.9, 176.7, 176.1, 175.1, 174.7, 174.7, 100.9, 79.4, 77.0,
77.0, 68.8, 66.3, 61.9, 55.0, 54.7, 53.9, 50.8, 49.9, 49.6, 32.5, αH
31.7, 30.9, 27.9, 22.0, 19.1, 17.7, 17.3; HRMS (ESI-LIT-orbitrap): (1H, q, J = 6.9 Hz; Lac-αH
m/z: calcd for C29H48N7O15: 734.3203 [M + H]+; found: (1H, s; Hanh-2), 3.92 (1H, s; Hanh-4), 3.83–3.80 (1H, m; H-6),
734.3216. 3.77–3.64 (4H, m; Hanh-6′, DAP 6-H, H-5, H-2), 3.59 (1H, s;
The synthetic procedures of 2b, 2d, 2f and 2h in Scheme 3 anh-3), 3.54–3.49 (1H, m; H-3), 3.40–3.39 (2H, m; H-6, H-4),
were shown in ESI.† 2.39–2.28 (2H, m; Gln-γCH2), 2.15–2.08 (1H, m; Gln-βCH), 1.99
Compound 2b. 1H NMR (700 MHz, D2O): δ 5.26 (0.62 H, d, (3H, s; –NHCOCH
3), 1.93–1.65 (5H, m; Gln-βCH, DAP 3-CH2
J = 3.6 Hz; H-1β), 4.62 (1H, q, J = 6.7 Hz; Lac-αH), 4.52 (1H, dd, 5-CH2), 1.44–1.30 (11H, m; DAP 4-CH2, Lac-βCH3, Ala-βCH3
J = 7.9 Hz, 7.9 Hz; H-1′), 4.23–4.27 (2H, m; Ala-αH, Gln-αH), D-Ala-βCH3
4.21–4.19 (1H, m; DAP 2-H), 4.06–4.02 (2H, m; –COCH2OH), 175.7, 175.6, 175.6, 174.7, 174.5, 174.3, 101.5, 100.3, 77.5, 76.8,
3.93–3.83 (3H, m; H-2, H-6, H-4), 3.80–3.64 (7H, m; H-2′, H-5, 76.7, 75.5, 74.7, 74.2, 70.4, 65.6, 61.6, 61.3, 56.3, 54.8, 54.3,
H-5′, H-6′, H-3, H-6, DAP 6-H), 3.57–3.51 (1H, m; H-3′), 53.4, 50.3, 49.6, 49.2, 32.1, 31.3, 30.5, 27.5, 22.9, 21.6, 18.7,
3.41–3.36 (2H, m; H-6′, H-4′), 2.39–2.37 (2H, m; Gln-γCH2), 17.4, 17.0; HRMS (LTQ-orbitrap MS): m/z: calcd for
2.16–2.13 (1H, m; Gln-βCH), 2.02 (3H, s; –NHCOCH
3), C37H60N8O20Na: 959.3816 [M + Na]+; found: 959.3821.
2.00–1.68 (5H, m; Gln-βCH, DAP 3-CH2, 5-CH2), 1.46–1.33 (8H, HEG1G cells bioassay for Nod1 activation. HEG1G cells
m; DAP 4-CH2, Ala-βCH3, Lac-βCH3
); 13C NMR (176 MHz, D2O): were cultured in Dullbecco’s modified Eagle Media (purchased
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲,
̲
̲
̲
̲ ,
̲
̲
̲
̲
,
̲
Compound 2h. 1H NMR (500 MHz, D2O): δ 5.41 (1H, s;
anh-1), 4.65 (2H, br s; Hanh-5, H-1), 4.31–4.26 (2H, m; D-Ala-
H
̲
̲
̲, DAP 2-H̲
̲
̲
̲
H
̲
̲
̲
̲
̲
̲
,
,
̲
̲
̲
̲
̲
̲
̲
); 13C NMR (125 MHz, D2O): δ 177.3, 176.5, 176.3,
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
δ 177.6, 176.0, 175.7, 175.3, 175.0, 174.7, 174.7, 174.5, 100.5, from Sigma), containing 10% heat-inactivated fetal calf serum
90.2, 79.2, 77.1, 76.1, 75.6, 73.6, 71.2, 70.3, 61.2, 61.0, 59.8, with 100 units per mL penicillin and 100 μg mL−1 strepto-
56.1, 54.5, 54.1, 53.5, 53.1, 49.9, 31.8, 30.8, 30.0, 26.9, 22.2, mycin (all culture reagents were purchased from Gibco). The
21.2, 18.4, 16.6; HRMS (ESI-LIT-orbitrap): m/z: calcd for assay procedure was as follows: 5 × 105 cells per mL HEG1G
C34H57N7O20Na: 906.3551 [M + Na]+; found: 906.3566.
cells were seeded to black flat-bottom 96-well plates; incubated
Compound 2d. 1H NMR (600 MHz, D2O): δ 5.21 (0.57 H, d, overnight before being treated with appropriate ligands for
J = 3.5 Hz; H-1β), 4.57 (1H, q, J = 6.5 Hz; Lac-αH), 4.49–4.46 24 h; NF-κB activation was detected by measuring the GFP
(1H, m; H-1′), 4.28–4.20 (3H, m; Ala-αH, Gln-αH, DAP 2-H), fluorescence at 535 nm with Infinite M200 pro.
̲
̲
̲
̲
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
HEK293T cells bioassay for Nod2 activation. Human
embryonic kidney (HEK) 293 T was cultured in Dullbecco’s
modified Eagle Media (purchased from Sigma), containing
10% heat-inactivated fetal calf serum with 100 units per mL
penicillin and 100 μg mL−1 streptomycin (all culture reagents
were purchased from Gibco). pMX2-HA-Nod2, pBxIV-luc and
pEF1BOS-β-gal plasmids were gifts from N. Inohara (University
8 N. Inohara, Y. Ogura, A. Fontalba, O. Gutierrez, F. Pons,
J. Crespo, K. Fukase, S. Inamura, S. Kusumoto,
M. Hashimoto, S. J. Foster, A. P. Moran, J. L. Fernandez-
Luna and G. Nunez, J. Biol. Chem., 2003, 278, 5509–5512.
9 S. E. Girardin, I. G. Boneca, J. Viala, M. Chamaillard,
A. Labigne, G. Thomas, D. J. Philpott and P. J. Sansonetti,
J. Biol. Chem., 2003, 278, 8869–8872.
of Michigan Medical School, Michigan, US). The assay pro- 10 S. K. Angala, J. M. Belardinelli, E. Huc-Claustre,
cedure was as follows: 0.5 × 105 cells per mL HEK293T cells
were seeded to transparent 24-well plate and incubated over-
W. H. Wheat and M. Jackson, Crit. Rev. Biochem. Mol. Biol.,
2014, 49, 361–399.
night before transfection; transfected plasmids Nod2 (20 ng 11 S. E. Girardin, I. G. Boneca, L. A. Carneiro, A. Antignac,
per well pMX2-HA-Nod2) and reporter plasmid: NF-κB-depen-
dent pBxIV-luc (5 ng per well) and control pEF1BOS-β-gal (20
ng per well) were transfected into cells by calcium phosphate
method; 16 h post-transfection, cells were treated with various
M. Jehanno, J. Viala, K. Tedin, M. K. Taha, A. Labigne,
U. Zahringer, A. J. Coyle, P. S. DiStefano, J. Bertin,
P. J. Sansonetti and D. J. Philpott, Science, 2003, 300, 1584–
1587.
ligands; 24 h later, ligand-dependent NF-κB activation was 12 J. B. Raymond, S. Mahapatra, D. C. Crick and M. S. Pavelka
determined by the luciferase reporter assay with FDSS.
Jr., J. Biol. Chem., 2005, 280, 326–333.
13 F. Coulombe, M. Divangahi, F. Veyrier, L. de Leseleuc,
J. L. Gleason, Y. Yang, M. A. Kelliher, A. K. Pandey,
C. M. Sassetti, M. B. Reed and M. A. Behr, J. Exp. Med.,
2009, 206, 1709–1716.
Acknowledgements
14 J. K. Iyer and K. M. Coggeshall, J. Immunol., 2011, 186,
3841–3845.
This work was supported in part by Grants-in-Aid for Scientific
Research (no. 20241053, 22310136, and 26282211) from the
Japan Society for the Promotion of Science, by ERATO Murata
Lipid Active Structure Project, by a funding program for Next
Generation World-Leading Researchers (NEXT Program;
LR025) from JSPS and CSTP, by Keio Gijuku Academic
Development Funds, by Yamada Science Foundation, and by
The Sumitomo Foundation.
15 J. E. Melnyk, V. Mohanan, A. K. Schaefer, C. W. Hou and
C. L. Grimes, J. Am. Chem. Soc., 2015, 137, 6987–6990.
16 S. Inamura, Y. Fujimoto, A. Kawasaki, Z. Shiokawa,
E. Woelk, H. Heine, B. Lindner, N. Inohara, S. Kusumoto
and K. Fukase, Org. Biomol. Chem., 2006, 4, 232–242.
17 N. Wang, C. Y. Huang, M. Hasegawa, N. Inohara,
Y. Fujimoto and K. Fukase, ChemBioChem, 2013, 14, 482–
488.
18 A. Kawasaki, Y. Karasudani, Y. Otsuka, M. Hasegawa,
N. Inohara, Y. Fujimoto and K. Fukase, Chem. – Eur. J.,
2008, 14, 10318–10330.
References
1 WHO. Global tuberculosis report 2014. World Health
Organization, 2014. http://www.who.int/tb/publications/ 19 A. Uehara, Y. Fujimoto, A. Kawasaki, S. Kusumoto,
global_report/en/ (accessed Dec 30, 2014).
2 C. V. Harding and W. H. Boom, Nat. Rev. Microbiol., 2010,
8, 296–307.
3 O. Takeuchi and S. Akira, Cell, 2010, 140, 805–820.
4 T. H. Mogensen, Clin. Microbiol. Rev., 2009, 22, 240–273.
5 J. V. Holtje, Microbiol. Mol. Biol. Rev., 1998, 62, 181–203.
K. Fukase and H. Takada, J. Immunol., 2006, 177, 1796–
1804.
20 C. L. Higginbotham and J. U. Izunobi, J. Mol. Struct., 2010,
977, 153–164.
21 Y. Saito, Y. Yoshimura, H. Wakamatsu and H. Takahata,
Molecules, 2013, 18, 1162–1173.
6 R. Caruso, N. Warner, N. Inohara and G. Nunez, Immunity, 22 M. Kunishima, C. Kawachi, K. Hioki, R. Terao and S. Tani,
2014, 41, 898–908. Tetrahedron, 2001, 57, 1551–1558.
7 M. Chamaillard, M. Hashimoto, Y. Horie, J. Masumoto, 23 J. Masumoto, K. Yang, S. Varambally, M. Hasegawa,
S. Qiu, L. Saab, Y. Ogura, A. Kawasaki, K. Fukase,
S. Kusumoto, M. A. Valvano, S. J. Foster, T. W. Mak,
G. Nunez and N. Inohara, Nat. Immunol., 2003, 4, 702–707.
S. A. Tomlins, S. Qiu, Y. Fujimoto, A. Kawasaki, S. J. Foster,
Y. Horie, T. W. Mak, G. Nunez, A. M. Chinnaiyan, K. Fukase
and N. Inohara, J. Exp. Med., 2006, 203, 203–213.
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem.