Towards the preparation of synthetic outer membrane vesicle models with micromolar affinity to wheat germ agglutinin using a dialkyl thioglycoside

Article abstract of DOI:10.3762/bjoc.15.90

A series of alkyl thioglycosides and alkyl thiodiglycosides bearing glucose and N-acetylglucosamine residues were prepared by thiol-ene coupling in moderate to good yields (40-85%). Their binding ability towards wheat germ agglutinin was measured by competitive enzyme-linked lectin assays. One of the synthetic compounds presenting two GlcNAc units showed the highest inhibitory effect of this study with an IC₅₀ of 11 μM corresponding to a 3182-fold improvement compared to GlcNAc. These synthetic molecules were used to produce giant vesicles, alone or in mixture with phospholipids, mimicking bacterial outer membrane vesicles (OMV) with potential antiadhesive properties.

Full text of DOI:10.3762/bjoc.15.90

Towards the preparation of synthetic outer membrane vesicle
models with micromolar affinity to wheat germ agglutinin
using a dialkyl thioglycoside
Dimitri Fayolle1, Nathalie Berthet‡2, Bastien Doumeche‡1, Olivier Renaudet2,
Peter Strazewski1 and Michele Fiore*,‡1
Full Research Paper
Open Access
Address:
Beilstein J. Org. Chem. 2019, 15, 937–946.
1Institut de Chimie et Biochimie Moléculaires et Supramoléculaires,
Université de Lyon, Claude Bernard Lyon 1, 43 blvd du 11 novembre
1918, 69622 Villeurbanne Cedex, France and 2Univ. Grenoble Alpes,
CNRS, DCM UMR 5250, F-38000 Grenoble, France
doi:10.3762/bjoc.15.90
Received: 10 December 2018
Accepted: 02 April 2019
Published: 17 April 2019
Email:
Michele Fiore* - michele.fiore@univ-lyon1.fr
Associate Editor: K. N. Allen
* Corresponding author ‡ Equal contributors
© 2019 Fayolle et al.; licensee Beilstein-Institut.
License and terms: see end of document.
Keywords:
glycolipids; outer membrane vesicles; synthetic vaccines
Abstract
A series of alkyl thioglycosides and alkyl thiodiglycosides bearing glucose and N-acetylglucosamine residues were prepared by
thiol–ene coupling in moderate to good yields (40–85%). Their binding ability towards wheat germ agglutinin was measured by
competitive enzyme-linked lectin assays. One of the synthetic compounds presenting two GlcNAc units showed the highest
inhibitory effect of this study with an IC50 of 11 µM corresponding to a 3182-fold improvement compared to GlcNAc. These syn-
thetic molecules were used to produce giant vesicles, alone or in mixture with phospholipids, mimicking bacterial outer membrane
vesicles (OMV) with potential antiadhesive properties.
Introduction
Outer membrane vesicles (OMV) [1], lipid bilayer vesicles re- [6,7] or Gram-negative bacteria [7] are challenges for the cur-
leased from the outer membrane of Gram-negative bacteria, are rent research in the field. Artificial OMV, composed of synthe-
considered today as attractive candidates for vaccine delivery. tic and non-toxic, non-immunogenic phospholipids and glyco-
However, they have some disadvantages, they are not easy to lipids are good candidates for drug or vaccine delivery. One of
produce and are difficult to characterize [2,3]. Moreover, toxic the most common reactions used to prepare monoalkyl glyco-
lipopolysaccharides (LPS) present in the outer membrane of sides is the Fisher reaction between pyranoses and fatty alco-
most of Gram-negative bacteria prevent the medical use of hols of different lengths [8,9]. Alkyl thioglycosides are known
OMV from natural sources [4]. Licensed vaccines based on for their properties as co-surfactants [10] and present interest-
crude OMV are currently available to contribute to the preven- ing antimicrobial activities [11], acting as glycosidase inhibi-
tion and to control at least twenty-five infections including pul- tors and being resistant towards glycoside hydrolases [12,13].
monary ones [5]. Developing synthetic vaccines against cancer Some of them have been obtained by coupling protected
937

Beilstein J. Org. Chem. 2019, 15, 937–946.
Results and Discussion
Preparation of alkyl glycosides from
glycosyl thiolates and n-alkyl halides [14-16]. Moreover,
mechanochemical thioglycosylation of glycosyl acetates was
used for the synthesis of n-alkyl 1-thio-α- -glycosides as
carbohydrate mesogens [17]. Unfortunately, the preparation of
alkyl glycosides cannot be carried out with unprotected thio-
glycosides, implying orthogonal protection and deprotection
steps in order to obtain unprotected glycolipids. We decided to
investigate the formation of artificial OMV composed of long
chain alkyl thioglycosides synthesized by the photoinduced
radical addition of thiols to alkenes – known as thiol–ene cou-
pling, TEC – a very efficient metal-free click reaction [18,19].
We prepared n-alkyl 1-thio-β- -glycosides from unprotected
sugar anomeric thiols, commercial n-alkenes and one synthetic
lipophilic scaffold presenting two reactive alkenyl ends. Al-
though thiyl-radical-mediated reactions have been extensively
investigated for the preparation of carbohydrate derivatives [20]
and some dithioether phospholipid and glycolipid analogues
unprotected sugars and lipophilic scaffolds
Thioglycolipids are not native in OMV [1], but present several
advantages compared to natural and synthetic glycolipids,
linked by a chemically and enzymatically non-stable acetal
function. Owing to their straightforward access, and the stability
of thioether conjugates, 1-thio-β- -glucopyranose (1a,
Figure 1), and 2-acetamido-2-deoxy-1-thio-β- -glucopyranose
(1b, Figure 1) have been selected for this study and were pre-
pared as previously reported [23]. Commercially available
1-decene (2a) and 1-tetradecene (2b) were selected for the syn-
thesis of n-alkyl 1-thio-β- -glycosides (3–5, Scheme 1),
under UV-A irradiation (λmax 365 nm), in the presence of
the photoinitiator 2,2-dimethoxy-2-phenylacetophenone
(DPAP).
[21,22], no examples were reported for the synthesis of n-alkyl The lipophilic scaffold hepta-1,6-dien-4-yl tetradecanotate (6,
thioglycosides by using thiol–ene coupling [18,19]. In addition Scheme 2) was obtained from hepta-1,6-dien-4-ol and served
to its high efficiency and selectivity, the TEC reaction does not for the preparation of compounds 7 and 8. Hepta-1,6-dien-4-ol
require any metal, a key feature for the preparation of potential was prepared according to experimental procedures previously
vaccines.
reported by others [24,25], then the myristoyl group (C14:0) was
Figure 1: Structure of the β-thiols 1a and 1b and of the commercial alkenes 2a and 2b.
Scheme 1: Synthesis of the n-alkyl thioglycosides 3–5, 7 and 8. Detailed reaction conditions are reported in the experimental part; i) 1a or 1b
(3 equiv), 2a or 2b (1 equiv), DPAP (0.3 equiv); ii) 1a or 1b (3 equiv), 6 (1 equiv), DPAP (0.6 equiv). TECs were performed at room temperature under
UV light (λmax 365 nm) for 30 min (with 2a and 2b) or 60 min (with 6) in methanol (2a) or DMF (2b and 6); DPAP = 2,2-dimethoxy-2-phenylaceto-
phenone.
Scheme 2: Synthesis of the lipophilic scaffold 6; DMAP = N,N-dimethylaminopyridine.
938

Beilstein J. Org. Chem. 2019, 15, 937–946.
added using myristoyl chloride in the presence of DMAP formed thioacetal bond, indicate the formation of 3. The aro-
(Scheme 1) [26].
matic protons of DPAP (δH = 8.00 and 7.20 ppm, not shown)
and the aliphatic chain signals (δH = 1.30 and 0.75 ppm) did not
shift during the irradiation time. The product was not isolated
and the reaction was repeated on a 50 mg scale in methanol,
giving 3 in good yields (75%). The poor solubility of 2b in
NMR monitoring
Following the formation of thioglycosides by
1H NMR analysis
A test reaction, based on 10 mg of 1a, was performed in deuter- methanol prompted us to use N,N-dimethylformamide (DMF)
ated methanol (MeOD) and the formation of compound 3 was for the synthesis of 4 and 5. Compounds 4 and 5 were isolated
monitored by periodical 1H NMR experiments, as illustrated in in good yields (55 and 75%, respectively) [27]. The syntheses
Supporting Information File 1, Figure S1. The signal corre- of compounds 7 and 8 were carried out in DMF as well. As sub-
sponding to the double bond of 2a (δH = 5.80–5.60 ppm) disap- stantial amounts of 6 were isolated after the reaction, indicating
peared together with the signals of the α-methylene of 2a an incomplete conversion to 8 (53%), the course of the reaction
(δH = 1.96 ppm). The appearance of a new doublet at 4.24 ppm was followed by 1H NMR in deuterated DMF (Figure 2). The
with a coupling constant (J = 9.7 Hz) typical for an anomeric formation of 8 occurred with a progressive disappearing of the
proton with β configuration, together with a new signal at signals of the alkenyl group of 6 (δH= 5.70–5.55 ppm) in one
2.60 ppm corresponding to the α-methylene of the newly hour. The anomeric proton of the thiol 1b (δH = 4.63 ppm)
Figure 2: Periodic monitoring by 1H NMR (300 MHz, DMF-d7) of the formation of product 8 from a mixture compounds 1b and 6. a) Spectrum of pure
1b; b) through d) crude reaction mixture after 0, 30 and 60 minutes of irradiation, respectively.
939

Beilstein J. Org. Chem. 2019, 15, 937–946.
progressively decreased in intensity with apparition of 30]. GVs were prepared by hydrating a thin film containing
new signals corresponding to the new anomeric protons 90:10 (mol %) of those glycoconjugates in mixture with a
(δH = 5.26 ppm) and the α-CH2 of the thioether bond natural phospholipid: 1-palmitoyl-2-oleoyl-sn-glycero-3-phos-
(δH = 1.70 ppm). Compound 8 was isolated by column chroma- phocholine (POPC). As expected, the mixture once hydrated
tography in 75% yield and no traces of starting material 6 were showed a high tendency to vesiculate (Figure 3, images a–e and
found. In addition, no intramolecular addition products neither h). Moreover, a thin layer of compounds 5 or 8 without POPC,
mono-glycosylated compounds were observed.
hydrated in the same conditions, gave similar giant vesicles
(Figure 3, images f,g and i). Microscopic observation suggested
that those supramolecular assemblies microscopically resemble
the well-known outer membrane vesicles (OMV) [1]. The tenta-
Preparation of vesicles upon hydration of a
thin film of phospholipids and glycolipids
Once synthesized, the compounds 5 and 8 bearing GlcNAc tive dissolution of compounds 5–8 in PBS (pH 7.4) produced a
residues were used to produce giant vesicles (GVs) upon hydra- slightly turbid suspension, suggesting that those molecules
tion with PBS (pH 7.4) by modifying reported procedures [28- formed a mixture of microscopically visible vesicles (diameter
Figure 3: Micrographs of giant vesicles and lipid aggregates obtained from the gentle hydration (in PBS, pH 7.4) of compounds 8 and 5, alone and in
binary and ternary mixtures with POPC. a–e) POPC/8, 9:1 molar ratio; f and g) pure 8; h) 5/8, 1:1 molar ratio; i) pure 5; All the samples were pre-
pared with 1 mM lipids and stained with NileRed® (1 mM in EtOH, 1 µL, λ[excit] = 561 nm) before the microscopic observation. Red arrows are used to
indicate small GVs. The scale bar is 20 µm for all micrographs in this figure.
940

Beilstein J. Org. Chem. 2019, 15, 937–946.
>0.5 µm), small vesicles (diameter <0.5 µm) and perhaps ples of biologically active carbohydrate-based compounds with
micelles (diameter in the nanometer range). However, due to high affinity to lectins and with antibiotic activities [34], lectin
our interest to study the interaction of OMV-like objects [1] we recognition by n-alkyl thioglycoside liposomes remains
prepared giant vesicles resembling the OMV by using mixtures unprecedented [33,34]. Moreover, the calculated distance be-
of fully synthetic compounds and natural occurring phospho- tween the two sugar moieties of compound 8 (up to 13 Å,
lipids. The vesicles showed stability over time; however, Figure 6) and the presence of flexible arms make these com-
osmotic stress, sudden temperature change and accidental pounds similar to other efficient multivalent glycoconjugates
drying of the hosting solution inevitably cause the disruption of obtained by TEC of sugar thiols with different multivalent scaf-
the supramolecular assembly [23]. Analyses under saline stress folds such as octasilsesquioxanes [35], cyclopeptides [36,37]
were not performed as the ultimate aim is to use of those OMV and polymers [38].
models under physiological conditions.
Biological evaluation of the synthesized com-
pounds
Competitive enzyme-linked lectin assays (ELLA) were used to
evaluate the binding properties of compounds 5 and 8 towards
wheat germ agglutinin (WGA), a lectin from Triticum vulgaris
that is specific for N-acetylneuraminic acid and N-acetylglucos-
amine. This assay consists in measuring the ability of the com-
pounds to inhibit the binding of WGA horseradish peroxidase-
labelled (WGA-HRP) to a GlcNAc-polyacrylamide conjugate
following the procedure described in Figure 4.
Figure 5: Inhibition curves for the binding of WGA-HRP to PAA-
GlcNAc by D-GlcNAc The symbols (■), ( ) and (○) represent the
monomer (GlcNAc) and the lipophilic thioglycoconjugates 8 and 5, re-
spectively.
Both compounds 5 and 8 showed low IC50 values (a tenth to
some hundreds of micromolars) where the GlcNAc monomer
displays only 35 mM (Table 1). Compounds 4 and 7 bearing
one and two glucose residues served as negative controls as
WGA has no affinity towards glucose. This result clearly shows
the high potential of the lipophilic n-alkyl thioglycoconjugates
for lectine binding. In particular, compound 8 showed the
highest inhibitory effect of this study with an IC50 of 11 µM
corresponding to a 3000-fold improvement compared to the
monomer control. This result suggests a higher participation of
Figure 4: A simplified (and not in scale) representation of the ELLA
sugar units in the lectin binding for this compound.
assay, to study the interaction between GVs or micelles of compound
8 or 5 and WGA. OPD: O-phenylenediamine dihydrochloride; PAA:
poly[N-(2-hydroxyethyl)acrylamide].
Table 1: Inhibition of the adhesion of WGA-HRP to GlcNAc-coated
microtiter plates as determined by ELLAa.
WGA is a homodimer in which each monomer is organized into
four domains (A–D) containing adjacent “primary” (B and C)
and “secondary” (A and D) domains binding sites with differ-
ent affinities to GlcNAc [31]. The proximity of adjacent binding
sites (≈14 Å) [32] makes this lectin an excellent candidate to in-
vestigate multivalent carbohydrate–protein interactions [33].
Although the scientific literature is rich of remarkable exam-
Compound
IC50 (μM)
rpb
-GlcNAc
35 000
142
1
5
8
246
3182
11
aExperiments were performed in triplicate except for compound 5. brp:
relative potency = IC50(lipophilic thioglycoconjugate)/IC50(GlcNAc).
941

Beilstein J. Org. Chem. 2019, 15, 937–946.
Docking calculation for model glycolipids
To appreciate the possible binding of 4, 5, 7 and 8 onto WGA, a
docking simulation was performed using the crystal structure of
WGA obtained in the presence of GlcNAc (PDB 2UVO) and of
other GlcNAc derivatives (PDB 2X52 and 4AML). The
docking experiments were performed using analogous mole-
cules bearing acetate and butanoate alkyl chains (namely 4C2,
4C4, 5C2, 5C4, 7C2, 7C4, 8C2 and 8C4 according to the alkyl
length, see Figures S5 and S6 in Supporting Information File 1)
instead of the long alkyl chains, for calculation simplicity.
The docking protocol was validated using GlcNAc and Glc as
ligands. In such a case, the poses found with GlcNAc were
strictly identical to the ones observed in the crystal structure
while no poses were obtained using Glc as ligand. Docking ex-
periments using Glc derivatives 4C2, 4C4, 7C2 and 7C4 do not
lead to any poses, suggesting, as observed in the ELLA experi-
ments, that WGA do not bind them.
About half of the poses obtained from compounds 5C2 and 5C4
(bearing a single sugar moiety) were found on the primary (BC)
and secondary (AD) binding sites with energies ranging from
−4.5 to −4.9 kcal mol−1 (pose A, Figure 6 and Supporting Infor-
mation File 1, Figure S5) [31,32,39,40]. The binding of 8C2 and
8C4 onto WGA lead to two main poses. The first one (energies
ranging from −5.3 to −6.2 kcal mol−1) corresponds to the
binding of both GlcNAc moieties onto the secondary binding
sites D2A1 and A2 (pose B, Figure 6). None of the poses were
found to occupy primary binding sites as it was observed for
5C2 and 5C4. The first GlcNAc moiety of 8C2 and 8C4 inter-
acts with Asp29A, Asp129B, Ser148B and Tyr159B and nearly
superimposes with GlcNAc in the WGA crystalline structure
(site D2A1). The second GlcNAc moiety is at the vicinity of
Ser19A and Tyr30A residues (A2 binding site), therefore
mimicking the binding of two GlcNAc molecules on WGA. The
distance between the two GlcNAc residues is 11–12 Å [41].
The alkyl chains (acetyl or butyl) fit in the non-polar environ-
ment provided by Trp150B and by Gly158B or by the acyl
moiety of Lys33A (pose B, Figure 6). Nevertheless, there is no
obvious involvement of the ester bond in the binding of 8C2 or
8C4 at these secondary sites. The other main poses of 8C2 and
8C4 present the ligands adopting a linear conformation in a cleft
between both monomers of WGA with similar energies (−5.3 to
−6.2 kcal mol−1, pose C, Figure 6 and Figure S6 in Supporting
Figure 6: Main poses obtained from docking experiments. WGA (PDB
2UVO) surface is shown in white for monomer A and in blue for mono-
mer B. Residues within 4 Å of the docked compound are in white stick
representation. A) Representative poses obtained for 5C2 (yellow
sticks) and 5C4 (orange sticks) at the B1C2 primary site superim-
posed with co-crystallized GlcNAc (white sticks). B) Representative
poses of 8C4 (orange and yellow sticks) obtained at the D2A1 and A2
binding superimposed with co-crystalized GlcNAc (white sticks).
C) Representative poses of 8C2 (yellow sticks) and 8C4 (orange sticks)
in the cleft between the two WGA monomers.
Information File 1). This cleft is defined by neutral residues with Asn143B, the carbonyl of Cys98B or Asn14A on the other
Asn14A and Asn101B but also by the non-polar Leu16A and side. The alkyl chains of 8C2 and 8C4 are found close to
Ile155B that provide hydrophobic environment for the heptanol Ile155B and head out of the cleft if a longer alkyl chain is used
moiety in 8C2 and 8C4. This region of the protein also lacks (as in 8). These observations suggest that the herein synthe-
charged residues that would prevent 8C2 or 8C4 from binding. sized bidentate analogues of GlcNAc would interact with WGA
GlcNAc moieties are proposed to interact with the hydroxy differently than compounds bearing one sugar residue. The
group of Ser8B and the carbonyl of Cys24A from one side and short distance between the two sugar moieties as well as the
942

Beilstein J. Org. Chem. 2019, 15, 937–946.
presence of the non-polar alkyl chains seems to prevent the solvents were from Fischer Scientific (Illkirch, France). Deuter-
binding to primary sites.
ated solvents for NMR were from Euriso-top (France). All
sensitive reactions were carried out in oven-dried glassware
under an inert argon atmosphere. Room temperature (rt) refers
Conclusion
A series of glucose and N-acetyl glucosamine-based thioglycol- to 20–25 °C. All thiol–ene couplings were performed with a
ipids were synthesized for the first time by using thiol–ene cou- Philips UVA lamp irradiating at 365 nm (Mgc typ838 150 W,
pling in the absence of protecting groups and in good yields. 6 tubes). Thin-layer chromatography (TLC) was carried out on
Monosaccharides where selected on the basis of similar aluminum sheets coated with silica gel 60 F254 (Merck, Kenil-
researches carried out in the field [38]. We reported a simpli- worth, USA). TLC plates were inspected by UV light
fied access to n-alkyl thioglycosides as building blocks for the (λ = 254 nm) and developed by treatment with a mixture of
preparation of OMV models. The obtained n-alkyl thioglyco- 10% H2SO4 in EtOH/H2O (1:1 v/v), followed by heating. Elec-
sides were tested for the first time for lectin recognition and trospray-ionization mass spectra (ESIMS) were recorded using
showed promising results, as their aggregates in PBS displayed a Bruker Q-Tof Micromass spectrometer. NMR spectra were re-
micromolar affinity with WGA. In addition, we produced multi- corded in DMSO-d6, CDCl3, CD3OD or DMF-d7 on a Bruker
lamellar giant vesicles of variable size from the glycolipids Avance 300 spectrometer at 300 and 75 MHz and on a Bruker
alone or in mixture with phospholipids. The vesicles we ob- Avance 400 at 400 and 100 MHz for 1H, and 13C, respectively.
tained showed similarities with OMV structures observed by The attached proton test experiment (APT) was performed to
Holst and co-workers (cf. Supporting Information File 1, Figure replace the 13C experiment when required. Chemical shifts of
S4) [1]. This work suggests that mixtures of phospholipids such solvents (CDCl3: δH = 7.26 and δC = 77.13; CD3OD: δH = 3.31
as POPC and small amounts of bioactive glycolipids should and δC = 49.50; DMF-d7 δH = 8.01, and δC = 163.15) served as
give a controlled bilayer. Such a construct could represent internal references. Signal shapes and multiplicities are abbrevi-
attractive carriers for vaccines once the sugar moiety of the ated as br (broad), s (singlet), d (doublet), t (triplet), q (quartet),
glycolipid is substituted by Tn or Tf antigens [42,43] or using quint (quintet) and m (multiplet). Where possible, a scalar cou-
natural glycolipids from Gram-negative bacteria such as lipid A pling constant J is given in hertz (Hz). Optical rotations were
[4]. This work represents the first step towards the formulation measured as CHCl3 solutions on a JASCO P-1010 digital
of an heterogeneous, stable systems (under biological condi- polarimeter and converted to specific rotations [α]D. Micro-
tions) that can be used for several biological purposes including graphs were recorded with an inverted microscope (Zeiss LSM
synthetic vaccines, supramolecular adjuvants, glycolipid–phos- 800) equipped with a 50× objective through the AxioCam soft-
pholipid drug delivery systems and for the formulation of GVs ware. Micrographs were used without any graphical treatment
that can be used as tools to bind to various bacterial lectins and the image size was adjusted respecting the x/y pixel propor-
depending on the mono- or disaccharides used. As a first and tions
relevant conclusion, in silico and in vitro studies demonstrated
that two of those compounds, bearing one or two N-acetyl Synthesis of compounds 3–8
glucosamine moieties are capable to bind strongly to the lectin Compound 3: A stirred solution of 1a (47 mg, 0.24 mmol,
WGA.
3 equiv), 2a (11.2 mg, 0.08 mmol, 1 equiv) and DPAP
(7.71 mg, 0.024 mmol, 0.3 equiv) in MeOH (660 μL) was irra-
diated at room temperature for 30 min and then concentrated.
The residue was purified by silica flash chromatography
Experimental
Materials and methods
Myristoyl chloride was from TCI-Europe (Antwerp, Belgium) (gradient: 100:0–80:20 CHCl3/MeOH v/v). Yield: 77%
and used without further purification. 1-Palmitoyl-2-oleoyl-sn- (20.7 mg, 0.06 mmol); [α]D25 +11.8 (c 0.1, CHCl3); 1H NMR
glycero-3-phosphate (POPC) was from Avanti Polar Lipids (300 MHz, CDCl3) δH 5.35 (d, J = 4.6 Hz, 1H), 4.42 (d, J =
(Alabaster, AL, USA). 2,3,4,6-Tetra-O-acetyl-α- -glucopyra- 9.4 Hz, 2H), 3.83 (s, 2H), 3.56 (dd, J = 7.9 Hz, 2H), 3.44 (s,
nosyl bromide and N-acetyl- -glucosamine were from 1H), 3.36 (dd, J = 8.9 Hz, 2H), 2.67 (t, J = 7.2 Hz, 2H), 1.58
Carbosynth (San Diego, CA, USA). Wheat germ agglutinin (dt, J = 7.4 Hz, 2H), 1.23 (s, 12H), 0.85 (t, J = 6.7 Hz, 3H);
(WGA), horseradish peroxidase-labelled WGA (WGA-HRP), 13C NMR (75 MHz, CDCl3) δC 14.3, 22.9, 29.2, 29.4, 29.6,
bovine serum albumin (BSA) and SIGMA FAST O-phenylene- 29.9, 30.3, 31.1, 61.7, 69.5, 72.9, 78.0, 79.7, 86.4; HRMS (m/z):
diamine dihydrochloride (OPD) were from Sigma-Aldrich. [M + H]+ calcd for C16H32O5S, 336.487; found, 336.493.
Polymeric N-acetyl-β- -glucosamine (PAA-GlcNAc) was from
Lectinity Holding, Inc., Moscow. DPAP (2,2-dimethoxy-2- Compound 4: A stirred solution of 1a, (50 mg, 0.25 mmol,
phenylacetophenone) and all the other reagents were from 3 equiv), 2b (16.4 mg, 21 µL, 0.085 mmol, 1 equiv) and DPAP
Sigma-Aldrich (Merck KGaA, Darmstadt, Germany). Reaction (5.9 mg, 0.025 mmol, 0.3 equiv) in DMF (1 mL) was irradiated
943

Beilstein J. Org. Chem. 2019, 15, 937–946.
at room temperature for 30 min and then concentrated. The CD3OD) δC 13.0, 22.3, 25.5, 28.9, 29.1, 29.2, 29.3, 29.4, 29.5,
residue was purified by silica flash chromatography (gradient: 31.7, 32.3, 33.3, 54.9, 61.6, 70.6, 76.0, 80.8, 84.3, 172.1; [α]D25
100:0–80:20 CHCl3/MeOH v/v). Yield: 55% (20.3 mg, 0.37 (c 0.1, MeOH); HRMS (m/z): [M + Na]+ calcd for
0.047 mmol). Chemical analyses are in agreement with what C33H62NaO12S2, 736.3502; found, 736.3508.
was previously reported [44].
Compound 8: A stirred solution of 1b (142.2 mg, 0.6 mmol), 6
Compound 5: A stirred solution of glycosyl thiol 1b (49.8 mg, (34.7 mg, 0.1 mmol, 1 equiv) and DPAP (15.0 mg, 0.06 mmol,
0.21 mmol, 3 equiv), 2b (13.7 mg, 0.07 mmol, 1 equiv) and 0.6 equiv) in DMF (660 μL) was irradiated at room tempera-
DPAP (5.1 mg, 0.02 mmol, 0.3 equiv) in DMF (660 μL) was ture for 120 min and then concentrated. The residue was puri-
irradiated at room temperature for 30 min and then concen- fied by silica flash chromatography (gradient: 100:0–75:25
trated. The residue was purified by silica flash chromatography CHCl3/MeOH v/v). Yield: 40% (31.8 mg, 0.04 mmol).
(gradient: 100:0–80:20 CHCl3/MeOH v/v). Yield: 85% 1H NMR (300 MHz, CD3OD) δH 5.34 (m, 4H), 5.08 (m, 1H),
(25.8 mg, 0.06 mmol); 1H NMR (400 MHz, MeOD) δH 4.46 (d, 4.34–4.21 (m, 2H), 3.71 (d, J = 4.0 Hz, 2H), 2.35–2.21 (m, 4H),
J = 12.0 Hz, 1H), 3.86 (dd, J = 12.0 Hz, 1H), 3.72 (t, J = 2.00 (d, J = 8.0 Hz, 4H), 1.61 (d, J = 4.0 Hz, 4H), 1.27 (d, J =
12.0 Hz, 1H), 3.70–3.64 (m, 2H), 3.32 (t, J = 12.0 Hz, 1H), 3.22 16.1 Hz, 34H), 0.87 (t, J = 8.0 Hz, 6H); 13C NMR (75 MHz,
(t, J = 12.0 Hz, 1H, partially masked by HOD signal), 2.70 (m, CDCl3) δC 22.7, 24.8, 24.9, 25.6, 27.0, 27.1, 27.2, 29.0–29.2,
2H) 1.97 (s, 3H), 1.61 (m, 2H), 1.29 (s, 14H), 0.90 (t, J = 29.3, 29.4, 29.7, 29.8, 31.9, 34.0, 34.3, 61.5, 62.0, 66.0, 72.1,
8.0 Hz, 3H); 13C NMR (100.0 MHz, MeOD) δ 12.9, 21.6, 22.3, 129.7, 130.0, 173.4, 173.8, 173.9; [α]D25 0.31 (c 0.1, MeOH);
25.6, 28.6, 28.9, 29.1, 29.2, 29.3–29.4, 54.9, 61.4, 70.6, 76.0, HRMS (m/z) calcd for C37H68N2O12S2, 796,4214; found,
80.8, 84.3; [α]D25 +11.8 (c 0.1, CHCl3); HRMS (m/z): [M + H]+ 796,4222.
calcd for C22H44NO5S, 434.2921; found, 434.2935.
Preparation and observation of giant vesicles: Giant vesicles
Compound 6: Hepta-1,6-dien-4-ol [24,25] (500 mg, 4.5 mmol), (GVs) were prepared by the natural swelling method [23].
myristoyl chloride (2.9 mL, 10.7 mmol) and DMAP (1.3 g, Lipids (mixture of commercial POPC and n-alkyl thioglyco-
10.7 mmol) were dissolved in 10 mL of dry CH2Cl2. The solu- sides 5 or 8) were dissolved in methanol (typically, 2 mL) in a
tion was stirred for 16 h, then transferred in a separation funnel 10 mL round-bottom flask. The solvent was completely evapo-
and washed with saturated NaHCO3 (2 × 50 mL) and brine rated under reduced pressure using a rotatory evaporator. The
(50 mL), dried over Na2SO4 and concentrated. The crude mix- resulting thin lipid film was further dried for 180 minutes at
ture was purified on a column of silica (isocratic, CH2Cl2) from 1 mbar/25 °C, and then hydrated for 16 hours – without shaking
which 3 was obtained as a transparent oil. Yield: 78% (1.12 g, – with the aqueous buffer, termed “I-solution” (composed of
3.50 mmol). 1H NMR (300 MHz, CDCl3) δH 5.80–5.66 (m, 200 mM sucrose in 50 mM of PBS buffer, pH 7.4) to obtain an
2H), 5.08 (d, J = 7.4 Hz, 2H), 5.03 (br s, 2H), 4.95 (dd, J = overall 1–2 mM lipid concentration. The hydration temperature
12.4, 5.9 Hz, 1H), 2.34–2.30 (m, 4H), 2.23 (t, J = 8.3 Hz, 2H), was 25 °C. Three volumes of the thus obtained GVs were
1.63–1.57 (m, 2H), 1.25 (br s, 20H), 0.87 (t, J = 6.7 Hz, 3H); diluted with one volume of an aqueous isotonic buffer solution
13C NMR (75 MHz, CDCl3) δC 14.1, 22.6, 25.0, 29.1, 29.2, termed “O-solution” (composed of 200 mM glucose in 50 mM
29.3, 29.4, 29.5–29.7, 31.9, 34.5, 38.0, 71.9, 117.7, 133.6, of PBS buffer, pH 7.4) and centrifuged at 5,000 rpm for
173.3; HRMS (m/z): [M + H]+ calcd for C21H38O2, 345.2769; 10 minutes in a bench-top Eppendorf mini-centrifuge. GVs
found, 345.2764.
were pelleted down in the Eppendorf tube due to the density
difference between the I-solution and the O-solution. The super-
Compound 7: A stirred solution of 1a (117.6 mg, 0.6 mmol, natant was carefully removed, and the pellet was re-suspended
6 equiv (3 × site), 6 (34.6 mg, 0.1 mmol, 1 equiv) and DPAP in 100 µL of fresh O-solution. Each sample was stained with a
(15.0 mg, 0.06 mmol, 0.6 equiv) in DMF (660 μL) was irradi- Nile Red® solution (1 mM in DMSO, 1 µL) before micro-
ated at room temperature for 60 min and then concentrated. The scopic observation.
residue was purified by silica flash chromatography (gradient:
100:0–80:20 CHCl3/MeOH v/v). Yield: 75% (51.5 mg, Enzyme-linked lectin assays: 96-well microtiter Nunc-
0.07 mmol). 1H NMR (400 MHz, CD3OD) δH (4.46 (d, J = Immuno plates (Maxi-Sorp) were coated with PAA-GlcNAc
10.3 Hz, 2H), 3.87 (dd, J = 12.0, 2.2 Hz, 2H), 3.73 (dd, J = (100 μL per well, diluted from a stock solution of 5 μg mL−1 in
10.3, 6.2 Hz, 2H) 3.71–3.66 (m, 2H), 3.65–3.63 (m, 2H), 3.54 50 mM carbonate buffer pH 9.6) for 1 h at 37 °C. The wells
(t, J = 6.7 Hz, 2H), 3.45–3.40 (m, 2H), 3.29–3.25 (m, 4H), were then washed with T-PBS (3 × 100 μL per well, PBS pH
2.78–2.61 (m, 2H), 2.04–1.55 (m, 2H) 1.66–1.48 (m, 2H), 1.27 7.4 containing 0.05% (v/v) Tween 20). This washing procedure
(br s, 10H), 0.87 (t, J = 6.0 Hz, 3H); 13C NMR (100 MHz, was repeated after each incubation step. The coated microtiter
944

Beilstein J. Org. Chem. 2019, 15, 937–946.
plates were then blocked with BSA in PBS (3% w/v, 1 h at The University of Lyon, “Claude Bernard” Lyon 1 and all the
37 °C, 100 μL per well). Serial two-fold dilutions of each inhib- members of ICBMS/LCO2-glyco and the support from the
itor were pre-incubated 1 h at 37 °C in PBS-DMSO (9:1 v/v, BQR grant (2014) are gratefully acknowledged as well. This
60 μL per well) in the presence of WGA-HRP (60 μL) at the work is dedicated to the memory of Océane Fiore (2015–2017).
desired concentration. The above solutions (100 μL) were then
ORCID® iDs
Dimitri Fayolle - https://orcid.org/0000-0002-1543-136X
transferred to the blocked microtiter plates which were incubat-
ed for 1 h at 37 °C. After incubation, the plates were washed
Nathalie Berthet - https://orcid.org/0000-0002-3479-1918
with T-PBS (3 × 100 μL per well) and then the color was de-
Bastien Doumeche - https://orcid.org/0000-0003-3458-1029
veloped using OPD (100 μL per well, 0.4 mg mL−1 in 0.05 M
Olivier Renaudet - https://orcid.org/0000-0003-4963-3848
phosphate-citrate buffer) and urea hydrogen peroxide
Peter Strazewski - https://orcid.org/0000-0001-8540-3279
(0.4 mg mL−1). The reaction was stopped after 10 min by
Michele Fiore - https://orcid.org/0000-0001-8176-9249
adding H2SO4 (30% v/v, 50 μL per well) and the absorbance
was measured at 490 nm. The percentage of inhibition was de-
References
1. Holst, J.; Martin, D.; Arnold, R.; Huergo, C. C.; Oster, P.;
termined using the following equation, where A is absorbance.
% Inhibition = [(A(no inhibitor) – A(with inhibitor)/A(no inhib-
O’Hallahan, J.; Rosenqvist, E. Vaccine 2009, 27 (Suppl. 2), B3–B12.
itor)] × 100. The percent of inhibition was plotted against the
doi:10.1016/j.vaccine.2009.04.071
logarithm of the concentration of the sugar derivatives. The
2. Feiring, B.; Fuglesang, J.; Oster, P.; Næss, L. M.; Helland, O. S.;
sigmoidal curves were fitted and the concentration at 50% inhi-
Tilman, S.; Rosenqvist, E.; Bergsaker, M. A. R.; Nøkleby, H.;
bition of WGA-HRP binding to PAA-GlcNAc coated plates
Aaberge, I. S. Clin. Vaccine Immunol. 2006, 13, 790–796.
was determined (IC50).
doi:10.1128/cvi.00047-06
3. Jackson, C.; Lennon, D. R.; Sotutu, V. T. K.; Yan, J.; Stewart, J. M.;
Reid, S.; Crengle, S.; Oster, P.; Ypma, E.; Aaberge, I.; Mulholland, K.;
Martin, D. R. Arch. Dis. Child. 2009, 94, 745–751.
doi:10.1136/adc.2007.132571
Docking experiments: WGA structure was obtained from the
Protein Data Bank (2UVO) and unbound ligands (water,
GlcNAc, glycerol and pyroglutamic acid) were removed. PDB
files for compounds 4C2, 4C4, 5C2, 5C4, 7C2, 7C4, 8C2 and
8C4 were prepared using OpenBabel (version 2.4.1). Docking
was performed using autodock vina implemented in the Samson
software (version beta 0.7.0, NANO-D Inria, https://
www.samson-connect.net). A grid (60 × 60 × 60 Å) including
the whole protein was used and 20 poses leading to the lowest
energies were kept as significant. Pictures were prepared using
PyMOL (version 0.99rc6, DeLano Scientific, San Carlos, CA,
700).
4. Molinaro, A.; Holst, O.; Di Lorenzo, F.; Callaghan, M.; Nurisso, A.;
D'Errico, G.; Zamyatina, A.; Peri, F.; Berisio, R.; Jerala, R.;
Jiménez-Barbero, J.; Silipo, A.; Martín-Santamaría, S. Chem. – Eur. J.
2015, 21, 500–519. doi:10.1002/chem.201403923
5. Global Vaccine Action Plan 2011-2020, World Health Organization:
Geneva, 2012.
6. Nativi, C.; Renaudet, O. ACS Med. Chem. Lett. 2014, 5, 1176–1178.
doi:10.1021/ml5003794
7. Collins, B. S. Discovery Med. 2011, 12, 7.
8. Robyt, J. F. Essentials of Carbohydrate chemistry; Springer Science &
Business media, 2012.
9. Taylor, M. E.; Drickamer, K. Introduction to glycobiology; Oxford
University Press: New York, USA, 2006.
10.El-Sukkary, M. M. A.; Syed, N. A.; Aiad, I.; Helmy, S. M.;
El-Azab, W. I. M. Tenside, Surfactants, Deterg. 2009, 46, 311–316.
doi:10.3139/113.110037
Supporting Information
Supporting Information File 1
Additional Figures S1–S6 and spectra of synthesized
glycolipids and starting materials.
11.Matsson, M. K.; Kronberg, B.; Claesson, P. M. Langmuir 2004, 20,
4051–4058. doi:10.1021/la035959p
12.Marino, C.; Mariño, K.; Miletti, L.; Manso Alves, M. J.; Colli, W.;
de Lederkremer, R. M. Glycobiology 1998, 8, 901–904.
doi:10.1093/glycob/8.9.901
[https://www.beilstein-journals.org/bjoc/content/
supplementary/1860-5397-15-90-S1.pdf]
13.Davis, C. B.; Hartnell, R. D.; Madge, P. D.; Owen, D. J.;
Thomson, R. J.; Chong, A. K. J.; Coppel, R. L.; von Itzstein, M.
Carbohydr. Res. 2007, 342, 1773–1780.
Acknowledgements
doi:10.1016/j.carres.2007.04.027
Michele Fiore warmly thanks Dr. Sébastian Vidal (ICBMS,
LCO2-Glyco) that encouraged exploring in this research field.
Dr. Beatrice Pelotier is gratefully acknowledged for the analyti-
cal sample of the compound hepta-1,6-dien-4-ol given. The
students of the class M1 “Fonctionnalisation en synthèse
organique (CHM1182M)” are gratefully acknowledged for their
support in the large-scale preparation of hepta-1,6-dien-4-ol.
14.Samoshin, A. V.; Dotsenko, I. A.; Samoshina, N. M.; Franz, A. H.;
Samoshin, V. V. Int. J. Carbohydr. Chem. 2014, No. 941059.
doi:10.1155/2014/941059
15.van Doren, H. A.; van der Geest, R.; Kellogg, R. M.; Wynberg, H.
Carbohydr. Res. 1989, 194, 71–77.
doi:10.1016/0008-6215(89)85007-4
945

Beilstein J. Org. Chem. 2019, 15, 937–946.
16.van Doren, H. A.; van der Geest, R.; van Bolhuis, F.; Kellogg, R. M.;
Wynberg, H. Carbohydr. Res. 1989, 194, 79–86.
38.Fiore, M.; Berthet, N.; Renaudet, O.; Barbier, V. Med. Chem. Commun.
2014, 5, 1202–1207. doi:10.1039/c4md00097h
doi:10.1016/0008-6215(89)85008-6
39.Wright, C. S.; Kellogg, G. E. Protein Sci. 1996, 5, 1466–1476.
doi:10.1002/pro.5560050803
17.Kumar, V.; Taxak, N.; Jangir, R.; Bharatam, P. V.; Kartha, K. P. R.
J. Org. Chem. 2014, 79, 3427–3439. doi:10.1021/jo5001753
18.Dondoni, A. Angew. Chem., Int. Ed. 2008, 47, 8995–8997.
doi:10.1002/anie.200802516
40.Schwefel, D.; Maierhofer, C.; Beck, J. G.; Seeberger, S.;
Diederichs, K.; Möller, H. M.; Welte, W.; Wittmann, V.
J. Am. Chem. Soc. 2010, 132, 8704–8719. doi:10.1021/ja101646k
41.This value is comparable with distances previously observed in WGA
crystals soaked in divalent and tetravalent GlcNAc ligands.
42.Fiore, M.; Thomas, B.; Duléry, V.; Dumy, P.; Renaudet, O.
New J. Chem. 2013, 37, 286–289. doi:10.1039/c2nj40972k
43.Richichi, B.; Thomas, B.; Fiore, M.; Bosco, R.; Qureshi, H.; Nativi, C.;
Renaudet, O.; BenMohamed, L. Angew. Chem., Int. Ed. 2014, 53,
11917–11920. doi:10.1002/anie.201406897
19.Dondoni, A.; Marra, A. Chem. Soc. Rev. 2012, 41, 573–586.
doi:10.1039/c1cs15157f
20.McSweeney, L.; Dénès, F.; Scanlan, E. M. Eur. J. Org. Chem. 2016,
2080–2095. doi:10.1002/ejoc.201501543
21.Zhou, C. Y.; Wu, H.; Devaraj, N. K. Chem. Sci. 2015, 6, 4365–4372.
doi:10.1039/c5sc00653h
22.Goyard, D.; Shiao, T. C.; Fraleigh, N. L.; Vu, H.-Y.; Lee, H.;
Diaz-Mitoma, F.; Le, H.-T.; Roy, R. J. Mater. Chem. B 2016, 4,
4227–4233. doi:10.1039/c6tb00344c
44.Saito, S.; Furumoto, T.; Ochiai, M.; Hosono, A.; Hoshino, H.;
Haraguchi, U.; Ikeda, R.; Shimada, N. Eur. J. Med. Chem. 1996, 31,
365–381. doi:10.1016/0223-5234(96)89163-x
23.Fiore, M.; Marra, A.; Dondoni, A. J. Org. Chem. 2009, 74, 4422–4425.
doi:10.1021/jo900514w
24.Dallanegra, R.; Pilgrim, B. S.; Chaplin, A. B.; Donohoe, T. J.;
Weller, A. S. Dalton Trans. 2011, 40, 6626. doi:10.1039/c1dt10503e
25.Glachet, T.; Fache, F.; Pelotier, B.; Piva, O. Synthesis 2017, 49,
5197–5202. doi:10.1055/s-0036-1588521
License and Terms
This is an Open Access article under the terms of the
Creative Commons Attribution License
26.Roodsari, F. S.; Wu, D.; Pum, G. S.; Hajdu, J. J. Org. Chem. 1999, 64,
7727–7737. doi:10.1021/jo990414e
27.Reactions performed in MeOH or MeOD and DMF or DMF-d7 gave
similar and reproducible yields. Although the conversion observed by
1H NMR was always 100%, with complete consumption of the starting
alkene, we encountered difficulties in the flash chromatography
purification step due to the amphiphilic character of the molecule and
its high affinity with the silica gel.
(http://creativecommons.org/licenses/by/4.0). Please note
that the reuse, redistribution and reproduction in particular
requires that the authors and source are credited.
The license is subject to the Beilstein Journal of Organic
Chemistry terms and conditions:
28.Fiore, M.; Madanamoothoo, W.; Berlioz-Barbier, A.; Maniti, O.;
Girard-Egrot, A.; Buchet, R.; Strazewski, P. Org. Biomol. Chem. 2017,
15, 4231–4240. doi:10.1039/c7ob00708f
(https://www.beilstein-journals.org/bjoc)
29.Fayolle, D.; Altamura, E.; D’Onofrio, A.; Madanamothoo, W.; Fenet, B.;
Mavelli, F.; Buchet, R.; Stano, P.; Fiore, M.; Strazewski, P. Sci. Rep.
2017, 7, 18106. doi:10.1038/s41598-017-18053-y
The definitive version of this article is the electronic one
which can be found at:
30.Fayolle, D.; Fiore, M.; Stano, P.; Strazewski, P. PLoS One 2018, 13,
e0192975. doi:10.1371/journal.pone.0192975
doi:10.3762/bjoc.15.90
31.Wright, C. S. J. Biol. Chem. 1992, 267, 14345.
32.Wright, C. S. J. Mol. Biol. 1987, 194, 501–529.
doi:10.1016/0022-2836(87)90678-4
33.Cecioni, S.; Imberty, A.; Vidal, S. Chem. Rev. 2015, 115, 525–561.
doi:10.1021/cr500303t
34.Bernardi, A.; Jiménez-Barbero, J.; Casnati, A.; De Castro, C.;
Darbre, T.; Fieschi, F.; Finne, J.; Funken, H.; Jaeger, K.-E.;
Lahmann, M.; Lindhorst, T. K.; Marradi, M.; Messner, P.; Molinaro, A.;
Murphy, P. V.; Nativi, C.; Oscarson, S.; Penadés, S.; Peri, F.;
Pieters, R. J.; Renaudet, O.; Reymond, J.-L.; Richichi, B.; Rojo, J.;
Sansone, F.; Schäffer, C.; Turnbull, W. B.; Velasco-Torrijos, T.;
Vidal, S.; Vincent, S.; Wennekes, T.; Zuilhof, H.; Imberty, A.
Chem. Soc. Rev. 2013, 42, 4709–4727. doi:10.1039/c2cs35408j
35.Lo Conte, M.; Staderini, S.; Chambery, A.; Berthet, N.; Dumy, P.;
Renaudet, O.; Marra, A.; Dondoni, A. Org. Biomol. Chem. 2012, 10,
3269. doi:10.1039/c2ob07078b
36.Fiore, M.; Berthet, N.; Marra, A.; Gillon, E.; Dumy, P.; Dondoni, A.;
Imberty, A.; Renaudet, O. Org. Biomol. Chem. 2013, 11, 7113.
doi:10.1039/c3ob41203b
37.Fiore, M.; Daskhan, G. C.; Thomas, B.; Renaudet, O.
Beilstein J. Org. Chem. 2014, 10, 1557–1563. doi:10.3762/bjoc.10.160
946