Oleanane-type glycosides from the roots of Weigela florida “rumba” and evaluation of their antibody recognition

Article abstract of DOI:10.1016/j.fitote.2018.04.017

Three triterpene glycosides were isolated from the roots of Weigela florida “rumba” (Bunge) A. DC.: two previously undescribed 3-O-β-D-xylopyranosyl-(1→2)-[β-D-xylopyranosyl-(1→4)]-β-D-xylopyranosyl-(1→4)-β-D-xylopyranosyl-(1→3)-α-L-rhamnopyranosyl-(1→2)-

Full text of DOI:10.1016/j.fitote.2018.04.017

Fitoterapia 128 (2018) 198–203
Contents lists available at ScienceDirect
Fitoterapia
journal homepage: www.elsevier.com/locate/fitote
Oleanane-type glycosides from the roots of Weigela florida “rumba” and
evaluation of their antibody recognition
Anne-Sophie Champy-Tixiera,b,c, Anne-Claire Mitaine-Offer , Feliciana Real Fernándezb,d,
Tomofumi Miyamoto , Chiaki Tanaka , Anna-Maria Papini , Marie-Aleth Lacaille-Dubois
a
e
e
b,c,f
a,⁎
a
PEPITE EA 4267, Laboratoire de Pharmacognosie, UFR Sciences de Santé, Université de Bourgogne Franche-Comté, BP 87900, 21079 Dijon Cedex, France
Interdepartmental Laboratory of Peptide and Protein Chemistry and Biology, University of Florence, 50019 Sesto Fiorentino, Italy
Department of Chemistry “Ugo Schiff”, University of Florence, 50019 Sesto Fiorentino, Italy
Department of Neurosciences, Psychology, Drug Research and Child Health (NeuroFarBa) - Section of Pharmaceutical Sciences and Nutraceutics, University of Florence,
0019 Sesto Fiorentino, Italy
b
c
d
5
e
Graduate School of Pharmaceutical Sciences, Kyushu University, Fukuoka, 812-8582, Japan
Laboratory of Chemical Biology EA 4505 & PeptLab@UCP, University of Cergy-Pontoise, 95031 Cergy-Pontoise Cedex, France
f
A R T I C L E I N F O
A B S T R A C T
Keywords:
Three triterpene glycosides were isolated from the roots of Weigela florida “rumba” (Bunge) A. DC.: two pre-
viously undescribed 3-O-β-D-xylopyranosyl-(1→2)-[β-D-xylopyranosyl-(1→4)]-β-D-xylopyranosyl-(1→4)-β-D-xy-
lopyranosyl-(1→3)-α-L-rhamnopyranosyl-(1→2)-α-L-arabinopyranosyloleanolic acid (1) and 3-O-β-D-xylopyr-
anosyl-(1→2)-[β-D-glucopyranosyl-(1→4)]-β-D-xylopyranosyl-(1→4)-β-D-xylopyranosyl-(1→3)-α-L-rhamnopyr-
anosyl-(1→2)-α-L-arabinopyranosyloleanolic acid (2), and one isolated for the first time from a natural source 3-
O-β-D-xylopyranosyl-(1→3)-α-L-rhamnopyranosyl-(1→2)-α-L-arabinopyranosyloleanolic acid (3). Their struc-
tures were elucidated mainly by 2D NMR spectroscopic analysis (COSY, TOCSY, NOESY, HSQC, HMBC) and
mass spectrometry. Compounds 2 and 3 were further evaluated as antigens in enzyme-linked immunosorbent
assay (ELISA) to recognize IgM antibodies in multiple sclerosis (MS) patients' sera.
Weigela florida “rumba”
Triterpene glycosides
NMR
ELISA
Multiple sclerosis
1. Introduction
glycosides, especially the genus Weigela. Saponins from three species,
Weigela hortensis [7], W. subsessilis [8], and W. stelzneri [9] were already
In Multiple Sclerosis (MS), the disease pathogenesis is still unclear
described in literature. The phytochemical study of a horticultural
cultivar of Weigela florida (Bunge) A. DC. was thus realized. Herein we
report the isolation, purification, and characterization of natural gly-
cosides from Weigela florida “rumba”. The further immunochemical
evaluation of two isolated compounds was performed by enzyme linked
immunosorbent assay (ELISA) and results were compared with the
CSF114(Glc)-based ELISA.
and the main hypothesis includes the role of viral, immunological, and
vascular factors [1]. Pathophysiological post-translational modifica-
tions of native antigens, mainly myelin glycoproteins, can trigger the
immune response generating antibodies [2]. In fact, autoantibodies
against aberrant glycosylated forms are proposed as biomarkers,
which are helpful in MS diagnosis [3–4]. The glucosylated peptide
CSF114(Glc) is able to evaluate MS-related autoAbs associated with
clinical assessment of MS activity and brain lesions MRI profile [3]. In
any case, despite the role of glycosylation has been previously in-
vestigated in MS pathogenesis [5], the nature of the native glycosylated
antigen has not been yet clarified. The interaction between glycosides,
as possible native antigens, and autoantibodies was previously eval-
uated in ELISA [6]. Natural triterpene glycosides showed good capacity
to recognize IgMs and can be good candidates as native glycoconju-
gates. A deep study of anti-saponin antibodies in MS can be helpful to
understand the role of carbohydrate structures in antibody recognition.
In this context, the Caprifoliaceae family is a rich source of triterpene
2. Experimental
2.1. General experimental procedures
Optical rotation values were recorded on an AA-10R automatic polari-
meter. NMR spectra were performed using a Varian INOVA 600 (Agilent
Technologies) at the operating frequency of 600 MHz. The operating condi-
1
tions were as follows: H: frequency, 600 MHz; sweep width, 8 kHz; sampling
point, 66 k; spectral width, 7804 Hz, accumulation, 32 pulses; temperature,
304 K. 13C: frequency, 150 MHz; sweep width, 32 kHz; sampling point, 160 k;
⁎
Corresponding author.
E-mail address: m-a.lacaille-dubois@u-bourgogne.fr (M.-A. Lacaille-Dubois).
https://doi.org/10.1016/j.fitote.2018.04.017
Received 12 February 2018; Received in revised form 22 April 2018; Accepted 27 April 2018
Available online 30 April 2018
0367-326X/ © 2018 Elsevier B.V. All rights reserved.

A.-S. Champy-Tixier et al.
Fitoterapia 128 (2018) 198–203
Table 1
1
3C NMR and 1H NMR data of the aglycone of compounds 1–3 in Pyridine-d
(δ in ppm, J in Hz).
5
Position
1
2
3
δ
C
δ
H
δ
C
δ
H
δ
C
δ
H
1
2
3
4
5
6
7
8
9
38.6
26.3
88.5
39.2
55.7
18.2,
32.8,
39.4
47.7
36.6
23.5
122.2
144.5
41.8
28.0
23.3
46.4
41.7
46.2
30.6
33.9
32.8
27.9
16.8
15.2
17.1
25.9
180.2
33.0
23.5
0.88 m, 1.43 m
1.78 m, 2.02 m
3.23
–
0.73 br d (12.2)
1.23, 1.47 m
1.22, 1.43 m
–
1.57 dd (14.1, 8.1)
–
1.83 m, 1.91
5.42
–
–
1.14 m, 2.08 m
1.91, 2.06 m
–
3.23
1.22, 1.73
–
1.11 m, 1.35 m
1.70, 1.89
1.21 s
1.03 s
0.76 s
0.90 s
1.23 s
–
0.89 s
0.93 s
38.5
26.2
88.5
39.2
55.6
18.2
32.9
39.3
47.7
36.7
23.4
122.2
144.4
41.8
27.8
23.3
46.4
41.6
46.1
30.6
nd
32.8
27.8
16.7
15.2
17.0
25.8
180.3
33.8
23.4
0.84 m, 1.42 m
1.74 m, 2.00 m
3.22
–
0.71 br d (11.6)
1.20, 1.45 m
1.19, 1.40 m
–
1.55 dd (15.0, 9.0)
–
1.82 m, 1.86
5.40 br t (3.0)
–
–
1.13 m, 2.03 m
1.89 m, 2.05 m
–
3.20
1.20, 1.72
–
38.6
23.3
88.5
39.4
55.7
18.1
32.8
39.2
47.7
36.7
23.5
122.2
144.5
41.8
28.0
26.3
46.4
41.6
46.2
30.6
33.9
32.8
27.9
16.8
15.2
17.1
25.9
180.2
33.0
23.5
0.88 m, 1.42 m
1.88 m, 2.04
3.22
–
0.72 br d (12.2)
1.45, nd
1.17, 1.38 m
–
1.57 dd (14.0, 8.6)
–
1.82 m, 1.86
5.44 br t (3.1)
–
–
1.12 m, 2.06 m
1.75 m, 2.01 m
–
3.20
1.19, 1.74
–
1.12 m, 1.37 m
1.71, 1.78
1.22 s
1.03 s
0.76 s
0.90 s
1.24 s
–
0.88 s
0.94 s
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
nd
1.73, nd
1.19 s
1.01 s
0.75 s
0.89 s
1.22 s
–
0.88 s
0.92 s
Overlapped proton signals are reported without designated multiplicity, nd = not determined.
spectral width, 30,000 Hz, accumulation, 8000 pulses; temperature, 304 K.
Each sample was dissolved in C N (200 μL) using 5 mm micro-sample tube
SHIGEMI Co., Ltd., Japan). Chemical shifts were referenced to solvent signal
7.22, δ 123.87). Conventional pulse sequences were used for gMQF-
of Pharmacognosy, Université de Bourgogne Franche-Comté, Dijon,
France.
5 5
D
(
(δ
H
C
2.3. Extraction and isolation
COSY, TOCSY, NOESY, gHSQC, and gHMBC. The mixing time in the NOESY
experiment was set to 500 ms. TOCSY spectra were acquired using the
standard MLEV17 spin-locking sequence and 60 ms mixing time. TOCSY,
NOESY and HSQC spectra were recorded using phase-sensitive mode. The
size of the acquisition data matrix was 2048 × 256 words in f2 and f1, re-
spectively, and zero filling up to 2 k in f1 was made prior to Fourier trans-
formation. Sine-bell or Shifted sine-bell window functions, with the corre-
sponding shift optimized for every spectrum, were used for resolution
enhancement and baseline correction was applied in both dimensions. HR-
ESIMS (positive-ion mode) was carried out on a Bruker micrOTOF mass
spectrometer. A R.E.U.S ultrasonic apparatus was used for the extraction.
Isolations of compounds were carried out using vacuum liquid chromato-
graphy (VLC) with reversed-phase RP-18 silica gel (75–200 μm, Silicycle).
Medium-pressure liquid chromatography (MPLC) was performed on silica gel
The dried, powdered roots of W. florida “rumba” (66.4 g) were
submitted to an ultrasound assisted extraction three times with EtOH/
O (7:3, 3 × 1 L) for 1 h. After the solvent evaporation under vacuum,
the resulting extract (2.9 g) was submitted to VLC (RP-18 silica gel,
O, MeOH/H O 50:50 and MeOH). The eluted fraction with MeOH
330 mg) was fractionated by MPLC on silica gel 60 (CHCl /MeOH/H
0:30:5). Ten fractions (F1-F10) with two pure compounds (1, 6.1 mg
and 2, 11.1 mg) were obtained. Then, the fraction F2 (37.7 mg) was
fractionated again by MPLC on silica gel 60 (CHCl /MeOH/H
0:30:5) to give compound 3 (20.1 mg).
H
2
H
(
7
2
2
3
2
O
3
2
O
7
2.4. 3-O-β-D-xylopyranosyl-(1→2)-[β-D-xylopyranosyl-(1→4)]-β-D-
xylopyranosyl-(1→4)-β-D-xylopyranosyl-(1→3)-α-L-rhamnopyranosyl-
(1→2)-α-L-arabinopyranosyloleanolic acid (1)
6
0 (Merck, 15–40 μm) with a Gilson M 305 pump (25 SC head pump, M 805
manometric module), a Büchi glass column (460 mm × 25 mm and
60 mm × 15 mm), and a Büchi precolumn (110 mm × 15 mm). Thin-layer
4
White amorphous powder; [α]D – 28° (c 0.75, MeOH); H and 13C
2
5
1
chromatography (TLC, Silicycle) and high-performance thin-layer chroma-
tography (HPTLC, Merck) were carried out on precoated silica gel plates
NMR data (600 MHz and 150 MHz, pyridine-d ), Tables 1 and 2; HR-ESIMS
5
m/z 1285.5334 [M + Na]+ (Calcd for C H O Na, 1285.6193).
6
0F254, solvent system CHCl
The spray reagent for saponins was vanillin reagent (1% vanillin in EtOH/
SO , 50:1).
3 2
/MeOH/H O/AcOH (70:30:5:1 and 60:32:7:1).
61 98 27
H
2
4
2
.5. 3-O-β-D-xylopyranosyl-(1→2)-[β-D-glucopyranosyl-(1→4)]-β-D-
xylopyranosyl-(1→4)-β-D-xylopyranosyl-(1→3)-α-L-rhamnopyranosyl-
2.2. Plant material
(1→2)-α-L-arabinopyranosyloleanolic acid (2)
25
– 24° (c 0.75, MeOH); H and 13C
), Tables 1 and 2; HR-ESIMS
28Na, 1315.6299).
1
Weigela florida “rumba” was provided in 2012 from Jardiland ®
Chenôve, France). The plant was properly identified with a voucher
specimen (N° 20121101) deposited in the herbarium of the Laboratory
White amorphous powder; [α]
NMR data (600 MHz and 150 MHz, pyridine-d
m/z 1315.5336 [M + Na]+ (Calcd for C62
O
100
D
(
5
H
199

A.-S. Champy-Tixier et al.
Fitoterapia 128 (2018) 198–203
Table 2
1
3C NMR and 1H NMR data of the sugar moieties of compounds 1–3 in Pyridine-d
(δ in ppm, J in Hz).
5
No
1
2
3
δ
C
δ
H
δ
C
δ
H
δ
C
δ
H
Ara-1
2
3
4
104.7
75.2
73.7
68.6
64.8
101.1
71.4
82.0
72.2
69.4
18.1
106.3
74.3
75.3
76.6
64.4
102.4
73.3
71.7
81.2
64.2
102.1
72.2
76.1
70.1
66.1
104.6
74.3
76.6
70.1
66.1
4.80 d (6.2)
4.46 dd (6.7, 6.2)
4.21
4.24 m
3.77 br d (10.7), 4.26
6.06 br s
104.6
75.1
73.5
68.5
64.7
101.0
71.3
82.0
72.2
69.4
18.0
106.1
74.7
75.2
76.6
64.1
102.3
73.1
72.2
82.0
64.1
101.8
72.2
75.7
69.9
65.9
4.79 d (6.2)
4.45 dd (6.9, 6.2)
4.20
4.24 m
3.76 br d (9.8), 4.27
6.03 br s
4.79 br s
4.59 dd (9.3, 2.7)
4.37
4.51 dq (9.5, 6.0)
1.49 d (6.0)
5.16 d (7.4)
3.99
104.7
75.1
73.7
68.6
64.9
101.0
71.4
82.0
72.3
69.4
18.1
106.6
75.0
77.6
70.5
66.8
4.79 d (6.0)
4.47 dd (6.9, 6.0)
4.19
4.23 m
3.75 br d (10.7), 4.25
6.08 br s
4.80 br s
4.63 dd (9.5, 2.9)
4.39 t (9.5)
4.53 dq (9.5, 6.2)
1.48 d (6.2)
5.22 d (7.6)
5
Rha-1
2
3
4
5
6
Xyl I-1
2
3
4
4.80 br s
4.62 dd (9.5, 2.3)
4.39 dd (9.5, 9.3)
4.53 dq (9.3, 6.0)
1.49 d (6.0)
5.19 d (7.1)
4.01 (dd, 7.8, 7.1)
4.07
4.01 dd (8.6, 7.6)
4.08 dd (8.6, 8.3)
4.10 m
4.05
4.07
4.10
5
3.52 t (10.7), 4.28
4.78 d (7.8)
3.93 dd (8.1, 7.8)
4.12
3.50 t (10.0), 4.28
4.76 d (7.4)
3.92 dd (8.6, 7.4)
4.14
3.58 t (10.5), 4.20
Xyl II-1
2
3
4
4.20
4.23
5
3.58, 4.34
4.86 d (6.9)
3.96 dd (7.4, 6.9)
4.09
3.56, 4.34
4.87 d (6.7)
3.98
4.06
4.07
Xyl III-1
2
3
4
4.07
5
3.60, 4.30
5.36 d (6.9)
4.02
4.08
4.07
3.58, 4.30
Xyl IV-1
2
3
4
5
3.60, 4.34
Glc I-1
104.5
74.7
77.6
70.8
78.1
62.0
5.27 d (7.9)
3.98
4.11
4.01
3.84 m
4.14, 4.37
2
3
4
5
6
Overlapped proton signals are reported without designated multiplicity.
2
.6. 3-O-β-D-xylopyranosyl-(1→3)-α-L-rhamnopyranosyl-(1→2)-α-L-
2.8. Peptides and ursolic acid
arabinopyranosyloleanolic acid (3)
The N-glucosylated peptide CSF114(Glc) was prepared by a micro-
wave-assisted solid-phase peptide synthesis as described elsewhere
[11]. Peptide was purified by solid-phase extraction and reversed phase
high pressure liquid chromatography (HPLC) and further characterized
by mass spectrometry and analytical HPLC. Ursolic acid ≥90% was
purchased to Sigma Aldrich.
2
5
– 15° (c 0.75, MeOH); H and 13C
), Tables 1 and 2; HR-
15Na, 889.4925).
1
White amorphous powder; [α]
D
NMR data (600 MHz and 150 MHz, pyridine-d
ESIMS m/z 889.4410 [M + Na] (Calcd for C46H O
74
5
+
2.7. Acid hydrolysis and GC analysis
2
.9. Immunoenzymatic assay of CSF114(Glc)
Each compound (3 mg) was hydrolyzed with 2 N aq. CF
5 mL) for 3 h at 95 °C. After extraction with CH Cl (3 × 5 mL), the
aqueous layer was repeatedly evaporated to dryness with MeOH until
neutral, and then analyzed by TLC over silica gel (CHCl /MeOH/H
:5:1) by comparison with authentic samples. Furthermore, the sugars
residue was dissolved in anhydrous pyridine (100 μL), and L-cysteine
methyl ester hydrochloride (0.06 mol/L) was added. The mixture was
stirred at 60 °C for 1 h, then 150 μL of HMDS-TMCS (hexamethyldisi-
lazane/trimethylchlorosilane 3:1) was added, and the mixture was
stirred at 60 °C for another 30 min. The precipitate was centrifuged off,
3
COOH
(
2
2
Sera from 19 Multiple Sclerosis patients collected at diagnosis and
10 healthy blood donors were obtained for diagnostic purposes and
stored at −20 °C until use. Multiple Sclerosis (MS) patient sera samples
were collected in two Multiple Sclerosis Centers: The Multiple Sclerosis
Clinical Care and Research Centre, Department of Neurosciences,
Reproductive Sciences and Odontostomatology, Federico II University
(Naples, Italy) and the Azienda Ospedaliera Universitaria Careggi,
Clinica Neurologica, University of Florence (Florence, Italy). All pa-
tients and healthy donors gave their informed consent. Multiple
Sclerosis patients were previously diagnosed after a lumbar puncture,
MRI examination, and cerebrospinal analysis. Samples were pre-se-
lected depending on their reactivity to CSF114(Glc). Antibody re-
sponses were determined in SP-ELISA.
3
2
O
8
and the supernatant was concentrated under a N
2
stream. The residue
was partitioned between n-hexane and H O (0.1 mL each), and the
2
hexane layer (1 μL) was analyzed by GC [10]. The absolute configura-
tions were determined by comparing the retention times with thiazo-
lidine derivatives prepared in a similar way from standard sugars
The glycopeptide CSF114(Glc) was coated on 96-well activated
polystyrene ELISA plates (NUNC Maxisorb) using a solution 10 μL/mL
(
Sigma–Aldrich).
200

A.-S. Champy-Tixier et al.
Fitoterapia 128 (2018) 198–203
in 10% FBS in 150 mM NaCl were incubated at +4 °C overnight
(
100 μL/well). After three washes, 100 μL of alkaline phosphatase
conjugated anti-human IgM (Invitrogen) (diluted 1:1500 in 10% FBS
washing buffer) were added to each well. After 3 h incubation at room
−
1
R
temperature and three washes, 100 μL/well of 1 mg.mL
trophenyl phosphate (Sigma–Aldrich) in carbonate buffer (pH 9.6) with
10 mM MgCl solution was added. After 30 min, the reaction was
p-ni-
1
2
3
S1
S2
S3
C8 OOH
2
2
stopped with 1 M NaOH (50 μL/well), and the absorbance was read at
405 nm. ELISA plates, coating conditions, reagent dilutions, buffers,
and incubation times were preliminary tested. Antibody levels are ex-
pressed as absorbance in arbitrary units at 405 nm.
3
OH
RO
O
HO
Ara
O
2.11. Statistical analysis
H C
HO
3
O
Rha
S1=
Xyl I
O
Statistical analysis was performed using Graph Pad prism 6.
O
OH
Xyl II
O
D'agostino-Parson test was employed to evaluate if data assumed a
Gaussian distribution. Mann-Whitney U tests were used to evaluate the
predictive antibody values.
HO
O
O
O
HO
HO
OH
HO
OH
OH
O
Xyl IV
O
HO
O
HO
3. Results and discussion
HO
OH
Ara
O
Xyl III
3
.1. Structural analysis
H C
3
O
S2=
HO
Rha
Xyl I
Three triterpene glycosides 1–3 (Fig. 1) were isolated from an
O
OH
O
Xyl II
OH
aqueous–ethanolic extract of the roots of Weigela florida “rumba” by
various solid/liquid chromatographic methods. Their structures were
determined by 2D NMR and mass spectrometry as follow.
O
O
HO
O
O
HO
HO
HO
OH
OH
HO
HO
Glc
O
O
The HR-ESIMS (positive-ion mode) spectrum of compound 1 es-
tablished its molecular formula as C61
H
98
O
27 with a pseudo-molecular
peak at m/z 1285.5334 [M + Na] , indicating a molecular weight of
262.
The H NMR spectrum of the aglycone part of 1 displayed signals
assignable to seven angular methyl groups at δ 0.76, 0.89, 0.90, 0.93,
1.03, 1.21 and 1.23 (s, each), one olefinic proton at δ 5.42 (H-12) and
one oxygen-bearing methine protons at δ 3.23 (H-3). The HMBC
+
OH
Xyl III
OH
1
1
O
HO
H
Ara
O
H
S3=
H C
H
3
O
2
3
HO
Rha
spectrum showed J and J couplings from the methyl protons (23, 24,
25, 26, 27, 29 and 30), which allowed the assignments of most carbons
and protons of this aglycone. The deshielded chemical shift of C-28
Xyl I
O
O
OH
HO
HO
C
observed at δ 180.2 suggested the presence of a carbonyl group of a
OH
carboxylic acid function. According to the literature data, this genin
was identified as oleanolic acid, already described in saponins from
Weigela stelzneri [9] (Table 1, Fig. 1).
Fig. 1. Structures of saponins 1–3.
in pure carbonate buffer 0.05 M at pH 9.6 (100 μL/well) and incubated
at +4 °C overnight. After three washes with 150 mM NaCl solution
containing 0.05% Tween 20 (washing buffer), non-specific binding sites
were blocked with 10% fetal bovine serum (FBS) in washing buffer
H
The HSQC spectrum of 1 displayed signals of six anomers at δ 4.78
(d, J = 7.8 Hz)/δ
C
102.4, δ
102.1, δ 5.19 (d, J = 7.1 Hz)/δ
C
104.6, and δ 6.06 (br s)/δ 101.1. The ring protons of
H
4.80 (d, J = 6.2 Hz) /δ
C
104.7, δ
H
4.86 (d,
J = 6.9 Hz)/δ
J = 6.9 Hz)/δ
C
H
C
106.3, δ
H
5.36 (d,
C
H
(
1
100 μL/well) at room temperature for 60 min. Sera diluted 1:100 in
the monosaccharide residues were assigned starting from the readily
identifiable anomeric protons by means of the 1H– H COSY, TOCSY,
HSQC and HMBC experiments. The monosaccharides obtained by acid
hydrolysis of 1 were identified by comparison on TLC with authentic
samples as xylose, arabinose and rhamnose. The absolute configura-
tions were determined to be D for xylose, and L for arabinose and
rhamnose by GC analysis according to a method previously described
1
0% FBS in washing buffer were incubated at +4 °C overnight. After
three washes, 100 μL of alkaline phosphatase conjugated with anti-
human IgM (diluted 1:1200 in 10% FBS in washing buffer) were added
to each well. After 3 h incubation at room temperature and three wa-
−
1
shes in washing buffer, 100 μL/well of 1 mg.mL
p-nitrophenylpho-
sphate (Sigma-Aldrich) in carbonate buffer (pH 9.6) were added. After
3
3
0 min, the reaction was stopped with 1 M NaOH (50 μL/well), and the
[10]. The relatively large JH-1,H-2 value of the Xyl and Ara (6.2–7.8 Hz)
absorbance was read in a multichannel ELISA reader (Tecan Sunrise,
Männedorf, Switzerland) at 405 nm. Antibody levels are expressed as
absorbance in arbitrary units at 405 nm.
in their pyranose form indicated a β anomeric orientation for Xyl, and
1
an α anomeric orientation for Ara. The large
JH-1,C-1 value of the Rha
(166 Hz) confirmed that the anomeric proton was equatorial (α-pyr-
anoid anomeric form).
Units of one α-L-arabinopyranosyl, one α-L-rhamnopyranosyl and
four β-D-xylopyranosyl were thus identified (Table 2). The mono-
desmosidic structure was elucidated by using mainly HMBC and NOESY
2.10. Immunoenzymatic assay of glycosides 2 and 3 and ursolic acid
Compounds 2 and 3 were coated on 96-well microplates BD Falcon™
spectra: the HMBC correlation at δ
NOESY correlation at δ 4.80 (Ara-1)/δ
heterosidic linkage between Ara and C-3 of the aglycone. The HMBC
cross-peak at δ 6.06 (Rha-1)/δ 75.2 (Ara-2) and the NOESY cross-
peak at δ 6.06 (Rha-1)/δ 4.46 (dd, J = 6.7, 6.2 Hz, Ara-2) proved the
H
4.80 (Ara-1)/δ
C
88.5 (C-3) and the
3.23 (H-3) confirmed the O-
separately using a solution 10 μg/ml in pure ethanol (100 μL/well). The
plates were then incubated at +4 °C overnight. The non-specific
binding sites were blocked with 10% FBS in 150 mM NaCl solution
without surfactant at room temperature for 60 min. Sera diluted 1:100
H
H
H
C
H
H
201

A.-S. Champy-Tixier et al.
Fitoterapia 128 (2018) 198–203
(
1
4
1→2) linkage between Rha and Ara. The correlation at δ
H
5.19 (Xyl I-
5.19 (Xyl I-1)/δ
.62 (dd, J = 9.5, 2.3 Hz, Rha-3) in the NOESY spectrum, confirmed the
4.78
4.78 (Xyl II-
4.10 (Xyl I-4) proved the (1→4) linkage between Xyl II and Xyl I.
The HSQC spectrum of 3 displayed signals of three anomers at δ
4.79 (d, J = 6.0 Hz)/δ 104.7, δ 5.22 (d, J = 7.6 Hz)/δ 106.6, and δ
6.08 (br s)/δ
were determined to be D for xylose, and L for arabinose and rhamnose
by GC analysis. After a total assignment by 2D NMR of the protons and
carbons of each sugar, units of one α-L-arabinopyranosyl, one α-L-
rhamnopyranosyl and one β-D-xylopyranosyl were identified (Table 2).
The structural analysis of the oligosaccharidic chain linked at the C-3 of
the aglycone was realized using mainly HMBC and NOESY spectra. The
H
)/δ
C
82.0 (Rha-3) in the HMBC spectrum, and at δ
H
H
C
H
C
H
C
101.0. The absolute configurations of these three sugars
(
(
1
1→3) linkage between Xyl I and Rha. The HMBC cross-peak at δ
Xyl II-1)/δ 76.6 (Xyl I-4) and the NOESY cross-peak at δ
)/δ
H
C
H
H
Finally, the structure analysis of the terminal sequence β-D-xylopyr-
anosyl-(1→2)-[β-D-xylopyranosyl-(1→4)]-β-D-xylopyranosyl was based
upon the HMBC correlation at δ
NOESY correlation at δ 5.36 (Xyl IV-1)/δ
HMBC cross-peak at δ 3.93 (dd, J = 8.1, 7.8 Hz, Xyl II-2)/δ
Xyl III-1) (Table 2, Fig. 1).
On the basis of the above results, the structure of the previously
undescribed compound 1 was elucidated as 3-O-β-D-xylopyranosyl-(1→
H
5.36 (Xyl IV-1)/δ
4.20 (Xyl II-4) and the
102.1
C
81.2 (Xyl II-4), the
H
H
HMBC correlation at δ
correlation at δ 4.79 (Ara-1)/δ
sidic linkage between Ara and C-3. The HMBC cross-peak at δ
(Rha-1)/δ 75.1 (Ara-2) and the NOESY cross-peak at δ 6.08 (Rha-1)/
4.47 (dd, J = 6.9, 6.0 Hz, Ara-2) proved the (1→2) linkage between
Rha and Ara. The correlation at δ 5.22 (Xyl I-1)/δ 82.0 (Rha-3) in the
HMBC spectrum and at δ 5.22 (Xyl I-1)/δ 4.63 (dd, J = 9.5, 2.9 Hz,
H
4.79 (Ara-1)/δ
3.22 (H-3) confirmed the O-hetero-
6.08
C
88.5 (C-3) and the NOESY
H
C
H
H
(
H
C
H
δ
H
2
)-[β-D-xylopyranosyl-(1→4)]-β-D-xylopyranosyl-(1→4)-β-D-xylopyr-
H
C
anosyl-(1→3)-α-L-rhamnopyranosyl-(1→2)-α-L-arabinopyr-
anosyloleanolic acid.
H
H
Rha-3) in the NOESY spectrum, confirmed the (1→3) linkage between
Xyl I and Rha (Table 2, Fig. 1).
The HR-ESIMS (positive-ion mode) spectrum of compound 2 es-
tablished its molecular formula as C62
H
100
O
28. It showed a pseudo-
On the basis of the above results, the structure of compound 3 was
elucidated as 3-O-β-D-xylopyranosyl-(1→3)-α-L-rhamnopyranosyl-(1→
2)-α-L-arabinopyranosyloleanolic acid. This compound has been al-
ready synthetized [12,13] but is isolated here for the first time from a
vegetal source.
This sequence was already found in saponins from Weigela stelzneri
[9]. Other species should be investigated to confirm this sequence as a
chemotaxonomic marker of the genus Weigela.
+
molecular peak at m/z 1315.5336 [M + Na] indicating a molecular
weight of 1292, differing from 1 of 30 uma.
1
13
The H and C NMR signals of compounds 1 and 2 were almost
surimposable excepted for the terminal hexose linked at the C-4 posi-
tion of Xyl II. The full NMR assignments of this remaining sugar and the
GC MS data, led to the identification of a β-D-glucopyranosyl moiety.
The NOESY correlation at δ
H
H
5.27 (Glc-1)/δ 4.23 (Xyl II-4) established
the structure of the previously undescribed compound 2 as 3-O-β-D-
xylopyranosyl-(1→2)-[β-D-glucopyranosyl-(1→4)]-β-D-xylopyranosyl-
3.2. Immunoenzymatic assay
(
1→4)-β-D-xylopyranosyl-(1→3)-α-L-rhamnopyranosyl-(1→2)-α-L-ara-
binopyranosyloleanolic acid.
After isolation and purification, triterpene glycosides 2 and 3 were
The HR-ESIMS (positive-ion mode) spectrum of compound 3 es-
chosen for the length of their oligasaccharidic chain, to be tested as
antigens to detect specific IgM antibodies in MS patients' sera and
normal blood donors (NBD) by ELISA. In parallel, the synthetic N-glu-
cosylated peptide CSF114(Glc) was employed as positive control an-
tigen. Moreover, the ursolic acid with a similar structure to the agly-
cone of 2–3, was tested to evaluate if antibody responses are directed to
glycosylated sites or/and the aglycone structure. IgM measurements
and statistical analysis were performed separately for saponins 2–3,
ursolic acid and CSF114(Glc). Data distribution of IgM antibody re-
sponses to CSF114(Glc), saponins 2–3 and ursolic acid can be observed
in Fig. 2.
tablished its molecular formula as C46
H
74
O
15, with a pseudo-molecular
peak at m/z 889.4410 [M + Na] , indicating a molecular weight of
66.
The H NMR spectrum of the aglycone part of 3 displayed signals
0.76, 0.88, 0.90, 0.94,
+
8
1
assignable to seven angular methyl groups at δ
.03, 1.22 and 1.24 (s, each), one olefinic proton at δ
J = 3.1 Hz), and one oxygen-bearing methine protons at δ
H
1
H
5.44 (br t,
3.22 (H-3).
H
1
3
C
In the C NMR spectrum, a signal at δ 180.2 suggested a carboxylic
acid function (C-28). As for 1 and 2, the genin of 3 was identified as
oleanolic acid [9] (Table 1, Fig. 1.).
1
1
0
0
.5
.0
.5
.0
Fig. 2. Data distribution of IgM antibody responses
to CSF114(Glc), saponins 2 and 3, and ursolic acid,
in sera of Multiple Sclerosis patients (MS) and
normal blood donors (NBD), determined by ELISA.
Data are reported as absorbance at 405 nm of sera
diluted 1:100. Data are presented as mean values
with the corresponding standard deviations.
M S p a tie n ts
C S F1 1 4 (Glc ) C S F1 1 4 (Glc ) s a p o n in
N B D
M S p a tie n ts
N B D
s a p o n in
M S p a tie n ts
s a p o n in
N B D
s a p o n in 3
M S p a tie n ts
u r s o lic a c id u r s o lic a c id
N B D
2
2
3
202

A.-S. Champy-Tixier et al.
Fitoterapia 128 (2018) 198–203
All compounds presented lower mean values for NBD than for MS
populations. The overall data distribution was statistically analyzed
using D'agostino-Parson test, and results showed that the ursolic acid
was the solely compound that passed the normality test (alpha = 0.05).
Then, antibody titers between MS patients and healthy controls were
evaluated using the Mann-Whitney U test. This cohort of MS patients
exhibited significantly greater anti-CSF114(Glc) antibodies when com-
pared to healthy subjects (P value < 0.05, two-tailed). In the other
hand, no significant differences in antibody levels against saponins 2
and 3, and ursolic acid was found between MS patients and healthy
controls (P value > 0.1, two tailed).
Conflict of interest
The authors declare that they have no conflict of interest.
Acknowledgements
We acknowledge the Multiple Sclerosis Clinical Care and Research
Center, Department of Neurosciences, Reproductive Sciences and
Odontostomatology, Federico II University (Naples, Italy) and the Azienda
Ospedaleria Universitaria Careggi, Clinica Neurologica, University of
Florence (Florence, Italy) for their support.
Indeed, saponins 2 and 3, monodesmosidic ones with oleanane-type
aglycone which differ by the length of their oligosaccharidic chain, are
not able to recognize IgMs in MS sera. Besides, ursolic acid, a com-
mercial triterpenic acid which has no sugar moiety also demonstrated
same results as saponins 2 and 3. The aim of this study was to define if
antibody responses are directed to glycosylated sites. Results obtained
by ursolic acid indicate that terpenoid structure only was no sufficient
for antibody recognition.
Peroni et al. [6] have previously tested five natural triterpene gly-
cosides for antibody recognition. From a structural point of view, these
compounds were bidesmosidic and monodesmosidic structures with
one glycosyl moiety at C-3 position of the aglycone. The glycosidic part
was composed of glucopyranosyl, sulfated glucopyranosyl, sulfated
quinovopyranosyl and xylopyranosyl moieties. In our study, mono-
desmosidic structures with three and six glycosyl moieties at C-3 posi-
tion of the aglycone were tested. The glycosidic part was composed of
arabinopyranosyl, rhamnopyranosyl, xylopyranosyl and glucopyr-
anosyl moieties. No relevant results were observed for saponins 2 and 3,
and the length of the oligosaccharidic chain and the type of sugars not
allowed the recognition of IgMs in MS sera. But according to previous
work [6] and our results, we can confirm the importance of specific
carbohydrate structures and hypothesize on the relevant role of the
ester carboxyl function at C-28 position in antibody recognition in MS
sera. New natural glycosides will be tested to try to establish structure/
activity relationships between natural compounds and IgM in sera of
patient affected by MS.
References
[1] J.S. Alexander, R. Zivadinov, A.H. Maghzi, V.C. Ganta, M.K. Harris, A. Minagar,
Multiple sclerosis and cerebral endothelial dysfunction: mechanisms,
Pathophysiology 18 (2011) 3–12.
[
2] H.A. Doyle, M.J. Mamula, Post-translational protein modification in antigen re-
cognition and autoimmunity, Trends Immunol. 22 (2001) 443–449.
3] F. Lolli, B. Mulinacci, A. Carotenuto, B. Bonetti, G. Sabatino, B. Mazzanti,
A.M. D'Ursi, E. Novellino, M. Pazzagli, L. Lovato, M.C. Alcaro, E. Peroni, M.C. Pozo-
Carrero, F. Nuti, L. Battistini, G. Borsellino, M. Chelli, P. Rovero, A.M. Papini, An N-
glucosylated peptide detecting disease-specific autoantibodies, biomarkers of mul-
tiple sclerosis, Proc. Natl. Acad. Sci. U. S. A. 102 (2005) 10273–10278.
4] C.B. Marta, A.R. Olivier, R.A. Sweet, S.E. Pfeiffer, N.H. Ruddle, Pathogenic myelin
oligodendrocyte glycoprotein antibodies recognize glycosylated epitopes and per-
turb oligodendrocyte physiology, Proc. Natl. Acad. Sci. U. S. A. 102 (2005)
13992–13997.
5] F. Nuti, E. Peroni, F. Real-Fernández, M.A. Bonache, A. Le Chavalier-Isaad,
M. Chelli, N. Lubin-Germain, J. Uziel, P. Rovero, F. Lolli, A.M. Papini, Post-trans-
lationally modified peptides efficiently mimicking neo-antigens: a challenge for
theragnostics of autoimmune diseases, Biopolymers: Pept. Sci. 94 (2010) 791–799.
6] E. Peroni, F. Real Fernández, C. Gheri, F. Nuti, A.C. Mitaine-Offer, F. Lolli,
M.A. Lacaille-Dubois, A.M. Papini, Natural triterpene glycosides for antibody re-
cognition, Planta Med. Lett. 3 (2016) e1–e6.
[
[
[
[
[7] T. Murayama, A. Kasahara, Y. Shiono, M. Ikeda, Structure elucidation of a tri-
terpene glycoside isolated from Weigela hortensis, Nat. Med. 57 (2003) 181–184.
[
8] Y.M. Won, Z.K. Seong, J.L. Kim, H.S. Kim, H.H. Song, D.Y. Kim, J.H. Kim, S.R. Oh,
H.W. Cho, J.H. Cho, H.K. Lee, Triterpene glycosides with stimulatory activity on
melanogenesis from the aerial parts of Weigela subsessilis, Arch. Pharm. Res. 38
(2015) 1541–1551.
[
9] A. Rezgui, A.C. Mitaine-Offer, T. Miyamoto, C. Tanaka, S. Delemasure, P. Dutartre,
M.A. Lacaille-Dubois, Oleanolic acid and hederagenin glycosides from Weigela
stelzneri, Phytochemistry 123 (2016) 40–47.
[
10] S. Hara, H. Okabe, K. Mihashi, Gas-liquid chromatographic separation of aldose
Summarizing, three new saponins (1–3) were extracted from the
roots of Weigela florida “rumba”, an ornamental shrub very common in
European gardens. Triterpene glycosides 2 and 3 which differ by their
oligosaccharidic chain were tested as antigens to detect specific IgM
antibodies in MS patient sera and normal blood donors (NBD) by ELISA.
Results showed no significant difference in antibody levels against sa-
ponins 2 and 3 between MS patients and healthy controls. New sapo-
nins with different structures and sugar moieties will be further tested
to confirm the importance of carbohydrate structures in the auto-
antibodies recognition.
enantiomers as trimethylsilyl ethers of methyl 2-(polyhydroxyalkyl)thiazolidine-
4
(R)-carboxylates, Chem. Pharm. Bull. 35 (1987) 501–506.
[11] F. Rizzolo, G. Sabatino, M. Chelli, P. Rovero, A.M. Papini, A convenient microwave-
enhanced solid-phase synthesis of difficult peptide sequences: case study of gra-
micidin a and CSF114(Glc), Int. J. Pept. Res. Ther. 13 (2007) 203–208.
[
12] Q. Liu, H. Liu, L. Zhang, T. Guo, P. Wang, M. Geng, Y. Li, Synthesis and antitumor
activities of naturally occurring oleanolic acid triterpenoid saponins and their de-
rivatives, Eur. J. Med. Chem. 64 (2013) 1–15.
[
13] L. Wang, Z. Wang, S. Su, Y. Xing, Y. Li, M. Li, J. Liu, S. Yang, Synthesis and cyto-
toxicity of oleanolic acid trisaccharide saponins, Carbohydr. Res. 442 (2017) 9–16.
203