Haller J. et al.
Previous investigations described the composition of epicuticular
waxes of S. pusilla (Lütz and Gülz, 1985) and the presence of
flavonoids in S. alpina (Kroslakova et al., 2016). The occurrence of
saponins in S. alpina has already been suspected more than 90
years ago (Luft, 1926). Since S. alpina has certain relevance in
traditional medicine, this present study investigated the
occurrence of saponins in this and three other Soldanella species.
HR-ESI-MS of isolated compounds
Experiments were performed on a micrOTOF-QII mass spectrometer
(Bruker Daltronics, Bremen, Germany) by directly infusing a methanolic
solution of each compound (0.25 mg/mL). MS conditions were set to
negative electrospray ionisation, a nebuliser pressure of 4.4 psi, a HV
capillary voltage of 3.5 kV, 180°C dry temperature, a dry gas flow of 4
L/min and a scan range from m/z 100 to 3000, respectively.
Sample preparation
Experimental
For preparation of the HPLC sample solutions the finely powdered plant
material (root or leaf, 50 mg) was extracted with methanol (3 mL) by
sonication (15 min). The mixture was centrifuged (5 min, 3500 rpm) and
the supernatant placed in a 10 mL volumetric flask. This extraction step
was repeated two more times, solutions were combined and the flask filled
to volume with methanol. Sample solutions are stable for at least two
weeks if stored at 4°C. Prior to HPLC analysis an aliquot of the leaf extract
(5.00 mL) or the root extract (10.00 mL) was evaporated to dryness under
reduced pressure, the residue was re-dissolved in methanol (1.00 mL) and
membrane filtered (GHP Acrodisc 13, 0.45 μm polypropylene membrane,
Pall, Port Washington, NY, USA).
Plant material and chemicals
Three individual samples (i.e. 5–10 entire plants each) of Soldanella alpina
(SA-1 to SA-3) as well as single samples of S. pusilla (SP-1), S. x transsylvanica
(ST-1) and S. montana (SM-1) were analysed. All of them, except SA-3
(Mediplant, Conthey, Switzerland) came from Gärtnerei Eschmann, Emmen,
Switzerland. They were harvested in May 2010 (SA-1), June 2011 (SA-2 and
SA-3) and May 2012 (SP-1, ST-1 and SM-1). The plant material was
authenticated by one of the authors (S. Schwaiger), separated into roots
(R) and leaves (L), and dried at ambient temperature. Voucher specimens
of all samples are deposited at the Institute of Pharmacy, Pharmacognosy,
University of Innsbruck.
All chemicals required for isolation and analysis were of analytical grade
(p.a.) quality and were purchased from Merck (Darmstadt, Germany).
Deuterated pyridine for NMR experiments came from Euroiso-top (Saint-
Aubin Cedex, France). HPLC grade water (18.0 MΩ/cm) was produced by
a Satorius Arium 611 water purification system (Göttingen, Germany).
Analytical conditions
HPLC experiments were performed on an Agilent 1200 system (Agilent,
Waldbronn, Germany) equipped with binary pump, autosampler and
column heater, fitted with a Sedex 85 LT ELSD detector from Sedere
(Alfortville Cedex, France). The best separation was achieved using a
Kinetex HILIC 2.6 μm 100 Å column (150 mm × 4.6 mm, Phenomenex,
Torrance, CA, USA), utilising a mobile phase comprising 20 mM ammonium
acetate in water (A), and a 9:1 mixture (v/v) of acetonitrile and 200 mM
ammonium acetate solution (B). Both solutions were adjusted to a pH 4.0
with acetic acid. The linear gradient started at 3% A/97% B to end after
25 min at 8% A/92% B. The column was washed for 5 min with 30%
A/70% B before re-equilibration for 20 min. Column temperature, flow rate
and sample volume were set to 30°C, 0.8 mL/min and 5 μL, respectively. The
evaporative light scattering detection (ELSD) settings were 40°C, a nebuliser
pressure of 3.6 bar (nitrogen) and gain 12.
Isolation of triterpene saponins from S. alpina
Sequential extraction (5 × 100 mL, 10 min ultrasonic bath for each solvent)
of dried and powdered S. alpina roots (8.5 g, sample SA-3-R) and
subsequent evaporation yielded n-hexane (45.8 mg), dichloromethane
(39.7 mg), ethyl acetate (18.6 mg) and methanol (2304.4 mg) extracts. The
methanolic extract (ca. 2 g) was dissolved in methanol/water (1:1; v/v, 4.0
mL) and mixed with acetone (8.0 mL). The resultant precipitate was
separated by decantation to yield fraction 1 (17 mg). The remaining
solution was mixed with more acetone (4 mL) and kept at room
temperature for 24 h. During that time the solution separated into an oily
syrup (at the bottom of the beaker) and a light yellow solution that were
separated and evaporated to dryness to give fraction 2 (904.0 mg) and
fraction 3 (605.6 mg), respectively. Fraction 3 was subsequently dissolved
in methanol (8.0 mL) and mixed with diethyl ether (8.0 mL) resulting in a
white precipitate (246.5 mg, fraction 3p) and the remaining supernatant,
which yielded fraction 3s (383.6 mg) after solvent removal. Fraction 3s
was further purified by centrifugal partition chromatography (FCPC A200,
Method validation
A standard stock solution of the three saponins (5 mg each) was prepared
with methanol (5.00 mL). Further dilutions were prepared in the ratio of 1:2
with the same solvent. The limits of detection (LOD) and limits of
quantitation (LOQ) were visually evaluated, defining concentrations with
a peak height of 3- or 10-times the baseline noise, respectively. Accuracy
was investigated by spiking sample SA-2-L with different concentrations
of the standards (high, medium and low spike). The quantitative results
were compared with the theoretically present amount and expressed as
recovery rate. Precision was investigated by analysing five individually
prepared solutions of sample SP-1-L on day 1. On days 2 and 3 the same
procedure was repeated, and the variation within one day (intra-day
precision) and within three days (inter-day precision) was calculated based
on peak area.
Kromaton, Angeres, France) using
a
solvent system of
chloroform/methanol/water/1-propanol (9:12:8:1; v/v/v/v). The lower phase
was pumped through the system (0.5 mL/min) as mobile phase in
descending mode with counter clockwise rotation of 800 rpm. Fractions
(2 mL) were collected and checked by HPTLC (LiChrospher silica gel 60
F254 plates, 7 μm, Merck, Darmstadt, Germany) using a developing solution
of chloroform/methanol/water/formic acid (6:3.2:0.8:1.2; v/v/v/v) and
visualised by spraying and heating with anisaldehyde-sulphuric acid
reagent. The pooled fractions (fraction 3sA) collected between 160 to 288
mL contained a mixture of two saponins, while the fraction eluting between
362 and 382 mL contained a pure saponin (13.8 mg, Sap3; hRf 50.0).
An aliquot (78.6 mg) of fraction 3sA was further separated by silica gel
column chromatography (diameter: 2.5 cm, length: 30 cm) using an
isocratic elution with chloroform/methanol (1:1, v/v). Fractions (3 mL) were
monitored using HPTLC yielding Sap1 (eluting between 180 and 225 mL,
36.3 mg, hRf 54.4) and Sap2 (eluting between 270 and 360 mL, 17.4 mg,
hRf 52.2). Purity and identity of the isolated compounds were verified by
HPTLC, LC-MS3, GC-FID, NMR and HR-ESI-MS (see Supporting Information
for LC-MSn and one-dimensional (1D)- and two-dimensional (2D)-NMR
spectra as well as details on hydrolysis/GC-FID experiments).
Results and discussion
Structural elucidation of saponins in S. alpina
Firstly, the three isolated saponins (Fig. 1) were studied by LC-ESI-
MS/MS. The parent ions ([M-H]˗) were observed at m/z 1059.9
(Sap1 and Sap2) and 1075.9 (Sap3). Taking Sap1 as an example,
further fragmentation showed signals at m/z 927.6 (MS2) and
765.1, 603.1 and 471.9 (MS3). The fragmentation pattern suggested
the loss of four sugar units as displayed by observed mass-
wileyonlinelibrary.com/journal/pca
Copyright © 2017 John Wiley & Sons, Ltd.
Phytochem. Anal. 2017