Journal of Natural Products
Article
A10 system (Molsheim, France). Reaction progress was monitored by
TLC on precoated silica gel plates (Kieselgel 60 F254, Merck) and
visualized by UV254 light. Flash column chromatography was
performed on silica gel (particle size 40−63 μm, Merck). Unless
otherwise stated, all reagents were obtained from Sigma-Aldrich and
used without further purification.
Determination of Solubility in Physiological pH Range.
Excess amounts of the chlorides, 1, 2, 4, and 5 were added to Na
phosphate buffer 0.1 M from pH 4.5 to 9 in a sealed vial. The samples
were left for 1 week under continuous stirring at 25 °C. Then, each
solution was filtrated with RC membrane 0.45 μm syringe filters and
appropriately diluted with mobile phase before injection in HPLC-
ESIMS/MS.
Synthesis of Compounds 2, 2a, and 4. Compound 2 was
prepared starting from 1, by pyrolysis in solvent-free conditions under
Determination of Lipophilicity by 1-Octanol/Water Partition
Coefficient and Distribution. The measurements of the 1-octanol/
water partition coefficients were carried out starting from an initial
concentration of 0.1 mM for all analytes in Na phosphate buffer 0.1 M
26
27
a vacuum or under microwave irradiation (Scheme S1, Supporting
Information). In contrast to the microwave-assisted synthetic
27
procedure described by Das et al., which used a commercial
microwave oven, in the present study the reaction was performed with
a scientific single-mode microwave apparatus, which allowed
reproducible conditions to be established. Compound 4 was
semisynthesized through the hydrolysis reaction of the acetal ring of
2
9
from pH 4.5 to 9. Then, 2 mL of this standard solution was added to
mL of 1-octanol presaturated with Na phosphate buffer, and the
samples were left to equilibrate for 1 week under continuous stirring at
5 °C. After centrifugation, the two phases were carefully separated.
2
2
2
6
The aqueous solution was collected and appropriately diluted with
mobile phase before injection in HPLC-ESIMS/MS. The general
equation used for the calculation of LogDo/w is reported in the
Supporting Information (eq S1).
Determination of Albumin Binding Association Constants
by Fluorescence Quenching. The measurements were performed
compound 1, in the presence of sulfuric acid and phloroglucin
(
Scheme S2, Supporting Information).
HPLC-ESIMS/MS. Liquid chromatography was performed using a
2690 Alliance system (Waters, Milford, MA, USA) equipped with a
built-in 120-position cooled autosampler. Analytical separation was
conducted on a Phenomenex Luna C18 (5 μm, 150 mm × 2.0 mm i.d.)
column. The mobile phases were 10 mM formic acid in water adjusted
to pH 4.0 with ammonia (solvent A) and a solution of acetonitrile−
methanol (95:5 v/v) (solvent B). Separation was achieved at a 0.15
−
5
−1
using bovine serum albumin (1.0 × 10 mol L ) in Tris-HCl buffer
−1
−1
solution (0.05 mol L Tris, 0.15 mol L NaCl, pH 7.4). The
emission spectra (excitation at 295 nm and emission wavelengths of
300−500 nm) were recorded with increasing concentrations of
−1
mL min flow rate under gradient elution conditions (5 min at 95% A
and 10 min at 40% A), followed by a column purge (5 min at 20% A)
and column re-equilibration (10 min at 0% A). The injected sample
volume was 10 μL. The analytical column was maintained at 30 °C. In
the optimized analytical conditions, the mean retention times were for
−5
−1
quencher (1, 2, 4, 5) from 0 to 2.4 × 10 mol L at increments
−5
−1
of 0.2 × 10 mol L at room temperature. The experimental results
published on the quenching mechanism of fluorescence of human
serum albumin (HSA) by 1 indicate that it is a static quenching
3
0
procedure and that the binding association can be determined by
plotting a modified Stern−Volmer equation (see eq S2, Supporting
Information).
1
10.1 ± 0.1 min, 2 10.6 ± 0.1 min, 4 9.4 ± 0.1 min, and 5 9.7 ± 0.1
min (n = 20). The column effluent was introduced into the ESI source,
operating in the positive-ionization mode, connected to a triple-
quadruple mass spectrometer (Quattro-LC, Micromass) operating in
the multiple reaction monitoring acquisition mode [m/z 336 → 320
Human Feeding Study. Berberine was administered to human
volunteers according to a clinical study protocol, approved on April 21,
2
009, by S. Orsola-Malpighi Hospital Review Board, No. 7-2209-U-
(1), m/z 322 → 307 (2), m/z 324 → 280 (4), m/z 338 → 323 (5),
SPER. The complete study protocol is described in the Supporting
and m/z 414 → 220 (R,S)-noscapine]. Nitrogen was used as nebulizer
gas at 75 L/h flow rate and as desolvation gas at 850 L/h. Ion source
block and desolvation temperatures were set to 130 and 250 °C,
respectively. Capillary voltage was 3.0 kV. The cone voltage was 45 kV.
MassLynx software version 4.0 was used for data acquisition and
processing.
Information. To evaluate the metabolic effects of berberine chloride
(
1) on cholesterol and biliary metabolisms, 12 consecutive euglycemic
subjects were enrolled, affected by mild to moderate hyper-
cholesterolemia (c-LDL >130 mg/mL and <190 mg/mL). All patients
received 15 mg/kg daily of berberine chloride (1) for three months.
The age range of the patients was between 18 and 70 years old. No
patient received any medication affecting biliary, glucose, or
cholesterol metabolism.
Quantification. A seven-point calibration curve was prepared daily
and injected in duplicate. Calibration samples were obtained in the
1.5−20 nM concentration range prepared in plasma for plasma matrix
samples and in mobile phase for physicochemical property samples
using (R,S)-noscapine as internal standard (IS, 2.5 nM). Linear
calibration curve parameters were obtained from the plot of the
analyte peak area/internal standard peak area versus analyte
concentration using a least-squares regression analysis (weight = 1/
ASSOCIATED CONTENT
Supporting Information
General synthetic procedures and spectroscopic character-
ization of compounds 2, 2a, 2b, and 4. General equations used
■
*
S
2
x ). Correlation coefficients were ≥0.996 for all analytes. The HPLC-
for the determination of LogDo/w and K . In vivo study
B
ESIMS/MS method was validated according to the current guide-
28
protocol and comparison between basal vs after treatment (V0
vs V3) of cholesterol (total, LDL, and HDL), total triglycerides,
total serum bile acid, fasting glucose, and fasting insulin. This
lines and has good sensitivity (LOQ 1.5 nM), selectivity, precision
(
CV% intra- and interassay <9%), and accuracy (bias % inter- and
intra-assay <7%) for all analytes.
Plasma Sample Preparation. Analytes were extracted from
plasma using an Oasis HLB (hydrophilic−lipophilic balance 200 mg, 6
mL) SPE column. The optimized extraction procedure utilized:
conditioning with 2 mL of MeOH and 2 mL of H O Milli-Q; loading
AUTHOR INFORMATION
Corresponding Author
2
■
with 780 μL of plasma (+10 μL of IS, 40 ng/mL) diluted with 2 mL of
ammonium formate (10 mM pH 7.0), washing with 1 mL of formic
acid (2%, v/v) and 2 mL of ammonium formate (10 mM pH 7.0),
elution with 2 mL of MeOH, followed by 1 mL of MeOH containing
1
% (v/v) CH COOH and 1 mL of MeOH containing 2% (v/v)
4
Notes
3
NH OH, vacuum drying, and reconstitution with 400 μL of mobile
phase. The recovery percentage was ≥98% for all analytes.
The authors declare no competing financial interest.
Determination of pK Values in Silico. The pK values were
a
a
ACKNOWLEDGMENTS
■
determined in silico using Epik module version 2.2 from Schro
Suite 2010, with water as the solvent. The Maestro interface of
Schrodinger Suite 2010 (Maestro, version 9.1, Schrodinger, LLC, New
York, NY, 2010) was used to build all molecules.
̈
dinger
We thank the University of Bologna for financial support. We
acknowledge Prof. A. Mazzanti and Dr. S. Grilli for fruitful
discussions on the spectroscopic characterization of the
̈
̈
7
71
dx.doi.org/10.1021/np400607k | J. Nat. Prod. 2014, 77, 766−772