C. Charrier et al. / Journal of Molecular Catalysis B: Enzymatic 102 (2014) 167–173
169
was used to obtain a better 13C signal-to-noise ratio. Chemical shifts
are given on a ␦-scale with digital resolution justifying the reported
values.
out with an Agilent DAD detector (200–400 nm) and a Quatro
Micromass spectrometer (Waters) operated in negative mode from
150 to 900 D.
In Paris (Univ. Descartes) 1H- and 13C-NMR spectra (500.13 MHz
and 125.77 MHz, respectively) were recorded on a Bruker Avance-
500 spectrometer in DMSO-d6 or acetone-d6 solvents. All NMR
spectra were measured at 20 ◦C, and proton and carbon chem-
ical shifts were referenced to the solvent signals: 2.50 ppm for
proton and 39.52 ppm for carbon in DMSO-d6, and 2.05 ppm for
proton and 29.85 ppm for carbon in (CD3)2CO. Two-dimensional
correlation spectroscopy (COSY) and total correlation spectroscopy
(TOCSY), rotating frame nuclear Overhauser effect spectroscopy
(ROESY), heteronuclear single quantum coherence (HSQC), and het-
eronuclear multiple bond correlation (HMBC) experiments were
performed using standard pulse sequences. Proton chemical shift
assignments were obtained by the analysis of COSY and TOCSY
spectra and carbon chemical shift assignments were made by the
analysis of HSQC and HMBC spectra. For reference, a complete pro-
ton and carbon peak assignment of the NMR spectra was made on
the parent compounds.
2.5. Cultivation of microorganism
The Streptomyces sp. strain M52104, isolated from soil (for the
full details of the strain used see [22–24]), was maintained on
ISP medium 2 agar (Difco) slants at 25–28 ◦C. Liquid medium for
cultures contained 2% D-(+)-glucose, 0.5% yeast extract (Difco),
0.5% soytone (Difco), 0.5% NaCl, 0.5% K2HPO4 in deionized water,
adjusted to pH 7 with HCl solution. Media were sterilized by auto-
claving at 121 ◦C for 20 min. Precultures of 50 mL were inoculated
with four drops of a bacterial cell suspension and incubated 72 h at
27 ◦C in the orbital shaker (200 rpm). Each preculture was used to
inoculate 1 L cultures, which were incubated at 27 ◦C (200 rpm) for
72 h.
2.6. Preparative syntheses and isolation of products
2.3. MALDI-TOF/TOF
Streptomyces sp. M52104 was grown at 27 ◦C with orbital shak-
ing (200 rpm) in 1L-conical flasks containing 250 mL-volume of
liquid medium during 72 h. Silybin substrates were added to the
culture (0.1 g/L final concentration) and incubations were con-
tinued for 96 h and monitored by HPLC/UV/MS. The cells were
removed by filtration on GF/A filter (Whatman) with celite fil-
ter aid, and the filtrate treated with Amberlite XAD-16. The
resin was repeatedly washed with water and then the glu-
curonides were eluted with ethanol. After evaporation in vacuo
to a small volume, the eluate residue was solubilized in a mini-
mal volume of acetonitrile and submitted to successive injections
(0.5–2 mL) on a semi-preparative Nucleodur (Macherey-Nagel)
C18 column (10 m, 32 × 250 mm) using a water-acetonitrile
gradient, containing 0.1% formic acid (40 mL/min) and detec-
tion at 280 nm. Collected fractions were evaluated for purity by
HPLC/UV/MS and those fractions that contained sufficient quantity
and purity of the glucuronides of interest were pooled, evapo-
rated and lyophilized. Final purity (in the 80–100% range) was
estimated from analytical HPLC and UV absorption between 230
and 320 nm using the aglycone molecular absorption as a refer-
ence.
MALDI–MS spectra were measured on
a MALDI-TOF/TOF
ultraFLEX III mass spectrometer (Bruker-Daltonics, Bremen, DE).
Positive/negative spectra were calibrated externally using the
monoisotopic [M + H]+ or [M − H]− ions of PepMixII calibrant
(Bruker-Daltonics). ␣-Cyano-4-hydroxycinnamic acid (5 mg/mL)
in 50% CH3CN/0.1% TFA was used as a MALDI matrix. A 0.4 L
of sample dissolved in MeCN was premixed with 0.5 L of the
matrix solution on the target and allowed to dry at ambient
temperature. The MALDI-TOF spectra were collected in reflectron
mode. The exact masses were measured using LTQ Orbitrap XL
hybrid mass spectrometer (Thermo Fisher Scientific, Waltham, MA,
USA) equipped with an electrospray ion source. The mobile phase
consisted of methanol/water (4:1), flow rate 30 L/min and the
samples were injected using a 2-L loop. Spray voltage, capil-
lary voltage, tube lens voltage and capillary temperature were
4.5 kV, −40 V, −120 V and 275 ◦C, respectively. The mass spectra
were internally calibrated using deprotonated stearic acid as lock
mass.
2.4. HPLC
3. Results
Analytical chromatography was carried out on the Shimadzu
Prominence UFLC system (Kyoto, JP) consisting of a DGU-20A
mobile phase degasser, two LC-20AD solvent delivery units, a
SIL-20ACHT cooling autosampler, a CTO-10AS column oven and
SPD-M20A diode array detector. The Chromolith Performance RP-
18e monolithic column (100 × 3 mm i.d., Merck, Darmstadt, DE)
coupled with a guard column (5 × 4.6 mm) (Merck, DE) was used
in all analyses. The PDA data were acquired in the 200–450 nm
range and the 285 nm signal was extracted. Binary gradient elu-
tion: mobile phase A (CH3CN/H2O/HCOOH (10/90/0.1; v/v/v);
mobile phase B (CH3CN/H2O/HCOOH (90/10//0.1; v/v/v)); gradi-
ent, 0–1.5 min, 100% A; 1.5–15 min, 0–100% B; 16–16.5 min, 100–0%
B. Flow rate was 1 mL/min at 25 ◦C. All samples were dissolved in
water.
The separation and quantification of glucuronide metabolites
was carried out using an Agilent model 1100 liquid chromato-
graph interfaced with UV and MS detection. Separations were
performed at 40 ◦C on Nucleodur (Macherey-Nagel) C18 columns
(5 m, 4 × 70 mm for preliminary separations and 3 × 250 mm for
final measurements), using a similar water-acetonitrile gradient,
containing 0.1% formic acid (0.5 mL/min). Detection was carried
As shown in Figs. 2 and 3, both silybin A (1a) and B (1b)
(100 mg each) are nearly completely converted by the Streptomyces
strain into three monoglucuronides, accounting for the presence
of 3 peaks (Figs. 2 and 3, and S25, Supplementary material) at
ca. 15.5 min (2a/2b), ca. 17.4 min (4a/4b) and 15.9 min (3a/3b)
with a similar mass of m/z 657 [M − H] and a similar fragment
at m/z 481 [M-H-176] corresponding to the aglycone fragment
(Fig. S22, Supplementary material). After separation and purifi-
cation by semipreparative HPLC, two of them, 2a/2b (34 and
26 mg, respectively), and 3a/3b (6 and 4 mg, respectively) were in
large excess compared to 4a/4b (less than 1 mg). Another minor
product, not isolated, was detected at a lower retention time (Rt
12.3 min) and exhibited mass peaks at m/z 833 (657 + 176) and
peak with m/z 655 and 479 (2 AMU lower than the major glu-
curonide peaks, Fig. S24, Supplementary material) was detected
with a higher Rt (19.3 min) and a higher UV absorption (see
Figs. 2 and 3).