Z.-P. Yu, et al.
Fitoterapia135(2019)99–106
and then excess (S)-(−)-1-phenylethylamine and NaBH3CN were
added. With a catalytic amount of acetic acid added, the mixture was
stirred at 40 °C for 3 h. After evaporation of the ethanol, the residue was
stirred with acetic anhydride in pyridine at room temperature for 5 h to
obtain the amino derivatives. Using the same method, the amino deri-
vatives of authentic L- and D-glucoses were obtained. These derivatives
were subjected to HPLC analyses using the following conditions [HPLC
column: Agilent ZORBAX SB-C18, 5 μm, 4.6 × 250 mm; Mobile phase:
MeCN-H2O (60%, v/v); Flow rate: 1.00 mL/min]. As shown in Fig. S34
[Supplementary material (SM)], the retention times of standard D- and
L-glucose derivatives were observed at tR 5.254 and 4.987 min, re-
spectively, while that of the sugar derivative of 1 was observed at tR
5.285 min, thus assigning D-configuration for the glucoses in 1.
m/z 375.1 [M - H]−
.
2.8. Cytotoxic assay
The cytotoxic activities of all the compounds were tested using SRB
method as we described previously [6]. Adriamycin was used as a po-
sitive control.
2.9. AChE inhibitory assay
The AChE inhibitory activities of all the compounds were tested
using modified Ellman's method as we described previously [9]. Tacrine
was used as a positive control.
2.6. Determination of D-glucoses, L-arabinose and L-rhamnose of 2
3. Results and discussion
Using the same procedure as 1, the amino derivatives of the hy-
drolytic sugars from 2 and their corresponding standard ones (D- and L-
glucoses, D- and L-arabinoses, and L-rhamnose) were prepared. These
derivatives were then subjected to HPLC analyses using the following
conditions [HPLC column: Agilent ZORBAX SB-C18, 5 μm,
4.6 × 250 mm; Mobile phase: MeCN-H2O (50%, v/v); Flow rate:
1.00 mL/min]. As shown in Fig. S35 (SM), the retention times for
standard D- and L-glucoses, D- and L-arabinoses, and L-rhamnose were at
tR 7.597, 7.219, 6.361, 6.257 and 8.652 min, respectively. The retention
times of hydrolytic samples were observed at tR 7.595, 6.247 and
8.632 min, and thus the D-glucoses, L-arabinose and L-rhamnose were
determined for 2.
Compound 1 was obtained as a white amorphous powder. Its mo-
lecular formula of C44H56N2O20 with 18 indices of hydrogen deficiency
was deduced from the protonated molecule ion peak at m/z 933.3497
[M + H]+ (calcd 933.3499) in (+)-HRESIMS analysis and 13C NMR
data. The IR spectrum showed strong absorption bands at 3416, 1690
and 1630 cm−1 attributable to hydroxyl, ester carbonyl and olefinic
groups, respectively. Two anomeric proton signals at δH 4.84 and 4.65
(both d, J = 7.9 Hz), along with the corresponding anomeric carbon
signals observed at δC 100.7 and 100.4, suggested the presence of two
monosaccharide units. Analysis of 1D TOCSY experiments by exciting
the aforementioned anomeric protons (Figs. S13 and S14, SM) returned
two separate coupling networks, which together with the 1He1H COSY
and HSQC data enabled the unambiguous assignments of 1H and 13C
NMR data for the sugar moieties (Table 1). Such information revealed
that both monosaccharide units were glucopyranoses [3,10], with β-
anomeric configurations being determined by the coupling constants of
anomeric protons (both J = 7.9 Hz) [11]. Further HPLC analyses of the
(S)-(−)-1-phenylethylamine derivatives from the hydrolyzed sugars
and authentic D- and L-glucoses (Fig. S34, SM) permitted the assignment
of D-configurations for both monosaccharide units. Besides those as-
signable to the sugar part, the 1H and 13C NMR data (Table 1) also
displayed signals for a carboxyl (δC 173.7), two ester carbonyls (δC
171.3, 169.4), an ortho disubstituted phenyl (δH 7.48, d, J = 7.9 Hz;
7.32, d, J = 8.2 Hz; 7.14, dd, J = 8.2, 7.0 Hz; 7.05, dd, J = 7.9, 7.0 Hz),
four double bonds including a monosubstituted terminal one (δH 5.40,
br d, J = 17.8 Hz; 5.28, br d, J = 10.6 Hz; 5.86, ddd, J = 17.8, 10.6,
7.4 Hz), two trisubstituted ones (δH 7.88, 7.43, each br s) and a tetra-
substituted one, two acetal protons (δH 5.93, d, J = 9.2 Hz; 5.27, d,
J = 5.2 Hz), three oxygenated or ammoniated methines (δH 5.30, 4.54,
3.90), and a secondary methyl (δH 1.10, d, J = 6.8 Hz). These func-
tionalities and the two glucopyranoses accounted for 13 out of 18
2.7. Alkaline hydrolysis
Pure compound 1 (5.0 mg) was digested with 5% NaOH solution
(3.0 mL) at 80 °C for 3 h. The reaction mixture was neutralized with 5%
HCl solution and then concentrated to dryness. The residue was sus-
pended with ethanol and then filtered to remove NaCl. Compounds 1a
(2.0 mg, tR = 12.3 min) and 1b (1.0 mg, tR = 8.1 min) were finally ob-
tained by HPLC purification using 30% MeCN-H2O (v/v) as mobile
phase.
2.7.1. 1a
White amorphous powder; UV (MeOH) λmax (log ε) 221 (3.96) nm;
ECD (c 0.05, MeOH) λ (Δε) 218 (1.09), 233 (−6.11) nm; (−)-ESIMS:
m/z 559.3 [M - H]−
.
2.7.2. 1b
White amorphous powder; UV (MeOH) λmax (log ε) 232 (3.91) nm;
ECD (c 0.04, MeOH) λ (Δε) 224 (−2.34), 253 (0.21) nm; (−)-ESIMS:
O
5'
9'
O
1H-1H
22
Unit A
Unit B
COSY
HMBC
23
6
11'
16
17
21
6'
15
14
OH
20
5
23
O
6
NOESY
3'
7'
19
3
O
O
B
C
5
O
4
b
d
22
NH
2
8'
1'
15
14
O
10'
17
8
c
3
1
A
2
9
O
O
1'''
OH
9
13
3'
19
18
5'''
6'''
4'''
NH
1
21
2'''
O
HO
a
3'''
12
1'
11
HO
5'
O
6'
HO
1''
OH
9'
8'
Unit A
Unit B
5''
6''
4''
2''
3''
HO
7'
HO
HO
Fig. 2. (A) 1He1H COSY and Key HMBC correlations for 1; (B) Key NOESY correlations for 1.
102