NATURAL PRODUCT RESEARCH
3
(kmax 233, 307 and 368 nm) indicated the presence of a conjugated system. And the
positive results for Mg-HCl reaction and Molish reagent suggested 1 to be a flavon-
oid glycoside.
13C-NMR spectrum (Table S1) of compound 1 displayed a group of signals (dC
100.5, 76.7, 75.9, 73.3, 69.7 and 60.7) belonging to a hexosyl unit, and the remaining
48 carbon signals attributing for a triflavonoid structure, among which are three car-
bonyl groups (dC 188.2, 185.4 and 176.3), three aromatic methyl groups (dC 7.4, 6.9
and 6.8), 11 CH (dC 122.1, 119.5, 118.1, 116.4, 116.2, 115.9, 115.6, 115.3, 114.9, 95.5 and
1
94.3) and 31 quaternary carbons. H-NMR spectrum (Table S1) of 1 showed signals for
three sets of typical ABX coupling systems [dH 7.13 (1H, d, J ¼ 8.7 Hz, H-20), 7.82 (1H, d,
J ¼ 2.0 Hz, H-50) and 6.18 (1H, dd, J ¼ 8.7, 2.0 Hz, H-60)], [dH 7.11 (1H, d, J ¼ 8.7 Hz, H-
0
0
0
ꢂ
ꢂ
ꢂ
2 ), 6.93 (1H, d, J ¼ 2.0 Hz, H-5 ) and 7.05 (1H, dd, J ¼ 8.7, 2.0 Hz, H-6 )], and [dH 6.73
(1H, d, J ¼ 8.7 Hz, H-2ꢀ0), 7.14 (1H, d, J ¼ 2.0 Hz, H-5ꢀ0) and 7.01 (1H, dd, J ¼ 8.7, 2.0 Hz,
H-6 )], corresponding to the 3 ,4 -dihydroxy-substituted rings B, B and Bꢀ of
a triflavonoid moiety, and a series of signals in the range dH 5.5-3.0 related to a sugar
moiety. Also two aromatic protons [dH 6.07 (1H, s, H-8) and 5.91 (1H, s, H-8ꢀ)] and
ꢀ0
0
0
ꢂ
ꢂ
three aromatic methyl signals [dH 1.91 (3H, s, Me-6), 2.10 (3H, s, Me-6 ) and 1.84 (3H,
s, Me-6ꢀ)] were presented. The above 1D-NMR data suggested 1 to be a triflavonoid
glycoside, and the aglycone moiety was composed of three molecules of flavonoids
with certain structural characteristics of 6-C-methylquercetin (Ibewuike et al. 1996;
Quang et al. 2008).
Comprehensive analysis of 1D-NMR and HSQC (Figure S4) spectra of 1 made assign-
ꢂ
ꢂ
ꢂ
ments of one flavonoid molecule (the segment II), referring to rings A , B and C ,
which were confirmed by HMBC correlations (Figure S1) from CH3-6 (dH 2.10) to [dC
ꢂ
0
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
161.0 (C-5 ), 102.9 (C-6 ) and 163.4 (C-7 )0], from H-2 (dH 7.11) to [dC 146.0 (C-2 0),
0
0
0
0
ꢂ
ꢂ
ꢂ
ꢂ
129.3 (C-1 ), 140.0 (C-3 ) and 140.5 (C-4 )], from H-5 (dH 6.93) to [dC 129.3 (C-1 ),
0
0
ꢂ
0
ꢂ
ꢂ
ꢂ
140.0 (C-3 ), 140.5 (C-4 ) and 122.1 (C-6 )], and from H-6 (dH 7.05) to [dC 146.0
(C-2 ) and 115.3 (C-2 )]. Compared with literature values of 6-C-methylquercetin
0
ꢂ
ꢂ
(Ibewuike et al. 1996;0 Quang et al. 2008),0 the segment II was further identified. The
ꢂ
ꢂ
carbon signals of C-3 (dC 140.4) and C-4 (dC 141.5) shifting dC 5.0-7.5 low-frequency
against those of 6-C-methylquercetin, suggested one available binding position for the
ꢂ
ꢂ
segment III, and the absence of an H-8 signal indicated a substituted C-8 , revealing
another binding position for the segment I.
In the segment III, except the signals assignable to rings A and B of a 6-C-methyl
quercetin molecule, the distinctive carbon signals [dC 188.2 (C-4ꢀ), 100.1 (C-2ꢀ), and
90.4 (C-3ꢀ)] suggested a dihydroflavanol structure and a 2ꢀ,3ꢀ-epoxide moiety, and the
carbon of C-3ꢀ provided a possible binding position for the segment II. Between seg-
ꢀ
0
ꢂ
ments II and III, the ether linkage (C-3 )-O-(C-4 ) was elucidated. The segment III,
refering to rings Aꢀ, Bꢀ, and Cꢀ, and the 2ꢀ,3ꢀ-epoxide moiety, was confirmed by
HMBC correlations (Figure S1) from CH3-6ꢀ (dH 1.84) to [dC 160.4 (C-5ꢀ), 104.8 (C-6ꢀ),
and 165.9 (C-7ꢀ)], from H-8ꢀ (dH 5.91) to [dC 104.8 (C-6ꢀ), 165.9 (C-7ꢀ), 156.5 (C-9ꢀ), and
99.3 (C-10ꢀ)], from H-2ꢀ0 (dH 6.73) to [dC 100.1 (C-2ꢀ), 146.8 (C-4ꢀ0), and 119.5 (C-6ꢀ0)],
from H-5ꢀ0 (dH 7.14) to [dC 124.6 (C-1ꢀ0), 146.6 (C-3ꢀ0), and 146.8 (C-4ꢀ0)], and from H-6ꢀ0
(dH 7.01) to [dC 124.6 (C-1ꢀ0) and 115.6 (C-5ꢀ0)], and was verified by related NMR data
of reported baeckeins C, D and E (Jia, Zhou, et al. 2011; Jia et al. 2013).