Biocatalytic generation of molecular diversity: Modification of ginsenoside Rb1 by β-1,4-Galactosyltransferase and Candida antarctica lipase

Article abstract of DOI:10.1002/1522-2675(200207)85:7<1943::AID-HLCA1943>3.0.CO;2-T

A series of specific derivatives of the complex protopanaxadiol glycoside ginsenoside Rb₁ (1) were prepared by catalysis of two unrelated enzymes: the β-1,4-galactosyltransferase from bovine colostrum (GalT) and the lipase B from Candida antarctica (Novozym 435). Both of the enzymes showed the expected regioselectivity towards specific glucose OH groups (i.e., OH-C(4) for GalT and preferentially the primary OH-C(6) for Novozym 435), accompanied by a nonpredictable 'site selectivity' for the gentiobiose disaccharide unit linked at C(20) of the dammarane skeleton. The galactosylated products 1a-e and the acetylated products 1f-h were isolated by HPLC and fully characterized by extensive MS and NMR analysis.

Full text of DOI:10.1002/1522-2675(200207)85:7<1943::AID-HLCA1943>3.0.CO;2-T

Helvetica Chimica Acta Vol. 85 (2002)
1943
Biocatalytic Generation of Molecular Diversity: Modification of Ginsenoside
Rb1 by b-1,4-Galactosyltransferase and Candida antarctica Lipase
Part 41)
by Steffen Gebhardta), Stefan Bihlera), Manfred Schubert-Zsilavecza), Sergio Riva*b), Daniela Montib),
Laura Falconeb)c), and Bruno Danielic)
a) Institut f¸r Pharmazeutische Chemie, Johann Wolfgang Goethe-Universit‰t, Marie-Curie-Strasse 9,
D-60439 Frankfurt am Main
b) Istituto di Chimica del Riconoscimento Molecolare, C.N.R., Via Mario Bianco, 9, I-20131 Milano
c
¡
) Dipartimento di Chimica Organica e Industriale, Universita degli Studi di Milano, Via Venezian, 21,
I-20133 Milano
A series of specific derivatives of the complex protopanaxadiol glycoside ginsenoside Rb1 (1) were
prepared by catalysis of two unrelated enzymes: the b-1,4-galactosyltransferase from bovine colostrum (GalT)
¾
and the lipase B from Candida antarctica (Novozym 435). Both of the enzymes showed the expected
regioselectivity towards specific glucose OH groups (i.e., OHÀC(4) for GalT and preferentially the primary
¾
OHÀC(6) for Novozym 435), accompanied by a nonpredictable −site selectivity× for the gentiobiose
disaccharide unit linked at C(20) of the dammarane skeleton. The galactosylated products 1a e and the
acetylated products 1f h were isolated by HPLC and fully characterized by extensive MS and NMR analysis.
Introduction. Natural products are still today an unsurpassed source of bioactive
compounds, which are useful to improve well-being and the quality of life of humans
and constitute a relevant economic resource for the pharmaceutical, cosmetic, and food
industries [2]. Chemical modification is currently largely applied to implement the
properties of existing materials by modulating their pharmacological activity and/or
bioavailability. This approach emulates some of the chemistry that occurs in biological
systems, where natural or nonnatural substrates are modified by the action of
biocatalysts to generate physiologically active compounds and/or their intermediates,
or are deactivated to be expelled from the organism. However, the semi-synthetic
chemical approach to natural-products derivatives runs into difficulties when the
molecules are labile or when they possess multiple functionalities, as in the case of
glycosides. In fact, chemical reactions involving these molecules have often indis-
criminate outcomes, resulting in the formation of mixtures of products and by-products.
In a way, these protocols can be considered a sort of combinatorial synthesis, offering
the opportunity to obtain a large number of derivatives from a single starting material.
However, on the other side, they are thwarted by difficulties in the isolation of
single products, which usually requires tedious and low-yielding chromatographic
procedures.
1
)
For Part 3, see [1].

1944
Helvetica Chimica Acta Vol. 85 (2002)
In the last years, it has been shown that biocatalysis offers a number of key
advantages over chemical synthesis when working on complex molecules, advantages
based on the chemo-, regio-, and stereoselectivity of enzymes and on the possibility to
carry out reactions under mild conditions [3]. Even in cases of simple transformations
(e.g., an acylation reaction), the chemo-and regioselectivity of the biocatalyst may
eliminate several protection-deprotection steps embodied in the chemical route, once
the enzyme and the reaction conditions have been carefully selected.
For many years, we have been studying the structural and functional elaboration of
several natural compounds, exploiting alternatively the regio-and stereoselectivity of
dehydrogenases [4], glycosidases [5], and glycosyltransferases [1][6], and the
regioselectivity of hydrolases [7]. As an extension of this work, we report here on
the use of two classes of enzymes (hydrolases and glycosyltransferases) for the
preparation of a series of specific derivatives of the complex protopanaxadiol glycoside
ginsenoside Rb1 (1), one of the saponins isolated from Panax Ginseng C.A. Meyer, a
plant widely used in traditional Chinese medicine [8]. This compound is a dammarane-
type triterpene, carrying a gentiobiose unit linked to C(20) and a sophorose unit linked
to C(3). The presence of four glucopyranose moieties, carrying 3 primary and 11
secondary OH groups, makes 1 an ideal candidate for the exploitation of biocatalysis to
generate a limited array of derivatives having a specific structure. The selected enzymes
will furnish derivatives with opposite characteristics, more hydrophilic in the case of
glycosyltransferases and more lipophilic in the case of hydrolases, thus allowing
generation of a molecular diversity that can be exploited to detect suitable modulations
of the bioactivity of the starting product.
Results and Discussion. We started our investigation by submitting ginsenoside
Rb1 (1) to the catalytic action of the b-1,4-galactosyltransferase from bovine milk
(GalT) [9]. It is well-known that this enzyme, naturally involved in the synthesis of
lactose or of N-acetyllactosamine from UDP-galactose and glucose or N-acetylglucos-
amine, respectively, is very selective concerning the formation of the new glycosidic
bond (b-1,4-linkage) and can accept a wide variety of aglycone-substituted glucopy-
ranosides. However, in the presence of more than one glucose moiety bearing a free
OHÀC(4), the reaction outcome is strongly dependant on the structure of the
substrate. For instance, we have shown that it is possible to have the selective
galactosylation of only one of the two or three possible glucose acceptors of stevioside
and steviolbioside [10], while, in the case of the more simple ginsenoside Rg1 (2) that
contains two glucose moieties, both saccharides were galactosylated, although at
different rates [1]. Due to the increased structural complexity of ginsenoside Rb1 (1),
this compound appeared to be a challenging substrate to get more information on the
performance of GalT.
The usual enzymatic cocktail (GalT, epimerase, alkaline phosphatase) was added to
a buffered solution of 1 containing UDP-glucose, a-lactalbumine, and MnCl2 (Scheme,
see Exper. Part) [1][10]. Due to the low solubility of 1, 20% DMSO was added because
we have demonstrated that this cosolvent is compatible with the activity of this enzyme
mixture [11]. The reaction was monitored by TLC and by HPLC, and a complex
chromatographic profile was obtained after 5 days of incubation at 378 (Fig. 1, 54%
overall conversion). The products corresponding to the five HPLC peaks (products

Helvetica Chimica Acta Vol. 85 (2002)
1945
Scheme. Multi-Enzymatic Protocol Used to Galactosylate Ginsenoside Rb1 (1)
1a e) were isolated by prep. HPLC and extensively analyzed by the usual 1D and 2D
NMR and MS techniques2).
The most abundant product 1d (tR 16.1 min) showed a quasi-molecular-ion peak
[M À H]À at m/z 1269 in its LSI-MS (neg. mode), indicating the presence of one
additional hexose moiety compared to 1 ([M À H]À at m/z 1107). The structure of 1d
was established as [b-d-Glcp-(1 ! 2)-b-d-Glcp-(1 ! O-3)]-[b-d-Galp-(1 ! 4)-b-d-
Glcp-(1 ! 6)-b-d-Glcp-(1 ! O-20)]-protopanaxadiol after extensive spectral analyses.
Accordingly, the 1H-NMR spectrum (CD3OD) of 1d showed 5 d at d 4.66, 4.58, 4.43, 4.39, and 4.34 due to
anomeric protons of the 5 sugar units, all with a b-d-configuration (vicinal coupling constant values between 7.2
and 7.6 Hz). Analysis of the spin systems by 1H,1H-COSYand TOCSYexperiments disclosed the glucopyranose
nature of the sugar units whose anomeric protons were at d 4.66, 4.58, 4.43, and 4.39, whereas the anomeric
proton at d 4.34 belonged to the newly introduced galactopyranose moiety because of its connection to the signal
at d 3.81 of the equatorial HÀC(4).
All the homo-and hetero-2D NMR experiments showed that the sophorose disaccharide moiety linked at
C(3) was unaffected, displaying anomeric proton signals at d 4.43 and 4.66 as in the starting material 1. The
attachment of the galactopyranose moiety to OHÀC(4) of the external glucose moiety of the gentiobiose
disaccharide unit was finally established as follows. The anomeric proton at d 4.58 was shown to be due to
HÀC(1)(Glc''') because of its connection to C(20) (d 84.99) of the dammarane skeleton (HMBC), and,
therefore, the second glucose moiety of the gentiobiose disaccharide unit had its anomeric proton at d 4.39
(HÀC(1)(Glc'''')). The anomeric proton of the galactopyranose moiety (d 4.34) (HÀC(1)(Glc''''')) was linked to
the last-mentioned signal via (C(4)(Glc'''') at d 80.84 and HÀC(4)(Glc'''') at d 3.54, as nicely detailed in Fig. 2,
which reports the relevant plots of the COSY, HMBC, HMQC, and TOCSYexperiments of 1d, and highlight the
connectivity from HÀC(1)(Gal''''') to HÀC(1)(Glc'''').
The thorough investigation of these products and of other ginsenoside derivatives by ESI-MSn has already
been reported [12].
2
)

1946
Helvetica Chimica Acta Vol. 85 (2002)
Fig. 1. HPLC Plot of the GalT-catalyzed galactosylation of 1. tR: 1a 6.2 min, 1b 8.5 min, 1c 12.2 min, 1d 16.1 min,
1e 17.5 min, and 1 24.8 min, at ca. 40% conversion.
The most retained product 1e (tR 17.5 min) was also a monogalactosylated
derivative, and NMR analysis similar to that of 1d showed that the anomeric proton
of the galactopyranose unit (d 4.50) was linked (HMBC) to the C(4)(Glc''')
(d 80.22) belonging to the −inner× glucopyranose moiety of the gentiobiose unit.
Furthermore, HÀC(4)(Glc''') (d 3.72) was connected (HMQC) to HÀC(1)(Glc''')
(d 4.60) via TOCSY. Product 1e was thus recognized as [b-d-Glcp-(1 ! 2)-b-d-Glcp-
(1 ! O-3)]-[b-d-Galp-(1! 4)-[b-d-Glcp-(1! 6)]-b-d-Glcp-(1! O-20)]-protopanaxadiol.
In the third mono-galactosylated derivative, product 1b (tR 8.5 min), the gentio-
biose disaccharide unit was found to be unaffected, and applying the above
spectroscopic approach, the galactopyranose moiety was shown to be linked to the
terminal glucose unit of the sophorose moiety (see data in the Exper. Part). Therefore,
1b was identified as [b-d-Galp-(1 ! 4)-b-d-Glcp-(1 ! 2)-b-d-Glcp-(1 ! O-3)]-[b-d-
Glcp-(1 ! 6)-b-d-Glcp-(1 ! O-20)]-protopanaxadiol.
The two remaining products, products 1a and 1c (tR 6.2 and 12.2 min, resp.), were
both di-galactosylated derivatives, as indicated by the quasi-molecular ion peak at m/z
1431 ([M À H]À) in their LSI-MS (neg. mode). Although the isolated amount of these
compounds was too low to record structurally significant 13C-NMR, HMQC, and
HMBC spectra and although both contained some unidentified impurities, we could
demonstrate that, in case of 1a, the two galactopyranose units were linked to the
terminal glucose units of the sophorose and gentiobiose moieties, respectively, by
comparison of its TOCSY plot with those of the fully characterized starting material 1,
of product 1b, and of product 1d. Fig. 3 reports the relevant portion of the TOCSY plots

Helvetica Chimica Acta Vol. 85 (2002)
1947
Fig. 2. Expanded 1H,1H-COSY, HMBC, HMQC, and TOCSY plots of 1d, showing the correlation between
HÀC(1)(Gal''''') (ˆ H-1(Gal''''')) and HÀC(1)(Glc'''') (ˆ H-1(Glc''''))

1948
Helvetica Chimica Acta Vol. 85 (2002)
and clearly shows that the contour profile of 1a is superimposable on that of 1b and 1d.
Similarly, structure elucidation for product 1c, which was by far the less-abundant
product, was based on the comparison of its TOCSY with those of 1, 1d, and 1e. This
indicated that the sophorose unit was unaffected and that the two galactose units
were attached to OHÀC(4) of each glucose moiety of the gentiobiose unit. In this
way, the structures of the di-galactosylated compounds 1a and 1c could be established
as [b-d-Galp-(1 ! 4)-b-d-Glcp-(1 ! 2)-b-d-Glcp-(1 ! O-3)]-[b-d-Galp-(1 ! 4)-b-d-
Glcp-(1 ! 6)-b-d-Glcp-(1 ! O-20)]-protopanaxadiol and as [b-d-Glcp-(1 ! 2)-b-d-
Glcp-(1 ! O-3)]-[b-d-Galp-(1 ! 4)-b-d-Glcp-(1 ! 6)-[b-d-Galp-(1 ! 4)]-b-d-Glcp-
(1 ! O-20)]-protopanaxadiol, respectively.
The structures of the isolated products allow us to deduce that GalT has an
overwhelming preference for the recognition of the gentiobiose unit linked at C(20): in
fact, four of the five isolated compounds carry at least one galactose unit at this −upper×
disaccharide moiety. Additionally, the mono-galactosylated compounds 1b and 1e were
produced in similar amount, despite that the formation of the latter requires the access
to an internal and, in principle, sterically much-more-hindered glucose moiety. Worth
mentioning is also the formation of product 1c, presumably derived from further
galactosylation of 1e, although in very small amounts.
A final comment can be made on the polarity of products 1a e. Clearly, the −lower×
portion of the derivatives of 1, which contains the sophorose disaccharide unit, is
mainly responsible for the polarity of the molecules. In fact, galactosylation of this
portion produces the less-retained compounds in HPLC, thus suggesting a weak
interaction of this region of the molecule with the apolar stationary phase.
To obtain new derivatives of ginsenoside Rb1 (1), this time possessing a more
hydrophobic character, the esterification of 1 by action of hydrolases in organic
solvents was investigated. Due to the insolubility of 1 in apolar or moderately polar
solvents, only few enzymes of this type could be used in a preliminary screening, and,
¾
after experimentation, only the lipase B from Candida antarctica (Novozym 435) [13]
proved to be able to acylate the substrate with almost complete conversion and
selectivity. This finding was in agreement with previous works showing that this enzyme
is particularly suitable for the regioselective esterification of complex natural glyco-
sides, the cognate Rg1 (2) [14] and the haemolytic saponine digitonin being just two
significant examples [15]. Accordingly, acylation reactions were performed by
dissolving compound 1 in DMF/THF 1:2 containing vinyl acetate. Novozym¾ 435
was added, and the suspension was shaken at 458. The reaction was very fast, and, after
only 4 h, 1 was almost completely converted into two acetylated products 1f and 1g,
which were isolated and purified by flash chromatography. At variance with the
previously reported galactosylated derivatives 1a e, an unambiguous successful
analysis of the NMR spectra of the new compounds could be performed only by
taking the spectra in (D6)DMSO, at 458 or at higher temperature. This solvent was
unique in suppressing extensive overlapping of signals and in securing good stability of
products at the working temperature.
The less-polar and less-abundant compound 1f was recognized as a di-O-
acetylginsenoside Rb1 on the basis of its LSI-MS (neg. mode; [M À H]À ion at 1191)
1
and of its H-NMR spectrum. Detailed spectral analyses revealed its structure to be
6'',6''''-di-O-acetylginsenoside Rb1 (1f).

Helvetica Chimica Acta Vol. 85 (2002)
1949
Fig. 3. TOCSY Plots of compounds 1, 1a, 1b, and 1d. The spin systems of the glucopyranose moieties carrying a
galactosylpyranosyl substituent are outlined.
The 1H-NMR spectrum of 1f showed, among other signals, 2 s at d 2.00 and 2.02 for the acetate groups and 2
AB portions of 2 ABX systems at low field, i.e., at d 4.25 and 4.02, and 4.19 and 3.98, respectively, for the
acetylated oxymethylene protons of two glucose units. To unambiguously locate the positions of attachment of
the acetyl residues, the usual detailed NMR study was carried out by means of 1D and 2D techniques, which
allowed us to establish the complete direct, vicinal, and long-range H,H and H,C connectivity of both
disaccharides 1f and 1g. The conclusion of these investigations was that the AB portion at d 4.25 and 4.02 of 1f
was due to CH2(6)(Glc''''), the oxymethylene protons of the external glucose unit of the −upper× gentiobiose

1950
Helvetica Chimica Acta Vol. 85 (2002)
disaccharide moiety, because these signals were linked to the anomeric HÀC(1)(Glc'''') at d 4.24. In turn, this
was connected to C(6)(Glc''') at d 68.81, the substituted oxymethylene of the internal glucose unit of the
gentiobiose moiety. The second acetate group of 1f was located on the external glucose unit of the −lower×
sophorose disaccharide moiety. In fact, the AB portion at d 4.19 and 3.98 was correlated to the anomeric proton
at d 4.47 possessing a long-range interaction with the substituted C(2)(Glc') whose 13C-NMR signal was at low
field (d 80.23) due to the ether linkage to the external glucose.
The more polar and more abundant product, apparently homogeneous on TLC, was
recognized as a monoacetate by LSI-MS analysis (neg. mode) showing an [M À H]À ion
at 1149. However, by analyzing its NMR spectrum, it turned out immediately that the
isolated product consisted of a mixture of two compounds 1g/1h in a ca. 9 :1 ratio (two
acetate signals at d 2.02 and 2.00). Separation of these two monoacetates proved to be a
difficult task and could be achieved, even if not completely, by prep. HPLC on a
Partisil-10-ODS-3 column in the isocratic mode. The major component 1g could be
isolated in pure form, whereas the minor 1h was always contaminated by 1g. The NMR
data of the pure, more abundant monoacetate were consistent with the structure of 6''''-
O-acetylginsenoside Rb1 (1g). By careful comparison of the NMR spectra of the
monoacetate mixture and of the pure monoacetate 1g, it was possible to detect
structurally significant signals (see Exper. Part) due to the less-abundant acetate 1h,
indicating that its external glucose unit of the sophorose moiety was acetylated at its
primary OH group. Therefore, the structure of 6''-O-acetylginsenoside Rb1 (1h) is
suggested for the minor acetate.3)
3
)
During a detailed investigation of the dammarane saponins of commercial American ginseng (Panax
quinquefolium L.), Tanaka and co-workers [16] isolated a new saponin, named quinquenoside R1 (Q-R1),
which was shown to be mono-O-acetyl-Rb1. By means of MS and with a view to the acetylation shift in the
13C-NMR spectra, the acetyl group was located at OHÀC(6) of the terminal glucose moiety of the b-
sophorose unit of Rb1. We observe that, because the 13C-NMR spectrum of Rb1 was taken in (D5)pyridine
and, in this solvent, the resonances of C(4), C(5), and C(6) of Glc', Glc'', and Glc'''' were superimposed (d
71.4, 78.0, and 62.6, resp.), the structure proposed by the authors should be the object of a more-detailed
investigation.

Helvetica Chimica Acta Vol. 85 (2002)
1951
The signals of the acetylated gentiobiose moiety of 1g, i.e., CH2(6)(Glc'''') at d 4.01 and 4.24,
HÀC(1)(Glc'''') at d 4.23, C(6)(Glc''') at d 68.75, CH2(6)(Gl''') at d 3.57 and 3.86, and HÀC(1)(Glc''') at d
4.41, were almost superimposible on the signals of the corresponding portion of the diacetate 1f, while those of
the unaffected sophorose unit were equivalent to the respective signals in the starting material 1.
¾
Therefore, Novozym 435 confirmed its regioselectivity for the acylation of primary
OH groups of ginsenosides [14], and, more significantly, the −upper× gentiobiose moiety
proved to be more reactive than the −lower× sophorose unit, as in the previously
described biocatalyzed galactosylation. These results, obtained with two completely
unrelated enzymes, are quite surprising when compared with the data previously collected
with the cognate ginsenoside Rg1 (2) [14]. In fact, in the case of 2, the glucose moiety
linked at C(20) was the less-reactive portion of the molecule, and both acetylation and
galactosylation took place preferentially at the glucose unit linked at C(6).
This is an additional support to the following general finding made by us and by
others: while the regioselectivity of both enzymes towards specific glucose OH groups
(strictly speaking, OHÀC(4) for GalT and preferentially the primary OHÀC(6) for
¾
Novozym 435) is the same in different substrates, the capability to discriminate the
same moiety (i.e., a glucopyranose) located in different parts of a molecule (−site
selectivity×) is strongly dependent upon more-subtle structural characteristics that are
presently far from well-understood. For instance, the tridimensional structure of the
lipase B from Candida antarctica has been determined by X-ray analysis [17], and, in
the presence of a covalently bound phosphonate inhibitor [18], it has been shown that a
narrow cleft limits access to its active site. We have found experimentally that this lipase
can accept a large variety of bulky (nonlinearly shaped) and sterically demanding
substrates (i.e., ginsenosides, see [14] and this work, and digitonin, see [15]), a clear
indication of the enzyme×s capability to dynamically modulate its shape around the
active site in order to host these compounds. This structural flexibility is obviously a
positive factor from a synthetic point of view, but, on the other hand, the present lack of
definite rationales for this phenomenon impairs efforts to predict enzymatic site
selectivity.
Indirect support for the different reactivities of the two disaccharide moieties of 1
was obtained by inspection of the behavior of the related disaccharide ginsenoside Rg3
(3), which contains only the −lower× sophorose moiety and is formed as a mixture of
(20R) and (20S) epimers by acid hydrolysis of 1 [19]. Galactosylation of 3 gave no
detectable reaction products after five days, whereas acetylation under the same
conditions as for 1 occurred very slowly, giving ca. 5% conversion after 4 h and
complete conversion to a monoacetate only after 72 h. The monoacetylation product
was recognized as the 6''-O-acetylginsenoside Rg3 (3a) on the basis of the usual
spectroscopic analysis (see Exper. Part).
In the same way, we examined the reactivity of the simple 3-O-b-d-glucosyl
protopanaxadiol (ˆ ginsenoside Rh2, 4), prepared by enzymatic hydrolysis of ginseno-
side Rg3 (3). The enzyme hesperidinase [20][21], b-glucosidase, cellulase, and b-
galactosidase were tested. Whereas cellulase was uneffective and hesperidinase and b-
glucosidase gave extensive hydrolysis of ginsenoside Rg3, to our surprise, b-galactosidase
promoted a controlled reaction resulting in the formation of a single product identified
as ginsenoside Rh2 (4). This compound was completely unaffected by GalT but gave the
¾
expected 6'-O-acetyl derivative 2a after 72 h incubation with Novozym 435.

1952
Helvetica Chimica Acta Vol. 85 (2002)
In conclusion, we have shown that enzymes from different classes can be
successfully used to selectively modify a complex natural compound carrying several
reactive functionalities. A limited array of different derivatives was generated from this
lead compound. This result derived from a combination of the expected regioselectivity
and of a significant and nonpredictable site selectivity of the biocatalysts employed.
Experimental Part
General. UDP-glucose, a-lactalbumin from bovine milk, UDP-galactose-4'-epimerase (EC 5.1.3.2; from
galactose-adapted yeast), alkaline phosphatase (EC 3.1.3.1; from bovine intestinal mucosa, type VII S), b-
galactosidase (EC 3.2.1.23; from A. oryzae) were from Sigma. b-1,4-Galactosyltransferase (EC 2.4.1.22; from
bovine colostrum) was either purchased from Sigma or purified as described elsewhere [22], and its activity was
¾
evaluated in a spectrophotometric assay [23]. Novozym 435 was a generous gift from Novo-Nordisk. TLC:
precoated silica gel 60 F254 plates (Merck); detection with the Komarowsky reagent [24]. Flash chromatography
(FC): silica gel 60 (70 230 mesh, Merck). Anal. HPLC: Jasco HPLC instrument (model 880-PU pump, model
870-UV/VIS detector, l 200 nm); Licrospher-100-RP-18 (5 mm; Merck) reversed-phase anal. column. Prep.
HPLC: Partisil-10-ODS-3 column (Whatman). M.p.: Kofler apparatus; uncorrected. NMR Spectra: Bruker 600,
AMX-500, and AC-200; standard 1D and 2D pulse sequences with appropriate parameters. Mass spectra:
Finnigan MAT 8500; 4.5 kV Cs beam, negative-ion mode, glycerol as matrix.
Ginsenoside Rb1 (ˆ [b-d-Glcp-(1 ! 2)-b-d-Glcp-(1 ! O-3)]-[b-d-Glcp-(1 ! 6)-b-d-Glcp-(1 ! O-20)]-
protopanaxadiol ˆ (3b-,12b)-20-[(6-O-b-d-Glucopyranosyl-b-d-glucopyranosyl)oxy]-12-hydroxydammar-24-
en-3-yl 2-O-b-d-Glucopyranosyl-b-d-glucopyranoside; 1). Ginsenoside Rb1 (1) was isolated by column
chromatography (CHCl3/MeOH/H2O 8 :4 :0.3) from a root extract of Korean ginseng supplied by Indena,
1
Milano (Italy). Amorphous solid. M.p. 186 1888. H-NMR (CD3OD, 500 MHz): aglycone moiety: 1.00, 1.72
(CH2(1)); 1.71, 1.98 (CH2(2)); 3.20 (HÀC(3)); 0.78 (HÀC(5)); 1.49, 1.57 (CH2(6)); 1.29, 1.52 (CH2(7)); 1.42
(HÀC(9)); 1.25, 1.79 (CH2(11)); 3.71 (HÀC(12)); 1.72 (HÀC(13)); 1.02, 1.57 (CH2(15)); 1.32, 1.73 (CH2(16));
2.29 (HÀC(17)); 1.00 (Me(18)); 0.91 Me(19); 1.36 (Me(21)); 1.53, 1.80 (CH2(22)); 2.05, 2.12 (CH2(23)); 5.13
(HÀC(24)); 1.68 (Me(26)); 1.62 (Me(27)); 1.06 (Me(28)); 0.91 (Me(29)); 0.85 (Me(30)); Glc': 4.43
(d, J(1',2') ˆ 7.8, H ÀC(1')); 3.56 (dd, J(2',3') ˆ 8.9, HÀC(2')); 3.54 (HÀC(3')); 3.27 (HÀC(4')); 3.25
(HÀC(5')); 3.64 (dd, J(6'a,6'b) ˆ 11.9, J(5',6'a) ˆ 5.2, HaÀC(6')); 3.86 (dd, J(6'a,6'b) ˆ 9.9, J(5',6'b) ˆ 2.0,
HbÀC(6')); Glc'': 4.66 (d, J(1'',2'') ˆ 7.7, H ÀC(1'')); 3.21 (dd, J(2'',3'') ˆ 8.9, HÀC(2'')); 3.35 (HÀC(3'')); 3.20
(HÀC(4'')); 3.25 (HÀC(5'')); 3.62 (dd, J(6''a,6''b) ˆ 11.8, J(5'',6''a) ˆ 2.7, HaÀC(6'')); 3.82 (dd, J(5'',6''b) ˆ 1.5,
HbÀC(6'')); Glc''': 4.57 (d, J(1''',2''') ˆ 7.8, HÀC(1''')); 3.10 (dd, J(2''',3''') ˆ 9.6, HÀC(2''')); 3.34 (HÀC(3'''));
3.35 (HÀC(4''')); 3.42 (HÀC(5''')); 3.79 (dd, J(6'''a,6'''b) ˆ 12.1, J(5''',6'''a) ˆ 7.3, HaÀC(6''')); 4.05
(dd, J(5''',6'''b) ˆ 1.5, HbÀC(6''')); Glc'''': 4.36 (d, J(1'''',2'''') ˆ 7.8, H ÀC(1'''')); 3.20 (HÀC(2'''')); 3.34

Helvetica Chimica Acta Vol. 85 (2002)
1953
(HÀC(3'''')); 3.33 (HÀC(4'''')); 3.27 (HÀC(5'''')); 3.63 (dd, J(6''''a,6''''b)ˆ 11.8, J(5'''',6''''a) ˆ 3.5, HaÀC(6'''')); 3.87
(dd, J(5'''',6''''b)ˆ 1.5, HbÀC(6'''')). 1H-NMR ((D6)DMSO, 600 MHz, 458): aglycone moiety: 0.91, 1.58 (CH2(1));
1.53, 1.86 (CH2(2)); 3.01 (HÀC(3)); 0.68 (HÀC(5)); 1.38, 1.48 (CH2(6)); 1.19, 1.48 (CH2(7)); 1.31 (HÀC(9));
1.07, 1.62 (CH2(11)); 3.56 (HÀC(12)); 1.55 (HÀC(13)); 0.92, 1.55 (CH2(15)); 1.45, 1.78 (CH2(16)); 2.198
(HÀC(17)); 1.44, 1.70 (CH2(22)); 1.932, 2.02 (CH2(23)); 5.10 (HÀC(24)); 0.76, 0.81, 0.83, 0.89 (Me(18),
Me(19), Me(29), Me(30)); 1.24 (Me(21)); 1.64 (Me(26)); 1.56 (Me(27)); 0.98 (Me(28)); Glc': 4.27 (d, J(1',2') ˆ
7.8, HÀC(1')); 3.32 (HÀC(2')); 3.38 (HÀC(3')); ca. 3.10 (HÀC(4'), HÀC(5')); 3.43 (dd, J(6'a,6'b) ˆ 11.8,
J(5',6'a) ˆ 3.5, HaÀC(6')); 3.66 (dd, J(5',6'b) ˆ 1.5, HbÀC(6')); Glc'': 4.43 (d, J(1'',2'') ˆ 7.7, H ÀC(1'')); 3.00
(dd, J(2'',3'') ˆ 8.9, HÀC(2'')); ca. 3.12 (HÀC(3'')); 3.04 (HÀC(4'')); ca. 3.15 (HÀC(5')); 3.49 (dd, J(6''a,6''b) ˆ
11.8, J(5'',6''a) ˆ 2.7, HaÀC(6'')); 3.62 (dd, J(5'',6''b) ˆ 1.5, HbÀC(6'')); Glc''': 4.43 (d, J(1''',2''') ˆ 8.2, HÀC(1'''));
2.89 (dd, J(2''',3''') ˆ 9.6, HÀC(2''')); 3.18 (HÀC(3''')); 3.06 (HÀC(4''')); 3.31 (HÀC(5''')); 3.58
(dd, J(6'''a,6'''b) ˆ 12.1, J(5''',6'''a) ˆ 7.3, HaÀC(6''')); 3.90 (dd, J(5''',6'''b) ˆ 1.5, HbÀC(6''')); Glc'''': 4.24
(HÀC(1'''')); 2.96 (HÀC(2'''')); ca. 3.13 (HÀC(3'''')); ca. 3.06 (HÀC(4'''')); ca. 3.12 (HÀC(5'''')); 3.43
(dd, J(6''''a,6''''b) ˆ 11.8, J(5'''',6''''a) ˆ 3.5, HaÀC(6'''')); 3.66 (dd, J(5'''',6''''b) ˆ 1.5, HbÀC(6'''
©)). 13C-NMR (CD3OD, 50 MHz; values with the same superscript might be interchangeable): aglycone moiety:
40.59a (C(1)); 27.25 (C(2)); 91.33 (C(3)); 40.26a (C(4)); 57.56 (C(5)); 19.25 (C(6)); 35.86 (C(7)); 40.99a (C(8));
51.10 (C(9)); 37.92 (C(10)); 30.82b (C(11)); 71.93 (C(12)); 49.85 (C(13)); 52.41c (C(14)); 30.84b (C(15)); 27.25
(C(16)); 52.41c (C(17)); 84.97 (C(20)); 21.85 (C(21)); 36.79 (C(22)); 23.89 (C(23)); 126.05 (C(24)); 132.28
(C(25)); 25.94 (C(26)); 18.02 (C(27)); 28.41 (C(28)); 16.32, 16.70, 16.73, 17.39 (C(18), C(19), C(29), C(30));
Glc': 105.40 (C(1')); 81.11 (C(2')); 78.52 (C(3')); 71.60d (C(4')); 77.71 (C(5')); 62.84e (C(6')); Glc'': 104.51
(C(1'')); 76.32 (C(2'')); 77.93 (C(3'')); 71.93 (C(4'')); 78.33 (C(5'')); 63.12e (C(6'')); Glc''': 98.10 (C(1''')); 75.29
(C(2''')); 77.93 (C(3''')); 71.68d (C(4''')); 76.80 (C(5''')); 70.25 (C(6''')); Glc'''': 104.99 (C(1'''')); 75.14 (C(2''''));
78.33 (C(3'''')); 71.68d (C(4'''')); 77.93 (C(5'''')); 61.84 (C(6'''')). 13C-NMR ((D6)DMSO, 150 MHz, 458; values
with the same superscript may be interchangeable): aglycone moiety: 38.76 (C(1)); 25.89 (C(2)); 88.35 (C(3));
38.96 (C(4)); 55.78 (C(5)); 18.73 (C(6)); 34.52 (C(7)); 39.23 (C(8)); 49.36 (C(9)); 36.39 (C(10)); 30.19 (C(11));
69.18 (C(12)); 48.63 (C(13)); 50.84 (C(14)); 30.19 (C(15)); 25.89 (C(16)); 50.72 (C(17)); 82.42 (C(20)); 21.85
(C(21)); 35.52 (C(22)); 22.40 (C(23)); 125.35 (C(24)); 130.31 (C(25)); 25.58 (C(26)); 17.78 (C(27)); 27.63
(C(28)); 15.70, 15.93, 16.04, 17.05 (C(18), C(19), C(29), C(30)); Glc': 103.68 (C(1')); 81.40 (C(2')); 77.23 (C(3'));
70.29a (C(4')); 76.30b (C(5')); 61.30c (C(6')); Glc'': 103.91 (C(1'')); 75.30d (C(2'')); 76.49 (C(3'')); 70.29a (C(4''));
76.73b (C(5'')); 61.16 (C(6'')); Glc''': 96.62 (C(1''')); 73.73e (C(2''')); 76.81b (C(3''')); 70.19a (C(4''')); 75.62
(C(5''')); 68.77 (C(6''')); Glc'''': 103.47 (C(1'''')); 73.90e (C(2'''')); 76.87b (C(3'''')); 70.19a (C(4'''')); 76.88b
(C(5'''')); 61.26c (C(6'''')). LSI-MS: 1107 ([M À H]À), 945, 783, 621, 459.
Enzymatic Galactosylation of Ginsenoside Rb1 (1). GalT (7 U), epimerase (2 mg, 18 U), alkaline
phosphatase (8 ml), and a-lactalbumine (15 mg) were added to 15 ml of 50 mm Tris buffer (pH 7.4) containing
20% DMSO, 165 mg of 1 (10 mm), 5 equiv. of UDP-glucose (475 mg, 50 mm), and MnCl2 (25 mm). The soln.
was stirred at 308 for five days, adjusting the pH daily with 0.2n NaOH. Reaction outcome was monitored
qualitatively by TLC (AcOEt/MeOH/H2O 9 :2 :0.5) and quantitatively by anal. HPLC (H2O/MeCN 82 :18, flow
rate 0.9 ml/min). Purification was performed by prep. HPLC (same eluent, injection of the reaction mixture in
several portions, flow rate 5 ml/min): 6 mg (5.2%) of 1a, 11 mg (5.9%) of 1b, 3 mg (4.8%) of 1c, 35 mg (18.7%)
of 1d, and 16 mg (8.6%) of 1e.
(3b,12b)-20-{[O-b-d-Galactopyranosyl-(1 ! 4)-O-b-d-glucopyranosyl-(1 ! 6)-b-d-glucopyranosyl]oxy}-
12-hydroxydammar-24-en-3-yl O-b-d-Galactopyranosyl-(1 ! 4)-O-b-d-glucopyranosyl-(1 ! 2)-b-d-glucopyra-
noside (ˆ [b-d-Galp-(1 ! 4)-b-d-Glcp-(1 ! 2)-b-d-Glcp-(1 ! O-3)]-[b-d-Galp-(1 ! 4)-b-d-Glcp-(1 ! 6)-b-d-
Glcp-(1 ! O-20)]-protopanaxadiol; 1a). Amorphous solid. M.p. 261 2658 (dec.). 1H-NMR (CD3OD,
500 MHz; values with the same superscript might be interchangeable): aglycone moiety: 1.00, 1.72 (CH2(1));
1.71, 1.98 (CH2(2)); 3.20 (HÀC(3)); 0.78 (HÀC(5)); 1.49, 1.57 (CH2(6)); 1.29, 1.52 (CH2(7)); 1.42 (HÀC(9));
1.25, 1.79 (CH2(11)); 3.71 (HÀC(12)); 1.72 (HÀC(13)); 1.02, 1.57 (CH2(15)); 1.32, 1.73 (CH2(16)); 2.29
(HÀC(17)); 1.00 (Me(18)); 0.91 (Me(19)); 1.36 (Me(21)); 1.53, 1.80 (CH2(22)); 2.05, 2.12 (CH2(23)); 5.13
(HÀC(24)); 1.68 (Me(26)); 1.62 (Me(27)); 1.06 (Me(28)); 0.91 (Me(29)); 0.85 (Me(30)); Glc': 4.43
(d, J(1',2') ˆ 7.8, H ÀC(1')); 3.53 (HÀC(2')); 3.53 (HÀC(3')); 3.26 (HÀC(4')); 3.24 (HÀC(5')); 3.64
(HaÀC(6')); 3.85 (HbÀC(6')); Glc'': 4.67 (d, J(1'',2'') ˆ 7.7, H ÀC(1'')); 3.28 (HÀC(2'')); 3.52 (HÀC(3'')); 3.48
(HÀC(4'')); 3.39 (HÀC(5'')); 3.77 (HaÀC(6'')); 3.89 (HbÀC(6'')); Gal''': 4.34a (d, J(1''',2''') ˆ 7.8, HÀC(1'''));
3.53 (HÀC(2''')); 3.46 (HÀC(3''')); 3.81 (HÀC(4''')); 3.58 (HÀC(5''')); 3.69 (HaÀC(6''')); 3.77 (HbÀC(6'''));
Glc'''': 4.58 (d, J(1'''',2'''') ˆ 7.8, HÀC(1'''')); 3.11 (HÀC(2'''')); 3.34 (HÀC(3'''')); 3.33 (HÀC(4'''')); 3.42
(HÀC(5'''')); 3.79 (dd, J(6''''a,6''''b)ˆ 11.8, J(5'''',6''''a) ˆ 3.5, HaÀC(6'''')); 4.04 (dd, J(5'''',6''''b)ˆ 1.8; HbÀC(6''''));
Glc''''': 4.39 (d, J(1''''',2''''') ˆ 7.8, HÀC(1''''')); 3.26 (HÀC(2''''')); 3.51 (HÀC(3''''')); 3.54 (HÀC(4''''')); 3.39

1954
Helvetica Chimica Acta Vol. 85 (2002)
(HÀC(5''''')); 3.82 (HaÀC(6''''')); 3.89 (HbÀC(6''''')); Gal'''''': 4.32a (d, J(1'''''',2'''''') ˆ 7.8, HÀC(1'''''')); 3.53
(HÀC(2'''''')); 3.46 (HÀC(3'''''')); 3.81 (HÀC(4'''''')); 3.58 (HÀC(5'''''')); 3.69 (HaÀC(6'''''')); 3.77
(HbÀC(6'''''')). LSI-MS: 1431 ([M À H]À, 1269, 1107, 945, 783.
(3b,12b)-20-[(6-O-b-d-Glucopyranosyl-b-d-glucopyranosyl)oxy]-12-hydroxydammar-24-en-3-yl O-b-d-
Galactopyranosyl-(1 ! 4)-O-b-d-glucopyranosyl-(1 ! 2)-b-d-glucopyranoside (ˆ [b-d-Galp(1 ! 4)-b-d-Glcp-
(1 ! 2)-b-d-Glcp-(1 ! O-3)]-[b-d-Glcp-(1 ! 6)-b-d-Glcp-(1 ! O-20)]-protopanaxadiol; 1b). Amorphous sol-
id. M.p. 255 2598 (dec.). 1H-NMR (CD3OD, 500 MHz): aglycone moiety: 1.00, 1.72 (CH2(1)); 1.71, 1.98
(CH2(2)); 3.20 (HÀC(3)); 0.78 (HÀC(5)); 1.49, 1.57 (CH2(6)); 1.29, 1.52 (CH2(7)); 1.42 (HÀC(9)); 1.25, 1.79
(CH2(11)); 3.71 (HÀC(12)); 1.72 (HÀC(13)); 1.02, 1.57 (CH2(15)); 1.32, 1.73 (CH2(16)); 2.29 (HÀC(17)); 1.00
(Me(18)); 0.91 (Me(19)); 1.36 (Me(21)); 1.53, 1.80 (CH2(22)); 2.05, 2.12 (CH2(23)); 5.13 (HÀC(24)); 1.68
(Me(26)); 1.62 (Me(27)); 1.06 (Me(28)); 0.91 (Me(29)); 0.85 (Me(30)); Glc': 4.43 (d, J(1',2') ˆ 7.8, HÀC(1'));
3.53 (dd, J(2',3') ˆ 7.8, HÀC(2')); 3.54 (HÀC(3')); 3.27 (HÀC(4')); 3.24 (HÀC(5')); 3.64 (dd, J(6'a,6'b) ˆ 11.9,
J(5',6'a) ˆ 5.2, HaÀC(6')); 3.85 (dd, J(5',6'b) ˆ 2.0, HbÀC(6')); Glc'': 4.67 (d, J(1'',2'') ˆ 7.7, H ÀC(1'')); 3.28
(dd, J(2'',3'') ˆ 8.9, HÀC(2'')); 3.52 (HÀC(3'')); 3.47 (HÀC(4'')); 3.39 (HÀC(5'')); 3.77 (dd, J(6''a,6''b) ˆ 11.8,
J(5'',6''a) ˆ 2.7, HaÀC(6'')); 3.89 (dd, J(5'',6''b) ˆ 1.5, HbÀC(6'')); Gal''': 4.32 (d, J(1''',2''') ˆ 7.8, HÀC(1''')); 3.52
(HÀC(2)); 3.47 (HÀC(3''')); 3.81 (HÀC(4''')); 3.58 (HÀC(5''')); 3.69 (dd, J(6'''a,6'''b) ˆ 11.8, J(5''',6'''a) ˆ 3.5,
HaÀC(6''')); 3.76 (dd, J(5''',6'''b) ˆ 1.5, HbÀC(6''')); Glc'''': 4.58 (d, J(1'''',2'''') ˆ 7.8, HÀC(1'''')); 3.11 (dd, J(2'''',3
''') ˆ 9.6, HÀC(2'''')); 3.34 (HÀC(3'''')); 3.35 (HÀC(4'''')); 3.42 (HÀC(5'''')); 3.78 (dd, J(6''''a,6''''b) ˆ 12.1,
J(5'''',6''''a) ˆ 7.3, HaÀC(6'''')); 4.05 (dd, J(5'''',6''''b) ˆ 1.5; HbÀC(6'''')); Glc''''': 4.35 (d, J(1''''',2''''') ˆ 7.8,
HÀC(1''''')); 3.19 (HÀC(2''''')); 3.34 (HÀC(3''''')); 3.33 (HÀC(4''''')); 3.26 (HÀC(5''''')); 3.64
(dd, J(6'''''a,6'''''b) ˆ 11.8, J(5''''',6'''''a) ˆ 3.5, HaÀC(6''''')); 3.86 (dd, J(5''''',6'''''b) ˆ 1.5, HbÀC(6''''')).
13C-NMR (CD3OD, 50 MHz; values with the same superscript might be interchangeable): aglycone moiety:
40.58a (C(1)); 27.28 (C(2)); 91.12 (C(3)); 40.58a (C(4)); 57.57 (C(5)); 19.29 (C(6)); 35.88 (C(7)); 41.00a (C(8));
51.12 (C(9)); 37.93 (C(10)); 30.79b (C(11)); 71.70 (C(12)); 49.85 (C(13)); 52.42c (C(14)); 30.76b (C(15)); 27.26
(C(16)); 52.93c (C(17)); 84.98 (C(20)); 22.50 (C(21)); 36.81 (C(22)); 23.90 (C(23)); 126.04 (C(24)); 132.25
(C(25)); 25.94 (C(26)); 18.02 (C(27)); 28.44 (C(28)); 16.33, 16.71, 16.80, 17.37 (C(18, C(19), C(29), C(30)); Glc':
105.36 (C(1')); 81.87 (C(2')); 78.42 (C(3')); 71.70d (C(4')); 77.66 (C(5')); 62.83e (C(6')); Glc'': 104.73 (C(1''));
76.11 (C(2'')); 76.25 (C(3'')); 81.29 (C(4'')); 76.91 (C(5'')); 62.52e (C(6'')); Gal''': 105.25 (C(1''')); 72.56 (C(2'''));
74.83 (C(3''')); 70.31 (C(4''')); 77.12 (C(5''')); 62.52e (C(6''')); Glc'''': 98.12 (C(1'''')); 75.32 (C(2'''')); 77.96
(C(3'''')); 71.70d (C(4'''')); 76.91 (C(5'''')); 62.52e (C(6'''')); Glc''''': 105.01 (C(1''''')); 75.16 (C(2''''')); 78.56
(C(3''''')); 71.70d (C(4''''')); 77.96 (C(5''''')); 62.83e (C(6''''')). LSI-MS: 1269 ([M À H]À), 1107, 945, 783.
(3b,12b)-20-({O-b-d-Galactopyranosyl-(1 ! 4)-O-b-d-glucopyranosyl-(1 ! 6)-O-[b-d-galactopyranosyl-
(1 ! 4)]-b-d-glucopyranosyl}oxy)-12-hydroxydammar-24-en-3-yl 2-O-b-d-Glucopyranosyl-b-d-glucopyrano-
side (ˆ [b-d-Glcp-(1 ! 2)-b-d-Glcp-(1 ! O-3)]-{b-d-Galp-(1 ! 4)-b-d-Glcp-(1 ! 6)-[b-d-Galp-(1 ! 4)]-b-d-
Glcp-(1 ! O-20)}-protopanaxadiol; 1c). Amorphous solid. 1H-NMR (CD3OD/(D5)pyridine, 500 MHz):
aglycone moiety: 0.98, 1.69 (CH2(1)); 1.65, 1.97 (CH2(2)); 3.16 (HÀC(3)); 0.74 (HÀC(5)); 1.46, 1.55
(CH2(6)); 1.26, 1.50 (CH2(7)); 1.41 (HÀC(9)); 1.22, 1.74 (CH2(11)); 3.75 (HÀC(12)); 1.75 (HÀC(13)); 0.99,
1.54 (CH2(15)); 1.29, 1.68 (CH2(16)); 2.28 (HÀC(17)); 0.91 (Me(18)); 0.91 (Me(19)); 1.35 (Me(21)); 1.51, 1.80
(CH2(22)); 2.01, 2.11 (CH2(23)); 5.11 (HÀC(24)); 1.62 (Me(26)); 1.58 (Me(27)); 1.04 (Me(28)); 0.86 (Me(29));
0.83 (Me(30)); Glc': 4.45 (d, J(1',2') ˆ 7.8, HÀC(1')); 3.70 (HÀC(2'), HÀC(3')); 3.42 (HÀC(4')); 3.39
(HÀC(5')); 3.74 (HaÀC(6')); 3.95 (HbÀC(6')); Glc'': 4.77 (d, J(1'',2'') ˆ 7.7, H ÀC(1'')); 3.38 (HÀC(2'')); 3.49
(HÀC(3'')); 3.39 (HÀC(4'')); 3.34 (HÀC(5'')); 3.73 (HaÀC(6'')); 3.91 (HbÀC(6'')); Glc''': 4.64 (d, J(1''',2''') ˆ
7.8, HÀC(1''')); 3.27 (dd, J(2''',3''') ˆ 9.6, HÀC(2''')); 3.58 (HÀC(3''')); 3.83 (HÀC(4''')); 3.61 (HÀC(5''')); 4.02
(HaÀC(6''')); 4.22 (HbÀC(6''')); Gal'''': 4.63 (d, J(1'''',2'''') ˆ 7.8, H ÀC(1'''')); 3.68 (HÀC(2'''')); 3.58
(HÀC(3'''')); 3.90 (HÀC(4'''')); 3.72 (HÀC(5'''')); 3.79 (HaÀC(6)); 3.84 (HbÀC(6'''')); Glc''''': 4.47
(d, J(1,2) ˆ 7.8, HÀC(1''''')); 3.40 (HÀC(2''''')); 3.56 (HÀC(3''''')); 3.59 (HÀC(4''''')); 3.35 (HÀC(5''''')); 3.94
(HaÀC(6''''')); 3.97 (HbÀC(6''''')); Gal'''''': 4.43 (d, J(1'''''',2'''''') ˆ 7.8, HÀC(1'''''')); 3.68 (HÀC(2'''''')); 3.58
(HÀC(3'''''')); 3.90 (HÀC(4'''''')); 3.72 (HÀC(5'''''')); 3.79 (HaÀC(6'''''')); 3.84 (HbÀC(6'''''')). LSI-MS: 1431
([M À H]À), 1269, 1107, 945, 783, 621, 459.
(3b,12b)-20-{[O-b-d-Galactopyranosyl-(1 ! 4)-O-b-glucopyranosyl-(1 ! 6)-b-d-glucopyranosyl]oxy}-12-
hydroxydammar-24-en-3-yl 2-O-b-d-Glucopyranosyl-b-d-glucopyranoside (ˆ [b-d-Glcp-(1 ! 2)-b-d-Glcp-
(1 ! O-3)]-[b-d-Galp-(1 ! 4)-b-d-Glcp-(1 ! 6)-b-d-Glcp-(1 ! O-20)]-protopanaxadiol; 1d). Amorphous sol-
id. M.p. 243 2468 (dec.). 1H-NMR (MeOD, 500 MHz): aglycone moiety: 1.00, 1.72 (CH2(1)); 1.71, 1.98
(CH2(2)); 3.20 (HÀC(3)); 0.78 (HÀC(5)); 1.49, 1.57 (CH2(6)); 1.29, 1.52 (CH2(7)); 1.42 (HÀC(9)); 1.25, 1.79
(CH2(11)); 3.71 (HÀC(12)); 1.72 (HÀC(13)); 1.02, 1.57 (CH2(15)); 1.32, 1.73 (CH2(16)); 2.29 (HÀC(17)); 1.00
(Me(18)); 0.91 (Me(19)); 1.36 (Me(21)); 1.53, 1.80 (CH2(22)); 2.05, 2.12 (CH2(23)); 5.13 (HÀC(24)); 1.68

Helvetica Chimica Acta Vol. 85 (2002)
1955
(Me(26)); 1.62 (Me(27)); 1.06 (Me(28)); 0.91 (Me(29)); 0.85 (Me(30)); Glc': 4.43 (d, J(1',2') ˆ 7.2, HÀC(1'));
3.55 (HÀC(2')); 3.53 (HÀC(3')); 3.26 (HÀC(4')); 3.24 (HÀC(5')); 3.64 (dd, J(6'a,6'b) ˆ 11.9, J(5',6'a) ˆ 5.3,
HaÀC(6')); 3.85 (dd, J(5',6'b) ˆ 1.8, HbÀC(6')); Glc'': 4.66 (d, J(1'',2'') ˆ 7.8, HÀC(1'')); 3.21 (HÀC(2'')); 3.35
(HÀC(3'')); 3.19 (HÀC(4'')); 3.23 (HÀC(5'')); 3.62 (HaÀC(6'')); 3.82 (HbÀC(6'')); Glc''': 4.58 (d, J(1''',2''') ˆ
7.8, HÀC(1''')); 3.11 (dd, J(2''',3''') ˆ 8.3, HÀC(2''')); 3.34 (HÀC(3''')); 3.33 (HÀC(4''')); 3.42 (HÀC(5''')); 3.79
(HaÀC(6''')); 4.04 (dd, J(6'''a,6'''b) ˆ 11.6, J(5''',6'''b) ˆ 1.8, HbÀC(6''')); Glc'''': 4.39 (d, J(1'''',2'''') ˆ 7.8,
HÀC(1'''')); 3.26 (HÀC(2'''')); 3.51 (HÀC(3'''')); 3.54 (HÀC(4'''')); 3.39 (HÀC(5'''')); 3.82
(dd, J(6''''a,6''''b) ˆ 12.2, J(5'''',6''''a) ˆ 3.5, HaÀC(6'''')); 3.89 (dd, J(5'''',6''''b) ˆ 1.5, HbÀC(6'''')); Gal''''': 4.34
(d, J(1''''',2''''') ˆ 7.6, HÀC(1''''')); 3.54 (HÀC(2''''')); 3.47 (HÀC(3''''')); 3.81 (HÀC(4''''')); 3.58 (HÀC(5'''''));
3.70 (dd,J( 6'''''a,6'''''b) ˆ 11.8, J(5''''',6'''''a) ˆ 3.5, HaÀC(6''''')); 3.76 (dd, J(5''''',6'''''b) ˆ 1.5, HbÀC(6''''')).
13C-NMR (CD3OD, 50 MHz; values with the same superscript might be interchangeable): aglycone moiety:
40.59a (C(1)); 27.26 (C(2)); 91.33 (C(3)); 40.27a (C(4)); 57.56 (C(5)); 19.24 (C(6)); 35.86 (C(7)); 40.99a (C(8));
51.11 (C(9)); 37.93 (C(10)); 30.83b (C(11)); 71.94 (C(12)); 49.85 (C(13)); 52.41c (C(14)); 30.83b (C(15)); 27.26
(C(16)); 52.93c (C(17)); 84.99 (C(20)); 22.48 (C(21)); 36.79 (C(22)); 23.90 (C(23)); 126.02 (C(24)); 132.30
(C(25)); 25.98 (C(26)); 18.04 (C(27)); 28.41 (C(28)); 16.33, 16.73 (double), 17.39 (C(18), C(19), C(29), C(30));
Glc': 105.41 (C(1')); 81.14d (C(2')); 78.53 (C(3')); 71.63e (C(4')); 77.71 (C(5')); 62.86f (C(6')); Glc'': 104.54
(C(1'')); 76.33 (C(2'')); 77.90 (C(3'')); 71.94e (C(4'')); 78.34 (C(5'')); 63.13 (C(6')); Glc''': 98.13 (C(1''')); 75.30
(C(2''')); 77.71 (C(3''')); 71.70e (C(4''')); 76.80 (C(5''')); 62.47f (C(6''')); Glc'''': 104.80 (C(1'''')); 75.11 (C(2''''));
77.87 (C(3'''')); 80.84d (C(4'''')); 76.48 (C(5'''')); 62.02f (C(6'''')); Gal''''': 105.19 (C(1''''')); 72.66 (C(2''''')); 74.89
(C(3''''')); 74.87 (C(4''''')); 76.84 (C(5''''')); 62.47f (C(6''''')). LSI-MS: 1269 ([M À H]À), 1107, 945, 783, 621, 459.
(3b,12b)-20-({O-b-d-Galactopyranosyl-(1 ! 4)-O-[b-d-glucopyranosyl-(1 ! 6)]-b-d-glucopyranosyl}oxy)-
12-hydroxydammar-24-en-3-yl 2-O-b-d-Glucopyranosyl-b-d-glucopyranoside (ˆ [b-d-Glcp-(1 ! 2)-b-d-Glcp-
(1 ! O-3)]-{b-d-Galp-(1 ! 4)-[b-d-Glcp-(1 ! 6)]-b-d-Glcp-(1 ! O-20)}-protopanaxadiol (1e). Amorphous
solid. M.p. 245 2508 (dec.). 1H-NMR (CD3OD, 500 MHz): aglycone moiety: 1.00, 1.72 (CH2(1)); 1.71, 1.98
(CH2(2)); 3.20 (HÀC(3)); 0.78 (HÀC(5)); 1.49, 1.57 (CH2(6)); 1.29, 1.52 (CH2(7)); 1.42 (HÀC(9)); 1.25, 1.79
(CH2(11)); 3.71 (HÀC(12)); 1.72 (HÀC(13)); 1.02, 1.57 (CH2(15)); 1.32, 1.73 (CH2(16)); 2.29 (HÀC(17)); 1.00
(Me(18)); 0.91 (Me(19)); 1.36 (Me(21)); 1.53, 1.80 (CH2(22)); 2.05, 2.12 (CH2(23)); 5.13 (HÀC(24)); 1.68
(Me(26)); 1.62 (Me(27)); 1.06 (Me(28)); 0.91 (Me(29)); 0.85 (Me(30)); Glc': 4.43 (d, J(1',2') ˆ 7.8, HÀC(1'));
3.55 (HÀC(2')); 3.54 (HÀC(3')); 3.27 (HÀC(4')); 3.25 (HÀC(5')); 3.65 (dd, J(6'a,6'b) ˆ 11.9, J(5',6'a) ˆ 5.4,
HaÀC(6')); 3.85 (dd, J(5',6'b) ˆ 2.0, HbÀC(6')); Glc'': 4.66 (d, J(1'',2'') ˆ 7.7, H ÀC(1'')); 3.21 (HÀC(2')); 3.35
(HÀC(3'')); 3.20 (HÀC(4'')); 3.24 (HÀC(5'')); 3.61 (HaÀC(6'')); 3.81 (HbÀC(6'')); Glc''': 4.60 (d, J(1''',2''') ˆ
7.8, HÀC(1''')); 3.20 (dd, J(2''',3''') ˆ 9.6, HÀC(2''')); 3.51 (HÀC(3''')); 3.72 (HÀC(4''')); 3.52 (HÀC(5''')); 3.91
(dd, j(6'''a,6'''b) ˆ 12.1, J(5''',6'''a) ˆ 3.7, HaÀC(5''')); 4.16 (dd, J(5''',6'''b) ˆ 1.5; HbÀC(6''')); Gal'''': 4.50
(d, J(1'''',2'''') ˆ 7.8, HÀC(1'''')); 3.52 (dd, J(2'''',3'''') ˆ 8.7, HÀC(2'''')); 3.49 (HÀC(3'''')); 3.80 (HÀC(4''''));
3.61 (HÀC(5'''')); 3.69 (dd, J(6''''a,6''''b) ˆ 12.1, J(5'''',6''''a) ˆ 7.3, HaÀC(6'''')); 3.75 (dd, J(5'''',6''''b) ˆ 1.5,
HbÀC(6'''')); Glc''''': 4.41 (d, J(1''''',2''''') ˆ 7.8, HÀC(1''''')); 3.22 (HÀC(2''''')); 3.34 (HÀC(3''''')); 3.33
(HÀC(4''''')); 3.27 (HÀC(5''''')); 3.63 (dd, J(6'''''a,6'''''b) ˆ 11.8, J(5''''',6'''''a) ˆ 3.5, HaÀC(6''''')); 3.87 (dd,
J(5''''',6'''''b) ˆ 1.5, HbÀC(6''''')). 13C-NMR (CD3OD, 50 MHz; values with the same superscript might be
interchangeable): aglycone moiety: 40.59a (C(1)); 27.26 (C(2)); 91.31 (C(3)); 40.30a (C(4)); 57.56 (C(5)); 19.27
(C(6)); 35.88 (C(7)); 41.00a (C(8)); 51.12 (C(9)); 37.94 (C(10)); 30.76b (C(11)); 71.95 (C(12)); 49.85 (C(13));
52.41c (C(14)); 30.76b (C(15)); 27.26 (C(16)); 52.75c (C(17)); 85.05 (C(20)); 22.37 (C(21)); 36.82 (C(22)); 23.83
(C(23)); 125.98 (C(24)); 132.29 (C(25)); 25.92 (C(26)); 18.03 (C(27)); 28.40 (C(28)); 16.34, 16.69 (double), 17.42
(C(18), C(19), C(29), C(30)); Glc': 105.40 (C(1')); 81.18 (C(2')); 78.54 (C(3')); 71.65d (C(4')); 77.71 (C(5'));
62.86e (C(6')); Glc'': 104.55 (C(1'')); 76.35 (C(2')); 77.87f (C(3'')); 71.95d (C(4'')); 78.34 (C(5'')); 63.14e (C(6''));
Glc''': 98.11 (C(1''')); 70.36 (C(2''')); 77.71f (C(3''')); 80.22 (C(4''')); 78.34 (C(5''')); 68.90 (C(6''')); Gal'''': 104.88
(C(1'''')); 72.66 (C(2'''')); 74.89 (C(3'''')); 70.36 (C(4'''')); 76.84 (C(5'''')); 62.46e (C(6'''')); Glc''''': 104.88
(C(1''''')); 75.11 (C(2'''')); 77.87f (C(3''''')); 71.75d (C(4''''')); 77.87f (C(5''''')); 62.86e (C(6''''')). LSI-MS: 1269
([M À H]À), 1107, 945, 783, 621, 459.
Enzymatic Acylation of Ginsenoside Rb1 (1). Compound 1 (200 mg) was dissolved in a mixture of DMF
¾
(8 ml), THF (16 ml) and vinyl acetate (6 ml). Novozym 435 (400 mg) was added and the suspension shaken at
458. The reaction was monitored by TLC (CHCl3/MeOH/H2O 8 :5 :0.4). After 4 h, the starting material was
almost completely converted to two products. The enzyme was filtered off, the solvent mixture evaporated, and
the crude residue purified by FC (CHCl3/MeOH/H2O 2 :1:0.1): 60 mg of 6'',6''''-di-O-acetylginsenoside Rb1
(1f) and 120 mg of a monoacetate fraction. The monoacetate fraction was recognized to be a 9 :1 mixture of 6''''-
O-acetylginsenoside Rb1 (1g) and 6''-O-acetylginsenoside Rb1 (1h) by anal. HPLC (H2O/MeCN 2 :1): tR

1956
Helvetica Chimica Acta Vol. 85 (2002)
16.46 min for 1g and 14.69 min for 1h. Prep. HPLC allowed the isolation of 46 mg of pure 1g, whereas 1h could
not be recovered as a pure compound.
6'',6''''-Di-O-acetylginsenoside Rb1 (ˆ(3b,12b)-20-{[6-O-(6-O-Acetyl-b-d-glucopyranosyl)-b-d-glucopyra-
nosyl]oxy}-12-hydroxydammar-24-en-3-yl 2-O-(6-O-Acetyl-b-d-glucopyranosyl)-b-d-glucopyranoside; 1f):
Amorphous solid. M.p. 177 1788. 1H-NMR ((D6)DMSO, 600 MHz, 458): aglycone moiety: 0.92, 1.58
(CH2(1)); 1.56, 1.88 (CH2(2)); 3.01 (HÀC(3)); 0.71 (HÀC(5)); 1.36, 1.49 (CH2(6)); 1.21, 1.48 (CH2(7)); 1.31
(HÀC(9)); 1.08, 1.64 (CH2(11)); 3.56 (HÀC(12)); 1.57 (HÀC(13)); 0.92, 1.57 (CH2(15)); 1.47, 1.77 (CH2(16));
2.18 (HÀC(17)); 0.71, 0.80, 0.83, 0.90 (Me(18), Me(19), Me(29), Me(30)); 1.25 (Me(21)); 1.47, 1.72 (CH2(22));
1.92, 2.04 (CH2(23)); 5.09 (HÀC(24)); 1.63 (Me(26)); 1.56 (Me(27)); 0.96 (Me(28)); Glc': 4.27 (d, J(1',2') ˆ 7.8,
HÀC(1')); 3.28 (HÀC(2')); 3.38 (HÀC(3')); ca. 3.11 (HÀC(4'), HÀC(5')); 3.45 (dd, J(6'a,6'b) ˆ 11.8,
J(5',6'a) ˆ 2.1, HaÀC(6')); 3.65 (dd, J(5'',6''b) ˆ 2.7, HbÀC(6')); Glc'': 4.47 (d, J(1'',2'') ˆ 7.8, HÀC(1'')); 3.03
(HÀC(2'')); 3.18 (HÀC(3'')); 3.09 (HÀC(4'')); 3.32 (HÀC(5'')); 3.98 (dd, J(6''a,6''b) ˆ 11.8, J(5'',6''a) ˆ 2.7,
HaÀC(6'')); 4.19 (dd, J(5'',6''b) ˆ 5.3, HbÀC(6'')); 2.02 (1 Ac); Glc''': 4.41 (d, J(1''',2''') ˆ 7.8, HÀC(1''')); 2.89
(dd, J(2''',3''') ˆ 9.6, HÀC(2''')); 3.19 (HÀC(3''')); 3.03 (HÀC(4''')); 3.37 (HÀC(5''')); 3.58 (dd, J(6'''a,6'''b) ˆ
10.5, J(5''',6'''a) ˆ 4.6, HaÀC(6''')); 3.86 (dd, J(5''',6'''b) ˆ 2.1, HbÀC(6''')); Glc'''': 4.24 (d, J(1'''',2'''') ˆ 8.2,
HÀC(1'''')); 2.98 (dd, J(2'''',3'''') ˆ 9.6, HÀC(2'''')); 3.16 (HÀC(3'''')); 3.07 (HÀC(4'''')); 3.32 (HÀC(5'''')); 4.03
(dd, J(6''''a,6''''b) ˆ 10.5, J(5'''',6''''a) ˆ 2.1, HaÀC(6'''')); 4.25 (dd, J(5'''',6''''a) ˆ 4.6, HbÀC(6'''')); 2.00 (1 Ac).
13C-NMR ((D6)DMSO, 125 MHz, 458; values with the same superscript might be interchangeable): aglycone
moiety: 38.56 (C(1)); 25.42 (C(2)); 87.74 (C(3)); 38.56 (C(4)); 55.10 (C(5)); 18.03 (C(6)); 34.58 (C(7)); 39.10
(C(8)); 48.70 (C(9)); 35.98 (C(10)); 28.98 (C(11)); 68.45 (C(12)); 48.41 (C(13)); 50.35 (C(14)); 30.01 (C(15));
25.05 (C(16)); 50.35 (C(17)); 82.25 (C(20)); 35.03 (C(22)); 22.40 (C(23)); 124.56 (C(24)); 130.63 (C(25)); 15.85,
16.05, 16.17, 16.27 (C(18), C(19), C(29), C(30)); 20.98 (C(21)); 24.70 (C(26)); 16.83 (C(27)); 26.63 (C(28));
Glc': 102.51 (C(1')); 87.78 (C(2')); 76.50a (C(3')); 68.96b (C(4')); 75.74c (C(5')); 60.34 (C(6')); Glc'': 103.07
(C(1'')); 74.29 (C(2'')); 75.79c (C(3'')); 68.96b (C(4'')); 72.88c (C(5'')); 62.97 (C(6'')); Glc''': 95.56 (C(1''')); 73.00
(C(2''')); 75.79c (C(3''')); 68.98b (C(4''')); 76.02a (C(5''')); 68.81 (C(6''')); Glc'''': 102.75 (C(1'''')); 72.63d (C(2''''));
75.79c (C(3'''')); 69.02a (C(4'''')); 72.88d (C(5'''')); 62.97 (C(6'''')); 19.95, 20.01 (MeCOO); 170.04, 170.08
(MeCOO). LSI-MS: 1191 ([M À H]À), 1149, 987, 945, 783, 621, 459.
6''''-O-Acetylginsenoside Rb1 (ˆ(3b,12b)-20-{[6-O-(6-O-Acetyl-b-d-glucopyranosyl)-b-d-glucopyranosyl]-
oxy}-12-hydroxydammar-24-en-3-yl 2-O-b-d-Glucopyranosyl-b-d-glucopyranoside; 1g): Amorphous solid. M.p.
189 1908. 1H-NMR ((D6)DMSO, 600 MHz, 458): aglycone moiety: 0.91, 1.59 (CH2(1)); 1.55, 1.88 (CH2(2)); 3.02
(HÀC(3)); 0.70 (HÀC(5)); 1.36, 1.46 (CH2(6)); 1.22, 1.48 (CH2(7)); 1.30 (HÀC(9)); 1.05, 1.64 (CH2(11)); 3.58
(HÀC(12)); 1.55 (HÀC(13)); 0.90, 1.52 (CH2(15)); 1.47, 1.75 (CH2(16)); 2.17 (HÀC(17)); 0.75, 0.81, 0.83, 0.89
(Me(18), Me(19), Me(29), Me(30)); 1.24 (Me(21)); 1.45, 1.74 (CH2(22)); 1.94, 2.01 (CH2(23)); 5.10
(HÀC(24)); 1.62 (Me(26)); 1.56 (Me(27)); 0.99 (Me(28)); Glc': 4.27 (d, J(1',2') ˆ 8.0, HÀC(1')); 3.32
(dd, J(2',3') ˆ 9.3, HÀC(2')); 3.38 (HÀC(3')); 3.10 (HÀC(4'), HÀC(5')); 3.43 (dd, J(6'a,6'b) ˆ 11.9,
J(5',6'a) ˆ 4.8, HaÀC(6')); 3.66 (dd, J(5',6b) ˆ 2.4, HbÀC(6')); Glc'': 4.45 (d, J(1'',2'') ˆ 7.7, H ÀC(1'')); 3.00
(HÀC(2'')); 3.13 (HÀC(3'')); 3.06 (HÀC(4'')); 3.18 (HÀC(5'')); 3.48 (dd, J(6''a,6''b) ˆ 11.8, J(5'',6''a) ˆ 5.9,
HaÀC(6'')); 3.62 (dd, J(5'',6''b) ˆ 2.5, HbÀC(6'')); Glc''': 4.41 (d, J(1''',2''') ˆ 8.2, HÀC(1''')); 2.89
(dd, J(2''',3''') ˆ 9.6, HÀC(2''')); 3.17 (HÀC(3''')); 3.02 (HÀC(4''')); 3.34 (HÀC(5''')); 3.57 (dd, J(6'''a,6'''b) ˆ
12.1, J(5''',6'''a) ˆ 7.3, HaÀC(6''')); 3.86 (dd, J(5''',6'''b) ˆ 2.3, HbÀC(6''')); Glc'''': 4.23 (d, J(1'''',2'''') ˆ 8.2,
HÀC(1'''')); 2.96 (HÀC(2'''')); 3.14 (HÀC(3'''')); 3.06 (HÀC(4'''')); 3.31 (HÀC(5'''')); 4.01 (dd, J(6''''a,6''''b) ˆ
11.5, J(5'''',6''''a) ˆ 7.5, H ÀC(6''''a)); 4.25 (dd, J(5'''',6''''b) ˆ 1.5, HÀC(6''''b)); 2.02 (Ac). 13C-NMR
((D6)DMSO, 150 MHz, 458; values with the same superscript might be interchangeable): aglycone moiety:
38.60 (C(1)); 25.73 (C(2)); 87.93 (C(3)); 38.81 (C(4)); 55.27 (C(5)); 18.68 (C(6)); 34.64 (C(7)); 39.18 (C(8));
48.950 (C(9)); 36.12 (C(10)); 30.08 (C(11)); 68.70 (C(12)); 48.58 (C(13)); 50.61 (C(14)); 30.12 (C(15)); 25.58
(C(16)); 50.64 (C(17)); 82.38 (C(20)); 20.98 (C(21)); 35.34 (C(22)); 22.48 (C(23)); 124.90 (C(24)); 130.42
(C(25)); 25.28 (C(26)); 17.23 (C(27)); 27.12 (C(28)); 16.01, 16.04, 16.18, 16.83 (C(18), C(19), C(29), C(30));
Glc': 102.70 (C(1')); 81.95 (C(2')); 77.01 (C(3')); 70.11a (C(4')); 76.18b (C(5')); 61.25 (C(6')); Glc'': 103.12
(C(1'')); 75.18c (C(2'')); 76.34 (C(3'')); 70.15a (C(4'')); 76.50b (C(5'')); 61.30 (C(6'')); Glc''': 95.34 (C(1''')); 73.58
(C(2''')); 76.25b (C(3''')); 70.09a (C(4''')); 75.23c (C(5''')); 68.75 (C(6''')); Glc'''': 102.73 (C(1'''')); 73.49 (C(2''''));
76.53b (C(3'''')); 70.02a (C(4'''')); 72.57 (C(5'''')); 63.10 (C(6'''')); 20.61 (MeCOO); 170.11 (MeCOO). LSI-MS:
1149 ([M À H]À), 1107, 987, 783, 621, 459.
6''-O-Acetylginsenoside Rb1 (ˆ(3b,12b)-20-[(6-O-b-d-Glucopyranosyl-b-d-glucopyranosyl)oxy]-12-hy-
droxydammar-24-en-3-yl 2-O-(6-O-Acetyl-b-d-glucopyranosyl)-b-d-glucopyranoside; 1h): Selected 1H-NMR
data ((D6)DMSO, 600 MHz, 458): Glc'': 4.47 (d, J(1'',2'') ˆ 8.2, HÀC(1'')); 3.02 (HÀC(2'')); 3.18 (HÀC(3''));
3.09 (HÀC(4'')); 3.30 (HÀC(5'')); 3.98 (dd, J(6''a,6''b) ˆ 11.8, J(5'',6''a) ˆ 5.9, HaÀC(6'')); 4.19 (dd, J(5'',6''b) ˆ

Helvetica Chimica Acta Vol. 85 (2002)
1957
2.5, HbÀC(6'')); Glc''': 4.42 (d, J(1''',2''') ˆ 8.0, HÀC(1''')); 2.89 (dd, J(2''',3''') ˆ 9.6, HÀC(2''')); 3.19
(HÀC(3''')); 3.06 (HÀC(4''')); 3.31 (HÀC(5''')); 3.56 (dd, J(6''',a,6'''b) ˆ 12.1, J(5''',6'''a) ˆ 7.3, HaÀC(6'''));
3.90 (dd, J(5''',6'''b) ˆ 2.3, HbÀC(6''')); 2.02 (Ac).
Ginsenoside Rg3 (ˆ(3b,12b,20R)- and (3b,12b,20S)-12,20-Dihydroxydammar-24-en-3-yl 2-O-b-d-gluco-
pyranosyl-b-d-glucopyranoside ˆ [b-d-Glcp-(1 ! 2)-b-d-Glcp-(1pO-3)]-protopanaxadiol; 3). Ginsenoside
Rb1 (1; 1 g) was dissolved in 40% AcOH/H2O (v/v; 40 ml), and the mixture was stirred for 2 h at 708. H2O
(50 ml) was added, the mixture extracted 4 times with AcOEt, and the org. phase dried (Na2SO4) and
evaporated. FC (AcOEt/MeOH/H2O 9 :2 :0.5) yielded 466 mg (65%) of 3 as a mixture of (20R) and (20S)
epimers. Amorphous solid. M.p. 146 1478. 1H-NMR ((D6)DMSO, 300 MHz, 808): selected data of the aglycone
moiety: 3.08 (HÀC(3)); 3.42 (HÀC(12)); 5.12, 5.14 (HÀC(24)); Glc': 4.31 (d, J(1',2') ˆ 8.0, HÀC(1')); 3.26
(t, J(2',3') ˆ 8.0, HÀC(2')); 3.40 (t, J(3',4') ˆ 8.0, HÀC(3')); 3.12 3.18 (HÀC(4'), HÀC(5')); 3.42
(dd, J(6'a,6'b) ˆ 11.8, J(5',6'a) ˆ 3.5, HaÀC(6')); 3.67 (dd, J(5',6'b) ˆ 1.5, HbÀC(6')); Glc'': 4.52 (d, J(1'',2'') ˆ
8.0, HÀC(1'')); 3.06 (t, J(2'',3'') ˆ 8.0, HÀC(2'')); 3.12 3.18 (HÀC(3''), HÀC(4''), HÀC(5')); 3.52
(dd, J(6''a,6''b) ˆ 11.8, J(5'',6''a) ˆ 2.7, HaÀC(6'')); 3.68 (dd, J(5'',6''b) ˆ 1.5, HbÀC(6'')). 13C-NMR ((D6)DMSO,
300 MHz, 808; values with the same superscript might be interchangeable): aglycone moiety: 38.87 (C(1)); 27.43
(C(2)); 88.41 (C(3)); 38.56a (C(4)); 55.83 (C(5)); 17.80 (C(6)); 34.57 (C(7)); 39.10a (C(8)); 49.59 (C(9)); 36.43
(C(10)); 31.01 (C(11)); 68.87 (C(12)); 47.67, 48.20 (C(13)); 51.04 (C(14)); 31.05, 31.06 (C(15)); 25.55, 26.48
(C(16)); 49.62, 53.41 (C(17)); 70.17, 72.11 (C(20)); 20.98 (C(21)); 34.90 (C(22)); 21.56 (C(23)); 125.26, 125.47
(C(24)); 130.05 (C(25)); 24.70 (C(26)); 16.83 (C(27)); 26.63 (C(28)); 15.85, 16.05, 16.17, 16.27 (C(18), C(19),
C(29), C(30)); Glc': 103.45 (C(1')); 81.02 (C(2')); 76.58 (C(3')); 70.36 (C(4')); 76.23 (C(5')); 61.31 (C(6')); Glc'':
103.45 (C(1'')); 74.95 (C(2'')); 76.60 (C(3'')); 70.38 (C(4'')); 76.23 (C(5'')); 61.31 (C(6'')). LSI-MS: 783
([M À H]À), 621, 459.
Enzymatic Acetylation of Ginsenoside Rg3 (3): 6''-O-Acetylginsenoside Rg3 (ˆ(3b,12b,20R)- and
(3b,12b,20S)-12,20-Dihydroxydammar-24-en-3-yl 2-O-(6-O-Acetyl-b-d-glucopyranosyl)-b-d-glucopyranoside;
3a). Compound 3 (200 mg) was dissolved in a mixture of DMF (5 ml), THF (15 ml), and vinyl acetate
¾
(5 ml). Novozym 435 (400 mg) was added and the suspension shaken at 458. After 72 h, 3 was almost
completely converted to one product (TLC monitoring (CHCl3/MeOH/H2O 6 :2 :0.3)). The enzyme was
filtered off, the solvent mixture evaporated, and the crude residue purified by FC (CHCl3/MeOH/H2O
2 :0.5 :0.1): 137 mg of 3a. Amorphous solid. M.p. 218 2198. 1H-NMR ((D6)DMSO, 600 MHz, 808): selected
data of the aglycone moiety: 3.02 (HÀC(3)); 3.38 (HÀC(12)); 5.06, 5.08 (HÀC(24)); Glc': 4.27 (d, J(1',2') ˆ 7.8,
HÀC(1')); 3.26 (t, J(2',3') ˆ 7.8, HÀC(2')); 3.37 (t, J(3',4') ˆ 7.8, HÀC(3')); 3.09 (HÀC(4'), HÀC(5')); 3.41
(dd, J(6'a,6'b) ˆ 11.5, J(5',6'a) ˆ 5.5, HaÀC(6')); 3.63 (dd, J(5',6'b) ˆ 1.0, HbÀC(6')); Glc'': 4.47 (d, J(1'',2'') ˆ 8.0,
HÀC(1'')); 3.01 (t, J(2'',3'') ˆ 8.0, HÀC(2'')); 3.17 (t, J(3'',4'') ˆ 8.0, HÀC(3'')); 3.09 (HÀC(4'')); 3.30
(ddd, J(4'',5'') ˆ 7.8, J(5'',6''a) ˆ 5.5, J(5'',6''b) ˆ 1.0, HÀC(5'')); 3.97 (dd, J(6''a,6''b) ˆ 11.5, HaÀC(6'')); 4.16
(dd, HbÀC(6'')); 2.00 (Ac). 13C-NMR ((D6)DMSO, 600 MHz, 808; values with the same superscript might be
interchangeable): aglycone moiety: 38.91 (C(1)); 27.43 (C(2)); 88.30 (C(3)); 38.56a (C(4)); 55.82 (C(5)); 17.80
(C(6)); 34.58 (C(7)); 39.10a (C(8)); 49.62 (C(9)); 36.47 (C(10)); 31.71 (C(11)); 69.89 (C(12)); 47.69, 48.27
(C(13)); 51.05 (C(14)); 31.05, 31.06 (C(15)); 25.85, 26.68 (C(16)); 50.02, 53.01 (C(17)); 70.57, 72.01 (C(20));
21.78 (C(21)); 34.75 (C(22)); 21.38 (C(23)); 125.22, 125.46 (C(24)); 130.05 (C(25)); 24.58 (C(26)); 16.03 (C(27));
27.13 (C(28)); 15.65, 16.15, 16.20, 16.38 (C(18), C(19), C(29), C(30)); Glc': 103.40 (C(1')); 82.01 (C(2')); 76.08
(C(3')); 70.17 (C(4')); 76.38 (C(5')); 61.30 (C(6')); Glc'': 103.88 (C(1'')); 74.94 (C(2'')); 76.80 (C(3'')); 69.89
(C(4'')); 73.86 (C(5'')); 63.74 (C(6'')); 20.57 (MeCOO); 170.20 (MeCOO). LSI-MS: 825 ([M À H]À), 783, 621,
459.
Ginsenoside Rh2 (ˆ(3b,12b,20R)- and (3b,12b,20S)-12,20-Dihydroxydammar-24-en-3-yl-b-d-Glucopyra-
noside; 4). A soln. of b-galactosidase (100 mg) in phosphate buffer (pH 4.5; 36 ml) was added dropwise into
MeOH (10 ml) containing 3 (450 mg). The mixture was shaken for 2 h at 378, then more b-galactosidase was
added. After 24 h, the mixture was extracted with 3 Â 20 ml AcOEt, the org. phase evaporated, and the residue
purified by FC (AcOEt/MeOH/H2O 9 :0.7:0.2): 129 mg (36%) of 4 as a mixture of (20R) and (20S) epimers.
Amorphous solid. M.p. 139 1448. 1H-NMR (CD3OD, 500 MHz): aglycone moiety: 1.02, 1.72 (CH2(1)); 1.71,
1.98 (CH2(2)); 3.21 (HÀC(3)); 0.78 HÀC(5)); 1.48, 1.56 (CH2(6)); 1.30, 1.51 (CH2(7)); 1.46 (HÀC(9)); 1.24,
1.83 (CH2(11)); 3.83 (HÀC(12)); 1.74 (HÀC(13)); 1.02, 1.56 (CH2(15)); 1.32, 1.74 (CH2(16)); 2.29 (HÀC(17));
0.85, 0.91, 0.92, 1.00 (Me(18), Me(19), Me(29), Me(30)); 1.29 (Me(21)); 1.37, 1.55 (CH2(22)); 1.99, 2.14
(CH2(23)); 5.13 (HÀC(24)); 1.66 (Me(26)); 1.61 (Me(27)); 1.07 (Me(28)); Glc': 4.31 (d, J(1',2') ˆ 7.8,
HÀC(1')); 3.18 (dd, J(2',3') ˆ 8.9, HÀC(2')); 3.33 (dd, J(3',4') ˆ 8.6, HÀC(3')); 3.28 (HÀC(4')); 3.25
(HÀC(5')); 3.65 (dd, J(6'a,6'b) ˆ 11.9, J(5',6'a) ˆ 5.2, HaÀC(6')); 3.83 (dd, J(5',6'b) ˆ 2.0, HbÀC(6')).
13C-NMR (CD3OD, 50 MHz; values with the same superscript might be interchangeable): aglycone moiety:

1958
Helvetica Chimica Acta Vol. 85 (2002)
40.63a (C(1)); 27.36 (C(2)); 90.62 (C(3)); 40.23a (C(4)); 57.56 (C(5)); 19.26 (C(6)); 35.95 (C(7)); 40.98a (C(8));
51.38 (C(9)); 37.97 (C(10)); 32.03 (C(11)); 71.69 (C(12)); 49.85 (C(13)); 52.59 (C(14)); 32.03 (C(15)); 27.19
(C(16)); 55.12 (C(17)); 74.40 (C(20)); 28.43 (C(21)); 35.95 (C(22)); 23.30 (C(23)); 126.19 (C(24)); 131.97
(C(25)); 25.90 (C(26)); 17.71 (C(27)); 28.43 (C(28)); 16.18, 16.81 (double), 17.15 (C(18), C(19), C(29), C(30));
Glc': 106.71 (C(1')); 72.14 (C(2')); 78.29 (C(3')); 71.69 (C(4')); 77.69 (C(5')); 62.84 (C(6')). LSI-MS: 621 ([M À
H]À), 459.
Enzymatic Acetylation of Ginsenoside Rh2 (4): 6'-O-Acetylginsenoside Rh2 (ˆ(3b,12b,20R)- and
(3b,12b,20S)-12,20-Dihydroxydammar-24-en-3-yl 6-O-Acetyl-b-d-Glucopyranoside; 4a). To a soln. of
4
¾
(3 mg) in THF (0.5 ml), vinyl acetate (0.1 ml) Novozym 435 (10 mg) were added. The mixture was shaken
72 h at 428 (TLC: (AcOEt/MeOH/H2O 9 :0.7:0.2): nearly 100% conversion to a single product). After filtration
of the enzyme, the mixture was evaporated: 4a as a mixture of (20R)-and (20 S)-epimers. Amorphous solid.
1H-NMR (CD3OD, 500 MHz): aglycone moiety: 1.02, 1.73 (CH2(1)); 1.71, 1.98 (CH2(2)); 3.22 (HÀC(3)); 0.78
(HÀC(5)); 1.48, 1.56 (CH2(6)); 1.30, 1.51 (CH2(7)); 1.46 (HÀC(9)); 1.24, 1.83 (CH2(11)); 3.83 (HÀC(12)); 1.74
(HÀC(13)); 1.02, 1.56 (CH2(15)); 1.32, 1.73 (CH2(16)); 2.29 (HÀC(17)); 0.84, 0.91, 0.92, 1.00 (Me(18), Me(19),
Me(29), Me(30)); 1.28 (Me(21)); 1.37, 1.55 (CH2(22)); 1.99, 2.14 (CH2(23)); 5.13 (HÀC(24)); 1.67 (Me(26));
1.61 (Me(27)); 1.07 (Me(28)); Glc': 4.30 (d, J(1',2') ˆ 7.8, HÀC(1')); 3.18 (dd, J(2',3') ˆ 8.9, HÀC(2')); 3.31
(dd, J(3',4') ˆ 8.6, HÀC(3')); 3.24 (HÀC(4')); 3.42 (HÀC(5')); 4.21 (HaÀC(6')); 4.33 (HbÀC(6')); 1.96 (Ac).
13C-NMR (CD3OD, 50 MHz; values with the same superscript might be interchangeable): aglycone moiety:
40.33a (C(1)); 27.17 (C(2)); 90.90 (C(3)); 41.06a (C(4)); 57.59 (C(5)); 19.27 (C(6)); 35.94 (C(7)); 40.32a (C(8));
51.30 (C(9)); 37.99 (C(10)); 30.90b (C(11)); 71.89 (C(12)); 49.86 (C(13)); 52.60 (C(14)); 32.04b (C(15)); 27.17
(C(16)); 55.14 (C(17)); 74.41 (C(20)); 28.40 (C(21)); 35.95 (C(22)); 23.30 (C(23)); 126.19 (C(24)); 131.98
(C(25)); 25.87 (C(26)); 17.69 (C(27)); 28.40 (C(28)); 16.18, 16.78 (double), 17.13 (C(18), C(19), C(29), C(30));
Glc': 106.74 (C(1')); 75.56 (C(2')); 78.14 (C(3')); 71.88 (C(4')); 77.69 (C(5')); 62.84 (C(6')); 20.72 (MeCOO);
172.68 (MeCOO). LSI-MS: 663 ([M À H]À), 621, 459.
We thank Prof. Enzio Ragg for recording NMR spectra on a Bruker AMX-600 instrument, Dr. Monica
Luisetti for her valuable technical assistence, and Dr. Ezio Bombardelli for a generous gift of Korean ginseng
extract. This work was supported financially by C.N.R. (Centro di Studio sulle Sostanze Organiche Naturali) and
DAAD (VIGONI program).
REFERENCES
[1] B. Danieli, L. Falcone, D. Monti, S. Riva, S. Gebhardt, M. Schubert-Zsilavecz, J. Org. Chem. 2001, 66, 262.
[2] G. M. Cragg, D. J. Newman, K. M. Snader, J. Nat. Prod. 1997, 60, 52.
[3] C. Walsh, Nature (London) 2001, 409, 226, and ref. cit. therein.
[4] G. Carrea, G. Ottolina, P. Pasta, S. Riva, Ann. N. Y. Acad. Sci. 1996, 799, 642.
[5] L. Husakova, S. Riva, M. Casali, S. Nicotra, M. Kuzma, Z. Hunkova, V. Kren, Carbohydr. Res. 2001, 331,
143; L. Weignerova, P. Sedmera, Z. Hunkova, P. Halada, V. Kren, M. Casali, S. Riva, Tetrahedron Lett.
1999, 40, 9297.
[6] S. Riva, D. Monti, M. Luisetti, B. Danieli, Ann. N. Y. Acad. Sci. 1998, 864, 70.
[7] G. Carrea, S. Riva, Angew. Chem., Int. Ed. 2000, 39, 2227; G. Roda, S. Riva, in −Methods in Non-Aqueous
Enzymology×, Ed. M. N. Gupta, Birkh‰user Verlag, Basel, 2000, p. 146; B. Danieli, G. Lesma, M. Luisetti, S.
Riva, Tetrahedron 1997, 53, 5855.
[8] O. Tanaka, R. Kasai, in −Progress in the Chemistry of Organic Natural Products×, Ed. W. Herz, H.
Griesebach, G. W. Kirby, and C. Tamm, Springer Verlag, Vienna-New York, 1984, Vol. 46, pp. 1 76.
[9] M. Amado, R. Almeida, T. Schwientek, H. Clausen, Biochim. Biophys. Acta 1999, 1473, 35.
[10] B. Danieli, M. Luisetti, M. Schubert-Zsilavecz, W. Likussar, S. Steurer, S. Riva, D. Monti, J. Reiner, Helv.
Chim. Acta 1997, 80, 1153.
[11] S. Riva, B. Sennino, F. Zambianchi, B. Danieli, L. Panza, Carbohydr. Res. 1998, 305, 525.
[12] M. Abdel Tawab, U. Bahr, B. Danieli, S. Gebhardt, M. Karas, S. Riva, M. Schubert-Zsilavecz, Helv. Chim.
Acta 2000, 83, 739.
[13] B. Danieli, M. Luisetti, G. Sampognaro, G. Carrea, S. Riva, J. Mol. Catal. B: Enzym. 1997, 3, 193.
[14] B. Danieli, M. Luisetti, S. Riva, A. Bertinotti, E. Ragg, L. Scaglioni, E. Bombardelli, J. Org. Chem. 1995, 60,
3637.

Helvetica Chimica Acta Vol. 85 (2002)
1959
[15] B. Danieli, M. Luisetti, S. Steurer, A. Michelitsch, W. Likussar, S. Riva, J. Reiner, M. Schubert-Zsilavecz, J.
Nat. Prod. 1999, 62, 670.
[16] H. Besso, R. Kasai, J. Wei, J.-F. Wang, Y. Saruwatari, T. Fuwa, O. Tanaka, Chem. Pharm. Bull. 1982, 30 (12),
4534.
[17] J. Uppenberg, M. T. Hansen, S. Patkar, T. A. Jones, Structure 1994, 2, 293.
[18] J. Uppenberg, N. Ohrner, M. Norin, K. Hult, S. Patkar, V. Waagen, T. Anthonsen, T. A. Jones, Biochemistry
1995, 34, 16838.
[19] S. Yahara, K. Kaji, O. Tanaka, Chem. Pharm. Bull. 1079, 27, 88; I. Kitagawa, T. Tanuyama, M. Yoshikawa,
Y. Ikenishi, Y. Nagakawa, Chem. Pharm. Bull. 1989, 37, 2961.
[20] S. Okada, K. Kishi, H. Higashihara, J. Fukumoto, J. Agric. Chem. Soc. Jpn. 1963, 37, 84.
[21] I. Sakamoto, H. Kohda, K. Murakami, O. Tanaka, Yakugazu Zasshi 1975, 95, 1507.
¡
[22] D. Monti, E. Giosue, S. Riva, L. Panza, Gazz. Chim. Ital. 1996, 126, 303.
[23] D. K. Fitzgerald, B. Colvin, R. Mawal, K. E. Ebner, Anal. Biochem. 1970, 36, 43.
[24] I. A. Macdonald, J. Org. Chem. 1977, 136, 348.
Received February 19, 2002