Synthesis of bisdesmosidic kryptogenyl saponins using the 'random glycosylation' strategy and evaluation of their antitumor activity

Article abstract of DOI:10.1016/j.bmcl.2006.09.071

A bisdesmosidic steroidal saponins library, composed of 16 novel kryptogenin glycosides, was set up via six random glycosylation procedures, wherein two compounds showed their antitumor activity against HeLa cell in the preliminary pharmacological researc

Full text of DOI:10.1016/j.bmcl.2006.09.071

Bioorganic & Medicinal Chemistry Letters 17 (2007) 156–160
Synthesis of bisdesmosidic kryptogenyl saponins using the ‘random
glycosylation’ strategy and evaluation of their antitumor activity
Yang Liu,a Dong-Mei Zhao,a Xue-Hua Lu,a Hui Wang,a Hong Chen,b Ying Ke,b
Ling Lengb and Mao-Sheng Chenga,*
aKey Lab of New Drugs Design and Discovery of Liaoning Province, School of Pharmaceutical Engineering,
Shenyang Pharmaceutical University, Shenyang 110016, PR China
bStaff Room of Pharmacognosy, Medical College of Chinese People’s Armed Police Force, Tianjin 300162, PR China
Received 9 July 2006; revised 20 September 2006; accepted 22 September 2006
Available online 10 October 2006
Abstract—A bisdesmosidic steroidal saponins library, composed of 16 novel kryptogenin glycosides, was set up via six random
glycosylation procedures, wherein two compounds showed their antitumor activity against HeLa cell in the preliminary pharmaco-
logical research.
Ó 2006 Elsevier Ltd. All rights reserved.
Steroidal saponins comprise a diverse class of plant gly-
cosides, which possess a broad range of promising bio-
logical property, such as antitumor activity.1,2 As an
important access to find novel steroidal glycoside struc-
tures or to scale up some interesting natural products
which are scarce in nature, chemical synthesis plays a
dominant role and receives increasing attention. Howev-
er, particular protecting group manipulation is often re-
quired to accomplish the regioselective glycosylation of
a certain multi-hydroxyl aglycone to produce bisdes-
mosidic or tridesmosidic saponins.3 Therefore, it is usu-
ally time-consuming for the lengthy course of stepwise
glycosylation work.
promising developments toward the same goal.6
Although the lack of regioselectivity in the two
approaches makes the target individual compound diffi-
cult to be isolated from the library,7 the ‘randomization’
strategy can accelerate the rate of finding new lead in the
initial phase of drug discovery.
To the best of our knowledge, the random glycosylation
strategy was mainly used in the reactions where sugar
donor glycosylated randomly with other unprotected su-
gar moiety.5 We, herein report the work of building a
bisdesmosidic steroidal saponins library, containing 16
diverse kryptogenin glycosides, by using different mono-
saccharide donors to glycosylate with kryptogenin ran-
domly. The preliminary antitumor activity of these
synthetic saponins was then investigated.
As an alternative approach, ‘random glycosylation’
4
strategy developed by Hindsgaul and co-workers was
proved to be effective for the rapid construction of oligo-
saccharide libraries. In the successful application of this
method, completely unprotected saccharide receptor
was glycosylated to generate a mixture of glycosides
containing all the possible products for further bioactiv-
ity screening research.5 In addition, a process centered
upon the inherent promiscuity of secondary metabo-
lite-associated glycosyltransferases is one of the latest
Kryptogenin 1, namely 3b,26-dihydroxy-25(R)-cholest-
5-en-16,22-dione, was used to be an intermediate in
the research of metabolism of bile acid in last century.8
As a natural sapogenin, 1 owns a basic chemical struc-
ture of cholesterol, which ensures its reliability of being
the mimetics in the biosynthesis research. Furthermore,
by far researches on cholestanol glycosides are not as
much as those on spirostanol or furostanol saponins,
either in synthetic method or in bioactivity evaluation.
Hence, we chose 1 as the aglycone to investigate the bio-
activity of the target bisdesmosidic cholestanol saponins
bearing different sugar moiety on both side chains. Since
1 is no longer commercially available and the reported
Keywords: Steroidal saponins; Kryptogenin glycosides; Random gly-
cosylation; Antitumor activity.
*
Corresponding author. Tel.: +86 24 23986413; fax: +86 24
23995043; e-mail: mscheng@syphu.edu.cn
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.09.071

Y. Liu et al. / Bioorg. Med. Chem. Lett. 17 (2007) 156–160
157
Scheme 1. Preparation of kryptogenin 1. Reagents and conditions: (a) BzCl, pyridine, CH2Cl2, rt, 12 h; (b) oxone, NaHCO3, acetone, H2O, CH2Cl2,
rt, 12 h, 90%; (c) Zn, KI, Ac2O, AcOH, 50 °C, 12 h, 66%; and (d) MeONa, MeOH, CH2Cl2, rt, 2 h, 98%.
preparation methods require harsh conditions,9 we
developed a new route to convert diosgenin 2 into 1 fac-
ilely based on our former work (Scheme 1).10 The
hydroxyl group at C-3 of 2 was first protected by BzCl
in dry pyridine—CH2Cl2 to give 3 quantificationally.
Under the stepwise oxidation with oxone–acetone sys-
tem in CH2Cl2–H2O using NaHCO3 as a buffer for con-
trolling pH ꢀ 7.5, 3 was converted to 5,6-epoxy-16a-ol
hemiketal 4 in 12 h with high yield of 90%. The reduc-
tion and the acetylation with Zn, KI and Ac2O–AcOH
in one pot transformed 4 into 16,22-dione 5 in a yield
of 66% after purification through silica gel column chro-
matography. Finally, the O-acetyl and the O-benzoyl
groups of 5 were removed by MeONa–MeOH in
CH2Cl2 to afford 1. The overall yield for the four steps
from 2 to 1 was 58%.
Using 1 as the aglycone, we attempted to attach different
sugar moieties to its hydroxyl groups at C-3 and C-26.
According to the traditionally sequential chemical meth-
ods, at least two steps (protection and de-protection of
either hydroxyl group) must be introduced additionally
to furnish the target bisdesmosidic saponins bearing var-
ious sugars on both sides. Furthermore, the ‘one by one’
synthetic manner makes the compound library difficult
to build. Alternatively, random glycosylation of 1 with
two sugar donors, after removal of the protecting
groups on the sugar moieties, should generate a mixture
of four bisdesmosidic saponins, which can be separated
completely via chromatography technique. Consequen-
tially, the latter strategy should show more superiority
in saving time and labor.
˚
Scheme 2. An example of random glycosylation. Reagents and conditions: (a) TMSOTf, CH2Cl2, 4 A MS, 0 °C to rt, 2 h; and (b) MeONa, MeOH,
rt, 2 h.

158
Y. Liu et al. / Bioorg. Med. Chem. Lett. 17 (2007) 156–160
A random glycosylation procedure was performed as
shown in Scheme 2. To a magnetically stirred solution
of 1 and two glycosyl trichloroacetimidates 6b and 6d
(0.8 equiv each) in redistilled CH2Cl2 under Nitrogen
As far as bioactivity evaluation is concerned, an individ-
ual sample can present a result more exactly and directly
than the mixture does. Therefore, we try to separate the
four compounds in the sublibrary for further studies.
Primary attempt using the column chromatography end-
ed in failure because of incomplete separation of the two
isomeric compounds 7h and 7n. Afterwards the recycling
preparative HPLC was applied to achieve collecting
every component clean and without loss.12 The chemical
structures of the products were characterized via ESI-
˚
at 0 °C, was added the powder of freshly activated 4 A
molecular sieves. After stirring for 15 min, a solution
of TMSOTf (0.1 equiv) in CH2Cl2 was added dropwise
to promote the reaction. The cooling bath was removed
30 min later and the stirring continued for 1.5 h until all
the glycosyl donors were consumed, when Et3N was
added to quench the reaction. The molecular sieves were
filtered off and the solvent was evaporated to give an oily
residue, which was treated on a flash silica gel column
chromatography with a step gradient to remove the
unreacted aglycone and minor monodesmosidic chemi-
cals.11 The collected bisdesmosidic saponins fractions
then were treated with a solution of MeONa in MeOH
and the mixture was stirred for 2 h before being neutral-
ized with Dowex 50 (H+) and filtered. The filtrate was
concentrated in vacuo and then anhydrous Et2O was
added to result in a white precipitate of the target mol-
ecules (7f, 7h, 7n, and 7p) as a sublibrary.
1
MS, H and 13C NMR, and the isomeric compounds
were identified by HMBC confirming the attached posi-
tion of the sugar residue.13
By this means, we chose four benzoyl-protected glycosyl
trichloroacetimidates (6a–d, Fig. 1), which were all read-
ily prepared according to the reported methods,14 to ap-
ply to the glycosylation with 1 in pairs.15 Thus, in every
entry only two chemical steps were required to produce
four bisdesmosidic saponins. Consequently 16 novel
kryptogenyl saponins (7a–p, Fig. 1) in high purity were
Figure 1. Four sugar donors (6a–d) chosen for random glycosylation and the library of the synthetic bisdesmosidic saponins (7a–p).
Table 1. Structures and physical data of the synthetic kryptogenyl saponinsa ,b
Compound
R1
R2
Formula
ESI-MS (M+Na)+
Mp (°C)
[a]D (MeOH)
7a
7b
7c
7d
7e
7f
a-L-Arap-
a-L-Arap-
a-L-Arap-
a-L-Arap-
b-D-Glcp-
b-D-Glcp-
b-D-Glcp-
b-D-Glcp-
a-L-Rhap-
a-L-Rhap-
a-L-Rhap-
a-L-Rhap-
b-D-Xylp-
b-D-Xylp-
b-D-Xylp-
b-D-Xylp-
a-L-Arap-
b-D-Glcp-
a-L-Rhap-
b-D-Xylp-
a-L-Arap-
b-D-Glcp-
a-L-Rhap-
b-D-Xylp-
a-L-Arap-
b-D-Glcp-
a-L-Rhap-
b-D-Xylp-
a-L-Arap-
b-D-Glcp-
a-L-Rhap-
b-D-Xylp-
C37H58O12
C38H60O13
C38H60O12
C37H58O12
C38H60O13
C39H62O14
C39H62O13
C38H60O13
C38H60O12
C39H62O13
C39H62O12
C38H60O12
C37H58O12
C38H60O13
C38H60O12
C37H58O12
717.3
747.6
731.6
717.6
747.6
777.7
761.6
747.6
731.6
761.7
745.7
731.7
717.6
747.6
708.6
717.6
237.7–239.0
244.3–245.5
236.5–238.3
231.4–233.6
249.0–250.7
260.5–262.2
255.8–257.6
254.9–256.1
244.8–246.7
257.5–259.0
234.3–235.8
232.0–233.8
232.8–234.0
245.9–247.0
239.3–240.6
227.2–229.1
ꢁ93.8 (c 0.16)
ꢁ110.7 (c 0.10)
ꢁ118.8 (c 0.15)
ꢁ114.7 (c 0.10)
ꢁ115.1 (c 0.13)
ꢁ119.6 (c 0.05)
ꢁ124.7 (c 0.07)
ꢁ120.0 (c 0.07)
ꢁ153.3 (c 0.08)
ꢁ150.0 (c 0.05)
ꢁ153.5 (c 0.05)
ꢁ152.7 (c 0.05)
ꢁ117.7 (c 0.05)
ꢁ119.5 (c 0.08)
ꢁ121.9 (c 0.07)
ꢁ119.2 (c 0.05)
7g
7h
7i
7j
7k
7l
7m
7n
7o
7p
a
¨
ESI-MS spectra were recorded with an Agilent 1100 mass spectrometer. Melting points were determined on a BUCHI Melting Point B-540. Optical
rotations were measured at room temperature with a Perkin-Elmer 241 MC polarimeter at the sodium D line.
b a-L-Arap-, a-L-arabinopyranosyl; b-D-Glcp-, b-D-glucopyranosyl; a-L-Rhap-, a-L-rhamnopyranosyl; b-D-Xylp-, b-D-xylopyranosyl.

Y. Liu et al. / Bioorg. Med. Chem. Lett. 17 (2007) 156–160
Table 2. Antitumor activity of the synthetic compounds against HeLa
tumor cell line
159
References and notes
1. Hostettmann, K.; Marston, A. Saponins; Cambridge
University Press: Cambridge, 1995; pp. 288–304.
2. Lasztity, R.; Hidvegi, M.; Bata, A. Food Rev. Int. 1998, 14,
371.
Compound
Inhibition rate at different
concentration (%)
IC50(lM)
10ꢁ5 M 10ꢁ6 M 10ꢁ7
M
10ꢁ8 M
1
22.34
18.47
62.02
9.71
17.29
11.03
4.34
13.80
12.05
1.64
9.49
8.33
0.26
3.25
6.36
0.98
0.87
0.46
3.42
7.95
1.35
0.16
7.19
13.64
0.68
17.47
15.84
6.00
>10
>10
4.25
>10
>10
>10
>10
>10
>10
>10
8.36
>10
>10
>10
>10
>10
>10
>10
3. For some examples reported, see (a) Hou, S.; Xu, P.;
Zhou, L.; Yu, D.; Lei, P. Bioorg. Med. Chem. Lett. 2006,
16, 2454; Wang, P.; Li, C.; Zang, J.; Song, N.; Zhang, X.;
Li, Y. Carbohydr. Res. 2005, 340, 2086; Zhang, Y.; Li, Y.;
Zhu, S.; Guan, H.; Lin, F.; Yu, B. Carbohydr. Res. 2004,
339, 1753; Suhr, R.; Lahmann, M.; Oscarson, S.; Thiem, J.
Eur. J. Org. Chem. 2003, 20, 4003.
4. Kanie, O.; Barresi, F.; Ding, Y.; Labbe, J.; Otter, A.;
Forsberg, L. S.; Ernst, B.; Hindsgaul, O. Angew. Chem.,
Int. Ed. Engl. 1995, 34, 2720.
5. Ding, Y.; Kanie, O.; Labbe, J.; Palcic, M. M.; Ernst, B.;
Hindsgaul, O. Adv. Exp. Med. Biol. 1995, 376, 261;
Ding, Y.; Labbe, J.; Kanie, O.; Hindsgdul, O. Bioorg.
Med. Chem. 1996, 4, 683; Boons, G. J.; Heskamp, B.;
Hour, F. Angew. Chem., Int. Ed. 1996, 35, 2845; Izumi,
M.; Ichikawa, Y. Tetrahedron Lett. 1998, 39, 2079; Yu,
B.; Li, B.; Xing, G.; Hui, Y. J. Comb. Chem. 2001, 3,
404.
5
7a
7b
7c
7d
7e
7f
7g
7h
7i
4.79
9.25
3.84
7.51
10.33
11.77
4.36
2.84
3.67
1.79
1.05
3.46
2.31
9.94
1.78
7.16
10.48
24.78
60.13
15.13
26.18
28.47
30.95
30.86
28.14
17.59
14.48
6.34
1.67
12.76
6.93
1.66
7j
7k
7l
21.19
20.70
10.64
22.97
26.30
16.27
14.28
20.06
5.45
7m
7n
7o
7p
19.26
22.29
14.23
6. Fu, X.; Albermann, C.; Zhang, C.; Thorson, J. S. Org.
Lett. 2004, 7, 1513; Yang, J.; Fu, X.; Jia, Q.; Shen, J.;
Biggins, J. B.; Jiang, J.; Zhao, J.; Schmidt, J. J.; Wang, P.
G.; Thorson, J. S. Org. Lett. 2003, 5, 2223.
7. The mixture generated via random glycosylation usually
composed of a series of isomeric compounds, which made
the isolation of individual chemicals extremely difficult.
Moreover, the deconvolution of the library was also a
complicated course if the active compounds needed to be
identified.
8. Shoda, J.; Axelson, M.; Sjovall, J. Steroids 1993, 58, 119;
Javitt, N. B.; Kok, E.; Lloyd, J.; Benscath, A.; Field, F. H.
Biomed. Mass Spectrom. 1982, 9, 61; Mui, M. M.; Kamat,
S. Y.; Elliott, W. H. Steroids 1974, 24, 239; Noll, B. W.;
Doisy, E. A.; Elliott, W. H. J. Lipid Res. 1973, 14, 391.
9. Alessandrini, L.; Ciuffreda, P.; Santaniello, E.; Terraneo,
G. Steroids 2004, 69, 789; Kessar, S. V.; Gupta, Y. P.;
Mahajan, R. K.; Joshi, G. S.; Rampal, A. L. Tetrahedron
1968, 24, 899.
generated through total six manipulations, whose struc-
tures are shown in Table 1.
The antitumor activity in vitro of these synthetic sapo-
nins against HeLa tumor cells was investigated by the
standard MTT assay. As shown in Table 2, only sapo-
nins 7a and 7i presented definite inhibitions against
HeLa cells below the concentration of 10 lM. However,
other 14 saponins, the ester form of kryptogenin 5, and
the aglycone 1 itself showed relatively lower effect at the
same condition. This result indicates that some sugar
moieties are important for inducing the antitumor activ-
ity of cholestanol saponins. Nevertheless, there is still a
large amount of work to do for probing such effective
sugar residues.
10. Cheng, M. S.; Wang, Q. L.; Tian, Q.; Song, H. Y.; Liu, Y.
X.; Li, Q.; Xu, X.; Miao, H. D.; Yao, X. S.; Yang, Z. J.
Org. Chem. 2003, 68, 3658.
In summary, a library of 16 novel bisdesmosidic steroi-
dal saponins was built for the evaluation of antitumor
activity. Applied to random glycosylation strategy, two
sugar donors and kryptogenin successfully generated
all four possible bisdesmosidic glycosides in a simpler
and faster manner. This idea should be useful for the
facile preparation of other multi-O-glycosyl saponins li-
brary. Since biological ligands are generally two to four
sugars in size, oligosaccharide donors need to be intro-
duced to steroidal aglycone randomly to prepare a
potentially useful library for bioactivity test and struc-
ture–activity relationship (SAR) research. Further stud-
ies in this area are in progress and will be reported in due
courses.
11. Petroleum ether–EtOAc (6:1, v/v) was used to collect the
fractions including all the possible bisdesmosidic glyco-
sides. Afterwards minor monodesmosidic saponins were
eluted by petroleum ether–EtOAc (3:1, v/v), followed by
the fraction of starting kryptogenin with a recovery of
10%.
12. A JAI-LC 9103 recycling preparative HPLC equipped
with an ODS reverse-phase column was applied to isolate
the components with MeOH–H2O (80:20, v/v) at a flow
rate of 9.0 ml/min. The relative abundance ratio of each
component was 27:23:18:31 (7f:7h:7n:7p).
13. NMR spectral data were listed as following: for 7h: 1H
NMR (600 MHz, pyridine-d5, ppm): d 5.28 (br s, 1H,
H-6), 5.06 (d, J = 7.7 Hz, 1H, H-10), 4.72 (d, J = 7.8 Hz,
1H, H-100), 4.59–4.57 (d, J = 11.5, 1H, CH2-60-1), 4.44–
4.42 (m, 1H, CH2-60-2), 4.38–4.35 (dd, J = 11.2, 5.2, 1H,
CH2-500-1), 4.31–4.29 (m, 2H, H-30, H-400), 4.24–4.23 (m,
1H, H-40), 4.18–4.15 (t, J = 8.6, 1H, H-300), 4.07 (m, 1H,
H-20), 4.02–4.00 (m, 2H, H-3, H-50), 3.96–3.93 (m, 2H, H-
200, CH2-26-1), 3.74–3.70 (t, J = 10.5, CH2-500-2), 3.65–3.63
(dd, J = 14.9, 4.0, CH2-26-2), 2.92–2.91 (m, 1H, H-20),
2.79–2.77 (m, 3H, H-17, CH2-4), 2.68–2.66 (m, 1H, CH2-
15-1), 2.45 (m, 1H, CH2-23-1), 2.14–2.12 (m, 1H, CH2-15-
Acknowledgment
The authors thank National Natural Science Founda-
tion of China (No. 20472054) for financial support of
this research.

160
Y. Liu et al. / Bioorg. Med. Chem. Lett. 17 (2007) 156–160
2), 2.05–2.01 (m, 3H, CH2-2-1, CH2-24-1, H-25), 1.86 (br
1.02-1.01 (d, J = 6.5 Hz, 3H, CH3-21), 0.95–0.93 (m, 2H,
CH2-1-2, H-9), 0.92 (s, 3H, CH3-19), 0.67 (s, 3H, CH3-18);
13C NMR (150 MHz, pyridine-d5, ppm): d 217.7 (C-16),
213.0 (C-22), 141.1 (C-5), 121.4 (C-6), 104.9 (C-100), 103.5
(C-10), 78.7 (C-300), 78.6 (C-500), 78.5 (C-30), 78.2 (C-3),
75.3 (C-200), 75.2 (C-26), 75.1 (C-20), 71.7 (C-400), 71.3
(C-40), 67.3 (C-50), 66.4 (C-17), 62.9 (C-600), 51.1 (C-14),
50.0 (C-9), 43.7 (C-20), 41.7 (C-13), 40.2 (C-15), 39.3
(C-4), 38.6 (C-12), 37.4 (C-23), 37.3 (C-1), 37.0 (C-10),
33.5 (C-25), 31.9 (C-2), 30.9 (C-7), 30.4 (C-8), 27.8 (C-24),
20.7 (C-11), 19.4 (C-19), 17.5 (C-27), 15.6 (C-18), 12.9
(C-21). In the HMBC spectrum of 7h, long range
correlations were observed between H-1 (d 5.06) of glucose
and C-3 (d 78.0) of the aglycone, and H-1 (d 4.72) of
xylose and C-26 (d 75.0) of the aglycone. While the HMBC
spectrum of 7n displayed long rang correlations between
H-1 (d 4.86) of xylose and C-3 (d 78.2) of the aglycone,
and H-1 (d 4.93) of glucose and C-26 (d 75.2) of the
aglycone. The similar coupling constants (J = 7.5–7.8 Hz)
s, 1H, CH2-2-2), 1.76–1.62 (m, 5H, CH2-7-1, CH2-1-1,
CH2-12-1, CH2-23-2, CH2-24-2), 1.45–1.43 (m, 2H, CH2-
11), 1.41–1.32 (m, 4H, H-8, CH2-7-2, H-14, CH2-12-2),
1.05–1.04 (d, J = 6.6 Hz, 3H, CH3-27), 1.02–1.01 (d,
J = 6.8 Hz, 3H, CH3-21), 0.96–0.93 (m, 2H, CH2-1-2, H-
9), 0.91 (s, 3H, CH3-19), 0.67 (s, 3H, CH3-18); 13C NMR
(150 MHz, pyridine-d5, ppm): d 217.7 (C-16), 212.9 (C-22),
141.1 (C-5), 121.4 (C-6), 105.4 (C-10), 102.6 (C-100), 78.7
(C-30), 78.6 (C-50), 78.5 (C-300), 78.0 (C-3), 75.4 (C-20), 75.0
(C-26), 74.8 (C-200), 71.7 (C-40), 71.2 (C-400), 67.3 (C-500),
66.4 (C-17), 62.9 (C-60), 51.1 (C-14), 49.9 (C-9), 43.7
(C-20), 41.7 (C-13), 40.1 (C-15), 39.3 (C-4), 38.6 (C-12),
37.3 (C-23), 37.1 (C-1), 37.0 (C-10), 33.5 (C-25), 31.9
(C-2), 30.9 (C-7), 30.2 (C-8), 27.7 (C-24), 20.7 (C-11), 19.4
(C-19), 17.5 (C-27), 15.6 (C-18), 12.9 (C-21); for 7n: 1H
NMR (600 MHz, pyridine-d5, ppm): d 5.31 (br s, 1H, H-
6), 4.93 (d, J = 7.5 Hz, 1H, H-100), 4.86 (d, J = 7.8 Hz, 1H,
H-10), 4.58–4.56 (d, J = 11.5, 1H, CH2-600-1), 4.42–4.38 (m,
2H, CH2-600-2, CH2-50-1), 4.29–4.25 (m, 3H, H-300, H-40,
H-400), 4.22-4.19 (t, J = 8.8, 1H, H-30), 4.07–4.02 (m, 2H,
H-200, H-3), 3.99–3.96 (m, 2H, H-500, H-20), 3.90–3.87 (m,
1H, CH2-26-1), 3.77–3.74 (t, J = 10.6, CH2-50-2), 3.62-3.60
(dd, J = 15.0, 3.8, CH2-26-2), 2.92–2.91 (m, 1H, H-20),
2.76–2.66 (m, 4H, H-17, CH2-4, CH2-15-1), 2.45 (m, 1H,
CH2-23-1), 2.14–2.12 (m, 1H, CH2-15-2), 2.06–1.96 (m,
3H, CH2-2-1, CH2-24-1, H-25), 1.85 (br s, 1H, CH2-2-2),
1.76–1.63 (m, 5H, CH2-7-1, CH2-1-1, CH2-12-1, CH2-23-
2, CH2-24-2), 1.47–1.34 (m, 6H, CH2-11, H-8, CH2-7-2,
H-14, CH2-12-2), 1.05–1.04 (d, J = 6.0 Hz, 3H, CH3-27),
1
of the anomeric protons signals in the H NMR spectra
indicated the b-configuration of anomeric carbons of
sugar units.
14. Schmidt, R. R. Modern Methods in Carbohydrate
Synthesis; Harwood Academic: Netherlands, 1996; pp.
20-s-54.
15. We also attempted to use three glycosyl donors in one
random reaction. Unfortunately, the corresponding sub-
library generated was too complex to be analyzed and
separated. So far, at best three components are suitable in
our research.