Structure-activity relationships of 3-O-β-chacotriosyl oleanic acid derivatives as entry inhibitors for highly pathogenic H5N1 influenza virus

Article abstract of DOI:10.1016/j.bmc.2017.06.025

Highly pathogenic H5N1 virus (H5N1) entry is a key target for the development of novel anti-influenza agents with new mechanisms of action. In our continuing efforts to identify novel potential anti-H5N1 entry inhibitors, a series of 3-O-β-chacotriosyl oleanolic acid analogs have been designed, synthesized and evaluated as H5N1 entry inhibitors based on two small molecule inhibitors 1 and 2 previously discovered by us. The anti-H5N1 entry activities were determined based on HA/HIV and VSVG/HIV entry assays. Compound 15 displayed the most promising anti-H5N1 entry activities with average IC₅₀ values of 4.05?μM and good selective index (22.9). Detailed structure-activity relationships (SARs) studies suggested that either the introduction of an additional oxo group to position 11 at OA or alteration of the C-3 configuration of OA from 3β- to 3α-forms can significantly enhance the selective index while maintaining their antiviral activities in vitro. Molecular simulation analysis confirmed that the compounds exert their inhibitory activity through binding tightly to hemagglutinin (HA2) protein near the fusion peptide and prevent virus entry.

Full text of DOI:10.1016/j.bmc.2017.06.025

Accepted Manuscript
Structure-activity relationships of 3-O-β-chacotriosyl oleanic acid derivatives
as entry inhibitors for highly pathogenic H5N1 influenza virus
Sumei Li, Xiuhua Jia, Xintian Shen, Zhuwen Wei, Zhiyan Jiang, Yixian Liao,
Yiming Guo, Xiaojun Zheng, Guohua Zhong, Gaopeng Song
PII:
S0968-0896(17)30720-4
DOI:
http://dx.doi.org/10.1016/j.bmc.2017.06.025
BMC 13808
Reference:
Bioorganic & Medicinal Chemistry
To appear in:
Received Date:
Revised Date:
Accepted Date:
4 April 2017
5 June 2017
13 June 2017
Please cite this article as: Li, S., Jia, X., Shen, X., Wei, Z., Jiang, Z., Liao, Y., Guo, Y., Zheng, X., Zhong, G., Song,
G., Structure-activity relationships of 3-O-β-chacotriosyl oleanic acid derivatives as entry inhibitors for highly
pathogenic H5N1 influenza virus, Bioorganic & Medicinal Chemistry (2017), doi: http://dx.doi.org/10.1016/j.bmc.
2017.06.025
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Structure-activity relationships of 3-O-β-chacotriosyl oleanic acid derivatives as
entry inhibitors for highly pathogenic H5N1 influenza virus
Sumei Lia,1, Xiuhua Jiac, 1, Xintian Shen d, Zhuwen Weib, Zhiyan Jiange, Yixian Liaob,
Yiming Guob, Xiaojun Zhengb, Guohua Zhonge*, Gaopeng Songb*
aDepartment of human anatomy, School of Medicine, Jinan University, Guangzhou, 510632,
China
bCollege of Materials and Energy, South China Agricultural University, Guangzhou, 510642,
China
cDepartment of Ophthalmology, The Third Affiliated Hospital, Sun Yat-sen University, Guangzhou,
510630, China
dDepartment of Pharmacy, Maternal and Child Health Hospital of Bao’an District, Shenzhen,
518133, China
eKey Laboratory of Natural Pesticide and Chemical Biology, Ministry of Education, and Lab of
Insect Toxicology, South China Agricultural University, Guangzhou, 510642, China
Abstract:
Highly pathogenic H5N1 virus (H5N1) entry is a key target for the development of novel
anti-influenza agents with new mechanisms of action. In our continuing efforts to identify novel
potential anti-H5N1 entry inhibitors, a series of 3-O-β-chacotriosyl oleanolic acid analogs have
been designed, synthesized and evaluated as H5N1 entry inhibitors based on two small molecule
inhibitors 1 and 2 previously discovered by us. The anti-H5N1 entry activities were determined
based on HA/HIV and VSVG/HIV entry assays. Compound 15 displayed the most promising
anti-H5N1 entry activities with average IC50 values of 4.05 µM and good selective index (22.9).
Detailed structure-activity relationships (SARs) studies suggested that either the introduction of an
additional oxo group to position 11 at OA or alteration of the C-3 configuration of OA from 3β- to
3α-forms can significantly enhance the selective index while maintaining their antiviral activities
in vitro. Molecular simulation analysis confirmed that the compounds exert their inhibitory
activity through binding tightly to hemagglutinin (HA2) protein near the fusion peptide and
prevent virus entry.
Keywords: 3-O-β-chacotriosyl saponins; H5N1 entry inhibitors; synthesis; structure-activity
relationships
1. Introduction
Highly pathogenic H5N1 virus has arisen unexpectedly to cause world-wide
* Corresponding author. Tel.: + 86-20-85280308; fax: + 86-20-85280308
E-mail address: guohuazhong@scau.edu.cn (G. Zhong)
Tel.: + 86-20-85280293; fax: + 86-20-85280292
E-mail address: vinsin1021@126.com(G. Song).
1These authors contributed equally to this work.
1

pandemics with markedly increased morbidity and mortality, which is maybe the most
severe pandemic threat in the near future.1-2 At present, only two classes of antiviral
drugs, M2 ion channel blockers and Neuraminidase (NA) inhibitors, are licensed for
the prevention and treatment of influenza virus, but their utility has been limited by
side effects and the rapid emergence of drug-resistant viral strains.3-5 Therefore, it is
the development of novel anti-influenza agents with new mechanisms of action that
represents a crucial defense strategy to combat seasonal and pandemic influenza
strains.
Virus entry inhibitors as a new class of antiviral agents have received considerable
attention by many synthetic chemists and biochemists, which can block effectively
virus entry into the target cells.6-8 Hemagglutinin (HA), the envelope protein of
influenza A virus, is the key protein mediating influenza virus entry.9-10 Thus, HA
represents a favorable target for drug discovery. Up to now, some H5N1 influenza
virus entry inhibitors have been discovered, which have become more and more
exciting research areas.11-14
Pentacyclic triterpenes are one of the most promising sources of novel
therapeutically active agents among the compounds of secondary metabolism
produced by plants. In several special studies, it has been demonstrated that
pentacyclic triterpenes show several antiviral activities against several viruses such as
HIV, IAV, and HCV.15-17 It is noted that some pentacyclic triterpenes such as
glycyrrhizin (GL) and compound Y3, an OA-acetyl galactose conjugate have
displayed good inhibitory activity against influenza A virus entry in vitro.18 Therefore,
2

pentacyclic triterpenes are interesting scaffolds for design of novel virus entry
inhibitors.
In our previous work, we used an efficient HIV-based pseudovirus system 19-20 to
screen a series of 3-O-β-chacotriosyl oleanolic acid analogs generated from
semisynthesis, and discovered the new two small molecule H5N1 viral entry
inhibitors
1
and
2
(Fig.
1),
which
contained
a
a
2,4-di-O-(α-L-rhamnopyranosyl)-β-D-glucopyranose trisaccharide known as
chacotriose. Compounds 1 and 2 exhibited strong inhibitory activity against H5N1
entry with IC50 of 4.74 and 4.82 µM, of which selectivity index was 3.6 and 13.4,
respectively.20 Further SARs studies indicate that introduction of fit bulk groups or a
disubstituted amide structure at the 17-COOH of oleanolic acid (OA) could enhance
selectivity index while keeping inhibitory activity against H5N1 entry.20 Notably, the
incorporation of benzyl group at the 17-COOH of OA (2) was more helpful to reduce
cytotoxicity against MDCK cells than that of disubstituted amide structure. 20
Based on the above results, in present communications, compounds 1 and 2 were
chosen as the lead compounds to design and synthesize a series of analogs to further
investigate the SARs of pentacyclic triterpenes with different sugar chains and
aglycones. For understanding the correlation between the sugar types, sugar
configuration, sugar chain length, as well as the sugar linkage and inhibitory activities,
we designed and synthesized compounds 3-12 (Fig. 1). In addition, compounds 13 -
16 were derived in order to judge the effect of either the C-3 configuration of benzyl
oleanolate as aglycone or the incorporation of a carbonyl group at the C-11 position of
3

benzyl oleanolate on the inhibitory activity. Mechanistic studies indicated that these
pentacyclic triterpenes could bind tightly to hemagglutinin (HA2) protein near the
fusion peptide. This study may establish the importance of pentacyclic triterpenes as
new lead compounds for the development of potential entry inhibitors of H5N1
influenza virus.
Fig. 1. Saponins designed for SARs studies
2. Results and discussion
2.1. Chemistry
The preparation of the known benzyl oleanolate 17 was followed our previous
procedure.20 Benzyl-3-epioleanolic acid 1821 was easily prepared by oxidation of 17
with pyridinium chlorochromate (PCC), followed via Meerwein-Ponndorf reduction.
As depicted in Scheme 1, benzyl esterification of 3-O-acetyl-11-oxo-18β-oleanolic
4

acid ether 1922 was done by the similar route as that for compound 17 and then was
hydrolyzed with NaOH in MeOH/THF to give 3β-O-11-oxo-benzyl oleanolate 20.
Scheme 1 Reagents and conditions: (a) (i) BnCl, K2CO3, DMF; (ii) 4 N NaOH, THF-MeOH
The synthetic route to the title compounds 3-8, 13 and 15 is outlined in Scheme 2.
Glycosylation of benzyl oleanolate 17 with 2,3,4,6-tetra-O-benzoyl-D-glucopyranosyl
trichloroacetimidate
2120
or
2,3,4,6-tetra-O-benzoyl-D-galactopyranosyl
trichloroacetimidate 2223 under the action of TMSOTf furnished the
3-O-β-glucopyranoside 2320 and 3-O-β-galactopyranoside 2423, respectively. Similarly,
the 3-O-β-glucopyranosides 25-26 were prepared through glycosylation of benzyl
3-epioleanolate 18 or 3β-O-11-oxo-benzyl oleanolate 20 with 21, respectively.
Hydrogenolysis of the benzoyl groups in 23-26 with NaOMe in MeOH yielded the
20
known compound 27 and 28-30 in good yields, respectively. Treatment of 27-30
with pivaloyl chloride in pyridine at - 10 °C protected selectively the 3, 6-OHs of the
β-glucopyranosyl or β-galactopyranosyl residues, leading to intermediates 31-34.
Subsequent glycosylation of the 2, 4-OHs in 31-34 with different monosaccharidal
trichloroacetimidates
including
2,3,4-tri-O-acetyl-L-fucopyranosyl
trichloroacetimidate 3524, 2,3,4-tri-O-acetyl-D-rhamnopyranosyl trichloroacetimidate
3624,
2,3,4-tri-O-acetyl-L-arabinosyl
trichloroacetimidate
3724,
5

2,3,4-tri-O-acetyl-D-arabinosyl
trichloroacetimidate
trichloroacetimidate
3825,
2,3,4-tri-O-acetyl-D-xylopyranosyl
3925
and
2,3,4-tri-O-acetyl-L-rhamnopyranosyl trichloroacetimidate 4020 under the “inverse
addition conditions” with trimethylsilyl trifluoromethanesulfonate (TMSOTf) as the
promoter, followed by deprotection of the acyl groups with NaOH, afforded title
saponins 3-8, 13 and 15, respectively. Hydrogenolysis of the 28-O-benzyl group in 8,
13 and 15 over palladium/carbon, followed by esterification with iodomethane (MeI)
in the presence of K2CO3 furnished the target saponins 9, 14 and 16, respectively.
Scheme 2 Reagents and conditions: (a) TMSOTf, CH2Cl2, 0 ºC; (b) MeONa, MeOH; (c) PivCl,
pyridine, - 10 ºC; (d) (i) TMSOTf, CH2Cl2, - 40 ºC; (ii) NaOH, THF-MeOH; (e) (i) 10% Pd/C, H2,
MeOH; (ii) MeI, K2CO3, DMF;
6

The synthetic route toward target saponins 10-12 is depicted in Scheme 3. Methyl
esterification of the known saponins 4120 and 4223 was done by the similar route as
that for compound 9 to provide the target compounds 10 and 11, respectively.
According to our previous strategy26, the synthesis of the target saponin 12 was
carried out employing the “one-pot sequential glycosylation” with the combined use
of 4-methyl-phenyl thioglycoside and trichloroacetimidate donors. Glycosylation of
thioglycoside 4326 and imidate 4426 was completed within 30 min with the use of
catalytic amount of TMSOTf at - 78 °C, furnishing the desired disaccharide
thioglycoside donor. Without purification, the reaction mixture was warmed to 0 °C,
and then saponin acceptor 4520 was added, followed by addition of
N-Iodosuccinimide (NIS) and silver trifluoromethanesulfonate (AgOTf), affording the
desired saponin 46. Final removal of the benzoyl groups in 46 with MeONa in MeOH
gave desired compound 12.
Scheme 3 Reagents and conditions: (a) MeI, K2CO3, DMF; (b) (i) TMSOTf, CH2Cl2, - 78 ºC; (ii)
NIS/AgOTf, CH2Cl2, 0 ºC; (c) MeONa, MeOH;
7

2.2. Inhibition of the infection of H5N1 pseudovirus
Due to the safety concerns in studying viral H5N1 pathogens, the single-cycle
pseudovirus was used instead of live H5N1 avian influenza virus to evaluate the
inhibitory activities of compounds 1-16 against H5N1 entry. A HA/HIV pseudotyped
virus model was carried out to test saponins on H5N1 viral entry inhibitory effects,
while VSVG/HIV pseudoviral transduction was used as a specificity control to
exclude inhibitory effect on post-entry for HIV infection.11-12 It is noted that only a
compound exhibited inhibitory activity on HA/HIV transduction, but not on
VSVG/HIV can be considered as an inhibitor of influenza viral entry. Our preceding
study has shown that 3-O-β-chacotriosyl oleanolic methyl ether (1) has displayed
good inhibition against H5N1, and is specific to hemagglutinin since it has no
inhibitory activity toward VSV-G pseudovirus.20
Compounds 1-16 were evaluated for their capability to inhibit the H5N1 (Influenza
strain A/ Thailand/ Kan353/2004) replication in MDCK cells based on an efficient
HIV-based pseudotyping system with luciferase report element established by us.11, 20
HA targeted compound CL-385319 was used as the reference drug. This pseudovirus
was enveloped with influenza glycoprotein HA and NA, and packed with
HIV-backbone.11 The results, expressed as IC50 (compound concentration required to
achieve 50% inhibition of replication of H5N1), CC50 (compound concentration
required to cause 50% death of uninfected MDCK cells) and SI (selectivity index), are
summarized in Table 1, respectively. As shown in Table 1, all the compounds except
compounds 4-7, 10 and 12 exhibited good inhibitory activity against H5N1
8

pseudovirus in the IC50 range of 4.0-20 µM. Notably, compound 15 showed stronger
activity and better selectivity when compared to the parent compound 1 (Table 1).
Table 1. ICa50 values (µM) of lead compounds 1-16 and CL-385319 screened in MDCK cells
infected with H5N1 pseudovirus and CCb cytotoxicity values (µM) of lead compounds 1-16
50
against MDCK cells (data derived from the mean of three independent assays)
compound
ICa50
4.74 0.14
CCb
SIc
50
16.85 0.51
23.50 0.88
27.35 0.42
52.40 1.22
26.42 0.60
48.25 0.56
20.82 0.48
82.35 1.35
45.62 0.68
22.35 0.52
23.55 0.72
25.32 0.28
>200
3.6
4.9
3.2
<1
1
4.82 0.42
8.65 0.83
>50
2
3
4
>50
<1
5
>50
<1
6
>50
<1
7
9.36 0.82
11.10 0.56
> 50
8.8
4.1
<1
8
9
10
16.28 0.35
> 50
1.4
<1
11
12
5.88 0.26
5.06 0.50
4.05 0.33
4.10 0.44
4.45 1.25
>34
>40
22.9
8.3
332.3
13
>200
14
92.70 1.52
33.83 0.58
1480 10
15
16
CL-385319
All data represent average values for three separate experiments. The variation of these results
under the standard procedures is below 10%.
a
IC50: compound concentration required to achieve 50% inhibition of replication of H5N1, as
determined by plaque reduction assays. Values shown are the mean SEM of at least three
experiments.
b
CC50: compound concentration required to cause 50% death of uninfected MDCK cells, as
9

determined by the MTT method. Values shown are the mean SEM of at least three experiments.c
SI: selectivity index as CC50/IC50
.
Representative compounds 3, 13 and 15 had no effect on VSV-G enveloped
pseudovirus (Fig. 2A), which was similar to both compound 1 and the reference drug
CL-38531911, suggesting that these compounds did not inhibit HIV-backbone and
luciferase reporting activity. In addition, compounds 3, 13 and 15 at 20 µM•L-1 did
not inhibit NA protein (Fig. 2B), which indicated that NA protein might not be the
primary target of 3-O-β-chacotriosyl oleanolic acid derivatives. Taken together, we
anticipated that all the active compounds might interfere with the entry of H5N1
influenza virus by targeting hemagglutinin, the only other glycoprotein enveloped in
the pseudovirus system.
Fig. 2 representative compounds did not inhibit VSVG pseudovirus (A) and neuraminidase
activity (B). Values shown are the mean SEM of at least three experiments.
2.3 SAR analysis of compounds 1-16
Our previous results showed that compounds 1 and 2 exhibited the good inhibitory
activity against the entry of H5N1 influenza virus but neither the trisaccharide residue
chacotriose nor oleanolic acid as the aglycone of compounds 1 and 2 showed H5N1
entry inhibitory activity. These results indicated that the active saponin 1 acted as an
10

integral structure but not the sugar chain or the aglycone alone.
In order to further understand the role of the chacotriosyl residue, structure-activity
relationships (SARs) exploration first focused on the influence of some other
trisaccharide residue instead of chacotriose on inhibitory activity. Table 1 showed that
substituents at C-2 and C-4 hydroxy groups of β-D-glucose exerted important effects
on the antiviral potency. It was observed that replacement of α-L-rhamnose on
chacotriose by α-L-fucose (compound 3) resulted in slight decreases in potency but
significantly decreasing cytotoxic activity against Madin-Darby canine kidney
(MDCK) cells. However, in the case of compound 4, changing L-rhamnopyranosyl
1
moiety with the normal C4 conformation at the C-2 and C-4 position of D-glucose
4
into D-rhamnopyranosyl moiety with the abnormal C1 conformation led to a
complete loss of potency. The same tendency was also observed when the two
L-rhamnosyl residues on chacotriose were replaced by other two pentoses with the
normal 4C1 conformation such as L-arabinosyls, D-arabinosyls and D-xylopyranosyls.
1
Structurally, L-rhamnose is a representative deoxyhexose with the C4 conformation
4
as L-fucose do, which is different from D-rhamnose and pentoses with the C1
conformation. We suppose that during the interaction of chacotriosyl residue with the
receptor, the conformation of chacotriose had an important effect on composition of
1
forces and deoxyhexose with the C4 conformation at the C-2 and C-4 position of
D-glucose was able to enhance the binding affinity of pentacyclic triterpenes to the
protein targets. In addition, conversion of β-D-glucopyranosyl moiety on chacotriose
of 1 or 2 to β-D-galactopyranosyl moiety resulted in a slight loss of potency but
11

enhanced
selectivity
index
(1
vs
9;
2
vs
8).
On a molecular level, D-galactose is structurally identical to D-glucose, except for an
inversion of the hydroxyl and hydrogen groups at the fourth carbon atom, indicating
that some certain modification of C4-OH of β-D-glucose was tolerated and helpful to
improve selectivity index.
The sugar linkage seemed to have a significant impact on potency. Adding
α-L-rhamnosyl moiety at the C2-OH and C3-OH of β-D-glucose obtained compound
10, which did not exhibit desirable anti-H5N1 activity. These results demonstrated
that the glycosylation site of hydroxyl groups of β-D-glucose is an important factor
for inhibitory activity against H5N1 entry. In addition, simplifying the β-chacotriosyl
moiety of compound 1 into α-L-rhamnopyranosyl-(1→2)-β-D-glucopyranosyl
(compound 11) or α-L-rhamnopyranosyl-(1→4)-β-D-glucopyranosyl moiety
(compound 12) resulted in slight decrease in activity against H5N1 or a complete loss
of potency, respectively, suggesting that α-L-rhamnose moiety linked to C2-OH of the
first monosaccharide β-D-glucose was critical to keep anti-H5N1 activity.
The effects of configuration of hydroxyl groups at C-3 of oleanolic acid were
further examined. When 3β hydroxyl group of oleanolic acid (1) was converted to 3α
hydroxyl group, the derived compound 14 displayed similar inhibitory activity but
lower cytotoxicity to MDCK cells than the 3β-forms 1. These results suggest that
pentacyclic triterpenes saponins execute inhibitory activity against H5N1 and
cytotoxic activity against MDCK cells in different mechanisms, which encourages to
develop pentacyclic triterpenes saponins into potent anti-H5N1 agents devoid of the
12

detrimental effect of cytotoxicity. Compared to compounds 1 and 2, the 11-oxo
derivatives (15 and 16) demonstrated more potent inhibitory activity against H5N1
and lower toxicity against MDCK lines, confirming that introduction of an additional
oxo group to position 11 at OA was beneficial to enhance anti-H5N1 activity and
selectivity index.
Compound 15 had better selectivity indexes and comparable potency in the
corresponding derivatives, indicating compound 15 is exploitable as a new lead
compound. Thus, compound 15 was chosen to investigate its mechanism of action
against H5N1.
2.4 Compound 15 does not block H5 HA adsorption to chicken RBCs
It is well known that the HA can absorb to chicken RBCs (red blood cells) leading
to hemagglutination through the interaction of the receptor-binding domain in HA1
subunit with the sialic acid receptor in the RBC membrane. Therefore, we performed
a hemagglutination inhibition assay to investigate whether compound 15 was able to
inhibit hemagglutination by interfering with the H5 HA adsorption to RBCs. It was
found that compound 15 could not inhibit the adsorption of H5 HA to chicken RBCs
even at the highest concentration of 20 µM, while the positive control antisera against
H5 hemagglutinin (Anti-H5) could effectively inhibit H5 HA adsorption to chicken
RBCs with a titer at 1:16 (Fig. 3). These results confirmed that compound 15 may not
target the receptor binding domain of the HA1. Taken together, we postulated that
3-O-β-chacotriosyl oleanolic acid derivatives could target to the membranes fusion
step medicated by the more conserved hemagglutinin transmembrane subunit HA2,
13

but not the receptor-binding domain of HA1.
Fig 3. Compound 15 did not inhibit hemagglutinination induced by H5 hemagglutinin
2.5 Binding kinetics by surface plasmon resonance (SPR)
Since HA-mediated fusion is mainly mediated by rearrangement of HA2 subunit at
low pH, we then used SPR experiment to study the kinetics and quantify the
interaction of compounds 13 and 15 with HA and also HA2 subunit. Compounds 13
and 15 could potently interact with HA protein in a dose-dependent manner, of which
the equilibrium dissociation constant (KD) of the interaction is 4.43 × 10-7 M and 6.17
-8
× 10 M, respectively (Fig 4A and Fig 4B). It was found that the stronger inhibitor
15 had better effect KD than 13, suggesting that the potency of the compound against
the influenza virus H5N1 was directly correlated to the inhibition of HA protein.
Further study revealed that compounds 13 and 15 potently interacted with HA2
subunit in a dose-dependent manner, of which the KD of the interaction is 3.86 × 10-9
M and 5.53 × 10-10 M, respectively (Fig 4C and Fig 4D). Compounds 13 and 15 did
not interact with negative control protein FKBP12 (FK506 Binding Protein 1A,
12kDa), indicated that interactions of HA2 with compounds 13 and 15 were specific
(data not shown). The affinity of these compounds to HA2 protein was correlated well
14

to their antiviral effects, supporting that HA2 protein was a potential target for
compounds 13 and 15.
Fig. 4 Interaction of compounds with HA or HA2 subunit detected by SPR. A (13 with HA), B
(15 with HA), C (13 with HA2) , D (15 with HA2), respectively.
2.6 Molecular modeling for compound 15 binding to H5 hemagglutinin
It is well known that molecular docking is a frequently used tool in computer-aided
structure-based rational drug design. In the previous work, we found that the HA2
sequence of A/Thailand/Kan353/2004 was 100% identical to that of
A/Vietnam/1194/2004, of which the neutral-pH crystal structure of the HA has been
well defined recently.11, 20 In this study, the newly synthesized compound 15 was
docked into the influenza HA protein of H5N1 (Influenza strain
A/Vietnam/1194/2004, PDB: 2IBX) 11, 19 using Surflex-Dock program of SYBYL 7.3.
We first performed a docking study of compound 15 to predict the binding mode
15

between compound 15 and HA. The computer-aided docking data suggested that
compound 15 exhibited a similar binding mode as the positive compound
CL-38531920, fitting inside the same cavity in the stem region of the HA2 near the
fusion peptide through Van der Waals interactions, hydrophobic contacts, and
hydrogen bonding networks. As shown in Fig. 5, the carbonyl oxygen atom in
17-COO moiety of OA formed a hydrogen bond with ASN812 on the B chain, and the
C2-OH and C4-OH of the L-rhamnose moiety linked to C2-OH of D-glucose forming
stable hydrogen bonds with ASN722 and GLU1031 on the B chain or A chain,
respectively (subscripts 1 and 2 refer to HA1 and HA2 subunits, respectively). For the
17-COO moiety of OA, besides the hydrophobic interaction, the C-28 O-benzyl
substitution of COOH occupied the narrow cavity and engaged in π-π stacking
interaction with LYS822 residue on the B chain, which also explained why the
introduction of the Bn group favored better activity. Among these, ASN81 and
ASN72 were critical for HA2 protein binding via hydrogen bonds, which is essential
for viral envelope fusion. These docking data was consistent well with results from
SAR studies, in particular the requirement for the L-rhamnose moiety linked to
C2-OH of D-glucose in the chacotriose residue and O-benzyl group in the 17-COOH
residue.
Docking studies indicated that the 2-hydroxyl group of the L-rhamnose moiety
linked to C2-OH of D-glucose in chacotriose, O-benzyl group of the17-COOH and
several other residues were critical in HA binding. In order to make sure the key
residues could be kept in proper distance and position, the OA aglycone played an
16

important role in inhibitory activity against H5N1 entry.
Fig. 5. Location of 15 docking site in the HA trimer of the simulated threedimensional structure of
the H5 HA. A) The surface of the docking pocket was presented. The modeling was based on the
published crystal structure of the A/Vietnam/1194/2004 (PDB: 2IBX).
Conclusion
In conclusion, we synthesized a series of 3-O-β-chacotriosyl oleanic acid
derivatives as H5N1 entry inhibitors based on our previously discovered small
molecule inhibitors 1 and 2. A systemic exploration of pharmacological activities of
the compounds indicated that conjugation of pentacyclic triterpene oleanic acid with
the trisaccharide residue chacotriose was the structural requirements for maintenance
of high activity. On the other hand, the above results indicated that either the
attachment of a fit bulk group to 17-COOH of OA or conversion of 3β-OH of
oleanolic acid to 3α-OH group can maintain antiviral activity and greatly enhance
selectivity index. In addition to this, it was clear based on our SAR results that
introduction of an additional oxo group to position 11 at OA was beneficial to
enhance anti-H5N1 activity and selectivity index. Taken together, compound 15
17

displayed better anti-H5N1 influenza virus activity than other types of triterpenes.
Docking studies suggested that these compounds impaired the virus entry step of cell
infection by inserting into the cavity in the stem region of the HA2 near the fusion
peptide, which was consistent with the SPR data. However, the detail mechanism of
action against H5N1 influenza virus need further study.
4. Experimental protocols
4.1. General methods
Solvents were purified in a conventional manner. Thin layer chromatography (TLC)
was performed on precoated E. Merck silica gel 60 F254 plates. Flash column
1
chromatography was performed on silica gel (200-300 mesh, Qingdao, China). H
NMR and 13C NMR spectra were taken on a JEOL JNM-ECP 600 spectrometer with
tetramethylsilane as an internal standard, and chemical shifts are recorded in ppm
values. Mass spectra were recorded on a Q-TOF Global mass spectrometer.
4.2 3-Hydroxy-11-oxo-olean-12-en-28-oic acid benzyl ester (20)
To a solution of 3-O-acetyl-11-oxo-18β-oleanolic acid ether 19 (3.00 g, 5.86 mmol)
in dry DMF (60 mL) was added potassium carbonate (2.02 g, 14.65 mmol). After 30
min of stirring at 50°C, benzyl chloride (1.01 mL, 8.79 mmol) was added and the
mixture was stirred at 50°C for an additional 4 h. The solvents were evaporated in
vacuo. The crude residue was dissolved in a mixture of CH2Cl2 (200 mL) and then
washed with water (50 mL × 3) and brine (60 mL), dried (Na2SO4) and concentrated
under reduced pressure. The crude products were treated with 4 N NaOH (20 mL) in
CH3OH-THF (1:1.5, 50 mL) at room temperature overnight. The solution was
18

neutralized with 3 N HCl, removed the organic solvent and then extracted with EtOAc
1
to provide compound 20 as a white solid (2.95 g, 90% for two steps). H NMR
(CDCl3): δ 7.31-7.38 (m, 5H, Ar-H), 5.65 (s, 1H, H-12), 5.12 (d, 1H, J = 12.4 Hz,
Ar-CH2-1), 5.07 (d, 1H, J = 12.4 Hz, Ar-CH2-2), 3.23 (dd, 1H, J = 11.0, 5.3 Hz, H-3),
3.05 (dd, 1H, J = 13.8, 4.0 Hz, H-1a), 2.83 (dt, 1H, J = 13.6, 3.4 Hz, H-18), 2.29 (s,
1H, H-9), 2.06 (td, 1H, J = 13.9, 4.1 Hz), 1.35, 1.08, 1.00, 0.94, 0.94, 0.80, 0.75 (each
s, each 3H, CH3); 13C NMR (CDCl3): δ 200.2 (C-11), 176.6 (C-28), 168.4 (C-13),
135.9, 128.5 (two), 128.3 (C-12), 128.2 (two), 128.0, 78.8, 66.3, 61.7, 55.0, 46.2, 44.9,
44.3, 43.5, 41.6, 39.1, 37.3, 33.7, 32.9 (two), 31.6, 30.7, 28.1, 27.7, 27.3, 23.5, 23.4,
22.9, 18.9, 17.4, 16.2, 15.6; HRESIMS calcd for C37H52O4Na 583.3763; found
583.3765.
4.3. General procedure for the preparation of 23-26
To
a
solution
of
compounds
17-18
and
20
(1
eq),
or
2,3,4,6-tetra-O-benzoyl-D-glucopyranosyl
trichloroacetimidate
21
2,3,4,6-tetra-O-benzoyl-D-galactopyranosyl trichloroacetimidate 22 (1.3 eq) and 4 Å
molecular sieves in dry CH2Cl2 was added TMSOTf (0.15 eq) at 0 °C under argon.
The reaction mixture was stirred for 45 min and warmed to room temperature for 30
min. The reaction was quenched by Et3N and concentrated in vacuo. The residue was
purified by silica gel column chromatography (petroleum ether-EtOAc, 8:1) to afford
compounds 23-26, respectively.
4.3.1 3α-O-(2,3,4,6-Tetra-O-benzoyl-β-D-glucopyranosyl)-olean-12-en-28-oic acid
benzyl ester (25)
19

1
Compound 25 was synthesized as a white solid in 89% yield; H NMR (CDCl3): δ
7.28-8.02 (m, 25H, Ar-H), 5.93 (t, 1H, J = 9.7 Hz, H-3΄), 5.66 (t, 1H, J = 9.7 Hz,
H-4΄), 5.58 (dd, 1H, J = 9.7, 7.9 Hz, H-2΄), 5.11 (t, 1H, J = 3.5 Hz, H-12), 5.08 (d, 1H,
J = 12.6 Hz, Ar-CH2-1), 5.03 (d, 1H, J = 12.6 Hz, Ar-CH2-2), 4.83 (d, 1H, J = 7.9 Hz,
H-1΄), 4.65 (dd, 1H, J = 11.9, 3.6 Hz, H-6΄-1), 4.52 (dd, 1H, J = 12.0, 5.4 Hz, H-6΄-2),
4.09-4.12 (m, 1H, H-5΄), 3.37 (t, 1H, J = 3.4 Hz, H-3), 2.87 (dd, 1H, J = 14.2.9, 4.6
Hz, H-18), 1.97 (td, 1H, J = 13.6, 4.0 Hz), 1.02, 0.96, 0.95, 0.94, 0.76, 0.75, 0.52
(each s, each 3H, CH3); 13C NMR (CDCl3): δ 177.5 (C-28), 166.1, 165.8, 165.3, 165.0,
143.3 (C-13), 136.5, 133.4, 133.2, 133.1, 133.0, 129.8 (three), 129.7 (two), 129.5,
129.0, 128.9, 128.4 (two), 128.3 (three), 127.9, 127.8, 122.4 (C-12), 98.4 (C-1΄), 82.3,
73.1, 72.0, 71.8, 70.4, 65.9, 63.2, 49.0, 47.1, 46.8, 46.2, 41.7, 41.5, 39.3, 36.7, 36.5,
33.9, 33.2, 32.9, 32.5, 32.4, 30.7, 28.5, 27.6, 25.7, 23.8, 23.1, 22.8, 22.1, 20.9, 17.9,
16.8, 15.2; HRESIMS calcd for C71H80O12Na 1147.5547; found 1147.5551.
4.3.2 3β-O-(2,3,4,6-Tetra-O-benzoyl-β-D-glucopyranosyl)-11-oxo-olean-12-en-28-oic
acid benzyl ester (26)
1
Similarly, compound 26 was synthesized as a white solid in 85% yield; H NMR
(CDCl3): δ 7.28-8.02 (m, 25H, Ar-H), 5.92 (t, 1H, J = 9.7 Hz, H-3΄), 5.65 (s, 1H,
H-12), 5.59 (t, 1H, J = 9.7 Hz, H-4΄), 5.57 (dd, 1H, J = 9.7, 8.0 Hz, H-2΄), 5.11 (d, 1H,
J = 12.4 Hz, Ar-CH2-1), 5.05 (d, 1H, J = 12.4 Hz, Ar-CH2-2), 4.88 (d, 1H, J = 7.9 Hz,
H-1΄), 4.66 (dd, 1H, J = 12.0, 3.2 Hz, H-6΄-1), 4.49 (dd, 1H, J = 12.0, 6.8 Hz, H-6΄-2),
4.14-4.17 (m, 1H, H-5΄), 3.14 (dd, 1H, J = 11.6, 4.8 Hz, H-3), 3.05 (dd, 1H, J = 13.9,
4.2 Hz, H-18), 2.78 (dt, 1H, J = 13.7, 3.3 Hz, H-1a), 1.13, 1.03, 0.94, 0.94, 0.70, 0.69,
20

0.67 (each s, each 3H, CH3); 13C NMR (CDCl3): δ 200.0 (C-11), 176.6 (C-28), 168.4,
166.2, 165.9, 165.3, 165.0, 163.6, 135.9, 133.5, 133.2, 133.1, 129.8 (two), 129.7 (two),
129.5, 129.4, 128.8 (two), 128.5, 128.4, 128.3 (two), 128.2 (two), 127.9, 103.1 (C-1΄),
91.9, 90.6, 73.0, 72.2, 72.0, 70.3, 66.4, 63.4, 61.7, 55.2, 46.2, 44.8, 44.3, 43.4, 41.6,
39.1, 39.0, 37.0, 33.7, 32.95, 32.8, 31.6, 30.7, 27.6, 25.8, 23.5, 23.4, 22.8, 18.9, 17.2,
16.2, 16.1; HRESIMS calcd for C71H78O13Na 1161.5340; found 1161.5346.
4.4. General procedure for the preparation of 27-30
Compounds 23-26 were dissolved in CH2Cl2 and CH3OH (V:V = 1:1) and then
NaOMe was added until pH = 10, respectively. After stirred at r.t. for 8 h, the solution
was neutralized with Dowex 50 × 8 (H+) resin until pH = 7, filtered and concentrated
in vacuo. Then the residue was purified by silica gel column chromatography
(CH2Cl2-MeOH, 10:1) to yield compounds 27-30, respectively.
4.4.1 3β-O-(β-D-Galactopyranosyl)-olean-12-en-28-oic acid benzyl ester (28)
1
Compound 28 was synthesized as a white solid in 98% yield; H NMR (CD3OD): δ
7.31-7.37 (m, 5H, Ar-H), 5.24 (t, 1H, J = 3.4 Hz, H-12), 5.08 (s, 2H, Ar-CH2), 4.29 (d,
1H, J = 7.7 Hz, H-1΄), 3.84 (dd, 1H, J = 3.4, 0.7 Hz, H-4΄), 3.73 (t, 2H, J = 5.8 Hz,
H-6΄), 3.53 (dd, 1H, J = 9.7, 7.7 Hz, H-2΄), 3.49 (td, 1H, J = 6.2, 0.9 Hz, H-5΄), 3.45
(dd, 1H, J = 9.7, 3.4 Hz, H-3΄), 3.18 (dd, 1H, J = 11.8, 4.4 Hz, H-3), 2.92 (dd, 1H, J =
13.9, 4.2 Hz, H-18), 2.05 (td, 1H, J = 14.8, 4.3 Hz, H-19), 1.15, 1.06, 0.94, 0.92, 0.92,
0.85, 0.60 (each s, each 3H, CH3); 13C NMR CD3OD): δ 177.7 (C-28), 143.5 (C-13),
136.3, 128.1 (two), 127.9 (two), 127.8, 122.6 (C-12), 105.9 (C-1΄), 89.3, 74.9, 73.8,
71.7, 68.8, 65.8, 60.9, 55.6, 46.7, 45.7, 41.4, 39.2, 38.7, 38.4, 36.4, 33.4, 32.6, 32.2,
21

32.1, 30.2, 27.3, 27.2, 25.6, 25.0, 23.1, 22.7, 22.6, 17.9, 16.3, 15.6, 14.5; HRESIMS
calcd for C43H64O8Na 731.4499; found 731.4503.
4.4.2 3α-O-(β-D-Glucopyranosyl)-olean-12-en-28-oic acid benzyl ester (29)
1
Similarly, compound 29 was synthesized as a white solid in 98% yield; H NMR
(CD3OD): δ 7.29-7.35 (m, 5H, Ar-H), 5.23 (t, 1H, J = 3.5 Hz, H-12), 5.06 (s, 2H,
Ar-CH2), 4.26 (d, 1H, J = 7.8 Hz, H-1΄), 3.85 (dd, 1H, J = 11.8, 2.3 Hz, H-6΄-1), 3.66
(dd, 1H, J = 11.8, 5.8 Hz, H-6΄-2), 3.44 (t, 1H, J = 2.2 Hz, H-3), 3.34 (t, 1H, J = 9.0
Hz, H-3΄), 3.27 (t, 1H, J = 8.8 Hz, H-4΄), 3.19-3.22 (m, 1H, H-5΄), 3.18 (dd, 1H, J =
9.1, 7.9 Hz, H-2΄), 2.90 (dd, 1H, J = 13.8, 4.3 Hz, H-18), 2.03 (td, 1H, J = 14.4, 3.4
Hz, H-19), 1.17, 0.96, 0.92, 0.92, 0.90, 0.86, 0.58 (each s, each 3H, CH3); 13C NMR
(CDCl3): δ 177.7 (C-28), 143.4 (C-13), 136.2, 128.1 (two), 127.9 (two), 127.8, 122.7
(C-12), 100.3 (C-1΄), 81.1, 77.0, 76.4, 73.7, 70.6, 65.8, 61.6, 49.4, 46.8, 46.7, 45.8,
41.6, 41.4, 39.3, 36.7, 36.5, 33.4, 32.7, 32.4, 32.3, 32.1, 30.2, 27.8, 27.3, 25.1, 23.0,
22.7, 22.6, 21.7, 20.3, 17.9, 16.4, 14.5; HRESIMS calcd for C43H64O8Na 731.4499;
found 731.4505.
4.4.3 3β-O-(β-D-Glucopyranosyl)-11-oxo-olean-12-en-28-oic acid benzyl ester (30)
1
Similarly, compound 30 was synthesized as a white solid in 99% yield; H NMR
(CD3OD): δ 7.31-7.37 (m, 5H, Ar-H), 5.53 (s, 1H, H-12), 5.15 (d, 1H, J = 12.2 Hz,
Ar-CH2-1), 5.06 (d, 1H, J = 12.2 Hz, Ar-CH2-2), 4.32 (d, 1H, J = 7.8 Hz, H-1΄), 3.85
(dd, 1H, J = 11.9, 2.3 Hz, H-6΄-1), 3.67 (dd, 1H, J = 11.9, 5.5 Hz, H-6΄-2), 3.34 (t, 1H,
J = 8.8 Hz, H-3΄), 3.30 (t, 1H, J = 8.8 Hz, H-4΄), 3.21-3.26 (m, 1H, H-5΄), 3.18-3.21
(m, 2H, H-2΄, H-3), 3.04 (dd, 1H, J = 13.1, 3.4 Hz, H-18), 2.71 (dt, 1H, J = 13.3, 3.1
22

Hz, H-1a), 1.39, 1.09, 1.07, 0.96, 0.96, 0.86, 0.69 (each s, each 3H, CH3); 13C NMR
(CD3OD): δ 201.4 (C-11), 176.7 (C-28), 170.4, 136.1, 128.2 (three), 127.9, 127.1,
105.3 (C-1΄), 89.0, 76.9, 76.3, 74.2, 70.2, 66.1, 61.6, 61.4, 55.1, 46.1, 45.0, 44.0, 43.5,
41.9, 39.1, 38.8, 36.8, 33.2, 32.6, 31.7, 31.4, 30.2, 27.3, 27.1, 25.6, 22.5, 22.4, 22.3,
18.3, 15.6, 15.4; HRESIMS calcd for C43H62O9Na 745.4292; found 745.4298.
4.5. General procedure for the preparation of 31-34
Compounds 27-30 (1 mmol) were dissolved in dry CH2Cl2 (100 mL) and pyridine
(V:V = 1:1) and then cooled to -15°C. To the cold solution, pivaloyl chloride (4 mmol)
was added dropwise over a period of 10 min and the reaction mixture was allowed to
stir for another 18 h. After the reaction was complete as judged by TLC, the mixture
was concentrated in vacuo. Then the residue was diluted with EtOAc (150 mL) and
washed with 1 M HCl (60 mL × 3), saturated aq NaHCO3 (60 mL × 3), and brine (60
mL). The organic layer was dried over Na2SO4, filtered, and concentrated under
diminished pressure. Purification of the crude reaction product by column
chromatography on silica gel [petroleum ether-EtOAc (4:1)] provided compounds
31-34, respectively.
4.5.1 3β-O-(3,6-Di-O-Pivaloyl-β-D-glucopyranosyl)-olean-12-en-28-oic acid benzyl
ester (31)
1
Compound 31 was synthesized as a white solid in 75% yield; H NMR (CD3COCD3):
δ 7.36-7.40 (m, 4H, Ar-H), 7.32 (tt, 1H, J = 7.9, 1.7 Hz, Ar-H), 5.26 (t, 1H, J = 3.6 Hz,
H-12), 5.10 (d, 1H, J = 12.5 Hz, Ar-CH2-1), 5.06 (d, 1H, J = 12.5 Hz, Ar-CH2-2),
4.94 (t, 1H, J = 9.4 Hz, H-3΄), 4.47 (dd, 1H, J = 11.6, 2.2 Hz, H-6-1΄), 4.46 (d, 1H, J =
23

7.8 Hz, H-1΄), 4.09 (dd, 1H, J = 11.6, 7.4 Hz, H-6-2΄), 3.60-3.63 (m, 1H, H-5΄), 3.45
(t, 1H, J = 9.4 Hz, H-4΄), 3.38 (t, 1H, J = 8.1 Hz, H-2΄), 3.17 (dd, 1H, J = 11.8, 4.7 Hz,
H-3), 2.95 (dd, 1H, J = 14.2, 4.6 Hz, H-18), 1.20, 1.19 (each s, each 9H, 2 × (CH3)3C),
1.16, 1.04, 0.93, 0.92, 0.92, 0.81, 0.64 (each s, each 3H, each CH3); 13C NMR
(CD3COCD3): δ 177.4, 177.3, 176.5, 143.8 (C-13), 136.8, 128.3 (two), 127.9 (two),
127.8, 122.4 (C-12), 105.5 (C-1΄), 89.2, 77.4, 73.8, 72.6, 69.3, 65.5, 63.7, 55.5, 47.6,
46.5, 45.7, 41.5 (two), 39.3, 38.9, 38.5, 38.4 (two), 36.6, 33.5, 32.6, 32.4, 32.3, 30.4,
27.4 (two), 26.6, 26.5, 25.8, 25.4, 23.2, 23.0, 22.9, 18.0, 16.6, 15.9, 14.9; HRESIMS
calcd for C53H80O10Na 899.5649; found 899.5653.
4.5.2 3β-O-(3,6-Di-O-Pivaloyl-β-D-galactopyranosyl)-olean-12-en-28-oic acid benzyl
ester (32)
1
Similarly, compound 32 was synthesized as a white solid in 78% yield; H NMR
(CDCl3): δ 7.30-7.36 (m, 5H, Ar-H), 5.30 (t, 1H, J = 3.5 Hz, H-12), 5.11 (d, 1H, J =
12.5 Hz, Ar-CH2-1), 5.05 (d, 1H, J = 12.5 Hz, Ar-CH2-2), 4.867 (dd, 1H, J = 10.1, 3.4
Hz, H-3΄), 4.39 (d, 1H, J = 7.7 Hz, H-1΄), 4.29 (d, 2H, J = 6.4 Hz, H-6΄), 3.97 (dd, 1H,
J = 3.4, 0.5 Hz, H-4΄), 3.88 (d, 1H, J = 10.1, 7.7 Hz, H-2΄), 3.75 (td, 1H, J = 6.4, 0.8
Hz, H-5΄), 3.16 (dd, 1H, J = 11.8, 4.5 Hz, H-3), 2.92 (dd, 1H, J = 14.0, 4.0 Hz, H-18),
1.99 (td, 1H, J = 13.6, 4.1 Hz, H-19), 1.26 (s, 9H, C(CH3)3), 1.21 (s, 9H, C(CH3)3),
1.13, 1.00, 0.93, 0.91, 0.89, 0.83, 0.61 (each s, each 3H, CH3); 13C NMR (CDCl3): δ
178.3, 178.0, 177.5, (C-28), 143.7 (C-13), 136.4, 128.4 (two), 128.0 (two),
127.9, 122.5 (C-12), 105.6 (C-1΄), 90.5, 74.5, 72.3, 70.1, 67.5, 65.9, 62.8, 55.5, 47.6,
46.7, 45.9, 41.7, 41.4, 39.3, 39.0 (two), 38.8, 38.4, 36.7, 33.8, 33.1, 32.7, 32.4, 30.7,
24

28.3, 27.6, 27.2, 27.1, 27.0, 25.8, 23.7, 23.4, 23.0, 18.2, 16.9, 16.7, 15.3; HRESIMS
calcd for C53H80O10Na 899.5650; found 899.5655.
4.5.3 3α-O-(3,6-Two-O-benzoyl-β-D-glucopyranosyl)-olean-12-en-28-oic acid benzyl
ester (33)
Similarly, compound 33 was synthesized as a white solid in 77% yield;1H NMR
(CDCl3): δ 7.30-7.36 (m, 5H, Ar-H), 5.29 (t, 1H, J = 3.6 Hz, H-12), 5.11 (d, 1H, J =
12.5 Hz, Ar-CH2-1), 5.05 (d, 1H, J = 12.5 Hz, Ar-CH2-2), 4.88 (t, 1H, J = 9.1 Hz,
H-3΄), 4.46 (dd, 1H, J = 11.9, 1.3 Hz, H-6΄-1), 4.32 (d, 1H, J = 7.7 Hz, H-1΄),
4.26-4.29 (m, 1H, H-6΄-2), 3.57 (dd, 1H, J = 9.4, 7.9 Hz, H-2΄), 3.50-3.53 (m 2H,
H-4΄, H-5΄), 3.37 (t, 1H, J = 2.1 Hz, H-3), 2.91 (dd, 1H, J = 14.1, 4.4 Hz, H-18), 1.99
(td, 1H, J = 13.7, 4.2 Hz, H-19), 1.26 (s, 9H, (CH3)3C), 1.22 (s, 9H, (CH3)3C), 1.15,
13
0.93, 0.91, 0.91, 0.90, 0.83, 0.62 (each s, each 3H, CH3); C NMR (CDCl3): δ 180.3,
178.5, 177.5 (C-28), 143.6 (C-13), 136.5, 128.4 (two), 128.0 (two), 127.9, 122.5
(C-12), 100.0 (C-1΄), 81.9, 78.1, 74.2, 71.9, 70.1, 65.9, 63.1, 49.9, 47.5, 46.8, 45.9,
41.8, 41.4, 39.5, 39.1, 38.9, 36.9, 36.8, 33.9, 33.5, 33.1, 32.5, 32.4, 30.7, 28.5, 27.6,
27.2, 27.1, 26.1, 23.6, 23.3, 23.1, 22.5, 21.1, 18.1, 16.9, 15.2; HRESIMS calcd for
C53H80O10Na 899.5649; found 899.5653.
4.5.4 3β-O-(3,6-Di-O-Pivaloyl-β-D-glucopyranosyl)-11-oxo-olean-12-en-28-oic acid
benzyl ester (34)
1
Similarly, compound 34 was synthesized as a white solid in 81% yield; H NMR
(CDCl3): δ 7.32-7.37 (m, 5H, Ar-H), 5.65 (s, 1H, H-12), 5.11 (d, 1H, J = 12.4 Hz,
Ar-CH2-1), 5.07 (d, 1H, J = 12.4 Hz, Ar-CH2-2), 4.87 (t, 1H, J = 9.3 Hz, H-3΄), 4.46
25

(dd, 1H, J = 12.0, 2.5 Hz, H-6΄-1), 4.41 (d, 1H, J = 7.8 Hz, H-1΄), 4.20 (dd, 1H, J =
12.0, 7.0 Hz, H-6΄-2), 3.55-3.58 (m, 2H, H-5΄, H-4΄), 3.47 (t-like, 1H, J = 9.3 Hz,
H-2΄), 3.17 (dd, 1H, J = 11.8, 4.6 Hz, H-3), 3.05 (dd, 1H, J = 13.9, 4.2 Hz, H-18),
2.81 (dt, 1H, J = 13.8, 3.5 Hz, H-1a), 1.26 (s, 9H, (CH3)3C), 1.21 (s, 9H, (CH3)3C),
1.34, 1.08, 1.01, 0.94, 0.94, 0.84, 0.74 (each s, each 3H, CH3); 13C NMR (CDCl3): δ
200.1 (C-11), 180.3, 178.7, 176.6, 168.4, 135.9, 128.5 128.3, 128.2, 128.0, 104.8
(C-1΄), 90.1, 78.0, 74.1, 72.7, 70.1, 66.4, 63.7, 61.7, 55.3, 46.2, 44.9, 44.3, 43.4, 41.6,
39.3, 39.1, 38.8, 37.0, 33.7, 32.9, 31.6, 30.7, 28.2, 27.6, 27.2, 27.1, 25.8, 23.4 (two),
22.9, 18.9, 17.3, 16.6, 16.1; HRESIMS calcd for C53H78O11Na 913.5442; found
913.5445.
4.6 General procedure for the preparation of 3-8, 13 and 15
To a mixture of 31-34 (1 eq) and 4 Å molecular sieves in dried CH2Cl2 (30 mL) at -
10 °C under argon was added TMSOTf (0.20 eq), followed by a solution of different
monosaccharidal
trichloroacetimidates
including
L-fucopyranosyl
trichloroacetimidate 35, D-rhamnopyranosyl trichloroacetimidate 36, L-arabinosyl
trichloroacetimidate 37, D-arabinosyl trichloroacetimidate 38, D-xylopyranosyl
trichloroacetimidate 39 and L-rhamnopyranosyl trichloroacetimidate 40 (5 eq) in
CH2Cl2 (5 mL). After stirring at -10 °C for 2 h and then at r.t. for 2 h, the reaction was
quenched with Et3N. The solid was filtered, and the filtrate was concentrated under
vacuum to furnish yellow oil. The oil was subjected to column chromatography on
silica gel (EtOAc-petroleum ether, 1:4) to give the desired crude trisaccharide. The
above trisaccharide mixture was dissolved in THF-CH3OH-H2O (V:V:V = 1:1:1) and
26

then NaOH (12 eq) was added. After stirred at 50 °C for 12 h, the solution was
neutralized with Dowex 50 × 8 (H+) resin until pH = 7, filtered and concentrated in
vacuo. Then the residue was purified by silica gel column chromatography
(CH2Cl2-MeOH, 6:1) to afford compounds 3-8, 13 and 15, respectively.
4.6.1 3β-O-[2,4-Di-O-(β-L-Fucopyranosyl)-β-D-glucopyranosyl]-olean-12-en-28-
oic acid benzyl ester (3)
1
Compound 3 was synthesized as a white solid in 62% yield for two steps; H NMR
(CD3OD): δ 7.31-7.37 (m, 5H, Ar-H), 5.24 (t, 1H, J = 3.2 Hz, H-12), 5.07 (d, 2H, J =
1.5 Hz, Ar-CH2), 4.64 (d, 1H, J = 7.6 Hz, Fuc-H-1), 4.45 (d, 1H, J = 7.6 Hz, Fuc-H-1),
4.54 (d, 1H, J = 7.8 Hz, H-1΄), 3.90 (dd, 1H, J = 12.3, 2.3 Hz, H-6΄-1), 3.79 (dd, 1H, J
= 12.2, 4.9 Hz, H-6΄-2), 3.71 (t, 1H, J = 9.1 Hz), 3.62-3.70 (m, 3H), 3.60-3.61 (m,
3H), 3.53 (t, 1H, J = 7.9 Hz), 3.52 (t, 1H, J = 8.0 Hz), 3.50-3.51 (m, 2H), 3.46 (dd, 1H,
J = 9.8, 3.2 Hz, Fuc-H-3), 3.34-3.35 (m, 1H), 3.24 (dd, 1H, J = 11.7, 4.3 Hz, H-3),
2.91 (dd, 1H, J = 13.7, 4.3 Hz, H-18), 1.28 (d, 3H, J = 6.4 Hz, Fuc-H-6), 1.27 (d, 3H,
J = 6.4 Hz, Fuc-H-6), 1.14, 1.07, 0.94, 0.92, 0.92, 0.85, 0.59 (each s, each 3H, CH3);
13C NMR (CD3OD): δ 177.7 (C-28), 143.5 (C-13), 136.3, 129.1, 128.1, 127.9, 127.8,
122.6 (C-12), 104.7 (C-1΄), 103.7 (Fuc-C-1), 101.9 (Fuc-C-1), 88.3, 79.0, 78.4, 75.1,
74.9, 74.7, 73.6, 73.5, 71.8, 71.7, 71.5, 71.2, 70.9, 70.6, 65.8, 61.4, 55.7, 48.2, 46.7,
45.7, 41.4 (two), 39.2, 38.7, 38.5, 36.4, 33.4, 32.6, 32.2, 32.1, 30.2, 29.3, 27.5, 27.3,
25.7, 25.0, 23.1, 22.7, 22.5, 17.9, 16.3, 15.8, 15.3, 14.5; HRESIMS calcd for
C55H84O16Na 1023.5652; found 1023.5655.
4.6.2 3β-O-[2,4-Di-O-(α-D-Rhamnopyranosyl)-β-D-glucopyranosyl]-olean-12-en-28
27

-oic acid benzyl ester (4)
1
Similarly, 4 was prepared as a white solid in 60% yield for two steps; H NMR
(CD3OD): δ 7.31-7.36 (m, 5H, Ar-H), 5.24 (t, 1H, J = 3.4 Hz, H-12), 5.23 (d, 1H, J =
1.7 Hz, Rha-H-1), 5.19 (d, 1H, J = 1.1 Hz, Rha-H-1), 5.07 (d, 2H, J = 1.1 Hz,
Ar-CH2), 4.42 (d, 1H, J = 7.9 Hz, H-1΄), 4.05-4.09 (m, 1H), 3.99 (dd, 1H, J = 3.1, 1.8
Hz, Rha-H-2), 3.92 (dd, 1H, J = 3.2, 1.5 Hz, Rha-H-2), 3.86 (dd, 1H, J = 11.8, 1.8 Hz,
H-6΄-1), 3.82-3.83 (m, 1H), 3.67-3.71 (m, 3H), 3.64 (dd, 1H, J = 9.6, 3.3 Hz,
Rha-H-3), 3.50-3.53 (m, 2H), 3.41 (t, 1H, J = 9.6 Hz), 3.40 (t, 1H, J = 9.2 Hz), 3.37 (t,
1H, J = 9.0 Hz), 3.19 (dd, 1H, J = 11.7, 4.3 Hz, H-3), 2.91 (dd, 1H, J = 13.7, 4.2 Hz,
H-18), 1.28 (d, 3H, J = 6.2 Hz, Rha-H-6), 1.25 (d, 3H, J = 6.2 Hz, Rha-H-6), 1.15,
1.02, 0.94, 0.92, 0.92, 0.84, 0.58 (each s, each 3H, CH3); 13C NMR (CD3OD): δ 177.6
(C-28), 143.4 (C-13), 136.3, 128.1 (two), 127.9 (two), 127.8, 122.6 (C-12), 105.0
(C-1΄), 101.9 (Rha-C-1), 100.3 (Rha-C-1), 88.8, 79.0, 77.1, 76.2, 75.1, 72.5, 72.4,
71.0, 70.9, 70.7, 69.2, 69.1, 68.6, 65.8, 61.3, 55.7, 48.2, 46.7, 45.7, 41.4 (two), 39.2,
38.6, 38.5, 36.4, 33.4, 32.5, 32.2, 32.1, 30.2, 27.3, 27.2, 25.8, 25.0, 23.1, 22.7, 22.5,
17.9, 16.6, 16.5, 16.3, 15.9, 14.5; HRESIMS calcd for C55H84O16Na 1023.5652; found
1023.5654.
4.6.3 3β-O-[2,4-Di-O-(α-L-Arabinopyranosyl)-β-D-glucopyranosyl]-olean-12-en-28
-oic acid benzyl ester (5)
1
Similarly, 5 was prepared as a white solid in 56% yield for two steps; H NMR
(CD3OD): δ 7.32-7.37 (m, 5H, Ar-H), 5.24 (t, 1H, J = 3.4 Hz, H-12), 5.08 (s, 2H,
Ar-CH2), 4.49 (d, 1H, J = 7.0 Hz, Ara-H-1), 4.47 (d, 1H, J = 7.9 Hz, H-1΄), 4.33 (d,
28

1H, J = 7.3 Hz, Ara-H-1), 3.93 (dd, 1H, J = 12.7, 2.9 Hz, Ara-H-5a), 3.87 (dd, 1H, J =
12.6, 2.9 Hz, Ara-H-5b), 3.86 (t, 1H, J = 3.2 Hz), 3.79-3.83 (m, 2H), 3.72 (t, 1H, J =
9.1 Hz), 3.66 (t, 1H, J = 9.1 Hz), 3.64 (t, 1H, J = 9.1 Hz), 3.60 (t, 1H, J = 9.9 Hz),
3.56-3.59 (m, 2H), 3.49-3.54 (m, 3H), 3.40 (dt, 1H, J = 9.7, 3.2 Hz), 3.17 (dd, 1H, J =
11.7, 4.3 Hz, H-3), 2.91 (dd, 1H, J = 13.8, 4.4 Hz, H-18), 1.15, 1.07, 0.94, 0.92, 0.92,
0.85, 0.59 (each s, each 3H, CH3); 13C NMR (CD3OD): δ 177.6 (C-28), 143.5 (C-13),
136.3, 128.1 (two), 127.9 (two), 127.8, 122.6 (C-12), 105.2 (C-1΄), 103.9 (Ara-C-1),
103.8 (Ara-C-1), 89.2, 81.8, 78.7, 74.9, 74.5, 72.9, 72.6, 72.5, 71.1, 68.5, 68.2, 66.4,
66.0, 65.8, 60.5, 55.7, 46.7, 45.7, 41.4, 39.2, 38.8, 38.5, 36.4, 33.4, 32.6, 32.2, 32.1,
30.2, 27.3, 27.0, 25.8, 25.0, 23.1, 22.7, 22.6, 17.9, 16.4, 15.3, 14.6; HRESIMS calcd
for C53H80O16Na 995.5344; found 995.5346.
4.6.4 3β-O-[2,4-Di-O-(α-D-Arabinopyranosyl)-β-D-glucopyranosyl]-olean-12-en-28
-oic acid benzyl ester (6)
1
Similarly, 6 was prepared as a white solid in 58% yield for two steps; H NMR
(CD3OD): δ 7.31-7.36 (m, 5H, Ar-H), 5.24 (t, 1H, J = 3.1 Hz, H-12), 5.07 (s, 2H,
Ar-CH2), 4.82 (d, 1H, J = 5.3 Hz, Ara-H-1), 4.56 (d, 1H, J = 7.1 Hz, Ara-H-1), 4.52 (d,
1H, J = 7.9 Hz, H-1΄), 4.00 (dd, 1H, J = 12.1, 5.3 Hz, Ara-H-5a), 3.87-3.92 (m, 3H),
3.79-3.81 (m, 1H), 3.76 (dd, 1H, J = 12.1, 5.0 Hz, Ara-H-5b), 3.71 (t, 1H, J = 9.2 Hz),
3.69 (t, 1H, J = 7.5 Hz), 3.48-3.61 (m, 7H), 3.33-3.35 (m, 1H), 3.23 (dd, 1H, J = 11.8,
4.3 Hz, H-3), 2.91 (dd, 1H, J = 13.9, 3.9 Hz, H-18), 1.15, 1.06, 0.94, 0.92, 0.85, 0.85,
0.59 (each s, each 3H, CH3); 13C NMR (CD3OD): δ 177.7 (C-28), 143.4 (C-13), 136.3,
128.1 (two), 127.9 (two), 127.8, 122.6 (C-12), 104.9 (C-1΄), 104.3 (Ara-C-1), 101.2
29