Synthesis and proteasome inhibition of glycyrrhetinic acid derivatives

Article abstract of DOI:10.1016/j.bmc.2008.05.078

This study discovered that glycyrrhetinic acid inhibited the human 20S proteasome at 22.3 μM. Esterification of the C-3 hydroxyl group on glycyrrhetinic acid with various carboxylic acid reagents yielded a series of analogs with marked improved potency. Among the derivatives, glycyrrhetinic acid 3-O-isophthalate (17) was the most potent compound with IC₅₀ of 0.22 μM, which was approximately 100-fold more potent than glycyrrhetinic acid.

Full text of DOI:10.1016/j.bmc.2008.05.078

Bioorganic & Medicinal Chemistry 16 (2008) 6696–6701
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Synthesis and proteasome inhibition of glycyrrhetinic acid derivatives
Li Huang a, Donglei Yu b, Phong Ho a, Keduo Qian b, Kuo-Hsiung Lee b, Chin-Ho Chen a,
*
a Department of Surgery, Duke University Medical Center, Box 2926, Durham, NC 27710-2926, USA
b Natural Products Research Laboratories, School of Pharmacy, University of North Carolina, Chapel Hill, NC 27599, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
This study discovered that glycyrrhetinic acid inhibited the human 20S proteasome at 22.3 lM. Esterifi-
cation of the C-3 hydroxyl group on glycyrrhetinic acid with various carboxylic acid reagents yielded a
series of analogs with marked improved potency. Among the derivatives, glycyrrhetinic acid 3-O-iso-
Received 14 March 2008
Revised 29 May 2008
Accepted 29 May 2008
Available online 5 June 2008
phthalate (17) was the most potent compound with IC50 of 0.22
lM, which was approximately 100-fold
more potent than glycyrrhetinic acid.
Ó 2008 Elsevier Ltd. All rights reserved.
Keywords:
Glycyrrhetinic acid
Proteasome inhibitor
Triterpene
1. Introduction
protein complex PA700 or PA28 binds to the a ring that opens the
channel in the 20S proteasome and allows the ubiquitin-tagged
proteins to access the proteolytic sites located inside the chamber.
Within the b rings, there are multiple proteolytic sites including
two chymotrypsin-like (ChT-L), two trypsin-like (T-L), and two cas-
pase-like (CA-L) proteolytic sites.10 The proteasome inhibitor
PS341 (bortezomib) was successfully developed into an anti-can-
cer drug for the treatment of multiple myeloma.11 Differential inhi-
bition of the three catalytic activities by bortezomib is believed to
be critical for clinical benefit in disease treatment.12 Recently, we
reported that betulinic acid (BA) and its derivatives could regulate
the proteasome activities.13 BA acted as an activator of the 20S pro-
teasome complex. On the other hand, some of the BA derivatives
were shown to be inhibitors of the 20S proteasome.
18b-Glycyrrhetinic acid (GLA) is the aglycone of glycyrrhizic
acid, a major component found in licorice root that has been
widely used in traditional medicines, especially in Asia.1 The
known pharmacological activities of GLA and glycyrrhizic acid in-
clude anti-inflammation, anti-ulcer, anti-allergenic, and anti-viral.
The GLA derivative, carbenoxolone, has been used in Europe as a
licensed medicine for the treatment of esophageal ulceration and
inflammation.2 GLA was shown to have multiple biological activi-
ties, including inhibition of 11-b-hydroxysteroid dehydrogenase
and induction of mitochondrial permeability transition.3–5 How
these biological activities contribute to the pharmacological effect
of GLA remains to be determined.
The proteasome–ubiquitin pathway is known for breaking
down proteins to remove the misfolded or damaged proteins in
eukaryotic cells.6,7 In the past decade, increasing evidence has indi-
cated that the proteasome–ubiquitin system plays an important
role in many cellular functions such as cell-stress response, cell-cy-
cle regulation, cellular differentiation, and antigenic peptide gener-
ation. Therefore, targeting the proteasome–ubiquitin pathway has
been considered a novel strategy for the treatment of various dis-
orders such as neurodegenerative diseases, cancers, and inflamma-
tory diseases.8
2. Results and discussion
In this study, we continued our effort of searching for new com-
pounds that possess proteasomal regulatory activity. Several triter-
penes with structures similar to BA were tested for their potentials
in regulating the proteasome, including glycyrrhetinic acid (GLA, 1),
oleanolic acid (OA), ursolic acid (UA), and moronic acid (MA) as
shown in Figure 1. The compounds were initially examined for their
effect on the ChT-L activity of the proteasome. The ChT-L activity is
the most critical enzymatic process in the proteasome. Genetic
studies suggested that impaired proteasome ChT-L activity caused
by mutations resulted in strong reduction in the degradation of
proteasomal substrates,14–17 whereas impaired T-L or CA-L activity
of proteasomes did not cause such a detrimental effect. Our current
study indicated that among these five triterpenes, GLA is the only
one that inhibited the ChT-L activity of the proteasome. GLA
The core structure of the proteasome is a barrel-shaped 20S
complex that has four stacked rings each with 7 subunits.9 The
two outer rings and two inner rings are named
a rings and b rings,
respectively. The 20S proteasome is activated when the regulatory
* Corresponding author. Tel.: +1 919 684 3819; fax: +1 919 684 3878.
E-mail address: chc@duke.edu (C.-H. Chen).
0968-0896/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2008.05.078

L. Huang et al. / Bioorg. Med. Chem. 16 (2008) 6696–6701
6697
H
H
E
H
H
COOH
COOH
COOH
28
A
H
H
H
3
HO
HO
HO
H
H
H
Ursolic acid (UA)
Oleanolic acid (OA)
Betulinic acid (BA)
COOH
30
20
H
H
O
COOH
H
H
O
H
HO
H
1
)
Glycyrrhetinic acid (GLA,
Moronic acid (MA)
Figure 1. Chemical structure of triterpenes.
Table 1
reduced the proteasome activity by 50% at 22.3 lM. MA did not in-
13C NMR spectral data (assignment) of compounds 8, 12, 17, and 18a
hibit or activate proteasome activity, possibly due to the lack of a C-
3 hydroxyl group. On the contrary, BA, UA, and OA all activated the
ChT-L activity of the proteasome (data not shown).
8
12
17
18
80.5 (C-3)
55.1 (C-5)
62.0 (C-9)
199.6 (C-11)
128.7 (C-12)
170.0 (C-13)
179.3 (C-30)
48.8
45.6
44.2
43.6
41.8
39.1
38.5
38.4
37.5
32.8
32.2
31.7
80.9 (C-3)
55.1 (C-5)
62.0 (C-9)
199.6 (C-11)
128.7 (C-12)
170.0 (C-13)
179.3 (C-30)
48.8
81.9 (C-3)
55.1 (C-5)
62.0 (C-9)
199.6 (C-11)
128.7 (C-12)
170.1 (C-13)
179.3 (C-30)
48.8
81.6 (C-3)
55.1 (C-5)
62.0 (C-9)
199.6 (C-11)
128.7 (C-12)
170.2 (C-13)
179.5 (C-30)
48.8
45.7
44.3
43.6
41.8
39.1
38.7
38.5
37.5
32.8
32.3
31.7
28.9
2.1. Synthesis and characterization of GLA derivatives
Our previous study indicated that BA derivatives with side-
chain modifications at the C-3 position transformed BA into pro-
teasome inhibitors. In an attempt to increase the potency of GLA’s
inhibitory effect on proteasome, a series of GLA derivatives func-
tionalized with a variety of C-3 side chains were synthesized. The
synthesis of GLA C-3 ester derivatives was accomplished by treat-
ing GLA with corresponding carboxylic acid reagents under general
esterification conditions. Compounds 2–7, 9, and 11 were obtained
with corresponding anhydrides in the presence of DMAP and pyr-
idine using microwave-assisted heating. The rest of the compounds
(8, 10, and 12–18) were synthesized by coupling the corresponding
carboxylic acid reagents in the presence of EDC/DMAP or DCC/
DMAP under microwave irradiation. Compounds 19–21 were mod-
ified at the C-30 position with amide moieties with or without
additional modification on their C-3 position. The synthesis of
these compounds has been previously reported.18
45.6
45.6
44.2
44.2
43.6
43.6
41.8
41.8
39.1
39.1
38.8
38.8
38.5
38.5
37.5
37.5
32.8
32.8
32.3
32.3
31.7
31.7
28.82
28.78
28.3
26.9
26.7
28.83
28.78
28.3
28.83
28.80
28.4
28.8
28.4
26.9
26.7
26.9
26.9
26.7
26.7
24.1
24.1
24.1
24.1
23.7
18.9
17.7
17.3
16.8
The structures of four of the most potent GLA derivatives, 8, 12,
17, and 18, were further confirmed with 13C NMR studies (Table 1).
The spectra have shown all 30 carbons of GLA triterpene skeletons
with unambiguous assignment for 7 of them based on comparison
of data reported before for GLA.19 The carbon signals of C-3 ester
side chain were also observed except for the three overlapping sig-
nals in compound 18. In summary, the spectra showed no or very
small difference of chemical shift (all within 0.2 ppm except for the
C-3) among the four compounds on the triterpene skeleton region
due to its structural rigidity and stability. The C-3 chemical shifts
showed more variety among these GLA esters and were found to
be more than 2 ppm downfield than that in GLA.
23.6
18.9
17.7
17.2
23.6
18.8
17.7
17.1
23.7
18.9
17.7
17.3
16.8
16.8
16.8
C on C-3 side chain
175.8 (C@O)
173.2 (C@O)
34.7 (CH2CO)
34.6 (CH2CO)
174.0 (C@O)
171.5 (C@O)
127.8 (CH@)
126.1 (CH@)
38.6 (@CHCH2)
38.5 (@CHCH2)
168.5 (C@O)
165.9 (C@O)
134.5 (Ar-C)
133.6 (Ar-C)
133.4 (Ar-C)
131.9 (Ar-C)
131.3 (Ar-C)
129.3 (Ar-C)
166.2b (C@O)
b
130.5c (Ar-C)
c
25.24 (CH2CH2CO)
25.21 (CH2CH2CO)
129.7d (Ar-C)
d
a
Recorded in pyridine-d5 at 125 MHz.
Overlapping signals.
2.2. Inhibition of the proteasome by GLA derivatives
b,c,d
The synthesized compounds were assayed at various concentra-
tions in order to determine their potency against the ChT-L activ-
ity.13 The known proteasome inhibitors Ac-Leu-Leu-Met-CHO
(LLM-F) and lactacystin were included in the assay as controls.
The results indicated that C-3 esterification of GLA increased the
potency of proteasome inhibition by 3- to 100-fold when com-
pared with unmodified GLA (Table 2). It appeared that compounds
with aromatic C-3 side chains as seen in 13–18 were, in general,
more potent (IC50 = 0.22–2.27 M) than compounds without the
aromatic side chain except for compounds 8 and 12. Compound
8 possessed an unbranched side chain similar to compounds 4
and 6. The length of the side chains of these compounds was pos-
l

6698
L. Huang et al. / Bioorg. Med. Chem. 16 (2008) 6696–6701
Table 2
and 8 increased from C4, C5, to C6, respectively, and this correlated
with the improved potency of these compounds. The 50% inhibi-
tory concentrations (IC50) of compounds 4, 6, and 8 were 2.63,
Inhibition of the proteasome by GLA and its derivatives
COR2
H
1.39, and 0.29
unsaturated side chain, also exhibited potent inhibitory activity
with IC50 equal to 0.35 M.
lM, respectively. Compound 12, containing an
O
l
H
3
The free carboxylic acid moiety in the C-3 side chain was not re-
quired for proteasome inhibition as shown with compounds 11, 13,
and 14, which exhibited moderate potency against proteasome.
However, a free carboxylic acid moiety appeared to enhance the
potency of the compound. For example, the carboxylic acid on
compounds 17 and 18 might be responsible for the 10-fold in-
crease in the anti-proteasomal activity when compared with com-
pound 13.
R1O
H
18-β-glycyrrhetinic acid derivatives
Compound R1
R2
IC50
(
l
M) IC50
(
l
M) IC50
(
l
M)
(ChT-L)a (T-L)a
(Ca-L)a
1
H
OH
OH
22.3
0.87
>40
—
>40
CH
3 O
O
2b
—
—
H3C
HOOC
CH3
Additional side-chain modification at C-30 significantly de-
creased the potency of the C-3 derivatives of GLA. The IC50 of the
3
OH
8.70
—
HOOC
C-3 derivative, 2, was 0.87
teasome. Addition of a C-30 side chain resulted in compound 21
that inhibited the proteasomal activity at 3.09 M.
lM against the ChT-L activity of the pro-
CH3
O
l
4
5
OH
OH
2.63
6.86
—
—
—
—
HOOC
HOOC
One of the possible reasons that bortezomib can selectively in-
hibit cancer cells could be due to its preferential inhibitory activity
against the ChT-L activity of the proteasome.12,20 To determine
whether GLA derivatives can also preferentially inhibit the ChT-L
activity of the proteasome, GLA and three of the most potent com-
pounds, 8, 12, and 17, were tested for their effects on the T-L and
CA-L activities of the proteasome. These two proteolytic activities
were analyzed using the assays previously described.13 In general,
the GLA derivatives were at least 10-fold less potent against the
CA-L or T-L activity when compared with their anti-ChT-L activity
of the proteasome (Table 2). These results suggested that all of the
three tested compounds exhibited preferential inhibitory activity
against the ChT-L activity of the proteasome.
To determine whether GLA derivatives could inhibit the protea-
some in the cells, 17, the most potent compound of the series, was
tested in a cell-based proteasome assay previously described.13 To
determine the effect of 17 on the ChT-L activity of the proteasome
in living cells, MT4 cells were treated with 17 or the known protea-
some inhibitors LLM-F and lactacystin. The ChT-L activity of the
proteasome in the cells was analyzed using a Promega cell-based
proteasome assay kit and protocol. In this study, compound 17
inhibited the ChT-L activity of the proteasome by 50% at approxi-
O
CH3
CH3
HOOC
6
7
OH
OH
1.39
1.93
0.29
—
—
—
—
O
HOOC
HOOC
H3C
O
O
8
OH
OH
3.32
5.02
—
O
9b
>40
—
—
O
H3C
10
OH
>40
1.92
—
O
H3C
H3C
11
12
OH
OH
—
—
O
O
H
0.35
2.08
19.3
26.5
HOOC
H
O
O
13
14
15
16
17
OH
OH
OH
OH
OH
—
—
—
—
—
O
2.27
1.05
1.78
0.22
—
O
CH3
O
—
mately 0.25 lM (Fig. 2). The inhibitory activity of 17 in the cell-
HOOC
—
HOOC
HOOC
O
O
120
100
80
60
40
20
0
2.91
3.56
HOOC
H
18
OH
0.31
—
—
—
—
O
NH
19b
>40
HOOC
NH
MeOOC
Compound 17
LLM-F
20b
>40
—
—
—
—
H3C
O
CH
3 O
Lactacystin
NH
21b
3.09
H3C
HOOC
HOOC
0.01
0.1
1
10
LLM-Fc
5.25
5.6
>25
>30
22.5
>30
Lactacystinc
Inhibitor (uM)
a
The inhibition of chymotrypsin-like (ChT-L), trypsin-like (T-L), and caspase-like
(Ca-L) activities of the 20S proteasome was determined in the presence of various
concentrations of the compounds as previously described.13 The value of IC50 was
averaged from three independent assays.
Figure 2. The GLA derivative 17 inhibited the chymotrypsin-like activity of the
proteasome in MT4 cells. A Promega cell-based proteasome assay was used to
determine the effect of compound 17 on the ChT-L activity of 20S proteasome.13 The
known proteasome inhibitors LLM-F and lactacystin were used as positive controls
for proteasome inhibition. MT4 cells are human T cells isolated from a patient with
adult T-cell leukemia. The ChT-L proteasome activity was detected as the relative
light unit (RLU) generated from the cleaved substrate in the assay. The percentage
control is derived using the formula: 100ꢀ (RLU with test compound)/(RLU from
the control without compound).
b
These compounds were previously reported for their anti-HIV activities.
LLM-F and lactacystin are known proteasome inhibitors.
c
itively correlated with the potency of the compounds against the
proteasome. For example, the C-3 side chains of compounds 4, 6,

L. Huang et al. / Bioorg. Med. Chem. 16 (2008) 6696–6701
6699
Table 3
based assay was comparable to the purified human 20S protea-
Purity of compounds determined by HPLC
some assay. On the other hand, lactacystin was approximately
one log10 more potent in the cell-based assay when compared with
that in the purified 20S proteasome assay. It is possible that the
metabolites of lactacystin in the cells are responsible for the in-
creased potency.21
Compound
Puritya (%)
Purityd (%)
Compound
Puritya (%)
Purityd (%)
2
3
4
5
6
7
8
10
98.9
95.6
98.1
99.6
95.4
98.0
91.8
100
94.8
95.1
99.5
99.6
98.3
100
11
12
13
14
15
16
17
18
97.5b
96.9
98.7c
96.4
98.4
97.1
96.4
99.1
97.2
91.6
99.4
97.8
98.4
96.6
99.5
100
2.3. Inhibition of IjB degradation in MT4 cells
91.2
98.4
Proteasome degradation of I
by tumor necrosis factor alpha (TNF
17 can inhibit proteasome degradation of I
treated with 17 (1.6 M) for 3 h before treating with TNF
j
B
a
is required for NF
). To determine if compound
, MT4 cells were
jB activation
a
a
Gradient of 88–95% of B in 22 min. Solvent A: 95% of acetonitrile/H2O and
0.045% of trifluoroacetic acid; Solvent B: acetonitrile/methanol/H2O = 85:10:5 and
0.045% of trifluoroacetic acid.
jBa
l
a
b
Gradient of 100% B in 30 min with same solvent system as described above.
Gradient of 90–100% B in 30 min with same solvent system as described above.
95% of acetonitrile/H2O in 30 min.
(10 ng/ml) for 30 min. The cytosolic proteins were analyzed in a
12% SDS–polyacrylamide gel. The proteins in the gel were trans-
c
d
ferred to a nitrocellulose paper for Western blot analyses. An I
monoclonal antibody (Abcam, Cambridge, UK) was used to detect
in the samples. TNF treatment resulted in a significant
reduction in I level when compared with the control in the ab-
sence of TNF (Fig. 3). The TNF -induced proteasome degradation
of I was significantly inhibited in the presence of 17. In con-
jBa
ProStar solvent delivery and PDA detector with Agilent Zorbax
ODS or C-8 columns (4.6 mmꢀ 25 cm and 9.4 mmꢀ 25 cm for ana-
lytical and semi preparative scales, respectively). The HPLC profile
of each new compound was obtained under two solvent systems
(Table 3).
IjBa
a
jBa
a
a
jBa
trast, b-actin level in the cells was not affected by 17. The results
suggested that 17 specifically inhibited the proteasome in the cells.
4.2. Procedures for coupling at the C-3 hydroxyl group
3. Conclusion
A mixture of GLA, anhydride (5 – 10 equiv), and DMAP (1 equiv)
in pyridine (anhydrous) was heated for 0.5–2 h at 100–130 °C in
microwave synthesizer. The mixture was then concentrated under
vacuum, re-dissolved in MeOH, and purified with Si-gel chroma-
tography or reverse phase HPLC. The yields ranged from 30% to
50%.
Alternatively, a solution of di-carboxylic acid (10–15 equiv),
DCC (5–10 equiv), and DMAP (1 equiv) in CH2Cl2 (anhydrous)
was added to GLA in pyridine and heated for 0.5–2 h at 100–
130 °C in microwave synthesizer. The yields ranged from 20% to
40%.
In summary, this study shows that GLA is a proteasome inhibi-
tor, and the potency of this inhibitory activity could be markedly
improved through chemical modification at the C-3 position of
the compound. GLA is a natural product and has abundant natural
sources. It can be obtained from the hydrolysis of the glycoside
glycyrrhizic acid (GL). GL from licorice has been used as a food
sweetener and in clinical treatment of HBV, where it appeared to
have low toxicity and side effects.22 GLA was found to be the major
component in serum after ingestion of GL.23 Therefore, develop-
ment of GLA derivatives as proteasome inhibitors might have the
potential to provide new therapeutics for diseases such as cancers
and inflammatory diseases.
4.2.1. 3b-O-(20,30-Dimethylsuccinyl)-18-b-glycyrrhetinic acid (3)
Positive FABMS m/z 599.5 (M+H)+; HR-FABMS calcd for
C
36H53O7 597.3791, found 597.3783. 1H NMR d 5.69 (1H, s, H-
4. Experimental
12), 4.52 (1H, dd, J = 4.5 Hz, J = 11.0 Hz, H-3), 2.74–2.89 (3H, m,
H2 and 2ꢀ CH–COO), 1.36 (3H, s, CH3-29), 1.24–1.26 (6H, m, 2ꢀ
a
4.1. General experimental procedures
CH3–CHCOO), 1.21 (3H, s, CH3-25), 1.16 (3H, s, CH3-23), 1.12 (3H,
s, CH3-24), 0.87, 0.88 (each 3H, each s, CH3-26 and CH3-27), 0.83
(3H, CH3-28).
The Biotage Initiator 2.0 was used for microwave-assisted syn-
thesis. All melting points were determined with a Fisher–Johns
melting point apparatus without correction. Positive and negative
FABHRMS were recorded on a Joel SX-102 spectrometer. 1H
(300 MHz) and 13C (125 MHz) NMR spectra were measured on a
Varian Mercury 300 spectrometer and Varian Inova 500 spectrom-
eter, respectively. Other than as noted, all samples were dissolved
in CDCl3 with TMS as internal standard. Silica gel chromatography
was carried out on a Biotage Horizon Flash chromatograph system
with pre-packed Si-gel column. HPLC was performed on a Varian
4.2.2. 3b-O-Succinyl-18-b-glycyrrhetinic acid (4)
Positive FABMS m/z 571.4 (M+H)+; HR-FABMS calcd for
C
34H49O7 569.3478, found 569.3465.%. 1H NMR (CDCl3/Py-d5) d
5.69 (1H, s, H-12), 4.52 (1H, dd, J = 5.7 Hz, J = 11.4 Hz, H-3), 2.75
(1H, d, J = 14.4 Hz, H2 ), 2.63 (4H, s, 2ꢀ CH2CO), 1.33 (3H, s, CH3-
a
29), 1.17 (3H, s, CH3-25), 1.10 (3H, s, CH3-23), 1.06 (3H, s, CH3-
24), 0.83 (6H, s, CH3-26 and CH3-27), 0.76 (3H, CH3-28).
4.2.3. 3b-O-(30,30-Dimethylglutaryl)-18-b-glycyrrhetinic acid (5)
Positive FABMS m/z 613.5 (M+H)+; HR-FABMS calcd for
C
37H55O7 611.3948, found 611.3930. 1H NMR (CDCl3/Py-d5) d
5.58 (1H, s, H-12), 4.48 (1H, dd, J = 5.2 Hz, J = 11.0 Hz, H-3), 2.64
(1H, d, J = 13.3 Hz, H2 ), 2.39–2.52 (4H, m, 2ꢀ COCH2), 1.29 (3H,
a
s, CH3-29), 1.23 (3H, s, CH3-25), 1.16 (3H, s, CH3-23), 1.14, 1.13
(each 3H, each s, [CH3]2C), 1.08 (3H, s, CH3-24), 0.83, 0.84 (each
3H, each s, CH3-26 and CH3-27), 0.68 (3H, CH3-28).
4.2.4. 3b-O-Glutaryl-18-b-glycyrrhetinic acid (6)
Positive FABMS m/z 585.5.5 (M+H)+; HR-FABMS calcd for
C
35H51O7 583.3635, found 583.3622. 1H NMR (CDCl3/Py-d5) d
Figure 3. The GLA derivative 17 inhibited degradation of IjB in MT4 cells.

6700
L. Huang et al. / Bioorg. Med. Chem. 16 (2008) 6696–6701
5.81 (1H, s, H-12), 4.61 (1H, dd, J = 5.5 Hz, J = 11.0 Hz, H-3), 2.91
(3H, s, CH3-23), 1.10 (3H, s, CH3-24), 0.80 (6H, s, CH3-26 and
CH3-27), 0.78 (3H, CH3-28).
(1H, d, J = 13.5 Hz, H2 ), 2.46–2.55 (4H, m, 2ꢀ COCH2), 2.01–2.12
a
(2H, m, CH2-30), 1.37 (3H, s, CH3-29), 1.26 (3H, s, CH3-25), 1.17
(3H, s, CH3-23), 1.09 (3H, s, CH3-24), 0.89 (6H, s, CH3-26 and
CH3-27), 0.80 (3H, CH3-28).
4.2.12. Glycyrrhetinic acid-m-phenylene monoacetate (15)
Positive FABMS m/z 647.5 (M+H)+; HR-FABMS calcd for
C
40H53O7 645.3791, found 645.3771. 1H NMR d 7.18–7.35 (5H, m,
4.2.5. 3b-O-Diglycolyl-18-b-glycyrrhetinic acid (7)
5ꢀ H-Ar), 5.67 (1H, s, H-12), 4.49 (1H, d, J = 4.2 Hz, H-3), 3.63
Positive FABMS m/z 587.4 (M+H)+; HR-FABMS calcd for
(4H, d, J = 11.0 Hz, 2ꢀ CH2-
U
), 2.76 (1H, d, J = 14.5 Hz, H2 ), 1.35
a
C
34H49O8 585.3427, found 585.3433. 1H NMR (CDCl3/Py-d5) d
(3H, s, CH3-29), 1.20 (3H, s, CH3-25), 1.13 (3H, s, CH3-23), 1.11
(3H, s, CH3-24), 0.81, 0.79 (each 3H, each s, CH3-26 and CH3-27),
0.76 (3H, CH3-28).
5.65 (1H, s, H-12), 4.54 (1H, dd, J = 5.5 Hz, J = 11.0 Hz, H-3), 4.19,
4.18 (each 2H, each s, 2ꢀ OCH2), 2.73 (1H, d, J = 13.5 Hz, H2 ),
a
1.27 (3H, s, CH3-29), 1.12 (3H, s, CH3-25), 1.04 (3H, s, CH3-23),
1.00 (3H, s, CH3-24), 0.78, 0.76 (each 3H, each s, CH3-26 and
CH3-27), 0.70 (3H, CH3-28).
4.2.13. Glycyrrhetinic acid-p-phenylene monoacetate (16)
Negative FABMS m/z 645.3 (MꢁH)ꢁ, HR-FABMS calcd for
C
40H53O7 645.3814, found 645.3791. 1H NMR d 7.26 (4H, s, 4ꢀ H-
4.2.6. 3b-O-Adipoyl-18-b-glycyrrhetinic acid (8)
Ar), 5.68 (1H, s, H-12), 4.53 (1H, m, H-3), 3.63 (4H, d, J = 15.5 Hz,
Negative FABMS m/z 597.3 (MꢁH)ꢁ, HR-FABMS calcd for
2ꢀ CH2-
U), 2.78 (1H, d, J = 13.0 Hz, H2 ), 1.36 (3H, s, CH3-29),
a
C
36H53O7 597.3791, found 597.3801. 1H NMR (CDCl3/Py-d5) d
1.21 (3H, s, CH3-25), 1.15 (3H, s, CH3-23), 1.12 (3H, s, CH3-24),
0.82 (6H, s, CH3-26 and CH3-27), 0.79 (3H, CH3-28).
5.71 (1H, s, H-12), 4.50 (1H, dd, J = 5.4 Hz, J = 11.5 Hz, H-3), 2.78
(1H, d, J = 13.5 Hz, H2 ), 2.32–2.34 (4H, m, 2ꢀ COCH2), 1.36 (3H,
a
s, CH3-29), 1.20 (3H, s, CH3-25), 1.13 (3H, s, CH3-23), 1.10 (3H, s,
CH3-24), 0.85 (6H, s, CH3-26 and CH3-27), 0.79 (3H, CH3-28).
4.2.14. Glycyrrhetinic acid monoisophthalate (17)
Negative FABMS m/z 617.3 (MꢁH)ꢁ, HR-FABMS calcd for
C38H49O7 617.3478, found 617.3483. 1H NMR (CDCl3/Py-d5) d 8.76
(1H, s, H-20-Ar), 8.28, 8.18 (each 1H, each d, J = 7.8 Hz, 2ꢀ H-Ar),
7.50 (1H, t, J = 7.8 Hz, H-50-Ar), 5.73 (1H, s, H-12), 4.78 (1H, dd,
4.2.7. Glycyrrhetinic acid propionate (10)
Negative FABMS m/z 525.3 (MꢁH)ꢁ, HR-FABMS calcd for
C
33H49O5 525.3580, found 525.3586. 1H NMR d 5.70 (1H, s, H-12),
J = 4.8 Hz, J = 11.0 Hz, H-3), 2.85 (1H, d, J = 14.0 Hz, H2 ), 1.38 (3H,
s, CH3-29), 1.20 (3H, s, CH3-25), 1.19 (3H, s, CH3-23), 1.12 (3H, s,
CH3-24), 1.04 (3H, s, CH3-26), 0.93 (3H, s, CH3-27), 0.80 (3H, CH3-28).
a
4.52 (1H, dd, J = 5.0 Hz, J = 11.0 Hz, H-3), 2.79 (1H, d, J = 13.5 Hz,
H2 ), 2.32 (2H, t, J = 7.5 Hz, CH2CH3), 1.37 (3H, s, CH3-29), 1.23
a
(3H, s, CH3-25), 1.17 (3H, s, CH3-23), 1.13 (3H, s, CH3-24), 1.13
(3H, d, J = 7.3 Hz, CH2CH3), 0.87, 0.88 (each 3H, each s, CH3-26
and CH3-27), 0.83 (3H, CH3-28).
4.2.15. Glycyrrhetinic acid monoterephthalate (18)
Negative FABMS m/z 617.3 (MꢁH)ꢁ, HR-FABMS calcd for
C
36H53O7 617.3478, found 617.3464. 1H NMR (CDCl3/Py-d5) d
4.2.8. Glycyrrhetinic acid isobutytate (11)
8.14, 8.05 (each 2H, each d, J = 8.5 Hz, 4ꢀ H-Ar), 5.71 (1H, s, H-
Negative FABMS m/z 539.3 (MꢁH)ꢁ, HR-FABMS calcd for
12), 4.74 (1H, dd, J = 4.5 Hz, J = 10.0 Hz, H-3), 2.83 (1H, d,
C
34H51O5 539.3765, found 539.3745. 1H NMR d 5.71 (1H, s, H-
J = 14.0 Hz, H2 ), 1.35 (3H, s, CH3-29), 1.19 (3H, s, CH3-25),
a
12), 4.50 (1H, dd, J = 4.4 Hz, J = 11.1 Hz, H-3), 2.79 (1H, d,
1.16(3H, s, CH3-23), 1.09 (3H, s, CH3-24), 0.99 (3H, s, CH3-26),
0.91 (3H, s, CH3-27), 0.78 (3H, CH3-28).
J = 13.6 Hz, H2 ), 2.54 (1H, qui, J = 7.0 Hz, CH–[CH3]2), 1.37 (3H, s,
a
CH3-29), 1.23 (3H, s, CH3-25), 1.18, 1.16 (each 3H, each d.
J = 6.7 Hz, [CH3]2CH), 1.17 (3H, s, CH3-23), 1.13 (3H, s, CH3-24),
0.89, 0.87 (each 3H, each s, CH3-26 and CH3-27), 0.84 (3H, CH3-28).
4.3. Biological assays
4.3.1. Proteasome assay
4.2.9. 3b-O-(trans-b-Hydromuconyl)-18-b-glycyrrhetinic acid
(12)
Proteasome assay kits were purchased from Calbiochem, San
Diego, CA. The effect of GLA and its analogs on the 20S proteasome
activity was assayed following the protocol provided by the man-
ufacturer. The major components of the assay mixture are human
20S proteasomes, fluorogenic peptide substrates and the protea-
some activator PA28. The assay was designed to measure hydroly-
sis of the fluorogenic substrates Suc-Leu-Leu-Val-Tyr-AMC, (Z)-
LLE-bNA, and Bz-VGR-AMC in the presence of the proteasome acti-
vator PA28. Suc-Leu-Leu-Val-Tyr-AMC is frequently used to detect
the chymotrypsin-like activity of 20S proteasomes. The trypsin-
like and caspase-like activities of the 20S proteasome were deter-
mined using the fluorogenic substrates Bz-VGR-AMC and (Z)-LLE-
bNA, respectively. Fluorescence generated from the proteolytic
reaction in the presence of various concentrations of GLA or its
analogs was measured using a Bio-Tek fluorometer (Winooski,
Vermont).
Negative FABMS m/z 595.3 (MꢁH)ꢁ, HR-FABMS calcd for
C
36H51O7 595.3635, found 595.3649. 1H NMR d 5.72 (1H, d,
J = 5.0 Hz, C@H-30), 5.69 (2H, s, C@H-20 and H-12), 4.52 (1H, dd,
J = 4.5 Hz, J = 11.5 Hz, H-3), 3.14, 3.09 (each 1H, each d, J = 5.4 Hz,
CH2-COO), 2.80 (1H, d, J = 14.0 Hz, H2 ), 1.37 (3H, s, CH3-29),
a
1.22 (3H, s, CH3-25), 1.16 (3H, s, CH3-23), 1.12 (3H, s, CH3-24),
0.87, 0.86 (each 3H, each s, CH3-26 and CH3-27), 0.83 (3H, CH3-28).
4.2.10. 3b-O-Benzoyl-18-b-glycyrrhetinic acid (13)
Negative FABMS m/z 573.3 (MꢁH)ꢁ, HR-FABMS calcd for
C
37H49O5 597.3592, found 573.3580. 1H NMR d 8.04 (2H, d,
J = 7.0 Hz, 2ꢀ H-o-Ar), 7.55 (1H, t, J = 7.2 Hz, H-p-Ar), 7.44 (2H, t,
J = 7.2 Hz, 2ꢀ H-m-Ar), 5.73 (1H, s, H-12), 4.76 (1H, dd, J = 5.0 Hz,
J = 10.8 Hz, H-3), 2.85 (1H, d, J = 13.5 Hz, H2 ), 1.40 (3H, s, CH3-
a
29), 1.23 (3H, s, CH3-25), 1.22(3H, s, CH3-23), 1.15 (3H, s, CH3-
24), 1.04 (3H, s, CH3-26), 0.95 (3H, s, CH3-27), 0.85 (3H, CH3-28).
For proteasome inhibition, various concentrations of GLA deriv-
atives were tested in the presence of 16 lg/ml of PA28. The known
proteasome inhibitors LLM-F (Boston Biochem, Cambridge, MA,
USA) and lactacystin (Sigma–Aldrich, St. Louis, MO) were used as
controls for the proteasome inhibition assays. The 50% inhibitory
concentration (IC50) is defined as the inhibitory concentration that
4.2.11. Glycyrrhetinic acid-4-acetylphenoxyl acetate (14)
Positive FABMS m/z 661.4 (M+H)+, HR-FABMS calcd for
C
41H57O7 661.4104, found 661.4097. 1H NMR d 7.92, 6.92, (each
2H, each d, J = 9.0 Hz, 4ꢀ H-Ar), 5.68 (1H, s, H-12), 4.63 (1H, dd,
reduces the reaction rate by 50%. The velocity of reaction,
(360/460 nm)/min, was plotted against the log-concentration of
the inhibitor to determine the IC50
DRFU
J = 5.5 Hz, J = 11.0 Hz, H-3), 2.79 (1H, d, J = 14.0 Hz, H2 ), 2.54
a
(3H, s, COCH3), 1.34 (3H, s, CH3-29), 1.20 (3H, s, CH3-25), 1.12
.

L. Huang et al. / Bioorg. Med. Chem. 16 (2008) 6696–6701
6701
4.3.2. Cell-based proteasome assay
References and notes
To determine the effect of GLA derivatives on proteasomes in
cells, a Promega cell-based assay was used in this study. MT4 cells
(4000 cells) were treated with GLA derivatives or known protea-
some inhibitors in serum-free medium at 37 °C for 3 h. MT4 cells
are human T cells isolated from a patient with adult T-cell leuke-
mia. MT4 cells were obtained from the NIH AIDS Research and Ref-
erence Reagent Program. The drug-treated MT4 cells were
incubated with the Promega Proteasome-Glo Cell-Based Assay Re-
agent (Promega Bioscience, Madison, WI) for 10 min. The chymo-
trypsin-like proteasome activity was detected as the relative light
unit (RLU) generated from the cleaved substrate in the reagent.
Luminescence generated from each reaction condition was de-
tected with a Perkin-Elmer Victor-3 luminometer (Shelton, CT,
USA).
1. Dalton, L. C&EN. 2002, 80, 37.
2. Davis, E.; Morris, D. Mol. Cell. Endocrinol. 1991, 78, 1–6.
3. Monder, C.; Stewart, P. M.; Lakshmi, V.; Valentino, R.; Burt, D.; Edwards, C. R.
W. Endocrinolgy 1989, 125, 1046–1053.
4. Buhler,H.; Perschel,F. H.;Hierholzer,K.Biochim.Biophys. Acta1991, 1075, 206–212.
5. Salvi, M.; Fiore, C.; Armanini, D.; Toninello, A. Biochem. Pharmacol. 2003, 66,
2375–2379.
6. Ciechanover, A. Nat. Rev. Mol. Cell Biol. 2005, 6, 79–86.
7. Hershko, A.; Ciechanover, A. Annu. Rev. Biochem. 1998, 67, 425–479.
8. Nalepa, G.; Rolfe, M.; Harper, J. W. Nat. Rev. Drug Disc. 2006, 5, 596–613.
9. Rechsteiner, M.; Hill, C. P. Trends Cell Biol. 2005, 15, 27–33.
10. Groll, M.; Heinemeyer, W.; Jager, S.; Ullrich, T.; Bochtler, M.; Wolf, D. H.; Huber,
R. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 10976–10983.
11. Kane, R. C.; Bross, P. F.; Farrell, A. T.; Pazdur, R. Oncologist 2003, 8, 508–513.
12. Kisselev, A. F.; Callard, A.; Goldberg, A. L. J. Biol. Chem. 2006, 281, 8582–8590.
13. Huang, L.; Ho, P.; Chen, C.-H. FEBS Lett. 2007, 581, 4955–4959.
14. Heinemeyer, W.; Kleinschmidt, J. A.; Saidowsky, J.; Escher, C.; Wolf, D. H. EMBO
J. 1991, 10, 555–562.
15. Gueckel, R.; Enenkel, C.; Wolf, D. H.; Hilt, W. J. Biol. Chem. 1998, 273, 19443–19452.
16. Richter-Ruoff, B.; Wolf, D. H.; Hochstrasser, M. FEBS Lett. 1994, 354, 50–52;
Richter-Ruoff, B.; Wolf, D. H.; Hochstrasser, M. FEBS Lett. 1995, 358, 104.
17. Seufert, W.; Jentsch, S. EMBO J. 1992, 11, 3077–3080.
18. Yu, D.; Sakurai, Y.; Chen, C. H.; Chang, F. R.; Huang, L.; Kashiwada, Y.; Lee, K. H.
J. Med. Chem. 2006, 49, 5462–5469.
19. Goldberg, A. L. Biochem. Soc. Trans. 2007, 35, 12–17.
20. Zhang, J.; Cheng, Z.-H.; Yu, B.-Y.; Cordell, G. A.; Qiu, S. X. Tetrahedron Lett. 2005,
46, 2337–2340.
21. Dick, L. R.; Cruikshank, A. A.; Grenier, L.; Melandri, F. D.; Nunes, S. L.; Stein, R. L.
J. Biol. Chem. 1996, 271, 7273–7276.
22. Sato, H.; Goto, W.; Yamamura, J.; Kurokawa, M.; Kageyama, S.; Takahara, T.;
Watanabe, A.; Shiraki, K. Antiviral Res. 1996, 30, 171–177.
23. Yamamura, Y.; Kawakami, J.; Santa, T.; Kotaki, H.; Uchino, K.; Sawada, Y.;
Tanaka, N.; Iga, T. J. Pharm. Sci. 1992, 81, 1042–1046.
Acknowledgments
The authors thank Shannon Meroney-Davis for her help on
editing this manuscript. The authors are grateful to Dr. George
Dubay of the Department of Chemistry and Dr. Anthony Ribeiro
of the NMR Spectroscopy Center of Duke University for their
help and assistance on mass and NMR spectroscopy data collec-
tion.. This work was supported in part by grant from Centers
for AIDS Research at Duke University (LH), the National Insti-
tutes of Health Grants AI65310 (CHC), and CA17625 and
AI33066 (KHL).