Garcinia xanthones as orally active antitumor agents

Article abstract of DOI:10.1021/jm301593r

Using a newly developed strategy whose key step is the regioselective propargylation of hydroxyxanthone substrates, 99 structurally diverse Garcinia natural-product-like xanthones based on gambogic acid were designed and synthesized and their in vitro antitumor activity was evaluated. A set of 40 related compounds was chosen for determination of their physicochemical properties including polar surface area, log D_7.4, aqueous solubility, and permeability at pH 7.4. In the light of the in vitro antitumor activity and the physicochemical properties, two compounds were advanced into in vivo efficacy experiments. The antitumor activity of compound 112, administered po, showed more potent in vivo oral antitumor activity than gambogic acid.

Full text of DOI:10.1021/jm301593r

Article
pubs.acs.org/jmc
Garcinia Xanthones as Orally Active Antitumor Agents
Xiaojin Zhang,†,‡,∥ Xiang Li,‡ Haopeng Sun,*,‡,§ Xiaojian Wang,# Li Zhao,⊥ Yuan Gao,⊥ Xiaorong Liu,§
Shenglie Zhang,‡,§ Yanyan Wang,‡,§ Yingrui Yang,‡,§ Su Zeng,∇ Qinglong Guo,*,⊥ and Qidong You*,†,‡
‡
§
†State Key Laboratory of Natural Medicines, Jiangsu Key Laboratory of Drug Design and Optimization, Department of Medicinal
∥
⊥
Chemistry, School of Pharmacy, Department of Organic Chemistry, School of Science, and Jiangsu Key Laboratory of
Carcinogenesis and Intervention, China Pharmaceutical University, Nanjing, 210009, China
#Institute of Materia Medica, Chinese Academy of Medical Science and Peking Union Medical College, Peking, 10005, China
∇Department of Pharmaceutical Analysis and Drug Metabolism, College of Pharmaceutical Sciences, Zhejiang University, Hangzhou,
310058, China
S
* Supporting Information
ABSTRACT: Using a newly developed strategy whose key step is the regioselective propargylation of hydroxyxanthone
substrates, 99 structurally diverse Garcinia natural-product-like xanthones based on gambogic acid were designed and synthesized
and their in vitro antitumor activity was evaluated. A set of 40 related compounds was chosen for determination of their
physicochemical properties including polar surface area, log D7.4, aqueous solubility, and permeability at pH 7.4. In the light of
the in vitro antitumor activity and the physicochemical properties, two compounds were advanced into in vivo efficacy
experiments. The antitumor activity of compound 112, administered po, showed more potent in vivo oral antitumor activity than
gambogic acid.
oxa-tricyclo[4.3.1.03,7]decan-2-one scaffold with a common
INTRODUCTION
■
xanthone backbone.5 This motif is further elaborated by
Natural products (NP) are a unique source of diverse bioactive
substitutions on the aromatic residue and peripheral oxidations
to produce a variety of structurally diverse caged Garcinia
lead compounds for drug discovery. More than 50% of the
drugs developed in the pharmaceutical industry over the past
30 years are NPs or inspired by NP structures.1,2 In the case of
xanthones shown in Figure 1, including gambogic acid (GA,
1),6,7 gambogellic acid (2),8 desoxymorellin (3),9 gaudichau-
anticancer drugs, the proportion is even higher; 74.8% of the
dione A (4),10 gaudichaudione G (5),10 and isobractatin (6).11
anticancer drugs approved worldwide between 1940 and
December 2010 owe their origins to NPs.1 The evolutionary
Among these, 1 is the most prominent and representative
example. Preclinical investigation identified 1 as a potent
selection processes has resulted in complex bioactive NPs with
antitumor agent that inhibits the growth of a broad panel of
diverse structures that are optimized to interact with various
cancer cell lines both in vitro and in vivo.12−15 Biological
biological systems. Due to their coevolution with the target
sites in biological systems, the quality of leads arising from NP
studies revealed that the antitumor efficacy of 1 arises through
multiple mechanisms involving apoptosis induction,16,17 cell
scaffolds is considered to be superior and often more
cycle regulation,18 angiogenesis inhibition,19,20 tumor cell
adhesion inhibition, and metastasis prevention.21 Biomolecular
targets of GA that have been reported include the bcl-2
family,22 p53/mdm2 complex,16 cytosolic thioredoxins,23
topoisomerase II,24 survivin,25 transferring receptor,26,27 heat
biologically friendly.3 NP scaffolds are in fact privileged
templates for rapid identification of promising new chemical
entities for therapeutic applications.
The gambogin resin secreted by the Garcinia genus of
tropical plants has been used as folk medicine for centuries in
Southeast Asia.4 Efforts to identify the bioactive constituents of
gambogin resin have revealed a growing family of natural
products known as “caged xanthones” based on their unique 4-
Received: October 29, 2012
Published: November 20, 2012
© 2012 American Chemical Society
276
dx.doi.org/10.1021/jm301593r | J. Med. Chem. 2013, 56, 276−292

Journal of Medicinal Chemistry
Article
Figure 1. Caged Garcinia xanthones with conservative D caged ring and structurally diverse AB-ring.
Figure 2. Designed Garcinia natural-product-like caged xanthones with a pharmacophoric caged core.
shock protein 90,28 stathmin,29 and IκBα kinase-β protein,30,31
which could explain some aspects of the mechanism of action.
In addition, GA showed minimal toxicity to normal cells and
has a potentially useful therapeutic window in cancer
treatment.14,15
exploration that has been reported of the planar ring B and
the caged ring D in simplified Garcinia analogues, we were
prompted to examine the Garcinia natural-product-like
xanthones shown in Figure 2. These compounds have diverse
structures based on the pharmacophoric caged core associated
with antitumor efficacy, and we conducted both detailed
structure−activity relationship (SAR) and structure−property
relationship (SPR) studies. Our synthetic strategy was mainly
based upon the regioselective propargylation of diverse
hydroxyxanthone substrates followed by a Claisen/Diels−
Alder cascade reaction to generate different planar B-rings
and caged D-rings. All the Garcinia natural-product-like caged
xanthones have been tested in vitro for antitumor activities
against the human hepatocellular carcinoma HepG2 cell line,
the human lung carcinoma A549 cell line, and the human
glioblastoma U251 cell line. Forty representative compounds
have been evaluated for intrinsic aqueous solubility, polar
surface area (PSA), log D, and permeability at pH 7.4 to
identify the compounds with potential to be advanced into in
vivo efficacy experiments. The compounds with effective in vivo
antitumor activity have been evaluated for their in vivo
Unfortunately, some perceived disadvantages of GA limit its
clinical application. These include limited natural sources and
complexities associated with its isolation in approximately 5%
yield from the gamboges resin,6,32 difficulties in the total
synthesis,33 very low aqueous solubility (about 0.01 mg/mL),34
and considerable loss of potency when administered orally.
Although several simpler analogues such as cluvenone (34) and
1-hydroxycluvenone (33) have been synthesized and reported
to possess the cytotoxicity exhibited by GA against some cancer
cell lines,35−38 these compounds still lacked the druglike
properties optimal for further preclinical development.
Preliminary SAR studies indicated that the α,β-unsaturated
ketone moiety of CD caged region13 in GA as well as its BC
planar region are both essential for the antitumor activity of
GA,31,37−39 and the peripheral moieties are suitable as sites for
diverse modification.13,32,40−44 In view of the limited
277
dx.doi.org/10.1021/jm301593r | J. Med. Chem. 2013, 56, 276−292

Journal of Medicinal Chemistry
Article
a
Scheme 1. Substrates Tested in the Propargylation Reaction and Their Probable Products
a
Reagents and conditions: (a) 3-chloro-3-methylbut-1-yne (5 equiv for substrate 7, 2.5 equiv for substrates 8 and 9), K2CO3 (3.3 equiv for substrate
7, 2.2 equiv for substrates 8 and 9), KI, 10 mol % CuI. For detailed reaction conditions see Table 1.
Table 1. Regioselectivity in Propargylation of Hydroxyxanthone Substrates
entry
substrate
solvent
Me2CO
T (°C)
t (h)
products (% yield)
substrate recovery (%)
1
7
7
7
7
7
7
7
7
7
7
7
7
8
8
9
9
reflux
reflux
35
6
2
6
6
6
6
2
6
6
6
6
6
6
6
6
6
7a (0)
7b (25)
7b (10)
7b (0)
7b (0)
7b (0)
7b (0)
7b (7)
7b (0)
7b (0)
7b (0)
7b (0)
7b (0)
8b (0)
8b (0)
9b (80)
9b (0)
7c (61)
7c (41)
7c (27)
7c (0)
7c (1)
7c (13)
7c (47)
7c (0)
7c (0)
7c (9)
7c (8)
7c (3)
8c (88)
8c (2)
9c (11)
9c (3)
0
2
Me2CO
7a (12)
7a (15)
7a (0)
28
50
98
82
60
28
99
95
50
43
82
0
3
Me2CO
4
Me2CO
20
5
CH3CN
20
7a (7)
6
CH3CN
35
7a (22)
7a (10)
7a (0)
7
CH3CN
50
8
THF
20
9
THF
35
7a (2)
10
11
12
13
14
15
16
CH3CH/THF (4:1)
CH3CH/THF (2:1)
CH3CH/THF (1:1)
Me2CO
35
7a (35)
7a (41)
7a (10)
8a (0)
35
35
reflux
35
CH3CN/THF (2:1)
Me2CO
8a (42)
9a (0)
50
0
reflux
35
CH3CN/THF (2:1)
9a (33)
52
pharmacokinetic parameters to determine which of them are
potentially orally active. In this paper, we report the preparation
and thorough examination of a library of 99 Garcinia natural-
product-like caged xanthones and the discovery of novel
Garcinia caged xanthones as orally active antitumor agents.
improve the druglike properties. This strategy was based on the
regioselective propargylation of the OH group at C6 of the 5,6-
dihydroxyxanthone substrates, leaving the C5-OH group free
for introduction of various allyl substituents to form diverse
rearrangement precursors producing diverse target caged
structures.
The 5,6-dihydroxyxanthones 7−9 (Scheme 1) were the
starting points for the propargylation reactions, which could be
conveniently effected according to our previously reported
methods.38,48 Using of 3-chloro-3-methylbut-1-yne in the
presence of KI and K2CO3 and CuI as catalyst was essential
for this alkylation.47 Refluxing substrate 7 in Me2CO for 6 h led
to alkylation, providing cyclized product 7c and the C5, C6-OH
propargylated product 7b but none of the desired C6-OH
propargylated product 7a (Table 1, entry 1). Although no 7a
was isolated at the end of the reaction, the formation of 7c
suggested that 7a might arise during the reaction because 7c
probably formed from 7a under CuI catalysis. With lower
temperatures (35 °C) (Table 1, entry 2) or shorter reaction
RESULTS AND DISCUSSION
■
Synthetic Chemistry. The unique 4-oxatricyclo[4.3.1.03,7]-
decan-2-one caged structure of these Garcinia natural-product-
like xanthones can be synthesized using the well-established
Claisen rearrangement/intramolecular Diels−Alder cascade
reaction,45−51 and the structural diversity of the target Garcinia
caged xanthone analogues depends on that of the precursors in
the rearrangement. However, previous synthetic approaches
were limited in practice to provide two main chemical types of
rearrangement precursors: bis-allyl and bis(2-methylbut-3-enyl)
compounds.33 In the present study, a new synthetic strategy
was developed to generate structural diversity in both the caged
and planar regions to obtain detailed SAR and SPR data and to
278
dx.doi.org/10.1021/jm301593r | J. Med. Chem. 2013, 56, 276−292

Journal of Medicinal Chemistry
Article
a
Scheme 2. Synthesis of Caged Xanthone Analogues 18−23 Missing One gem-Dimethyl Group in the Caged Region
a
Reagents and conditions: (a) H2, Pd/BaSO4, EtOH, 25 °C, 20 min, 82−89%; (b) H2, Pd/BaSO4, EtOH, 25 °C, 40 min, 90%; (c) allyl bromide,
K2CO3, Me2CO, 35 °C, 4 h, 79−87%; (d) DMF, 120 °C, 2 h, 71−86%.
a
Scheme 3. Synthesis of Caged Xanthone Analogues 28−34 with Two gem-Dimethyl Groups in the Caged Region
a
Reagents and conditions: (a) 3-chloro-3-methylbut-1-yne, K2CO3, KI, 10% CuI, Me2CO, reflux, 30 min, 85%; (b) H2, Pd/BaSO4, EtOH, 25 °C, 40
min, 80−89%; (c) DMF, 120 °C, 2 h, 57−82%.
times (Table 1, entry 3), relatively low yields of the C6-OH
propargylated product 7a were obtained, but it was not
produced by the reaction carried out at 20 °C in Me2CO
(Table 1, entry 4). Different solvents such as CH3CN, THF,
and their mixtures were also evaluated in order to improve the
yield and selectivity of 7a (Table 1, entries 5−12), and it was
found that in CH3CN, 7a was formed even at 20 °C (Table 1,
entry 5). A significant proportion of 7 was not converted to 7a
in this reaction as a result of its low solubility in CH3CN at
such relatively low reaction temperatures (Table 1, entries 5, 6),
and increases in reaction temperature led to an increased ratio
of the cyclized product 7c. Meanwhile, the solvent THF, in
which the solubility of 7 was enhanced, was found to seriously
hinder the propargylation process (Table 1, entries 8, 9).
Subsequently, use of CH3CN/THF mixtures (v:v = 2:1) as
solvent was found to give an improved yield (41%) of 7a with
279
dx.doi.org/10.1021/jm301593r | J. Med. Chem. 2013, 56, 276−292

Journal of Medicinal Chemistry
Article
acceptable regioselectivity between C6-OH and C5-OH groups
(Table 1, entry 11). In a similar manner, using CH3CN/THF
as solvent at 35 °C, substrates 8 and 9 selectively provided the
C6-OH monopropargylated products 8a and 9a, respectively
(Table 1, entries 14, 16), but no C6-OH monopropargylated
products were observed when refluxing 7, 8, or 9 in Me2CO
(Table 1, entries 13, 15). Despite the moderate C6-OH
propargylation yields (33−42%) and the recovery of reactants,
this strategy offered the most expedient selective route to the
C6-OH propargylated products 7a−9a, the key intermediates
to generate the diverse target Garcinia caged xanthones.
diallylation to form 35 and, subsequently, Claisen/Diels−Alder
reaction in decalin at 180 °C (Scheme 4).
Subsequently, a number of D-ring-modified caged xanthones
(46−130) were designed to probe the caged region and to
determine the optimal substitution for antitumor activity as well
as druglike properties (Scheme 5). Compound 10 was allylated
with (E)-1,4-dibromobut-2-ene in the presence of K2CO3 in
Me2CO, generating the intermediate 37, derivatization of which
with various aliphatic or aromatic amines, alcohols, or acids in
the presence of K2CO3 in DMF led to the precursor 38, which
underwent the Claisen/Diels−Alder reaction to provide the
caged xanthones 46−52 and 53−59 as isomer pairs, with a
pyran ring in the planar region and a modified caged region
(Scheme 5). Similarly, D-ring-modified caged xanthones 60−81
and 82−103 were obtained from compound 12, which
possessed a prenyl group in its planar region (Scheme 5).
Treatment of 12 with various acrylic acids in the presence of 1-
ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride
(EDC·HCl) and 4-dimethylaminopyridine (DMAP) in DCM
led to the ethers 41a−d. Heating precursors 41a−d in DMF at
120 °C produced the caged xanthones 104−107 with a
carbonyl group in the caged region, by replacing the gem-
dimethyl group and a C4 substituted prenyl group in the planar
region (Scheme 5). In this case, hardly any of the C2
prenylated isomer was isolated. Furthermore, by use of
compounds 14 and 16 as synthetic entries instead of 12, the
D-ring-modified caged xanthones 108−130 with a simple
planar region were similarly prepared as a single isomer
(Scheme 5).
Next, a set of caged xanthones varying in both the planar and
caged regions was synthesized to probe the substituent groups
around the pharmacophoric caged core (Schemes 2−4). With
Scheme 4. Synthesis of Caged Xanthone Analogues 36
Missing the Two gem-Dimethyl Groups in the Caged
a
Region
a
Reagents and conditions: (a) allyl bromide, K2CO3, Me2CO, 35 °C, 4
h, 95%; (b) decalin, 180 °C, 3 h, 51%.
Antiproliferative Activities. The antiproliferative activity
of 1 and the 99 synthetic Garcinia natural-product-like caged
xanthones were assessed with the tetrazolium-based MTT assay
using the human hepatocellular carcinoma HepG2 cell line, the
human lung carcinoma A549 cell line, or the human
glioblastoma U251 cell line. Cancer cells were treated with
increasing concentrations of the compounds for 24 h, and then
the cell viability was evaluated through measurements of
mitochondrial dehydrogenase activity. The antiproliferative
activities, expressed as IC50 values, are summarized in Tables
2−6. Compound 1 was used as positive control for the in vitro
assay and exhibited IC50 values of 0.89, 1.10, and 2.56 μM
against HepG2, A549, and U251 cells, respectively. In general,
most of the caged xanthones showed potent antiproliferative
activities against HepG2 and A549 cells, with IC50 values at low
micromolar levels. All the compounds tested were found to be
less sensitive to U251 cells than to HepG2 and A549 cells.
Among the 99 synthetic caged xanthones, compound 33
showed the most potent inhibitory activities in vitro against
HepG2 and A549 cells, with an IC50 value of 0.81 and 0.70 μM,
respectively, while compound 78 was the most cytotoxic against
U251 cells, with an IC50 value of 6.56 μM.
Structure−Activity Relationship Studies. First, a set of
caged xanthones (Table 2) varying in both the planar region
(seven structural types) and caged region (four structural
types) was designed to gain an understanding of the SAR
around the pharmacophoric caged core and especially to
explore the optimal structure of the caged region for antitumor
activity. As shown in Table 2, all the compounds seemed to be
more active toward HepG2 and A549 cells but less active or
even not active toward U251 cells at similar concentrations. In
the cases of HepG2 and A549 cells, generally, compounds with
type A (containing two gem-dimethyl groups) and type B
(containing one gem-dimethyl group) caged regions were found
the C6-OH propargylated xanthones 7a−9a in hand, caged
xanthone analogues 18−23 missing one gem-dimethyl group in
the caged region were synthesized through routes similar to
that in Scheme 2. Regioselective monohydrogenation of the
acetylenic moiety at C6 in 7a in the presence of Lindlar catalyst
provided the corresponding olefin 10 by curtailing the reaction
time to 20 min. With further prolongation of the reaction time
to 40 min, both the C5 and C6 acetylenic moieties in 7a were
reduced to the olefin 12. Subsequent alkylation of 10 and 12 in
the presence of allyl bromide and K2CO3 gave rise to 11 and
13, respectively. Drawing on previous experience with related
structures50 we heated 11 in DMF at 120 °C utilizing the
Claisen/Diels−Alder/Claisen reaction and obtained the caged
xanthones 18 and 19 as isomers with a pyran ring merged into
planar region, that at the C2 and C4 positions, respectively
(Scheme 2). Meanwhile, in the case of 13, caged xanthones 20
and 21 were formed as isomers with a prenyl group at the C2
and C4 positions, respectively (Scheme 2). In a similar manner,
intermediates 8a and 9a underwent Lindlar reduction and
allylation followed by Claisen/Diels−Alder cascade reaction to
form, respectively, the caged xanthones 22 and 23 with a simple
planar region (Scheme 2). On the basis of the caged compound
28 previously developed in our laboratory,49 the caged
xanthone analogues 29 and 30, with a pyran ring in the planar
region and both the gem-dimethyl groups in the caged region,
were obtained by propargylation of the C3-OH group using 3-
chloro-3-methylbut-1-yne followed by Claisen rearrangement
cyclization (Scheme 3). Caged xanthones 31 and 32
(forbesione) were prepared as shown in Scheme 3 according
to published reactions,31,47 and compounds 33 (cluvenone)
and 34 (1-hydroxycluvenone) were prepared according to our
previously reported procedure (Scheme 3).38 In addition, caged
xanthone 36, from which the two gem-dimethyl group in the
caged region are absent, was prepared from the xanthone 8 by
280
dx.doi.org/10.1021/jm301593r | J. Med. Chem. 2013, 56, 276−292

Journal of Medicinal Chemistry
Article
a
Scheme 5. Synthesis of D-Ring-Modified Caged Xanthone Analogues 46−130 with Diverse Planar Region
a
Reagents and conditions: (a) (E)-1,4-dibromobut-2-ene, K2CO3, Me2CO, 35 °C, 4 h, 74−77%; (b) different amide or alcohol or acid, K2CO3,
DMF, 35 °C, 1 h, 99−100%; (c) substituted acrylic acid, EDC·HCl, DMAP, DCM, 25 °C, 2 h, 71−80%; (d) DMF, 120 °C, 2 h, 24−32% for 46−52,
35−43% for 53−59, 20−32% for 60−81, 30−44% for 82−103, 31−37% for 104−107, and 51−70% for 108−130. For the detailed R groups of
compounds 46−130, see Tables 3−6.
to be more active than that with type C (containing one gem-
dimethyl group and one carbonyl group) or type D caged
structures (missing two gem-dimethyl groups). For instance,
compounds 22 and 33 possessed a low micromolar activity
(0.70−3.79 μM) against HepG2 and A549 cells and were much
more potent than compound 36, which contains the same
planar region. Similarly, compared with compounds 21 and 32,
compound 104, with a carbonyl group in the caged region, was
about 10−20-fold less active against A549 cells and inactive at
100 μM against HepG2 and U251 cells. Meanwhile,
compounds 18−21 with type A caged structure displayed
antitumor activity similar to that of the corresponding
compounds 29−32 with type B caged moiety that lacked a
gem-dimethyl group. The planar region with diverse structures
including types A−G was found to be tolerated, whereas type G
without any substituents contributed the least to their
antitumor activity. Compound 22 for example, bearing a C1
hydroxyl group on the B-ring had IC50 values of 2.65 and 3.79
μM against HepG2 and A549 cells, respectively, while
compounds 18−21, containing the same caged structure as
22, possessed similar IC50 values ranging from 2.04 to 8.43 μM.
On the other hand, compound 23, formed by the removal of
the C1 hydroxyl group, when compared to 22, was found to be
almost 4−10-fold less active. Similar decreases in cytotoxicity
were also observed between compounds 33 and 34. These
results suggest that (a) the gem-dimethyl group directly linked
to the caged scaffold is not necessary for the bioactivity, (b)
replacement of the gem-dimethyl group linked to the caged
scaffold with a carbonyl group leads to substantial loss of
activity, (c) the gem-dimethyl group of the prenyl moiety in the
caged region contributes significantly to the antitumor activity,
and (d) the prenylation and pyran fusion in the B-ring planar
region are well-tolerated and the C1 hydroxyl group on B-ring
benefits the cytotoxicity.
On the basis of the above SAR around the pharmacophoric
caged core, a large set of D-ring-modified caged xanthones 46−
281
dx.doi.org/10.1021/jm301593r | J. Med. Chem. 2013, 56, 276−292

Journal of Medicinal Chemistry
Article
Table 2. SAR of the Planar Region and gem-Dimethyl Group in the Caged Region
a
structure
IC50 (μM)
compd
planar region
caged region
HepG2
A549
U251
18
19
20
21
22
23
29
30
31
32
28
33
34
104
36
type A
type B
type C
type D
type F
type G
type A
type B
type C
type D
type E
type F
type G
type D
type F
type A
type A
type A
type A
type A
type A
type B
type B
type B
type B
type B
type B
type B
type C
type D
2.99 0.24
4.40 0.14
3.46 0.23
3.13 0.33
2.65 0.21
12.4 1.1
2.00 0.07
4.72 0.30
6.27 0.58
4.07 0.09
3.39 0.11
0.81 0.04
6.77 0.46
>100
2.00 0.27
8.43 0.52
3.78 0.25
2.04 0.09
3.79 0.50
36.1 3.4
2.12 0.18
3.15 0.13
14.4 1.6
3.58 0.22
2.67 0.39
0.70 0.09
1.93 0.36
37.9 4.6
18.7 3.8
12.2 1.08
8.29 0.76
18.7 2.1
14.3 1.9
23.9 2.0
30.0 2.9
9.62 1.44
9.01 0.66
22.4 2.5
8.57 0.94
7.32 0.87
9.05 1.17
18.7 2.2
>100
9.14 1.33
51.7 4.3
a
1 is the internal control for in vitro assays: HepG2, 0.89 0.08 μM; A549, 1.10 0.12 μM; and U251, 2.56 0.25 μM.
130 (Table 3-6) with different planar regions was designed and
synthesized to further explore in detail the caged region with
various substituents and to improve the druglike properties by
introducing hydrophilic groups.
with a pyran ring in the planar region, the introduction of
additional amino groups to the D caged ring contributed little
to improvement of the cytotoxicity.
b. Garcinia Natural-Product-Like D-Ring-Modified Caged
Xanthones 60−81 with a C2 Prenyl Group in the Planar
Region. To gain insight on how modification of the D caged
ring can affect cytotoxicity in compounds with a planar region
featuring a type C structure, HepG2, A549, and U251 cells
were incubated with the designed compounds 60−81. As
shown in Table 4, these compounds were generally more
potent than the compounds mentioned above with pyran ring
in the planar region. They were also found to be less sensitive
to U251 cells than to HepG2 and A549 cells. In HepG2 cells,
compared with compound 20, seven compounds (60, 63−65,
69, 72, 81) had increased potency with IC50 values ranging
from 1.74 to 3.16 μM. These compounds all had an alkylamino
group substitution such as pyrrolidinyl, piperazinyl, methyl-
piperazinyl, 4-oxopiperidinyl, (n-Pr)2N, (HOCH2CH2)2N, or
phenylpiperazinyl groups. Another set of compounds (61−62,
67−68, 71) with alkylamino group substitution had activities
comparable to that of 20. Introduction of an i-PrO- group, as in
compound 73, gave a similar cytotoxicity when compared to
20. An increase of the volume of the substituted alkyloxy group
as in compounds 75 and 76 led to a decrease in activity
compared to 73. Introduction of ester groups caused a 3−6-fold
a. Garcinia Natural-Product-Like D-Ring-Modified Caged
Xanthones 46−59 with a Pyran Ring in the Planar Region.
To gain insight on how modification of the D caged ring can
affect cytotoxicity when the planar region features type A and B
structures, HepG2, A549, and U251 cells were treated with the
designed compounds 46−59. As shown in Table 3, all the
compounds were found to be less sensitive to U251 cells than
to HepG2 and A549 cells. Compared with compound 18, all of
the compounds with pyran ring at C2 (compounds 46−52)
were less cytotoxic against the three cancer cell lines with the
exception of 52 (IC50 = 0.91 μM), which was 3.2 times more
active than 18 against HepG2 cells. Compared with compound
19, all of the compounds with a pyran ring at C4 (compounds
53−59) were less cytotoxic against HepG2 and U251 cells. The
introduction of morpholino, (n-Pr)2N, (i-Pr)2N, or (n-Bu)2N
groups (compounds 55, 57−59) led to a 1.2−3.0-fold increase
in activity as compared to 19, with IC50 values ranging from
2.16 to 7.25 μM against A549 cells, but they were all less active
than 1. The presence of a pyrrolidinyl or Et2N group in the
caged region resulted in complete loss of cytotoxicity against
the three cancer cell lines. Generally, in the case of compounds
282
dx.doi.org/10.1021/jm301593r | J. Med. Chem. 2013, 56, 276−292

Journal of Medicinal Chemistry
Article
Table 3. SAR Focused on the Caged Region Containing the Pyran Ring
a
IC50 (μM)
compd
planar region
R
HepG2
A549
U251
46
47
48
49
50
51
52
53
54
55
56
57
58
59
type A
type A
type A
type A
type A
type A
type A
type B
type B
type B
type B
type B
type B
type B
pyrrolidin-1-yl
piperidin-1-yl
morpholino
Et2N
71.3 5.7
71.3 4.3
8.82 1.08
50.4 3.6
5.05 0.97
7.93 0.27
0.93 0.08
>100
70.3 6.5
70.3 6.2
1.79 0.18
>100
70.3 5.4
59.1 3.9
12.2 1.7
>100
(n-Pr)2N
36.7 3.2
17.8 1.1
29.6 2.6
>100
16.8 1.5
>100
(i-Pr)2N
(n-Bu)2N
pyrrolidin-1-yl
piperidin-1-yl
morpholino
Et2N
>100
>100
91.5 4.0
9.35 0.82
>100
24.9 2.1
2.82 1.7
>100
>100
13.2 1.6
>100
(n-Pr)2N
5.93 0.33
14.5 1.2
9.44 0.54
2.16 0.06
7.25 0.45
6.01 1.07
40.4 4.2
23.8 2.3
10.7 1.4
(i-Pr)2N
(n-Bu)2N
a
Compounds 1, 18, and 19 are internal controls. For 1: HepG2, 0.89 0.08 μM; A549, 1.1 0.12 μM; U251, 2.56 0.25 μM. For 18: HepG2,
2.99 0.24 μM; A549, 2.00 0.27 μM; U251, 12.2 1.08 μM. For 19: HepG2, 4.40 0.14 μM; A549, 8.43 0.52 μM; U251, 8.29 0.76 μM.
Table 4. SAR Focused on the Caged Region Containing the Prenyl Side Chain at C2
a
IC50 (μM)
compd
R
HepG2
A549
U251
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
pyrrolidin-1-yl
piperidin-1-yl
morpholino
piperazin-1-yl
4-methylpiperazin-1-yl
4-oxopiperidin-1-yl
3,4-dioxopyrrolidin-1-yl
Me2N
2.42 0.09
3.97 0.15
6.27 1.02
3.16 0.97
2.28 0.34
1.74 0.05
52.0 7.3
4.28 0.22
3.86 0.30
3.13 0.25
11.0 1.3
4.77 0.31
2.94 0.18
3.93 0.40
15.7 1.9
4.84 0.72
20.8 1.4
20.5 2.2
10.8 0.8
12.6 1.1
20.1 1.7
3.10 0.06
3.10 0.24
6.95 0.98
35.9 2.8
8.58 1.03
10.4 1.3
17.3 2.0
16.6 1.4
17.3 0.9
18.0 2.1
11.7 1.5
3.25 0.22
19.4 2.1
9.85 0.68
12.9 0.7
58.2 5.6
5.18 0.55
9.12 0.51
50.7 4.5
7.34 0.63
6.02 0.71
27.1 2.3
3.23 0.17
58.6 4.4
25.1 3.0
36.4 4.2
31.0 4.7
64.6 2.9
58.0 4.3
83.1 7.8
45.4 5.1
26.0 2.4
79.2 6.1
37.1 2.6
>100
Et2N
(n-Pr)2N
(i-Pr)2N
(n-Bu)2N
(HOCH2CH2)2N
i-PrO
62.7 5.2
11.7 1.5
43.2 2.9
17.2 1.4
10.4 1.1
21.6 1.6
6.56 0.47
13.8 1.5
27.4 3.1
63.4 5.9
HOCH2CH2O
PhCH2O
PhCH2CH2O
AcO
PhCOO
PhCH2COO
PhCHCHCOO
4-phenylpiperazin-1-yl
a
Compounds 1 and 20 are internal controls. For 1: HepG2, 0.89 0.08 μM; A549, 1.10 0.12 μM; U251, 2.56 0.25 μM. For 20: HepG2, 3.46
0.23 μM; A549, 3.78 0.25 μM; U251, 18.7 2.1 μM.
283
dx.doi.org/10.1021/jm301593r | J. Med. Chem. 2013, 56, 276−292

Journal of Medicinal Chemistry
Article
Table 5. SAR Focused on the Caged Region Containing a Prenyl Side Chain at C4
a
IC50 (μM)
compd
R
HepG2
A549
U251
82
pyrrolidin-1-yl
piperidin-1-yl
morpholino
piperazin-1-yl
4-methylpiperazin-1-yl
4-oxopiperidin-1-yl
3,4-dioxopyrrolidin-1-yl
Me2N
4.71 0.17
8.37 0.39
4.42 0.22
2.56 0.15
4.65 0.40
10.6 0.7
2.33 0.12
3.92 0.33
2.00 0.02
4.60 0.50
2.90 0.14
3.96 0.44
3.08 0.29
2.20 0.09
16.7 2.5
2.22 0.13
13.4 2.0
2.95 0.13
2.13 0.20
3.28 0.25
21.2 2.4
1.96 0.12
>100
8.41 0.44
5.14 0.32
8.06 0.68
17.9 1.5
10.5 0.7
6.30 0.55
5.48 0.34
6.30 0.47
7.61 0.08
4.36 0.11
3.45 0.15
5.66 0.63
3.83 0.31
7.92 0.66
28.5 1.9
2.50 0.09
4.78 0.67
10.7 1.2
1.13 0.10
2.13 0.09
30.2 2.7
3.07 0.21
37.9 3.6
34.9 3.1
>100
17.1 0.9
27.8 1.7
33.1 2.2
35.1 2.1
71.6 4.6
>100
83
84
85
86
87
88
>100
89
21.2 3.1
25.1 2.8
24.3 2.2
>100
90
Et2N
91
(n-Pr)2N
92
(i-Pr)2N
93
(n-Bu)2N
17.3 2.1
33.9 2.7
7.48 0.83
43.2 3.2
22.1 1.6
17.5 2.3
13.2 1.8
16.7 2.4.
9.46 1.07
41.4 3.5
37.2 3.5
>100
94
(HOCH2CH2)2N
i-PrO
95
96
HOCH2CH2O
PhCH2O
97
98
PhCH2CH2O
AcO
99
100
101
102
103
104
105
106
107
PhCOO
PhCH2COO
PhCHCHCOO
4-phenylpiperazin-1-yl
H
Me
>100
>100
CH3OOC
>100
>100
Ph
>100
>100
>100
a
Compounds 1 and 21 are internal controls. For 1: HepG2, 0.89 0.08 μM; A549, 1.10 0.12 μM; U251, 2.56 0.25 μM. For 21: HepG2, 3.13
0.33 μM; A549, 2.04 0.09 μM; U251, 14.3 1.9 μM.
decrease in activity. In A549 cells, most of the compounds were
found to be less active when compared to compound 20, and
only three compounds (60, 70, 81) were more cytotoxic. In
U251 cells, five compounds (73, 75−76, 78−79) with
hydrophobic group substitution in the caged region showed
greater activity than compound 20. These results suggest that in
the case of compounds with a C2 prenyl group in the planar
region, the introduction of hydrophilic alkylamino groups to
the caged D-ring generally increased cytotoxicity against
HepG2 cells while introduction of hydrophobic and bulky
groups led to a decrease in activity. The presence of
substituents in the caged region was generally unfavorable to
antitumor activity against A549 and U251 cells.
c. Garcinia Natural-Product-Like D-Ring-Modified Caged
Xanthones 82−107 with a C4 Prenyl Group in the Planar
Region. To gain insight on how modification of the D caged
ring can affect cytotoxicity when the planar region features a
type D structure, HepG2, A549, and U251 cells were incubated
with the designed compounds 82−107. As shown in Table 5,
compared with compound 21, 10 compounds (85, 88, 90, 92,
94−95, 97, 99−100, 103) had increased potency against
HepG2 cells with IC50 values ranging from 1.96 to 3.08 μM,
and most of these compounds showed decreased cytotoxicity
against A549 and U251 cells. The substitution of the gem-
dimethyl group linked to the caged scaffold by an electron-
withdrawing carbonyl group interfered significantly with the
cytotoxicity, in all three cancer cell lines, as observed for
compounds 104−107. Among compounds 82−103, the
substitution of the alkylamino group led to increased or
maintained cytotoxicity against HepG2 cells. When introducing
some alkyloxy or ester groups, compounds 95, 97, 99, 100 were
obtained and had improved activity against HepG2 cells. The
cytotoxicity was improved when the substituent groups
contained an aromatic ring, as observed for compounds 97
and 100−101 with IC50 values ranging from 1.13 to 3.28 μM
against HepG2 and A549 cells. However, replacement of
substituents by the long-chain alkyloxy or ester groups as in
compounds 98 and 102 caused a significant decrease in activity
against HepG2 and A549 cells. Compound 112 with a
phenylpiperazinyl group was synthesized to determine if the
length of the substituents affected the cytotoxicity and whether
the alkylamino group may improve the physicochemical
property of the molecules. Compound 112 showed the most
potent cytotoxicity against HepG2 cells with an IC50 value of
284
dx.doi.org/10.1021/jm301593r | J. Med. Chem. 2013, 56, 276−292

Journal of Medicinal Chemistry
Article
all the compounds were nearly inactive against U251 cells even
at a concentration of 100 μM. This series of compounds was
generally less active than the corresponding analogues that have
prenyl substitution in the planar region. Although compared
with compound 22, most of the compounds showed less
potency against HepG2 and A549 cells, these simplified
analogues generally were inhibitory at micromolar levels and
had improved druglike properties (see Table 7). Notably, 15 of
them had IC50 values against HepG cells below 10 μM, while
the number was decreased to 10 in the case of A549 cells.
Introduction of hydrophilic alkylamino groups, such as
methylpiperazinyl, 4-oxopiperidinyl, and (n-Pr)2N groups, as
in compounds 112, 113, and 117, respectively, led to a similar
cytotoxic profile when compared to compound 22. Compound
112 showed comparable cytotoxicity against HepG2 and A549
cells with an IC50 value of 3.25 and 3.60 μM, respectively.
Meanwhile, compounds 121−123, with alkyloxy substituents,
also exhibited cytotoxicity, whereas further introducing long-
chain alkyloxy group as in compound 124 resulted in major loss
of activity against HepG2 cells. A similar decrease in
cytotoxicity was also observed in compound 128 as compared
to 126 and 127. The bulky substituent seemed to increase the
activity against A549 cells, since compound 119 with a (n-
Bu)2N group showed the most potent activity against A549
cells among all these series of compounds. Introduction of
long-chain aromatic moieties as in compounds 124, 127−129
did not interfere significantly with the cytotoxicity to A549
cells. Removal of the C1 hydroxyl substituent in compound 112
produced the significantly less active compound 130. This was
consistent with the finding that the presence of a hydroxyl
group at C1 improved the cytotoxicity. These results suggest
that in the case of compounds with a simplified planar region,
the introduction of alkylamino, alkyloxyl, or ester groups to
caged D-ring generally led to sustained antitumor activity and
facilitated improvement of the physicochemical properties such
as water solubility and cell member permeability while having
little effect on cytotoxicity, since the structures of these natural-
product-like caged xanthones were greatly simplified compared
to the natural product GA.
Structure−Property Relationship Studies. In parallel
with the antitumor activity testing, the physicochemical
properties of 40 representative diverse caged compounds
were determined in an effort to identify potential druglike
compounds prior to the time-consuming and costly develop-
ment and optimization of analogues that may ultimately fail in
in vivo efficacy experiments. Compound 1 was used as a
reference compound for comparison of polar surface area, log
D7.4, intrinsic aqueous solubility, and permeability (Table 7).
Digoxin and propranolol were used as internal standards for
permeability. The polar surface area (PSA) was calculated by
Discovery Studio 3.0 using the ADMET protocol. The partition
coefficient (log D) at pH 7.4 and aqueous solubility were
determined according to the method of Avdeef and Tsinman52
on a Gemini Profiler instrument (pION) by the “gold-
standard” Avdeef−Bucher potentiometric titration method.53
Permeability coefficients were determined using a standard
parallel artificial membrane permeability assay (PAMPA) on a
PAMPA Explorer instrument (pION).
Table 6. SAR Focused on the Caged Region Containing a
Simplified B-Ring
a
IC50 (μM)
compd
R
HepG2
A549
U251
>100
108
109
110
111
112
pyrrolidin-1-yl
piperidin-1-yl
morpholino
8.82 0.77
12.1 1.1
7.72 0.45
7.50 0.43
3.25 0.22
12.5 1.0
10.2 0.8
8.51 0.34
4.24 0.27
3.60 0.19
>100
>100
piperazin-1-yl
>100
4-methylpiperazin-
1-yl
71.0 5.7
113
114
4-oxopiperidin-1-yl
2.59 0.31
10.2 0.6
9.14 0.48
38.0 2.58
>100
>100
3,4-dioxopyrrolidin-
1-yl
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
Me2N
8.59 0.71
7.89 0.73
3.69 0.26
33.1 2.08
7.83 0.63
6.53 0.48
5.70 0.35
5.21 0.32
7.21 0.48
>100
38.2 2.41
20.5 2.17
7.82 0.33
23.5 2.26
1.25 0.06
9.68 0.70
14.8 1.4
37.3 3.5
5.83 1.01
4.46 0.29
36.3 4.0
10.9 1.1
7.58 0.33
14.9 1.3
2.15 0.17
>100
Et2N
>100
(n-Pr)2N
(i-Pr)2N
>100
>100
(n-Bu)2N
(HOCH2CH2)2N
MeO
94.9 3.3
>100
>100
i-PrO
>100
PhCH2O
PhCH2CH2O
AcO
93.1 5.1
>100
23.9 1.7
11.8 0.8
6.65 0.12
24.8 2.2
>100
PhCOO
>100
PhCH2COO
PhCHCHCOO
>100
>100
4-phenylpiperazin-1- 7.34 0.34
yl
>100
130
4-methylpiperazin-
1-yl
41.6 4.3
54.0 3.6
>100
a
Compounds 1 and 22 are internal controls. For 1: HepG2, 0.89
0.08 μM; A549, 1.10 0.12 μM; U251, 2.56 0.25 μM. For 22:
HepG2, 2.65 0.21 μM; A549, 3.79 0.50 μM; U251, 23.9 2.0
μM.
1.96 μM and also inhibited A549 cells with an IC50 value of
3.07 μM. These results suggest that in the case of compounds
with a C4 prenyl group in the planar region, the introduction of
hydrophilic alkylamino groups to the caged D-ring can also
increase cytotoxicity against HepG2 cells. The presence of
aromatic substitutents in the caged region seemed to be
favorable for antitumor activity against HepG2 and A549 cells,
while the length and volume of the substituted moiety in the
caged region affected the cytotoxicity significantly. Further-
more, replacement of the gem-dimethyl group linked to the
caged scaffold by a carbonyl group caused loss in activity
against all the three cancer cells.
d. Garcinia Natural-Product-Like D-Ring-Modified Caged
Xanthones 108−130 Containing a Simplified Planar Region.
To gain insight on how modification of the D caged ring can
affect the cytotoxicity when the planar region features a type E
structure, HepG2, A549, and U251 cells were incubated with
the designed compounds 108−129. As shown in Table 6, a
better selectivity for HepG2 and A549 cells was achieved, since
As shown in Table 7, all the synthetic caged compounds
displayed aqueous solubility that was improved over that of 1.
Among all the tested compounds, compound 120 was the most
soluble with an intrinsic solubility of 10.37 mM, almost 1000
times more soluble than 1. Generally, the introduction of
285
dx.doi.org/10.1021/jm301593r | J. Med. Chem. 2013, 56, 276−292

Journal of Medicinal Chemistry
Article
Table 7. SPR Focused on both the Planar and Caged Regions
structure
a
compd planar region caged region
R
PSA_2D (Å2) log D, pH 7.4 intrinsic solubility, (mM) Pe, pH 7.4 (10−6 cm/s)
18
type A
type A
type A
type B
type B
type B
type C
type C
type C
type C
type C
type C
type C
type D
type D
type D
type D
type D
type D
type D
type D
type D
type D
type D
type D
type E
type F
type G
type F
type F
type F
type F
type F
type F
type F
type F
type F
type F
type F
type F

type A
type E
type E
type A
type E
type E
type A
type E
type E
type E
type E
type E
type E
type A
type B
type E
type E
type E
type E
type E
type E
type E
type E
type E
type E
type B
type B
type B
type E
type E
type E
type E
type E
type E
type E
type E
type E
type E
type E
type E


82.2
94.5
85.6
82.2
94.5
85.6
94.1
97.4
106.4
97.4
97.4
139.1
100.8
94.1
94.1
97.4
106.4
97.4
94.4
139.1
100.8
103.0
103.0
120.3
120.3
94.1
73.3
52.5
76.6
85.6
89.4
80.0
76.6
76.6
118.3
80.0
82.2
82.2
99.5
99.5
120.3
4.0
3.8
6.2
4.3
4.0
6.0
3.6
3.9
3.1
2.4
4.7
2.6
4.5
3.5
3.9
3.7
3.2
2.6
5.1
2.0
4.7
3.6
4.9
2.8
5.1
2.3
2.9
0.32
0.84
0.028
0.17
0.43
0.021
0.57
0.98
1.27
2.19
0.71
4.52
0.38
0.50
0.34
0.85
1.25
2.45
0.092
4.32
0.34
0.32
0.16
0.83
0.21
1.41
0.84
16.8
26.4
8.1
48
morpholino
(n-Bu)2N

52
19
13.6
36.3
8.2
55
morpholino
(n-Bu)2N

59
20
21.9
26.2
28.4
12.2
22.7
9.1
60
pyrrolidin-1-yl
morpholino
Me2N
62
67
71
(n-Bu)2N
(HOCH2CH2)2N
4-phenylpiperazin-1-yl

72
81
13.2
12.1
10.3
15.6
27.3
41.9
11.8
19.0
20.6
23.5
11.0
34.2
9.8
21
32

82
pyrrolidin-1-yl
morpholino
Me2N
83
89
93
(n-Bu)2N
(HOCH2CH2)2N
4-phenylpiperazin-1-yl
i-PrO
94
103
95
97
PhCH2O
99
AcO
101
28
PhCH2COO

27.5
77.6
29.3
108.1
153.2
37.2
155.7
189.1
50.1
197.2
77.3
123.2
82.5
143.3
67.3
26.5
33

b
b
34



108
110
111
112
115
119
120
129
122
123
125
127
1
pyrrolidin-1-yl
morpholino
piperazin-1-yl
4-methylpiperazin-1-yl
Me2N
2.9
2.0
0.44
1.8
1.7
4.6
1.2
3.6
2.9
3.5
1.9
3.6
3.6
2.00
3.88
5.79
4.80
7.01
0.19
10.37
0.62
1.06
0.57
5.44
0.64
0.015
(n-Bu)2N
(HOCH2CH2)2N
4-phenylpiperazin-1-yl
i-Pr
PhCH2O
AcO
PhCH2COO

a
b
Digoxin (30.0 × 10−6 cm/s) and propranolol (541.5 × 10−6 cm/s) are internal standards in permeability determinations. Undetermined
hydrophilic groups such as morpholino, pyrrolidin-1-yl,
piperazin-1-yl, and 4-methylpiperazin-1-yl resulted in a
significant increase in solubility. For instance, comparing the
solubility of 52 and 48, 55 and 19, 67 and 20, 120 and 33,
showed that all the former compounds were more soluble than
the latter compounds. Introduction of hydrophobic aliphatic
chains such as dibutylamino and benzyl groups with bulk
volume led to poor solubility. For examples, compounds 52, 59,
71, and 119 with dibutylamino substitution, compounds 97 and
123 with a benzyloxy group, and compounds 101 and 127 with
a phenylacetoxy group all displayed reduced aqueous solubility
compared to the corresponding analogues lacking these
substituents. Furthermore, it was shown that compounds with
a simplified planar region (type F) were far more soluble than
compounds with a prenyl substituent (types C and D) or fused
pyran ring (types A and B) in the planar region. Compounds of
type A and B planar regions showed the poorest solubility,
probably a result of the increased rigidity of the molecule due to
the additional pyran ring. These results suggest that
compounds with a smaller framework and with hydrophilic
groups are likely to have better aqueous solubility.
Permeability is another important property reflecting the
ability of molecules to diffuse through the cell membrane.
Generally, as shown in Table 7, the Pe values of all compounds
were less than that of the positive control compound,
propranolol. Compound 115, the most permeable of the
caged xanthone analogues, had a Pe value of 189.1 × 10−6 cm/s.
Compounds with simple B-ring, such as those with a type F
planar region, possessed better permeability when compared to
those with prenyl or pyran substituents in the planar region. In
addition, the introduction of hydrophilic groups to the caged
region of the molecules also led to an improvement in
286
dx.doi.org/10.1021/jm301593r | J. Med. Chem. 2013, 56, 276−292

Journal of Medicinal Chemistry
Article
Figure 3. Plot of log D (pH 7.4) versus PSA_2D.
Table 8. Inhibitory Effect of 103, 112, 33, and 1 with iv Administration on the Growth of Heps Transplanted Mice
body weight (g)
group
dose (iv) (mg/kg)
predose
postdose
weight of tumor (g)
inhibitory rate (%)
control
103
103
103
112
112
112
33
0
20.75 1.20
20.88 1.69
21.75 1.92
21.50 1.32
21.10 1.30
22.00 1.56
21.56 0.68
20.67 1.41
21.44 2.87
20.13 1.27
22.50 1.20
19.50 1.32
29.75 3.03
28.50 1.50
27.43 4.53
26.50 2.50
29.25 1.56
27.71 4.27
27.22 2.53
27.33 5.54
28.11 4.33
26.43 3.50
28.00 4.56
21.17 4.14
1.27 0.20
0.88 0.51
0.00
20
40
80
5
30.74
48.77
60.20
45.67
56.06
57.75
25.21
34.49
40.66
69.11
70.31
b
0.65 0.16
b
0.51 0.22
b
0.62 0.18
b
10
20
10
20
40
10
25
0.50 0.26
b
0.48 0.15
a
0.95 0.27
b
33
0.83 0.19
b
33
0.75 0.39
b
1
0.35 0.17
b
5-FU
0.38 0.22
a
b
P < 0.05. P < 0.01 vs control.
permeability. Compounds such as 52, 59, 97, 101, 119, and
123, with bulky hydrophobic substituents, were found to have
poor permeability when compared to the analogues lacking
these substituents. These results suggest that the trends of
permeabilities of these compounds are similar to the
solubilities; compounds with smaller frameworks and hydro-
philic groups are also likely to have better permeability.
Whether a drug can be well-absorbed intestinally after oral
administration depends on the intrinsic properties of the
molecule, and the polar surface area and distribution coefficient
are the two main aspects. Egan et al. developed an intestinal
absorption model using 182 compounds with descriptors that
include AlogP and 2D polar surface area (PSA_2D).54,55 In this
model, the compounds that are well-absorbed will generally
reside within the elliptical regions of 95% (in red) and 99% (in
green) confidence levels (Figure 3). Analysis of a plot of the
measured log D versus the calculated PSA_2D of the selected
cage xanthones (Figure 3) showed that most of the compounds
with type F structure in the planar region fell within the ellipse
regions of 95% while GA fell outside the ellipse regions of 95%.
These results are consistent with the permeability data and
suggest that the compounds with good permeability such as
compounds with type F planar region are likely to have
favorable intestinal absorption.
In vivo Antitumor Activity with iv Administration.
Both the in vitro antitumor activity and physicochemical
properties have been taken into account when determining
whether a compound is druglike and has potential to be active
in vivo. For instance, compound 52, which features a pyran ring
in the plannar region, possessed a low micromolar activity
against HepG2 cells (IC50 = 0.93 μM), but it had poor aqueous
solubility (0.028 mM) and cell member permeability (8.1 ×
10−6 cm/s), thus compound 52 fell out of favor for further in
vivo evaluation. Compound 103, which features a prenyl group
in the plannar region, showed potent in vitro activity against
HepG2 cells (IC50 = 1.96 μM) and had an improved aqueous
solubility (0.34 mM) as compared to 1, and as a consequence,
compound 103 was selected to be advanced into in vivo
efficacy experiments. Although compound 112 had moderate
potency in vitro against HepG2 cells (IC50 = 3.25 μM), it
287
dx.doi.org/10.1021/jm301593r | J. Med. Chem. 2013, 56, 276−292

Journal of Medicinal Chemistry
Article
features a much smaller molecular size and possessed greatly
improved solubility (4.8 mM) and permeability (155.7 × 10−6
cm/s) as compared to 1. Besides, statistic data showed that the
polar surface area (PSA) for 1590 orally administered non-CNS
drugs that have reached at least phase II efficacy studies mainly
distribute between 60 and 80 Å2 and that compounds with large
PSA (>120 Å2) are hardly absorbed by the passive transcellular
route.56 The PSA value of compound 112 (80.0 Å2) fitted well
with the requirement of orally administered drugs. On the basis
of the above consideration, compound 112 was expected to be
potentially active in vivo.
In the light of the in vitro antitumor activity and
physicochemical properties, we selected compounds 103 and
112 to determine their in vivo efficacy with iv administration on
the growth of hepatoma solidity (Heps) in mice using 1, 5-
fluorouracil (5-FU), and a previously studied38 caged xanthone
analogue 33 for comparison. As shown in Table 8, the tumor
growth inhibitory rates of 1 and 5-FU were 69.11% and 70.31%,
at 10 and 25 mg/kg twice daily doses, over a period of 8 days,
respectively. Compounds 103 and 112 displayed significant
inhibitory effect on the growth of inoculated Heps in mice in a
dose-dependent manner, with 48.77% and 56.06% inhibitory in
tumor growth, at 40 and 10 mg/kg twice daily doses,
respectively. Whereas the inhibitory rate was only 40.66%
when administrated with compound 33 at 40 mg/kg twice daily
dose. Although 103 (IC50 = 1.96 μM) and 112 (IC50 = 3.25
μM) were less active than 33 (IC50 = 0.81 μM) against HepG2
cells in vitro, they both were found to be more effective than 33
in vivo. Moreover, compound 112, with the lowest in vitro
activity of three caged analogues against HepG2 cells, turned
out to the most potent compound in vivo on inhibition of the
growth of inoculated Heps in mice. This may be attributed to
the improved physicochemical properties such as aqueous
solubility and cell membrane permeability of 112 compared to
103 and 33. No vascular irritation or weight loss was observed
in any of these in vivo tests.
volume. Compound 112 had a t1/2 value similar to that of 1 and
significantly increased distribution volume, indicating the
possibility of wide distribution. With oral dosing, shorter tmax
values were observed in compounds 33 and 103 (0.25 h)
compared to 1 (0.5 h), while compound 112 had the same tmax
value as 1. Oral bioavailability of the simplified xanthone 33
(2.59%) was much less than that of 1 (13.4%). Introduction of
hydrophilic amino moieties led to an increase in oral
bioavailability as observed in compounds 103 and 112
compared to 33. Compound 112, which has a simplified
planar region and hydrophilic methylpiperazinyl substituents in
the caged region, had the best oral bioavailability, about 12%
better than that of 1.
In vivo Antitumor Activity with po Administration.
Considering both the in vivo antitumor activity with iv
administration and the pharmacokinetic data described above,
we selected compound 112 to determine further its in vivo
efficacy with po administration on the growth of hepatoma
solidity (Heps) in mice using 1 as comparison control. As
shown in Table 10, compound 112 displayed much more
inhibitory potency on the growth of inoculated Heps in mice,
while the tumor weight in the control group was unchanged
and neither vascular irritation nor weight loss was observed in
any of the groups. For compound 112, 57.74% inhibition in
tumor growth was observed at 100 mg/kg daily oral dose, over
a period of four days, while the inhibitory rate of 1 was 36.36%
at the same dose. Thus, compound 112 was identified as the
first orally active Garcinia natural-product-like caged xanthone
and also a promising novel antitumor agent for further clinical
development.
CONCLUSIONS
■
Using regioselective propargylation of diverse hydroxyxanthone
substrates, we have developed a novel synthetic approach to
Garcinia xanthones with diverse caged regions. With this
synthetic strategy, we have synthesized 99 Garcinia structurally
diverse natural-product-like xanthones containing the pharma-
cophoric caged core. These synthetic compounds have been
used to further explore the properties of the planar B-ring and
the caged D-ring of Garcinia analogues. All the synthesized
compounds have been evaluated for their in vitro antitumor
activity against HepG2, A549, and U251 cells, and detailed SAR
around the pharmacophoric caged core have been determined.
The intrinsic aqueous solubility, polar surface area (PSA), log
D7.4, and permeability at pH 7.4, and the detailed structure−
property relationships (SPR) of a select set of 40 Garcinia
caged xanthones have also been determined. Various hydro-
philic groups have been introduced to the caged region in an
effort to improve the druglike properties of the compounds.
The important SAR and SPR information of these compounds
is comprehensively presented in Figure 4. In the light of the in
vivo antitumor activity and the measured druglike properties,
two compounds, 103 and 112, were selected for testing in in
vivo efficacy experiments with iv administration. The results
showed that 103 and 112 displayed a significant and dose-
dependent inhibitory effect on the growth of inoculated Heps
in mice, while neither vascular irritation nor weight loss was
observed in any of the dose groups. In vivo pharmacokinetic
parameters such as tmax, t1/2, AUC, and F were determined.
Compound 112 had the best bioavailability and was chosen
for determination of its antitumor potency with po
administration. The results showed that 112 has more potent
in vivo orally antitumor activity than GA. This novel Garcinia
Pharmacokinetic Data. The compounds with effective in
vivo antitumor activity were evaluated for their in vivo
pharmacokinetic parameters to determine which compounds
are potentially orally active. The natural product GA was tested
for comparison. The full in vivo PK parameters for 103, 112,
33, and 1 in CD-1 mice are presented in Table 9. Compared
with 1, the simplified xanthones 33 and 103 showed relative
shorter t1/2 values, with a higher clearance and distribution
Table 9. Pharmacokinetics of 103, 112, 33, and 1
a
PK parameters
103
112
33
1
iv
dose (mg/kg)
t1/2 (h)
5
5
5
5
1.60
1.44
22615
52.2
205
20
4.34
1.81
20521
128.3
242
1.41
0.53
3338
6.8
5.06
2.26
2461
19.7
2005
20
MRT (h)
Cl (mL/h/kg)
Vd (L/kg)
AUC (ng h/mL)
dose (mg/kg)
tmax (h)
1477
20
po
20
0.25
14.9
2.62
2.28
40
0.50
79.6
1.12
1.63
151
0.25
1.46
1.36
0.947
153
0.50
191
Cmax (ng/mL)
t1/2 (h)
3.29
4.51
1072
13.4
MRT (h)
AUC (ng h/mL)
F (%)
4.88
15.6
2.59
a
Determinations with three mice per time point, average value.
288
dx.doi.org/10.1021/jm301593r | J. Med. Chem. 2013, 56, 276−292

Journal of Medicinal Chemistry
Article
Table 10. Inhibitory Effect of 112 and 1 with po Administration on the Growth of Heps Transplanted Mice
body weight (g)
group
dose (po) (mg/kg)
predose
postdose
weight of tumor (g)
1.27 0.20
inhibitory rate (%)
control
112
1
0
20.75 1.20
22.00 1.18
21.60 0.49
29.75 3.03
27.22 2.53
25.13 2.98
0.00
b
100
100
0.48 0.15
57.74
36.36
a
0.73 0.24
a
b
P < 0.05. P < 0.01 vs control.
Figure 4. Graphical depiction of the general SAR for cell cytotoxicity based on the IC50 results on HepG2 cells and SPR for water solubility and cell
membrane permeability.
out in triplicate in a parallel manner for each concentration of target
compounds used, and the results are presented as mean SE. Control
cells were given only culture media. After incubation for 24 h, the
absorbance (A) was measured at 570 nm. Survival ratio (%) was
natural-product-like caged xanthone is therefore an orally active
antitumor agent that is promising for further clinical develop-
ment.
calculated using the following equation: survival ratio (%) = (Atreatment
/
EXPERIMENTAL SECTION
All reagents were purchased from commercial sources and, unless
otherwise noted, were used without further purification. Organic
solutions were concentrated in a rotary evaporator (Buchi Rotavapor)
■
Acontrol) × 100%. IC50 was taken as the concentration that caused 50%
inhibition of cell viabilities and was calculated by the SigmaPlot
software (Systat Software Inc.).
̈
PAMPA Assay. Permeability (Pe) was determined by a standard
parallel artificial membrane permeability assay (PAMPA by pION).
PAMPA was conducted on a PAMPA Explorer instrument (pION
Inc., Woburn, MA) with PAMPA Explorer command software
(Version 3.6.0.6) as follows: test compound stock was diluted with
system solution buffer, pH 7.4 (pION Inc., Woburn, MA) to make
diluted test compound at 2 mM concentration. Then 150 μL of
diluted test compound was transferred to a UV plate (pION
Inc.,Woburn, MA), and the UV spectrum was read as the reference
plate. The membrane on a preloaded PAMPA sandwich (pION Inc.,
Woburn, MA) was painted with 4 μL of GIT lipid (pION Inc.,
Woburn, MA). The acceptor chamber was then filled with 200 μL of
acceptor solution buffer (pION Inc., Woburn, MA), and the donor
chamber was filled with 150 μL of diluted test compound. The
PAMPA sandwich was assembled, placed on the Gut-Box controlled
environment chamber, and stirred for 30 min. The aqueous boundary
layer was set to 40 μm for stirring. The UV spectrum (200−500 nm)
of the donor and the acceptor were read. The permeability coefficient
was calculated using PAMPA Explorer command software (Version
3.6.0.6) based on the AUC of the reference plate, the donor plate, and
the acceptor plate. All compounds were tested in triplicate, and the
data are presented as an average value. Standards for this assay include
digoxin (Pe = 30.0 × 10−6 cm/s at pH = 7.4) and propranolol (Pe =
541.5 × 10−6 at pH = 7.4).
In Vivo Pharmacokinetic Assay. Fifty-four healthy male CD-1
mice (weight ranging from 18 to 30 g, 27 mice for po, and 27 mice for
iv) were administered with 103, 112, 33, or 1. For the po study, a
single dose of 20 mg/kg body weight of 103, 112, 33, and 1 was
administered orally to mice. The dosing volume was 10 mL/kg.
Animals were fasted approximately 16 h prior to dosing until 4 h
postdosing but water was available to all animals at all times. For iv
study, a single dose of 5 mg/kg body weight of 103, 112, 33, and 1 was
administered iv to mice. The dosing volume was 10 mL/kg. Food and
water were available to animals at all times during the experiment.
below 45 °C under reduced pressure. Reactions were monitored by
thin-layer chromatography (TLC) on 0.25 mm silica gel plates
(GF254) and visualized under UV light. Silica gel (60 Å, 300−400
mesh) was used for flash column chromatography. Melting points
were determined with a Melt-Temp II apparatus and were reported
without correction. IR spectra were recorded on a Nicolet iS10 Avatar
1
FT-IR spectrometer using KBr film. The H NMR and 13C NMR
spectra were measured on a Bruker AV-300 or Bruker AV-500
instrument using deuterated solvents with tetramethylsilane (TMS) as
internal standard. Multiplicities were defined as s (singlet), d
(doublet), t (triplet), or m (multiplet). EI-mass spectra were recorded
with a Shimadzu GCMS-2010 mass spectrometer. High-resolution
mass spectra (HRMS) were recorded on a Water Q-Tof micro mass
spectrometer. The purity (≥95%) of the compounds was verified by
the HPLC study performed on an Amethyst C18−P (4.6 × 150 mm, 5
μm, Merck) column using a mixture of solvent methanol/water or
acetonitrile/water at the flow rate of 2 mL/min and peak detection at
240 nm under UV.
Cancer cell lines were obtained from Cell Bank of Shanghai,
Institute of Biochemistry and Cell Biology, Chinese Academy of
Sciences. Cells were cultured in 90% RPMI 1640 medium (GIBCO,
Invitrogen Corp.) supplemented with 10% fetal bovine serum
(Sijiqing, Zhejiang, China), 100 U/mL benzyl penicillin, and 100
μg/mL streptomycin in a humidified environment with 5% CO2 at 37
°C. Gambogic acid was isolated from Gamboge resin according to the
previously reported protocols.6 Samples containing >95% Gambogic
acid were used in the experiments. All the tested compounds were
dissolved in DMSO at a concentration of 0.01 mol/L and stored at −4
°C.
Chemistry. For experimental procedures and spectral character-
ization data see the Supporting Information.
MTT Assay. Cell viabilities were measured by a colorimetric assay
using 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide
(MTT, Roche) as described previously.38 Experiments were carried
289
dx.doi.org/10.1021/jm301593r | J. Med. Chem. 2013, 56, 276−292

Journal of Medicinal Chemistry
Article
Three mice were bled by cardiac puncture at each time point (0, 0.25,
0.5, 1, 1.5, 2, 4, 6, and 24 h for the po group and 0, 0.083, 0.25, 0.5, 1,
2, 4, 6, and 24 h for the iv group). Blood was collected into K3EDTA
tubes and centrifuged immediately at 4 °C at 3000 rpm for 10 min.
Plasma samples were transferred and stored at −20 °C prior to
analysis. The plasma samples were determined by LC−MS/MS assays.
In Vivo Tumor Growth Inhibition Assay. Kunming mice with
body weight of 18−22 g were transplanted with hepatoma solidity
(Heps) according to protocols of transplant tumor research; 24 h after
tumor transplantation, animals were weighed and randomly divided
into five groups. All test drugs were given through injections 24 h after
tumor transplantation. The treated groups were iv injected or po
administration of different doses of test compounds, respectively. The
negative control group received 0.9% normal saline. Treatments were
done at a frequency of one iv dose per 2 days or one po dose per day
for a total of four treatments. All animals were sacrificed 24 h after the
fourth treatment. All data are presented as the mean SD, and the
statistical significance was evaluated by the t-test.
REFERENCES
■
(1) Newman, D. J.; Cragg, G. M. Natural products as sources of new
drugs over the 30 years from 1981 to 2010. J. Nat. Prod. 2012, 75,
311−335.
(2) Butler, M. S. Natural products to drugs: Natural product-derived
compounds in clinical trials. Nat. Prod. Rep. 2008, 25, 475−516.
(3) Mishra, B. B.; Tiwari, V. K. Natural products: An evolving role in
future drug discovery. Eur. J. Med. Chem. 2011, 46, 4769−4807.
(4) Auterhoff, H.; Frauendorf, H.; Liesenklas, W.; Schwandt, C. The
chief constituents of gamboges resin. 1. Chemistry of gamboge. Arch.
Pharm. 1962, 295, 833−846.
(5) Han, Q. B.; Xu, H. X. Caged Garcinia xanthones: Development
since 1937. Curr. Med. Chem. 2009, 16, 3775−3796.
(6) Ollis, W. D.; Ramsy, M. V. J.; Sutherland, I. O.; Mongkolsuk, S.
The constitution of gambogic acid. Tetrahedron 1965, 21, 1453−1470.
(7) Ren, Y.; Yuan, C.; Chai, H. B.; Ding, Y.; Li, X. C.; Ferreira, D.;
Kinghorn, A. D. Absolute configuration of (−)-gambogic acid, and
antitumor agent. J. Nat. Prod. 2011, 74, 460−463.
(8) Asano, J.; Chiba, K.; Tada, M.; Yoshii, T. Cytotoxic xanthones
from Garcinia hanburyi. Phytochemistry 1996, 41, 815−820.
(9) Han, Q. B.; Wang, Y. L.; Yang, L.; Tso, T. F.; Qiao, C. F.; Song, J.
Z.; Xu, L. J.; Chen, S. L.; Yang, D. J.; Xu, H. X. Cytotoxic
polyprenylated xanthones from the resin of Garcinia hanburyi. Chem.
Pharm. Bull. 2006, 54, 265−267.
(10) Cao, S. G.; Sng, V. H. L.; Wu, X. H.; Sim, K. Y.; Tan, B. H. K.;
Pereira, J. T.; Goh, S. H. Novel cytotoxic polyprenylated xanthonoids
from Garcinia gaudichaudii (Guttiferae). Tetrahedron 1998, 54,
10915−10924.
(11) Thoisen, O.; Fahy, J.; Dumontet, V.; Chiaroni, A.; Riche, C.;
Tri, M. V.; Sevenet, T. Cytotoxic prenylxanthones from Garcinia
bracteata. J. Nat. Prod. 2000, 63, 441−446.
ASSOCIATED CONTENT
■
S
* Supporting Information
Experimental procedures for synthesis and spectral data for
characterization. This material is available free of charge via the
Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*Q.Y.: fax and phone, +86-25-83271351; e-mail, youqidong@
gmail.com. Q.G.: fax and phone, +86-25-83271440; e-mail:
anticancer_drug@yahoo.com.cn. H.S.: fax and phone, +86-25-
83271216; e-mail, sunhaopeng@163.com.
(12) Wu, Z. Q.; Guo, Q. L.; You, Q. D.; Zhao, L.; Gu, H. Y.
Gambogic acid inhibits proliferation of human lung carcinoma SPC-A1
cells in vivo and in vitro and represses telomerase activity and
telomerase reverse transcriptase mRNA expression in the cells. Biol.
Pharm. Bull. 2004, 27, 1769−1774.
Notes
The authors declare no competing financial interest.
(13) Zhang, H. Z.; Kasibhatla, S.; Wang, Y.; Herich, J.; Guastella, J.;
Tseng, B.; Drewe, J.; Cai, S. X. Discovery, characterization and SAR of
gambogic acid as a potent apoptosis inducer by a HTS assay. Bioorg.
Med. Chem. 2004, 12, 309−317.
(14) Yang, Y.; Yang, L.; You, Q. D.; Nie, F. F.; Gu, H. Y.; Zhao, L.;
Wang, X. T.; Guo, Q. L. Differential apoptotic induction of gambogic
acid, a novel anticancer natural product, on hepatoma cells and normal
hepatocytes. Cancer Lett. 2007, 256, 259−266.
(15) Zhao, L.; Zhen, C.; Wu, Z.; Hu, R.; Zhou, C.; Guo, Q. General
pharmacological properties, developmental toxicity, and analgesic
activity of gambogic acid, a novel natural anticancer agent. Drug Chem.
Toxicol. 2010, 33, 88−96.
(16) Gu, H.; Wang, X.; Rao, S.; Wang, J.; Zhao, J.; Ren, F. L.; Mu, R.;
Yang, Y.; Qi, Q.; Liu, W.; Lu, N.; Ling, H.; You, Q.; Guo, Q. Gambogic
acid mediates apoptosis as a p53 inducer through down-regulation of
mdm2 in wild-type p53-expressing cancer cells. Mol. Cancer Ther.
2008, 7, 3298−3305.
(17) Huang, H.; Chen, D.; Li, S.; Li, X.; Liu, N.; Lu, X.; Liu, S.; Zhao,
K.; Zhao, C.; Guo, H.; Yang, C.; Zhou, P.; Dong, X.; Zhang, C.
Guanmei; Dou, Q. P.; Liu, J. Gambogic acid enhances proteasome
inhibitor-induced anticancer activity. Cancer Lett. 2011, 301, 221−228.
(18) Yu, J.; Guo, Q. L.; You, Q. D.; Zhao, L.; Gu, H. Y.; Yang, Y.;
Zhang, H. W.; Tan, Z.; Wang, X. Gambogic acid-induced G2/M phase
cell-cycle arrest via disturbing CDK7-mediated phosphorylation of
CDC2/p34 in human gastric carcinoma BGC-823 cells. Carcinogenesis
2007, 28, 632−638.
(19) Qiang, L.; Yang, Y.; You, Q. D.; Ma, Y. J.; Yang, L.; Nie, F. F.;
Gu, H. Y.; Zhao, L.; Lu, N.; Qi, Q.; Liu, W.; Wang, X. T.; Guo, Q. L.
Inhibition of glioblastoma growth and angiogenesis by gambogic acid:
An in vitro and in vivo study. Biochem. Pharmacol. 2008, 75, 1083−
1092.
ACKNOWLEDGMENTS
■
We are thankful for the financial support of the National
Natural Science Foundation of China (No. 90713038, No.
21072231), National Major Science and Technology Project of
China (Innovation and Development of New Drugs, No.
2009ZX09501-003, No. 2010ZX09401-401), and “863” Pro-
gram of China (No. 2012AA020301).
ABBREVIATIONS USED
■
NP, natural product; GA, gambogic acid; Bcl-2, B-cell
lymphoma 2; IκBα, nuclear factor of kappa light polypeptide
gene enhancer in B-cells inhibitor alpha; SAR, structure−
activity relationship; SPR, structure−property relationship;
PSA_2D, 2D polar surface area; THF, tetrahydrofuran; DMF,
dimethylformamide; DCM, dichloromethane; DMSO, dimethyl
sulfoxide; EDC·HCl, 1-ethyl-3-(3-dimethylaminopropyl)-
carbodiimide hydrochloride; DMAP, 4-dimethylaminopryidine;
MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium
bromide; IC50, half-maximal inhibitory concentration; Me,
methyl; Et, ethyl; Pr, propyl; Bu, butyl; Ph, phenyl; Ac, acetyl;
Me2CO, acetone; ADMET, aborption distribution metabolism
excretion toxicity; PAMPA, parallel artificial membrane
permeation assay; Pe, permeability coefficient; CNS, central
nervous system; iv, intravenous injection; 5-FU, 5-fluorouracil;
po, per os (oral) ; t1/2, half-life; MRT, mean residence time; Cl,
clearance; Vd, volume of distribution; AUC, area under curve;
C
max, peak concentration; F, bioavailability; Heps, hepatoma
(20) Yi, T.; Yi, Z.; Cho, S. K.; Luo, J.; Pandey, M. K.; Aggarwal, B. B.;
solidity.
Liu, M. Gambogic acid inhibits angiogenesis and prostate tumor
290
dx.doi.org/10.1021/jm301593r | J. Med. Chem. 2013, 56, 276−292

Journal of Medicinal Chemistry
Article
growth by suppressing vascular endothelial growth factor receptor 2
signaling. Cancer Res. 2008, 68, 1843−1850.
(21) Li, C.; Lu, N.; Qi, Q.; Li, F.; Ling, Y.; Chen, Y.; Qin, Y.; Li, Z.;
Zhang, H.; You, Q.; Guo, Q. Gambogic acid inhibits tumor cell
adhesion by suppressing integrin β1 and membrane lipid rafts-
associated integrin signaling pathway. Biochem. Pharmacol. 2011, 82,
1873−1883.
pharmacophoric motif of the caged Garcinia xanthones. Org. Biomol.
Chem. 2009, 5, 4886−4894.
(38) Wang, X.; Lu, N.; Yang, Q.; Gong, D.; Lin, C.; Zhang, S.; Xi, M.;
Gao, Y.; Wei, L.; Guo, Q.; You, Q. Studies on chemical modification
and biology of a natural product, gambogic acid (III): Determination
of the essential pharmacophore for biological activity. Eur. J. Med.
Chem. 2011, 46, 1280−1290.
(22) Zhai, D.; Jin, C.; Shiau, C. W.; Kitada, S.; Satterthwait, A. C.;
Reed, J. C. Gambogic acid is an antagonist of antiapoptotic Bcl-2
family proteins. Mol. Cancer Ther. 2008, 7, 1639−1646.
(23) Yang, J.; Li, C.; Ding, L.; Guo, Q.; You, Q.; Jin, S. Gambogic
acid deactivates cytosolic and mitochondrial thioredoxins by covalent
binding to the functional domain. J. Nat. Prod. 2012, 75, 1108−1116.
(24) Qin, Y.; Meng, L.; Hu, C.; Duan, W.; Zuo, Z.; Lin, L.; Zhang, X.;
Ding, J. Gambogic acid inhibits the catalytic activity of human
topoisomerase IIalpha by binding to its ATPase domain. Mol. Cancer
Ther. 2007, 6, 2429−2440.
(39) Kuemmerle, J.; Jiang, S.; Tseng, B.; Kasibhatla, S.; Drewe, J.;
Cai, S. X. Synthesis of caged 2,3,3a,7a-terahydro-3,6-methanobenzo-
furan-7(6H)-ones: Evaluating the minimum structure for apoptosis
induction by gambogic acid. Bioorg. Med. Chem. 2008, 16, 4233−4241.
(40) Wang, J.; Ma, J.; You, Q.; Zhao, L.; Wang, F.; Li, C.; Guo, Q.
Studies on chemical modification and biology of a natural product,
gambogic acid (II): Synthesis and bioevaluation of gambogellic acid
and its derivatives. Eur. J. Med. Chem. 2010, 45, 4343−4353.
(41) Wang, X.; Lu, N.; Yang, Q.; Dai, Q.; Tao, L.; Guo, X.; Guo, Q.;
You, Q. Spectacular modification of gambogic acid on microwave
irradiation in methanol: Isolation and structure identification of two
products with potent anti-tumor activity. Bioorg. Med. Chem. 2010, 20,
2438−2442.
(25) Wang, T.; Wei, J.; Qian, X.; Ding, Y.; Yu, L.; Liu, B. Gambogic
acid, a potent inhibitor of survivin, reverses docetaxel resistance in
gastric cancer cells. Cancer Lett. 2008, 262, 214−222.
(26) Kasibhatla, S.; Jessen, K. A.; Maliartchouk, S.; Wang, J. Y.;
English, N. M.; Drewe, J.; Qiu, L.; Archer, S. P.; Ponce, A. E.; Sirisoma,
N.; Jiang, S.; Zhang, H. Z.; Gehlsen, K. R.; Cai, S. X.; Green, D. R.;
Tseng, B. A role for transferrin receptor in triggering apoptosis when
targeted with gambogic acid. Proc. Natl. Acad. Sci. U S A 2005, 102,
12095−12100.
(27) Pandey, M. K.; Sung, B.; Ahn, K. S.; Kunnumakkara, A. B.;
Chaturvedi, M. M.; Aggarwal, B. B. Gambogic acid, a novel ligand for
transferring receptor, potentiates TNF-induced apoptosis through
modulation of the nuclear factor-κB signaling pathway. Blood 2007,
110, 3517−3525.
(28) Davenport, J.; Manjarrez, J. R.; Peterson, L.; Krumm, B.; Blagg,
B. S.; Matts, R. L. Gambogic acid, a natural product inhibitor of
Hsp90. J. Nat. Prod. 2011, 74, 1085−1092.
(42) Li, X.; Zhang, X.; Sun, H.; Zhang, L.; Gao, Y.; Wang, J.; Guo, Q.;
You, Q. Synthesis and anti-tumor evaluation of novel C-37 modified
derivatives of gambogic acid. Chin. J. Chem. 2012, 30, 1083−1091.
(43) Sun, H.-P.; Liu, Z.-L.; Xue, X.; Gao, Y.; Zhang, L.; Guo, Q.-L.;
You, Q.-D. Studies on chemical structure modification and structure−
activity relationship of derivatives of gambogic acid at C(39). Chem.
Biodiv. 2012, 9, 1579−1590.
(44) Zhang, X.-J; Li, X.; Yang, Y.-R.; Sun, H.-P.; Gao, Y.; Zhang, L.;
Wang, J.-X.; Guo, Q.-L.; You, Q.-D. Studies on chemical structure
modification and structure−activity relationship of derivatives of
gambogic acid at C(34). Chem. Biodiv. 2012, 9, 2295−2308.
(45) Nicolaou, K. C.; Li, J. “Biomimetic” cascade reactions in organic
synthesis: Construction of 4-oxatricyclo[4.3.1.0]decan-2-one systems
and total synthesis of 1-O-methylforbesione via tandem Claisen
rearrangement/Diels−Alder reaction. Angew. Chem., Int. Ed. 2001, 40,
4264−4268.
(29) Wang, X.; Chen, Y.; Han, Q. B.; Chan, C. Y.; Wang, H.; Liu, Z.;
Cheng, C. H.; Yew, D. T.; Lin, M. C.; He, M. L.; Xu, H. X.; Sung, J. J.;
Kung, H. F. Proteomic identification of molecular targets of gambogic
acid: Role of stathmin in hepatocellular carcinoma. Proteomics 2009, 9,
242−253.
(46) Tisdale, E. J.; Li, H.; Vong, B. G.; Kim, S. H.; Theodorakis, E. A.
Regioselective synthesis of the tricyclic core of lateriflorone. Org. Lett.
2003, 5, 1491−1494.
(30) Palempalli, U. D.; Gandhi, U.; Kalantari, P.; Vunta, H.; Arner, R.
J.; Narayan, V.; Ravindran, A.; Prabhu, K. S. Gambogic acid covalently
modifies IkappaB kinase-beta subunit to mediate suppression of
lipopolysaccharide-induced activation of NF-kappaB in macrophages.
Biochem. J. 2009, 419, 401−409.
(31) Sun, H.; Chen, F.; Wang, X.; Liu, Z.; Yang, Q.; Zhang, X.; Zhu,
J.; Qiang, L.; Guo, Q.; You, Q. Studies on chemical modification and
biology of a natural product, Gambogic acid (IV): Identification of IκB
Kinase-beta (IKKβ) as the molecular target and determination of the
pharmacophoric scaffold. Eur. J. Med. Chem. 2012, 51, 110−123.
(32) Wang, J.; Zhaoaaa, Li.; Hu, Y.; Guo, Q.; Zhang, L.; Wang, X.; Li,
N.; You, Q. Studies on chemical modification and biology of a natural
product, gambogic acid (I): Synthesis and biological evaluation of
oxidized analogues of gambogic acid. Eur. J. Med. Chem. 2009, 44,
2611−2620.
(33) Chantarasriwong, O.; Batova, A.; Chavasiri, W.; Theodorakis, E.
A. Chemistry and biology of the caged Garcinia xanthones. Chem.
Eur. J. 2010, 16, 9944−9962.
(34) He, L.; Ling, Y.; Fu, L.; Yin, D.; Wang, X.; Zhang, Y. Synthesis
and biological evaluation of novel derivatives of gambogic acid as anti-
hepatocellualar carcinoma agents. Bioorg. Med. Chem. 2012, 22, 289−
292.
(35) Batova, A.; Lam, T.; Wascholowski, V.; Yu, A. L.; Giannis, A.;
Theodorakis, E. A. Synthesis and evaluation of caged Garcinia
xanthones. Org. Biomol. Chem. 2007, 5, 494−500.
(47) Nicolaou, K. C.; Sasmal, P. K.; Xu, H. Biomimetically inspired
total synthesis and structure activity relationships of 1-O-methyllateri-
florone. 6π Electrocyclizations in organic synthesis. J. Am. Chem. Soc.
2004, 126, 5493−5501.
(48) Tisdale, E. J.; Slobodov, I.; Theodorakis, E. A. Unified synthesis
of caged Garcinia natural products based on a site-selective Claisen/
Diels−Alder/Claisen rearrangement. Proc. Natl. Acad. Sci. U. S. A.
2004, 101, 12030−12035.
(49) Li, N.-G.; Wang, J.-X.; Liu, X.-R.; Lin, C.-J.; You, Q.-D.; Guo,
Q.-L. A novel and efficient route to the construction of the 4-oxa-
tricyclo[4.3.1.0]decan-2-one scaffold. Tetrahedron Lett. 2007, 48,
6586−6589.
(50) Liu, Z.-L.; Wang, X.-J.; Li, N.-G.; Sun, H.-P.; Wang, J.-X.; You,
Q.-D. Total synthesis of aldehyde-containing Garcinia natural products
isomorellin and gaudichaudione A. Tetrahedron 2011, 67, 4774−4779.
(51) Li, X.; Zhang, X.; Yu, Z.; Liu, X.; You, Q.; Guo, Q. Microwave-
assisted Claisen rearrangement/Diels−Alder cascade reaction for the
synthesis of caged Garcinia natural products and analogues. J. Chem.
Res. 2012, 35, 630−633.
(52) Avdeef, A.; Tsinman, O. Miniaturized rotating disk intrinsic
dissolution rate measurement: Effect of buffer capacity in comparisons
to traditional Wood’s apparatus. Pharm. Res. 2008, 25, 2613−2627.
(53) Avdeef, A.; Bucher, J. J. Accurate measurements of the
concentration of hydrogen ions with a glass electrode: Calibrations
using the Prideaux and other universal buffer solutions and a
computer-controlled automatic titrator. Anal. Chem. 1987, 50,
2137−2142.
(36) Zhang, X.; Li, X.; Sun, H.; Jiang, Z.; Tao, L.; Gao, Y.; Guo, Q.;
You, Q. Synthesis and evaluation of novel aza-caged Garcinia
xanthones. Org. Biomol. Chem. 2012, 10, 3288−3299.
(37) Chantarasriwong, O.; Cho, W. C.; Batova, A.; Chavasiri, W.;
Moore, C.; Rheingold, A. L.; Theodorakis, E. A. Evaluation of the
(54) Egan, W. J.; Merz, K. M.; Baldwin, J. J. Prediction of drug
absorption using multivariate statistics. J. Med. Chem. 2000, 43, 3867−
3877.
291
dx.doi.org/10.1021/jm301593r | J. Med. Chem. 2013, 56, 276−292

Journal of Medicinal Chemistry
Article
(55) Egan, W. J.; Lauri, G. Prediction of intestinal permeability. Adv.
Drug Delivery Rev. 2002, 54, 273−289.
(56) Kelder, J.; Grootenhuis, P. D.; Bayada, D. M.; Delbressine, L. P.;
Ploemen, J.-P. Polar molecular surface as a dominating determinant for
oral absorption and brain penetration of drugs. Pharm. Res. 1999, 16,
1514−1519.
(57) Hollingshead, M. G. Antitumor efficacy testing in rodents. J.
Nat. Cancer Inst. 2008, 100, 1500−1510.
292
dx.doi.org/10.1021/jm301593r | J. Med. Chem. 2013, 56, 276−292