Insights into the in vitro antitumor mechanism of action of a new pyranoxanthone

Article abstract of DOI:10.1111/j.1747-0285.2010.00978.x

Naturally occurring xanthones have been documented as having antitumor properties, with some of them presently undergoing clinical trials. In an attempt to improve the biological activities of dihydroxyxanthones, prenylation and other molecular modificati

Full text of DOI:10.1111/j.1747-0285.2010.00978.x

ª 2010 John Wiley & Sons A/S
doi: 10.1111/j.1747-0285.2010.00978.x
Chem Biol Drug Des 2010; 76: 43–58
Research Article
Insights into the In Vitro Antitumor Mechanism
of Action of a New Pyranoxanthone
Andreia Palmeira1,2,3, Ana Paiva1,2, Emı´lia
Key words: antiproliferation, apoptosis, Bcl-xL, leukemia, structure-
based screening, xanthone
Sousa1,2, Hugo Seca3, Gabriela M.
Almeida3, Raquel T. Lima3,4, Miguel X.
Received 21 October 2009, revised 11 March 2010 and accepted for
publication 16 March 2010
Fernandes5, Madalena Pinto1,2 and
M. Helena Vasconcelos2,3,4,
*
1Department of Chemistry, Laboratory of Organic and
Pharmaceutical Chemistry, Faculty of Pharmacy, University of Porto,
Rua Anibal Cunha 164, 4050-047 Porto, Portugal
A large number of naturally occurring and synthetic xanthones with
interesting biological and pharmacological activities have been
reported in the past few years, namely as antitumor agents (1). The
most remarkable examples undergoing clinical trials as antitumor
agents are psorospermin (1), a fused furanoxanthone, a promising
antileukemic agent isolated from the roots and bark of the tropical
African plant Psorospermum febrifugum (2), and dimethylxanthe-
none-4-acetic acid (DMXAA, 2) which causes a fast vascular col-
lapse and tumor necrosis and was discovered in a structure–activity
relationship study involving a series of flavone acetic acid analogues
(3,4) (Figure 1). Natural dehydroxanthones namely artobiloxanthone
(5) and caged xanthones like gaudichaudione A (6) are also potent
antileukemia agents. The significant antitumor properties reported
for the dehydrofuranoxanthone psorospermin (1) as well as for the
acridone alkaloid acronycine (3, Figure 1), both linear tricyclic ring
systems fused to a fourth oxygenated ring, had already encouraged
the investigation into isosteric acronycine xanthone derivatives that
were even more potent than acronycine (3) in L1210 leukemia cell
line (7–9). Nevertheless, acronycine (thio)xanthone derivatives with
methoxy group in the 1-position were found to be highly insoluble
and did not improve the overall activity (8,10).
2Research Center of Medicinal Chemistry (CEQUIMED-UP), University
of Porto, Rua Anibal Cunha 164, 4050-047 Porto, Portugal
3Cancer Biology Group, IPATIMUP – Institute of Molecular Pathology
and Immunology of the University of Porto, Rua Dr. Roberto Frias,
S ⁄ N, 4200-465 Porto, Portugal
4Department of Biological Sciences, Laboratory of Microbiology,
Faculty of Pharmacy, University of Porto, Rua Anibal Cunha 164,
4050-047 Porto, Portugal
5Centro De Quꢀmica da Madeira, Universidade da Madeira, Campus
da Penteada, 9000-390 Funchal, Portugal
*Corresponding author: M. Helena Vasconcelos,
hvasconcelos@ipatimup.pt
Naturally occurring xanthones have been docu-
mented as having antitumor properties, with some
of them presently undergoing clinical trials. In an
attempt to improve the biological activities of
dihydroxyxanthones, prenylation and other mole-
cular modifications were performed. All the com-
pounds reduced viable cell number in a leukemia
cell line K-562, with the fused xanthone 3,
4-dihydro-12-hydroxy-2,2-dimethyl-2H,6H-pyrano[3,
2-b]xanthen-6-one (5) being the most potent. The
Xanthones with prenyl units, either fused in a ring or as an open-
chain unit, have been reported to mediate some important biologi-
cal activities, concerning a large variety of targets with therapeutic
value (1,11). In our research group, we have previously investigated
the effect of several prenylated derivatives of 1,3-dihydroxy-2-
methylxanthone and 1,3-dihydroxyxanthone on the in vitro growth of
some human tumor cell lines, with particular attention on breast
adenocarcinoma (MCF-7) (12). It was found that, in some cases, the
presence of prenyl side chains enhanced activity when compared
with their non-prenylated analogs (12,13).
pyranoxanthone
5 was particularly effective in
additional leukemia cell lines (HL-60 and BV-173).
Furthermore, the pyranoxanthone 5 decreased cel-
lular proliferation and induced an S-phase cell
cycle arrest. In vitro, the pyranoxanthone
5
increased the percentage of apoptotic cells which
was confirmed by an appropriate response at the
protein level (e.g., PARP cleavage). Using a com-
puter screening strategy based on the structure of
several anti- and pro-apoptotic proteins, it was
verified that the pyranoxanthone 5 may block the
binding of anti-apoptotic Bcl-xL to pro-apoptotic
Bad and Bim. The structure-based screening
revealed the pyranoxanthone 5 as a new scaffold
that may guide the design of small molecules with
better affinity profile for Bcl-xL.
Considering previous results and the fact that 3,4-dihydroxyxanthone
(4), interesting as a PKC inhibitor (14) but with limited growth inhib-
itory effects on human tumor cell lines (15) is more water-soluble
than the previously investigated oxygenated scaffolds (8,10,12), the
synthesis of both open-chain and fused prenylated derivatives of
compound 4 was initially planned to look for antileukemic agents
(Figure 2). Herein, we firstly describe the synthesis and effect on
the viable cell number of a leukemia cell line, of a dihydropyrano
43

Palmeira et al.
Figure 1: Structures of
antitumor xanthones in clinical
trials (1–2) and of acronycine (3).
Figure 2: Synthesis of 3,4-di-
hydroxyxanthone (4) analogues 5–
8. (a) Et2O, AlCl3, room temp,
22 h; (b) MeOH, H2O, NaOH, reflu-
x, 47 h; (c) C6H5CH3, AlCl3, 70 ꢀC,
13 h; (d) BrCH2CHC(CH3)2, MW
200 W, Montmorillonite K10 (20
equiv.), CHCl3, 40 min; (e) BrCH2C-
HC(CH3)2, K2CO3, DMF, room temp,
48 h; (f) Br(CH2)6Br (drop-wise,
30 min), K2CO3, DMF, room temp,
24 h; (g) K3[Fe(CN)6], Me2CO:H2O
(1:1), room temp; (h) K2CO3, H2O,
room temp.
(5) and a diprenylated derivative (6), as well as of the bromoalky-
lated (7) and the fused lignoid (8) derivatives of 3,4-dihydroxyxanth-
one (4). The effect of the most potent compound, the fused
xanthone 3,4-dihydro-12-hydroxy-2,2-dimethyl-2H,6H-pyrano[3,2-b]xan-
then-6-one (5), on viable cell number was also confirmed in addi-
tional leukemia cell lines.
using an Ethos MicroSYNTH 1600 Microwave Labstation (Mile-
stone). Melting points were obtained in a Kçfler microscope
(Wagner and Munz, Munich, Germany) and are uncorrected. The
optical rotation was zero for 8 and was obtained on a Polartronic
Universal polarimeter. IR spectra were measured on an ATI Mattson
Genesis series FTIR (software: WinFirst v.2.10) spectrophotometer in
1
KBr microplates. H and 13C NMR spectra were taken in CDCl3 or
To gain further insights into the mechanism of action of the fused
xanthone derivatives, the antiproliferative and apoptotic effects of
compound 5 were evaluated, and docking studies were further
performed.
DMSO-d6 at room temperature, on Bruker Avance 300 instrument
(300.13 MHz for 1H and 75.47 MHz for 13C). Chemical shifts are
expressed in d (ppm) values relative to tetramethylsilane (TMS) as
an internal reference; 13C NMR assignments were made by 2D
heteronuclear single quantum correlation (HSQC) and Hetronuclear
Multiple Bond Correlation (HMBC) experiments (long-range C, H
coupling constants were optimized to 7 and 1 Hz). Mass spectro-
metry (MS) spectra were recorded as electronic impact (EI) mode
on a VG Autospec Q spectrometer, and high resolution mass spec-
trometry (HRMS) results were obtained in the services of C.A.C.T.I.
(Vigo, Spain). The following materials were synthesized and purified
by the described procedures.
Materials and Methods
Chemistry
K3[Fe(CN)6] was supplied by Merck (Darmstadt, Germany). Prenyl
bromide, dibromohexane, and coniferyl alcohol (9) were obtained
from Sigma (Steunheim, Germany). Purifications of compounds 4–8
were performed by column chromatography (CC) using Merck silica
gel 60 (0.040–0.063 mm) and preparative thin layer chromatography
(TLC) using Merck silica gel 60 (GF254) plates. Microwave reactions
were performed using set-up for atmospheric-pressure reactions
and also 120 mL closed reactors (internal reaction temperature
measurement with a fiber-optic probe sensor) and were carried out
3,4-Dihydroxyxanthone (4) The synthesis of compound 4
(Figure 2) was carried out according to the method of Gottlieb et al.
(16). To the crude product obtained from the demethylation of 3,4-
dimethoxyxanthone, HCl 5N (80 ml) was added, and the mixture
was extracted with ethyl acetate. The organic layer was collected
44
Chem Biol Drug Des 2010; 76: 43–58

Antitumor Mechanism of Action of a Pyranoxanthone
and dried (Na2SO4), and the solvent was evaporated to dryness.
Compound 4 was isolated by crystallization from acetone ⁄ petroleum
ether (40–60 ꢀC boiling fraction) (4:1 v ⁄ v) as slightly yellow crystals
(70%) and characterized according to the described procedure (14).
(C-4b), 151.1 (C-3), 139.5 (C-4a), 138.6 (C-3¢), 136.4 (C-4), 135.4 (C-
3¢¢), 134.4 (C-6), 126.6 (C-8), 123.8 (C-7), 122.1 (C-2¢¢), 121.5 (C-8a),
120.0 (C-2¢), 119.1 (C-2¢¢), 118.8 (C-8a), 118.6 (C-9a), 118.2 (C-
5),118.0 (C-1), 116.5 (C-2), 69.8 (C-1¢), 66.1 (C-1¢¢), 25.8 (3¢-CH3),
25.7 (3¢¢-CH3), 18.3 (3¢-CH3), 17.9 (3¢¢-CH3) ppm; IR (KBr): m = 3408,
2959, 2920, 2852, 1655, 1603, 1463, 1436, 1375, 1330, 1282, 1261,
1222, 1077, 756 cm)1; MS (EI, 70 eV) m ⁄ z (%): 364 (54), 349 (74),
239 (100), 225 (78), 211 (63), 197 (35), 121 (20), 65 (6); HRMS-ESI
m ⁄ z calcd for C23H25O4 [M-H]+ 365.1756; found: 365.1747.
3,4-Dihydro-12-hydroxy-2,2-dimethyl-2H,6H-
pyrano[3,2-b]xanthen-6-one (5)
A mixture of 4 (4.4 mmol), prenyl bromide (8.8 mmol), and anhy-
drous K2CO3 (9.0 mmol) in anhydrous dimethylformamide (DMF)
(20 mL) was refluxed for 8 h (150 ꢀC). After cooling, the solid was
filtered and the solvent removed under reduced pressure and affor-
ded the crude product that was purified by flash chromatography
(SiO2; hexane ⁄ AcOEt, 6:4 v ⁄ v) and by preparative TLC (SiO2; hex-
ane ⁄ AcOEt, 5:5 v ⁄ v). Yellow crystals of 5 were obtained from ace-
tone (3%). This procedure was optimized to get procedure f: A
slurry of the K10 clay (20 equiv of weight) in CHCl3 (ca. 30 mL) was
treated with the 3,4-dihydroxyxanthone (4, 3.55 mmol), followed by
the addition of prenyl bromide (14.20 mmol) in CHCl3, in a 120-mL
microwave reactor. The mixture under stirring was irradiated at
200 W for 2 · 20 min, and the final temperature was 108 ꢀC. The
reaction mixture was filtered under vacuum, washed with CH2Cl2,
Me2CO, and MeOH, and the filtrate was concentrated under vacuum.
The recovered clay was reactivated by washing with MeOH. The
crude product was purified by chromatography flash cartridge (Grace
Resolvꢁ, Grace Company, Deerfield, IL, USA) in hexane ⁄ AcOEt (7:3
v ⁄ v). Yellow crystals of compound 5 were obtained from acetone
3-[(6¢-Bromohexyl)oxy]-4-hydroxy-9H-xanthen-9-
one (7) (17)
To a solution of compound 4 (2.2 mmol) in dry DMF (10 mL), was
added K2CO3 (2.2 mmol), and the mixture was stirred at room tem-
perature for 15 min. To this suspension was added, in 30 min, a
solution of 1,6-dibromobutane (4.4 mmol) in dry DMF (5 mL). The
reaction mixture was stirred for 8 h at room temperature and then
poured into ice-water (50 mL). The brown solid thus obtained was
filtered and purified by CC [petroleum ether (40–60 ꢀC boiling frac-
tion) ⁄ diethyl ether, several proportions]. A white solid was obtained
from the collected fractions from petroleum ether ⁄ diethyl ether (4:6
v ⁄ v). Crystallization from ethyl acetate ⁄ petroleum ether (3:1 v ⁄ v)
gave compound 7 (345 mg, 40%): mp 173 ꢀC (ethyl acetate ⁄ petro-
leum ether); 1H-NMR (DMSO-d6): d = 8.34 (1H, dd, J = 8.0 and
1.6 Hz, H-8), 7.91 (1H, d, J = 8.9 Hz, H-1), 7.73 (1H, ddd, J = 7.0,
7.0, and 1.5 Hz, H-6), 7.60 (1H, d, J = 7.9 Hz, H-5), 7.38 (1H, ddd,
J = 7.0, 7.0, and 1.5 Hz, H-7), 6.99 (1H, d, J = 8.9 Hz, H-2), 5.75
(1H, s, HO-C(4)), 4.23 (2H, t, H-1¢), 3.44 (2H, t, H-6¢), 1.93 (4H, m,H-
2¢ and H-5¢), 1.59 (4H, m, H-3¢ and H-4¢) ppm; 13C NMR (DMSO-d6):
d = 175.1 (C-9), 166.0 (C-10a), 150.3 (C-3), 144.9 (C-4a), 134.6 (C-6),
133.5 (C-4), 126.7 (C-8), 123.9 (C-7), 121.6 (C-8a), 118.2 (C-5), 118.2
(C-1), 116.6 (C-9a), 108.3 (C-2), 69.5 (C-1¢), 33.7 (C-6¢), 32.9 (C-2¢),
29.7 (C-5¢), 28.0 (C-3¢), 25.1 ppm (C-4¢); IR (KBr): m = 1645, 1606,
1455, 1334, 1250, 1184, 1081, 757 cm)1; HRMS-ESI m ⁄ z calcd for
C19H20BrO4 [M-H]+ 391.05395, found: 391.05375.
1
(518 mg, 40%): mp 199–202 ꢀC (Me2CO); H-NMR (CDCl3): d = 8.33
(1H, dd, J = 8.0 and 1.6 Hz, H-7), 7.70 (1H, ddd, J = 7.7, 8.0, and
1.6 Hz, H-9), 7.70 (1H, s, H-5), 7.58 (1H, dd, J = 8.0 and 1.0 Hz, H-8),
7.35 (1H, ddd, J = 7.3, 7.3, and 1.0 Hz, H-10), 5.74 (1H, br s, OH);
2.93 (2H, t, J = 6.7, H-4), 1.92 (2H, t, J = 6.7, H-3), 1.45 (6H, s, H-1¢)
ppm; 13C-NMR (CDCl3): d = 176.7 (C-6), 156.1 (C-10a), 146.3 (C-12a),
143.2 (C-11a), 134.3 (C-9), 132.4 (C-12), 126.6 (C-7), 123.6 (C-8),
121.5 (C-6a), 118.2 (C-10), 117.9 (C-5), 117.0 (C-4a), 115.4 (C-5a),
76.5 (C-2), 32.6 (C-3), 27.0 (C-4), 21.7 (C-1¢) ppm; IR (KBr): m = 3315,
2867, 2924, 2851, 1653, 1611, 1455, 1370, 1253, 1222, 1190, 1154,
1117, 757 cm)1; MS (EI, 70 eV) m ⁄ z (%): 296 (98) [M]+, 241 (100),
209 (28), 156 (15), 128 (15), 73 (5); HRMS-ESI m ⁄ z calcd for
C18H16O4 [M-H]+ 296.1049, found: 296.1051.
( )-(2R*,3R*)- and ( )-(2S*,3S*)-2-Hydroxymethyl-
3-(4-hydroxy-3-methoxyphenyl)-1,4-dioxane[5,6-
c]xanthones (8)
To a stirred solution of 4 (11.5 mmol) and 9 (11.1 mmol) in a 1:1
mixture of Me2CO ⁄ H2O (250 mL), a AcONa solution (44 mmol, in
100 mL of H2O) and then a K3[Fe(CN)6] solution (120 mmol in H2O)
were added, at room temp. After 5 h, the mixture was slightly acid-
ified with 10% HCl and the precipitate filtrated. The product
obtained, corresponding to a mixture of trans and cis isomers of 8
(3 g), was suspended in H2O (250 mL) and basified with K2CO3. The
suspension was stirred at room temp, for 1 day, until complete con-
version of cis isomer into the trans isomer of 8. The product was
filtered, washed with H2O, dried, and crystallized from 1,4-dioxane
to get a white powder (2.5 g, 55%): mp 243–245 ꢀC (1,4-dioxane);
1H-NMR (DMSO-d6): d = 9.25 (1H, s, C(4¢)-OH), 8.20 (1H, dd,
J = 7.7 and 1.6 Hz, H-8), 7.87 (1H, ddd, J = 8.3, 7.4, and 1.6 Hz, H-
10), 7.70 (1H, d, J = 8.8 Hz, H-6), 7.69 (1H, d, J = 8.3 Hz, H-11),
7.49 (1H, ddd, J = 7.7, 7.4, and 0.7 Hz, H-9), 7.07 (1H, d,
J = 8.8 Hz, H-5), 7.07 (1H, d, J = 1.7 Hz, H-2¢), 6.92 (1H, dd,
J = 8.1 and 1.7 Hz, H-6¢), 6.83 (1H, d, J = 8.1 Hz, H-5¢), 5.15 (1H, d,
3,4-bis-(3-Methylbut-2-enyloxy)-9H-xanten-9-one
(6)
To a solution of 4 (2.2 mmol) in dry DMF (15 mL) and K2CO3
(4.4 mmol), prenyl bromide (4.4 mmol) was added drop-wise, with
stirring at room temperature. After 8 h, the reaction mixture was
poured into ice-water (50 mL), and the solid was filtered. From this
crude product, the compound 6 was isolated by preparative TLC
(SiO2; hexane ⁄ AcOEt, 6:4 v ⁄ v). Prenylated xanthone 6 was crystal-
lized as a brown powder from EtOH (20 mg, 3%): mp > 330 ꢀC
1
(EtOH); H-NMR (CDCl3): d = 8.32 (1H, dd, J = 8.0 and 1.8 Hz, H-8),
8.06 (1H, d, J = 9.0 Hz, H-1), 7.70 (1H, ddd, J = 7.0, 9.0, and
1.6 Hz, H-6), 7.56 (1H, dd, J = 8.3 and 1.0 Hz, H-7), 7.37 (1H, ddd,
J = 6.0, 7.5, and 1.0, H-5), 7.00 (1H, d, J = 9.0 Hz, H-2), 5.65 (1H,
t, J = 6.6 Hz, H-2¢), 5.53 (1H, t, J = 7.4 Hz, H-2¢¢), 4.73 (2H, d,
J = 6.6 Hz, H-1¢), 4.67 (2H, J = 7.4 Hz, H-1¢¢), 1.80 (6H, s, H-3¢),
1.74 (6H, s, H-3¢¢) ppm; 13C-NMR (CDCl3): d = 177.4 (C-9), 157.5
Chem Biol Drug Des 2010; 76: 43–58
45

Palmeira et al.
J = 7.9 Hz, H-2), 5.13 (1H, t, J = 6.1 Hz, CH2OH), 4.36-4.40 (1H, m,
H-3), 3.79 (3H, s, 3¢-OCH3), 3.71-3.76 (1H, m, CH2OH), 3.40-3.50 (1H,
m, 2-CH2) ppm; 13C-NMR (DMSO-d6): d= 175.1 (C-7), 155.4 (C-11a),
148.8 (C-4a), 147.7 (C-3¢), 147.2 (C-4¢), 145.9 (C-12a), 135.2 (C-10),
131.8 (C-12b), 126.7 (C-1¢), 126.0 (C-8), 124.5 (C-9), 121.0 (C-7a),
120.7 (C-6¢), 118.1 (C-11), 117.5 (C-6), 115.7 (C-6a), 115.4 (C-5¢),
113.9 (C-5), 111.8 (C-2¢), 78.0 (C-3), 76.5 (C-2), 59.9 (2-CH2), 55.7
(3¢-OCH3) ppm; IR (KBr): m = 3398, 1642, 1605, 1562, 1449, 1338,
1264, 1025, 861 cm)1; MS (EI, 70 eV) m ⁄ z (%): 406 (6) [M+], 356
(26), 341 (22), 327 (17), 306 (18), 292 (12), 281 (100), 261 (38), 253
(5), 228 (13), 215 (8), 180 (10), 165 (8), 141 (10), 115 (18), 105 (9),
91 (17), 77 (18), 63 (10), 55 (13); HRMS-ESI m ⁄ z calcd for C23H19O7
[M-H]+ 407.1131, found: 407.1130.
antibody (1:100, Dako). For nuclear counter staining, Vectashield
mounting medium for fluorescence with 4¢,6-diamidino-2-phenylin-
dole (DAPI; Vector Laboratories Inc., Peterborough, UK) was used.
Cells were observed in a LEICA DM2000 microscope (LEICA), and a
semiquantitative evaluation was performed by counting a minimum
of 300 cells per slide.
Cell cycle analysis
For the analysis of cell cycle phase distribution of DNA, K-562 cells
were treated as indicated earlier with complete medium (blank), the
solvent of the compounds (DMSO 0.2% v ⁄ v, control), or with com-
pound 5 (10 or 20 lM) and processed at 48 and 72 h after incubation
with the compounds. Cells were pelleted and fixed in 70% ethanol
for 10 min at room temperature. After centrifugation, cells were
stained for 30 min on ice in propidium iodide (40 lg ⁄ mL) and RNase
in PBS (100 lg ⁄ mL). At least 50 000 cells per sample were counted
and analyzed by flow cytometry (Epics XL-MCL; Beckman Coulter,
Brea, CA, USA), and the percentage of cells in the G1, S, and G2 ⁄ M
phases of the cell cycle was determined using the FlowJo 7.2 soft-
ware (Tree Star Inc., Ashland, OR, USA) after cell debris exclusion.
Cell lines
K-562 (human chronic myelogenous leukemia, erythroblastic), HL-60
(human acute myelocytic leukemia), BV-173 (human chronic myeloid
leukemia, lymphoid), and MRC-5 (immortalized fibroblasts) cell lines
were routinely maintained in RPMI-1640 (with Hepes and Glutamax;
Gibcoꢁ Invitrogen Cell Culture, Barcelona, Spain), with 10% fetal
bovine serum (FBS; Gibco) and incubated in a humidified incubator
at 37 ꢀC with 5% CO2 in air. All experiments were performed with
cells in exponential growth with viabilities over 90% and repeated
at least three times.
Cellular apoptosis analysis by the TUNEL assay
The levels of apoptosis of the K-562 cells treated with compound 5
were analyzed by the TUNEL assay, using the In Situ Cell Death
Detection Kit, Fluorescein (Roche Applied Science, Amadora, Portu-
gal). Briefly, K-562 cells were treated for 72 h with complete med-
ium (blank), the solvent of the compounds (DMSO 0.2% v ⁄ v,
control), or compound 5 (10 or 20 lM), as described previously.
Cells were then harvested, fixed in 4% paraformaldehyde in PBS,
and cytospins were prepared. Cells were permeabilized (0.1% Triton
X-100 in 0.1% sodium citrate) and incubated with TUNEL reaction
mix, according to the optimized procedure recommended by the
manufacturer (enzyme dilution 1:20). Vectashield mounting medium
for fluorescence with DAPI (Vector Laboratories, Inc) was used for
nuclear counter staining. Cells were observed in a LEICA DM2000
microscope (LEICA), and a semiquantitative evaluation was per-
formed by counting a minimum of 300 cells per slide.
Cellular response to the compounds in terms of
viable cell number
In the experiments to determine the cellular response of K-562 cells
to different compounds, cells were plated (2 · 105 cells ⁄ mL in 24-
well plates) and treated with complete medium (blank), with the
solvent of the compounds (DMSO 0.2% v ⁄ v, control), or with the
following compounds: 4, 6–8 (10 and 20 lM) and 5 (1, 2.5, 5, 10
and 20 lM). For further verification of the effects of the pyranox-
anthone 5, the HL-60 and BV-173 cell lines were plated under the
same conditions described for K-562 cells and were treated with
compound 5 (5, 10 and 20 lM). For the experiments with the adher-
ent cell line, MRC-5, compound 5 (10 or 20 lM) was added to the
cells only 24 h after plating (2 · 105 cells ⁄ mL in 6-well plates). For
all cell lines, viability and cell number were assessed following 24,
48, and 72 h of incubation with the compounds, using the trypan
blue exclusion assay. The equivalent volume of the compounds sol-
vent (DMSO) was added to the cells and analyzed at the same
time-points, as a control.
Protein expression analysis by Western blot
For analysis of protein expression, K-562 cells were treated with
complete medium (blank), the solvent of the compound (DMSO
0.2% v ⁄ v, control) or with 10 or 20 lM of compound 5, as
described previously, and processed 72 h after incubation. Cells
were lysed in Winman's buffer (1% NP-40, 0.1 M Tris-HCl pH 8.0,
0.15 M NaCl and 5 mM EDTA) with EDTA-free protease inhibitor
cocktail (Boehringer, Mannheim). Proteins were quantified using the
DC Protein Assay kit (BioRad, Hercules, CA, USA) and separated in
8% or 12% tris-glycine sodium dodecyl sulfate (SDS)-polyacrylamide
gels (18). Proteins were then transferred to a nitro-cellulose mem-
brane (GE Healthcare, Madrid, Spain). The membranes were incu-
bated with the following primary antibodies for poly (ADP-ribose)
polymerase (PARP) (1:4000, Santa Cruz Biotechnology Inc., Heidel-
berg, Germany), Actin (1:2000, Santa Cruz Biotechnology Inc), Bcl-xL
(1:200, Santa Cruz Biotechnology Inc), Bid (1:200, Santa Cruz Bio-
technology Inc), and further incubated with the respective secondary
Cellular proliferation analysis by the BrdU
incorporation assay
To assess the effect of the pyranoxanthone 5 on K-562 cellular pro-
liferation, BrdU incorporation was analyzed 72 h after incubation
with complete medium (blank), with the solvent of the compounds
(DMSO 0.2% v ⁄ v, control), or with compound 5 (10 or 20 lM). After
a pulse of BrdU (10 lM, Sigma) for 1 h, cells were washed, fixed
in 4% paraformaldehyde in phosphate buffer saline (PBS), and cyto-
spins prepared. After incubation in 2M HCl for 20 min, cells were
incubated with mouse anti-BrdU (1:10, Dakoꢁ, Glostrup, Denmark)
and further incubated with fluorescein-labeled rabbit anti-mouse
46
Chem Biol Drug Des 2010; 76: 43–58

Antitumor Mechanism of Action of a Pyranoxanthone
antibodies- horseradish peroxidase (HRP) conjugated (1:2000, Santa
Cruz Biotechnology). The signal was detected with the Amersham
ECL kit (GE Healthcare), Hyperfilm ECL (GE Healthcare) and Kodak
GBX developer and fixer twin pack (Sigma) as previously described
(19–21). The intensity of the bands obtained in each film was fur-
ther analyzed using the software Quantity One – 1D Analysis (Bio-
Rad).
2,2¢-dioxygenated benzophenone. The demethylation of the corre-
sponding dimethoxylated xanthone (Figure 2 c) afforded compound
4 in moderate yield (70%) (27). The synthetic approach used to
obtain the pyranoxanthone 5 was by treatment of 4 with prenyl
bromide in alkaline medium under reflux (12). At room tempera-
ture under similar conditions, the same building block furnished
the diprenyl derivative 6 (Figure 2 e). Recently, the combination of
microwave-assisted synthesis with the use of inorganic solid sup-
ports as catalysts, either with solvent or under solvent-free condi-
tions, provided the synthesis of prenylated xanthones with
enhanced reaction rates and high yields (28). Coupling of micro-
wave irradiation and montmorillonite K10 clay was successfully
applied herein to the synthesis of compound 5, with a drastic
increase in the yield of product as well as the decrease in the
reaction time (Figure 2 f and experimental conditions). Also the
synthesis of linear O-prenylxanthones such as compound 6 can be
optimized by applying microwave irradiation (28). 3-[(6-Bro-
mohexyl)oxy]-4-hydroxyxanthone (7) was obtained as the major
product of the drop-wise addition of dibromohexane to 4 (Figure 2
g). The synthetic approach for the xanthonolignoid 8 was based
on the oxidative coupling of 4 with coniferyl alcohol (9) (Figure 2
h). The mixture of trans and cis isomers of (€)-3¢-hydroxymethyl-
2¢-(4¢¢-hydroxy-3¢¢-methoxyphenyl)-1¢,4¢-dioxane[5¢,6¢-c]xanthone thus
obtained was treated with anhydrous K2CO3 to yield exclusively
the trans isomer 8 (Figure 2 i) (29).
Docking studies
Structures of several apoptosis pathway proteins were collected
from Protein Data Bank, namely 1ysw (structure of the anti-apoptotic
protein Bcl-2 complexed with an acyl-sulfonamide-based ligand),
1ysn (anti-apoptotic protein Bcl-xL complexed with an acyl-sulfon-
amide-based ligand), 1g5j (Bcl-xL in complex with Bad), 1pq1 (Bcl-xL
in complex with Bim), 3fdl (Bcl-xL in complex with BH3 Bim), 2yxj
(Bcl-xL in complex with ABT-737), and 3d7v (crystal structure of
Mcl-1 in complex with an Mcl-1 selective BH3 ligand); 2k7w and
1f16 (pro-apoptotic protein Bax), 2yv6 (pro-apoptotic protein Bak),
2bid (pro-apoptotic protein Bid), and 2jk7 (XIAP BIR3 bound to a
SMAC mimetic). Virtual screening was carried out on a workstation
with Intel Pentium processor, 3.00 GHz, 512 KB cache, 120 GB hard
drive and Nvidia Geforce 6600 GT graphic card running Linux Ubuntu
6.06. eHiTS 2009 (22,23) from SimBioSys Inc. was used for active
site detection and docking. Hyperchem (24) was used to draw and
optimize the structures of the pyranoxanthone 5 and of several
known Bcl-2 family antagonists (Table 1). Open Babel (25) was used
for manipulating the various chemical formats of ligands. PyMol
from DeLano Scientific (26) and Chimera from UCSF (27) were used
for visual inspection of results and graphical presentations.
The structures of compounds were established by 2D NMR tech-
niques such as Correlation Spectroscopy (COSY), HSQC, and
HMBC, which established the connectivities of the substituents on
the xanthone scaffold (5–8) as well as the orientation on the
1,4-dioxane ring (8). Nuclear Overhauser Effect (NOE) experiments
were used to determine the stereochemistry of compound 8.
No special preparation of the 3D structures was carried out
because eHiTS automatically evaluates all of the possible proton-
ation states for ligands and receptor. eHiTS ran using the crystallo-
graphic ligand or the whole protein as clip file. In the first case,
the program automatically detected the ligand and selected the part
of enzyme within a 7-ꢀ margin around this ligand to be the active
site. The known Bcl-2 antagonists and the pyranoxanthone 5 were
then docked to the active site using the intermediate mode of dock-
ing (accuracy set to 2). The affinity of each molecule to the recep-
tor was determined by a scoring function, eHiTS_Score, which is
included in the eHiTS software package.
Effects of compounds 4–8 on viable cell
number
The effect of the xanthones 4–8 was tested in the K-562 cell line,
derived from a blastic phase of human chronic myelogenous leuke-
mia, by verifying the number of viable cells using the trypan blue
exclusion assay. These effects were assessed with concentrations
ranging from 0 to 20 lM for 24, 48, or 72 h. Results were com-
pared with those obtained with appropriate controls: DMSO control
(solvent of the compounds, 0.2% v ⁄ v) and blank treatment (cells
incubated with complete medium). All the compounds caused some
degree of reduction in the number of viable cells (Figure 3). Results
also reveal that compound 5 is the most potent, causing a dose-
and time-dependent reduction in viable cell number and presenting
an IC50 (concentration that inhibits 50% of viable cell number) of
20 lM upon a 72-h treatment (Figure 4A).
Results
Synthesis
The generation of appendance diversity was achieved by different
synthetic pathways for 3,4-dihydroxyxanthone (4) holding a pyrano
ring (5), prenyl groups (6), a bromohexyl chain (7), and a conyferil
unit (8), as represented in Figure 2. The structures of compounds
4–8 were established by spectroscopic methods (IR, 1H and 13C
NMR, HMBC and HSQC) and mass spectrometry.
Furthermore, the effect of the pyranoxanthone 5 on viable cell num-
ber was investigated in two additional leukemia cell lines: HL-60
and BV-173 (Figure 4B and C, respectively). The results were com-
pared to those previously obtained in the K-562 cell line (Figures 3
and 4A). Compound 5 was capable of reducing the number of via-
ble cells in HL-60 (IC50 ꢀ 6–7 lM at both 48 and 72 h, Figure 4B)
and BV-173 (IC50 ꢀ 14 lM following 72 h, Figure 4C), presenting
IC50 values lower than those obtained in the K-562 cell line.
Compound 4 (Figure 2) was prepared through condensation by
Friedel–Crafts acylation (Figure 2 a) of an appropriate substituted
benzoyl chloride with a phenolic derivative followed by the nucle-
ophilic addition–elimination cyclization step (Figure 2 b) of the
Chem Biol Drug Des 2010; 76: 43–58
47

Palmeira et al.
Table 1: The structures of non-peptidic antagonists of Bcl-2 and Bcl-xL
Cl
N
CH3
H3C
CH3
OH
N
O
CH3
O
H3C
HO
O
HO
HO
O
O
N+ O–
NH
OH
OH
O
O
NH
S
OH
NH
O
CH3
NH
O
O
O
O
S
CH3
H3C
CH3
H3C N
CH3
ABT-737
Antimycin A3
Apogossypol
Cl
I
HO
H2C
O
O
H3C
CH3
O
OH
O
NH
N
S
N+ N+
H3C
S
Cl
Br
HO
OH
CH3
S
O
O
Dibenzodiazocine
derivative
BH3I-1
BH3I-2
OH
OH
HO
H3C
CH3
HN
H3
C
H3C
OH
O
OH
OH
OH
NH
OH
OH
N
C
O
O
HO
HO
OH
CH3
H3C
O
HO
O
H3
H3C
CH3
OH
Epigallocatechin
gallate (EGCG)
Gossypol
GX015-070
O
OH
O
OH
OH
HO
CH3
NH2
O
CH3
O
O
CH3
OH
H3C
O
H3C
O
CH3
CH3
O
O
HO
HO
OH
O
Br
HA14-1
Terphenyl
Theaflavanin
48
Chem Biol Drug Des 2010; 76: 43–58

Antitumor Mechanism of Action of a Pyranoxanthone
Figure 3: Response of K-562 cells to compounds 4–8. K-562 cells were treated with several concentrations of 4–8, with complete med-
ium (blank), or with DMSO control for 24, 48 or 72 h. Viable cell number was determined by the trypan blue exclusion assay. Results are
expressed as the mean € SE of three independent experiments.
Additionally, compound 5 was further tested in an immortalized fibro-
blast cell line (MRC-5) and no alterations in the viable cell number
were observed during 72 h of treatment at 10 or 20 lM (data not
shown), suggesting that this compound preferentially reduces the
growth of these leukemia cells.
The effect of the pyranoxanthone 5 on cellular proliferation of K-
562 cells was determined by using the BrdU incorporation assay.
The proliferation levels were calculated as the percentage of cells
undergoing proliferation. Results show that cells treated with 10 or
20 lM of compound 5 for 72 h incorporated less BrdU than the
appropriate control cells. The levels of proliferation decreased from
30.8% to 26.2% or 16.9%, after treatment with 10 or 20 lM of
compound 5, respectively (Figure 5).
Further characterization of the cellular effects
of 3,4-dihydro-12-hydroxy-2,2-dimethyl-2H,6H-
pyrano[3,2-b]xanthen-6-one (5)
To understand the mechanism of action of the pyranoxanthone 5,
cellular proliferation, cell cycle, and apoptosis were studied in the
K-562 cell line.
Analysis of the effect of the pyranoxanthone 5 on cell cycle was
performed by flow cytometry. Compound 5 showed a dose- and
time-dependent increase in the percentage of cells in the S-
phase of the cell cycle, with a concomitant decrease in cells in
Chem Biol Drug Des 2010; 76: 43–58
49

Palmeira et al.
that increasing concentrations of this xanthone derivative caused
increased DNA fragmentation, with the levels of apoptosis rising
from 1.3% (DMSO control) to 4.1% or 7.0% after treatment with 10
or 20 lM of compound 5, respectively (Figure 7).
A
The effect of the pyranoxanthone 5 on the expression of some
proteins involved in the apoptotic process was determined by
Western blot. Results show that incubation with this compound
leads to a dose-dependent decrease in PARP with concomitant
increase of the cleaved form of this protein (cleaved PARP, Fig-
ure 8A). Additionally, a small decrease in the expression levels of
anti-apoptotic Bcl-xL (Figure 8B) and an increase in the expression
of the pro-apoptotic protein Bid (Figure 8C) were also detected
upon compound 5 treatment. Any possible alteration in the levels
of expression of Bcl-2, Bak, and XIAP was not conclusive (data
not shown).
B
To further understand the possible mechanism of action of the pyra-
noxanthone 5 and its involvement in the apoptotic intrinsic path-
way, a docking study was performed using the software eHiTS
2009 (22,23). For each protein to be analyzed, we employed crystal
structures retrieved from the Protein Data Bank, PDB (30), and
Table 2 lists the docking scores (kJ ⁄ mol) of the ligands against the
Bcl-2 family proteins studied (along with the PDB IDs) Table S1 (in
Supporting Information) presents the docking scores (kJ/mol) of the
pyranoxanthone 5 against other proteins involved in the apoptotic
intrinsic pathway.
C
The docking scores of known Bcl-2 family antagonists (Tables 1 and
2) were used as positive controls. Fifty unrelated molecules, ran-
domly chosen from the NCI compound database, were used as neg-
ative controls (Table S2, Supporting Information). Docking is a
method that predicts the preferred orientation of a ligand to a pro-
tein, when bound to each other, to form a stable complex. In turn,
knowledge of the preferred orientation may be used to predict the
binding affinity between them, using scoring functions. In general,
the docking score is (or can be related to) the DG for complex for-
mation. A negative score indicates that the complex formation is
energetically favorable and the more negative the value, the better
would be the binding affinity. The 50 unrelated compounds used as
'decoys' (Table S2, Supporting information) showed higher scores
than any of the positive controls and test molecule (>)15 kJ ⁄ mol).
The pyranoxanthone 5 once docked to Bcl-xL holds scores similar to
known Bcl-xL inhibitors (Table 2). Compound 5 docks to the same
binding pocket as the known Bcl-xL inhibitor ABT-737 (pdb code
2yxj) (Figure 9). Additionally, compound 5 establishes interactions
within the cavity where BH3-only proteins such as Bim (3fdl) and
Bad (1g5j) bind (Figure 10), suggesting that it may block the forma-
tion of a complex between anti-apoptotic Bcl-xL and pro-apoptotic
proteins. Scoring results for compound 5 were also in the same
order of magnitude as the controls (known Bcl-2 family inhibitors)
for other members of the Bcl-2 family (Table 2).
Figure 4: Response of three leukemia cell lines to the pyranox-
anthone 5. K-562, HL-60 and BV-173 cells were treated with com-
pound 5 for 24, 48 and 72 h. Viable cell number was analyzed with
the trypan blue assay. Plots represent the percentage of viable cells
(as a % of the blank cells) against the concentration of compound
5. Data for the K-562 cell line correspond to that shown in Fig-
ure 2. The broken line indicates the IC50. (A) Effects of compound 5
on K-562 cell line, (B) on HL-60 cell line, and (C) on BV-173 cell
line. Each data point represents the mean € SE obtained from three
independent experiments. Appropriate controls with DMSO were
included in the experiments (data not shown).
the G2 ⁄ M phase (Figure 6). The pyranoxanthone 5 therefore
seems to be a strong inducer of S-phase cell cycle arrest, after
72 h of treatment, exhibiting an almost complete loss of cells in
G2 ⁄ M.
Discussion
Additionally, it was investigated whether compound 5 induced
apoptosis in the K-562 cell line, by the TUNEL assay. Results show
A large group of xanthones has been described as having antitumor
properties namely in leukemia models (1). Based on natural
50
Chem Biol Drug Des 2010; 76: 43–58

Antitumor Mechanism of Action of a Pyranoxanthone
A
B
Figure 5: Effects of the pyranoxanthone 5 in proliferation of K-562 cells. K-562 cells were treated for 72 h with complete medium (blank),
DMSO (control) or with 10 or 20 lM of compound 5. The BrdU incorporation assay was used to detect proliferation and cells were analyzed
under a fluorescent microscope. (A) Representative photographs of the treated cells. Nuclei are labeled with DAPI. A typical BrdU-positive cell
is labeled with the arrow 1 and a typical BrdU-negative cell is labeled with the arrow 2. (B) Semi-quantitative evaluation was carried out by
counting a minimum of 300 cells per treatment in each independent experiment. Results are expressed as the mean € SE of three indepen-
dent experiments. Statistical significance was tested by paired t-test using the DMSO as a negative control. ***Indicates p < 0.001;
*Indicates 0.01 < p < 0.05.
prenylated xanthones (12,31), and other analogues with antitumor
activity (16,29), structural modifications were produced on the hit
compound, 3,4-dihydroxyxanthone (4) (15), to prepare four active
compounds (5–8). The present work shows that these derivatives
have improved the ability to reduce the viable cell number of K-562
leukemia cells when compared to 3,4-dihydroxyxanthone (4). Com-
pound 5, which includes a pyrano ring at the [3,2-b], was the most
potent compound. The synthesis of compound 5 was improved by
solid-phase catalysis and microwave-assisted synthesis, to get bet-
ter yield and shorter reaction time (Figure 2, experimental section).
cells, blastic phase human chronic myelogenous leukemia. The rea-
son for this is most likely related to either: (i) the different uptake
of the compound by the cells or; (ii) the mechanism of action of the
compound. Several differences can be found among the leukemia
cell lines used, such as documented lack of expression of wt p53
tumor suppressor gene in the HL-60 and K-562 cells (32–34), or
constitutive activation of the tyrosine kinase (Abl) in both K-562
and BV-173 cells (26), among others. These differences, along with
the mechanism of action of the compound, most likely explain the
different IC50 values found in the three cell lines. Because the
HL-60 cell line is a model of acute promyeloid leukemia, the rela-
tively low IC50 value obtained for this cell line is particularly inter-
esting regarding future applications for this compound or its new
derivatives. Furthermore, the pyranoxanthone 5 is a promising mole-
cule considering that etoposide, currently used in the clinic, has
The results reported here demonstrate that the pyranoxanthone 5
can reduce the number of viable cells of other leukemia cell lines.
Compound 5 is particularly active in HL-60 cells, a model of human
acute promyelocytic leukemia, when compared to BV-173 and K-562
Chem Biol Drug Des 2010; 76: 43–58
51

Palmeira et al.
previously given us an IC50 of 9 lM upon 48-h treatment of the
K562 cell line (35).
G2 ⁄ M phase of the cell cycle. On the other hand, the linear ana-
logues induced a partial accumulation of cells in the G1 phase of
the cell cycle (36). Results presented herein clearly show that treat-
ment of K-562 cells with compound 5, a linear pyran-fused xan-
The pyran-fused xanthones have been previously found to act selec-
tively at specific phases of the cell cycle (8). In general, the angular
analogues, which possess a structural similarity with acronycine (3),
have been shown to induce a partial accumulation of cells in the
thone, caused
a concentration- and time-dependent delayed
progression of the K-562 cells through the S-phase of the cell cycle,
resulting in an S-phase cell cycle arrest with an almost complete
disappearance of cells in G2 ⁄ M.
A
Further studies of this compound demonstrated that it does not
have a detectable effect on the viable cell number of an immortal-
ized fibroblast cell line, within the 72-h incubation period. This find-
ing suggests that leukemia cells are more sensitive to this
compound, probably because of the faster growth rate of leukemia
cells when compared to the fibroblast cell line. The aforementioned
results agree with this hypothesis, because the pyranoxanthone 5
is capable of reducing cellular proliferation by arresting cells in the
S-phase of the cell cycle, which will be more evident in cell lines
with faster doubling time (growth rate). However, this preference of
the compound for leukemia cells should be further confirmed in
other 'normal' cellular models.
B
Several compounds previously showed induction of S-phase arrest,
in different types of human cells, including the well-known antican-
cer agents like cisplatin (37,38), mitomycin C (39,40), hydroxyurea
(41,42), and the recent polyphenolic anticancer agent resveratrol
(43,44). Several anticancer agents are known to inhibit topoisomer-
ase activity and other key enzymes in DNA duplication (45,46). Fur-
thermore, simple prenylated xanthones have proven to inhibit
topoisomerases I and II (47–49) namely psorospermin (1) which is,
a topoisomerase poison (50,51). Therefore, it would be interesting
to further investigate mechanisms of S-phase arrest induced by the
pyranoxanthone 5. Future work may also include research of the
levels of cyclin-dependent kinases (CDKs, which govern the progres-
sion through the cell cycle) and of the levels of CDK inhibitors (such
as p21 or p27), to deeper understand the mechanism of action of
this compound.
C
a
Figure 6: Cell cycle analysis of K-562 cells treated with the pyr-
anoxanthone 5. K-562 cells were cultured with 10 or 20 lM of
compound 5 for 48 or 72 h. Untreated cells were used as control
(blank). A DMSO treatment was also included (control). Cell cycle
analysis was performed by flow cytometry following propidium
iodide staining and a minimum of 50 000 cells were analyzed. (A)
Dot plot representing region R1 which was used to trace the histo-
gram. The dot plot is representative of triplicate experiments. (B)
Representative histograms of the cell cycle profile of three indepen-
dent experiments. (C) Graphs representing the differences in the cell
cycle distribution of K-562 cells at (a) 48 h and (b) 72 h following
treatment. Results were obtained with the FlowJo 7.2 software.
Means € SE were calculated from triplicates experiments. Statisti-
cal significance was tested by paired t-test using DMSO as a nega-
tive control. ***Indicates p < 0.001; **Indicates 0.001 < p < 0.01;
*Indicates 0.01 < p < 0.05; ns indicates not significant, i.e.,
p > 0.05 (n = 3).
b
52
Chem Biol Drug Des 2010; 76: 43–58

Antitumor Mechanism of Action of a Pyranoxanthone
A
B
Figure 7: Effects of the pyranoxanthone 5 on programmed cell death of K-562 cells. K-562 cells were treated for 72 h with complete
medium (blank), DMSO (control) or with 10 or 20 lM of compound 5. Programmed cell death was detected by the TUNEL assay. (A) Represen-
tative photographs of the treated cells. Nuclei are labeled with DAPI. A typical TUNEL-positive cell is labeled with the arrow 1 and a typical
TUNEL-negative cell is labeled with the arrow 2. (B) Semi-quantitative evaluation was carried out by counting a minimum of 300 cells per
treatment in each independent experiment. Results are the mean € SE of three independent experiments. Statistical significance was tested
by paired t-test using the DMSO as a negative control. ***Indicates p < 0.001; *Indicates 0.01 < p < 0.05.
An increase in programmed cell death was also found following
treatment with pyranoxanthone 5, even though the determined lev-
els of apoptosis were not very high, possibly attributable to the
method of detection used (TUNEL). To verify the effective induction
of apoptosis, a Western blot analysis showed an evident increase
in PARP cleavage following treatment with compound 5 (Figure 8A),
indicating that this compound induces cell death by apoptosis in
K-562 cells. In accordance with this, it is worth noticing that treat-
ment with compound 5 slightly decreases Bcl-xL (anti-apoptotic)
(Figure 8B) and increases Bid (pro-apoptotic) (Figure 8C) levels,
which supports the proposal that this compound induces cellular
death by apoptosis, namely by inhibition of some Bcl-2 family pro-
teins, directly or indirectly. There was no significant consistent
alteration in the levels of expression of Bcl-2, Bak and XIAP (data
not shown). Therefore the TUNEL and Western blot assays indicate
that the decrease in viable cell number previously observed after
treatment with compound 5 was caused, at least in part, by pro-
grammed cell death.
A virtual screening on members of the mitochondrial pathway
involved in programmed cell death was also performed, to find
which ones could be affected by compound 5. This would provide
an indication about compound 5's potential as Bcl-2 family anta-
gonist. A docking study of compound 5 against several known Bcl-2
inhibitors and several anti-apoptotic proteins was performed. The
pro-survival proteins such as Bcl-2, Bcl-xL and Mcl-1 share four
domains of sequence homology known as Bcl-2 homology 1 (BH1),
BH2, BH3, and BH4 (52). Other key players that orchestrate apopto-
sis are the pro-apoptotic BH3-only proteins, such as Bad, Bim and
Bid (53,54). The structure of Bcl-xL in complex with the Bad and
Bim BH3 domains (Figure 10) sets the foundations for many of the
current models for the life and death switch. In fact, in recent
Chem Biol Drug Des 2010; 76: 43–58
53

Palmeira et al.
A
B
C
Figure 8: The pyranoxanthone 5 alters the expression of some apoptotic proteins. K-562 cells were treated for 72 h with complete med-
ium (Blank), DMSO (control) or with 10 or 20 lM of compound 5. Proteins were extracted and analyzed by Western blot. Actin was used as
a loading control. The left panel shows representative results and the right panel shows the results of the densitometry analysis, as
mean € SE of at least three independent experiments. (A) Treatment with compound 5 caused cleavage of PARP, (B) a slight reduction in Bcl-
xL and (C) increase in Bid.
Table 2: Scoring results (kJ ⁄ mol) for the pyranoxanthone 5 and
years, several screening studies have been undertaken to find small
molecules that could be used as anticancer drugs by binding to Bcl-
2 family of proteins (55–60). Results from these studies include the
BH3I class of compounds, HA14, ABT-737, gossypol and even the
natural product antimycin A. From the obtained results, we hypo-
thesize that the pyranoxanthone 5 may block the binding of anti-
apoptotic Bcl-xL to pro-apoptotic Bad or Bim (Figure 10), thus
blocking the Bcl-xL/BH3 binding groove (56). Therefore, compound 5
could bind to Bcl-xL protein, triggering apoptosis. However, the indi-
cations arising from the results of docking studies must be con-
firmed experimentally, for example by immunoprecipitation, to
determine whether compound 5 is acting directly on Bcl-xL or in
any other protein of the apoptotic pathway leading indirectly to
decreased levels of Bcl-xL and increased levels of Bid. As shown in
Figure 9A–C, compound 5 fits along the ABT-737 binding cavity in
spite of being a much smaller molecule than the known inhibitor
ABT-737, an organic ligand with high binding affinity for the hydro-
phobic groove of Bcl-xL (61,62). Therefore, for the design of novel
several Bcl-2 family antagonists obtained with eHits
Bcl-2
Bcl-xL
Bcl-xL
Bcl-xL
Mcl-1
(1ysw)
(1g5j)
(3fdl)
(2yxj)
(3d7v)
ABT-737
)41.69
)23.55
)28.23
)20.67
)24.62
)19.39
)37.26
)22.45
)34.12
)20.54
)21.77
)19.72
)38.10
)23.98
)33.14
)23.59
)24.48
)20.56
)38.35
)24.63
)32.24
)23.77
)24.70
)20.03
)35.02
)22.40
)38.53
)22.02
)23.19
)19.10
Antimycin A3
Apogossypol
BH3I-1
BH3I-2
Dibenzodiazocine
derivative
EGCG
Gossypol
)24.65
)28.81
)22.58
)20.96
)30.76
)25.84
)21.58
)25.85
)30.57
)21.42
)19.69
)37.74
)25.07
)23.52
)25.29
)37.47
)23.48
)22.45
)37.23
)25.58
)25.13
)24.70
)30.92
)22.32
)24.95
)33.48
)26.18
)25.17
)23.40
)37.89
)22.56
)18.09
)33.00
)26.64
)25.98
GX015-070
HA14-1
Terphenyl derivative
Theaflavanin
Pyranoxanthone 5
54
Chem Biol Drug Des 2010; 76: 43–58

Antitumor Mechanism of Action of a Pyranoxanthone
A
A
B
C
B
D
Figure 9: Docking poses of the pyranoxanthone 5 on Bcl-xL. (A)
Several poses of compound 5 (light gray) docked on the Bcl-xL
antagonists docking site; crystallographic inhibitor ABT-737 is repre-
sented in black. (B) Surface representation of Bcl-xL and the pyra-
noxanthone 5 best docking pose; broken white lines represent non-
covalent interactions with the receptor. (C) Ribbon representation of
Bcl-xL, compound 5 best ranking score (stick) and cristalographic
ABT-737 ligand (line). (D) Compound 5 best docking pose on Bcl-xL
and important residues involved in the binding.
Figure 10: Comparison of the pyranoxanthone 5 and pro-apop-
totic binding modes on Bcl-xL. Compound 5 best docking pose on
Bcl-xL binding site; transparent surface representing the binding of
pro-apoptotic protein (A) Bad and (B) Bim.
xanthone derivatives, a good strategy to get even better IC50 values
could be the molecular extension of the pyranoxanthone 5 to
occupy the ABT-737 binding cavity. These docking results represent
preliminary data for future optimization studies and the design of
novel analogues.
one (4) in leukemia cells. New insights into the antitumor molecular
mechanism of action of a new pyranoxanthone were achieved.
3,4-Dihydro-12-hydroxy-2,2-dimethyl-2H,6H-pyrano[3,2-b]xanthen-6-one
(5) exhibited antiproliferative and apoptotic effects in leukemia cell
lines. The docked geometries of the pyranoxanthone 5, supported
by experimental data, suggest that this molecule may possibly
target Bcl-xL.
Conclusion
The results reported here show that the synthetic approaches used
were successful in improving the cytotoxicity of 3,4-dihydroxyxanth-
Chem Biol Drug Des 2010; 76: 43–58
55

Palmeira et al.
12. Castanheiro R.A., Pinto M.M., Silva A.M., Cravo S.M., Gales L.,
Damas A.M., Nazareth N., Nascimento M.S., Eaton G. (2007)
Dihydroxyxanthones prenylated derivatives: synthesis, structure
elucidation, and growth inhibitory activity on human tumor cell
lines with improvement of selectivity for MCF-7. Bioorg Med
Chem;15:6080–6088.
13. Pinto M.M., Castanheiro R. (2008) Natural prenylated xanthones:
chemistry and biological activities. In: Brahmachari G., editor.
Natural Products: Chemistry, Biochemistry and Pharmacology.
New Delhi, India: Narosa Publishing House Pvt. LTD; p. 520–
675.
Acknowledgments
FCT (I&D, no. 226 ⁄ 2003 and I&D, no. 4040 ⁄ 07), FEDER, POCI, U.
Porto and Caixa Geral de Depꢁsitos for financial support. The
authors thank Professor Clara Sambade for helpful comments and
critical review of the manuscript and Sara Cravo for technical sup-
port in microwave equipment.
References
1. Pinto M.M., Sousa M.E., Nascimento M.S. (2005) Xanthone
derivatives: new insights in biological activities. Curr Med
Chem;12:2517–2538.
14. Saraiva L., Fresco P., Pinto E., Sousa E., Pinto M., Goncalves J.
(2003) Inhibition of protein kinase C by synthetic xanthone deriv-
atives. Bioorg Med Chem;11:1215–1225.
2. Kupchan S.M., Streelman D.R., Sneden A.T. (1980) Psorospermin,
a new antileukemic xanthone from Psorospermum febrifugum. J
Nat Prod;43:296–301.
3. Rewcastle G.W., Atwell G.J., Li Z.A., Baguley B.C., Denny W.A.
(1991) Potential antitumor agents. 61. Structure-activity relation-
ships for in vivo colon 38 activity among disubstituted 9-oxo-9H-
xanthene-4-acetic acids. J Med Chem;34:217–222.
4. Jameson M.B., Thompson P.I., Baguley B.C., Evans B.D., Harvey
V.J., Porter D.J., McCrystal M.R., Small M., Bellenger K., Gum-
brell L., Halbert G.W., Kestell P. (2003) Clinical aspects of a
phase I trial of 5,6-dimethylxanthenone-4-acetic acid (DMXAA),
a novel antivascular agent. Br J Cancer;88:1844–1850.
5. Hakim E.H., Asnizar, Yurnawilis, Aimi N., Kitajima M., Takayama
H. (2002) Artoindonesianin P, a new prenylated flavone with
cytotoxic activity from Artocarpus lanceifolius. Fitotera-
pia;73:668–673.
6. Wu X., Cao S., Goh S., Hsu A., Tan B.K. (2002) Mitochondrial
destabilisation and caspase-3 activation are involved in the
apoptosis of Jurkat cells induced by gaudichaudione A, a cyto-
toxic xanthone. Planta Med;68:198–203.
7. Elomri A., Mitaku S., Michel S., Skaltsounis A.L., Tillequin F.,
Koch M., Pierrꢂ A., Guilbaud N., Lꢂonce S., Kraus-Berthier L.,
Rolland Y., Atassi G. (1996) Synthesis and cytotoxic and antitu-
mor activity of esters in the 1,2-dihydroxy-1,2-dihydroacronycine
series. J Med Chem;39:4762–4766.
8. Ghirtis K.P.N., Marakos P., Skaltsounis A.L., Leonce S., Caignard
D.H., Atassi G. (2000) Synthesis and Conformational Analysis of
Some New Pyrano[2,3-c]xanthen-7-one and Pyrano[3,2-b]xanthen-
6-one Derivatives with Cytotoxic Activity. Heterocycles;53:93.
9. Kostakis I.K., Magiatis P., Pouli N., Marakos P., Skaltsounis A.L.,
Pratsinis H., Lꢂonce S., Pierrꢂ A. (2002) Design, synthesis, and
antiproliferative activity of some new pyrazole-fused amino
derivatives of the pyranoxanthenone, pyranothioxanthenone, and
pyranoacridone ring systems: a new class of cytotoxic agents. J
Med Chem;45:2599–2609.
10. Kostakis I.K., Pouli N., Marakos P., Mikros E., Skaltsounis A.L.,
Leonce S., Atassi G., Renard P. (2001) Synthesis, cytotoxic activ-
ity, NMR study and stereochemical effects of some new
pyrano[3,2-b]thioxanthen-6-ones and pyrano[2,3-c]thioxanthen-
7-ones. Bioorg Med Chem;9:2793–2802.
11. Pinto M.M., Sousa E.P. (2003) Natural and synthetic xanthonolig-
noids: chemistry and biological activities. Curr Med Chem;10:1–
12.
15. Pedro M., Cerqueira F., Sousa M.E., Nascimento M.S., Pinto M.
(2002) Xanthones as inhibitors of growth of human cancer cell
lines and their effects on the proliferation of human lympho-
cytes in vitro. Bioorg Med Chem;10:3725–3730.
16. Gottlieb O.R.M., Oliveira G.G., Melo M.T. (1970) Xanthones from
Kielmeyera speciosa. Phytochemistry;9:2537–2544.
17. Sousa E.P., Nazareth N., Gales L., Damas A., Nascimento M.S.J.,
Pinto M. (2009) Bromoalkoxyxanthones as promising antitumor
agents: synthesis, crystal structure and effect on human tumor
cell lines. Eur J Med Chem;44:1–6.
18. Sambrook J., Fritsch E.F., Maniatis T. (1989) Molecular Cloning:
A Laboratory Manual. NY: Cold Spring Harbor.
19. Lima R.T., Martins L.M., Guimaraes J.E., Sambade C., Va-
sconcelos M.H. (2004) Specific downregulation of bcl-2 and
xIAP by RNAi enhances the effects of chemotherapeutic agents
in MCF-7 human breast cancer cells. Cancer Gene
Ther;11:309–316.
20. Vasconcelos M.H., Beleza S.S., Quirk C., Maia L.F., Sambade C.,
Guimaraes J.E. (2000) Limited synergistic effect of antisense oli-
gonucleotides against bcr-abl and transferrin receptor mRNA in
leukemic cells in culture. Cancer Lett;152:135–143.
21. Lima R.T.G., Vasconcelos M.H. (2007) Overcoming K562Dox resis-
tance to STI571 (Gleevec) by downregulation of P-gp expression
using siRNAs. Cancer Ther;5:67–76.
22. Zsoldos Z., Reid D., Simon A., Sadjad B.S., Johnson A.P. (2006)
eHiTS: an innovative approach to the docking and scoring func-
tion problems. Curr Protein Pept Sci;7:421–435.
23. Zsoldos Z., Reid D., Simon A., Sadjad S.B., Johnson A.P. (2007)
eHiTS: a new fast, exhaustive flexible ligand docking system.
J Mol Graph Model;26:198–212.
24. Froimowitz M. (1993) HyperChem: a software package for com-
putational chemistry and molecular modeling. BioTech-
niques;14:1010–1013.
25. O'Boyle N.M., Hutchison G.R. (2008) Pybel: a Python wrapper for
the OpenBabel cheminformatics toolkit. Chem Cent J;9:5.
26. Ordog R. (2008) PyDeT, a PyMOL plug-in for visualizing geomet-
ric concepts around proteins. Bioinformation;2:346–347.
27. Pettersen E.F., Goddard T.D., Huang C.C., Couch G.S., Greenblatt
D.M., Meng E.C., Ferrin TE. (2004) UCSF Chimera-a visualization
system for exploratory research and analysis.
Chem;25:1605–1612.
J Comput
28. Castanheiro R.P., Pinto M.M., Cravo S., Pinto D., Silva A., Kijjoa
A. (2009) Improved methodologies for synthesis of prenylated
xanthones by microwave irradiation and combination of hetero-
56
Chem Biol Drug Des 2010; 76: 43–58

Antitumor Mechanism of Action of a Pyranoxanthone
geneous catalysis (K10 clay) with microwave irradiation. Tetrahe-
dron;65:3848–3857.
tor NF-kappaB, inhibits proliferation, causes S-phase arrest, and
induces apoptosis of acute myeloid leukemia cells.
Blood;102:987–995.
29. Sousa E.P.S., Pinto M.M., Pedro M.M., Cerqueira F.A.M.,
Nascimento M.S.J. (2002) Isomeric Kielcorins and Dihydroxyx-
anthones: Synthesis, Structure Elucidation, and Inhibitory Activ-
ities of Growth of Human Cancer Cell Lines and on the
Proliferation of Human Lymphocytes In Vitro. Helv Chim
Acta;85:2862–2876.
44. Tyagi A., Singh R.P., Agarwal C., Siriwardana S., Sclafani R.A.,
Agarwal R. (2005) Resveratrol causes Cdc2-tyr15 phosphorylation
via ATM ⁄ ATR-Chk1 ⁄ 2-Cdc25C pathway as a central mechanism
for S phase arrest in human ovarian carcinoma Ovcar-3 cells.
Carcinogenesis;26:1978–1987.
30. Berman H.M., Westbrook J., Feng Z., Gilliland G., Bhat T.N.,
Weissig H., Shindyalov I.N., Bourne P.E. (2000) The protein data
bank. Nucleic Acids Res;28:235–242.
45. Maga G., Hubscher U. (2008) Repair and translesion DNA
polymerases as anticancer drug targets. Anticancer Agents Med
Chem;8:431–447.
31. Pouli N., Marakos P. (2009) Fused xanthone derivatives as anti-
proliferative agents. Anticancer Agents Med Chem;9:77–98.
32. Sugimoto K., Toyoshima H., Sakai R., Miyagawa K., Hagiwara
K., Ishikawa F., Takaku F., Yazaki Y., Hirai H. (1992) Frequent
mutations in the p53 gene in human myeloid leukemia cell lines.
Blood;79:2378–2383.
33. Gahmberg C.G., Andersson L.C. (1981) K562-a human leukemia
cell line with erythroid features. Semin Hematol;18:72–77.
34. Law J.C.R., Yalowich J.C., Leder G.H., Ferrell R.E. (1993) Muta-
tional inactivation of the p53 gene in the human erythroid leuke-
mic K562 cell line. Leuk Res;17:1045–1050.
35. Lima R.T., Martins L.M., Guimar¼es J.E., Sambade C., Vasconce-
los M.H. (2006) Chemosensitization effects of XIAP downregula-
tion in K562 leukemia cells. J Chemother;18:98–102.
36. Ghirtis K.P., Marakos P., Skaltsounis A.L., Leonce S., Atassi G.,
Caignard D.H. (2001) Design and synthesis of some new pyr-
anoxanthenones with cytotoxic activity. J Heterocycl Chem;38:
147.
37. Gong J.G., Costanzo A., Yang H.Q., Melino G., Kaelin W.G. Jr,
Levrero M., Wang J.Y.J. (1999) The tyrosine kinase c-Abl regu-
lates p73 in apoptotic response to cisplatin-induced DNA dam-
age. Nature;399:806–809.
46. Palchaudhuri R., Hergenrother P.J. (2007) DNA as a target for
anticancer compounds: methods to determine the mode of bind-
ing and the mechanism of action. Curr Opin Biotechnol;18:497–
503.
47. Zhu F.L., Zhou S.N., Vrijmoed L.L.P. (2004) Xanthone Derivatives
Isolated from two Mangrove EndophyticFungi #2526 and #1850
from the South China Sea. Nat Prod Res Dev;16:406.
48. Froelich-Ammon S.J., Osheroff N. (1995) Topoisomerase poisons:
harnessing the dark side of enzyme mechanism.
Chem;270:21429–21432.
J Biol
49. Tosa H.I., Tanaka T., Nozaki H., Ikeda S., Tsutsui K., Tsutsui
K., Yamada M., Fujimori S. (1997) Inhibitory Activity of Xan-
thone Derivatives Isolated from some Guttiferaeous Plants
Against DNA Topoisomerases
Bull;45:418–420.
I
and II. Chem Pharm
50. Permana P.A., Ho D.K., Cassady J.M., Snapka R.M. (1994) Mech-
anism of action of the antileukemic xanthone psorospermin:
DNA strand breaks, abasic sites, and protein-DNA cross-links.
Cancer Res;54:3191–3195.
51. Kwok Y., Hurley L.H. (1998) Topoisomerase II site-directed alkyl-
ation of DNA by psorospermin and its effect on topoisomerase
II-mediated DNA cleavage. J Biol Chem;273:33020–33026.
52. Adams J.M., Cory S. (1998) The Bcl-2 protein family: arbiters of
cell survival. Science;281:1322–1326.
38. Almeida G.M., Duarte T.L., Farmer P.B., Steward W.P., Jones
G.D. (2008) Multiple end-point analysis reveals cisplatin damage
tolerance to be a chemoresistance mechanism in a NSCLC
53. Fernandez-Luna J.L. (2008) Regulation of pro-apoptotic BH3-only
proteins and its contribution to cancer progression and chemore-
sistance. Cell Signal;20:1921–1926.
model: implications for predictive testing. Int
cer;122:1810–1819.
J
Can-
39. Kang S.G., Chung H., Yoo Y.D., Lee J.G., Choi Y.I., Yu Y.S. (2001)
Mechanism of growth inhibitory effect of Mitomycin-C on cul-
tured human retinal pigment epithelial cells: apoptosis and cell
cycle arrest. Curr Eye Res;22:174–181.
54. Zha J., Harada H., Osipov K., Jockel J., Waksman G., Korsmeyer
S.J. (1997) BH3 domain of BAD is required for heterodimeriza-
tion with BCL-XL and pro-apoptotic activity.
Chem;272:24101–24104.
J
Biol
40. Yun J., Zhong Q., Kwak J.Y., Lee W.H. (2005) Hypersensitivity of
Brca1-deficient MEF to the DNA interstrand crosslinking agent
mitomycin C is associated with defect in homologous recombi-
nation repair and aberrant S-phase arrest. Oncogene;24:4009–
4016.
41. Hammond E.M., Green S.L., Giaccia A.J. (2003) Comparison of
hypoxia-induced replication arrest with hydroxyurea and aphidic-
olin-induced arrest. Mutat Res;532:205–213.
55. Kang M.H., Reynolds C.P. (2009) Bcl-2 inhibitors: targeting mito-
chondrial apoptotic pathways in cancer therapy. Clin Cancer
Res;15:1126–1132.
56. Lessene G., Czabotar P.E., Colman P.M. (2008) BCL-2 family
antagonists for cancer therapy. Nat Rev Drug Discov;7:989–
1000.
57. Yip K.W., Reed J.C. (2008) Bcl-2 family proteins and cancer.
Oncogene;27:6398–6406.
42. Zhang G.S., Liu D.S., Dai C.W., Li R.J. (2006) Antitumor effects
of celecoxib on K562 leukemia cells are mediated by cell-cycle
arrest, caspase-3 activation, and downregulation of Cox-2
expression and are synergistic with hydroxyurea or imatinib. Am
J Hematol;81:242–255.
43. Estrov Z., Shishodia S., Faderl S., Harris D., Van Q., Kantarjian
H.M., Talpaz M., Aggarwal B.B. (2003) Resveratrol blocks inter-
leukin-1beta-induced activation of the nuclear transcription fac-
58. Vogler M., Dinsdale D., Dyer M.J., Cohen G.M. (2009) Bcl-2
inhibitors: small molecules with a big impact on cancer therapy.
Cell Death Differ;16:360–367.
59. Azmi A.S., Mohammad R.M. (2009) Non-peptidic small molecule
inhibitors against Bcl-2 for cancer therapy.
iol;218:13–21.
J Cell Phys-
60. Del Gaizo Moore V., Letai A. (2008) Rational design of therapeu-
tics targeting the BCL-2 family: are some cancer cells primed for
Chem Biol Drug Des 2010; 76: 43–58
57

Palmeira et al.
death but waiting for a final push? Adv Exp Med Biol;615:159–
175.
Table S1. Scoring results (kJ ⁄ mol) for the pyranoxanthone 5 and
Bax, Bid, Bak and XIAP proteins obtained with eHits.
61. Chauhan D., Velankar M., Brahmandam M., Hideshima T., Podar
K., Richardson P., Schlossman R., Ghobrial I., Raje N., Munshi
N., Anderson K.C. (2007) A novel Bcl-2 ⁄ Bcl-X(L) ⁄ Bcl-w inhibitor
ABT-737 as therapy in multiple myeloma. Oncogene;26:2374–
2380.
62. Del Gaizo Moore V., Schlis K.D., Sallan S.E., Armstrong S.A., Le-
tai A. (2008) BCL-2 dependence and ABT-737 sensitivity in acute
lymphoblastic leukemia. Blood;111:2300–2309.
Table S2. Scoring results (kJ ⁄ mol) obtained with eHits for 50
'decoy' molecules randomly picked from NCI database.
Please note: Wiley-Blackwell is not responsible for the content or
functionality of any supporting materials supplied by the authors.
Any queries (other than missing material) should be directed to the
corresponding author for the article.
Supporting Information
Additional Supporting Information may be found in the online ver-
sion of this article:
58
Chem Biol Drug Des 2010; 76: 43–58