New thioderivatives of gossypol and gossypolone, as prodrugs of cytotoxic agents

Article abstract of DOI:10.1016/S0968-0896(03)00066-X

New dithiane or dithiolane derivatives of gossypol and gossypolone were synthesized with dithiolethane or dithiolpropane in the presence of BF₃/Et₂O. These thioderivatives exhibited low toxicity on KB cells (human epidermoid carcinom

Full text of DOI:10.1016/S0968-0896(03)00066-X

Bioorganic & Medicinal Chemistry 11 (2003) 2001–2006
New Thioderivatives of Gossypol and Gossypolone, as Prodrugs of
Cytotoxic Agents
Vi-Thuy Dao,a Michael K. Dowd,b Christiane Gaspard,a Marie-Therese Martin,a
Julie Hemez,a Olivier Laprevote,a Michel Mayera and Robert J. Michelota,*
aInstitut de Chimie des Substances Naturelles, CNRS, 91198 Gif sur Yvette, France
bUnited State Department of Agriculture, Southern Regional Research Center, New Orleans Louisiana 70179, USA
Received 15 November 2002; accepted 23 January 2003
Abstract—New dithiane or dithiolane derivatives of gossypol and gossypolone were synthesized with dithiolethane or dithiolpro-
pane in the presence of BF3/Et2O. These thioderivatives exhibited low toxicity on KB cells (human epidermoid carcinoma cells of
the mouth). They react easily with electrophiles in aprotic solvents to regenerate gossypolone or to form dehydrogossypoldithianes
and dehydrogossypoldithiolanes, which display higher toxicity on KB cells. In addition, the low toxicity of gossypol thioderivatives
was reversed by nitric oxide donors in physiological media. These experiments suggest that gossypol and gossypolone dithianes and
dithiolanes can be used as prodrugs that target tumor cells surrounded by high concentrations of nitric oxide.
# 2003 Elsevier Science Ltd. All rights reserved.
Introduction
compounds were found to have less toxicity and less
Gossypol is a natural product of the cotton plant that
has been studied as a potential male contraceptive. In
addition to its antifertility properties, gossypol exhibits
other interesting biological effects,1 and it may be useful
as an anti-neoplastic drug.2 However, the concentration
of gossypol needed to effect a specific drug response
relative to its overall toxicity may preclude its direct use
as a therapeutic agent. Consequently, there is interest in
synthesizing and testing derivatives and analogues of
gossypol for their activity and toxicity.
biological activity. Also important for maintaining
activity is the retention of the 6, 60, 7, and 70 hydroxyl
groups. These observations were recently confirmed
with gossypol, gossypolone, gossypol Schiff’s bases,
gossypolone Schiff’s bases, methylated gossypol, and
gossypolol (reduced gossypol).8
In this work, we report on the formation of dithiane and
dithiolane aldehyde derivatives of gossypol and gossy-
polone. We also determined the relative toxicity of these
compounds against KB cells (human epidermoid carci-
noma cells of the mouth). As expected, the chemistry was
readily reversed in the presence of electophiles,9 and we
tested the possibility that these masking groups could be
deprotected with NO in physiological media. The results
suggest that this class of analogues can be used as less-
toxic prodrugs that become activated near tumor cells
surrounded by high concentrations of NO.10ꢀ14
The activity of gossypol has been often attributed to the
presence of the aldehyde groups,3 which are capable of
reacting with nucleophiles. Of particular interest is gos-
sypol’s ability to bind to the e-amino groups of lysine in
peptides or proteins4 and the likelihood that these com-
plexes interfere with cellular function. Several deriva-
tives of the aldehyde group have been synthesized,
including periacylated gossylic nitriles,5 gossylic imino-
lactones,6 and a series of gossypol Schiff’s bases7 (Fig.
1). These derivatives were found to exhibit a varying
degree of activity relative to gossypol. In most cases, the
Results and Discussion
Chemistry
The potential for forming stable aldehyde derivatives of
gossypol is limited because the aldehydes groups inter-
*Corresponding author. Fax: +33-1-69-82-30-84; e-mail: michelot@
icsn.cnrs-gif.fr
0968-0896/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0968-0896(03)00066-X

2002
V.-T. Dao et al. / Bioorg. Med. Chem. 11 (2003) 2001–2006
Figure 1.
Figure 3.
act with the 1 and 10 hydroxyl groups to formlactol in
neutral or basic conditions. However, dithiane deriva-
tives of aldehydes (or ketones) are usually formed in
acidic conditions by the acid-catalyzed condensation of
thiols. These aldehyde modifications have not yet been
reported for gossypol. To formthese derivatives, gossy-
pol was treated with dithioethane and dithiopropane in
the presence of various acid catalysts, such as con-
centrated aqueous HCl, ZnI2, TiCl4, SiCl4, and
BF3/Et2O and the reaction was monitored by HPLC.
The bridging reaction appeared to proceed by a simple
displacement of the lactol ethers by the electron-rich
sulphur atoms (Fig. 4a). The length of an extended
unrestrained dithioethane molecule is just long enough
to cover the distance between the acetal carbon atoms if
the naphthalene rings are slightly distorted fromper-
pendicular. Consequently, this is the smallest dithio-
bridge possible. In comparison, gossypol binaphthalene
bridges have also been formed by condensation reac-
tions with diamines (Fig. 4b).15 These reactions proceed
via the aldehyde tautomer with aliphatic bridging chains
greater than or equal to four carbon atoms.15 The
shorter thio-based bridges are possible because of the
longer C–S bond lengths compared to the C–N and
C=N bond lengths and because the distance between
the acetal carbon atoms of the lactols is shorter than the
distance between the carbonyl carbon atoms of the
aldehydes.
The formation of the derivatives was highly dependent
on the catalyst. With concentrated aqueous HCl or ZnI2/
dioxane, the reaction did not proceed. With TiCl4 and
SiCl4 the reaction led to tarring, probably due to the
complexation with the gossypol phenolic hydroxyl
groups. Only BF3/Et2O proved to be an efficient catalyst
for these reactions. We obtained dithiolanes and
dithiane derivatives of gossypol (1, 2) and gossypolone
(3, 4) in yields >60% (Fig. 2).
Under typical acid or base conditions, dithiolanes and
dithianes are very stable, which has led to their synthetic
use as carbonyl protecting groups or as synthons. For
gossypol and gossypolone however, these products can-
not be used as intermediates without prior protection of
the phenolic hydroxyl groups, which are easily oxidized.
In an attempt to form protected thioderivatives, 7, 70, 8,
80-tetramethylgossypol and hexamethylgossypol were
treated with dithioethane and dithiopropane in the pre-
sence of BF3/Et2O. However, instead of forming the
expected dithiolanes or dithianes, we obtained 8,80
bridged thioacetals (5–7) in yields ranging from20 to
60% (Fig. 3).
Because it is well established that (ꢀ)-gossypol posseses
greater antitumor activity than (+)-gossypol,16,17 the
potential of these bridges to stabilize the gossypol
enantiomers is of particular interest. Research on these
thio-bridged gossypol derivatives is continuing
In the presence of strong electrophiles, such as Ag+,18
or NO+,19 aldehyde groups can be recovered from
dithianes. Reaction of 4 with catalytic amounts of
NO+.BFꢀ4 produced the expected gossypolone (Fig. 5a).
Reaction of 2 at the same conditions resulted in a new
compound, dehydrogossypoldithiane (8). The structure
of 8 was determined by HPLC-electron spray ionization
mass spectroscopy, which indicated the loss of four
hydrogen atoms (m/z=694). The 1H and 13C NMR
spectra were consistent with the coexistence of two iso-
mers 8a and 8b (Fig. 5b), 8a being the more abundant
form( ꢀc=o=187.8 ppm). The unexpected reactivity was
attributed to the presence of the 1 and 10 phenolic
hydroxyl groups in 2, which can interact intramolecu-
larly with the enones at the 8 and 80 positions. The same
product was formed when the dithiane derivative of
gossypol was exposed to NO in DMF with Fe3+
.
Biological studies
The toxicity of the synthesized derivatives (1–8) on KB
cells is reported in Table 1. The results confirmthat
masking the aldehyde groups decreases gossypol and
gossypolone toxicity (100-fold for 2 compared to gos-
sypol and 10-fold for 4 compared to gossypolone). The
methylated gossypol derivatives were also less toxic.
Figure 2.

V.-T. Dao et al. / Bioorg. Med. Chem. 11 (2003) 2001–2006
Table 1. Toxicity of the thioderivatives on KB cells
2003
Compound
IC50
Gossypol
1
2
Gossypolone
5.7 10ꢀ6M
4.5 10ꢀ4M
1.2 10ꢀ4M
2.4 10ꢀ6M
2.0 10ꢀ5M
3.5 10ꢀ5M
>25 10ꢀ6M
Atoxic
3
4
Tetramethoxy Gossypol
5
6
Hexamethoxy Gossypol
7
8
Atoxic
>25 10ꢀ6M
Atoxic
5.5 10ꢀ6M
Table 2. Toxicity of 2 in the presence of DPTA NONOate
Concentration
of 2
Without DPTA
NONOate
With DPTA NONOate
10ꢀ4 M
5 10ꢀ5
M
10ꢀ5
M
5 10ꢀ5 M
29%
0%
59%
41%
22%
11%
11%
0%
10ꢀ5
M
pounds and the chemistry resulting from the exposure
to nitric oxide essentially restored the toxicity.
Figure 4.
In biological media, nitronium ions can be formed from
NO binding to the heme of cytochromes. This may serve
as a potential source of nitroniumions to initiate the
formation of nitrosothiols.
Â
3þÃ
Â
þÃ
NO þ Fe3þ
!
NO-Fe
!
NO þ Fe2þ
NOþ þ R-Sꢀ ! RS-NO
The low toxicity of the gossypol thioderivatives promp-
ted us to examine their toxicity in physiological media in
the presence of a NO donor. The experiment was per-
formed by incubating 2 on a KB cell culture systemwith
dipropylenetriamine nonoate (DPTA NONOate). We
observed that the presence of DPTA NONOate (at
concentrations that were not toxic to KB cells) sig-
nificantly increased the toxicity of 2 (Table 2). The
result suggests that at physiological conditions NO
could release the sulphur moieties from dithiolane and
dithiane modified compounds to increase the toxicity of
the underlying drugs. More experimentation is needed
to confirmthe feasibility of this approach for use in
drug therapy.
Figure 5.
Cytotoxicities of gossypol thioderivatives in the presence
of NO
Conclusion
The dithianes and dithiolanes were found to react read-
ily with a nitroniumdonor in organic solvent. When the
reaction mixture was poured into water, gossypolone
dithiane was deprotected to give gossypolone, and gos-
sypol dithiane was rearranged to give 8, which has a
toxicity on KB cells similar to gossypol (Table 1). For
both classes of compounds, the toxicity of the thioder-
ivatives was significantly less than for the original com-
The toxicity of gossypol derivatives has often been
attributed to the presence of the aldehyde groups, which
can potentially bind or cross-link enzymes and nucleic
acids. To further study the influence of the aldehyde
moieties of gossypol on activity, several thioderivatives
of gossypol were prepared, characterized, and tested for
toxicity on KB human tumor cells. The results support

2004
V.-T. Dao et al. / Bioorg. Med. Chem. 11 (2003) 2001–2006
the general hypothesis of the role of the gossypol alde-
hydes on cell toxicity.
equiv of BF3-diethylether complex was added dropwise
to the cooled solutions. For 1 and 2, the solutions were
warmed to room temperature and allowed to stand for
up to 15 h resulting in precipitation of the product. The
precipitate was filtered off, washed with Et2O, and
vacuumdried. For 3 and 4, the progress of the reaction
was monitored by HPLC. Upon completion, the solu-
tions were washed successively with a 5% solution of
NaHCO3, water, and brine, dried over Na2SO4, and the
solvent evaporated. The products were precipitated
fromEt 2O/hexane as green solids. Compounds 5–7 were
obtained as white solids by following the procedure for
1 and 2.
In addition, because NO is generated in high concen-
tration by activated macrophages as part of the immune
response against bacteria, viruses, and tumor cells, it is
particularly abundent near solid tumors, which are
highly vascularized, and it plays a predominant role in
angiogenesis and vascular permeability. Thus, high local
concentrations of NO could be used as a deprotection
reagent (i.e., a chemical trigger) to activate the toxicity
of therapeutic agents. The results suggest a new
approach to targeted drug therapy.
8,80-Bis-[1,3]dithiolan-2-yl-5,50-diisopropyl-3,30-dimethyl-
[2,20]binaphthalenyl-1,6,7,10,60,70-hexaol (1). Yield 67%,
yellow solid, HPLC: k0(II): 9,17; ESI-MS: m/z 669
Experimental
1
Reagents and solvents were purchased fromFluka Che-
mie AG. Gossypol (as gossypol-acetic acid), gossypol
methyl ethers, and gossypolone were obtained as pre-
viously described.8 DPTA NONOate was fromCayman
Chemical Co., Boron trifluoride diethyl ether complex
99% was fromLancaster Co. Mass spectra were deter-
mined at the Service de Spectrometrie de Masse de l’In-
stitut de Chimie des Substances Naturelles on AEI
MS50 or Navigator-Thermoquest spectrometers. NMR
spectra were recorded in CDCl3 or DMSO-d6 on Bruker
AM300 or AM400 spectrometers. The structures of 1–8
were checked by HMQC and HMBC 2-D NMR.
(M–Hꢀ); H NMR (400 MHz, DMSO) d 1.54 (d, 12H,
2HC-(CH3)2), 2.00 (s, 6H, 2 Ar-CH3), 3.31, 3.65 (m,
8H, 2 C–S–CH2–CH2–S), 3.94 (m, 2H, 2 HC–(CH3)2),
7.51 (s, 2H, 2 ArꢀH), 8.13 (s, 2H, 2 S–HCꢀS), 7.96,
8.29, 8.63 (s, br, 6H, 6 OH at 1,10,6,60,7,70 positions);
13C NMR (100 MHz, DMSO) d 20.4 (Ar–CH3), 20.6
(HC–(CH3)2), 26.2 (HC–(CH3)2), 39.7 (C–S–CH2–CH2–
S), 50.6 (S–HC–S); 111.7 (C-8), 115.4 (C-4), 117.4 and
117.6 (C-2 and C-9), 125.7 (C-5), 129.2 (C-10), 132.3
(C-3), 143.9 (C-6), 146.0 (C-7), 150.7 (C-1).
8,80 -Bis-[1,3]dithian-2-yl-5,50 -diisopropyl-3,30 dimethyl-
[2,20]binaphthalenyl-1,6,7,10,60,70-hexaol (2). Yield 93%.
HPLC: k0(II): 10,25; ESI-MS: m/z 721 (M+Na); 1H
NMR (400 MHz, DMSO) d 1.47 (d, 12H, 2HC–
(CH3)2), 1.96 (s, 6H, 2 Ar–CH3), 1.74, 2.15 (m, 4H, 2 C–
S–CH2–CH2–CH2–S), 2.93 (m, 8H, 2 C-S–CH2–CH2–
CH2–S), 3.87 (m, 2H, 2 HC–(CH3)2), 7.55 (s, 2H, 2 Ar-
H), 7.83 (s, 2H, 2 S–HC–S), 7.77, 8.25, 8.63 (s, br, 6H, 6
OH at 1,10,6,60, 7,70 positions); 13C NMR (100 MHz,
DMSO) d 20.4 (Ar–CH3), 20.5 (HC–(CH3)2), 26.2 (HC–
(CH3)2), 25.0 (C–S–CH2–CH2–CH2-S), 31.6 (C–S–
CH2–CH2–CH2-S), 45.7 (S–HC–S), 111.4 (C-8), 115.6,
115.7 (C-4 and C-9), 117.9 (C-2), 125.6 (C-5), 129.4 (C-
10), 132.4 (C-3), 143.7 (C-6), 146.7 (C-7), 150.7 (C-1).
Analytical-scale HPLC work was performed with a
SpectraSYSTEM P1000XR pumping system and a
reverse-phase Alltima C-18 column (4.6 mm i.d.ꢁ250
mm) with isocratic elution at a flow rate of 1 mL/min.
Two mobile phases were used: (I) 10% A and 90% B
and (II) 20% A and 80% B, where A was 95:5 (v/v)
water/CH3CN acidified with TFA to a final concen-
tration of 0.1% and B was CH3CN also acidified with
TFA to a final concentration of 0.1%. Compounds were
detected with a Waters Corp. Model 996 photo-diode
array detector operated from254 to 400 nm. Pre-
parative work was performed with a Waters Corp. 515
HPLC pumping unit, reverse-phase PrepPak C-18 car-
8,80-bis-[1,3]dithiolan-2-yl-6,7,60,70-tetrahydroxy-5,50-dii-
sopropyl-3,30-dimethyl-[2,20]binaphthalenyl-1,4,10,40-tetra-
one (3). Yield 65%. HPLC: k0(II): 4,70; ESI-MS: m/z
˚
tridge column (10 mmparticles with 125 A pores; 25
mm i.d.ꢁ100 mm) with isocratic elution and a flow rate
of 5 mL/min. The mobile phase was 20:80 (v/v)
CH3CN/water. A Waters Corp. Model 2487 dual
wavelength UV detector operated at 254 nmwas used
for detection.
1
697 (M–Hꢀ). H NMR (300 MHz, CDCl3) d 1.43 (dd,
12H, 2HC–(CH3)2), 1.99 (s, 6H, 2 Ar–CH3), 3.40, 3.58
(m, 8H, 2 C–S–CH2–CH2–S), 4.03 (m, 2H, 2 HC–
(CH3)2), 7.15 (s, 2H, 2 S–HC–S), 6.63, 8.89 (s, br, 4H, 4
OH at 6,60,7,70 positions); 13C NMR (75MHz, CDCl3) d
14.4 (Ar–CH3), 19.4, 19,6 (HC–(CH3)2), 28.2 (HC–
(CH3)2), 39.8, 40.0 (C–S–CH2–CH2–S), 50.1 (S–HC–S),
119.9 (C-8), 125.5 (C-9), 128.8 (C-10), 136.8 (C-5), 139.4
(C-2), 145.8 (C-3), 147.2 (C-7), 149.7 (C-6), 185.4 (C-1),
187.6 (C-4).
General procedure for the preparation of 1,3 dithiane or
1,2-dithiolane dervivatives of gossypol and gossypolone
and bridged thioacetal of gossypol methyl ethers 1–7. A
0.1–1.0 M solution of either gossypol, gossypolone, or a
methylether derivative of gossypol in chloroform was
combined with an equimolar amount of either 1,2-
dithiolethane or 1,3-dithiolpropane at roomtempera-
ture. For reaction with gossypol or gossypolone, the
solution was kept at roomtemperature for 1 h before
cooling to ꢀ20 ꢂC. For reactions with the methyl ether
gossypol derivatives, the solutions were cooled on an ice
bath immediately after mixing the components. A 0.1
8,80-bis-[1,3]dithian-2-yl-6,7,60,70-tetrahydroxy-5,50-diiso-
propyl-3,30 -dimethyl-[2,20]binaphthalenyl-1,4,10,40 -tetra-
one (4). Yield 89%. HPLC: k0(II): 5,97; ESI MS: m/z
1
749 (M+Na); H NMR (400 MHz, CDCl3) d 1.43 (dd,
12H, 2HC–(CH3)2), 1.99 (s, 6H, 2 Ar–CH3), 1.87, 2.15
(m, 4H, 2 C–S–CH2–CH2–CH2–S), 2.85, 3.10 (m, 8H, 2

V.-T. Dao et al. / Bioorg. Med. Chem. 11 (2003) 2001–2006
2005
C–S–CH2–CH2–CH2–S), 4.02 (m, 2H, 2 HC–(CH3)2),
7.23 (s, 2H, 2 S–HC–S), 6.57, 8.18 (s, 4H, 4 OH at
6,60,7,70 positions); 13C NMR (75 MHz, CDCl3) d 14.5
(Ar–CH3), 19.9, 20.0 (HC–(CH3)2), 28.2 (HC–(CH3)2),
24.8 (S–CH2–CH2–CH2–S), 31.2 (S–CH2–CH2–CH2–S),
43.1 (S–HC–S), 121.5 (C-8), 123.4 (C-9), 128.8 (C-10),
136.8 (C-5), 139.4 (C-2), 145.9 (C-3), 147.0 (C-7), 149.5
(C-6), 185.3 (C-1), 187.8 (C-4).
temperature, the reaction, which was monitored by
HPLC, gave compound 8 in a 68% yield. The product
was recovered by preparative HPLC. The correspond-
ing reaction with 4 gave gossypolone in a 40% yield
after standing 4 days.
80-Dimercaptomethylene-8-[1,3]dithian-2-ylidene-1,6,10,60-
tetrahydroxy - 5,50 - diisopropyl - 3,30 - dimethyl - 8H,80H -
[2,20]binaphthalenyl-7,70-dione (8a) (see Fig. 5). Yield
14%. HPLC: k0(II): 3,19; ESI-MS: m/z 694 (M); 1H
NMR (400 MHz, CDCl3) d 1.53 (d, 12H, 2HC–(CH3)2),
1.99 (s, 6H, 2 Ar–CH3); 13C NMR (100 MHz, CDCl3) d
20.3–20.7 (Ar–CH3), 21.3–21.4 (HC–(CH3)2), 28.1–28.2
(HC–(CH3)2), 27.0 (C–S–CH2–CH2–CH2–S), 29.6–30.0
(C–S–CH2–CH2–CH2–S), 41.9, 65.1 (S–HC–S), 107.9–
149.9, 173.0 and 187.8 (naphthalene nucleus).
5,50-Diisopropyl-2,20-dithioethane-4,40-dimethoxy-7,70-di-
methyl-2H,20H-[8,80] bi[naphtho [1,8- bc]furanyl]-3,30-
diol (5). Yield 24.2%. HPLC: k0(I): 3.82; EI–MS: m/z
605 (MH+); 1H NMR (300 MHz, CDCl3) d 1.41 (d,
12H, 2HC–(CH3)2), 2.30 (s, 6H, 2 Ar–CH3); 2.61, 2.95
(d, 4H, C–S–CH2–CH2–S), 3.66 (m, 2H, 2 HC–(CH3)2),
4.10 (s, 6H, 2 OCH3 at 7,70 positions) 6.18 (s, 2H, 2 OH
at 6,60 positions), 6.98 (s, 2H, 2 O–HC–S), 7.30 (s, 2H, 2
Ar–H); 13C NMR (75 MHz, CDCl3) d 20.7 (HC–
(CH3)2, Ar–CH3), 26.9 (HC-(CH3)2), 30.7 (C–S–CH2–
CH2–S), 59.3 (OCH3 at 7,70 positions), 90.2 (O-HC–S),
109.4 (C-2), 114.2 (C-4), 118.2 (C-8), 121.2 (C-9), 125.8,
126.1 (C-5, C-10), 137.2 (C-3), 139.8 (C-7), 146.4 (C-6),
156.5 (C-1).
Reaction of 1,3 dithiane gossypol 2 with NO in DMF. A
saturated solution of NO in DMF was prepared by
bubbling NO gas in DMF previously degassed with
argon. The NO concentration was determined by add-
ing 2,20-azino-bis(2-ethylbenzthiazoline-6-sulfonic acid)-
(ABTS) with detection by UV–vis spectroscopy at 660
and 750 nm, corresponding to the absorption of a stable
cation radical ABTSꢃ+.20 One hundred microliters of a
NO/DMF solution (0.53 mmol/L) was poured into 500
mL of DMF containing 10 mg of 2 at roomtemperature.
The reaction mixture was monitored by HPLC and, after
20 min, 2 displayed no transformation. The same
experiment performed in the presence of 50 mL of 1%
(w/w) aqueous FeCl3 led to the immediate formation of 8.
5,50-Diisopropyl-2,20-dithiopropane-4,40-dimethoxy-7,70-
dimethyl-2H,20H-[8,80] bi[naphtho [1,8-bc]furanyl]-3,30-
diol (6). Yield 60.3%. HPLC: k0(I): 4,09; ESI-MS: 619
1
(MH+); H NMR (300 MHz, DMSO) d 1.43 (d, 12H,
2HC–(CH3)2), 2.14 (s, 6H, 2 Ar–CH3), 1.67 (m, 4H, S–
CH2–CH2–CH2–S), 2.98, 3.02 (m, 4H, S–CH2–CH2–
CH2–S), 3.72 (m, 2H, 2 HC–(CH3)2), 4.09 (s, 6H, 2
OCH3 at 7,70 positions), 7.31 (s, 2H, 2 Ar–H), 7.39 (s,
2H, 2 O–HC–S), 8,60 (s, 2H, 2 OH at 6,60 positions);
13C NMR (75 MHz, CDCl3) d 20.3 (Ar–CH3), 20.5HC–
(CH3)2, 25.9 (HC–(CH3)2), 27.0 (S–CH2–CH2–CH2–S),
27.6 (S–CH2–CH2–CH2-S), 58.2 (OCH3 at 7,70 posi-
tions), 88.6 (O-HC-S), 108.9 (C-2), 113.4 (C-4), 118.1
(C-8), 120.6 (C-9), 125.1, 125.2 (C-5, C-10), 136.2 (C-3),
140.5 (C-7), 147.1 (C-6), 155.1 (C-1).
Biological studies
KB cells were originally obtained fromthe American
Type Culture Collection. The cell line was grown in
minimal essential media with Earle’s salt solution (pur-
chased fromSeromed) containing 10% fetal calf serum,
2 m Ml-glutamine, 60 mg/mL penicillin G, 60 mg/mL
streptomycin sulfate, and 40 mg/mL gentamycin. Cells
were grown as monolayers in Nunc 24-well plastic plates
(ꢄ25,000 cells were seeded per well in 1 mL of media).
5,50 - Diisopropyl - 2,20 - dithiopropane - 3,30,4,40 - tetrame-
thoxy-7,70-dimethyl-2H,20H-[8,80] bi[naphtho [1,8- bc]fur-
anyl] (7). Yield 27.3%. HPLC: k0(I): 10.71; ESI-MS: m/
1
z, 647 (MH+); H NMR (400 MHz, CDCl3) d 1.51 (d,
For toxicity assays, the compounds of interest were dis-
solved in DMSO and were serially diluted with the same
solvent. Immediately after plating the cells, the test
compounds were added to the cultures (<10 mL) such
that the final concentration of DMSO was <1%. Con-
trol cultures received an equal dilution with DMSO.
The cultures were incubated at 37 ꢂC in a 95:5 (v/v) air/
CO2 humidified incubator. After incubating for 3 days,
cell viability was determined by adding 100 mL of a
0.02% solution of neutral-red vital dye in the growth
media to each well and incubating the plate for an
additional 8-16 h. The wells were then washed with
phosphate buffered saline, and the cells were lysed with
a 1% solution of sodiumlauryl-dodecyl sulfate. Extrac-
ted dye was measured at 540 nm with a Uniskam-II
microplate reader (Labosystems, Life Science Interna-
tional).
12H, 2HC–(CH3)2), 2.24 (s, 6H, 2 Ar–CH3), 1.76 (m,
2H, C–S–CH2–CH2–CH2–S), 3.04, 3.34 (m, 4H, C–S–
CH2–CH2–CH2–S), 3.82 (m, 2H, 2 HC–(CH3)2), 3.87 (s,
6H, 2 OCH3 at 6,60positions), 4.13 (s, 6H, 2 OCH3 at
7,70 positions), 7.12 (s, 2H, 2 O–HC–S), 7.44 (s, 2H, 2
Ar–H); 13C NMR (75 MHz, CDCl3) d 20.9 (Ar–CH3),
22.0 (HC–(CH3)2), 27.1 (HC–(CH3)2, S-CH2–CH2–
CH2–S), 27.5 (S–CH2–CH2–CH2–S), 61.4 (OCH3 at 6,60
positions), 59.2 (OCH3 at 7,70 positions), 89.5 (O–HC–
S), 111.0 (C-2), 115.7 (C-4), 121.4 (C-8), 124.0 (C-9),
125.9 (C-10), 136.0 (C-5), 137.1 (C-3), 145.2 (C-7), 151.0
(C-6), 155.5 (C-1).
Reaction of 1,3 dithiane gossypol (2) and 1,3 dithiane
gossypolone (4) with nitrosonium tetrafluoroborate
(NO+, BFꢀ4 ). One hundred and fifty microliters of a 50-
mg/mL solution of NO+BF4ꢀ in CH2Cl2 and 50 mL of
DMF were added to 15 mL of a DMF solution con-
taining 150 mg of 2. After standing 24 h at ambient
Most of the compounds were tested at concentrations
between 0.5 and 10 mM. Each concentration was added

2006
V.-T. Dao et al. / Bioorg. Med. Chem. 11 (2003) 2001–2006
to duplicate wells on the same plate. The percentage of
cell death was plotted against the compound concen-
tration, and IC50 values were determined from the plots.
8. Dao, V. T.; Gaspard, C.; Mayer, M.; Werner, G. H.;
Nguyen, N. S.; Michelot, R. J. Eur. J. Med. Chem. 2000, 35,
805.
9. Meyers, A. I. Organic Synthesis; John Wiley Ed: New
York, 1974.
To study the effect of DPTA NONOate on the toxicity
of 2, KB cells were first grown as given above except
that the incubation period was reduced to 24 h. Com-
pound 2 and DPTA NONOate were added simulta-
neously to the cultures. Control samples were also run
in the same assays with no compound addition, with
only 2 added, and with only DPTA NONOate added.
After incubating the cells at 37 ꢂC for 48 h, cell destruc-
tion was determined by the neutral-red viability test
described above.
10. Thomsen, L. L.; Miles, D. W.; Happerfield, L.; Bobrow,
L. G.; Knowles, R. G.; Moncada, S. Br. J. Cancer 1995, 72,
41.
11. Thomsen, L. L.; Lawton, F. G.; Knowles, R. G.; Beesley,
J. E.; Riveros-Moreno, V.; Moncada, S. Cancer Res. 1994, 84,
1352.
12. Cobbs, C. S.; Brenman, J. E.; Aldape, K. D.; Bredt, D. S.;
Israel, M. A. Cancer Res. 1995, 55, 727.
13. Jenkins, D. C.; Charles, I. G.; Thomsen, L. L.; Moss,
D. W.; Holmes, L. S.; Baylis, S. A.; Rhodes, P.; Westmore, K.;
Emson, P. C.; Moncada, S. Proc. Natl. Acad. Sci. U.S.A 1995,
92, 4392.
14. Prazma, J.; Petrusz, P.; Mims, W.; Ball, S. S.; Weissler,
M. C. Otolaryngol. Head Neck Surg 1995, 113, 541.
15. Kunesch, N.; Barbier, M. P.; Dalkafouki, A.; Poisson,
J. E. Ann. Pharmaceutiques. Francaises 1990, 48, 1.
16. Shelley, M. D.; Hartley, L.; Fish, R. G.; Groundwater, P.;
Morgan, J. J. G.; Mort, D.; Mason, M.; Evans, A. Cancer
Lett.ers 1999, 135, 171.
References and Notes
1. Vander Jagt, D. L.; Deck, L. M.; Royer, R. E. Cur. Med.
Chem. 2000, 7, 479.
2. Reidenberg, M. M. Reproductive medicine: A Millennium
review 1999, 305.
3. Wichmann, K.; Krusius, L.; Sinervirta, R.; Puranen, J.;
Jaene, J. Contraception 1986, 33, 519.
4. Adams, R. C.; Geissman, T. A.; Edwards, E. Chem. Rev.
1960, 60, 555.
5. Royer, R. E.; Deck, L. M.; Campos, N. M.; Hunsaker,
L. A.; Vander Jagt, D. L. J. Med. Chem. 1986, 29, 1799.
6. Deck, L. M.; Vander Jagt, D. L.; Royer, R. E. J. Med.
Chem. 1991, 34, 3301.
7. Liang, X. S.; Rogers, A. J.; Webber, C. L.; Ormsby, T. J.;
Tiritan, M. E.; Matlin, S. A.; Benz, C. C. Investigational New
Drugs 1995, 13, 181.
17. Liu, S.; Kulp, S. K.; Sugimoto, Y.; Jiang, J.; Chang, H.-
L.; Dowd, M. K.; Wan, P.; Lin, Y. C. Anticancer Res. 2002,
22, 33.
18. Satchell, D. P. N.; Satchell, R. S. Chem. Soc. Rev. 1990,
19, 55.
19. York, C.; Surya Prakash, G. K.; Olah, G. A. Tetrahedron
1996, 52, 9.
20. Nims, R.W.; Cook, J.C.; Krishna, M.C.; Christodoulou,
D.; Poore, C.M.B.; Miles, A.M.; Grisham, M.P.; Wink, D.A.
In Methods in Enzymology, Lester Packer Ed. Acad. Press,
1996, 268, Part A, 93