Novel disulfides with antitumour efficacy and specificity

Article abstract of DOI:10.1071/CH03105

Some disulfides have previously been shown to possess antifungal and/or antileukaemic activity. Importantly, this cytotoxicity can be selective. We have previously shown that a subset of these compounds does not block the proliferative potential of normal

Full text of DOI:10.1071/CH03105

Full Paper
CSIRO PUBLISHING
Aust. J. Chem. 2005, 58, 128–136
www.publish.csiro.au/journals/ajc
Novel Disulfides with Antitumour Efficacy and Specificity
Rebecca Griffiths,A,D W. Wei-Lynn Wong,A,B,D Stephen P. Fletcher,C Linda Z. Penn,A,B,E
and Richard F. LanglerC,F
A Ontario Cancer Institute, University Health Network, 610 University Avenue, Toronto M5G 2M9, Canada.
B Department of Medical Biophysics, University of Toronto, Toronto M5G 2M9, Canada.
C Department of Chemistry, Mount Allison University, Sackville, New Brunswick E4L 1G8, Canada.
D These authors contributed equally to the work.
E Corresponding author (biological testing). Email: lpenn@oci.utoronto.ca
F Corresponding author (organic chemistry). Email: rlangler@mta.ca
Some disulfides have previously been shown to possess antifungal and/or antileukaemic activity. Importantly,
this cytotoxicity can be selective. We have previously shown that a subset of these compounds does not block the
proliferative potential of normal, non-transformed cells. Based on these results and proposed mechanisms of action,
a new set of structurally modified organosulfur compounds, including α-substituted disulfides and a thiosulfonate
ester, have been prepared and evaluated for their potential as antileukaemic agents. Compounds were screened for
antiproliferative activity against a panel of human cells derived from acute lymphocytic and acute myelogenous
leukaemia, as well as non-transformed cells. We have identified five new disulfides and a thiosulfonate that can
trigger tumour cells to undergo cell death by an apoptotic mechanism in a sensitive and specific manner.
Manuscript received: 22 April 2003.
Final version: 16 June 2004.
CH3SO2CH2SCH2SSCH2SO2CH3
Introduction
1
Triggering the apoptotic pathway is a desirable cell-death
pathway to target with anticancer agents.[1,2] Apoptosis is a
highly regulated molecular mechanism that is characterized
by cell shrinkage, DNA fragmentation, membrane blebbing,
and separation of the cell into apoptotic bodies.[3] Unlike
other forms of cell death, such as necrosis, apoptosis does
not trigger an inflammatory response or result in damage to
the surrounding tissue. Although anticancer agents currently
in use may induce apoptosis, both malignant and healthy
non-transformed cells are affected. Consequently, the devel-
opment of novel therapeutic agents with greater specificity
for malignant cells is critical. To this end, we have prepared
low molecular weight compounds and screened them for their
ability to specifically induce transformed cells to undergo
apoptosis.
Scheme 1.
RSO2CH2SSRꢀ
2
Scheme 2.
CH3SSCH2CO2CH3
CH3SO2CH2CH2SSCH3
3
4
Scheme 3.
In pursuit of such novel apoptotic agents, naturally-
occurring organosulfur compounds (OSCs) have been of
great interest.[4–12] We began structure–activity work by
modifying the structure of dysoxysulfone 1 (Scheme 1),
which was originally isolated from the Fijian medicinal plant
Dysoxylum richii and shown to possess both antifungal and
anticancer activity.[13,14]
In an effort to identify compounds with superior speci-
ficity and potency, we previously synthesized seven struc-
tural relatives of dysoxysulfone and screened them for
antifungal and antitumour activity. Initial antileukaemic test-
ing showed a correlation between disulfide functionality
and anticancer activity. In particular, α-sulfone disulfides 2
(Scheme 2) induced apoptosis in both non-transformed and
leukaemia cells.
More interestingly, structure–activity analysis revealed
disulfides 3 and 4 (Scheme 3) as promising antileukaemic
agents that were not cytotoxic to non-transformed cells.[15]
Two modes of behaviour were proposed to rationalize the
biological activity displayed (see Schemes 4 and 5). For the
© CSIRO 2005
10.1071/CH03105
0004-9425/05/020128

Novel Disulfides with Antitumour Efficacy and Specificity
129
present study, we exploited both Schemes 4 and 5 and the
results from our previous antileukaemic study[15] to design a
second-generation set of potentially selective antileukaemic
disulfides. Six disulfides and a thiosulfonate were prepared
and screened for their ability to trigger apoptosis in a tumour-
specific manner (see Scheme 6). Six of these compounds
induced apoptosis with specificity for leukaemic cells.
rings serve to optimize the balance between hydrophilicity
and hydrophobicity, which in turn enhances pharmacological
activity.[18]
In light of the results from our previous study,[15] we chose
to examine a pair of structures, namely 6 and 8, in which the
S S bond was less aggressively activated for nucleophilic
attack. In particular, S R in 8 is less nucleofugal than SO2R
in 2, while the thiosulfonate 6 presents a more sterically
congested sulfenyl sulfur atom than that of 2. In this way,
we hoped to maintain the antileukaemic properties of the
disulfides while significantly diminishing their toxicity to
normal cells.
Results and Discussion
Structural Design
In the current study, we evaluated the apoptotic activity of
seven OSCs (see Scheme 6). Scheme 6 includes a represen-
tative example of each of our newly established, biologically
active disulfide types, namely an α-sulfone disulfide 5, an
α-ester disulfide 7 (ester group is attached to the disulfide
framework by a C O bond), and an aryl methyl disulfide 11.
Each of these disulfides exploits a specifically different man-
ner in which it activates the disulfide linkage toward nucleo-
philic attack (see Schemes 4 and 5[16,17]). The α-sulfone
disulfide 5 features a relatively long, unbranched alkyl sub-
stituent. Unbranched C5–C9 alkyl substituents or phenyl
Our initial study[15] established that ester disulfide 3
(Scheme 3) showed significant antileukaemic activity with-
out any substantial accompanying activity against normal
cells. Hence, the α,αꢀ-diester disulfide 9 was targeted for
study. Furthermore, as Scheme 5 rationalizes the antitu-
mour activity of 3, and ketones are better carbon acids than
esters, the α-ketodisulfide 10 was prepared to test for specific
activity as an antileukaemic agent in vitro.
Growth Inhibitory Effect of OSCs
Toassessthepotentialtumour-specificantiproliferativeactiv-
ity of the OSCs (six disulfides and a thiosulfonate, Scheme
6), a panel of leukaemic cell lines was screened to assess
each line’s sensitivity to each sulfur compound. Two cell
lines derived from acute lymphocytic leukaemia (ALL), KK
and B1, acute myelogenous leukaemia (AML), OCI-AML-3
(hereafter referred to asAML-3), and NB-4, as well as a non-
transformed human diploid fibroblast cell line WI38 were
exposed to increasing concentrations of OSCs and assayed
for proliferation using the 3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide (MTT) assay. The MTT assay
measures the ability of cellular mitochondrial dehydrogenase
to convert the yellow MTT substrate into a purple formazan
product. The resultant light absorbency values were normal-
ized to untreated controls. Exposure to each OSC resulted in
a dose-dependent decrease in proliferation in all leukaemic
cell lines (Table 1). In contrast, the ability of each OSC to
decrease proliferation of the normal diploid cell line was less
robust, with the exception of compound 11. The MTT50 val-
ues (the dose required to reduce proliferation to 50%) suggest
that compounds 5, 6, 7, 8, 9, and 10 possess antiprolifer-
ative activity that is specific for leukaemic cell lines, while
compound 11 is a general cytotoxic agent (seeTable 1). Com-
pounds 9 and 10 appear to be more specific for three of the
four leukaemic cell lines (AML-3, NB-4, B1). Furthermore,
compounds 5, 6, and 7 show greater potency in comparison
X
CH2
S
S
ꢁ
ꢁ
CH2S
X
R
S
Nu
Nu
Rꢀ
X ꢂ O2CCH3 or O2SCH3
Scheme 4. Proposed chemical rationale for biological activity of
disulfides in which the α-carbon bears a leaving group.
S
Rꢀ
H
Y
C
H
S
YCH
S
S
Rꢀ
ꢁ
(a) Nu
(b) B
H
B
YCH2SNu ꢁ B
Y ꢂ PhC(O) or CO2CH3
Scheme 5. Proposed rationale for biological activity of disulfides in
which α-protons are acidified but the α-carbon does not bear a leaving
group.
CH3SO2CH2CH2CH2CH2CH3
CH3SO2SPh
5
6
PhSSCH2OC(O)CH2CH3
CH3SCH2SSCH3
CH3OC(O)CH2SSCH2C(O)OCH3
7
8
9
PhC(O)CH2SSCH3
p-O2N(C6H4)SSCH3
10
11
Scheme 6.

130
R. Griffiths et al.
to compounds 8, 9, and 10; this demonstrates that the choice
of α-substituent does affect disulfide activity.
cycle, or undergo apoptosis.[19] In addition, fixed PI permits
measurement of cells that are undergoing apoptosis follow-
ing exposure to a compound, by quantifying the percentage
of cells in the pre-G1 region of the profile.[20]
OSC-Induced Changes in Cell-Cycle Profile
and Apoptosis
Treatment of leukaemic cell lines with the seven com-
pounds (5–11) resulted in an increase in cells with a DNA
content less than G1 (Table 2). Treatment of the WI38 non-
transformed fibroblast cells with six of the compounds (5–10)
resulted in minimal pre-G1 regions of 1–5% (Fig. 1a, data
not shown). However, compound 11, previously noted in
the MTT assay as a general cytotoxin, caused a consid-
erable percentage of the non-transformed cells to undergo
apoptosis. Interestingly, exposure of the WI38 cells to OSCs
5–10 resulted in a change in the cell-cycle profile, with an
accumulation of cells in the G2 phase. This suggests that
OSCs can trigger a G2 cell-cycle arrest in non-transformed
cells (Fig. 1a, data not shown). Clearly, five of the OSCs
(5–7, 9, 10) are considerably more cytotoxic to ALL and
AML cell lines than non-transformed fibroblast cells; this
suggests that these OSCs possess tumour-specific activity
(Fig. 1b). Interestingly, a robust accumulation of cells in the
pre-G1 region was not as evident with compound 8, yet the
To determine whether the decrease in proliferation was the
result of apoptosis, and to further evaluate tumour specificity,
leukaemic and non-transformed fibroblast cells were exposed
to OSCs at two different concentrations for 48 h. The cellu-
lar DNA content was then measured by staining the cells
with propidium iodide (PI), a DNA intercalating dye. The
amount of DNA within the cell was then analyzed by flow
cytometry. By this approach (referred to as fixed PI), the
percentage of cells in a given phase of the cell cycle can
be evaluated. Briefly, the cell cycle can be divided into four
sequential phases: G1, the cell prepares to duplicate DNA; S,
DNA is duplicated; G2, the cell prepares to divide; and M, cell
division occurs. Cell cycle checkpoints, which exist before
replication of the DNA, during the G1 phase, and before
separation of the chromosomes during the G2 phase govern
whether the cell will progress through the cycle, arrest in cell
Table 1. The dose (µM) required to reduce the MTT activity to 50% (MTT50) of the
non-transformed diploid cell line (WI38), the AML cell lines (AML-3, NB-4), and the ALL
cell lines (KK, B1)
The MTT50 value was determined by exposing the cells to a wide concentration range (0–200 µM)
of OSCs for 48 h and conducting MTT assays. Results shown are the mean MTT50 for three
independent experiments performed with three replicates in each experiment
Compound
tested
Cell line
NB-4
WI38
AML-3
B1
KK
5
6
7
8
9
10
11
59.6 11.2
57.4 18.3
65.0 9.3
>200
155.9 27.4
144.7 26.7
10.1 1.9
11.3 2.8
13.6 5.8
16.1 3.7
97.7 20.0
62.2 7.6
71.9 19.4
13.9 3.5
12.3 1.9
25.3 5.1
16.8 5.5
71.6 10.6
76.8 17.4
38.9 9.5
14.5 3.1
10.3 2.1
16.4 10.1
16.6 4.6
97.7 35.2
77.8 10.2
99.9 36.1
17.0 5.9
13.3 3.5
12.8 6.5
24.2 5.1
118.6 17.4
130.9 29.8
155.4 23.5
10.0 2.2
Table 2. Percentage of apoptotic (pre-G1 as determined by fixed PI) leukaemic and fibroblast
cells after 48 h exposure to OSCs (mean standard error of mean)
Compound
tested
Dose
Cell line
NB-4
WI38
AML-3
B1
KK
11
5
25 µM
50 µM
25 µM
50 µM
25 µM
50 µM
25 µM
50 µM
50 µM
100 µM
50 µM
100 µM
50 µM
100 µM
30.8 7.1 47.0 3.9
33.4 3.1 35.8 7.7
0.5 0.8 21.2 0.2
1.3 0.2 53.6 17.1 58.3 5.2 37.4 4.3
0.5 0.1 11.6 3.3 15.2 1.2 7.5 1.6
1.3 0.3 45.6 10.2 36.5 3.3 27.2 2.2
0.4 0.2 14.4 7.5
1.7 0.3 29.0 7.9
1.6 0.5 10.1 2.8
5.3 1.5 10.2 2.6
47.0 7.1 53.5 20.2 66.0 13.9
62.7 10.3 39.9 10.3 45.4 10.0
32.4 5.5 22.9 5.5
10.4 2.3
28.7 7.5
19.9 8.6
25.1 5.2
36.6 12.4
39.2 5.8
7.9 1.3
6
7
25.2 8.0 23.3 5.3
64.6 5.8 49.6 5.9
27.2 3.0 14.5 4.0
8
34.0 1.9
10.0 1.3
8.7 1.8
8.8 6.7
6.9 0.6
36.9 15.4
9
0.7 0.3
2.6 0.7
1.9 0.7 22.4 10.6 26.8 3.2 26.9 11.3 37.8 17.6
10
1.2 0.2
9.9 0.7
23.8 1.3
6.8 0.6
27.8 19.7
38.8 14.6
2.2 0.4 47.3 18.1 40.2 2.1 33.1 4.3

Novel Disulfides with Antitumour Efficacy and Specificity
131
(a)
WI38
AML-3
B1
350
280
210
140
70
Acetone 150
350
280
210
140
70
1.2%
0.5%
0.6%
120
90
60
30
0
0
0
0
0
0
0
0
200 400 600 800 1000
1.4%
0
0
0
0
0
200 400 600 800 1000
27.8%
0
200 400 600 800 1000
30.2%
5
70
60
50
40
30
20
10
0
120
100
80
60
40
20
0
120
100
80
60
40
20
0
50 ꢃM
200 400 600 800 1000
3.4%
200 400 600 800 1000
17.7%
0
0
0
0
200 400 600 800 1000
8.9%
8
60
50
40
30
20
10
0
50
40
30
20
10
0
120
100
80
60
40
20
0
100 ꢃM
200 400 600 800 1000
3.0%
200 400 600 800 1000
34.7%
200 400 600 800 1000
40.5%
10
60
50
40
30
20
10
0
120
100
80
60
40
20
0
90
80
70
60
50
40
30
20
10
0
100 ꢃM
200 400 600 800 1000
26.6%
200 400 600 800 1000
21.2%
200 400 600 800 1000
22.0%
120
100
80
60
40
20
0
150
120
90
60
30
0
11
40
30
20
10
0
50 ꢃM
200 400 600 800 1000
200 400 600 800 1000
Propidium iodide
200 400 600 800 1000
(b)
75
50
25
0
WI38
B1
5
6
7
8
9
10
11
Compound
Fig. 1. OSCs alter leukaemic and WI38 cell-cycle profiles. Cells were exposed to acetone (solvent control) and two concentrations of OSCs for
48 h then stained with propidium iodide and analyzed by flow cytometry: (a) representative cell-cycle profiles in which the percentage of cells
stained in the pre-G1 region (apoptotic) is displayed in the upper right corner of each histogram; (b) a histogram comparing the average percentage
of cells stained in the pre-G1 region (apoptotic) after exposure of the non-transformed diploid cell line WI38 and a representative leukaemic cell line
B1 to OSCs 5–7 and 11 at 50 µM, and OSCs 8–10 at 100 µM. The results shown are an average calculated from three independent experiments
standard error of the mean.

132
R. Griffiths et al.
Table 3. Percentage of TUNEL-positive cells after 48 h exposure to OSCs
(mean standard error of mean)
Compound
Dose
Cell line
AML-3
25 µM 76.9 1.0
NB-4
33.8 1.3
B1
KK
11
5
74.5 0.2
70.5 3.6
26.1 4.6
50.6 5.4
39.2 2.9
75.4 0.7
22.1 4.3
60.4 8.1
20.7 9.0
33.7 3.9
46.1 5.7
45.5 28.2
25.9 6.7
54.6 5.0
67.1 0.2
62.8 7.8
29.7 2.1
40.5 8.8
29.7 1.6
74.5 6.5
31.8 7.1
45.9 7.5
20.8 8.3
33.7 3.9
30.1 20.9
47.3 8.8
31.4 12.8
49.5 17.3
50 µM 63.0 10.4 51.7 8.2
25 µM 86.3 3.3
50 µM 77.1 3.1
25 µM 55.2 38.8
50 µM 84.8 0.2
25 µM 12.4 3.8
57.9 22.1
73.1 3.8
9.4 2.5
24.9 6.6
12.4 3.8
6
7
50 µM 50.1 12.4 42.4 13.8
8
50 µM 14.0 6.0
100 µM 26.6 7.3
50 µM 11.0 1.0
100 µM 91.4 0.6
50 µM 16.1 5.5
12.0 2.7
38.6 10.5
6.8 0.5
28.6 2.8
21.7 7.7
9
10
100 µM 48.2 17.4 37.9 13.6
MTT results clearly showed that compound 8 was as potent
as the other five compounds in reducing proliferation exclu-
sively in leukaemia cells. Indeed, this result is reminiscent of
our previous analysis of compounds 3 and 4. In particular,
it showed that apoptosis could not be readily detected using
a fixed PI approach as OSCs 3 and 4 were found to induce
cells to undergo apoptosis from the G2/M phase of the cell
cycle.[15] Hence, we required a specific method that would
readily detect cells undergoing apoptosis in the G2/M phase
of the cell cycle.
Conclusions
Chemistry
Our initial report[15] disclosed that a pair of α-sulfonyl disul-
fides, one a methyl disulfide and one a phenyl disulfide,
were too toxic to normal cells. The current results show that
α-sulfonyl pentyl disulfide 5 has significantly diminished
toxicity towards non-transformed cells and is a competitive
antileukaemic in comparison to the compounds examined
herein (see Table 1). It is now clear that appropriately sub-
stituted α-sulfone disulfides have significant promise as
antileukaemic agents. Results for 8 (see Table 1) support
our contention that less aggressively activated α-substituted
disulfides, which may behave in accordance with Scheme 1,
are significantly more selective antileukaemics than the
α-sulfone disulfides examined in our earlier study.[15]
Tables 2 and 3 show results for compound 5 and sev-
eral other promising disulfides, including 6 and 7, which
can be correlated with antileukaemic efficacy using the
Scheme 1 mechanism. The encouraging results (Tables 2
and 3) for compounds 9 and 10 offer support for the notion
that the mechanism in Scheme 5 may serve as a basis for
the development of a second class of α-substituted disulfide
antileukaemics.
Apoptosis Detection Through DNA Fragmentation
We employed the terminal deoxynucleotidyl transferase-
mediated dUTP nick end-labelling (TUNEL) assay to eval-
uate whether the OSCs induced apoptosis in the leukaemia
cell lines. This assay allows characterization of drug-induced
apoptosis on a single-cell basis by directly measuring DNA
fragmentation.[21] ALL and AML cells were incubated in the
presence of each OSC for 48 h, fixed, and then the presence
of fragmented DNA was evaluated by labelling the 3ꢀ ends
of the fragmented DNA with a biotin-dUTP/fluorescein iso-
thiocyanate (FITC)-avidin system. Flow cytometry was then
applied to quantify the percentage of FITC-labelled cells.
Using this approach, we confirmed that the OSCs exam-
ined in this study induced the leukaemic cells to undergo
apoptosis. The cells were exposed to two concentrations
of each compound in order to evaluate the dose response.
The percentage of positive staining cells increased with an
increase in concentration of the sulfur compound and the
results were consistent across a panel of four leukaemic cell
lines (see Table 3, Fig. 2).
Biology
Our novel disulfides show antiproliferative and apoptotic
activity with specificity for leukaemic cell lines. The
compounds reduced the viability of the leukaemic cell lines
in a dose-dependent manner. From the MTT50 values, com-
pound 11 was determined to be a general cytotoxin, while
compounds 5–10 showed specificity for transformed cell
lines.
The fixed PI and TUNEL results indicate that growth
inhibitory effects are associated with an induction of apop-
tosis. The cell-cycle profiles suggest that leukaemic cell
lines undergo apoptosis, while the non-transformed WI38
fibroblast cell line undergoes cell cycle arrest in response
The cells were counterstained with PI during the TUNEL
approach in order to identify the phase of the cell cycle
in which apoptosis occurred. Indeed, as proposed, OSC 8
appears to target the G2/M phase of the cell cycle. Therefore,
although not fully detectable by fixed PI as was the case for
the two lead compounds 3 and 4,[15] OSC 8 does actually
induce apoptosis in leukaemic cells.

Novel Disulfides with Antitumour Efficacy and Specificity
133
Acetone
50 ꢃM
104
104
103
102
101
100
11
1.8%
41.7%
103
102
101
100
0
200 400 600 800 1000
Propidium iodide
0
200 400 600 800 1000
Propidium iodide
25 ꢃM
50 ꢃM
50 ꢃM
100 ꢃM
104
103
102
101
100
104
103
102
101
100
104
103
102
101
100
104
103
102
101
100
5
6
7
8
27.6%
49.3%
29.4%
33.6%
0
0
0
200 400 600 800 1000
0
200 400 600 800 1000
0
0
0
200 400 600 800 1000
0
200 400 600 800 1000
Propidium iodide
Propidium iodide
104
103
102
101
100
104
104
103
102
101
100
104
9
28.1%
67.9%
9.2%
38.6%
103
102
101
100
103
102
101
100
200 400 600 800 1000
0
200 400 600 800 1000
200 400 600 800 1000
0
200 400 600 800 1000
Propidium iodide
Propidium iodide
104
103
102
101
100
104
104
103
102
101
100
104
10
31.8%
38.2%
35.9%
46.0%
103
102
101
100
103
102
101
100
200 400 600 800 1000
0
200 400 600 800 1000
200 400 600 800 1000
0
200 400 600 800 1000
Propidium iodide
Propidium iodide
Fig. 2. OSCs induce apoptosis in leukaemic cell lines. TUNEL staining of representative leukaemic KK cells. The cells were exposed to acetone
(solvent control), OSC 11 at 50 µM (positive control), and to each of the other OSCs 5–7 at 25 and 50 µM, and 8–10 at 50 and 100 µM, and analyzed
by flow cytometry. The percentage of cells stained TUNEL positive (apoptotic) is displayed in the top right quadrant of each profile. Results shown
are representative of three independent experiments.
to compounds 5–10. Increased sensitivity of the transformed
cells to disulfide-induced apoptosis in comparison to the non-
transformed cell line WI38 is consistent with the concept that
cells which harbour genetic abnormalities that deregulate
cell-cycle checkpoints are sensitized to undergo apoptosis
when exposed to various cytotoxic or cytostatic agents.[22]
Therefore, the OSCs may provoke the leukaemic cell lines to
undergo apoptosis as a consequence of deregulated cell-cycle
checkpoints in the transformed cells.
Although the biological mechanism of the apoptosis
induction by these OSCs remains unknown, compound 8
appears to induce apoptosis from the G2/M phase of the cell
cycle. However, it is not known whether growth arrest of
cells in the G2/M phase of the cell cycle actually occurs and

134
R. Griffiths et al.
apoptosis is simply a consequence of the arrest in the G2/M
phase. Indeed, this profile is similar to that of OSCs 3 and 4
reported previously.[15] The induction of G2/M phase arrest
appears to be a characteristic of other disulfides under inves-
tigation for antitumour activity, such as ajoene, diallyl disul-
fide, and imidazolyl disulfides.[23–26] Exposure of the cells
to the six remaining compounds revealed that death occurred
from all phases of the cell cycle. It will be illuminating to
elucidate the mechanisms of disulfide activity and compare
them with those of other cell-cycle inhibitors. Furthermore,
disulfides such as ajoene and diallyl disulfide or dithiosul-
finates have been associated with the induction of reactive
oxygen species, inhibition of metabolic enzymes, DNA mod-
ification, and disruption of signalling pathways.[27–36] It will
be of interest to compare and contrast the mechanism of
apoptosis induced by the disulfides in this study.
The functional screen approach employed in this study
revealed that selected disulfides have potent apoptotic
activity, andthatchangestotheregionsthatflankthedisulfide
moiety modulate the toxicity and specificity of each com-
pound toward transformed cells. Clearly, structural relatives
of compounds 5–10 have the potential to be exploited fur-
ther to gain increased sensitivity in inducing apoptosis while
maintaining specificity for transformed cells. Experimental
evidence provided in this study suggests that disulfides have
therapeutic and clinical potential as novel anticancer agents.
Prism 3.0 (GraphPad Software Inc., San Diego, CA). A dose–response
curve for each compound was repeated, independently, three times.
Fixed PI Staining
Approximately 1 × 106 leukaemia cells and 3.5 × 105 WI38 cells
were seeded for PI staining in a six-well dish (Falcon). Cells were then
exposed for 48 h to an acetone control and two concentrations of each
compound: 5–7, 11 (25 and 50 µM), and 8–10 (50 and 100 µM). The
cells were subsequently harvested, washed, and fixed with 80% ethanol.
Cells were stained with 50 µg mL−1 PI as described previously[37] and
approximately 2.0 × 104 cells were analyzed using a FACScalibur flow
cytometer (Becton Dickinson, San Jose, CA). Cell-cycle parameters
were measured with Cellquest software (Becton Dickinson, San Jose,
CA). Each sample was repeated, independently, three times.
TUNEL
Approximately 1 × 106 leukaemia cells were seeded into a six-well
dish (Falcon). The cells were then exposed to an acetone control and
two concentrations of each compound: 5–7, 11 (25 and 50 µM), and
8–10 (50 and 100 µM). After 48 h, cells were harvested and fixed with
4% formaldehyde for 15 min. The cells were washed in cold D-PBS and
stored at −20◦C in 70% ethanol. Samples were labelled with FITC using
the TUNEL method as described previously.[15] Briefly, about 1 × 106
treated cells were incubated in labelling mixture that contained biotin-
dUTP (Roche) and 12.5 U TdT enzyme (Boehringer Mannheim) for
45 min at 37◦C. Samples were washed and spun at 1000 rpm for 5 min.
The cell pellet was resuspended in 200 µL of 1 : 1000 fluorescein iso-
thiocyanate (FITC)-conjugated avidin (Sigma) in 4 × SSC, 5% skim
milk powder, and 0.05% Tween-20 (Sigma) for 1 h. Cells were then
washed and stained with 10 µg mL−1 PI for 30 min. Approximately
2 × 104 cells were analyzed for FITC-positive cells using the FACScal-
ibur flow cytometer (Becton Dickinson, San Jose, CA). Samples were
analyzed using Cellquest software (Becton Dickinson, San Jose, CA).
Each sample was repeated, independently, three times.
Experimental
Biology
Cell Culture and Cell Lines
Chemistry
The human ALL cell lines KK and B1, as well as the human AML
cell lines AML-3 and NB-4 were derived from patient samples as
described previously.[37] The human fibroblast cell line WI38 was pur-
chased fromAmericanType Culture Collection (Manassas,VA).All cell
lines were assayed as asynchronously growing cells and maintained in α-
minimal essential medium (α-MEM; Princess Margaret Hospital Media
Services) supplemented with 10% fetal bovine serum (FBS; Sigma,
St. Louis, MO) and 1% penicillin/streptomycin at 37◦C under 5% CO2.
The WI38 cell line was plated subconfluently 24 h before treatment.
General
Details have been provided earlier.[16]
Thiosulfonate and Disulfide Syntheses
Some OSCs (see Scheme 6) were prepared as described earlier,
namely the sulfone disulfide 5,[40] the thiosulfonate 6,[41] the nitro-
phenyl disulfide 11,[17] the diester disulfide 9,[42] and the α-ester
disulfide 7.[43]
Biological Testing: General Sample Preparation
Preparation of 2,3,5-Trithiahexane 8
Approximately 20 mg of each compound was dissolved inACS grade
acetone (5 mL, Sigma). Stock solutions of the compounds were stored
in the dark at 4◦C. Compounds were diluted, as indicated, immediately
before each experiment.
Sodium metal (0.02 g, 0.87 mmol) was dissolved in methanol (2 mL)
and methanethiol (20 mL) was bubbled slowly through the solution. The
solvent was evaporated and the residue was dried under vacuum before
being dissolved in DMSO (10 mL). A portion (1 mL) of the resultant
solution was added to a solution of dimethyl disulfide (12 mL) and
2,4,5,7-tetrathiaoctane[44] (2.0 g, 10.7 mmol). The reaction mixture was
stirred at ambient temperature for 8 days. Hydrochloric acid (70 mL,
2.5% v/v) was added and the resultant mixture was extracted with
diethyl ether (3 × 50 mL). The combined organic fractions were dried
(MgSO4), filtered, andthesolventwasevaporated.Theresiduewasrecti-
fied at reduced pressure to afford unchanged dimethyl disulfide (3.52 g)
and 2,3,5-trithiahexane 8 (2.3 g, 78%), bp 75–85◦C/18 Torr. Isolated
compound 8 [δC (68 MHz, CDCl3) 65.1, 73.4, 94.2] was identical to
previously described material.[44]
MTT Assays
As described previously,[15] leukaemic and WI38 fibroblast cells
were seeded at a density of 2.7 × 105 and 6.7 × 104 cells mL−1, respec-
tively, in 96-well plates (Falcon, Mississauga, ON). The cells were
exposed to an acetone control as well as a dose range of each com-
pound in a total volume of 150 µL and assayed in triplicate. After a
48 h incubation under 5% CO2 at 37◦C, 40 µL of a 5 mg mL−1 solution
of MTT substrate (Sigma) in Dulbecco’s phosphate-buffered saline (D-
PBS) was added to each well. The resultant violet formazan precipitate
went into solution overnight after addition of 80 µL of a 0.01 M HCl
and 10% sodium dodecyl sulfate (SDS; Sigma) solution and a further
4 h of incubation at 37◦C under 5% CO2. The plates were then analyzed
using the BioRad Benchmark Microplate Reader (BioRad Laboratories,
Hercules, CA) at 570 nm to determine the optical density of the sam-
ples. MTT data was analyzed by the Chou–Talalay method[38,39] using
Preparation of Phenacyl Thiolacetate
A solution of thiolacetic acid (0.56 g, 7.3 mmol) in dry pyridine
(4 mL) was added to a solution of phenacyl chloride (1.0 g, 6.5 mmol)
and the resultant reaction was maintained at 83◦C for 1.5 h. Chloro-
form (200 mL) was added to the orange reaction mixture, which was

Novel Disulfides with Antitumour Efficacy and Specificity
135
washed with 5% v/v hydrochloric acid (150 mL) and extracted with
2.5% w/v aqueous sodium hydroxide (100 mL). The organic layer was
dried (MgSO4), filtered, and the solvent was evaporated. The crude
product was purified by chromatography on silica gel (5 g) employing
light petroleum (400 mL) as eluent. The solvent was evaporated and the
residue was rectified at reduced pressure to afford phenacyl thiolacte-
tate (0.96 g, 75%), bp 137–139◦C/2.4 Torr. νmax/cm−1 1700, 1680. δH
(270 MHz, CDCl3; Me4Si) 2.38 (3H, s, CH3), 4.39 (2H, s, CH2), 7.46
(2H, t, ArH), 7.58 (1H, t, ArH), 7.97 (2H, d, ArH). δC (68 MHz; CDCl3)
30.2, 36.6, 128.4, 128.7, 133.7, 135.5, 193.1, 194.1. m/z (GLC/MS) 194
(0.5%, M+•), 152 (7.4), 105 (100).
[2] L. Z. Penn, Curr. Opin. Investig. Drugs 2001, 5, 684.
[3] M. Zornig,A. Hueber, W. Baum, G. Evan, Biochim. Biophys.Acta
2001, 1551, F1.
[4] F. R. G. Braunschweig, Cancer Lett. 1990, 53, 103.
doi:10.1016/0304-3835(90)90201-8
[5] K. Sakomoto, L. D. Lawson, J. A. Milner, Nutr. Cancer 1997,
29, 152.
[6] S. G. Sundaram, J. A. Milner, Biochim. Biophys. Acta 1996,
1315, 15.
[7] G. Sigounas, J. L. Hooker, W. Li,A.Anagnostou, M. Steiner, Nutr.
Cancer 1997, 28, 153.
[8] G. Sigounas, J. Hooker, A. Anagnostou, M. Steiner, Nutr. Cancer
1997, 27, 186.
Preparation of Diphenacyl Disulfide
[9] H. Nakagawa, K. Tsuta, K. Kiuchi, H. Senzaki, K. Tanaka,
K. Hioki, A. Tsubura, Carcinogenesis 2001, 22, 891.
doi:10.1093/CARCIN/22.6.891
[10] S. G. Sundaram, J. A. Milner, Carcinogenesis 1996, 17, 669.
[11] V. M. Dirsch,A. L. Gerbes,A. M.Vollmar, Mol. Pharmacol. 1998,
53, 402.
[12] S. G. Sundaram, J. A. Milner, J. Nutr. 1996, 126, 1355.
[13] E. Block, R. DeOrazio, M. Thiruvazhi, J. Org. Chem. 1994,
59, 2273.
[14] M. K. Jogia, R. J. Andersen, E. K. Mantus, J. Clardy, Tetrahedron
Lett. 1989, 30, 4919. doi:10.1016/S0040-4039(01)80542-6
[15] W. W. Wong, S. MacDonald, R. F. Langler, L. Z. Penn,
Anticancer Res. 2000, 20, 1367.
[16] F. J. Baerlocher, R. F. Langler, M. U. Frederiksen, N. M. Georges,
R. D. Witherell, Aust. J. Chem. 1999, 52, 167. doi:10.1071/
C98141
[17] F. J. Baerlocher, M. O. Baerlocher, R. F. Langler, S. L. MacQuarrie,
M. E. Marchand, Aust. J. Chem. 2000, 53, 1. doi:10.1071/
CH99139
Potassium carbonate (19.0 g) was added to methanol (160 mL) and
the resultant mixture was cooled in an ice-water bath for 1 h before
phenacyl thiolacetate (4.2 g, 21.4 mmol) was added. The cold reaction
was then stirred for 0.5 h. Diethyl ether (205 mL) and water (200 mL)
were added and the ice-water bath was maintained for a further 0.5 h.
Iodine (3.59 g) was then added in small portions over 30 min fol-
lowed by an aqueous solution of saturated sodium thiosulfate (16 mL)
and diethyl ether (85 mL). The organic layer was separated, washed
with water (2 × 200 mL), dried (MgSO4), filtered, and the solvent was
evaporated. The residue was purified by chromatography on silica gel
(400 g) employing chloroform (100 mL fractions) as eluent. Fractions
16–29 were concentrated and combined. Purified diphenacyl disulfide
was recrystallized twice from benzene/methanol to afford a granular
solid (0.18 g, 3%), mp 68–72◦C (Found: C 63.2, H 4.6. C16H14O2S2
requires C 63.5, H 4.7%). νmax/cm−1 1680. δH (270 MHz, CDCl3;
Me4Si) 4.20 (2H, s, CH2), 7.49 (2H, t, ArH), 7.59 (1H, t, ArH), 7.94
(2H, d, ArH). δC (68 MHz, CDCl3) 45.4, 128.7, 128.8, 133.7, 135.4,
194.3. m/z (GLC/MS) 105 (100%), 77 (52).
[18] R. B. Silverman, The Organic Chemistry of Drug Design and
Drug Action 1992, p. 16 (Academic Press: New York, NY).
[19] M. Malumbres, M. Barbacid, Nat. Rev. Cancer 2001, 1, 222.
doi:10.1038/35106065
[20] I. Nicoletti, G. Migliorati, M. C. Pagliacci, F. Grignani,
C. Riccardi, J. Immunol. Methods 1991, 139, 271. doi:10.1016/
0022-1759(91)90198-O
[21] S. Nagata, Exp. Cell Res. 2000, 256, 12. doi:10.1006/EXCR.
2000.4834
[22] G. I. Evan, A. H. Wyllie, C. S. Gilbert, T. D. Littlewood, H. Land,
M. Brooks, C. M. Waters, L. Z. Penn, D. C. Hancock, Cell 1992,
69, 119. doi:10.1016/0092-8674(92)90123-T
[23] L. M. Knowles, J. A. Milner, Nutr. Cancer 1998, 30, 169.
[24] D. J. Frantz, B. G. Hughes, D. R. Nelson, B. K. Murray,
M. J. Christensen, Nutr. Cancer 2000, 38, 255. doi:10.1207/
S15327914NC382_15
[25] M. Li, J. R. Ciu, Y. Ye, J. M. Min, L. H. Zhang, K. Wang,
M. Gares, J. Cros, M. Wright, J. Leung-Tack, Carcinogenesis
2002, 23, 573. doi:10.1093/CARCIN/23.4.573
[26] A. Vogt, K. Tamura, S. Watson, J. S. Lazo, J. Pharmacol. Exp.
Ther. 2000, 294, 1070.
Preparation of Phenacyl Methyl Disulfide 10
Sodium metal (0.019 g, 0.82 mmol) was dissolved in methanol
(10 mL) and methanethiol (20 mL) was slowly bubbled into the reaction
mixture. The solvent was evaporated and the residue was dried under
vacuum. DMSO (10 mL) was added and the mixture was stirred at ambi-
ent temperature for 0.5 h. A portion (1 mL) was then added to a solution
of diphenacyl disulfide (1.68 g, 5.5 mmol) in dimethyl disulfide (12 mL)
and the resultant reaction mixture was stirred at ambient temperature for
8 days. Hydrochloric acid (70 mL, 2.5% v/v) was added and the resul-
tant mixture was extracted with diethyl ether (3 × 50 mL).The combined
organic layers were dried (MgSO4), filtered, and the solvent was evap-
orated. The crude product was purified by chromatography on silica
gel (200 g) employing 3 : 7 chloroform/light petroleum (100 mL frac-
tions) as eluent. Fractions 12–19 were combined and concentrated under
reducedpressure.Thepurifieddisulfidewasdistilledatreducedpressure
to afford phenacyl methyl disulfide 10 (1.12 g, 51%), bp 145–150◦C/1.7
Torr. νmax/cm−1 1690. δH (270 MHz, CDCl3; Me4Si) 2.37 (3H, s,
SSCH3), 4.09 (2H, s, CH2), 7.47 (2H, t, ArH), 7.59 (1H, t, ArH), 7.97
(2H, d,ArH). δC (68MHz, CDCl3)22.4, 43.7, 128.1, 128.2, 132.9, 134.7,
193.9. m/z (GLC/MS) 198 (18%, M+•), 153 (3), 105 (100), 77 (43).
[27] M. Focke, A. Feld, K. Lichtentaler, FEBS Lett. 1990, 261, 106.
doi:10.1016/0014-5793(90)80647-2
Acknowledgments
[28] R. Gebhardt, H. Beck, Lipids 1996, 12, 1269.
[29] S. Lee, S. Park, J. W. Oh, C. Yang, Planta Med. 1998, 64, 303.
[30] H. Nakagawa, K. Tsuta, K. Kiuchi, H. Senzaki, K. Tanaka,
K. Hioki, A. Tsubura, Carcinogenesis 2001, 22, 891. doi:
10.1093/CARCIN/22.6.891
[31] V. Robert, B. Mouille, C. Mayeur, M. Michaud, F. Blachier,
Carcinogenesis 2001, 22, 1155. doi:10.1093/CARCIN/22.
8.1155
[32] V. M. Dirsh, D. S. Antisperger, H. Hentze, A. M. Vollmar,
Leukemia 2002, 16, 74. doi:10.1038/SJ.LEU.2402337
[33] V. M. Dirsch,A. L. Gerbes,A. M.Vollmar, Mol. Pharmacol. 1998,
53, 402.
This work was supported by MedInnova Partners.The authors
acknowledge technical assistance from P. O’Connor. High-
field NMR spectra were obtained by D. Durant. Mass spectra
wererunbyR. Smith.TheauthorsalsothankDrM. D. Minden
and Dr M. H. Freedman for providing our AML and ALL
cell lines, respectively. W.W.-L.W. is supported by a CIHR
Doctoral Award.
References
[1] J. C. Reed, Trends Mol. Med. 2001, 7, 314. doi:10.1016/
[34] G. Filomeni, K. Aquilano, G. Rotilio, M. R. Ciriolo, Cancer Res.
2003, 63, 5940.
S1471-4914(01)02026-3

136
R. Griffiths et al.
[35] D. S. Antisperger, V. M. Dirsch, D. Ferreira, J. L. Su,
M. L. Kuo, A. M. Vollmar, Oncogene 2003, 22, 582. doi:
10.1038/SJ.ONC.1206161
[40] F. J. Baerlocher, M. O. Baerlocher, R. F. Langler,
S. L. MacQuarrie, P. E. O’Connor, Sulfur Lett. 2002, 25, 135.
doi:10.1080/02786110213980
[36] L. M. Knowles, J. A. Milner, J. Nutr. 2003, 133, 2901.
[37] T. C. Chou, P. Talalay, Adv. Enzyme Reg. 1984, 22, 27.
doi:10.1016/0065-2571(84)90007-4
[41] F. J. Baerlocher, M. O. Baerlocher, C. L. Chaulk, R. F. Langler,
S. L. MacQuarrie, Aust. J. Chem. 2000, 53, 399. doi:10.1071/
CH00030
[38] The Median-Effect Principle and the Combination Index for
Quantitation of Synergism andAntagonism (Ed.T. C. Chou) 1991,
p. 61 (Academic Press: San Diego, CA).
[42] F. J. Baerlocher, M. O. Baerlocher, K. D. Guckert, R. F. Langler,
S. L. MacQuarrie, P. E. O’Connor, G. C. Y. Sung, Aust. J. Chem.
2001, 54, 397. doi:10.1071/CH01046
[39] J. Dimitroulakos, D. Nohynek, K. L. Backway, D. W. Hedley,
H. Yeger, M. H. Freedman, M. D. Minden, L. Z. Penn, Blood
1999, 93, 1308.
[43] F. J. Baerlocher, M. O. Baerlocher, C. L. Chaulk, R. F. Langler,
E. M. O’Brien, Sulfur Lett. 2000, 24, 101.
[44] P. Dubs, R. Stuessi, Helv. Chim. Acta 1978, 61, 2351.