Synthesis, biological evaluation, and molecular modeling of novel thioacridone derivatives related to the anticancer alkaloid acronycine

Article abstract of DOI:10.1016/j.bmc.2004.10.051

The well-reported, but moderate antitumor activity of the acronycine alkaloid led us to synthesize a novel series of thioacridone compounds related to acronycine, as potential anticancer agents. Compounds were designed either as DNA intercalating agents, or as DNA intercalating agents with covalent bond forming potential. Bathochromic shifts of the compounds upon complexation with salmon testis DNA suggested intercalation as the mode of DNA binding. The binding interaction of the compounds was found to be approximately 10 ² M^-1, with that of the most potent compound 1-(2-dimethylaminoethylamino)-9(10H)-thioacridone, 10⁴ M ^-1. In vitro cytotoxic activity (IC₅₀) against HL-60 cells was found to range between 3.5 and 22 μg/mL. QSAR analyses yielded a multiple linear regression equation with an r² of 0.847 for DNA binding and an r² of 0.575 for cytotoxicity. The physicochemical parameters used in the QSAR analyses were log P, polar surface area, and calculated molar refractivity. Docking studies were also performed to compare the binding of the most potent and least potent compounds in the study in order to predict desirable chemical characteristics for further exploitation in drug design efforts. The thioacridone compounds in this series demonstrate cytotoxic activity in vitro that merit future in vivo evaluation.

Full text of DOI:10.1016/j.bmc.2004.10.051

Bioorganic & Medicinal Chemistry 13 (2005) 689–698
Synthesis, biological evaluation, and molecular modeling of novel
thioacridone derivatives related to the anticancer alkaloid
acronycine
James P. Dheyongera,a Werner J. Geldenhuys,a,c Theodor G. Dekkerb and
Cornelis J. Van der Schyfa,c,
*
aPharmaceutical Chemistry, North West University, Potchefstroom 2520, South Africa
bResearch Institute for Industrial Pharmacy, North West University, Potchefstroom 2520, South Africa
cDepartment of Pharmaceutical Sciences, Texas Tech University Health Sciences Center, School of Pharmacy, Amarillo,
TX 79106, USA
Received 2 September 2004; revised 19 October 2004; accepted 25 October 2004
Available online 21 November 2004
Abstract—The well-reported, but moderate antitumor activity of the acronycine alkaloid led us to synthesize a novel series of thio-
acridone compounds related to acronycine, as potential anticancer agents. Compounds were designed either as DNA intercalating
agents, or as DNA intercalating agents with covalent bond forming potential. Bathochromic shifts of the compounds upon com-
plexation with salmon testis DNA suggested intercalation as the mode of DNA binding. The binding interaction of the compounds
was found to be approximately 102 Mꢀ1, with that of the most potent compound 1-(2-dimethylaminoethylamino)-9(10H)-thioacri-
done, 104 Mꢀ1. In vitro cytotoxic activity (IC50) against HL-60 cells was found to range between 3.5 and 22lg/mL. QSAR analyses
yielded a multiple linear regression equation with an r2 of 0.847 for DNA binding and an r2 of 0.575 for cytotoxicity. The physic-
ochemical parameters used in the QSAR analyses were logP, polar surface area, and calculated molar refractivity. Docking studies
were also performed to compare the binding of the most potent and least potent compounds in the study in order to predict desirable
chemical characteristics for further exploitation in drug design efforts. The thioacridone compounds in this series demonstrate cyto-
toxic activity in vitro that merit future in vivo evaluation.
ꢀ 2004 Elsevier Ltd. All rights reserved.
1. Introduction
origin. These findings were further developed and
reviewed in elegant studies by Tillequin and his group.4
Acronycine (Fig. 1), also known as acronine, is an acri-
done alkaloid which was first isolated in 1948 from the
bark of the Australian tree Acronychia baueri.1 In sub-
sequent investigations it was found that acronycine
possess anticancer activity against a wide spectrum
of experimental neoplasms in laboratory (murine) ani-
mals.2 Dorr et al.3 studied the in vitro cytotoxicity of
acronycine in a variety of solid malignancies against
human tumor cells isolated from patients. The results
of early studies showed that acronycine possesses activi-
ty against a variety of solid human tumors, including
ovarian cancer and metastatic tumors of unknown
The structure–activity relationship (SAR) in the acrony-
cine group of compounds appears to be very inflexible
considering that most of the acronycine derivatives
tested to date against cancer, showed much lower activ-
ity than acronycine itself. Schneider and co-workers5 at-
tempted to synthesize structural analogs by modifying
the side chain in the 6-position on acronycine. However,
compound 1 (Fig. 1), containing an oxy-dimethylamino-
ethyl side chain in lieu of the methoxy group in the 6-
position of acronycine, exhibited significant antitumor
activity. This compound was, therefore, a strong moti-
vating factor in the design of the compounds reported
in the current study. Furthermore, in addition to com-
pound 1, several acronycine analogs, in which the
methoxy group at C-6 has been replaced by dialkyl-
aminoalkylamino side chains (including a dimethyl-
aminoethyl analog) have recently been tested in a
Keywords: Thioacridones; Acronycine; DNA intercalation; Docking;
QSAR; Bathochromic shift.
*
Corresponding author. Tel.: +1 806 356 4015x329; fax: +1 806 356
4034; e-mail: neels.vanderschyf@ttuhsc.edu
0968-0896/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2004.10.051

690
J. P. Dheyongera et al. / Bioorg. Med. Chem. 13 (2005) 689–698
O
OCH3
H3CO
NHSO2CH3
8
6
7
9
5
NH
10
4
3
N
O
12
11
CH3
CH3
1
CH3
2
N
Amsacrine
Acronycine
O
O(CH2)2N(CH3)2
N
O
CH3
CH3
CH3
1
Figure 1. Structure of compounds discussed in the text, including the lead compound, 1.5
Cl
Cl
O
Cl
O
Cl
CO OH
COOH
H2SO4
K2CO3
CH3I
X
H2N
N
H
N
H
N
2
X = Cl or Br
R
R
CH3
R
R
20
18
R = H, Cl, CH3
19
POCl3
POCl3
Cl
Cl
S
Cl
Cl
Cl
S
Cl
NaHS
Na2S2SO3
(PO2Cl2)
N
N
H
N
N
22
R
R
12:R=H
13:R=Cl
14:R=CH3
CH3
21
R
CH3
R
15:R=H
16:R=Cl
17:R=CH3
H2N(CH2)2N(CH3)2
S
NH(CH2CH2OH)2
S
NH2(CH2)2N(CH3)2
NH(CH2)2N(CH3)2
N(CH2CH2OH)2
S
NH(CH2)2N(CH3)2
8
1
9a
4a
8a
9
7
2
3
N
H
N
10
N
6
10a
5
4
R
H
R
23
CH3
R
6:R=H
SOCl2
7:R=Cl
3:R=H
4:R=Cl
8:R=CH3
S
N(CH2CH2Cl)2
5:R=CH3
N
H
9:R=H
R
10:R=Cl
11:R=CH3
Scheme 1. Synthesis of compounds used in this study.
cytotoxic assay.6 This modification resulted in a signifi-
cant increase of the cytotoxic activity toward L-1210
cells compared with acronycine.6
1), as well as alkylation (compounds 9–11 in Scheme 1).
There are two characteristics common to many DNA
intercalators, a planar polycyclic chromophore able to
intercalate into DNA, and the presence of a basic side
chain.7 This side chain can both increase DNA binding
affinity and, depending on its hydrophilicity, may in-
crease the solubility of a compound. The fused planar
We were interested to introduce structural modifications
into the acronycine molecule with two modes of action
in mind: DNA intercalation (compounds 3–8 in Scheme

J. P. Dheyongera et al. / Bioorg. Med. Chem. 13 (2005) 689–698
691
tricyclic moiety contained in acronycine that is necessary
for DNA binding by intercalation was, therefore, main-
tained and a basic side chain was introduced in the 6-
position by replacing the methoxy group with a 2-
(dimethylamino)ethylamino chain. The nonplanar por-
tion of the polycyclic ring structure (the pyran ring)
was eliminated since it was reported not to be essential
for the intercalation process,8 although Elomri et al.9
did suggest that the 1,2-double bond on the pyran ring
may be an essential structural requirement for cytotoxic
activity.
priate N-phenylanthranilic acids 18 with sulfuric acid
following previously described methods.11,12
3. Results and discussion
3.1. Spectral shift determination
UV–vis absorption spectroscopy represents one of the
most convenient and facile ways to identify compounds
that intercalate with DNA. The technique involves
measuring the bathochromic shift (wavelength absorp-
tion to longer wavelengths) that is encountered as a
compound intercalates with DNA. The values of the
bathochromic shifts measured, are listed in Table 1
and an example shown in Figure 2.
Furthermore, the carbonyl oxygen in the 7-position was
replaced with sulfur. The replacement of oxygen
(as found in the acronycine-derived series reported by
other investigators, including Smolders et al.10) with sul-
fur, introduces a new element in the structure–activity
equation. Sulfur is an atom with a larger atomic radius
than that of oxygen. Consequently, the presence of sul-
fur imparts lower electronegativity and lower ionization
energy to the molecule, and such molecules have a low-
ered tendency to form ionic bonds. Therefore, the C@S
group is less polar than the C@O group. Thus, thioke-
tone compounds are expected to be less water soluble
than their corresponding carbonyl congeners. However,
replacement of oxygen by sulfur as proposed above,
results in compounds that are not merely weakly
basic acridines, but such modification also drastically
alters many electronic properties including the molecu-
lar dipole moment and electronic charge distribution.
These changes might have a considerable influence
on the biological activity of the resulting compounds.
Steric and electronic influences were also investi-
gated using chloro and methyl substituent groups at
the 4-position. Compounds with the general structure
of 9–11, were synthesized with a side chain that
likely will display DNA cross-linking characteristics
(Scheme 1).
Compounds 3–11 underwent bathochromic shifts simi-
lar to that of the reference compound acriflavine. The
spectra in the visible range of the new compounds
showed bathochromic shifts in the order of 4–22nm
when the compounds bound to DNA. These results sug-
gest that these compounds interact with DNA through
intercalation. The 1-chlorothioacridone intermediates
12–14 and 15–17 showed virtually no bathochromic shift
after binding with DNA, suggesting that these com-
pounds do not intercalate with DNA.
3.2. DNA binding and intercalation studies
The results of ethidium binding are given in Table 2 as a
C50 value of the compounds. The C50 value of a drug is
Table 1. Bathochromic shifts (Dk) of 1-aminothioacridones upon
addition of DNA
Compound
kmax (nm)
Dk (nm)
3
418
410
431
403
385
418
454
463
479
445
+4
+5
4
5
+8
6
+12
+10
+13
+4
2. Chemistry
7
8
The substrate N-phenylanthranilic acids 18, used as
starting compounds (Scheme 1) were prepared through
condensation of an appropriate 2-halobenzoic acid with
an excess amount of the appropriate aromatic amine.
Cyclization of compounds 18 with phosphoryl chloride
yielded the dichloroacridines 22, which upon reaction
with sodium hydrogen sulfide afforded the 1-chlorothio-
acridones 12–14. Treatment of the latter compounds
with an excess amount of dimethylaminoethylamine
then afforded the target compounds 6–8. The prepara-
tion of compounds 3–5 involved the reaction of the
appropriate 1-chloro-N-methylthioacridones (15–17)
with 2-dimethylaminoethylamine. The necessary start-
ing thioacridone compounds 15–17 were obtained via
the reaction of sodium thiosulfate with 9-chloroacri-
dium salts 21, which were produced from 1-chloro-N-
methylacridones (20) and phosphoryl chloride. The re-
quired N-methylacridones 20 in turn were obtained by
N-methylation of the corresponding 1-chloroacridones
19 using methyl iodide. The 1-chloroacridones required
for this study were prepared via cyclization of the appro-
9
10
+7
11
Acriflavine
+22
+19
Figure 2. UV absorption spectrum showing the bathochromic effect of
DNA addition on the spectrum of 6.

692
J. P. Dheyongera et al. / Bioorg. Med. Chem. 13 (2005) 689–698
Table 2. C50 values of 1-aminothioacridones
compounds seem to have lower binding affinities to
DNA than their corresponding desmethyl (6–8 and
12–15) congeners.
Compound
C50 (lM)
3
23.2
42.6
38.2
8.7
4
With the exception of compound 6, the interaction of
the target compounds with DNA was found to be rela-
tively weak, with binding affinity constants in the region
of 102 Mꢀ1 compared with most other DNA intercalat-
ing drugs which show binding affinities in the range
of 105–1010 Mꢀ1.8 This attenuated DNA binding charac-
teristic could impart a broad-spectrum anticancer
activity to these compounds. Anticancer drugs require
optimization of the distributive properties of the drug
in order to penetrate to remote tumor sites in sufficient
concentrations to be effective. For DNA intercalation
agents, compounds that will distribute most efficiently
are likely to be those with the lowest DNA binding
constants, since a larger proportion of unbound drug
will be available at equilibrium. Consequently, the new
1-aminothioacridone compounds are expected to have
a broad scope of anticancer activity including solid
tumors, as well as leukemia.
5
6
7
35.3
26.8
43.1
46.4
50.2
8
9
10
11
the concentration necessary to reduce the fluorescence of
initially DNA-bound ethidium by 50% under standard
assay conditions.13 A reduction in the fluorescence of
DNA–ethidium complexes was found upon addition of
the 1-aminothioacridone compounds 3–11. Such a
reduction in fluorescence is indicative of intercalative
binding into DNA. Only marginal displacement of ethi-
dium was observed with the 1-chlorothioacridone inter-
mediates 12–17. As a result, the C50 values of these
compounds could not be determined. The results in
Table 1 show that compound 6 binds to DNA with
the highest affinity. The results also show that com-
pounds 3–8—carrying an NH(CH2)2N(CH3) side
chain—bind to DNA with higher affinities than com-
pounds 9–11, which all contain a nitrogen mustard
moiety. This is in qualitative agreement with the binding
constant values obtained for these compounds, as
shown in Table 3 below.
3.3. Cytotoxicity evaluation
The results of the cytotoxic evaluation are given in Table
3, expressed as the concentration of an individual com-
pound that is required to kill 50% of the cells (IC50) after
4 days of drug exposure. The cytotoxicity of the new
thioacridone compounds was compared with that of
amsacrine (see Fig. 1), a potent cytotoxic agent and also
structurally related to the target compounds 3–11
(Scheme 1, Fig. 1). Amsacrine has been shown to bind
to DNA both by intercalation and by external binding.15
The binding constant of the thioacridone compounds
with DNA was measured to determine the degree of
intercalation of the compounds with DNA in vitro.
Scatchard plots were used to determine the DNA bind-
ing constants (K, Table 3). The results show that all tar-
get compounds examined (3–11), interact with DNA,
albeit with different affinities. The results further indi-
cate that 1-chlorothioacridone intermediates 12–17 are
also capable of forming complexes with DNA. Their
binding affinities, however, were much lower than those
measured for the compounds carrying a side chain
(compare 12–14 with 6–8, or 15–17 with 3–5). The addi-
tion of a basic chain to the thioacridone chromophore
improved the DNA binding properties of the system,
likely by stabilization of the intercalation through ionic
interaction between the protonated side chain nitrogen
and the ionic phosphate groups of DNA. On the other
hand, the ring nitrogen of the 1-chlorothioacridone pre-
cursors 12–17 is not basic and, therefore, not protonated
at pH7.0, and is also not available to stabilize the inter-
calation of the tricyclic ring.
Our results show that compounds 3–11 exhibit signifi-
cant in vitro cytotoxic activity with IC50 values ranging
from ꢁ2 to 15lg/mL. Compound 6 was the most active
with an IC50 value of 2.3lg/mL. The activity of this
compound, however, was still lower than that of amsa-
Table 3. Binding constants (K) for 1-aminothioacridones 3–11 and
their 1-chlorothioacridone precursors 12–17, and in vitro cytotoxicity
against HL-60 human promyelocytic leukemia cells
Compound
K (Mꢀ1
)
IC50 (lg/mL)
3
8.6 · 102
1.1 · 102
2.3 · 102
2.6 · 104
1.2 · 102
2.7 · 102
1.1 · 102
0.9 · 102
0.6 · 102
0.008 · 102
0.006 · 102
0.007 · 102
0.003 · 102
0.004 · 102
0.005 · 102
—
3.5
5.6
4
4
5
6
2.3
8.5
15
7
8
9
10
8.7
10.2
17
Among compounds with a side chain, those carrying a
substituent at the 4-position of the thioacridone struc-
ture bind to DNA less strongly than those in which
the 4-position is unsubstituted (compare 7 and 8 with
6). This observation may be the result of a possible
impeding effect experienced by rings substituted with
either a chloro or methyl group at the 4-position upon
intercalative entry of the thioacridone nucleus into the
DNA molecule. The N-methylated (3–5 and 15–17)
11
12
22.6
>26
>26
8.9
6.7
6
13
14
15
16
17
Amsacrine
0.01a
a Taken from Su et al.14

J. P. Dheyongera et al. / Bioorg. Med. Chem. 13 (2005) 689–698
Table 4. QSAR descriptors of the compounds studied
693
crine, which has a reported IC50 value of 0.01lg/mL
against human leukemia HL-60 cells.14 It must be noted,
though, that the IC50 value for amsacrine was obtained
using a different technique and under different condi-
tions13 than those utilized in the current study.
Compound
PSA
logP
CMR
3
56.865
55.242
55.028
79.739
75.659
64.261
39.316
40.590
32.989
66.547
65.349
55.924
43.371
41.347
41.347
2.95
3.51
3.44
2.71
3.27
3.20
4.92
5.48
5.41
3.78
4.34
4.26
4.01
4.57
4.50
9.979
10.470
10.443
4
5
6
9.5154
10.006
9.979
Compared with the target compounds 3–8, the corre-
sponding 1-chlorothioacridone precursors 12–17 exhib-
ited decreased cytotoxic activity (Table 3). These
results suggest that a 2-(dimethylamino)ethylamino side
chain at the 1-position of the thioacridone structure is
essential for high selectivity in this series of compounds.
The NH(CH)2N(CH3)2 side chain seems to confer high-
er activity than does the nitrogen mustard moiety (9–
11). This finding suggests that there may be differences
in the mechanism of interaction of these compounds
with DNA or that these compounds may interact with
additional, or alternative, biochemical targets.
7
8
9
10.129
10.620
10.593
7.414
10
11
12
13
14
15
16
17
7.905
7.878
7.878
8.369
8.341
Abbreviations are: polar surface area (PSA) and calculated molar
refractivity (CMR).
No direct correlation could be found between DNA
binding and cytotoxicity. For example, based on IC50
values, compound 6 (IC50 = 2.3lg/mL) displayed the
highest toxicity of all compounds tested, and also dis-
played the highest binding affinity (K = 2.6 · 104). This
might suggest that DNA binding is a factor in the cyto-
toxic activity of compound 6. In contrast, compound 8,
which binds more strongly to DNA than compounds 4,
5, and 7, has much less cytotoxic potency than the latter
compounds. It would appear that affinity for DNA is
not the sole determinant of cytotoxic potency in this ser-
ies of compounds. It is possible that there exists a dual
mechanism whereby DNA binding contributes as one
mechanism only. Further study is needed to determine
more precisely the exact mechanism of action of these
compounds. SAR for cytotoxicity observed indicated
that the presence of a 2-(dimethylamino)ethylamino side
chain at the 1-position of the thioacridone structure sig-
nificantly increases the cytotoxic activity of the basic
intercalation system (compounds 3–8). The nitrogen
mustard moiety is less effective in this respect (com-
pounds 12–14). Substitution at the 4-position with either
chloro or methyl decreases cytotoxicity of the 1-amino-
thioacridone compounds 3–11.
polar surface area (PSA), logP, and the calculated mo-
lar refractivity (CMR). The results are given as Eq. 1
and 2 (above) as well as in the graphs represented in Fig-
ures 3 and 4. Multiple linear regression analysis was
needed, since no linear correlation could be found for
one single descriptor, and biological activity. In the case
of DNA binding, Eq. 1 suggests that logP, PSA, and
CMR affect the antitumor activity. A resultant r2 of
5
4
3
2
4
0
-1
1
2
3
4
0
5
-1
Experimental log (1/k)
Figure 3. Correlation between experimental and predicted potency.
3.4. QSAR analysis
logð1=KÞ ¼ 0:0331 PSA ꢀ 0:331 log P
þ 1:051 CMR ꢀ 8:89
30
20
10
0
R2 ¼ 0:847 SE ¼ 0:665
ð1Þ
ð2Þ
IC50 ¼ 0:454 log P þ 8:943 PSA ꢀ 2:741 CMR
ꢀ 23:692
R2 ¼ 0:575 SE ¼ 5:9106
0
10
Experimental IC50
30
20
A classical QSAR study was performed using our data.
Both the DNA binding and cytotoxic activities were cor-
related with molecular descriptors (Table 4), including
Figure 4. Correlation between experimental and predicted cyto-
toxicity.

694
J. P. Dheyongera et al. / Bioorg. Med. Chem. 13 (2005) 689–698
0.847 was found for Eq. 1. The negative sign of logP
suggests that the more hydrophilic compounds bind bet-
ter to DNA. This finding correlates with the notion that
a common characteristic of intercalative drugs is that
they need at least one basic side chain to increase
DNA binding affinity.7 Eq. 2 suggests that logP, PSA,
and CMR affect the cytotoxicity of these compounds.
The resultant r2 was found to be 0.575. This equation
predicts that a more negative CMR will influence cyto-
toxic activity in this group of compounds.
4. Conclusions
The current study was performed to evaluate a new ser-
ies of 1-aminothioacridone compounds for cytotoxic
activity. Our results demonstrate that these compounds
possess cytotoxic activity in vitro. However, no clear
relationship between cytotoxicity and DNA binding
could be established, suggesting that DNA might not
be the sole molecular target for the compounds in this
series. Of all the compounds tested, the dimethylamino-
ethylamino-thioacridone compound 6 exhibited the
most promising biological activity and may be useful
as a lead compound in the search for more potent anti-
cancer agents.
3.5. Docking studies
Docking studies were also performed to investigate the
DNA binding differences of the most potent (6), and
the least potent (15) compounds. The results are shown
in Figure 5. The planar ring system of compound 6 fits
comfortably between the DNA base pairs (Fig. 5A,B),
with the basic side chain oriented in a favorable position
to undergo hydrogen bonding. In contrast, compound
15 can still undergo intercalation, but the methyl group
may result in steric hindering (Fig. 5C,D) and conse-
quently a decreased binding potential. In addition, the
absence of the basic side chain in compound 15, when
compared with 6, may also explain the lower potency
of this compound. The relative binding energies calcu-
lated in FLEXIDOCKꢁ for the two compounds were
found to be ꢀ375 for compound 6 and ꢀ75 for 15, cor-
relating well with the experimental DNA binding values
reported in Table 3. The more negative the relative bind-
ing energy, the more potent the binding is between lig-
and and target.
5. Experimental
5.1. Spectral shift determinations
Solutions of the test compounds at specified concentra-
tions were prepared in a 7.5mM phosphate buffer,
pH7.0, containing 1mM ethylenediaminetetraacetic
acid (EDTA). For compounds 3–11, compounds were
first dissolved in the least amount of ethanol, and then
2–5 drops of 1M HCl were added followed by buffer.
In the case of 12–17, 2–5 drops of 1M NaOH were
added to the ethanolic solution before the addition of
buffer. The DNA solution was prepared by dissolving
salmon testis DNA in the buffer to give a solution of
1mg/mL. Absorption spectrum of the test solution in
the absence of DNA was recorded over a wavelength
Figure 5. Docking studies of the most potent DNA binding compound, 6 (A,B), and least potent, 15 (C,D), into DNA base pairs (AT). The
thioacridone compounds are shown in green, to distinguish them from the DNA base pairs.

J. P. Dheyongera et al. / Bioorg. Med. Chem. 13 (2005) 689–698
695
range of 600–300nm. This gave the absorbance of
totally free ligand A1. Spectrophotometric titration
was then carried out by sequential addition of aliquots
of the DNA solution to 3mL of the test compound in
buffer at room temperature. The absorbance A of the
mixture was recorded after each addition. The titration
was continued until no further decrease in absorption
could be observed. This value was taken as the absorb-
ance of totally bound ligand A2.
were first dissolved in the least amount of DMSO, and
then 2–5 drops of 1M HCl were added followed by
the buffer. In the case of 12–17 2–5 drops of 1M NaOH
were added to the DMSO solution before the addition
of the buffer. To determine their cytotoxic activity, com-
pounds were added to the cells in microplates and
incubated for 4 days in the humidified incubator.
Cytotoxicity was assessed by determining the cell num-
bers with a Coulterꢁ counter, and their viability by
microscopic examination for exclusion of 0.1% trypan
blue. The IC50 for each compound was determined to
be the drug concentration that produced 50% growth
inhibition. Each experiment was done in duplicate.
5.2. DNA binding constant determination
The fraction of ligand bound to DNA (a) after each
addition of DNA during the titration was calculated
from the expression a = A1ꢀA/A1ꢀA2.16 If a is known,
the values c (molar concentration of free ligand) and r
(molar concentration of bound ligand/DNAP) can then
be calculated. DNAP refers to the molar concentration
of DNA expressed in terms of phosphate. A Scatchard
plot (r/c = KnꢀKr) was then constructed for each com-
pound, where K is the binding constant for formation
of the bound ligand/DNA complex, and n is the number
of binding sites/DNAP. The slope of the line on the
Scatchard plot is ꢀK. Acriflavine, a known DNA inter-
calator, was used as a standard under identical experi-
mental conditions.
5.6. Chemistry
All melting points were determined on a Buchi 510 melt-
¨
1
ing point apparatus and are uncorrected. The H and
13C spectra were recorded on a Varian VXR 300 spec-
trometer. Samples were dissolved in a deuterated solvent
(either DMSO-d6 or CDCl3, as indicated) with tetra-
methylsilane (Me4Si) as an internal standard. Infrared
spectra were recorded as a neat film on KBr plates with
a Shimadzu FT IR 4200 spectrophotometer. Mass spec-
tra and HR-MS were recorded at 70eV (EI), or in FAB
mode, on a VG 7070E spectrometer. Elemental analyses
were performed on a Perkin–Elmer model 240 instru-
ment at the Department of Chemistry, Cape Town
University and the Council for Scientific and Industrial
Research (CSIR), Pretoria, South Africa, and data were
all within acceptable limits ( 0.04%). Reactions were
monitored by thin layer chromatography on Merck Sil-
ica gel 60 F254 precoated aluminum plates, using a
CHCl3/CH3OH (4:1) mobile phase system. Spots were
visualized under UV or by iodine vapor.
5.3. Ethidium displacement determination
Ethidium displacement determination was carried out
according to Cain et al.13 Briefly, ethidium bromide
was added to 3mL buffer solution (2mM HEPES,
8mM NaCl, 0.05mM EDTA, pH7.0) to obtain a final
concentration of 2lM. The optimal amount of DNA
was determined by slowly adding DNA (calf thymus)
to the solution to obtain near saturation of ethidium
bromide with DNA as determined by the enhancement
of fluorescence. The test compound was then added in
1lL portions and the fluorescence recorded after each
addition. The temperature was maintained at 25ꢂC dur-
ing the assays. Titrations were performed in duplicate
and the C50 values (representing the concentration of
compound needed to decrease the DNA-bound ethi-
dium fluorescence by 50%) were determined from a plot
of fluorescence (% max) against added test compound
concentration (lM).
5.6.1. 1-(2-Dimethylaminoethylamino)-10-methyl-9(10H)-
thioacridone (3). A mixture of 1-chloro-10-methylthioac-
ridone (15) (500mg, 1.9mmol), anhydrous potassium
carbonate (150g, 1.1mmol), and 2-dimethylaminoethyl-
amine (2mL, 18mmol) was stirred for 30min and then
left at room temperature for 24h. Diethyl ether
(50mL) was added and the product extracted with 1M
HCl (3 · 100mL). The combined aqueous layer was fil-
tered and the solution made basic with 1M NaOH. The
product was then extracted with diethyl ether
(3 · 50mL). The combined organic layer was dried over
anhydrous calcium chloride, and filtered. Removal of
the solvent by evaporation gave 220mg (36% yield) of
3 as a viscous yellow liquid with darkens on exposure
5.4. Cell culture
HL-60 cells were routinely maintained as exponentially
proliferating suspension cultures in RPMI-1640 medium
supplemented with 10% fetal calf serum (FCS) contain-
ing penicillin (100IU/mL) and streptomycin (100lg/
mL). Incubations were carried out at 37ꢂC in a humid-
ified 5% CO2 incubator.
to light. IR: mmax = 1569, 1219, 1069, 752, 720cmꢀ1
;
FAB-MS: m/z = 314 (M+3H)+, 276, 269, 255, 241,
233, 228, 219; 1H NMR (CDCl3): d = 10.46, 8.47,
7.67–7.18, 6.48, 6.33, 3.38, 2.71, 2.36; 13C NMR
(CDCl3): d = 152.62, 145.19, 141.87, 134.98, 132.98,
127.09, 123.01, 120.73, 114.10, 108.19, 100.75, 99.17,
58.10, 45.65, 41.15, 34.30. Elemental analysis, calcd for
C18H21N3S: C, 69.42; H, 6.80; N, 13.49. Found: C,
69.47; H, 6.92; N, 13.52.
5.5. Cytotoxicity
Cytotoxicity studies were carried out in 5mL liquid cul-
tures established at 2 · 105 cells/mL. The compounds
tested were dissolved in DMSO and the same volume
of DMSO was used as control, not exceeding 0.2% for
all the treatments. For compounds 3–11, compounds
Compounds 4 and 5 were obtained in an analogous
manner from 16 (500mg, 1.7mmol) and 17 (500mg,
8mmol), respectively.

696
J. P. Dheyongera et al. / Bioorg. Med. Chem. 13 (2005) 689–698
5.6.2. 1-(2-Dimethylaminoethylamino)-4-chloro-10-meth-
yl-9(10H)-thioacridone (4). Compound 4 was isolated as
a yellow viscous liquid, which darkens on exposure to
air; yield 32%; IR: mmax = 1565, 1240, 1195, 1066, 750,
m
max = 3308, 1600, 1551, 1505, 1460, 1325cmꢀ1; FAB-
MS: m/z = 314 (M+3H)+, 276, 269, 255, 241, 233, 228,
1
219, 205; H NMR (CDCl3): d = 8.13, 8.04, 7.79, 7.64,
7.36–7.22, 3.75, 2.80, 2.50, 2.32; 13C NMR (CDCl ):
3
1
723cmꢀ1; MS: m/z = 348 (M+3H)+, 265, 172, 159; H
d = 14.09, 149.48, 136.69, 130.04, 129.48, 128.07,
125.60, 125.27, 125.06, 121.78, 116.62, 113.74, 59.05,
49.05, 44.97, 19.26. Elemental analysis, calcd for
C18H21N3S: C, 69.42; H, 6.80; N, 13.49. Found: C,
69.57; H, 6.82; N, 13.52.
NMR (CDCl3): d = 10.35, 8.30, 7.61, 7.45–7.19, 6.33,
3.90, 3.35, 2.68, 2.35; 13C NMR (CDCl3): d = 150.96,
145.32, 144.85, 137.76, 133.13, 126.38, 123.82, 121.55,
116.12, 111.31, 105.55, 103.14, 57.97, 45.41, 43.97,
41.15. Elemental analysis, calcd for C18H20ClN3S: C,
62.61; H, 5.80; N, 12.17. Found: C, 62.52; H, 5.97; N,
12.23.
5.6.7. 1-Bis(2-chloroethyl)-amino-9(10H)-thioacridone (9).
Thionyl chloride (5mL, 68.5mmol) was carefully
added to 1-bis(2-hydroxyethyl)-amino-9(10H)-thioacri-
done (23, R@H, 200mg, 0.637mmol) in an ice-cooled
flask. The mixture was allowed to stir overnight at room
temperature and then poured into anhydrous diethyl
ether (25mL). An orange precipitate formed. The sol-
vent was distilled off, more ether was added, and the
product filtered. The product was washed several times
with ether to remove any residual thionyl chloride.
The product was then dissolved in hot absolute ethanol
and precipitated with ether. Filtration and vacuum des-
iccation afforded 170mg (60%) of the product as a yel-
low orange powder; mp 237ꢂC; IR: mmax = 3409, 1607,
1559, 1516, 1456, 1262cmꢀ1; EI-MS: m/z = 350 (M+),
265, 172, 159, 119, 105, 91; 1H NMR (CDCl3):
d = 12.22, 8.05–7.92, 7.64–7.54, 7.28–7.12, 3.78, 3.62;
13C NMR (CDCl3): d = 150.96, 145.32, 144.85, 137.76,
133.19, 126.38, 123.82, 121.55, 116.12, 111.31, 105.55,
103.14, 55.61, 41.15. Elemental analysis, calcd for
C17H16Cl2N2S: C, 58.29; H, 4.57; N, 8.00. Found: C,
57.96; H, 4.92; N, 8.32.
5.6.3.
1-(2-Dimethylaminoethylamino)-4,10-dimethyl-
9(10H)-thioacridone (5). Compound 5 was isolated as a
pale yellow viscous liquid, which darkens on exposure
to air; yield 30%; IR: mmax = 1597, 1306, 1250, 1167,
1075, 753, 723cmꢀ1; MS: m/z = 328 (M+3H)+, 311,
1
282, 268, 236, 222; H NMR (CDCl3): d = 10.15, 8.34,
7.62, 7.37–7.17, 6.35, 3.77, 3.37, 2.70, 2.46, 2.36; 13C
NMR (CDCl3): d = 150.38, 147.80, 145.60, 139.29,
126.46, 123.81, 120.94, 115.81, 110.71, 109.84, 102.19,
58.15, 45.64, 43.48, 41.15, 22.04. Elemental analysis,
calcd for C19H23N2S: C, 70.15; H, 7.08; N, 12.92.
Found: C, 70.16; H, 7.12; N, 12.87.
5.6.4. 1-(2-Dimethylaminoethylamino)-9(10H)-thioacri-
done (6). A mixture of 1-chloro-9(10H)-thioacridone
(12) (100mg, 0.41mmol), anhydrous potassium carbon-
ate (30mg, 0.21mmol), and 2-dimethylaminoethylamine
(2mL, 18mmol) was magnetically stirred for 15min and
then left at room temperature for 10h (6h in the case of
7). Diethyl ether (20mL) was added and the product ex-
tracted with 1M HCl (3 · 50mL). The combined ex-
tracts were filtered and excess ice cold 1M NaOH
(200mL) was added to the filtrate with shaking. The pre-
cipitate was collected by filtration, washed with water,
and air dried to give 6 as a yellow fine powder in a yield
of 50mg (41%); mp >360ꢂC; IR: mmax = 3409, 1607,
1559, 1516, 1456, 1262cmꢀ1; FAB-MS: m/z = 300
(M+3H)+, 255, 241, 229, 214; 1H NMR (CDCl3):
d = 8.11–7.99, 7.65, 7.34, 7.22, 3.83, 2.59; 13C NMR
(CDCl3): d = 152.01, 149.96, 135.96, 130.47, 129.50,
127.94, 124.93, 123.61, 123.20, 122.80, 116.54, 114.62,
58.49, 46.64, 44.84. Elemental analysis, calcd for
C17H19N3S: C, 68.65; H, 6.44; N, 14.13. Found: C,
68.43; H, 6.59; N, 14.18.
Compounds10 and 11 were obtained in an analogous
manner from their corresponding precursors 23.
5.6.8. 1-Bis(2-chloroethyl)-amino-4-chloro-9(10H)-thio-
acridone (10). Compound 10 was isolated as an orange
powder; yield 85%; mp 222ꢂC; IR: mmax = 3422, 1547,
1507, 1408, 1329cmꢀ1; MS: m/z = 384 (M+), 357, 292,
1
264, 188, 159, 131; H NMR (CDCl3): d = 12.20, 8.53,
8.32, 7.73, 7.47–7.34, 3.91, 3.76; 13C NMR (CDCl3):
d = 152.62, 145.19, 141.87, 134.98, 134.98, 132.98,
127.09, 123.01, 114.10, 108.19, 100.75, 99.17, 55.56,
41.15. Elemental analysis, calcd for C17H15Cl3N2S: C,
53.13; H, 3.91; N, 7.29. Found: C, 53.22, H, 4.03; N,
7.30.
5.6.5. 1-(2-Dimethylaminoethylamino)-4-chloro-9(10H)-
thioacridone (7). Compound 7 was isolated as a fine yel-
low powder; yield 52mg (44%); mp 80–82ꢂC; IR:
5.6.9. 1-Bis(2-chloroethyl)-amino-4-methyl-9(10H)-thio-
acridone (11). Compound 11 was isolated as an orange
powder; yield 85%; mp 254 ꢂC; IR: mmax = 3308, 1600,
1551, 1505, 1460, 1325cmꢀ1; MS: m/z = 364 (M+), 292,
264, 188, 159; 1H NMR (CDCl3): 12.28, 8.54, 8.0,
7.82, 7.57–7.37, 6.50, 3.88, 3.68; 13C NMR (CDCl3):
d = 150.38, 147.80, 145.60, 139.29, 132.85, 126.46,
123.81, 120.94, 115.81, 110.71, 109.84, 102.19, 55.60,
m
max = 3422, 1547, 1507, 1408, 1329cmꢀ1; FAB-MS: m/
z = 334 (M+3H)+, 298, 289, 275, 263, 253, 248; 1H
NMR (CDCl3): d = 8.05–7.92, 7.64–7.54, 7.28–7.12,
3.82, 2.56, 2.35; 13C NMR (CDCl3): d = 154.55,
132.51, 130.80, 129.58, 128.25, 127.11, 125.31, 125.24,
125.17, 123.98, 122.02, 116.14, 58.71, 48.62, 44.84. Ele-
mental analysis, calcd for C17H18ClN3S: C, 61.53; H,
5.47; N, 12.66. Found: C, 61.68; H, 5.32; N, 12.62.
41.15, 18.80.
Elemental
analysis, calcd
for
C18H18Cl2N2S: C, 59.34; H, 4.95; N, 7.69. Found: C,
59.34; H, 5.03; N, 7.73.
5.6.6. 1-(2-Dimethylaminoethylamino)-4-methyl-9(10H)-
thioacridone (8). Compound 8 was isolated as a yellow
powder; yield 48mg (40%); mp 312–314ꢂC; IR:
5.6.10. 1-Chloro-9(10H)-thioacridone (12). NaHS (24g,
0.43mol) and 96% ethanol (500mL) was added to 1,9-
dichloroacridine (22, R@H, 10g, 0.04mol). The mixture

J. P. Dheyongera et al. / Bioorg. Med. Chem. 13 (2005) 689–698
697
was heated under reflux for 30min, filtered whilst warm,
and acidified with dilute HCl until complete precipita-
tion of the product was effected. The crude product
(8g, 81%) was recrystallized from pyridine/water to give
light brown crystals (mp 210–212ꢂC); IR: mmax = 3438,
3260, 1617, 1578, 1466, 1210cmꢀ1; MS: m/z = 245
5.6.14. 1,4-Dichloro-10-methylthioacridone (16). Com-
pound 16 was synthesized by mixing, at room tempera-
ture, 1,4-dichloro-10-methylacridone (20, R@Cl, 2.5g,
0.009mol), phosphoryl chloride (15mL, 0.175), and
150mL of a 1M sodium thiosulfate solution. Yield
70%; mp 125–126ꢂC; IR: mmax = 1570, 1483, 1218,
758cmꢀ1; MS: m/z = 277, 258, 242, 164; 1H NMR
(CDCl3): d = 8.60, 7.71–7.23, 3.87; 13C NMR (CDCl3):
d = 205.26, 140.17, 137.00, 136.18, 134.71, 132.90,
131.26, 129.48, 126.48, 126.37, 122.90, 114.08, 113.55,
96.98, 36.08. Elemental analysis, calcd for C14H9C12NS:
C, 57.16; H, 3.08; N, 4.76. Found: C, 57.17; H, 3.12; N,
4.62.
1
(M+), 210, 201, 177, 166, 140, 105; H NMR (DMSO-
d6): d = 13.62, 9.38–9.28, 7.64–7.48, 7.37–7.26; 13C
NMR (DMSO-d6): d = 199.21, 139.42, 137.18, 136.31,
135.51, 133.61, 132.60, 130.26, 128.80, 118.02, 116.50.
Elemental analysis, calcd for C13H8ClNS: C, 63.54; H,
3.28; N, 5.70. Found: C, 63.75; H, 3.19; N, 5.72.
Compounds 13 and 14 were obtained according to the
same procedure from their corresponding 1,9-dichloro-
acridines 22.
5.6.15. 1-Chloro-4,10-dimethylthioacridone (17). Com-
pound 17 was synthesized by mixing, at room tempera-
ture, 1-chloro-4,10-dimethylacridone (20, R@Me, 25g,
0.01mol), phosphoryl chloride (10mL, 0.11mol), and
150mL of 1M sodium thiosulfate solution. Yield 75%;
5.6.11. 1,4-Dichloro-9(10H)-thioacridone (13). Com-
pound 13 was isolated as light brown crystals: yield
86%: mp 198–200ꢂC (from xylene); IR: mmax = 3439,
3296, 2924, 2359, 2359, 1607, 1474, 1192cmꢀ1; MS: m/
z = 279 (M+), 263, 244, 209, 200, 164; 1H NMR
(DMSO-d6): d = 11.30, 8.44, 4.40, 8.17, 7.74, 7.31; 13C
NMR (DMSO-d6): d = 202.11, 135.40, 134.18, 133.97,
133.57, 132.73, 131.90, 128.75, 127.27, 126.06, 124.11,
120.20, 118.50. Elemental analysis, calcd for
C13H7Cl2NS: C, 55.73; H, 2.52; N, 5.00. Found: C,
55.77; H, 2.51; N, 5.01.
mp 120–122ꢂC; IR: mmax = 1566, 1469, 1240, 751 cmꢀ1
;
1
MS: m/z = 274 (M+H), 238, 224, 154, 136; H NMR
(DMSO-d6): d = 8.33, 7.61; 7.37–7.17, 3.79, 2.65; 13C
NMR (DMSO-d6): d = 208.17, 142.59, 139.51, 139.38,
135.02, 133.41, 132.46, 130.77, 128.26, 127.16, 126.02,
122.87, 115.50, 43.73, 22.68. Elemental analysis, calcd
for C15H12ClNS: C, 65.81; H, 4.42; N, 5.12. Found: C,
65.84; H, 4.31; N, 5.16.
5.7. Molecular modeling
5.6.12. 1-Chloro-4-methyl-9(10H)-thioacridone (14).
Compound 14 was isolated as dark brown crystals; yield
91%; mp 228–230ꢂC (from xylene); IR: mmax = 3316,
2922, 1611, 1514, 1226, 754cmꢀ1; MS: m/z = 259
Molecular modeling studies were either performed on ꢁa
Silicon Graphics Octaneꢁ computer using SYBYL
6.9.117 software, orꢁon a Dell Dimenꢁsionꢁ P4 desktop
running ChemDraw 8.18 With SYBYL , the compounds
were drawn using SKETCHꢁ and energy minimized
(100 iterations) using the Triposꢁ force field with atomic
1
(M+), 243, 225, 180; H NMR: d = 10.87, 8.51, 7.92,
7.69, 7.43, 7.26, 3.32; 13C NMR: d = 201.65, 137.32,
134.35, 133.35, 132.94, 132.55, 131.21, 128.77, 126.46,
125.87, 123.34, 118.17, 17.92. Elemental analysis, calcd
for C14H10ClNS: C, 64.74; H, 3.88; N, 5.39. Found: C,
65.04; H, 3.91; N, 5.35.
charges added using the Gasteiger–Huckel method.
¨
ꢁ
SYBYL was used to calculate polar surface area (PSA)
and perform the docking studies with FLEXIDOCKꢁ.
The DNA molecule was built using the BIOPOLY-
MER/BUILDꢁ module and energy minimized
using the AMBER7 force field. ChemDrawꢁ 8 was
used to calculate logP as well as the molecular
refractivity.
5.6.13. 1-Chloro-10-methylthioacridone (15). A mixture
of 1-chloro-10-methylacridone (20, R@H, 2.5g,
0.01mol) and phosphoryl chloride (10mL, 0.11mol)
was heated under reflux for 1h. The solution that
formed was then cooled with ice after which 150mL of
a 1M sodium thiosulfate solution was added. A brown-
ish sticky material formed almost immediately. The
material was stirred vigorously for about 5min. Dilute
ammonia solution (50mL) was added and stirring con-
tinued for another 5min. The granular precipitate that
formed was then filtered off, dried, and recrystallized
twice from ethanol to give a light brown powder,
Acknowledgements
This study was funded in part by the National Research
Foundation (South Africa). Opinions expressed and
conclusions arrived at, are those of the authors and
not necessarily to be attributed to the NRF.
mp 114–116ꢂC; IR: mmax = 1570, 1480, 1215, 755cmꢀ1
.
MS: m/z = 259 (M+), 244, 215, 201, 168; 1H NMR
(CDCl3): d = 8.60, 7.71–7.22, 3.65; 13C NMR (CDCl3):
d = 204.99, 140.05, 136.83, 134.58, 132.87, 131.22,
129.39, 126.30, 122.84, 114.10, 113.58, 36.03. Elemental
analysis, calculated for C14H10ClNS: C, 64.74; H, 3.88;
N, 5.39. Found: C, 64.71; H, 3.92; N, 5.40.
References and notes
1. Hughes, G. K.; Lahey, F. N.; Price, J. R.; Webb, L. J.
Nature 1948, 162, 223–224.
2. Svoboda, G. H.; Poore, G. A.; Simpson, P. J.; Border, G.
B. J. Pharm. Sci. 1966, 55, 758–768.
Compounds 16 and 17 were obtained in an analogous
manner from their corresponding 1-chloro-methylacri-
dones 20.
3. Dorr, R. T.; Liddil, J. D.; Von Hoff, D. D.; Soble, M.;
Osborne, C. K. Cancer Res. 1989, 49, 340–344.
4. Tillequin, F. Ann. Pharm. Fr. 2002, 60, 246–252.

698
J. P. Dheyongera et al. / Bioorg. Med. Chem. 13 (2005) 689–698
5. Schneider, J.; Evans, E. L.; Grunberg, E.; Fryer, R. I. J.
Med. Chem. 1972, 15, 266–270.
11. Lehmstedt, K.; Schrader, K. Ber. Dtsch. Chem. Ges. 1937,
70, 838–849.
6. Costes, N.; Elomri, A.; Dufat, H.; Michel, S.; Seguin, E.;
Koch, M.; Tillequin, F.; Pfeiffer, B.; Renard, P.; Leonce,
S.; Pierre, A. Oncol. Res. 2003, 13, 191–197.
12. Nisbet, H. B. J. Chem. Soc. 1933, 1372–1373.
13. Cain, B. F.; Baguley, B. C.; Denny, W. A. J. Med. Chem.
1978, 21, 656–668.
7. Miri, R.; Javidnia, K.; Hemmateenejad, B.; Azarpira, A.;
Amirghofran, Z. Bioorg. Med. Chem. 2004, 12, 2529–2536.
8. Pindur, U.; Haber, M.; Sattler, K. J. Chem. Edu. 1993,
263–272.
9. Elomri, A.; Mitaku, S.; Michel, S.; Skaltsounis, A. L.;
Tillequin, F.; Koch, M.; Pierre, A.; Guilbaud, N.; Leonce,
S.; Kraus-Berthier, L.; Rolland, Y.; Atassi, G. J. Med.
Chem. 1996, 39, 4762–4766.
14. Su, T. L.; Chou, T. C.; Kim, J. Y.; Huang, J. T.;
Ciszewska, G.; Ren, W. Y.; Otter, G. M.; Sirotnak, F. M.;
Watanabe, K. A. J. Med. Chem. 1995, 38, 3226–3235.
15. Waring, M. J. Eur. J. Cancer 1976, 12, 995–1001.
16. Double, J. C.; Brown, J. R. J. Pharm. Pharmacol. 1975,
27, 502–507.
ꢁ
17. SYBYL 6.9.1 Tripos Inc., 1699 South Hanley Rd., St.
Louis, Missouri, 63144, USA.
10. Smolders, R. R.; Hanuise, J.; Coomans, R.; Proietto, V.;
Voglet, N.; Waefelaer, A. Synthesis 1982, 493–494.
18. ChemDrawꢁ 8 CambridgeSoft Corporation, 100 Cam-
bridge Park Drive, Cambridge, MA 02140, USA.