Cytotoxic 1,3-diarylidene-2-tetralones and related compounds

Article abstract of DOI:10.1016/S0223-5234(02)01402-2

A number of 1,3-arylidene-2-tetralones 1, 2 and 4 were synthesised and demonstrated cytotoxic activity towards murine P388 and L1210 cells as well as human Molt 4/C8 and CEM T-lymphocytes. In general, the related 1-arylidene-2-tetralones 3 possessed lower potencies in these screens than the compounds in series 1 and 4. Approximately, half of the compounds were evaluated against a panel of human tumour cell lines. In this screen, most of the enones were more cytotoxic than the established anticancer agent melphalan and some demonstrated selective toxicity towards leukemic and colon cancer cells. The modes of action of representative compounds include interfering with the biosyntheses of nucleic acids and proteins as well as altering redox potentials. The compounds were well tolerated when administered intraperiteonally to mice. Thus these novel enones are promising prototypic molecules due to their potent cytotoxic properties and lack of significant murine toxicity.

Full text of DOI:10.1016/S0223-5234(02)01402-2

European Journal of Medicinal Chemistry 37 (2002) 813Á824
/
www.elsevier.com/locate/ejmech
Original article
Cytotoxic 1,3-diarylidene-2-tetralones and related compounds
Jonathan R. Dimmock a,*, Maniyan P. Padmanilyam a, Gordon A. Zello a, J. Wilson
Quail b, Eliud O. Oloo b, Jared S. Prisciak b, Heinz-Bernhard Kraatz b,
Arten Cherkasov c, Jeremy S. Lee c, Theresa M. Allen d, Cheryl L. Santos d, Elias
K. Manavathu e, Erik De Clercq f, Jan Balzarini f, James P. Stables g
a College of Pharmacy and Nutrition, University of Saskatchewan, 110 Science Place, Saskatoon, Saskatchewan, Canada S7N 5C9
b Department of Chemistry, University of Saskatchewan, Saskatoon, Saskatchewan, Canada S7N 5C9
c Department of Biochemistry, University of Saskatchewan, Saskatoon, Saskatchewan, Canada S7N 5E5
d Department of Pharmacology, University of Alberta, Edmonton, Alberta, Canada T6G 2H7
e Department of Medicine, Wayne State University, Detroit, MI 48201-1908, USA
f Rega Institute for Medical Research, Katholieke Univesiteit Leuven, B-3000 Leuven, Belgium
g National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD 20892-9020, USA
Received 19 March 2002; received in revised form 11 July 2002; accepted 11 July 2002
Abstract
A number of 1,3-arylidene-2-tetralones 1, 2 and 4 were synthesised and demonstrated cytotoxic activity towards murine P388 and
L1210 cells as well as human Molt 4/C8 and CEM T-lymphocytes. In general, the related 1-arylidene-2-tetralones 3 possessed lower
potencies in these screens than the compounds in series 1 and 4. Approximately, half of the compounds were evaluated against a
panel of human tumour cell lines. In this screen, most of the enones were more cytotoxic than the established anticancer agent
melphalan and some demonstrated selective toxicity towards leukemic and colon cancer cells. The modes of action of representative
compounds include interfering with the biosyntheses of nucleic acids and proteins as well as altering redox potentials. The
compounds were well tolerated when administered intraperiteonally to mice. Thus these novel enones are promising prototypic
molecules due to their potent cytotoxic properties and lack of significant murine toxicity.
´
# 2002 Published by Editions scientifiques et me´dicales Elsevier SAS.
Keywords: 2-Tetralones; Cytotoxicity; Murine toxicity; Redox potentials
1. Introduction
structural motif into the compounds under considera-
tion was planned. Such observations led to the decision
to incorporate the 1,5-diaryl-1,4-dien-3-oxo and stilbene
moieties into series 1 and 2 (Figs. 1 and 2).
Previous studies have revealed the promising cyto-
toxic properties of various 2,6-bis(arylidene)cyclohexa-
nones [1] and 3,5-bis(arylidene)-4-piperidones [2]. Both
groups of compounds contain the 1,5-diaryl-1,4-dien-3-
oxo group. These compounds were designed as thiol
alkylators, since a,b-unsaturated ketones have an affi-
nity for thiols in contrast to amino and hydroxy groups
[3] which are found in nucleic acids. Hence enones may
lack genotoxic properties. In addition, a number of
cytotoxics such as tamoxifen [4] and various combre-
tastatins [5] are stilbenes and the incorporation of this
There was however a further reason for preparing
series 1. A number of studies have shown that lowering
the concentrations of chemoprotectant thiols in cells
prior to the administration of anticancer drugs is more
deleterious to cancer cells than normal tissue [6,7]. The
compounds in series 1 and 2 have two olefinic centres in
different electronic and steric environments permitting
the sequential attack of thiols. In other words, an initial
thiol alkylation could occur followed by a subsequent
interaction with thiols at the second methine group of 1
and 2. If such a process is important in eliciting
cytotoxic activity, the monoarylidene compounds 3
should have significantly less than half of the potencies
* Correspondence and reprints
E-mail address: dimmock@skyway.usask.ca (J.R. Dimmock).
´
0223-5234/02/$ - see front matter # 2002 Published by Editions scientifiques et me´dicales Elsevier SAS.
PII: S 0 2 2 3 - 5 2 3 4 ( 0 2 ) 0 1 4 0 2 - 2

814
J.R. Dimmock et al. / European Journal of Medicinal Chemistry 37 (2002) 813Á824
/
into rings B and C in series 1. The bioevaluations of the
tetralones 1aÁc,g,j permit a Topliss analysis [9] which
/
may indicate the structure of further analogues with
increased potencies. Compound 1k possesses a 3,4,5-
trimethoxyphenyl group which is present in various
combretastatins [5] and some naturally occurring anti-
neoplastic agents such as podophyllotoxin [10]. In the
case of the 3,5-bis(arylidene)-4-piperidones, the highest
potencies were found in the compound possessing a 4-
nitro substituent [2] which may have been due to
reduction to one or more toxic species [11]. Conse-
quently, in 1pÁr, a nitro group was placed in different
/
positions of the aryl rings. Various nitrophenols cause
uncoupling of oxidative phosphorylation [12] and hence
the aryl substitution pattern of 1s was suggested.
Compound 2 was designed in order to evaluate the
effect on potency of attaching groups at the ortho and
meta positions of rings B and C of 1a. The choice of
substituents in ring B of series 3 was made for similar
reasons as those indicated for the enones 1.
The evaluation of representative compounds for
anticonvulsant activity and neurotoxicity was also
planned. Compounds which penetrate the central ner-
vous system (CNS) could be of value in treating tumours
of the CNS; conversely, such compounds may have
unwanted side effects.
In summary, the aims of the present investigation
were to prepare a series of bis and mono arylidene 2-
tetralones as candidate antineoplastic agents coupled to
an evaluation of their murine toxicity.
Fig. 1. The structures of the compounds in series 1Á4. The aryl
substitution patterns are indicated in Table 1.
/
of the analogues in series 1. Furthermore, in order to
exacerbate the magnitude of the differences between the
fractional positive charges on the olefinic carbon atoms,
series 4 was proposed in which the substituents in rings
B and C of each compound were divergent.
The choice of substituents in series 1 was predicated
by the following considerations. A variety of groups
were placed in the aryl rings with varying s and p
values; in fact, substituents in all four quadrants of a
Craig plot for para substituents [8] were employed. In
addition to a consideration of the s and p values of the
aryl substituents, groups of varying sizes were inserted
2. Chemistry
Acid catalysed condensation was undertaken between
the appropriate aryl aldehyde and either 2-tetralone
Fig. 2. The 1,5-diaryl-1,4-dien-3-oxo and stilbene portions of series 1 and 4.

J.R. Dimmock et al. / European Journal of Medicinal Chemistry 37 (2002) 813Á
/
824
815
leading to series 1 and 2 or a monoarylidene-2-tetralone
which gave rise to series 4. The enones 3 were
synthesised by the Knoevenagel method. 1H-NMR
spectroscopy revealed the stereohomogeneity of the
molecules. X-ray crystallography of 1b,c,g,j and 3h
indicated that the configuration of the olefinic bonds
was E and the assumption was made that this stereo-
chemistry was displayed in the analogues of these
compounds which is consistent with studies involving
2,6-bis(arylidene)cyclohexanones [1,13], a 2-arylidenecy-
clohexanone [14] and 3,5-bis(arylidene)-4-piperidones
[2]. The redox potentials of 1a,k,n,p,q,r,t, 3a,g and 4b
were determined; these data are portrayed in Table 4.
this overview is that the general structures 1 and 4 are
clearly promising novel molecular templates.
Some comments will be made regarding the potencies
of the 1,3-bis(arylidene)-2-tetralones 1 and 4 in relation
to the 2,6-bis(arylidene)cyclohexanones and 3,5-bis(ar-
ylidene)-4-piperidones mentioned earlier. In the first
place, this study has provided further evidence that the
1,5-diaryl-1,4-dien-3-oxo-group is an important toxo-
phore thereby providing a justification for its incorpora-
tion into candidate cytotoxic and anticancer agents.
Secondly, in the case of the 2,6-bis(arylidene)cyclohex-
anones, in general different substituents were placed in
each of the two aryl rings. Hence a comparison of the
IC50 values of these compounds with the enones 1 was
not possible. However, the average IC50 values of 13 2,6-
bis(arylidene)cyclohexanones in the four screens was
3. Bioevaluations
27.0 (4.29Á
/
283) mM while for the compounds 1aÁ
/
l,oÁt
/
All of the compounds in series 1Á4 were evaluated for
/
and 4aÁc the figure was 9.81 (1.17Á
/
/
47.0) mM. (The
cytotoxic activity against murine P388 and L1210
leukemic cells as well as human Molt 4/C8 and CEM
T-lymphocytes. These data are presented in Tables 1
and 2. Approximately, half of the compounds were
examined for their effect on a panel of human tumour
cell lines and the results obtained are portrayed in Table
7. The inhibitory effects of 1a,p,r, 3a,g on DNA, RNA
and protein biosyntheses in L1210 cells was undertaken
and the results are presented in Table 8. In addition,
1a,p, 3a,g were evaluated for DNA binding properties
while the antifungal properties of 1b, 3b and 3f were
assessed against Aspergillus fumigatus and Candida
albicans. Approximately, two-thirds of the compounds
were examined for anticonvulsant activity and neuro-
toxicity in mice.
esters 1m and 1n were omitted from consideration due to
a lack of precise IC50 figures in some of the screens.)
Hence one may tentatively draw the conclusion that the
presence of ring A in series 1 and 4 may contribute to
cytotoxicity by creating van der Waals’ bonds at a
binding site. The average IC50 of seven 3,5-bis(aryli-
dene)-4-piperidones was 60.1 (3.03Á
same aryl substitution was present in 1aÁ
which the median IC50 value was 17.9 (2.11Á
/
174) mM [2]. The
d,j,r,t for
47) mM.
/
/
The increase in potencies of the analogues in series 1
compared to the 4-piperidones may be due to the
presence of ring A in 1 as well as the absence of a basic
centre in the tetralones.
The next phase of the investigation was directed to
evaluating whether cytotoxicity was correlated with one
or more physicochemical properties of the aryl sub-
stituents in series 1 and 3. Linear and semilogarithmic
plots were obtained between the Hammett s and/or Taft
s* values, the Hansch p figures and molar refractivity
(MR) constants, which reflect the electronic, hydropho-
bic and steric properties of the aryl groups, respectively,
and the IC50 concentrations generated for each cell line.
Compounds 1m,n and 3f were omitted from this
evaluation since specific IC50 concentrations were not
available in each test. Negative correlations were
established between the s/s* constants and cytotoxicity
in the P388 screen for the compounds in series 1 when
4. Results and discussion
All of the compounds in series 1Á4 were assessed
/
against P388, L1210, Molt 4/C8 and CEM cells; these
data are portrayed in Tables 1 and 2. Approximately,
40Á50% of the compounds had one-fifth or more of the
/
potency of the established antineoplastic alkylating
agent melphalan. In particular, the following com-
pounds possessed equal or greater cytotoxicity than
melphalan (cell line in parentheses), namely 1q (P388),
1dÁ
/
f,h,k,p, 4aÁ
/c
(L1210), 1a,b,dÁ
/
f,h,i,k,l,o,p, 4aÁ
/c
both linear (P B
/
0.001) and semilogarithmic (P B0.001)
/
(Molt 4/C8), and 1a,b,dÁ
/
f,h,i,k,l,p, 4aÁ
/
c (CEM). These
plots were made. No other correlations were noted in
the other screens for both series 1 and 3, suggesting that,
in general, one single physicochemical parameter is
unlikely to contribute significantly to bioactivity.
During the investigation, the cytotoxicity of 2 was
determined. The data in Table 1 reveal that the potency
of this compound was substantially lower than 1a
suggesting that the attachment of bulky groups at the
ortho and meta positions of rings B and C was
comparisons revealed that a greater number of com-
pounds displayed equal or superior activity to melpha-
lan in the T-lymphocyte screens rather than towards
murine cell lines. In addition, the average IC50 figures
towards all four cell lines were computed. The most
potent compounds (average IC50 value in mM in
parentheses) were 1e (1.17), 1h (1.28), 1p (1.46), 4a
(2.01), 4b (1.53) and 4c (1.70) which compare favourably
with melphalan (2.02). The conclusion to be drawn from

816
J.R. Dimmock et al. / European Journal of Medicinal Chemistry 37 (2002) 813Á824
/
Table 1
Cytotoxicity of the compounds in series 1Á3 against murine P388 and L1210 cells as well as Molt 4/C8 and CEM T-lymphocytes
/
Compound
R1
R2
R3
R4
IC50 (mM)
P388
L1210
Molt 4/C8
CEM
1a
1b
1c
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
NO2
H
H
H
H
H
H
Cl
H
1.7990.2
6.2390.10
2.4990.24
41.491.3
4.4092.72
1.6790.15
3.8692.04
45.091.2
1.2690.40
7.8790.01
38.496.1
4.8893.21
8.5690.13
ꢀ500
5.3892.70
2.4590.62
47.090.9
1.7190.06
1.4490.05
3.7091.94
46.297.6
1.7090.03
1.6390.09
8.7691.69
1.6490.09
1.9890.01
ꢀ500
2.4490.59
2.3090.08
50.190.9
1.7490.18
1.3090.16
4.6793.04
40.094.0
1.6990.06
2.0090.06
9.5990.28
1.6590.03
3.1591.09
ꢀ500
H
H
1.2090.03
0.84990.05
1.0490.04
0.28190.03
0.54890.01
4.0090.2
Cl
H
Cl
F
H
1d
1e
H
F
H
Br
F
1f
H
H
1g
1h
1i
H
CH3
CF3
OH
H
H
H
0.47190.02
3.5590.34
3.8590.08
0.69390.03
4.5890.32
ꢀ5091.8
1.4890.23
0.73290.04
0.9890.06
0.18790.03
0.50890.04
1.0290.06
3.7990.25
ꢀ5091.3
14.690.92
7.0090.93
3.5790.38
7.1591.12
7.4990.34
17.690.9
H
H
1j
H
OCH3
OCH3
H
1k
1l
OCH3
H
OCH3
H
OCOCH3
OCOCHÄCHC6H5
COOCH3
COOC3H5
H
1m
1n
1o
1p
1q
1r
H
H
H
H
431998
ꢀ500
ꢀ500
H
H
6.0590.49
1.8590.03
14.599.1
40.995.78
12.191.3
61.8923.6
32394
44.492.2
39.691.3
38.890.7
45.991.0
50.091.8
436990
6.4290.36
39.490.0
52.393.3
67.493.9
27.7912.7
2.1390.03
4.7791.88
1.6290.15
6.6691.65
16.696.37
17.098.9
50.2923.3
ꢀ500
39.091.5
31.392.1
14.292.3
36.090.6
32.996.5
ꢀ500
7.1291.17
34.095.9
41.291.8
64.4915.4
7.8090.13
3.2490.79
6.2090.36
1.3890.07
7.1591.27
23.395.81
18.695.7
72.2919.0
349931
39.295.6
25.498.8
25.4912.7
43.892.3
18.793.0
401948
5.9590.06
25.5914.8
30.493.0
50.694.6
8.2190.24
2.4790.30
H
H
NO2
H
H
NO2
H
H
1s
NO2
H
OH
N(CH3)2
H
1t
H
2
Á/
H
Á/
H
Á/
H
Cl
Á/
H
3a
3b
3c
H
H
H
H
Cl
H
Cl
CH3
H
3d
3e
H
H
H
H
OCH3
OCOCHÄCHC6H5
H
NO2
H
3f
H
H
H
3g
3h
3i
NO2
H
H
H
1.9390.43
12.390.6
7.0390.07
19.390.3
4.7990.4
0.2290.01
H
H
H
H
N(CH3)2
COOH
OCH3
H
3j
H
H
OCH3
H
OCH3
3k
Melphalan
H
a
Á/
Á/
Á/
Á/
a
These data are reproduced from Ref. [1], p. 970.
disadvantageous. Attention was therefore directed to the
influence of para substituents in rings B and C.
other parts of the molecule with a binding site. In order
to explore this possibility, two approaches were used
involving those compounds possessing a single para
substituent. In the first place, linear and logarithmic
The possibility exists that the aryl ring B (in series 1
and 3) and also ring C in series 1 align at complementary
areas on binding sites. Thus the presence of substituents
in these rings could increase potencies compared to the
unsubstituted compound 1a by interacting with cellular
constituents at an auxiliary binding site adjacent to the
locus of action with the aryl rings B and C. Alterna-
tively, the aryl substituents could impede alignment of
plots between the IC50 values of 1b,d,fÁ
/
j,l,o,r,t as well as
j for each cell line and the MR figures of the R3
3b,d,e,hÁ
/
substituents were made in each series (1m,n, 3f were
eliminated from consideration since specific IC50 values
were not available in each screen). No correlations (P ꢀ
/
0.1) were observed. Second, the IC50 values of each of
Table 2
Evaluation of the cytotoxicity of 4aÁc against murine P388 and L1210 cells and human Molt 4/C8 and CEM T-lymphocytes
/
Compound
R1 R2
R3
R4
R5 R6
R7
R8
IC50 (mM)
P388
L1210
Molt 4/C8
CEM
4a
4b
4c
H
H
H
H
NO2
NO2
H
H
H
H
H
Cl
OCH3
NO2
H
0.45490.02
1.0690.2
0.38390.03
4.1892.16
1.9490.12
3.3192.21
1.7290.03
1.5690.01
1.5590.08
1.7090.01
1.5490.08
1.5590.20
H
OCH3
H
OCH3
OCH3
H
OCH3
H
OCH3

J.R. Dimmock et al. / European Journal of Medicinal Chemistry 37 (2002) 813Á
/
824
817
the compounds 1b,d,fÁ
/
j,lÁ
/
o,r,t and 3b,dÁ
/
f,hÁ/j in the
Table 3
The s and s* values of different atoms and groups in rings B and C of
4aÁc and 1b,r,k
four cytotoxicity assays were compared with the IC50
figures of the unsubstituted compounds, namely 1a and
3a, respectively. In the case of the thirteen analogs in
series 1, compounds with a single para substituent were
more potent than 1a in 6, 3, 4 and 1 cases when the
P388, L1210, Molt 4/C8 and CEM data, respectively,
were considered, i.e. overall in 27% of the comparisons
made. When the seven enones in series 3 were examined,
greater activity was found in 5, 2, 3 and 2 compounds
when comparisons with 3a were made in the four
screens, i.e. overall in 39% of the comparisons made.
The conclusion to be drawn from these evaluations is
that, in general, when only topological features of the
molecules are considered, substitution in the para
location of rings B and C decreases cytotoxic potencies.
This observation suggests that sites of steric inhibition
may be adjacent to the para position of the aryl rings.
An examination was made of the theory that com-
pounds possessing two sites for thiol alkylation should
have an enhanced capacity to kill malignant cells
compared to analogues causing a single chemical attack
with critical biomacromolecules. In other words, the 2-
tetralones in series 1 (two alkylation sites) should be
more than twice as potent as the related compounds in
series 3 (one alkylation site). Comparisons were there-
fore made between the potencies of the following pairs
of compounds which had the same aryl substituents,
namely 1a/3a, 1b/3b, 1c/3c, 1g/3d, 1j/3e, 1k/3k, 1p/3g, 1r/
3h and 1t/3i in each screen. No comparisons were made
between 1m and 3f since specific IC50 concentrations
were not available in every assay. The potencies of the
compounds in series 1 were more than twice those of the
analogs in series 3 in 5/9, 4/9, 5/9 and 4/9 of the cases
when data in the P388, L1210, Molt 4/C8 and CEM
screens, respectively, were considered. Overall the theory
was validated in 50% of the comparisons although
greater potencies were displayed by the analogs in series
1 than 3 in 69% of the cases. These observations suggest
that the mode of action of the compounds in series 1 is
not due exclusively to sequential interaction with
cellular thiols although this pathway may be followed
in some cases.
/
Compound
Combined s and s* values
a
b
Ring B
Ring C
Total
4a
4b
4c
1b
1r
1.27
1.27
0.43
0.73
1.27
0.43
0.73
0.43
1.27
0.73
1.27
0.43
2.00
1.70
1.70
1.46
2.54
0.86
1k
a
The figures refer to the R1 ÁR4 atoms and groups.
These values are for the R5 ÁR8 atoms and groups for 4aÁ
R4 atoms and groups in 1b,r,k.
/
b
/
/
c and
R1 Á
/
c as well as the greater divergence of the s/s* values in
rings B and C in 4aÁc. Hence comparisons were made
only between the IC50 values of 4aÁc (a s/s*ꢀ1.70 or
2.00) with 1r (a s/s*ꢀ2.54) in the four cytotoxicity
screens. Of the comparisons made, 4aꢀ1r, 4bꢀ1r and
4cꢀ1r using the L1210, Molt 4/C8 and CEM cell lines
while in the P388 screen, 4aꢀ1r, 1rꢀ4b and 4cꢀ1r.
/
/
/
/
/
/
/
/
/
/
Thus the hypothesis was validated in 10 of the 12
comparisons made (83%). Bearing in mind the very high
potencies of 4aÁc, the expansion of this series of
/
compounds in which the differences between the elec-
tronic properties of the aryl substituents are exaggerated
still further is clearly warranted.
Analyses of the data for 1aÁ
potency order table [9] did not reveal any parameter
dependencies. The application of the decision tree
/
c,g,j and 3aÁe using a
/
approach [15] using the available data for 1aÁ
well as 3aÁc suggested that the synthesis of further
/c,g as
/
compounds in series 1 and 3 should possess strongly
electron-attracting substituents in the aryl rings in order
for potency to be increased.
The insertion of two additional methoxy groups to 1j
leading to 1k was accompanied by an approximately
sixfold increase in cytotoxicity in each of the four cell
lines suggesting the importance of the 3,4,5-trimethox-
yphenyl motif in these clusters of compounds. In fact 1k
was more potent than melphalan when the two human
T-lymphocytes were considered. The three structurally
Series 4, in which the groups in rings B and C are
different, was prepared in order to examine the hypoth-
esis that substantial divergences between the electronic
properties of the substituents in rings B and C were
greater contributors to cytotoxicity than the magnitude
of the s/s* values of the aryl substituents. The
isomeric mononitro compounds 1pÁr displayed good
activity towards P388 cells. However, only the 2-nitro
analogue 1p was markedly toxic to the other three cell
/
lines. In general, the order of activity was 1pꢀ
/
1qꢀ1r.
/
Insertion of a hydroxy group into 1q to give a potential
mitochondrial poison 1s was accompanied by a reduc-
tion in cytotoxicity in three of the four cell lines.
Previous studies have revealed correlations between
the redox potentials of different chalcones and cytotoxi-
city [16]. In the present case, compounds 1a,k,n,p,q,r,t,
3a,g and 4b were chosen which have a wide range of IC50
values in the four screens. Their redox potentials are
electronic nature of the R1Á
and R5ÁR8 groups in ring C for 4aÁ
relevant data for the related compounds 1b, 1r and 1k
are presented in Table 3. If the potencies of 4aÁc were
/
R4 substituents in ring B
/
/c as well as the
/
greater than 1b and 1k, the result could be attributed to
two indistinguishable features, namely the greater mag-
nitude of the s/s* figures of the aryl substituents in 4aÁ
/

818
J.R. Dimmock et al. / European Journal of Medicinal Chemistry 37 (2002) 813Á824
/
Table 4
Redox potentials of 1a,k,n,p,q,r,t, 3a,g and 4b
Compound
Aryl substituents
Redox potential (mV)
1a
1k
1n
1p
1q
1r
H
1759
1573
1764
1860
1834
1780
1930
1845
1950
1518
3,4,5-(OCH3)3
4-COOCH3
2-NO2
3-NO2
4-NO2
1t
4-N(CH3)2
H
2-NO2
3a
3g
4b
4-NO2/3,4,5-(OCH3)3
portrayed in Table 4. A comparison between the
unsaturated ketones in series 1 and 3 possessing the
same aryl substituents, namely 1a and 3a, as well as 1p
and 3g, indicated that the compound in series 1 had
greater reducing properties and higher potencies. In
order to evaluate the generality of a possible cytotoxi-
city/redox potential relationship, linear and logarithmic
plots were made between the IC50 values of 1a,k,p,q,r,t,
3a,g and 4b for each cell line and the redox potentials. A
positive correlation was noted in the case of Molt 4/C8
Fig. 3. ORTEP diagram of 1g.
Table 5
Torsion angles u1 and u2 (esd) of 1b,c,g,j and 3h determined by X-ray
crystallography and some relative potency scores
cells (Pꢀ
/
B0.1 in both the linear and logarithmic plots)
/
a
Compound u1 (8)
u2 (8)
Relative potency score
and a positive trend was noted with CEM cells (Pꢀ
/
1b
1c
1g
1j
ꢁ44.1 (3)
41.5 (7)
ꢁ46.1 (3) 15
ꢁ13.8 (7) 8.5
0.105 and 0.106 in the linear and logarithmic plots,
respectively). No other correlations were noted. These
data indicate that the role of this physicochemical
parameter in mediating cytotoxicity is dependent upon
the cell line under consideration.
43.7 (7)
ꢁ33.4 (7)
ꢁ42.5 (12)
ꢁ10.1 (8)
31.3 (8)
7
9.5
3h
Á/
Á/
a
The relative potency scores were obtained by allocating 4, 3, 2 and
The structures of 1b,c,g,j and 3h were determined by
X-ray crystallography. In all cases the E stereochemistry
pertaining to the olefinic double bonds was noted. In the
present study, the lack of coplanarity between the planes
of rings B and C with the adjacent olefinic groups as
determined by X-ray crystallography were found and
designated u1 and u2, respectively. An ORTEP diagram
[17] of a representative compound 1g is presented in Fig.
1 points for the relative potencies for the four compounds in each
screen, i.e. 4 for the most potent analogue, 3 for the next most
cytotoxic compound and so forth. Standard deviations of the IC50
figures were taken into consideration. The total points available in
each screen were invariably 10.
A number of investigators have described correlations
between bioactivity and the torsion angles between aryl
rings and adjacent unsaturated groups [18Á20]. In the
/
3; the u1 and u2 values are the torsion angles C1Ã
C12ÃC13 and C3ÃC18ÃC19ÃC20, respectively. The
/
C11Ã
/
present investigation, the relative potencies of 1b,c,g,j in
each cell line were computed, allocating a score of 4 to
the most potent compound, 3 to the next in terms of
potency, then 2 and finally 1 to the compound display-
ing the lowest bioactivity. A maximum of 10 points was
available for the four compounds when each cell line
was considered. The standard deviations of the IC50
values were taken into consideration in calculating the
relative potency scores. Thus if the IC50 figures of the
two least potent compounds in a screen were indis-
tinguishable statistically, each compound would be
allocated a score of 1.5 since three points were available
for the two least cytotoxic analogues. The data in Table
5 indicate that the order of potencies of the four
/
/
/
/
average figures for u1 and u2 in 1b,c,g,j were 40.7 and
25.3, respectively. The larger u1 values may have been
due to the non-bonded interactions between the protons
on carbon atoms 9 and 13. The u1 value of 3h is similar
in magnitude to the figures obtained in the 1,3-diaryli-
dene-2-tetralones 1b,c,g,j indicating that the 3-arylidene
group has little or no effect upon the topography of the
1-arylidene substituent. The negative signs in Table 5
indicate that the torsion angles are in an anticlockwise
direction. The enone 1c possesses an unsymmetrical aryl
substitution pattern and the X-ray crystallographic data
revealed that the chloro atoms are located on carbon
atoms 15, 16, 22 and 23.
compounds was 1bꢀ
/
1jꢀ
/
1cꢀ1g. It is of interest to
/

J.R. Dimmock et al. / European Journal of Medicinal Chemistry 37 (2002) 813Á
/
824
819
note the relative potencies of 1b,c,g,j was positively
correlated with the u2 figures while no correlation
between cytotoxicity and u1 values was detected. The
relative potency scores revealed the marked superiority
of the cytotoxic properties of 1b compared to the three
analogues 1c,g,j. The figures in Table 5 reveal that the
orientation of both aryl rings in 1b,c,g,j were similar
only in the case of 1b which may be a significant
contributor to its potency. For example, potent cyto-
toxicity may be dependent on the specific orientations of
both aryl rings in series 1 fitting into clefts on a binding
site. Thus expansion of this series of compounds in
which ortho substituents are placed in the aryl rings may
lead to molecules having the same orientations and
potent cytotoxicity.
A further experiment was conducted with small
groups of compounds in order to obtain an estimate
of the electron densities at the predicted sites of
thiolation, namely the olefinic carbon atoms adjacent
to rings B and C, and were designated the CB and CC
atoms, respectively. Both PM3 and Mulliken charges
were obtained and the data are presented in Table 6.
The following conclusions may be drawn from these
results. First, differences between the electron densities
on the CB and CC atoms exist in these compounds. In
series 1 and 4, thiolation would likely occur at the
olefinic carbon atom bearing the lower electron density
which was either the CB and CC atoms depending on the
compound using PM3 calculations while the Mulliken
charges indicated the CC atoms in most cases. Second,
the differences in atomic charges between the CB and CC
atoms are generally greater in the compounds in series 4
than 1 as predicted. Third, the differences in electron
densities of the CB atoms in 1 and 3 were noted when the
same aryl substituents were present in ring B. The
disparities between 1b and 3b, 1k and 3k and 1r and 3h
were 0.029, 0.024 and 0.036 esu, respectively (PM3
charges) and 0.041, 0.029 and 0.046 esu, respectively
(Mulliken calculations). These data indicate that the 3-
arylidene group in series 1 exerts an electron-attracting
influence on the CB atoms. Finally, there were no
correlations between the atomic charges and the s/s*
values which may have been due to the strong resonance
effects present in these molecules. This observation may
explain, at least in part, the lack of positive correlations
between the s/s* figures and cytotoxicity vide supra.
In view of the promising cytotoxic activity of many of
these compounds in these four screens, representative
enones were evaluated against a panel of 55Á60 human
/
tumour cell lines from nine different neoplastic diseases,
namely leukemia and melanoma, as well as non-small
cell lung, colon, CNS, ovarian, renal, prostate and
breast cancers. The number of cell lines in each subpanel
of a specific type of cancer varied ranging from two in
the case of prostate cancer to nine when non-small cell
lung tumours were used. These results are portrayed in
Table 7. The data revealed that without exception, the
compounds were more potent than the reference drug
melphalan when all cell lines are considered. Of parti-
cular note is the observation that 4b possesses 40 times
the potency of this clinically useful drug. The average
IC50 values of the compounds in series 1, 3 and 4 were
4.93, 7.06 and 1.59 mM, respectively, which are sub-
stantially lower than the figures for the two reference
drugs. Although not all compounds in each of the series
1, 3 and 4 were evaluated against the human tumour cell
lines, the average IC50 figures suggest that the order of
potencies were 4ꢀ
/
1ꢀ3 which is a similar observation
/
to the data presented in Table 1.
Previous studies from this laboratory revealed the
selective toxicity of a variety of enones towards leukemic
and/or colon cancer cell lines [1,21]. In the present study,
a selectivity index (SI) figure was calculated for each
compound whereby the IC50 values towards either
leukemic or colon cancer cells were compared with the
potencies towards all cell lines. Compounds displaying
Table 6
Atomic charges at the olefinic carbon atoms of 1a,b,k,r, 3b,h,k and 4aÁc
/
a
Compound
PM3 charges
CB
Mulliken charges
s/s* Values
CC
CCÃCB
CB
CC
CCÃCB
Ring B
Ring C
1a
1b
1k
1r
ꢁ0.020
ꢁ0.006
ꢁ0.012
ꢁ0.032
ꢁ0.035
ꢁ0.068
ꢁ0.036
ꢁ0.042
ꢁ0.051
ꢁ0.011
ꢁ0.014
ꢁ0.019
ꢁ0.028
ꢁ0.038
Á/
Á/
ꢁ0.006
0.013
0.016
ꢁ0.105
ꢁ0.106
ꢁ0.114
ꢁ0.115
ꢁ0.147
ꢁ0.161
ꢁ0.143
ꢁ0.120
ꢁ0.133
ꢁ0.103
ꢁ0.095
ꢁ0.096
ꢁ0.097
ꢁ0.103
Á/
Á/
ꢁ0.010
ꢁ0.010
ꢁ0.017
ꢁ0.012
Á/
Á/
0.49
0.73
0.43
1.27
0.73
1.27
0.43
1.27
1.27
0.43
0.49
0.73
0.43
1.27
0.006
3b
3h
3k
4a
4b
4c
Á/
Á/
Á/
Á/
Á/
Á/
0.73
0.43
1.27
Á/
Á/
Á/
ꢁ0.006
ꢁ0.020
ꢁ0.053
ꢁ0.036
ꢁ0.031
0.042
ꢁ0.090
ꢁ0.097
ꢁ0.114
ꢁ0.030
ꢁ0.036
0.011
a
The s/s* values are for the substituents R1 Á
/
R4 in ring B in series 1, 3 and 4 and for ring C, these figures are of the R1 Á
/
R4 groups in series 1 and
the R5 ÁR8 substituents in series 4.
/

820
J.R. Dimmock et al. / European Journal of Medicinal Chemistry 37 (2002) 813Á
/
824
Table 7
Evaluation of various compounds against a panel of human tumour cells lines
a
Compound
All cell lines
Leukemic cells
Colon cancer cells
b
b
IC50 (mM)
IC50 (mM)
SI
IC50 (mM)
SI
1a
1b
2.24
2.09
10.23
2.75
2.69
2.95
14.79
3.47
1.35
2.34
9.33
11.75
5.25
4.17
2.45
0.589
1.74
23.5
32.6
1.58
2.20
2.30
2.37
1.91
1.45
6.22
1.14
0.528
1.90
4.49
6.37
3.91
1.94
1.01
0.301
0.482
4.65
27.6
1.4
1.0
4.5
1.2
1.4
2.0
2.4
3.0
2.6
1.2
2.1
1.8
1.3
2.2
2.4
2.0
3.6
5.1
1.2
1.90
1.83
8.91
2.29
2.37
2.17
7.43
3.32
1.10
1.92
8.43
12.88
4.41
3.48
2.10
0.372
1.17
44.6
7.90
1.2
1.1
1.2
1.2
1.1
1.4
2.0
1.1
1.2
1.2
1.1
0.9
1.2
1.2
1.2
1.6
1.5
0.5
4.1
1c
1d
1f
1i
1j
1l
1p
1r
1s
3c
3g
3k
4a
4b
4c
Melphalan
5-Fluorouracil
a
A total of 55Á60 cell lines were used in this screen from nine different groups of neoplasms namely leukemia, melanoma and non-small cell lung,
/
colon, CNS, ovarian, renal, prostate and breast cancers. The data for melphalan and 5-fluorouracil were taken from Ref. [1], p. 971.
b
SI, selectivity index, i.e. the IC50 values for all cell lines divided by the IC50 figures of either leukemic or colon cancer cells.
selectivity for one or more groups of cancers may exert a
preferential antineoplastic activity towards malignant
rather than normal cells. These data are also presented
in Table 7 and so are the relevant data for melphalan,
which is used clinically in treating leukemia [22], and 5-
fluorouracil which is used in the treatment of carcinoma
of the colon [23]. A SI value of 1.5 was arbitrarily chosen
to be an indicator of promising selectivity.
the L1210 screen. The effects of 2-tetralones on macro-
molecular biosyntheses in L1210 cells are presented in
Table 8. Linear and logarithmic plots between the extent
of inhibition of each of the macromolecular syntheses
and the IC50 figures obtained in the L1210 screen were
undertaken. Using linear plots, positive correlations
were noted between the IC50 values obtained in the
L1210 screen and the inhibition of the biosyntheses of
The evaluation against leukemic cells revealed that all
of the compounds were more potent than melphalan
except for 1j and 3c. In particular, 1p, 4b and 4c had
IC50 values of less than 1 mM and these three
compounds were, on average, 11 times more potent
than melphalan. No less than 65% of the compounds
had SI values of greater than 1.5 when assessed against
leukemic cells and maximum selectivity was displayed
by 1c and 4c. The compounds prepared in this study
were more cytotoxic to colon cancers than 5-fluorouracil
except for 1c, 1s and 3c. Maximum activity was
displayed by 4b having 21 times the potency of 5-
fluorouracil. This compound, as well as 1j and 4c, had
SI values of 1.5 or greater.
The next stage of the investigation sought to explore
whether the modes of action of these compounds could
be attributed, at least in part, to interference with the
biosynthesis of DNA, RNA and proteins as well as
binding to DNA. Initially possible inhibiting effects on
macromolecular biosyntheses were considered. Five
representative compounds were chosen having IC50
values of ca. 2 (1p), 6 (1a, 3g) and 43 (1r, 3a) mM in
DNA (P B
0.05). The logarithmic plots gave similar results except
that a P value of B0.05 in the DNA assay was
obtained. These results afford strong evidence that the
mode of action of the compounds prepared in this study
includes interference with the synthesis of cellular
macromolecules.
An ethidium displacement assay revealed that com-
pounds 1a,p and 3a,g did not bind to calf thymus DNA
and it is unlikely therefore that DNA binding contri-
butes to the cytotoxicity of these 2-tetralones.
Finally, the question was addressed as to whether
these compounds were likely to be general biocidal
agents or to display some selectivity towards neoplastic
cells. Two approaches were utilised in this regard,
namely the evaluation of representative compounds
against certain fungi as well as toxicity studies in mice.
Compounds 1b, 3b and 3f have substantially divergent
cytotoxicity towards various neoplastic cells as the data
in Table 1 reveal. At the highest concentration utilised,
namely 500 mM, fungicidal activity was absent. Thus the
possibility that the compounds prepared in this study
/
0.1), RNA (P B
/
0.05) and proteins (P B
/
/

J.R. Dimmock et al. / European Journal of Medicinal Chemistry 37 (2002) 813Á
/
824
821
Table 8
Inhibition of 1a,p,r, 3a,g on DNA, RNA and protein biosyntheses in murine L1210 cells
Compound
IC50 (mM)
DNA synthesis
RNA synthesis
Protein synthesis
L1210 proliferation
1a
1p
1r
13.6
13.8
333
17.5
15.5
300
13.2
12.4
102
6.23
1.85
40.9
44.4
6.42
3a
3g
169
283
83.5
19.4
55.5
58.5
have low toxicities suggested a more extensive in vivo
study. Compounds 1a,cÁh,j,k,oÁq,s,t, 2, 3d,e,g,h,k, 4b
number of the compounds towards leukemic and colon
cancer cells is particularly noteworthy. The mode of
action of these compounds includes interference with
macromolecular biosyntheses and probably cellular
redox reactions. The observation that the compounds
are well tolerated by mice enhances the need to
investigate these molecules further.
/
/
and 4c were injected intraperitoneally into mice using
doses of up to and including 300 mg kgꢁ1. The animals
were observed 0.5 and 4 h after administration of the
compounds for pathological symptoms including neu-
rotoxicity. Furthermore, in order to explore the possi-
bility of CNS penetration by these 2-tetralones, these
compounds were examined in the maximal electroshock
(MES) and subcutaneous pentylenetetrazole (scPTZ)
screens for possible protection against seizures. Any
activity noted in either screen implies CNS penetration
which could be beneficial in suggesting a utility of these
compounds in treating CNS tumours; conversely, this
property may be detrimental in causing unwanted
toxicity. No mortalities were noted and marginal
neurotoxicity was noted only with 1f,h,o,s. None of
the compounds were active in the MES screen and only
1q afforded protection in some of the mice at the
maximum dose of 300 mg kgꢁ1 in the scPTZ test. In
order to evaluate the tolerability of these compounds in
vivo, 1a,d,p,q, 3g,h were administered orally to rats
using doses of 30 mg kgꢁ1 or higher. Observations were
6. Experimental
6.1. Chemistry
M.p.s are unconnected and are quoted in 8C. Ele-
mental analyses were undertaken on 1aÁ
/
o, 2 and 3aÁ
/
f,j,k (C, H) and 1pÁt, 3gÁi and 4aÁc (C, H, N) by K.
/
/
/
Thoms, Department of Chemistry, University of Sas-
katchewan and are within 0.4% of the calculated values.
1H-NMR spectra were determined on all compounds in
series 1Á
(500 MHz). An EnrafÁ
was used for the collection of X-ray crystallographic
data. The recrystallisation solvents were CHCl3ÁMeOH
(A), CHCl3ÁMeOH (1:9, B), etherÁMeOH (C), MeCN
(D), CHCl3 (E), dioxane (F) and CHCl3ÁMeOH (2:8,
/4 using a Bruker AM 500 FT NMR machine
/Nonius CAD-4 diffractometer
/
made in general over the 0.25Á4 h time span. No deaths
/
/
/
of the animals were reported and toxicity was unde-
tected. However, in contrast to the mouse intraperito-
neal screen, the dienones 1a,d,p,q gave protection in the
MES screen suggesting that CNS penetration had
occurred. The conclusion to be drawn is that, under
the experimental conditions utilised, these compounds
were well tolerated in vivo.
/
G).
6.1.1. Synthesis of series 1 and 2
The compounds in series 1 and 2 were prepared by the
following general method. Hydrogen chloride gas was
passed into a solution of 2-tetralone (10 mM) and the
appropriate aryl aldehyde (21 mM) in Et2O (20 mL) for
0.25 h and the resultant mixture was stirred at room
temperature (r.t.) for 12 h. The precipitate was removed
5. Conclusions
by filtration, washed with Et2O (2ꢂ
recrystallised to give the desired product.
The m.p.s (8C), yields (%) and recrystallisation
solvents, respectively, were as follows: 1a: 156Á157, 65,
A; 1b: 166, 65, A; 1c: 202, 59, A; 1d: 166Á167, 70, B; 1e:
175Á177, 65, B; 1f: 185Á186, 72, B; 1g: 197Á198, 69, A;
1h: 137Á139, 71, B; 1i: 210Á212, 70, C; 1j: 194Á
A; 1k: 139, 70, B; 1l: 155Á156, 90, D; 1m: 189Á
A; 1n: 206Á208, 55, E; 1o: 150Á
A; 1q: 222Á224, 61, E; 1r: 229Á230, 68, E; 1s: 202Á
/
25 mL) and
This study has revealed that a number of 1,3-
diarylidene-2-tetralones possess significant cytotoxicities
towards various murine and human cancer cell lines. In
particular, the compounds in series 4, and especially 4b,
were more potent that melphalan in most of the assays
utilised. The bioevaluation of these compounds revealed
that their design to undergo sequential reactions with
cellular constituents warrants further investigation. In
addition, the demonstration of selective toxicity by a
/
/
/
/
/
/
/
/
195, 72,
/
/190, 85,
/
/
152, 49, B; 1p: 185, 30,
/
/
/204,

822
J.R. Dimmock et al. / European Journal of Medicinal Chemistry 37 (2002) 813Á824
/
1
51, F; 1t: 200, 79, E; 2: 271Á
spectrum (CDCl3) of a representative compound 1i was
as follows: 4.20 (s, 2H, CH2), 6.74Á6.76 (d, 2H, Jꢀ8.6
Hz, aryl H), 6.88Á6.89 (d, 2H, Jꢀ8.6 Hz, aryl H), 7.05Á
7.06 (m, 1H, methine H), 7.18 (m, 2H, aryl H), 7.24Á
7.25 (d, 1H, Jꢀ7.4 Hz, aryl H), 7.36Á7.38 (d, 2H, Jꢀ
8.6 Hz, aryl H), 7.45Á7.50 (m, 3H, aryl H), 7.70 (s, 1H,
methine H).
/
273, 63, E. The H-NMR
7.44 (2d, 4H, aryl H), 7.58 (s, 1H, methine H), 7.60Á
(d, 2H, Jꢀ8.8 Hz, aryl H), 7.81 (s, 1H, methine H),
8.13Á8.15 (d, 2H, Jꢀ8.8 Hz, aryl H).
/
7.61
/
/
/
/
/
/
/
/
/
6.1.4. X-ray crystallographic determinations of 1b,c,g,j
and 3h
/
/
/
/
Suitable crystals for X-ray crystallography were
obtained by vapour diffusion. A mixture of CHCl3
and MeOH (3:7) was used for 1b,c and 1j while a
solution of CHCl3 and MeOH (1:4) and MeOH were
used for 1g and 3h, respectively. Data were collected
using v scans. Structures were solved using NRCVAX
[24] and refined using SHELXL [25]. Atomic scattering
factors and anomolous dispersion corrections were
taken from the literature [26]. Hydrogen atoms were
placed by geometry and are not refined. Specific details
of the X-ray crystallographic data of these compounds
may be obtained from the authors.
6.1.2. Syntheses of series 3
A mixture of 2-tetralone (10 mM), the appropriate
aryl aldehyde (10 mM), AcOH (50 mg), piperidine (50
mg), 4A molecular sieves (2.5 g) and C6H5CH3 (30 mL)
was stirred at r.t. for 24 h. The molecular sieves, which
were removed by filtration, were washed with EtOAc
(25 mL). The combined organic extracts were stirred
with an aqueous solution of sodium metabisulphite (5%
w/v, 25 mL) for 0.25 h. The separated organic phase was
washed with water (25 mL) and brine (25 mL) and dried
(anhydrous Mg2SO4). The solvents were removed to
produce a residue which was purified by passage
6.1.5. Measurement of redox potentials
The redox potentials of 1a,k,n,p,q,r,t, 3a,g and 4b
were determined by a previously reported methodology
[2]. In brief, solutions of the compounds in C3H6O (0.1
M) were prepared and the determinations utilised a
glassy carbon working electrode and a Pt wire counter
electrode.
through a column of silica gel (60Á200 mesh) using
/
solvents of C6H14 and subsequently a mixture of EtOAc
in C6H14 (2% v/v). The products obtained were further
purified by recrystallisation.
The m.p.s (8C), yields (%) and recrystallisation
solvents, respectively, were as follows: 3a: 116Á
A; 3b: 138, 62, A; 3c: 106Á107, 35, A; 3d: 96Á97, 84, A;
3e: 83Á84, 64, A; 3f: 179, 53, A; 3g: 124, 68, A; 3h: 154Á
155, 65, A; 3i: 110Á111, 38, A; 3j: 231Á233, 40, A; 3k:
143, 65, G. The 1H-NMR spectrum (CDCl3) of a
representative compound 3e was as follows: 2.58Á2.59
(t, 2H, Jꢀ6.7 Hz, 4-CH2), 2.99Á3.02 (t, 2H, Jꢀ6.6 Hz,
3-CH2), 3.79 (s, 3H, OCH3), 6.76Á6.78 (d, 2H, Jꢀ9.7
Hz, aryl H), 7.04 (m, 1H, aryl H), 7.18 (bs, 1H, aryl H),
7.39Á7.40 (m, 1H, aryl H), 7.41Á7.42 (d, 1H, aryl H),
7.42Á7.43 (d, 2H, Jꢀ9.5 Hz, aryl H), 7.62 (s, 1H,
methine CH).
/
117, 75,
/
/
6.1.6. Charge density calculations
/
/
The geometry of each of the molecules 1a,b,k,r, 3b,h,k
/
/
and 4aÁ
computational methodology. All charges from the
Mulliken populations were calculated by the HartreeÁ
Fock approach using the 6Á31** basis. Hyperchem 6.0
/c were optimised using the semiempirical PM3
/
/
/
/
/
/
/
/
software was used for geometry optimisations and
charge calculations.
/
/
/
/
6.1.7. Statistical analyses
The s values were taken from the literature [27] and
the s* figures were obtained from a reference source
[28]. The aryl p and the MR values have been described
previously [29]. When performing the statistical analysis
6.1.3. Synthesis of series 4
A stream of dry hydrogen chloride gas was passed
through a solution of the appropriate 1-arylmethylene-
2-tetralone (2.5 mM) and aryl aldehyde (2.8 mM) in
CHCl3 (10 mL) for 10 min. After stirring at r.t. for 12 h,
CHCl3 and hydrogen chloride were removed by eva-
poration. The resultant residue was purified using a
using the s/s*, p and MR values, all of the R1ÁR4
/
substituents in series 1 and 3 were taken into considera-
tion. For example, the electronic effect of these groups
in 3b was calculated to be 0.73 (R1ꢀ
/
Hꢀ
0.24). Similarly, the MR value
of this compound was 9.12 (R1ꢀR2ꢀR4ꢀ
Hꢀ1.03ꢂ
3, R3ꢀ
Clꢀ6.03). In the case of series 1, the figures of
each physicochemical constant of the R1ÁR4 group were
/
0.49, R2ꢀ
/
R4ꢀ
/
Hꢀ
/
0, R3ꢀ
/
Clꢀ
/
chromatography column of silica gel (60Á/200 mesh) and
/
/
/
/
/
eluants of EtOAc in C6H14 (2% v/v increasing to 3% v/v)
to yield products which were further purified by
recrystallisation.
/
/
/
doubled prior to undertaking the statistical evaluation.
The m.p.s (8C), yields (%) and recrystallisation
solvents, respectively, were as follows: 4a: 193Á
/
194, 55,
6.2. Screening
B; 4b: 185Á186, 52, B; 4c: 176Á
/
/
178, 68, B. The 1H-NMR
spectrum (CDCl3) of a representative compound 4a was
as follows: 4.15 (s, 2H, 4-CH2), 7.05 (t, 1H, aryl H), 7.18
6.2.1. Cytotoxicity evaluations
The evaluation of the compounds described in the
study against P388D1 cells was undertaken by a
(d, 1H, aryl H), 7.25 (d, 2H, Jꢀ
/
8.1 Hz, aryl H), 7.42Á
/

J.R. Dimmock et al. / European Journal of Medicinal Chemistry 37 (2002) 813Á
/
824
823
reported procedure [30], while the screening against
L1210, Molt 4/C8 and CEM cells used a method which
has been described previously [31].
The 2-tetralone 1q protected 2/5 mice in the scPTZ test
using a dose of 300 mg kgꢁ1 at the end of 0.5 h. In
addition, 1g caused toxic extension in the scPTZ screen
at the end of 0.5 h when a dose of 30 mg kgꢁ1 was used.
Six compounds were administered per os to rats using
doses of 30 (1a,p, 3g,h), 30 and 125 (1d) and 50 mg kgꢁ1
(1q). Observations for toxicity were made at the end of
0.25, 0.5, 1, 2 and 4 h for 1a,d (30 mg kgꢁ1 dose),p,q
(although no observation was made at the end of 4 h),
3g and 3h. In the case of 1d (125 mg kgꢁ1 dose),
evaluations were made after 2, 4, 6 and 8 h after
administration of the compound. In this test, evidence
of toxicity was noted when ataxia, abnormal gait and
stance occurred. None of the compounds displayed
toxicity. Evaluation of 1a,d,p,q, 3g,h in the oral MES
screen was also undertaken using a dose of 30 mg kgꢁ1
and the animals were observed at the end of 0.25, 0.5, 1,
2 and 4 h. Protection was afforded by the following
compounds (number of rats protected out of four, time
in h): 1a (1, 1; 2, 2), 1d (2, 0.25; 1, 0.5; 1, 1; 3, 2; 1, 4), 1p
(1, 0.5; 1, 0.5; 1, 1; 3, 2; 4, 4) and 1q (1, 0.25; 1, 0.5; 1, 4).
Elevation of the dose of 1d to 100 mg kgꢁ1 led to
protection of 1, 2, 1 and 2 rats out of four at the end of
1, 2, 4 and 6 h, respectively. An oral dose of 50 mg kgꢁ1
of 1q administered to rats afforded protection in the
scPTZ screen in one of four rats, while no activity was
observed at the end of 0.5, 1 and 2 h.
The cytotoxicity of various compounds was examined
towards human tumour cell lines using a literature
procedure [32]. This screen utilised a concentration
range of 10ꢁ4Á
/
10ꢁ8 M and 58 (range of 55Á
60) cell
/
lines. The compounds inhibited the growth of all of the
cell lines by more than 50% with the following excep-
tions (number of cell lines in which the IC50 of the
compound was greater than 10ꢁ4 M/total number of
cell lines), namely 1c (2/58), 1j (1/58) and 4c (3/55). The
figures listed in Table 7 are IC50 concentrations for
leukemic and colon cancer cell lines.
6.2.2. DNA binding studies
Compounds 1a,p, 3a and 3g were evaluated for any
binding properties by an ethidium displacement assay
[33]. In brief, excess of the compound was incubated
with 50 mg mLꢁ1 of calf thymus DNA at 20 8C for 24 h
in tromethamine hydrochloride buffer, pH 8.0. Aliquots
of the DNA were removed and assessed for their ability
to bind ethidium.
6.2.3. Antifungal evaluation
The enones 1b, 3b and 3f were evaluated against three
isolates of A. fumigatus (ATCC 208995, ATCC 208996
and ATCC 208997) and one isolate of C. albicans
(ATCC 90028) using the broth dilution method which
has been described previously [34].
Acknowledgements
6.2.4. Effects of various compounds on macromolecular
biosyntheses
In each well of a 96-well microtiter plate were placed
The authors thank the following organisations for
providing financial assistance for this study, namely
Purdue Neuroscience Inc., USA (J.R.D., M.P.P.), the
Natural Sciences and Engineering Research Council
(J.W.Q., E.O.O., H.-B.K.), the Canadian Institutes of
Health Research (A.C., J.S.L.), the National Cancer
Institute (T.M.A.) and the Flemish Fonds voor Gen-
eeeskundig Wetenschappelijk Onderzoek (E.D.C., J.B.).
The excellent technical assistance of L. van Berckelaer
(L1210, Molt 4/C8 and CEM assays) is gratefully
acknowledged, as is the US National Cancer Institute,
which generated the biodata in table 7.
L1210 cells (ꢁ
/
2.5ꢂ
105), different concentrations of
/
1a,p,r, 3a or 3g and 0.25 mCi of [methyl-3H]deoxythy-
midine, [5-3H]uridine or [3,4-3H]leucine. After 20 h
incubation at 37 8C, the trichloroacetic acid-insoluble
fractions of the cell cultures were analysed for radio-
activity using a liquid scintillation counter.
6.2.5. Evaluations in the murine toxicity, MES and
scPTZ screens
Compounds 1a,cÁ
/
h,j,k,oÁq,s,t, 2, 3d,e,g,h,k, 4b and
/
4c were examined in the initial murine toxicity, MES
and scPTZ screens [35]. In brief, doses of 30, 100 and
300 mg kgꢁ1 of each compound were injected intraper-
itoneally into mice and the animals were observed at the
end of 0.5 and 4 h. Neurotoxicity was measured by the
rotorod procedure [36] and was observed with the
following compounds (number of animals displaying
neurotoxicity/total number of animals used, time of
observation in h, dose in mg kgꢁ1): 1f (1/4, 0.5, 300), 1h
(1/4, 0.5, 300), 1o (1/8, 0.5, 100; 2/4, 0.5, 300; 1/4, 4, 100;
1/2, 4, 300) and 1s (2/4, 0.5, 300). None of the
compounds afforded protection in the MES screen.
References
[1] J.R. Dimmock, P. Kumar, A.J. Nazarali, N.L. Motoganahalli,
T.P. Kowalchuk, M.A. Beazely, J.W. Quail, E.O. Oloo, T.M.
Allen, J. Szydlowski, E. De Clercq, J. Balzarini, Eur. J. Med.
Chem. 35 (2000) 967Á977.
/
[2] J.R. Dimmock, M.P. Padmanilayam, R.N. Puthucode, A.J.
Nazarali, N.L. Motaganahalli, G.A. Zello, J.W. Quail, E.O.
Oloo, H.-B. Kraatz, J.S. Prisciak, T.M. Allen, C.L. Santos, J.
Balzarini, E. De Clercq, E.K. Manavathu, J. Med. Chem. 44
(2001) 586Á
/
593.
[3] A. Baluja, A.M. Municio, S. Vega, Chem. Ind. (1964) 2053Á
/
2054.

824
J.R. Dimmock et al. / European Journal of Medicinal Chemistry 37 (2002) 813Á824
/
[4] H. Singh, V.K. Kapoor, Medicinal and Pharmaceutical Chem-
istry, Vallabh Prakashan, Delhi, India, 1996, p. 522.
[5] G.R. Petit, S.B. Singh, M.R. Boyd, E. Hamel, R.K. Petit, J.M.
[22] E.M. Greenspan, H.W. Bruckner, in: E.M. Greenspan (Ed.),
Clinical Cancer Chemotherapy, Raven Press, New York, 1975, p.
37.
Schmidt, F. Hogan, J. Med. Chem. 38 (1995) 1666Á
[6] J.B. Mitchell, A. Russo, Br. J. Cancer (Suppl. VIII) 55 (1987) 96Á
104.
[7] K. Tsutsui, C. Komuro, K. Ono, T. Nishidai, Y. Shibamoto, M.
Takahashi, M. Abe, Int. J. Radiat. Oncol. Biol. Phys. 12 (1986)
/
1672.
[23] W.A. Remers, in: J.N. Delgado, W.A. Remers (Eds.), Wilson and
Gisvold’s Textbook of Organic Medicinal and Pharmaceutical
Chemistry, 10th ed., Lippincott-Raven, Philadelphia, 1998, p.
367.
/
[24] E.J. Gabe, Y. LePage, J.-P. Charland, P.S. White, J. Appl.
1183Á
[8] P.N. Craig, J. Med. Chem. 14 (1971) 680Á
[9] J.G. Topliss, J. Med. Chem. 20 (1977) 463Á
[10] H. Singh, V.K. Kapoor, Medicinal and Pharmaceutical Chem-
istry, Vallabh Prakashan, Delhi, 1996, p. 519.
/
1186.
Crystallogr. 22 (1989) 384Á
/
387.
[25] G.M. Sheldrick, SHELXL-97. Programme for the Refinement of
/
684.
/469.
Crystal Structures, University of Gottingen, Germany, 1997.
¨
[26] International Tables for X-ray Crystallography, vol. IV, Kynoch
Press, Birmingham, UK, 1974.
[11] T.A. Connors, in: J.W. Gorrod (Ed.), Drug Toxicity, Taylor and
Francis Ltd, London, 1979, p. 175.
[27] D.D. Perrin, B. Dempsey, E.P. Serjeant, pKa Prediction for
Organic Acids and Bases, Chapman and Hall, London, 1981, pp.
[12] P.G. Heytler, in: M. Erecin´ska, D.F. Wilson (Eds.), Inhibitors of
Mitochondrial Functions, Pergamon Press, Oxford, 1981, pp.
109Á112, 114.
/
[28] R.W. Taft, Jr., in: M.S. Newman (Ed.), Steric Effects in Organic
Chemistry, John Wiley and Sons, Inc, New York, 1956, p. 591.
[29] C. Hansch, A.J. Leo, Substituent Constants for Correlation
Analysis in Chemistry and Biology, John Wiley and Sons, New
202Á
[13] Z. Jia, J.W. Quail, V.K. Arora, J.R. Dimmock, Acta Crystallogr.,
Sect. C 45 (1989) 285Á289.
[14] Z. Jia, J.W. Quail, J.R. Dimmock, Acta Crystallogr., Sect. C 46
(1990) 2467Á2468.
[15] J.G. Topliss, J. Med. Chem. 15 (1992) 1006Á
/
205.
/
York, 1979, pp. 49Á
[30] O.A. Phillips, L.A. Nelson, E.E. Knaus, T.M. Allen, R. Fathi-
Afshar, Drug Des. Deliv. 4 (1989) 121Á127.
[31] J. Balzarini, E. De Clercq, M.P. Mertes, D. Shugar, P.F.
Torrence, Biochem. Pharmacol. 31 (1982) 3673Á3682.
[32] M.R. Boyd, K.D. Paull, Drug Dev. Res. 34 (1995) 91Á
[33] A.R. Morgan, J.S. Lee, D.E. Pulleyblank, N.L. Murray, D.E.
Evans, Nucl. Acids Res. 7 (1979) 547Á569.
[34] E.K. Manavathu, S.C. Vashishtha, G.J. Alangaden, J.R. Dim-
mock, Can. J. Microbiol. 44 (1998) 74Á79.
/50.
/
/1011.
/
[16] J.R. Dimmock, N.M. Kandepu, M. Hetherington, J.W. Quail, U.
Pugazhenthi, A.M. Sudom, M. Chamankhah, P. Rose, E. Pass,
T.M. Allen, S. Halleran, J. Szydlowski, B. Mutus, M. Tannous,
E.K. Manavathu, T.G. Myers, E. De Clercq, J. Balzarini, J. Med.
/
/109.
Chem. 41 (1998) 1014Á1026.
/
/
[17] C.K. Johnson, ORTEP II. Report ORNL-5138, Oak Ridge
National Laboratory, Tennessee, USA, 1976.
[18] W.H. Laarhoven, R.J.F. Nivard, E. Havinga, Experientia 17
/
[35] J.P. Stables, H.J. Kupferberg, in: G. Avanzini, P. Tanganelli, M.
(1961) 214Á
[19] D.W. Adamson, P.A. Barrett, J.W. Billingshurst, A.F. Green,
T.S.G. Jones, Nature 168 (1951) 204Á205.
/
215.
Avoli (Eds.), Molecular and Cellular Targets for Antiepileptic
Drugs, John Libbey and Company Ltd, London, 1997, pp. 191Á
/
/
[20] A. Albert, Selective Toxicity, 7th ed., Chapman and Hall,
London, 1985, p. 514.
198.
[36] M.S. Dunham, T.A. Miya, J. Am. Pharm. Ass. Sci. Ed. 46 (1957)
[21] S.C. Vashishtha, T.M. Allen, S. Halleran, J. Szydlowski, C.L.
Santos, E. De Clercq, J. Balzarini, J.R. Dimmock, Pharmazie 56
208Á209.
/
(2001) 390Á393.
/