6-Benzylidene-2-[4-(pyridin-3-ylcarboxy)benzylidene]cyclohexanones: A novel cluster of tumour-selective cytotoxins

Article abstract of DOI:10.1016/j.bmcl.2017.02.016

Novel cytotoxins 3–5 containing the 1,5-diaryl-3-oxo-1,4-pentadienyl pharmacophore are disclosed. The compounds in series 3 and 5 have the potential to liberate niacin which may reduce some of the side effects of antineoplastic compounds. 3a-c emerged as the most potent cytotoxic compounds with IC₅₀ values in the low micromolar range against human Molt4/C8 and CEM CD₄⁺ T-lymphocytes as well as murine L1210 leukemia cells. QSAR studies revealed that cytotoxic potencies were negatively correlated with the magnitude of the Hammett sigma values of the aryl substituents. The compounds 3a-e displayed tumour-selective toxicity against human HL-60, HSC-2, HSC-3 and HSC-4 neoplasms as compared to human HGF, HPC and HPLF nonmalignant cells. A representative potent compound 3a caused PARP1 cleavage and G₀/G₁ cell cycle arrest in HSC-2 cells. These compounds are well tolerated in mice at doses up to and including 300 mg/kg of the compounds and no mortalities were noted after 4 h. The stability studies undertaken did not reveal that a representative compound 3a underwent hydrolysis to the related phenol 2a. Some guidelines for further analog development of the novel esters 3 were made.

Full text of DOI:10.1016/j.bmcl.2017.02.016

Accepted Manuscript
6-Benzylidene-2-[4-(pyridin-3-ylcarboxy)benzylidene]cyclohexanones: A
novel cluster of tumour-selective cytotoxins
Atulya K. Panda, Umashankar Das, Naoki Umemura, Hiroshi Sakagami,
Masami Kawase, Jan Balzarini, Erik De Clercq, Stephen G. Dimmock, Praveen
K. Roayapalley, Jonathan R. Dimmock
PII:
S0960-894X(17)30134-8
DOI:
http://dx.doi.org/10.1016/j.bmcl.2017.02.016
BMCL 24688
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
5 January 2017
6 February 2017
7 February 2017
Please cite this article as: Panda, A.K., Das, U., Umemura, N., Sakagami, H., Kawase, M., Balzarini, J., De Clercq,
E., Dimmock, S.G., Roayapalley, P.K., Dimmock, J.R., 6-Benzylidene-2-[4-(pyridin-3-
ylcarboxy)benzylidene]cyclohexanones: A novel cluster of tumour-selective cytotoxins, Bioorganic & Medicinal
Chemistry Letters (2017), doi: http://dx.doi.org/10.1016/j.bmcl.2017.02.016
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

6-Benzylidene-2-[4-(pyridin-3-ylcarboxy)benzylidene]cyclohexanones: A
novel cluster of tumour-selective cytotoxins
Atulya K. Pandaa, Umashankar Dasa,*, Naoki Umemurab, Hiroshi Sakagamib, Masami Kawasec, Jan
Balzarinid, Erik De Clercqd, Stephen G. Dimmocke, Praveen K. Roayapalleya, Jonathan R. Dimmocka,*
aDrug Discovery and Development Research Group, College of Pharmacy and Nutrition, University of
Saskatchewan, Saskatoon, Saskatchewan S7N 5E5, Canada
bDivision of Pharmacology, Meikai University School of Dentistry, Sakado, Saitama 350-0283, Japan
cFaculty of Pharmaceutical Sciences, Matsuyama University, 4-2 Bunkyo-cho, Matsuyama, Ehime 790-
8578, Japan
dRega Institute of Medical Research, KU Leuven, 3000 Leuven, Belgium
eDepartment of Finance, Nanyang Technological University, Singapore 639798, Singapore
*Corresponding authors.
E-mail addresses: umashankar.das@usask.ca (U. Das), jr.dimmock@usask.ca (J. R. Dimmock).
1

ABSTRACT
Novel cytotoxins 3-5 containing the 1,5-diaryl-3-oxo-1,4-pentadienyl pharmacophore are
disclosed. The compounds in series 3 and 5 have the potential to liberate niacin which may reduce
some of the side effects of antineoplastic compounds. 3a-c emerged as the most potent cytotoxic
+
compounds with IC50 values in the low micromolar range against human Molt4/C8 and CEM CD4 T-
lymphocytes as well as murine L1210 leukemia cells. QSAR studies revealed that cytotoxic potencies
were negatively correlated with the magnitude of the Hammett sigma values of the aryl substituents.
The compounds 3a-e displayed tumour-selective toxicity against human HL-60, HSC-2, HSC-3 and
HSC-4 neoplasms as compared to human HGF, HPC and HPLF nonmalignant cells. A representative
potent compound 3a caused PARP1 cleavage and G0/G1 cell cycle arrest in HSC-2 cells. These
compounds are well tolerated in mice at doses up to and including 300 mg/kg of the compounds and no
mortalities were noted after 4 hrs. The stability studies undertaken did not reveal that a representative
compound 3a underwent hydrolysis to the related phenol 2a. Some guidelines for further analog
development of the novel esters 3 were made.
Keywords; conjugated unsaturated ketones / cytotoxins / tumour-selective toxicity / structure-activity
relationships / QSAR / niacin
2

One of the long-term objectives of this laboratory is the creation of molecules which are
cytotoxic to neoplastic cells and yet have the capacity to liberate a compound which may diminish the
pathological side effects which often occur in cancer chemotherapy. This goal is planned in various
stages and the present report describes the concept and the initial studies.
The design of the candidate cytotoxins was based on the following observations. First, the
choice was made for a conjugated unsaturated keto group being the cytotoxic pharmacophore, since
various studies revealed that conjugated unsaturated ketones have a preferential or exclusive affinity
for thiols, in contrast to hydroxyl and amino groups which are present in nucleic acids.1-3 Hence, the
genotoxic properties of a number of contemporary anticancer drugs which lead to tumour formation4,5
should be avoided. Second, in the case of the 1,5-diaryl-3-oxo-1,4-pentadienes, sequential interaction
with thiols can take place. In other words, a reaction occurs initially at the most electron-deficient
olefinic carbon atom and subsequently at the remaining olefinic group. Several studies revealed that an
initial lowering of cellular thiol concentrations followed by a second chemical insult is more
detrimental to neoplasms than normal cells.6,7 In the present study, the 1,5-diaryl-3-oxo-1,4-pentadienyl
group was mounted on a cyclohexyl ring since a variety of 2,6-bis(benzylidene)cyclohexanones such
as 1 (Figure 1) have demonstrated significant cytotoxic potencies.8,9
Figure 1. Compound 1 carrying a cytotoxic 1,5-diaryl-3-oxo-1,4-pentadienyl pharmacophore.
One of the effects of cancer chemotherapy is a deficiency of niacin (pyridine-3-carboxylic acid,
nicotinic acid, vitamin B3)10 which can lead to a number of adverse clinical conditions such as pellagra
and dermatitis.11 Conversely, niacin supplementation decreases the incidence of nonlymphocytic
leukemia in rats treated with a carcinogenic nitrosourea.12
The problem therefore is to create molecules which possess a cytotoxic pharmacophore and yet
have the capacity to ameliorate or eliminate the problem of niacin deficiency. Such considerations led
to the designing of series 3 and 5 with the consideration that niacin will be released upon hydrolysis in
situ (Scheme 1). In the case of series 3, the rate of niacin release will be controlled by the electronic
nature of the substituents in ring A. On the other hand, if the compounds are stable any bioactivity
observed will be due to the intact molecules per se. In this case, the effect on cytotoxic potencies of
replacing the pyridyl group of 3a,b by a phenyl ring to give 4a,b was considered of interest. In
addition, if the release of niacin leads to beneficial results such as increases in selective toxicity for
neoplasms, then compound 5 may prove to be a useful antineoplastic agent. The objective of the
present study is to examine whether the novel compounds in series 3-5 (1) display favourable cytotoxic
properties, (2) demonstrate tumour-selective toxicity as compared to non-malignant cells, (3) release
niacin upon hydrolysis, and (4) are tolerated in mice. In addition if potent cytotoxicity is demonstrated,
some of the ways whereby bioactivity is mediated will be investigated.
3

Scheme 1. The syntheses of the compounds in series 2 – 5. The aryl substituents are as follows: a: R = H; b : R = CH3; c: R
= OCH3; d : R = Cl; e: R = NO2; f: R = OH. The reagents used in these syntheses are i = HCl / CH3COOH; ii = 2a-e /
nicotinoyl chloride; iii = 2a,b / benzoyl chloride; iv = 2f / nicotinoyl chloride.
The compounds in series 3 – 5 were synthesised by the route indicated in Scheme 1. Various 2-
benzylidenecyclohexanones, which have been described previously9,13,14, were condensed with 4-
hydroxybenzaldehyde to produce 2a-d, f9 and 2e.15 Acylation of 2a-e with nicotinoyl chloride led to
the esters 3a-e. Reaction of 2a, b with benzoyl chloride produced the corresponding esters 4a, b, while
reaction of 2f with nicotinoyl chloride led to the formation of 5.
Table 1
Evaluation of 3a-e, 4a,b and 5 for their cytotoxic activity against Molt 4/C8, CEM and L1210 cells
Average
IC50(µM)
Molt 4/C8
IC50(µM) Δ2a-e/3a-e
CEM
L1210
Compound
b
b
b
IC50(µM)
5.82±0.83
Δ2a-e/3a-e
IC50(µM)
7.32±0.87
Δ2a-e/3a-e
3-5
2
2aa
2ba
2ca
2da
2ea
5.17±1.83
5.22±1.72
4.18±1.92
4.90±0.67
9.41±0.18
5.52±2.12
1.62±0.26
2.88±1.51
1.19±0.78
7.89±1.70
38.1±9.3
10.4±0.6
5.70±1.56
22.6±15.3
>500
---
---
---
---
---
---
4.31±1.16
5.68±1.65
2.69±0.39
7.27±0.93
6.84±0.14
7.30±0.91
9.49±1.14
9.85±2.33
17.4±9.3
48.1±18.5
40.4±6.3
10.6±2.2
169±58
---
7.41±0.86
6.85±1.18
6.44±1.14
9.95±0.35
13.5±2.3
8.71±0.5
12.8±10.3
7.63±0.83
29.6±1.3
61.1±10.2
46.7±2.0
37.9±7.6
213±42
---
---
---
---
---
---
---
---
---
---
---
---
---
---
---
---
---
---
---
---
2fa
---
---
---
3.19
1.82c
3.51
0.62
0.25
---
0.80c
0.45
0.58
0.16
0.15
---
0.84
0.58c
0.90c
0.22
0.16
---
5.88 6.10
8.39 5.65
6.22 5.57
18.3 4.68
49.1 8.88
3a
3b
3c
3d
3e
32.5
18.1
135
---
---
---
4a
---
---
---
4b
---
---
---
5
Niacin
Melphalan
---
>500
---
>500
---
>500 ---
2.61 ---
3.24±0.56
---
2.47±0.21
---
2.13±0.02
---
aData reported previously.9,15 bThe Δ2a-e/3a-e values are the quotients of the IC50 figures towards a specific cell line
c
of the members of series 2 and 3 which have the same aryl substituents in ring A. The difference between the
IC50 values is not statistically significant.
4

The compounds in series 3-5 were evaluated against human Molt 4/C8 and CEM CD4+ T-
lymphocytes as well as murine L1210 leukemic cells. The data are presented in Table 1. A number of
the compounds have IC50 values in the low micromolar range with 42% of these figures being less than
10 µM. The cell lines vary in their sensitivity to these dienones. Thus the average IC50 values of the
compounds in series 3 -5 towards Molt4/C8, CEM and L1210 cells are 11.3, 39.0 and 52.2 µM,
respectively. This cell line specificity may indicate compounds with greater toxicities towards
neoplasms than nonmalignant cells, i.e., they may display tumour-specific toxicity and this possibility
was examined, vide infra.
Melphalan is an alkylating agent used in cancer chemotherapy and a comparison was made
between its potencies and the compounds in series 3 – 5 which were designed as thiol alkylators. The
biodata generated in the Molt4/C8 bioassay revealed that the enones 3a and 3c are 2.00 and 2.72 times
more potent than melphalan, while 3b is equipotent with this established drug.
An examination was made of the biodata in terms of finding how cytotoxic potencies varied
with alterations in the structure of the compounds. Previous disclosures revealed the IC50 values of 2a-
d9 and 2e15 in the Molt4/C8, CEM and L1210 evaluations. These values are presented in the
supplemental section. A comparison was made between the potencies of 2a-e with the analog in series
3 which has the same aryl substituent in ring A. Thus the IC50 value of 2a was compared with 3a in the
Molt4/C8 bioassay and so forth with standard deviations being taken into account. These comparisons
are expressed as Δ2a-e / Δ3a-e figures in Table 1. The introduction of a 3-pyridylcarbonyl moiety to the
compounds in series 2 leads to the analogs 3a-e with retention of cytotoxic potencies. In the Molt4/C8
bioassay, 3a and 3c have increased potency compared to 2a and 2c in this screen while 3b (Molt4/C8
assay), 3a (CEM screen) and 3b, c (L1210 bioassay) are equipotent with the analogs in series 2. Clearly
3a-c are more potent than 3d, e and, as indicated in Table 1, 3a-c have average IC50 values similar to
2a-c.
The question arises whether the relative potencies of 3a-e, 4a,b and 5 are the same in each of
the three bioassays. If this is the case, then the average IC50 values may be calculated and comparisons
made between the potencies of the dienones. Alternatively if the relative potencies vary, then
comments on structure-activity relationships need to be made for each cell line. In order to address this
problem, Kendall’s coefficient of concordance16 was used. In this analysis, if the relative potencies are
identical, then Kendall’s coefficient of concordance will be 1; conversely if there is no correlation, the
coefficient will be 0. In this investigation, the coefficient of concordance is 0.947 (p = 0.006). Hence
the conclusion drawn is despite the variation in sensitivities of the cell lines to the compounds, the
chemical and physicochemical properties of 3a-e, 4a,b and 5 influence potencies in a similar fashion.
Replacement of the 3-pyridyl group of 3a, b by a phenyl ring led to 4a, b. The aryl ester 4a has
lower potency than 3a in the three screens, 3b and 4b are equipotent in the Molt4/C8 and CEM
bioassays and 4b has a higher IC50 value than 3b in the L1210 screen. The logP values of 3a, b and 4a,
b are 5.26, 5.71, 6.84 and 7.28, respectively. Hence the lower IC50 values of 3a, b in general may have
been due, at least partially, to their lower lipophilicity. Hence future structural modifications should
retain the 3-pyridylcarbonyl group in aryl ring B in this cluster of compounds. The replacement of the
small groups attached to ring A of 3a-e by a 3-pyridylcarboxy group present in 5 lowered potencies.
Hence among series 3 -5, it is the niacin esters 3 which should be pursued further and in particular 3a-
c. Subsequent discussion and experimentation therefore is focussed principally on series 3.
The aryl substituents present in series 3 are found in all four quadrants of a Craig diagram for
para substituents.17 Linear and semilogarithmic plots were made between the Hammett σ, Hansch π
and molar refractivity (MR) constants of the aryl substituents of 3a-e and the IC50 values generated in
the Molt4/C8, CEM and L1210 evaluations. Positive correlations (p<0.05) were noted between the σ
values of the aryl substituents present in 3a-e and the IC50 values in the Molt4/C8, CEM and L1210
evaluations. No correlations or trends to a correlation (p<0.1) were noted between the cytotoxic
potencies and the π and MR constants. Hence in the future groups with strongly electron-donating
5

Table 2
Evaluation of the compounds in series 3 for their cytotoxicity against human tumour and human nonmalignant cells
Human Tumour Cell Lines
Compound
Human Non-malignant Cells, CC50 (μM)a
HL-60
SIb
HSC-2
SIb
HSC-3
SIb
HSC-4
SIb
Ave
CC50
11.9
Ave
SIb
23.0
HGF
HPC
HPLF
Ave
CC50 (μM)a
CC50 (μM)a
CC50 (μM)a
CC50 (μM)a
1.5±0.61
1.9±0.31
2.3±0.21
10.5±1.1
>200
70.7
12.5±0.86 8.48
17.7±1.5
98±2.5
16.0±1.0
33.7±1.5
>200
5.99
15.9±0.9
6.67
116±15.1
184±4.0
>200
190±10
>200
142±16
>200
>200
179±18
>200
>200
60±31
163±15
>200
154±9.3
>200
106
3a
3b
3c
3d
3e
>95.8 90.3±16.0 >2.02
>87.0 30.3±13.9 >6.60
16.6
~1.0
~1.0
75.6
>1.86 94.3±12.0 >1.93 71.1 >25.4
>182
>200
174
>200
>200
67.3
>12.5
5.16
~1.0
~1.0
1.46
22.7±1.5
37.3±1.5
>200
>200
46±6.0
>8.81 17.8 >28.7
30.5±10.0 5.71
112±12.5 >1.79
4.67
~1.0
~1.0
1.46
28.0
>178 ~1.20
>200
26.2
8.04
Niacin
Melphalan
>200
0.89±0.17
>200
12±2.7
~1.0
5.61
>200
46±31
~1.0
21.0
>200
73±15
>200
92±7.8
37±7.3
aThe CC50 values represent the concentrations of the compounds which kill 50% of the cells.
bThe letters SI refer to the selectivity index figures. The values are obtained by dividing the average CC50 figures of the non-malignant cells by the CC50 value
of a specific malignant cell line.
6

properties should be placed in ring A such as the 4-dimethylamino and 4-isopropyloxy group which
have σ values of -0.83 and -0.45, respectively.18
The next phase of the investigation was to determine whether the compounds in series 3 display
tumour-selective cytotoxicity, i.e., do they have greater potency towards neoplasms than non-malignant
cells? Hence 3a-e were evaluated against HL-60, HSC-2, HSC-3 and HSC-4 neoplastic cell lines as
well as HGF, HPC and HPLF nonmalignant cells. These data are presented in Table 2.
The biodata in Table 2 reveal that in general the compounds in series 3 are lethal to HL-60,
HSC-2, HSC-3 and HSC-4 cells. The relative average potencies of 3a-e are 3a > 3c > 3d > 3b > 3e
which are similar to the average potencies in Table 1 viz 3a > 3c > 3b > 3d > 3e, i.e., the greatest
cytotoxicity is displayed by 3a and 3c while 3e is the least potent. In particular, 3a, c, d are more potent
than melphalan towards HSC-4 cells while the following esters are equipotent with this drug (cell line
in parentheses) namely 3a, b (HL-60), 3a (HSC-2) and 3a, c, d (HSC-3).
The compounds in series 3 were also evaluated against HGF, HPC and HPLF non-malignant
cells. The low average CC50 values of 3b-e are noteworthy and reveal that in general, series 3 is well
tolerated by nonmalignant cells. In fact, in 47 % of the bioassays, the CC50 values are in excess of 200
µM. All of these esters are less toxic than melphalan for these three cell lines except 3a and melphalan
have the same toxicity towards HPLF cells. In order to quantify the greater toxicity of 3a-e for
neoplasms than normal cells, selectivity index (SI) figures were calculated. Individual tumours are
surrounded by different nonmalignant cells and hence the SI values are the quotients of the average
CC50 value towards HGF, HPC and HPLF cells and the CC50 figure of the compound against a specific
neoplastic cell line. The SI figures of 3a-c generated using HL-60 cells are huge and in addition these
three compounds have SI values in excess of 1 against HSC-2, HSC-3 and HSC-4 cells. The average SI
values of 3a-c are greater than is displayed by melphalan. Furthermore the virtual lack of potency of
the 4-nitro analog 3e to both neoplastic and nonmalignant cells is noteworthy.
In order to identify compounds which have both good potencies and are selectively toxic to
neoplasms, the potency-selectivity expression (PSE) values of the compounds in series 3 were
calculated. This property is the product of the reciprocal of the average CC50 value and the average SI
figure multiplied by 100.19 Thus the PSE values of 3a-e are 193, >35.7, >161, 28.7 and >0.67,
respectively, in contrast to a figure of 80.2 for melphalan. These data reinforce the identification of 3a
and 3c as important lead molecules.
Figure 2. Evaluation of 3a (7µM) for PARP1 cleavage in HSC-2 and HSC-3 cells for 24 h.
An investigation was launched to gain some idea of the way in which a lead compound 3a
exerts its lethal effect on various neoplasms. The enzyme poly(ADP-ribose)polymerase 1 (PARP1)
repairs DNA single-stranded breaks20 and hence compounds which cleave PARP1 may be of
considerable value in cancer chemotherapy. The ability of 3a to cleave PARP1 in HSC-2 and HSC-3
cells was investigated and the results are depicted in Figure 2. This compound caused PARP1 cleavage
in HSC-2 cells and hence its antineoplastic effect against this cell line is partially due to this mode of
action. Under the same conditions, 3a did not cause cleavage of PARP1 in HSC-3 cells which may
contribute to this compound displaying greater toxicity to HSC-2 cells than to the HSC-3 neoplasms.
7

Another way in which anticancer drugs exert their bioactivity is by interfering with different stages of
the cell cycle. Consequently HSC-2 cells were incubated with 3a (7µM) for 24 hours and the
distribution of the cells determined at the end of the time period. The percentage of cells in the G0/G1, S
and G2/M phases (the figures for untreated cells are in parentheses) are 86 (73), 9.7 (13) and 1.5 (11),
respectively. Thus another way that 3a exerts its antineoplastic effect on HSC-2 cells is by causing
G0/G1 phase arrest. In summary, the cytotoxicity of 3a, and probably structurally related compounds,
includes PARP1 cleavage and cell cycle arrest.
A major problem in treating cancer patients with drugs is the resultant toxicity of these
compounds.21 Hence the stimulus for developing these compounds will be enhanced if they are well-
tolerated in mammals. Consequently short-term toxicology experiments were initiated. All of the
compounds in series 3 – 5 were administered intraperitoneally to mice using doses of 30, 100 and 300
mg/kg and the animals were examined for survival and neurological deficit after 0.5 and 4 hours. No
mortalities were observed and after 0.5 hour only 3a caused neurotoxicity in less than half of the
animals receiving a dose of 300 mg/kg. These symptoms were absent after 4 hours. Doses of 50 and
30 mg/kg of 3a and 3b, respectively, were administered orally to rats and the animals examined after
0.25, 0.5, 1, 2 and 4 hours did not reveal any overt toxicity. Bearing in mind that the LD50 value of
melphalan is 21.7 mg/kg in mice21, these animal data for series 3–5 are encouraging.
Table 3
Evaluation of 3a-c for drug-like properties
Compound
Physicochemical parametera
M.W.
logP
HBA
HBD
RB
TPSA
395.46
5.26
4
3a
0
0
0
5
6
7
56.27
56.27
65.50
<140
409.48
425.48
<500
5.21
5.32
4
5
3b
3c
Ideal
<5.00
<10
<5
<10
compound
aMW: molecular weight, log P: (the logarithm of the partition coefficient), HBA: the number
of hydrogen bond acceptor atoms, HBD: the number of hydrogen bond donor atoms, RB: the
number of rotatable bonds (RB) and TPSA: the total polar surface area (TPSA). The characteristics
of an ideal compound are based on guidelines proposed in the literature.22,23
In view of the promising biodata, drug-likeness properties of 3a-c were assessed. Various
studies have revealed that in general for good oral absorption, organic molecules should have a
molecular weight of less than 500, a logP value below 5, less than 10 hydrogen bond acceptor atoms
and fewer than 5 hydrogen bond donor atoms.22 In addition, the number of rotatable bonds should be
less than 10 and the total polar surface area should be below 140Å.23 The relevant physicochemical
parameters of 3a-c are presented in Table 3 which reveals that the compounds meet the guidelines
except the logP values are above 5. In fact logP figures of 0-3 are considered to be optimal for oral
bioavailability.24 Hence future development will need to introduce hydrophilic groups into the
molecule as well as electron-repelling groups vide supra. Examples of suitable aryl substituents which
have both hydrophilic and electron-donating properties (π and σ constants in parentheses) are hydroxyl
(-0.67, -0.37), amino (-1.23, -0.66) and methylamino (-0.47, -0.84).25
Finally, various experiments were performed to assess the likelihood of the compounds in series
3 undergoing hydrolysis. A representative compound 3a was incubated under different conditions for
48 hours and at 37° C which are the time and temperature used in the in vitro assays. First, a solution of
1
3a was incubated in a mixture of deuterium oxide and deuterated dimethylsulfoxide. However the H
NMR spectrum revealed the presence of unreacted 3a. Second, 3a was incubated in a mixture of buffer
8

pH 7.4 and dimethylsulfoxide and with the aid of a water suppressor technique, the 1H NMR spectrum
revealed only the presence of intact 3a. Third, 3a was incubated in a mixture of rat serum and
dimethylsulfoxide and subsequently extracted with ethyl acetate. The experiment was repeated in the
absence of 3a and both extracts were examined by ¹H NMR spectroscopy. By creating difference
spectra, unreacted 3a was detected but not the hydrolysis product 6-benzylidene-2-(4-
hydroxybenzylidene)cyclohexanone. One or more products were also observed which may have been
due to the reaction of 3a with compounds in the rat serum. While an exhaustive probe as to the stability
of 3a and related compounds is beyond the span of this study, the evidence to hand suggests that in the
future electron attracting substituents should be placed in the aryl ring B to enhance hydrolysis.
In conclusion, three novel series of compounds 3-5 were developed as candidate antineoplastic
agents and among these derivatives 3a-c display excellent cytotoxic properties towards Molt4/C8,
CEM and L1210 cells. In addition, these dienones have CC50 values of approximately 2 µM towards
human HL-60 leukemia cells and substantially lower toxicity to HGF, HPC and HPLF nonmalignant
cells. SAR and QSAR studies reveal that cytotoxic potencies are enhanced by the incorporation of
electron-donating groups into ring A of series 3. A representative potent compound 3a cleaved PARP1
and caused cell cycle arrest in HSC-2 cells. The compounds in series 3 as well as 4 and 5 are well
tolerated in mice and in general possess drug-like properties. Three different stability studies of a
representative compound 3a revealed that it did not undergo hydrolysis to the corresponding phenol 2a.
Various suggestions for future analog development have been outlined.
Acknowledgements
The authors thank the following agencies for their financial support of this study namely the
Canadian Institutes of Health Research for a grant to J. R. Dimmock, the Ministry of Education,
Science, Sports and Culture for a grant-in-aid to H. Sakagami and the Belgian Concerted Research
Actions (GOA 10/014) who provided funds to J. Balzarini. L. van Berckelaer kindly undertook the
bioevaluations presented in Table 1. Mr. J. P. Stables and some of the staff at the National Institute of
Neurological Disorders and Stroke, USA are thanked for undertaking the in vivo evaluation of 3-5 in
mice and 3a and 3b in rats.
Supplementary data
Supplementary data are available which describes the syntheses of 3a-e, 4a,b and 5, the
methods for determining physicochemical properties of the compounds and bioevaluations.
9

References and notes
1. Mutus, B.; Wagner, J. D.; Talpas, C. L.; Dimmock, J. R.; Phillips, O. A.; Reid, R.S. Anal. Biochem.
1989, 177, 237-243.
2. Baluja, G.; Municio, A, M.; Vega, S. Chem. Ind. 1964, 2053-2054.
3. Schaefer, W. H.; Harris, T. M.; Guengerich, F. P. Arch. Biochem. Biophys. 1987, 257, 186-193.
4. Boffetta, P.; Kaldor, J.M. Acta Oncol. 1994, 33, 591-598.
5. Levine, E. G.; Bloomfield, C.D. Semin. Oncol. 1992, 19, 47-84.
6. Chen, G.; Waxman, D. J. Biochem. Pharmacol. 1994, 47, 1079-1087.
7. Tsutsui, K.; Komuro, C.; Ono, K.; Nishidia, T.; Shibamoto, Y.; Takahashi, M.; Abe, M. Int. J.
Radiat. Oncol. Biol. Phys. 1986, 12, 1183-1186.
8. Dimmock, J.R.; Padmanilayam, M.P.; Zello, G. A.; Nienaber, K. H.; Allen, T.M.; Santos, C. L.; De
Clercq, E.; Balzarini, J.; Manavathu, E. K.; Stables, J. P. Eur. J. Med. Chem. 2003, 38, 169-177.
9. Dimmock, J. R.; Kumar, P.; Nazarali, A.J.; Motaganahalli, N. L.; Kowalchuk, T.P.; Beazely, M. A.;
Quail, J. W.; Oloo, E. O.; Allen, T. M.; Szydlowski, J.; De Clercq, E.; Balzarini, J. Eur. J. Med.
Chem. 2000, 35, 967-977.
10. Dreizen, S.; McCredie, K. B.; Keating, M. J.; Andersson, B. S. Postgrad. Med. 1990, 87, 163-167.
11. McCarter, D. N.; Holbrook, J. In Textbook of Therapeutics Drug and Disease Management, 7th
edition.; Herfindal, E. T.; Gourley, D. R. Eds., Lipincott Williams and Wilkins, Philadelphia, 2000,
p. 183.
12. Bartleman, A. P.; Jacobs, R.; Kirkland, J. B. Nutr. Cancer, 2008, 60, 251-258.
13. Dimmock, J. R.; Sidhu, K. K.; Chen, M.; Li, J.; Quail, J. W.; Kao, G. Y. J. Pharm. Sci. 1994, 83,
852-858.
14. Vieweg, H.; Wagner, G. Pharmazie 1979, 34, 785-788.
15. Das, U.; Gul, H. I.; Alcorn, J.; Shrivastav, A.; George, T.; Sharma, R. K.; Nienaber, K. H.; De
Clercq, E.; Balzarini, J.; Kawase, M.; Kan, N.; Tanaka, T.; Tani, S.; Werbovetz, K. A.; Yakovich,
A. J .; Manavathu, E. K.;. Stables, J. P.; Dimmock, J. R. Eur. J. Med. Chem. 2006, 41, 577-585.
16. Sheshkin, D.; Handbook of parametric and nonparametric statistical procedures, Chapman and
Hall, London, 2004 pp. 1093-1107.
17. Craig, P. N. J. Med. Chem. 1971, 14, 680-684.
18. Hansch, C.; Leo, A. J. Substituent Constants for Correlation Analysis in Chemistry and Biology,
John Wiley and Sons, New York, 1979, p.50.
19. Das, S.; Das, U.; Sakagami, H.; Umemura, N.; Iwamoto, S.; Matsuta, T.; Kawase, M.; Molnár, J.;
Serly, J.; Gorecki, D. K. J. ; Dimmock, J. R. Eur. J. Med. Chem. 2012, 51, 193-199.
20. Malanga, M.; Althaus, F. R. Biochem. Cell Biol. 2005, 83, 354-364.
21. Quinn, F. R.; Milne, G. W. A. Fundam. Appl. Toxicol. 1986, 6, 270-277.
22. Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P. J. Adv. Drug Del. Rev. 1997, 23 3-25.
23. Veber, D.F.; Johnson, S. R.; Cheng, H. –Y.; Smith, K. W.; Ward, Kopple, K.D. J. Med. Chem.
2002, 45, 2615-2623.
24. Kerns, E. H.; Di, L. Drug-like properties: concepts, structure design and methods: from ADME to
toxicity optimization, Academic Press, Burlington, MA, 2008, p.45.
25. Hansch, C.; Leo, A. J. Substituent Constants for Correlation Analysis in Chemistry and Biology,
John Wiley and Sons, New York, 1979, p.49.
10

GRAPHICAL ABSTRACT
11

Highlights
 Novel series of cytotoxins displaying tumour-selective toxicity.
 Development of SAR and QSAR correlations.
 The compounds possess drug-like properties.
 The compounds are well tolerated in a short term screen in mice.
 The mode of actions include PARP1 cleavage and G0/G1 cell cycle arrest.
12