Cink4T, a quinazolinone-based dual inhibitor of Cdk4 and tubulin polymerization, identified via ligand-based virtual screening, for efficient anticancer therapy

Article abstract of DOI:10.1016/j.ejmech.2019.01.011

Inhibition of cyclin dependent kinase 4 (Cdk4) prevents cancer cells from entering the early G₀/G₁ phase of the cell division cycle whereas inhibiting tubulin polymerization blocks cancer cells’ ability to undergo mitosis (M) late in the cell cycle. We had reported earlier that two non-planar and relatively non-toxic fascaplysin derivatives, an indole and a tryptoline, inhibit Cdk4 with IC₅₀ values of 6.2 and 10 μM, respectively. Serendipitously, we had also found that they inhibited tubulin polymerization. The molecules were efficacious in mouse tumor models. We have now identified Cink4T in a 59-compound quinazolinone library, designed on the basis of ligand-based virtual screening, as a compound that inhibits Cdk4 and tubulin. Its IC₅₀ value for Cdk4 inhibition is 0.47 μM and >50 μM for inhibition of Cdk1, Cdk2, Cdk6, Cdk9. Cink4T inhibits tubulin polymerization with an IC₅₀ of 0.6 μM. Molecular modelling studies on Cink4T with Cdk4 and tubulin crystal structures lend support to these observations. Cancer cell cycle analyses confirm that Cink4T blocks cells at both G₀/G₁ and M phases as it should if it were to inhibit both Cdk4 and tubulin polymerization. Our results show, for the very first time, that virtual screening can be used to design novel inhibitors that can potently block two crucial phases of the cell division cycle.

Full text of DOI:10.1016/j.ejmech.2019.01.011

Accepted Manuscript
Cink4T, a quinazolinone-based dual inhibitor of Cdk4 and tubulin polymerization,
identified via ligand-based virtual screening, for efficient anticancer therapy
Vinay Sonawane, Mohd Usman Mohd Siddique, Surender Singh Jadav, Barij Nayan
Sinha, Venkatesan Jayaprakash, Bhabatosh Chaudhuri
PII:
S0223-5234(19)30017-0
DOI:
https://doi.org/10.1016/j.ejmech.2019.01.011
EJMECH 11018
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 18 October 2018
Revised Date: 5 January 2019
Accepted Date: 5 January 2019
Please cite this article as: V. Sonawane, M.U. Mohd Siddique, S.S. Jadav, B.N. Sinha, V. Jayaprakash,
B. Chaudhuri, Cink4T, a quinazolinone-based dual inhibitor of Cdk4 and tubulin polymerization,
identified via ligand-based virtual screening, for efficient anticancer therapy, European Journal of
Medicinal Chemistry (2019), doi: https://doi.org/10.1016/j.ejmech.2019.01.011.
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.

ACCEPTED MANUSCRIPT
Graphical Abstract
M
G /M
G0
2
Tubulin
Polymerization
G2
G /G
0 1
G1
Cdk4/
CyclinD
S
GLU94
Cink4T
HIS95
GLU144
VAL96
ASP97
ARG101
THR102

ACCEPTED MANUSCRIPT
Cink4T, a quinazolinone-based dual inhibitor of Cdk4 and tubulin
polymerization, identified via ligand-based virtual screening, for efficient
anticancer therapy
a, †
b, †
c
Vinay Sonawane, Mohd Usman Mohd Siddique, Surender Singh Jadav, Barij Nayan
b
b,*
a,*
Sinha, Venkatesan Jayaprakash, Bhabatosh Chaudhuri
a
Leicester School of Pharmacy, De Montfort University, Leicester, LE1 9BH, UK
b
Department of Pharmaceutical Sciences Technology, Birla Institute of Technology, Mesra,
Ranchi 835215, India
c
CSIR-Indian Institute of Chemical Technology, Tarnaka, Hyderabad 500007, India
†
Contributed equally to this work.
*
Authors for correspondence.
Venkatesan Jayaprakash (VJ): Email: venkatesanj@bitmesra.ac.in; Tel: +91-9470137264
Bhabatosh Chaudhuri (BC): Email: bchaud00@gmail.com; Tel: 00-44-7805230121
Declaration of interest: None
1
| P a g e

ACCEPTED MANUSCRIPT
Abstract
Inhibition of cyclin dependent kinase 4 (Cdk4) prevents cancer cells from entering the
early G /G phase of the cell division cycle whereas inhibiting tubulin polymerization blocks
0
1
cancer cells’ ability to undergo mitosis (M) late in the cell cycle. We had reported earlier that
two non-planar and relatively non-toxic fascaplysin derivatives, an indole and a tryptoline,
inhibit Cdk4 with IC values of 6.2 and 10 µM, respectively. Serendipitously, we had also
5
0
found that they inhibited tubulin polymerization. The molecules were efficacious in mouse
tumor models. We have now identified Cink4T in a 59-compound quinazolinone library,
designed on the basis of ligand-based virtual screening, as a compound that inhibits Cdk4 and
tubulin. Its IC value for Cdk4 inhibition is 0.47 µM and >50 µM for inhibition of Cdk1,
5
0
Cdk2, Cdk6, Cdk9. Cink4T inhibits tubulin polymerization with an IC of 0.6 µM. Molecular
5
0
modelling studies on Cink4T with Cdk4 and tubulin crystal structures lend support to these
observations. Cancer cell cycle analyses confirm that Cink4T blocks cells at both G /G and
0
1
M phases as it should if it were to inhibit both Cdk4 and tubulin polymerization. Our results
show, for the very first time, that virtual screening can be used to design novel inhibitors that
can potently block two crucial phases of the cell division cycle.
Keywords: Quinazolinone; Dual inhibitor; Cdk4; Tubulin polymerization; G /G arrest;
0
1
G /M arrest
2
2
| P a g e

ACCEPTED MANUSCRIPT
1. Introduction
Cyclin-dependent kinases (Cdks) are serine/threonine-specific kinases that regulate cell
cycle check points through initiation of phosphorylation of proteinaceous substrates, using
adenosine triphosphate (ATP) as a phosphate donor [1]. In order to maintain homeostasis of
normal cells undergoing cellular proliferation, activities of Cdks are tightly controlled. Any
abnormality in the normal regulation of the cell cycle process leads to uncontrolled cell
growth, which is often regarded as one of the hallmarks of cancer cells [2, 3]. Interruption of a
deregulated cell cycle, as a consequence of overactivation of a Cdk, has been accepted as a
strategy for treatment of proliferative diseases, such as cancer [4-7].
Identification of several small molecule inhibitors of the Cdk family of kinases and their
roles in the treatment of diseased states have been published. Several pharmacophores have
been reported as potent Cdk1, Cdk2, Cdk4 and/or Cdk9 inhibitors. Flavopiridol (alvocidib)
was one of the first-generation ATP-competitive Cdk inhibitors tested in clinical trials [8].
Olomoucine [9], roscovitine [10], purvalanol [11], paullones [12], butyrolactone [13],
oxindoles [14], aminothiazoles [15], benzocarbazoles [16] and pyrimidine [17] derivatives
have also been recognized as Cdk inhibitors. Recently, palbociclib has been approved as part
of a combination of two drugs that could be used simultaneously for patients with estrogen
receptor positive advanced breast cancer [18].
Sometime ago we had found that the natural pigment fascaplysin, isolated from the marine
sponge Fascaplysinopsis Bergquist sp [19], was a potent inhibitor of Cdk4 [20]. It blocks the
cell cycle at the G /G phase in both normal and tumor cells [20]. However, it was thought
0
1
that the planar fascaplysin molecule would be of limited use as an anticancer agent because of
its ability to intercalate DNA which causes high toxicity [21]. In our attempts to find non-
planar derivatives of fascaplysin, which do not intercalate DNA yet inhibit Cdk4, we had
3
| P a g e

ACCEPTED MANUSCRIPT
identified indole and tryptoline derivatives as Cdk4 inhibitors [22-24]. When performing cell
cycle analyses with some representative compounds, we had fortuitously found that certain
non-planar analogues of fascaplysin also prevent tubulin polymerization [25, 26]. These
molecules not only block at the G /G phase of the cell cycle due to Cdk4 inhibition, but also
0
1
at the G /M phase due to inhibition of tubulin polymerization. Since they showed high
2
efficacy in mouse tumor models, we had commented that dual inhibitors of Cdk4 and tubulin
polymerization, which block cells at two distinctive phases of the cell cycle, should be
invaluable in the treatment of cancer [25, 26]. Cdk4 and tubulin polymerization govern the
two most important checkpoints (i.e. at the G /G and G /M phases) in the cell division cycle.
0
1
2
The Cdk4 enzyme is considered a target for anticancer therapy because it is deregulated in
most, if not all, cancer cells [4-18]. The importance of tubulin polymerization has been
exemplified by the regulatory approval of tubulin binding molecules for clinical use,
molecules such as the family of compounds known as taxanes and the general use of the vinca
alkaloids (one of which is vinblastine) in anticancer therapy [27-30].
Here, we report the construction of a chemical library of 59 quinazolinone derivatives
designed on the basis of virtual screening which employed as templates (i.e. ‘queries’) two
molecules which have shown dual inhibitory potential to Cdk4 enzyme and tubulin
polymerization [25, 26]. The aim of virtual screening was to identify molecules that may have
the potential to bind simultaneously to both Cdk4 and tubulin so that, on the basis of
identified structures, chemical libraries could be synthesized to seek molecules that would
empirically be capable of inhibiting two crucial targets in the cancer cell division cycle. A
Cdk4-cyclin D1 protein kinase assay was initially used to determine whether there were any
compounds in a chemical library of quinazolinones, synthesized on the basis of results
obtained from virtual screening, which inhibited Cdk4 enzyme. A quinazolinone derivative
(
i.e. compound 40; Table 1) was found that inhibits Cdk4 with an IC value of 0.47 µM. The
50
4
| P a g e

ACCEPTED MANUSCRIPT
IC values for inhibition of Cdk2-cyclin A2, Cdk1-cyclin B1, Cdk6-cyclin D1 and Cdk9-
5
0
cyclin T1 were found to be >50 µM. Compound 40 also inhibits tubulin polymerization with
an IC of 0.6 µM. We have named compound 40 as Cink4T (i.e. Compound that inhibits
5
0
Cdk4 and Tubulin polymerization) in analogy to Cink4, a Cdk4-specific inhibitor with an IC50
value of 1.5 µM that we had identified through a high-throughput screen and which was also
shown to be efficacious in in vivo mouse tumor models [31]. Molecular modelling data tend
to support our experimental findings that Cink4T is indeed a dual inhibitor of Cdk4 and
tubulin.
At 1x EC concentration (EC being the concentration of a drug that gives half-maximal
50
50
response, in this case inhibition of cancer cell growth), Cink4T blocks growth of pRb
retinoblastoma protein)-positive T47D breast cancer cells at the G /G phase of the cell
(
0
1
cycle. T47D cells are deficient in the mitotic-spindle checkpoint (MSC)[32], MSC
representing the G /M cell cycle checkpoint that is affected by drugs that interfere with
2
microtubule (i.e. tubulin) polymerization [33, 34]. Cells deficient in MSC are totally
unaffected by such drugs. Hence, Cink4T’s inhibitory effects on tubulin polymerization
would not be manifested in T47D cells. However, as would be expected of a Cdk4 inhibitor
which mediates its cell proliferative function through phosphorylation of pRb [35, 36],
Cink4T arrests pRb-positive T47D cells at the early G /G phase of the cell cycle and also
0
1
inhibits Cdk4-specific pRb phosphorylation which is responsible for progression of cells from
G /G , through the restriction point, further into the cell cycle [37-39]. Cink4T also reflects
0
1
the ability of a true Cdk4 inhibitor by maintaining a block at G /G in serum-starved T47D
0
1
and Calu-1 non-small cell lung carcinoma (NSCLC) cells, both cell types being pRb-positive
but MSC-deficient [32, 40]. In stark contrast, we see that Cink4T is unable to arrest MDA-
MB-468 breast cancer cells, which are pRb-null but MSC-deficient [41], at G /G even at 2x
0
1
EC concentration, since the absence of cellular pRb renders a Cdk4 inhibitor (i.e. Cink4T)
50
5
| P a g e

ACCEPTED MANUSCRIPT
non-functional and, thus, completely ineffective. In parallel, in MSC-proficient and pRb-
positive A549 NSCLC cells [40], at 1x EC concentration, Cink4T reveals its Cdk4 and
50
tubulin dual inhibitory potential by trying to block cells at both G /G and G /M. However, a
0
1
2
G /M phase block is mainly seen at 2x EC concentration reflecting Cink4T’s ability to
2
50
inhibit tubulin polymerization.
The results presented here demonstrate, for the very first time, that virtual screening based
on already discovered templates [25, 26] can surprisingly lead to the identification of novel
molecules that can potently inhibit two crucial phases of the cancer cell division cycle.
2. Results and Discussion
2
.1. Virtual screen of quinazolinones that could potentially identify a dual inhibitor of
Cdk4 and tubulin polymerization. Shape and electrostatic similarity based virtual screening
of ligands can be used as tools in drug discovery, from enriching a huge database of
compounds to identification of potential ‘hits’ [42-45]. It has been claimed that ligand-based
virtual screening, used for the identification of dual inhibitors of human immunodeficiency
virus-1 (HIV-1) reverse transcriptase and ribonuclease H activities, was the first application
of such an approach in the context of HIV-1 research [46].
We have already reported that certain non-planar analogues of fascaplysin can act as dual
inhibitors of two important targets of the cancer cell division cycle, the Cdk4 enzyme and
tubulin polymerization [25, 26]. Originally, we had only been seeking non-planar molecules,
derived from fascaplysin, which would have the ability to inhibit Cdk4 specifically [47].
Quite unintentionally, we had found that some molecules that inhibit Cdk4 also inhibit
polymerization of tubulin. These molecules include CA224 [25, 48] and CA199 [49], which is
6
| P a g e

ACCEPTED MANUSCRIPT
an analogue of BPT [26]. They do not possess fascaplysin’s DNA intercalating property [21]
and, hence, are devoid of the toxicity that is associated with fascaplysin.
In an attempt to identify new chemical scaffolds that simultaneously target inhibition of
Cdk4 enzyme and tubulin polymerization, shape and electrostatic similarity based virtual
screening was performed using OpenEye tools, OMEGA, ROCS and EON. A drug-like
subset in the ZINC13 database, having 13,195,609 molecules, was screened with the
OpenEye tools. The two non-planar fascaplysin analogues, CA199 [49] and CA224 [25, 48]
(depicted in Figure 1), already reported as dual inhibitors of Cdk4 enzyme and tubulin
polymerization, were used as templates (i.e. queries). The ‘hits’ obtained with both the
queries were merged and the top 30,000 molecules were picked to provide an enriched library
for further molecular docking studies. Protein Preparation Wizard, Ligprep and Glide
modules from Maestro-8.5 (Schrodinger LLC) were used to run molecular docking
simulation. The enriched-library was subjected to molecular docking against ATP biding site
of Cdk4 using X-ray crystal structure of the Cdk4 mimic of Cdk2 (PDB: 1GII)[50] using the
XP-protocol in the Glide module. The top 10% ‘hits’ were then selected for molecular
docking against the colchicine ATP-binding pocket of tubulin using the X-ray crystal
structure of tau-tubulin kinase in complex with a small molecule inhibitor (PDB: 4BTK) [51].
The top ten ‘hits’, from molecular docking against both Cdk4 and tubulin, are also shown in
Figure 1. They belong to seven different chemical scaffolds which, structurally, are quite
different from the fascaplysin analogues reported by us earlier.
2
.2. Syntheses of the quinazolinone chemical library. Compounds 5 and 9 (shown in
Figure 1), possessing the quinazolinone nucleus, were amongst the top ten ‘hits’ that showed
affinity towards both Cdk4 and tubulin in the virtual screen. In a completely separate drug
design program, we had previously identified the quinazolinone scaffold [52] as a mimic of α-
7
| P a g e

ACCEPTED MANUSCRIPT
naphthoflavone (ANF), using isosteric replacement strategy. In vitro, ANF is known to inhibit
potently both CYP1A1 and CYP1B1 isozymes [53] belonging to the drug-metabolizing
cytochrome P450 family of enzymes [54, 55]. On the basis of such a screen, we had reported
that certain quinazolinone derivatives, based on compounds 5 and 9 (Figure 1), can act as
potent and specific inhibitors of CYP1B1 enzyme which is known to have a role in the
progression of different tissue-specific cancers [52].
A library of fifty-nine compounds, twenty (14-20 and 22-34) which had been reported
earlier [52] and thirty-nine new compounds, were synthesized as per the reactions outlined in
Scheme 1. All the quinazolinone derivatives were formed through condensation, in DMF, of
anthranilamide with a corresponding benzaldehyde in the presence of sodium acetate, as base,
and iodine as oxidizing agent. Benzaldehydes used for the synthesis of compounds 14-34
(Table 1) were synthesized as reported earlier [52]. Benzaldehydes required for the synthesis
of compounds 35-72 (Table 1) were synthesized by refluxing, in acetone, respective 2-chloro-
N-aryl-aceatamides with corresponding hydroxy benzaldehydes (3-hydroxybenzaldehyde, 4-
hydroxybenzaldehyde, vanillin or isovanillin) in the presence of potassium carbonate and
potassium iodide. The products thus obtained were purified by washing with an ethyl
acetate:petroleum ether (40:60) mixture. Structures of the synthesized compounds (Table 1)
1
13
were confirmed by H-NMR, C-NMR and MS spectral data (see 4. Experimental section).
8
| P a g e

ACCEPTED MANUSCRIPT
Figure 1. (A) The two ‘queries’ used for virtual screening. (B) The ten ‘hits’ obtained from
virtual screening.
Scheme 1. A general scheme for the syntheses of compounds 14-72.
9
| P a g e

ACCEPTED MANUSCRIPT
Table 1. Structure of quinazolinone derivatives (i.e. compounds 14 to 72) and their
percentage inhibition of Cdk4-cyclin D1 protein kinase, at 10 µM concentration of each
compound.
%
inhibition of Cdk4-cyclin D1
Code
R
R1
X
a
at 10 µM
1
1
1
1
4
5
6
7
-H
-NO
-Br
-H
-H
-H
-H
N
36ꢀ±ꢀ1.6
0ꢀ±ꢀ0.3
31ꢀ±ꢀ1.5
30ꢀ±ꢀ1.2
2
CH
-OH
CH
CH
1
8
-H
13ꢀ±ꢀ0.5
1
9
-H
CH
CH
30 ±ꢀ1.2
2
0
-H
42ꢀ±ꢀ1.7
2
2
1
2
-H
-H
CH
CH
35ꢀ±ꢀ1.4
30ꢀ±ꢀ1.2
2
2
2
3
4
5
-H
-OCH
-OH
-OH
-OH
CH
CH
CH
-6ꢀ±ꢀ0.5
0ꢀ±ꢀ0.3
39ꢀ±ꢀ1.2
3
-OCH
3
2
6
-OCH
3
CH
55ꢀ±ꢀ1.8
2
7
-OCH
3
3
CH
34ꢀ±ꢀ1.1
2
8
-OCH
CH
36ꢀ±ꢀ1.3
2
3
9
0
-OCH
3
3
CH
CH
78ꢀ±ꢀ2.3
38ꢀ±ꢀ1.3
-OCH
3
3
1
2
-OCH
3
3
CH
CH
46ꢀ±ꢀ1.7
29ꢀ±ꢀ1.0
-OCH
3
3
-OCH
3
CH
45ꢀ±ꢀ1.5
1
0 | P a g e

ACCEPTED MANUSCRIPT
%
inhibition of Cdk4-cyclin D1
Code
R
R1
X
a
at 10 µM
3
4
-OCH
3
CH
76ꢀ±ꢀ2.2
3
3
5
6
-H
-H
CH
CH
20ꢀ±ꢀ0.7
45ꢀ±ꢀ1.6
3
3
7
8
-H
-H
-H
CH
CH
-11±0.7
5ꢀ±ꢀ0.3
3
9
0
CH
CH
CH
-13ꢀ±ꢀ0.7
95 ± 3.1
26 ± 0.9
-27 ± 0.8
28 ± 1.1
-8 ± 0.6
0 ± 0.3
4
-H
(
Cink4T)
4
4
4
4
4
1
2
3
4
5
-H
-H
-H
-H
-H
CH
CH
CH
CH
4
4
6
7
-H
-H
CH
CH
5 ± 0.4
28 ± 0.9
4
4
8
9
-H
CH
CH
-37 ± 1.1
48 ± 1.4
-OCH
3
5
0
-OCH
3
CH
89 ± 2.8
1
1 | P a g e

ACCEPTED MANUSCRIPT
%
inhibition of Cdk4-cyclin D1
Code
R
R1
X
a
at 10 µM
5
5
5
5
5
1
2
3
4
5
-OCH
3
CH
16 ± 0.7
41 ± 1.4
0 ± 0.4
7 ± 0.4
-3 ± 0.3
-OCH
-OCH
-OCH
-OCH
3
CH
CH
CH
CH
3
3
3
5
6
-OCH
3
3
CH
0 ± 0.3
5
5
5
7
8
9
-OCH
CH
CH
CH
16 ± 0.5
20 ± 0.5
21 ± 0.5
-OCH
3
3
-OCH
6
6
6
0
1
2
-OCH
3
3
CH
CH
5 ± 0.3
14 ± 0.5
3 ± 0.4
-OCH
-OCH
-OCH
-OCH
3
3
3
CH
CH
CH
6
6
3
4
-5 ± 0.4
0 ± 0.3
6
5
-OCH
3
CH
-4 ± 0.3
6
6
6
7
-OCH
3
3
CH
CH
41 ± 1.1
0 ± 0.3
-OCH
1
2 | P a g e

ACCEPTED MANUSCRIPT
%
inhibition of Cdk4-cyclin D1
Code
R
R1
X
a
at 10 µM
6
6
7
7
8
9
0
1
-OCH
3
CH
-12 ± 0.5
-3 ± 0.4
1 ± 0.3
-OCH
-OCH
-OCH
3
CH
CH
CH
3
3
-10 ± 0.5
7
2
-OCH
3
CH
39 ± 0.8
a
Percentage inhibition represent mean and standard deviations from three independent
experiments.
2
.3. Screening of the quinazolinone chemical library in the Cdk assays. The library of
fifty-nine quinazolinones, at 10 µM concentration, was at first screened in the Cdk4-cyclin D1
kinase assay using a C-terminal fragment of pRb (i.e. RBCTF) [56] as substrate, as suggested
by the supplier of the assay (ProQinase, Germany). Out of the fifty-nine quinazolinones
synthesized, fifty-six compounds were derived from hydroxy benzaldehydes (3-
hydroxybenzaldehyde, 4-hydroxybenzaldehyde, vanillin or isovanillin) and their derivatives.
nd
The aryl group of these benzaldehydes constitutes the pendent phenyl ring at 2 position of
quinazolinones (compounds 17 to 72; Table 1). Hydroxyl groups of these four benzaldehydes
provided an opportunity to extend the sidechain of pendent phenyl ring at its meta- and para-
positions. 4-hydroxy derivative (compound 17) displayed 30% inhibition at 10 µM
concentration. Substitution of 2,4-diflurobenzyl-oxy group (compound 20) marginally
improved its activity to 42%. An additional methoxy functional group ortho- to 4th-position
did not show any significant improvement in activity with reference to compounds bearing
2
,4-dichlorophenyl-oxy (compounds 19, 27) and methyl sulphonyl (compounds 21, 30)
substitutions. Whereas, compounds bearing 2,4-diflurobenzyl-oxy substitution (compounds
0, 29), have significantly improved activity, that is,78% inhibition at 10 µM concentration.
2
1
3 | P a g e

ACCEPTED MANUSCRIPT
Swapping the position of these two functional groups provided compound 34 with similar
activity (76% inhibition at 10 µM). Introduction of -CO-NH- between -OCH2- and -R
provided an oxy-acetanilide side chain (compounds 35 to 72; Table 1). Compound 36
carrying oxy-acetanilide with unsubstituted phenyl ring exhibited 45% inhibition at 10 µM
concentration. Substitution at ortho- (compounds 37, 39) and para- (compounds 38, 41, 42,
4
4) positions of phenyl ring of oxy-acetanilide group (compounds 37, 39, 40), ring fusion
(compounds 43, 45, 46), replacing phenyl with Ar-alkyl (compounds 47, 48) did not provide
better activity in comparison with compound 36. While, chloro- substitution at meta- position
of phenyl ring of oxy-acetanilide group, provided the best compound 40, in this series, with
9
5% inhibition at 10 µM concentration. In order to comment on the effects of additional
methoxy group ortho- to 4-oxyacetanilide substitution would require further library
enumeration as current set of compounds 49-57 do not share similarity at 4th position when
comparing with compounds 35-48. But a compound having ortho-methoxy substitution on
phenyl ring of oxy-acetanilide (50) has showed 89% inhibition at 10 µM concentration.
Swapping the position of methoxy and oxy-acetanilide substitutions (58-72) decreased the
activity. In summary, (i) diflurobenzyl-oxy substitution at 3rd or 4th position of pendent
phenyl ring with an ortho-methoxy functional group, (ii) oxy-acetanilide with 3-chloro
substitution in its phenyl ring at 4th position of pendent phenyl ring and (iii) oxy-acetanilide
with 2-methoxy substitution on its phenyl ring at 4th position of pendent phenyl ring along
with an ortho-methoxy functional group are favorable for Cdk4 inhibitory activity.
Four compounds 29, 34, Cink4T (40) and 50 were identified as having >75% inhibition at
1
0 µM concentration (Table 1). These compounds were further screened for percentage
inhibition, at 10 µM concentration, in the Cdk2-cyclin A, Cdk1-cyclin B1, Cdk9-cyclin T1,
and Cdk6-cyclin D1 kinase assays (Table 2), using a C-terminal fragment of pRb (RBCTF) or
RBER CHKtide as substrates, depending on the Cdk, as suggested by the supplier (ProQinase,
1
4 | P a g e

ACCEPTED MANUSCRIPT
Germany). RBER CHKtide consists of an N-terminal GST tag linked to amino acids S773-
K928 of pRb (National Center for Biotechnology Information, NCBI Accession #
NP_000312) which is followed by 11 Arg residues and the 23-amino acid peptide sequence,
KKKVSRSGLY RSPSMPENLN RPR (CHKtide).
After determination of percentage inhibition of the four different Cdk enzymes (Table 2),
IC values, with respect to inhibition of the four kinases, were ascertained for the four
5
0
compounds 29, 34, Cink4T and 50 (Table 2). All four compounds were found to be selective
inhibitors of Cdk4-cyclin D1 enzyme with low micromolar Cdk4 IC values, ranging from
5
0
0
.47 µM to 3.8 µM (Table 2). Amongst the four compounds, Cink4T was the most potent
with an IC value of 0.47 µM which was similar to that observed earlier with fascaplysin
5
0
(
0.41 ± 0.04 µM)[20] and much lower than the IC values of fascaplysin’s two published
50
non-planar analogues BPT (10 ± 1.2 µM) [26] and CA224 (6.2 ± 0.9 µM) [25], which had
also been found to be efficacious in mouse tumor models. Cink4T (i.e. compound 40) and
compound 50 were also tested for inhibition of CYP1A1, CYP1A2 and CYP1B1 enzymes.
They had IC values of >5 µM for each of the three enzymes in microsomal enzyme assays
5
0
[
52], microsomal enzymes being endoplasmic reticular membrane-bound CYP enzymes
which are isolated from recombinant insect or yeast cells. Compounds 29 and 34 have never
been reported before. They also have IC values of >5 µM in the three CYP1 microsomal
5
0
assays (unpublished observations).
Table 2. IC values of the four most potent Cdk4 inhibitors, as identified in Table 1, in
5
0
Cdk4-cyclin D1, Cdk2-cyclin A2, Cdk1-cyclin B1, Cdk9-cyclin T1 and Cdk6-cyclin D1
kinase assays. Results demonstrate Cdk4-specificity of the four compounds vis-à-vis the three
kinases Cdk2-cyclin A2, Cdk1-cyclin B1 and Cdk9-cyclin T1. Pablociclib, the US Food and
Drug Administration (FDA)-approved Cdk4 inhibitor [57], was used as a control.
1
5 | P a g e

ACCEPTED MANUSCRIPT
%
Inhibition
% Inhibition
Cdk2/cycA2,
% Inhibition
Cdk1/cycB1,
% Inhibition
Cdk9/cycT1,
%Inhibition
Cdk6/cycD1,
Cdk4/cycD1,
Code
at 10 µM
at 10 µM
at 10 µM
at 10 µM
at 10 µM
a
a
a
a
a
(
IC , µM)
(IC , µM)
(IC , µM)
(IC , µM)
(IC , µM)
50
50
50
50
50
9
5 ± 3.1
0.47 ± 0.06)
8 ± 2.3
3.5 ± 0.2)
6 ± 2.2
3.8 ± 0.3)
9 ± 2.8
-18 ± 0.8
(>100)
-8 ± 0.4
(>100)
-5 ± 0.2
(>100)
14 ± 0.7
(>50)
8 ± 0.3
(>50)
13 ± 0.6
(>50)
10 ± 0.4
(>50)
14 ± 0.6
(>50)
-15 ± 0.7
(>100)
-5 ± 0.2
(>100)
8 ± 0.3
(>50)
-22 ± 1.1
(>100)
-15 ± 08
(>100)
-18 ± 0.9
(>100)
-14 ± 0.7
(>100)
Cink4T (40)
(
7
29
34
50
(
(
(
7
8
-22 ± 0.8
(>100)
1.5 ± 0.1)
95 ± 2.5
(0.013 ±
0.002)
Pablociclib
-97 ± 0.7
(0.018 ±
0.002)
42 ± 1.7
5 ± 0.4
(>100 µM)
-4 ± 0.5
(>100 µM)
(
obtained
(
>10 µM)
values)
Pablociclib
published
values)[57]
(
(IC : 0.011)
(IC : >10 µM) (IC : >10 µM)
(IC : ND)
(IC : 0.016)
5
0
50
50
50
50
a
Percentage inhibition of the Cdk4, Cdk2, Cdk1, and Cdk9 enzymes and the IC values for
5
0
inhibition of enzyme activities represent mean and standard deviations (where applicable)
from three independent experiments. ‘ND’ denotes not determined.
2
.4. Comparison of the ability of the four Cdk4 inhibitors, compounds 29, 34, Cink4T
and 50, to inhibit tubulin polymerization in vitro, at 5 µM concentration. Cink4T had
been identified in the chemical library that was designed on the basis of a virtual screen which
aimed to find dual inhibitors of Cdk4 and tubulin. The templates (‘queries’) that were used for
this virtual screen were two known dual inhibitors of Cdk4 and tubulin polymerization [25,
4
9]. Hence, we investigated next the action of Cink4T on tubulin polymerization in vitro.
The original procedures [58, 59] for determination of tubulin polymerization depend on
light being scattered by microtubules in a way that light scattering is proportional to increase
in microtubule polymer concentration. The curve, that results from observations of light
scattering of tubulin polymerization over a period of time, is representative of the three phases
of microtubule polymerization: (i) nucleation, (ii) growth and (iii) steady state equilibrium.
The fluorescence-based tubulin polymerization assay, used in our studies, has been adapted
from a protocol described earlier [60]. It is a quick one-step procedure for determining the
effects of drugs on tubulin polymerization and it compares well with the absorbance-based
1
6 | P a g e

ACCEPTED MANUSCRIPT
assay (Tubulin polymerization assay kit; Cytoskeleton Inc; Cat # BK011P;
https://www.cytoskeleton.com/bk011p). Tubulin polymerization is tracked by fluorescence
enhancement due to the incorporation of a fluorescent reporter into microtubules that occurs
with the progression of polymerization. Porcine neuronal tubulin, used for the experiments,
was of the highest available purity (>99% pure, Cytoskeleton, Cat. # T240).
Representative polymerization curves of paclitaxel (an enhancer of tubulin
polymerization), compounds 29, 34, 50 and Cink4T are shown in Figure 2. Compounds that
interact with tubulin alter one or more of the characteristic phases of polymerization. For
example, at 3 µM concentration paclitaxel eliminates the nucleation phase and enhances the
Vmax of the growth phase. In contrast, vinblastine, a microtubule destabilizing drug (i.e. a
tubulin polymerization inhibitor belonging to the family of vinca alkaloids), causes a drastic
decrease in Vmax and reduction in final polymer mass (Tubulin polymerization assay kit;
Cytoskeleton Inc; Cat # BK011P; https://www.cytoskeleton.com/bk011p) [61]. Our results
indicate that, at 5 µM concentration, Cink4T interacts with tubulin more strongly than
compounds 29, 34 and 50, suggesting that it is the most potent tubulin polymerization
inhibitor identified in our studies. The Cink4T curve resembles the one obtained with 3 µM
vinblastine same fluorescence-based assay. The ‘Control’ curve, with no drug, relates to the
reaction when tubulin is incubated alone (i.e. in the absence of any drug).
1
7 | P a g e

ACCEPTED MANUSCRIPT
Figure 2. Tubulin polymerization using the fluorescence-based tubulin polymerization assay
(BK011P; Cytoskeleton Inc). Tubulin was incubated (a) alone (Control), (b) with paclitaxel,
and (c) with compounds 29, 34, 50 and Cink4T. Each curve represents the average of results
performed four times. Polymerization was measured at excitation wavelength of 360 nm and
emission at 420 nm. The three phases of tubulin polymerization are marked for the Control
(no drug) polymerization curve; I: nucleation, II: growth, III: steady state equilibrium. ‘Cpd’
signifies ‘compound’.
2
.5. The ability of the Cdk4 inhibitor, Cink4T, to inhibit tubulin polymerization at
different concentrations. Cink4T was then tested over a range of concentrations for its
ability to inhibit tubulin polymerization in the same fluorescence-based assay. Again,
paclitaxel was used in the assay as a known enhancer of tubulin polymerization. Figure 3
shows the dose response of Cink4T in the process of polymerization of tubulin.
1
8 | P a g e

ACCEPTED MANUSCRIPT
Figure 3. Dose response of Cink4T to inhibition of tubulin polymerization. Tubulin
polymerization was determined using the fluorescence-based tubulin polymerization assay
(BK011P; Cytoskeleton Inc). Tubulin was incubated alone (Control, i.e. ‘no drug’), and with
(
a) the polymerization enhancer paclitaxel, (b) the depolymerization agent CaCl , and (c)
2
Cink4T at four different concentrations (0.1, 0.5, 0.75, 1 µM). Each curve represents the
average of results performed four times. Polymerization was measured at excitation
wavelength of 360 nm and emission at 420 nm. The three Phases of tubulin polymerization
are marked for the Control (no drug) polymerization curve; I: nucleation, II: growth, III:
steady state equilibrium. The IC for inhibition of tubulin polymerization by Cink4T was
5
0
calculated to be 0.6 µM, vinblastine’s IC being also 0.6 µM (Tubulin polymerization assay
5
0
kit; Cytoskeleton Inc; Cat # BK011P).
2
.6. Molecular modeling studies with Cink4T and Cdk4 or tubulin. Cink4T exhibited
~212-fold selectivity towards ATP binding pocket of Cdk4 in comparison with Cdk2.
Molecular docking simulation was carried out with AutoDock-4.2. High-scoring
conformation from the largest cluster was considered for the analysis of the molecular
interaction. Cink4T was found to interact with ATP-binding pocket of Cdk4 in an inverted
bend conformation in comparison with a linear conformation in the ATP-binding pocket of
Cdk2 (Figure 4). Quinazolinone N3 amino hydrogen and C4 carbonyl oxygen establishes two
H-bonds with backbone carbonyl oxygen and amino hydrogen of Val96. These interactions
1
9 | P a g e

ACCEPTED MANUSCRIPT
keep the quinazolinone nucleus in the pocket with the sidechain oxy-acetamido group
extending to the entrance of the pocket. Sidechain amino hydrogen of the Arg101’s planar
guanidine group establishes an additional H-bonding interaction with sidechain carbonyl
oxygen of Cink4T, thus forcing the molecule to adopt a cis-conformation for maximizing the
hydrophobic interaction of Cink4T with Ile12, Val20, Ala33, Lys35, His95, Asp97, Gln98,
Asp99, Glu144, Leu147 and Asp158. Interaction of Cink4T with Cdk2 in comparison with
Cdk4 revealed that it misses two crucial interactions with Lys88 (Arg101, Cdk4) and
Gln131(Glu141, Cdk4) and that may be the reason for its selectivity towards Cdk4. Figure 5
displays 2D-interaction plot for Cink4T with Cdk4 and Cdk2.
Figure 4. Interactions of Cink4T with Cdk2 and Cdk4. Cdk2 active site residues in orange
ribbon with sidechains represented by orange tubes and heteroatoms colored by atom type;
Cink4T in Cdk2’s pocket is represented by magenta tubes and heteroatoms are colored by
atom type. Cdk4 active site residues are in cyan ribbon with sidechains represented by cyan
tubes and heteroatoms colored by atom type; Cink4T in the Cdk4 pocket is represented by
forest green tubes and heteroatoms are colored by atom type.
2
0 | P a g e

ACCEPTED MANUSCRIPT
Figure 5. 2D-plot of Cink4T with Cdk4 (top) and Cdk2 (bottom). Hydrogen bonds are
represented as green dotted lines; hydrophobic interactions are shown by red lines and
equivalent residues are shown as red circles.
2
1 | P a g e

ACCEPTED MANUSCRIPT
Cink4T’s interaction with tubulin has been studied at the binding pocket in the interface of
tubulin-ꢁ/ꢂ subunit. No H-bonding interaction was found. Instead, Cink4T exhibited
hydrophobic interactions with the following residues in theꢀꢂ-subunit: Val238, Cys241,
Leu242, Leu248, Ala250, Lys254, Leu255, Asn258, Ala316, Lys352, Thr353 and Ala354.
The following four residues from the ꢁ-subunit, Asn101, Ser178, Thr179 and Ala180,
displayed hydrophobic interactions with the sidechain oxy-acetanilide portion of Cink4T.
Figures 6 shows the interactions of Cink4T with active site residues of tubulin.
Figure 6. Interaction of Cink4T with tubulin. Sidechains of residues from ꢂ-subunit lining the
pocket are in orange wires with heteroatoms colored by atom type; sidechains of residues
from ꢁ-subunit lining the pocket are in cyan wires with heteroatoms colored by atom type.
Cink4T is represented by forest green tubes with heteroatoms colored by atom type.
2
.7. Cancer cell growth inhibition. EC values for inhibition of cell growth by the four
50
compounds 29, 34, Cink4T (40) and 50 were then determined. The four cancer cell lines:
A549 (mitotic spindle checkpoint-(MSC)-proficient, pRb-positive, p53-positive), Calu-1
(MSC-deficient, pRb-positive, p53-null), MDA-MB-468 (MSC-deficient, pRb-null, p53
mutant), and T47D (MSC-deficient, pRb-positive, p53 mutant) were chosen for these studies.
The EC values (an EC value being the concentration of a drug that gives half-maximal
5
0
50
response, in this case specifically inhibition of growth of cancer cells) obtained are presented
2
2 | P a g e

ACCEPTED MANUSCRIPT
in Table 3. Compounds 29, 34 and 50 displayed cell proliferation inhibitory activity ≤ 10 µM
concentration in the three pRb-positive cell lines studied (i.e. T47D, Calu-1 and A549), pRb
being the sole mediator of active Cdk4-triggerred cell proliferation [35, 36], while Cink4T
(40) inhibited cell proliferation at sub-micromolar concentrations in the pRb-positive cell
lines.
Table 3. Growth inhibition of four cancer cell lines treated with the four Cdk4 inhibitors
which were identified in the quinazolinone chemical library, designed on the basis of ligand-
based virtual screening.
A549,
Calu-1,
MDA-MB-468,
T47D,
Compound
Cink4T (40)
a
a
a
a
EC (µM)
EC (µM)
EC (µM)
EC (µM)
5
0
50
50
50
0.8 ± 0.05
7 ± 0.6
0.9 ± 0.05
9 ± 0.7
10.85 ± 0.6
>50
0.75 ± 0.06
8 ± 0.7
29
34
50
7.5 ± 0.7
3 ± 0.2
9.8 ± 0.6
6 ± 0.4
>50
8.2 ± 0.65
4 ± 0.3
>50
Palbociclib
0.4 ± 0.02
33 ± 2.2
1.9 ± 0.2
0.35 ± 0.03
a
Percentage inhibition represent mean and standard deviations from three independent
experiments.
2
.8. Cink4T blocks growth of asynchronous MSC-deficient, pRb-positive T47D breast
cancer cells at G /G . Upon treatment with 1x EC (0.75 µM) concentration of Cink4T for
0
1
50
1
8 h, T47D cells were found to be blocked at G /G (compare Figures 7A and 7B). T47D
0 1
cells are not arrested at G /M by Cink4T because they are mitotic spindle checkpoint (MSC)-
2
deficient[32]; MSC-proficiency is required for a G /M block mediated by a compound that
2
has a role in tubulin polymerization [25, 26, 33, 62]. The block at G /G phase observed in
0
1
cells treated with Cink4T would suggest inhibition of cellular Cdk4 which acts at G /G .
0
1
Further cell cycle analysis showed that treatment of cells with 2x EC (1.5 µM) concentration
50
of Cink4T leads to induction of apoptosis, which is displayed by the sub-G1 peak in Figure
2
3 | P a g e

ACCEPTED MANUSCRIPT
7
C. It has been observed that a compound which induces block of cancer cells at G /G often
0
1
induces apoptosis [31, 63, 64].
The phosphorylation status of Cdk4-specific serine residues, Ser780, Ser795 and
Ser807/811 in pRb (i.e. serine residues in pRb that are specifically phosphorylated by Cdk4-
cyclin D1) [38, 39, 65] was assessed in untreated T47D cells together with cells treated with
Cink4T, at 1x EC (0.75 µM) and 2x EC (1.5 µM) concentrations, via Western blotting
50
50
(Figures 7D to 7G). Phosphorylation of these residues modify the conformation of pRb to
allow progression of cells in the cell cycle [39]. The Western blots (Figures 7E and 7F)
demonstrate that Cink4T, at both 1x EC and 2x EC concentrations, inhibits the
50
50
phosphorylation of pRb at the Ser780 and Ser795 residues, thus suggesting Cdk4-specific
activity of Cink4T. Although Cink4T fails to prevent phosphorylation of pRb at the Ser807/11
residues at 1x EC concentration, it does so at the higher 2x EC concentration (Figure 7G).
50
50
Lower levels of pRb at 2x EC concentration of Cink4T (Figure 7D) may suggest that cells
50
were apoptotic which was actually seen from the sub-G peak in the cell cycle analysis
1
(Figure 7C).
2
4 | P a g e

ACCEPTED MANUSCRIPT
+
Figure 7. (A) to (C): Cell cycle analyses of T47D (MSC-deficient, pRb ) cells which are
treated with 1x EC (0.75 µM) and 2x EC (1.5 µM) concentrations of Cink4T for 18 h. The
50
50
ratio between G /G and S phases of the cell cycle (i.e. G :S ratio), increases dramatically
0
1
1
upon treatment with Cink4T which is indicative of a block at G /G . The sub-G peak
0
1
1
indicates apoptosis. (D) to (G): Western blot analysis of proteins from untreated cells (i.e.
Control) and T47D cells treated for 18 h with 1x EC (0.75 µM)) and 2x EC (1.5 µM)
50
50
concentrations of Cink4T, using phospho-specific antibodies (Ser780, New England Biolabs,
Cat # 9307; Ser795, New England Biolabs, Cat # 9301; Ser807/811, New England Biolabs,
Cat # 9308) and an antibody to pRb (New England Biolabs, Cat # 9309). An appropriate
HRP-conjugated secondary antibody was used to illuminate the different forms of pRb
(phosphorylated and unphosphorylated).
2
.9. Cink4T maintains block of cell growth at G /G , induced by serum starvation, of
0 1
MSC-deficient, pRb-positive T47D breast cancer and Calu-1 non-small cell lung
carcinoma (NSCLC) cells. Cdk4 is the first cyclin dependent kinase that is activated upon
entry of cells in the cell division cycle, that is, at the G /G phase transition. The activation of
0
1
Cdk4 by cyclin D1 initiates the phosphorylation of pRb that ultimately causes its dissociation
2
5 | P a g e

ACCEPTED MANUSCRIPT
from the E2F family of transcription factors leading to a cascade of events that result in the
entry of cells into the S phase [66]. The transition from G to G phase of the cell cycle is
0
1
dependent on serum growth factors present in the cell culture medium. When cells are starved
of serum, they undergo arrest at the G /G boundary but when fresh serum is added to these
0
1
cells, the arrested cells re-enter the cell cycle. We hypothesized that if Cink4T truly inhibits
Cdk4-cyclin D1 enzyme in cells, serum-starved cells would not re-enter the cell cycle if they
were released in the presence of Cink4T. In order to block the cells at early G phase, T47D
1
and Calu-1 cells were serum starved by incubating in medium containing 0.1% FBS (fetal
bovine serum) for 24 h. They were then released in complete growth medium supplemented
with 10% FBS in the presence or absence of test compound for a further 18 h.
Flow cytometric analysis of T47D and Calu-1 cells show that, upon serum starvation,
~
85% and ~88% percentage of cells reside at the G /G phase of the cell cycle (Figures 8A
0 1
and 7E). Serum-starved cells, when released just in the presence of fresh medium, entered the
cell cycle normally with 60% and 69% cells at G /G and percentage increase in S phase of
0
1
around 11% (Figures 8B and 8F). Serum-starved cells, synchronized at G /G , released in the
0
1
presence of 1x EC concentration of Cink4T is prevented from fully re-entering the cell cycle
5
0
(
Figures 8C and 8G). The G /G block is somewhat maintained with 77% and 75% cells being
0 1
found in the G /G phase. When the serum starved cells were released in the presence of 2x
0
1
EC concentration of Cink4T (Figures 8D and 8H), G /G arrest was fully maintained in
50
0
1
Calu-1 cells but apoptosis was induced in T47D cells. 18.6% of T47D cells were detected in
the sub-G peak which indicates apoptotic cells (Figure 8D) suggesting that at the higher
1
concentration of Cink4T induction of apoptosis is seen only in T47D cells, which carry a p53
mutant allele, at the G /G phase of the cell cycle. Calu-1 cells, which completely lack the p53
0
1
gene (i.e. are p53-null), are resistant to apoptosis [67]. Overall, cell cycle analyses show that
2
6 | P a g e

ACCEPTED MANUSCRIPT
Cink4T displays the typical characteristic of a true Cdk4 inhibitor by maintaining G /G arrest
0
1
in the cell cycle when cells are released in fresh medium after serum starvation [25, 26].
Figure 8. Cell cycle analyses of the response of T47D and Calu-1 cells, synchronized through
serum starvation, to treatment with Cink4T. (A), (E): synchronized, serum-starved T47D and
Calu-1 cells; (B), (F): serum-starved cells released in fresh medium for 18 h; (C), (G): serum-
starved cells released into fresh medium in the presence of 1x EC concentration of Cink4T
50
for 18 h; and (D), (H): serum-starved cells released into fresh medium in the presence of 2x
EC concentration of Cink4T for 18 h.
50
2
.10. Cink4T is unable to block growth of asynchronous MSC-deficient but pRb-null
MDA-MB-468 breast cancer cells at G /G . Inhibition of Cdk4-mediated block of cell
0
1
growth at G /G depends on the presence of pRb within the cells [35, 36], while cells which
0
1
are mitotic spindle checkpoint (MSC)-deficient are unaffected by tubulin polymerization
inhibitors[33, 62]. Hence, it would be expected that MDA-MB468 cells which lack the
retinoblastoma gene, pRb, and are also MSC-deficient [41] are mainly unaffected by
increasing concentrations of Cink4T (compare Figure 9A with 9B and 9C). Indeed, MDA-
MB468 cells do not undergo cell cycle block at G /G or undergo apoptosis as is indicated by
0
1
the ratio of the peaks observed at the G /G and S phases (G :S ratio), and the sub-G peak,
0
1
1
1
2
7 | P a g e

ACCEPTED MANUSCRIPT
upon treatment of cells with 1x EC (10.85 µM) and 2x EC (22 µM) concentrations of
50
50
Cink4T for 18 h (Figures 9B and 9C).
Figure 9. (A) to (C): Cell cycle analyses of MDA-MB-468 (MSC-deficient, pRb-null) cells
which are treated with 1x EC (10.85 µM; B) and 2x EC (22 µM; C) concentrations of
50
50
Cink4T for 18 h. The ratio between G /G and S phases of the cell cycle (G :S ratio), with
0
1
1
increasing concentrations of Cink4T, are shown within each cell cycle analysis, (A) to (C).
The sub-G peak, which is ‘zero’, would indicate that MDA-MB-468 cells do not undergo any
1
apoptosis upon treatment with Cink4T.
2
.11. Cink4T blocks growth of asynchronous MSC-proficient, pRb-positive A549
non-small cell lung carcinoma (NSCLC) cells at G /M. In MSC-proficient A549 cells
2
which also carry a functional pRb gene [40], cells undergo an increase in the G /S ratio (from
1
2
.7 to 6.5) together with an increase in the percentage of cells at G /M (from 18% to 28%)
2
when treated, for 18 h, with 1x EC (0.8 µM) concentration of Cink4T (compare Figures 10A
50
and 10B). However, a G /M block becomes obvious, with percentage of cells at G /M being
2
2
7
0%, when A549 cells are treated with 2x EC (1.6 µM) concentration of Cink4T (Figure
50
1
0C). The results show the accumulation of A549 cells, in which the mitotic spindle
checkpoint is intact, predominantly at mitosis (i.e. G /M) only when cells are treated with the
2
higher concentration of Cink4T. They also indicate that Cink4T, designed on the basis of a
virtual screen, not only arrests cells at the early G /G phase of the cell cycle but also at
0
1
mitosis (i.e. G /M phase of the cell cycle).
2
2
8 | P a g e