New insights into the structural requirements for pro-apoptotic agents based on 2,4-diaminoquinazoline, 2,4-diaminopyrido[2,3-d]pyrimidine and 2,4-diaminopyrimidine derivatives

Article abstract of DOI:10.1016/j.ejmech.2011.05.060

As a continuation of our work on new anti-tumoral derivatives with selective pro-apoptotic activity in cancer cells, we describe the synthesis and the preliminary evaluation of the cytotoxic and pro-apoptotic activities of a series of pyrimidin-2,4-diamine derivatives that are structurally related to quinazolin-2,4-diamine and pyrido[2,3-d]pyrimidin-2,4-diamine derivatives. We also describe the structure-activity relationship studies carried out on four series' of quinazolin-2,4-diamine, 2-(alkylsulfanyl)-N-alkyl- and 2-(alkylsulfanyl)-N-alkylarylpyrido[2,3-d]pyrimidine and pyrimidin-2,4-diamine derivatives. The proposed preliminary pharmacophore consists of a flat heterocyclic ring, preferably a pyrido[2,3-d]pyrimidine, with two equivalent alkylarylamine chains, preferably N-benzyl- or N-ethylphenylamine, located in positions 2 and 4 of the ring, and with a preferred ALogP in the range 4.5-5.5. The nitrogen present in the central ring can act as hydrogen bond acceptors (HBA) whereas the amino group in the 4-position can act as a donor (HBD) or an HBA and the amino group in the 2-position can act as an HBD. On the basis of the analyzed structural profiles, different mechanisms of action can be suggested for the quinazolin-2,4-diamine, the 2-(alkylsulfanyl)-N-alkylpyrido[2,3-d] pyrimidin-4-amine and the pyrido[2,3-d]pyrimidin-2,4-diamine derivatives.

Full text of DOI:10.1016/j.ejmech.2011.05.060

European Journal of Medicinal Chemistry 46 (2011) 3887e3899
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
New insights into the structural requirements for pro-apoptotic agents based
on 2,4-diaminoquinazoline, 2,4-diaminopyrido[2,3-d]pyrimidine
and 2,4-diaminopyrimidine derivatives
,
a
b
b
María Font a , Álvaro González , Juan Antonio Palop , Carmen Sanmartín
*
a Sección de Modelización Molecular, Dpto de Química Orgánica y Farmacéutica, Facultad de Farmacia, Universidad de Navarra, Irunlarrea 1, E-31008 Pamplona, Spain
b Sección de Síntesis, Dpto de Química Orgánica y Farmacéutica, Facultad de Farmacia, Universidad de Navarra, Irunlarrea 1, E-31008, Pamplona, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
As a continuation of our work on new anti-tumoral derivatives with selective pro-apoptotic activity in
cancer cells, we describe the synthesis and the preliminary evaluation of the cytotoxic and pro-apoptotic
activities of a series of pyrimidin-2,4-diamine derivatives that are structurally related to quinazolin-2,4-
diamine and pyrido[2,3-d]pyrimidin-2,4-diamine derivatives. We also describe the structureeactivity
relationship studies carried out on four series’ of quinazolin-2,4-diamine, 2-(alkylsulfanyl)-N-alkyl- and
2-(alkylsulfanyl)-N-alkylarylpyrido[2,3-d]pyrimidine and pyrimidin-2,4-diamine derivatives.
The proposed preliminary pharmacophore consists of a flat heterocyclic ring, preferably a pyrido[2,3-
d]pyrimidine, with two equivalent alkylarylamine chains, preferably N-benzyl- or N-ethylphenylamine,
located in positions 2 and 4 of the ring, and with a preferred ALogP in the range 4.5e5.5. The nitrogen
present in the central ring can act as hydrogen bond acceptors (HBA) whereas the amino group in the
4-position can act as a donor (HBD) or an HBA and the amino group in the 2-position can act as an HBD.
On the basis of the analyzed structural profiles, different mechanisms of action can be suggested for the
quinazolin-2,4-diamine, the 2-(alkylsulfanyl)-N-alkylpyrido[2,3-d]pyrimidin-4-amine and the pyrido
[2,3-d]pyrimidin-2,4-diamine derivatives.
Received 21 January 2011
Received in revised form
10 May 2011
Accepted 24 May 2011
Available online 31 May 2011
Keywords:
Cytotoxicity
Apoptosis
Structureeactivity relationships
Quinazolin-2,4-diamine
Pyrido[2,3-d]pyrimidin-2,4-diamine
Pyrimidin-2,4-diamine
Ó 2011 Elsevier Masson SAS. All rights reserved.
1. Introduction
tumor cell lines, thus affecting to a lesser extent the non-malignant
cells [5e11].
The molecular players involved in the development of apoptosis,
the cellular suicide mechanism that enables living organisms to
control abnormal cell division, are becoming increasingly well-
characterized and represent a set of new targets for the develop-
ment of more effective and selective anti-tumor agents [1].
The development of new anti-tumor therapeutic tools has
advanced greatly in the past decade; thus, approaches for the
treatment of cancer have moved towards targeting the specific
molecular alterations that occur in tumor cells, concentrating
essentially on the development of both small molecules and bio-
logical agents that have shown remarkable clinical activity without
the toxicity associated with conventional chemotherapy [2e4].
With this objective in mind, we have in recent years developed
a broad research program aimed at obtaining new pro-apoptotic
anti-tumor compounds that also show high selectivity over
As previously reported [9,10], the design of the target
compounds was carried out according to a molecular fragmentary
approach [12e14] based on the available bibliographic informa-
tion. This process involved the construction of a starting general
structure (Fig. 1a), with a planar heterocyclic ring (quinazoline or
pyrido[2,3-d]pyrimidine ring), selected as the central fragment
that can act as a scaffold to carry two functionalized branches in
positions 2 and 4, which are equivalent or different with the aim of
evaluating the possible influence of the symmetry/asymmetry on
the target activity [15]. The selected starting ring systems are
involved in numerous biological activities e mainly concerning
cancer e with action mechanisms related to folate metabolism
inhibition, topoisomerase inhibition, the inhibition of diverse
tyrosine kinases and apoptosis induction, amongst others [16e19].
As an example, of the eight small molecules currently marketed
against oncology indications, three (Gefitinib, Fig. 1b; Erlotinib,
Fig. 1c and Lapatinib, Fig. 1d) are from the 4-anilinoquinazoline
class [20,21].
* Corresponding author. Tel.: þ34 948 425 600X6509; fax: þ34 948 425 648.
E-mail address: mfont@unav.es (M. Font).
0223-5234/$ e see front matter Ó 2011 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2011.05.060

3888
M. Font et al. / European Journal of Medicinal Chemistry 46 (2011) 3887e3899
a
b
c
CN
X
Cl
O
F
N
N
HN
N
d2
X
N
X
HN
N
d3
O
O
CH3
CH3
N
O
O
d1
O
O
N
CH3
d4
d
e
f
CF3
Cl
CH3
O
O
CH3 NH2
O
F
N
HN
HN
N
N
N
O
NH2
N
O
H3C
O
N
N
N
H
N
H
O
S
H
N
CH3
.
Fig. 1. (a) Preliminary pharmacophore (b) Gefitinib (c) Erlonitib. (d) Lapatinib. (e) Piritrexim (f) Aurora A kinase inhibitor. Structural modifications carried out on the central
heteroaromatic nucleus taken as a scaffold (dotted lines): scaffold electronic modification, from quinazoline to pyrido[2,3-d]pyrimidine; scaffold simplification, from bicyclic to
monocyclic system.
The importance of the lateral branches’ length and volume,
along with the presence of polar groups (like hydroxyl) at the end of
the chain on the cytotoxic/pro-apoptotic activity were assessed
[11,22]. We have also explored two different linkages to connect the
side chains to the central ring. The first linkage was the amino
group (eNHe), which was inserted in positions 2 and 4. Secondly,
a sulfur atom was introduced in the 2-position [2-(alkylsulfanyl)-N-
alkylpyrido[2,3-d]pyrimidin-4-amine derivatives] [11] maintaining
an amino NH linkage at 4-position.
As a result of this approach we selected two new compounds,
N,N0-dibenzyl aminoquinazolin-2,4-diamine (3e, Fig. 2) and N,N0-
bis(4-methoxyphenyl)methyl pyrido[2,3-d]pyrimidin-2,4-diamine
(6ac, Fig. 2), as leads for the subsequent design of other new
compounds. Concerning the biological profile for these compounds,
which showed good profiles as apoptosis inducers, we have
demonstrated their activity as caspase-3 activators and inductors of
cell lines. Tests were carried out to determine the possible
mechanism of action for these selected compounds. The initial data
highlighted how the observed dose-dependent cytostatic and pro-
apoptotic effects resulted from the activation of two different
signaling pathways, namelya pathway leading to cell cycle arrest and
a transcription-independent route leading to rapid apoptosis [9,10].
In an effort to gain new insights into the structureeactivity
relationships for these compounds, we designed a series of new
pyrimidine derivatives (compound 8). This new derivatives, struc-
turally related to the previously reported examples, implies a scaf-
fold simplification associated with a significant diminution in the
volume of the central nucleus as well as a change in the lipophilic
profile. Three complementary new pyrido[2,3-d]pyrimidin-2,4-
diamine (compounds 6v, 6x, 6y) were also prepared. The
proposed structural modifications on the scaffold were aimed at
gaining some insight into the influence of parameters like the
electronic distribution, lipophilicity and the total number of
DNA fragmentation in the micromolar (mM) range in several cancer
Fig. 2. H-bond donor (sphere) or acceptor (arrow) distribution pattern for compounds (a) Piritrexim, (b) 3e (previously reported as JRF-12), (c) 6ac (previously reported as HC-6).

M. Font et al. / European Journal of Medicinal Chemistry 46 (2011) 3887e3899
3889
nitrogens present in this central fragment [23], particularly in
terms of the variation in the number of centers potentially involved
in hydrogen bond formation.
We subsequently carried out a global molecular modeling study
on the four series evaluated, i.e., quinazoline-2,4-diamine, pyrido
[2,3-d]pyrimidin-2,4-diamine, 2-(alkylsulfanyl)-N-alkylpyrido[2,3-
d]pyrimidin-4-amine and pyrimidin-2,4-diamine derivatives.
highest IC50 calculated in the three tumoral lines was selected as
the test concentration for assays on non-tumoral cells.
The complete biological screening data, including the previously
published data, are summarized in Table 1 in order to facilitate the
understanding of the structureeactivity analysis carried out on all
four series of derivatives.
The data obtained for the evaluated new compound 6 confirmed
our starting hypothesis, according to which the presence of
a hydroxyl group at the end of the lateral branches (compounds 6x
and 6y) leads to the disappearance of the cytotoxic activity. Simi-
larly, the length of the lateral branches clearly has an influence on
the results. Thus compound 6v (with two hexylamine substituents
in the 2- and 4-positions) is inactive.
With respect to compound 8, all of the evaluated compounds
showed cytotoxic activity apart from compound 8i, which is inac-
tive. Generally, the pyrimidine derivatives proved to be less active
than the previously reported analogs derived from quinazoline and
pyrido[2,3-d]pyrimidine [9,10]. In spite of this fact, five compounds
(8a, 8c, 8d, 8e and 8f) proved to be cytotoxic and showed IC50
2. Results and discussion
2.1. Chemistry
The synthesis of the analyzed derivatives was carried out
according to previously reported methods, which were adapted to
the target structures (see Supplementary data for details). Thus, for
the quinazoline derivatives 3 (synthesis not previously reported),
2,4-(1H,3H)-quinazolinedione was reacted in refluxing phosphorus
oxychloride to afford 2,4-dichloroquinazoline. In a second step the
chloro-substituents were replaced by the corresponding amines in
the presence of ethanol as solvent (Scheme 1a).
The pyrido[2,3-d]pyrimidin-2,4-diamine derivatives 6 (Scheme
1b) were prepared from 2-aminonicotinic acid by condensation
with excess urea to afford the pyrido[2,3-d]pyrimidin-2,4-diol. The
hydroxyl groups were replaced by chloro-substituents by treat-
ment with refluxing phosphorus oxychloride and N,N-dime-
thylformamide as catalyst. In the third step, the chloro-substituents
were replaced by the corresponding amines in the presence of
equimolecular amounts of triethylamine with ethanol as solvent.
Finally, the pyrimidin-2,4-diamine derivatives 8 were synthe-
sized from the commercially available 2,4-dichloropyrimidine by
treatment with an excess of the selected amines under reflux in
ethanol (Scheme 1c).
values below 10 mM in at least one cell line. In addition, 8a showed
values as an apoptosis inducer between 2.3 and 2.8 times higher
than that of Camptothecin, which was used as the reference. In
general, and in a similar way to the data collected for the previously
studied series, it was observed that compounds had a more potent
effect against the breast cancer cell line. Nevertheless, the active
cytotoxic compound 8 did not show any selectivity in the tests
carried out with the non-tumoral cell lines.
2.3. Molecular modeling
Three molecular modeling approaches were applied in the study
described here. Firstly, it was envisaged that some descriptive
parameter at a molecular level could be obtained e such as the log
of the partition coefficient (ALogP 98 [24,25]), the dipole moment
and the molecular volume e that would allow discrimination
between active and inactive compounds. Secondly, we planned to
study a number of quantum parameters, i.e., HOMO and LUMO
orbitals, charge and molecular electrostatic potential (MEP) distri-
bution. Thirdly, we planned to study the conformational behavior
of the compounds in order to identify their preferred conforma-
tions and to perform a distribution analysis on the pharmacophoric
elements. In fact, a given compound can be considered as a collec-
tion of multiple instances, one for each conformation. An active
compound contains at least one active instance, characterized by
a set of chemical functions with a certain spatial arrangement and
molecular shape, while an inactive compound does not [7,26,27].
On the assumption that all of the active compounds bind to the
same site, we would expect their binding conformations to have
similar shapes.
The new synthesized compounds have been characterized
unequivocally by means of IR, 1H NMR and 13C NMR spectroscopy
techniques (see Supplementary data for details).
2.2. Biology
The method of evaluation of the new compounds was similar to
that followed for the previously reported series of quinazoline and
pyridopyrimidine derivatives [8e11]. The study of the biological
profiles of the compounds involved the determination of the
cytotoxic activity in three tumoral cell lines [breast (MD-MBA-231),
bladder (T-24), and colon (HT-29)] using the neutral red assay at the
screening concentrations of 100 and 20
the reference (IC50 ¼ 0.291 M in MD-MB-231 cell line, 0.014
HT-29, and 0.006 M in T-24). The survival percentage was deter-
mM, with camptothecin as
m
mM in
m
mined after a period of 72 h using the survival percentage obtained
from the cells treated only with the solvent (dimethylsulfoxide,
DMSO at 0.5%) as a reference. The results are expressed as the
average of assays carried out in triplicate. IC50 values were calcu-
lated for those compounds that showed survival levels of 45%
In the molecular modeling studies we used the geometric data
obtained for some structurally analogous molecules (Fig. 3) as
a reference; these data were obtained from the Cambridge Struc-
tural Database [28] (CSD) and from the Brookhaven Protein Data
Bank [29,30] (PDB).
(100 mM) in activity assays. The ability of the selected compounds
to induce DNA fragmentation, which is considered a hallmark of
apoptosis, was assessed using the Cell Death Detection ELISA Plus
Kit (Roche) after 24 h of incubation with the compounds. The DNA
degradation level measured in the control culture was taken as 1.
The compounds were also subjected to a caspase-3 activation
assay. The enzyme level was measured at 14, 24 and 48 h using flow
cytometry. The results are expressed by (þ) when an increase of
approximately 50% was detected and (þþ) when an increase of
100% or more was observed.
The compounds FARRUM [31] (Gefitinib, Fig. 3a), GIBHUU [32]
(Piritrexim, Fig. 3b), Trimetrexate (Fig. 3c) isolated from PDB
refcode 1pd9 [33,34], and the ligand isolated from PDB refcode
2no3 [35] (Fig. 3d), among others, were taken as templates to build
the quinazolin-2,4-diamine, pyrido[2,3-d]pyrimidin-2,4-diamine
and pyrido-2,4-diamine derivatives. The bicyclic skeleton in the
quinazoline or pyrido[2,3-d]pyrimidine derivatives and the
pyrimidine ring in the pyrimidine derivatives were first super-
imposed onto the corresponding unit in the selected reference
compounds. The lateral chains were constructed initially in an
extended conformation, with the selected dihedrals (abcd and
With regard to selectivity, cytotoxicity was determined in cell
cultures of two non-tumoral lines (CRL-7899 and CRL-11233). The

3890
M. Font et al. / European Journal of Medicinal Chemistry 46 (2011) 3887e3899
Scheme 1. (a) Synthesis route for quinazoline derivatives 3. (b) Synthesis route for pyrido[2,3-d]pyrimidine derivatives 6. (c) Synthesis route for pyrimidine derivatives 8 (N,N-
DMF ¼ N,N-dimethylformamide; EtOH ¼ Ethanol; TEA ¼ N,N,N-triethylamine).
a‘b’c‘d’, Fig. 3) in the cis configuration. The initial geometries were
fully minimized with the Dreiding Minimize tool implemented in
the Discovery Studio v2.5 suite (DS v2.5) to an energy gradient
below 10e6 kcal mole1 Åe1 and the ALogP 98 value was determined
on these preliminary geometries obtained after the first minimi-
zation (Table 2).
3m. In addition, an ample tolerance for these values was observed
since some products with an ALogP 98 value higher than 7.0 showed
remarkable cytotoxic activity, although generally the products with
values below 3.0 were inactive in the cytotoxicity tests. This is the
case for compounds like 3k (ALogP 98 ¼ 7.372 and IC50 ¼ 3.32, 3.47
and 10.36 mM in T-24, HT-29 and MD-MBA-231 cell lines, respec-
As predicted, the proposed structural variations caused a gradual
modification in the lipophilicity of the compounds; thus, a wide
range of ALogP 98 values was obtained, from 0.418 for 6x, with
hydroxyl groups present at the end of the side chains, to 7.700 for
tively) or compound 6ai (ALogP 98 ¼ 7.462 and IC50 ¼ 1.40, 5.00 and
1.10 M in T-24, HT-29 and MD-MBA-231 cell lines, respectively).
m
The modification from quinazoline to pyrido[2,3-d]pyrimidine
leads to a reduction in the ALogP values and this appears favorable

M. Font et al. / European Journal of Medicinal Chemistry 46 (2011) 3887e3899
3891
Table 1
Biological profile of compounds.
HN
N
n
R
R''
N
R'
X
z
m
Comp.a
X
n
m
Z
R
R0
R00
IC50
(
m
M)b
Apoptosis in cell lineb
Caspase-3c
Percent
Percent
survivalb survivalb
(a)
(b)
(c)
(a)
(b)
(c)
(d) (e)
66
3a
3b
3c
3d
3e
3f
3g
3h
3i
3j
3k
3l
3m
6a
6b
6c
6d
C
C
C
C
C
C
C
C
C
C
C
C
C
N
N
N
N
2
3
4
5
1
1
1
1
2
2
2
3
4
1
3
4
5
2
3
4
5
1
1
1
1
2
2
2
3
4
1
1
1
1
NH CH3
NH CH3
NH CH3
NH CH3
NH C6H5
NH 4-CH3-C6H4
NH 4-Cl-C6H4
NH 4-OCH3-C6H4 4-OCH3-C6H4
NH C6H5
CH3
CH3
CH3
CH3
C6H5
4-CH3-C6H4
4-Cl-C6H4
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
28.95 18.92
4.43 2.0
1.09 n.a.
3.13 1.8
7.56 n.a.
1.79 n.a.
5.47 4.3
4.26 2.0
3.87 5.0
2.18 n.a.
4.65 n.a.
n.a.d
1.5
1.7
n.a.
3.6
5.2
n.a.
8.5
n.a.
n.a.
n.a.
n.a.
2.0
n.a.
1.2
5.0
8.1
n.a.
n.a.
2.5
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
2.4
n.a.
n.a.
n.a.
2.0
2.2
6.6
3.3
n.a.
n.a.
0
19.32
5.25
23.33
5.96
15.40
6.41
3.87
2.86
2.17
3.32
3.45
11.90
6.96
3.57
5.81
4.72
4.42
5.89
2.88
2.30
1.64
4
4
1
5
þ(T-24)
e
2
0
þþ(T-24)
89
13
0
100
33
0
n.a.
n.a.
þ(HT-29)
0
92
0
100
100
0
C6H5
þ(T-24)
NH 4-CH3-C6H4
NH 4-Br-C6H4
NH C6H5
4-CH3-C6H4
4-Br-C6H4
C6H5
C6H5
CH3
CH3
CH3
CH3
þ(HT-29)
3.47 10.36 n.a
2.20
3.78
þ(T-24); þ(HT-29)
11
87
6
1
2
11
e
4.18 n.a.
1.22 n.a.
þþ(T-24); þþ(HT-29)
NH C6H5
þþ(T-24)
S
S
S
S
CH3
CH3
CH3
CH3
20.10 26.20 18.70 1.3
20.90 25.30
4.30 5.10
n.a.
e
9.20 1.5
1.30 3.6
1.40 1.5
n.a.
e
e
þþ(T-24); þ(HT-29)
88
100
n.d.
5.90 22.50
þþ(HT-29);
n.d.e
þþ(MD-MBA-231)
6e
6f
6g
6h
6i
6j
6k
6l
6m
6n
6o
6p
6q
6r
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
e
e
e
e
e
e
e
6
3
4
5
3
4
5
6
3
3
4
5
3
4
5
4
5
3
4
1
1
1
1
2
2
2
2
3
4
1
1
1
1
2
3
4
1
1
1
1
1
1
1
1
1
3
3
3
4
4
4
4
5
3
4
1
1
1
1
2
2
2
2
3
4
1
1
1
1
2
3
4
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
CH3
OH
OH
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
H
H
H
H
5.30 19.90
n.a.
n.a.
n.a.
2.60 n.a.
n.a.
n.a.
n.a.
2.4
e
2.7
e
þþ (T-24)
93
e
100
e
n.a.
n.a.
n.a.
e
e
e
e
e
e
e
e
e
OH
e
e
e
e
e
CH3
CH3
CH3
CH3
OH
CH3
CH3
CH3
CH3
CH3
CH3
CH3 12.30 32.10 15.70 n.a.
n.a.
4.2
n.d.c
n.a.
n.a.
n.a.
1.9
6.4
e
1.3
2.2
2.9
n.d.
n.a.
n.d.
1.7
9.3
e
e
e
e
CH3
CH3 15.80 20.20
CH3 8.00 24.40
CH3 n.a. n.a.
7.90 25.20
5.60 1.7
4.50 1.6
7.00 1.5
þ(T-24); þþ(HT-29)
100
e
66
e
n.a.
n.a.
n.a.
n.a.
e
e
n.a.
e
e
e
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
10.60 13.60
8.90 12.60
8.40 3.7
6.30 n.a.
1.20 4.3
e
e
þ(T-24)
84
95
e
76
99
e
4.50
n.a.
n.a.
18.50 18.40
2.80
n.a.
n.a.
þþ(T-24); þþ(HT-29)
n.a.
n.a.
e
e
e
n.a.
n.a.
0.9
e
e
e
e
e
6s
6u
6v
6x
6y
9.00 n.a.
1.20 4.4
n.d.
2.8
e
n.a.
n.a.
e
e
NH CH3
NH CH3
NH OH
NH OH
NH C6H5
NH 4-CH3-C6H4
NH 4-Cl-C6H4
NH 4-OCH3-C6H4 4-OCH3-C6H4
NH C6H5
NH 4-Br-C6H4
NH 2-pyridyl
NH 3-indolyl
NH C6H5
NH C6H5
NH C6H5
NH 4-CH3-C6H4
NH 4-Cl-C6H4
NH 4-OCH3-C6H4 4-OCH3-C6H4
NH C6H5
NH C6H5
NH C6H5
0.70
4.60
0.58
4.23
100
e
54
e
CH3
OH
OH
C6H5
4-CH3-C6H4
4-Cl-C6H4
2.10
n.a.
n.a.
e
e
e
e
n.a.
n.a.
n.a.
n.a.
e
e
e
e
e
e
e
e
e
e
6z
13.40 11.00 10.30 4.6
2.5
n.a.
n.a.
2.7
9.2
7.2
e
n.a.
4.0
1.3
1.5
n.a.
n.a.
e
þ(T-24)
85
95
0
85
0
90
100
100
100
0
6aa
6ab
6ac
6ad
6ae
6af
6ag
6ah
6ai
8a
8b
8c
8d
8e
8f
8g
5.00
7.50
3.20 2.7
8.30 1.9
2.90 2.3
1.20 1.2
2.70 1.5
n.d.
11.00 15.00
n.a.
5.00
8.10
11.50
n.a
8.90
2.50
4.10
þ(HT-29)
C6H5
þþ(T-24); þ(HT-29)
4-Br-C6H4
2-pyridyl
3-indolyl
C6H5
C6H5
C6H5
n.a.
e
90
e
100
e
n.a.
n.a
e
7.70
13.20
1.40
3.00
1.70
5.00
5.10 n.a
5.80 3.1
1.10 4.6
3.0
6
n.a.
1.6
n.a.
n.a.
1.5
2.2
2.4
1.3
3.2
e
þþ(HT-29)
þ(HT-29)
þ(HT-29)
n.a.
n.a.
n.a.
56
0
86
22
6
0
0
5
10
e
n.d.
0
82
0
1.4
7.4
2.0
2.7
1.9
n.a.
n.a.
e
6.10 17.20
22.17 14.94 19.09 4.8
8.50 9.80 2.0 n.a.
18.70 12.60
9.14 6.97
8.67 12.39
n.a. n.a.
1.90
6
4-CH3-C6H4
4-Cl-C6H4
1
n.d.
n.d.
5
1.50 1.5
3.01 1.2
0.36 1.7
n.a.
n.a.
C6H5
C6H5
C6H5
þ(HT-29)
1
n.a.
e
e
e
a
Previously reported [9,10] as 3a ¼ JRF3; 3b ¼ JRF4; 3c ¼ JRF5; 3d ¼ JRF6; 3e ¼ JRF12; 3f ¼ JRF13; 3g ¼ JRF14; 3h ¼ JRF15; 3i ¼ JRF7; 3j ¼ JRF11; 3k ¼ JRF10; 3l ¼ JRF8;
3m ¼ JRF9; 6u ¼ VD14; 6v ¼ VD22; 6z ¼ HC1; 6aa ¼ HC2; 6ab ¼ HC3; 6ac ¼ HC6; 6ad ¼ XS1; 6ae ¼ XS4; 6af ¼ ME43; 6ag ¼ ME44; 6ah ¼ XS2; and 6ai ¼ XS3.
b
Cell lines (a) T-24; (b) HT-29; (c) MD-MBA-231; (d) CRL-8799; and (e) CRL-11233.
c
(þ) an increase of 50% of caspase-3 level with respect to the control; and (þþ) an increase of 75% of caspase-3 level with respect to the control.
d
n.a. ¼ No activity observed after 48 h incubation.
e
n.d. ¼ No data.

3892
M. Font et al. / European Journal of Medicinal Chemistry 46 (2011) 3887e3899
Fig. 3. Some crystallographic structures taken as geometric references (a) Gefitinib, CSD Refcode ¼ FARRUM. (b) Piritrexim, CSD Refcode ¼ GIBHUU. (c) Trimetrexate, ligand isolated
from Brookhaven Protein Data Bank (PDB) Refcode ¼ 1pd9. (d) Ligand isolated from PDB Refcode ¼ 2no3. (e) and (f) Selected dihedrals (in solid) taken as a pattern value for
conformational family distribution (A, abcd ¼ a‘b’c‘d’ z 0ꢀ; B, abcd z 0ꢀ, a‘b’c‘d’ z 180ꢀ; C, abcd z 180ꢀ, a‘b’c‘d’ z 180ꢀ; D, abcd ¼ 0ꢀ, a‘b’c‘d’ ¼ 180ꢀ).
for the target activity. For example, comparison of the activities for
pairs of homologous compounds, e.g., 3c and 6u or 3d and 6v, with
ALogP ¼ 5.636 and 4.942 for 3c and 6u, respectively, and 6.548 and
5.854 for 3d and 6v, shows that compound 6, with smaller ALogP
value, has superior activity in the cytotoxicity and apoptosis tests
carried out on the three cell lines. Similar observations were made
for the pairs 3f/6aa and 3h/6ac.
Modification of the scaffold from pyrido[2,3-d]pyrimidine to
pyrimidine also leads to a reduction in the value of ALogP 98.
Comparison of the activity of homologous compounds 6 and 8
shows that the pyridopyrimidine derivatives have higher cytotoxic
and pro-apoptotic activities.
The high ALogP values found for some active compounds
contravene Lipinski’s Rule of Five [36], which indicates the “drug-
likeness” and according to which orally active drugs should have:
(a) a molecular weight under 500 Daltons, (b) limited lipophilicity
(expressed by Log P < 5, with P ¼ [drug]org./[drug]aq.), (c)
a maximum number of 5 HBD (expressed as the sum of OH and NH
groups) and (d) a maximum number of 10 HBA (expressed as the
sum of Os and Ns).
The corresponding conformational analysis was carried out,
having selected the rotatable bonds, by applying the Diverse
Conformation Generation protocol implemented in the DS 2.5v
suite. A set of 25e30 representative low energy conformations for
each analyzed compound was selected (the energy differences
between the different conformations analyzed for each trajectory
were in the range 2e5 kcal; data not shown for the sake of brevity).
The low energy conformers were superimposed onto the central
nucleus of quinazoline, pyrido[2,3-d]pyrimidine or pyrimidine, as
appropriate, with the N1, N2 and C2 of the pyrimidine ring taken as
tethers. The effectiveness of the superimposed models was evalu-
ated in terms of the root mean square (rms) values obtained.
The molecular volume was then calculated from the confor-
mational trajectories and the mean values obtained for each
analyzed compound were recorded (see Supplementary data). A
wide range was observed for the mean molecular volumes. For the
selected quinazoline derivatives that show the target activity, the
average volumes are between 1000 and 1250 Å3. In the case of the
selected pyrido[2,3-d]pyrimidine derivatives, the average volumes
are generally lower (869 Å3 for compound 6c and 1172 Å3 for

M. Font et al. / European Journal of Medicinal Chemistry 46 (2011) 3887e3899
3893
Table 2
abcd and a‘b’c‘d’ (Fig. 3), which were selected as patterns for the
analysis. The overall values for the dihedral angles allowed us to
distinguish four conformational families: A, abcd ¼ a‘b’c‘d’ z 0ꢀ; B,
abcd z 0ꢀ, a‘b’c‘d’ z 180ꢀ; C, abcd z 0ꢀ, a‘b’c‘d’ z 180ꢀ; and D,
abcd z 180ꢀ, a‘b’c‘d’ z 0ꢀ. Withrespecttothe1,4-diaminoderivatives,
the preferred conformation for the majority of compounds is type A
and this is followed by type B. Despite this conformational distribu-
tion, the lowest energy conformation for the most active compounds
surprisinglycorresponds tothe type B family (seeSupplementary data
for details). As a representative example, the images obtained in the
conformational study carried out on compounds 3e and 6ac are
shown in Fig. 4. The H located on the N that acts as a linker for the
lateral branch in position 4 is projected towards the pyridine ring in
a similar way to the situation observed for the reference compounds
(Fig. 3), whereas the H on the N acting as a linker for the branch in
position 2 is projected towards N3 of the pyrimidine ring.
The preferred conformation in the 2-(alkylsulfanyl)-N-alkyl-
pyrido[2,3-d]pyrimidin-4-amine derivatives is type A in all cases,
with a value for the abcd dihedral close to 0o. The conformational
variability of the pyrimidin-1,4-diamino derivatives is notable,
since practically all of the studied compounds display a significant
number of conformations assigned to the four families A, B, C and D.
In general, the lateral alkyl or alkylarylamino chains adopt an
extended conformation, thus exploiting the molecular planarity.
Exceptions include compounds like 6r and 8g, as illustrative exam-
ples, in which the lateral branches preferentially adopt a folded
conformation (see Fig. 4). As a general pattern for the more active
compounds, the aromatic rings located at the end of the side chain
are projected differently with respect to the central ring plane; while
the ring on the chain in position 2 preferentially forms an angle of
ꢁ90o with this plane, the ring in position 4 is located in a plane that
formsanangleofꢁ30o(see Fig. 4). Whenthechainissufficientlylong
(n > 3) the presence of OH groups causes the chain to fold, which in
turn allows the formation of intramolecular H-bonds that involve the
pyrimidine ring nitrogens and the sulfur atom in the 2-position, as
well as the possibility of forming H-bonds that affect the OH group
(Fig. 5). This framework of intramolecular H-bonds can stabilize the
folded forms and can also lead to an increase in the value of ALogP.
Molecular descriptors.a
n
R
HN
N
1
N
R''
2
m
R'
X
z
5
4
3
Dipoleb
MEPc
HOMOd
LUMOd
D
EL ꢂ EH
d
Comp.
AlogP 98
3a
3b
3c
3d
3e
3f
3g
3h
3i
3j
3k
3l
3m
6a
6b
6c
6d
6e
6f
3.811
4.723
5.636
6.548
5.233
6.205
6.562
5.200
5.875
6.848
7.372
6.788
7.700
2.803
3.783
4.239
4.695
5.151
1.977
2.557
3.013
4.160
4.616
5.073
5.529
2.354
4.763
5.219
5.670
5.219
6.131
4.942
5.854
0.418
1.578
4.539
5.511
5.868
4.506
5.181
6.678
3.164
5.768
6.094
7.462
3.896
4.868
5.225
3.863
4.538
5.451
3.411
3.319
3.243
3.420
2.966
3.934
2.155
3.942
2.817
3.401
2.500
2.999
3.060
5.146
5.311
5.160
5.151
5.144
5.057
5.129
5.427
5.179
5.074
5.274
5.199
4.846
5.193
5.290
5.239
5.332
5.055
5.625
5.460
5.003
4.063
5.091
5.826
3.832
5.692
5.382
3.749
6.549
6.167
5.096
5.390
2.589
3.424
1.678
3.920
2.510
2.448
1042.9
896.5
888.0
600.7
828.9
1032.8
859.3
1002.9
914.2
614.9
873.7
645.9
755.2
606.7
636.4
564.9
727.7
641.8
848.6
769.1
692.9
631.9
592.1
496.8
509.4
800.4
521.8
538.3
553.6
639.6
561.6
992.7
875.5
1336.5
1354.6
977.4
1085.0
975.3
1219.4
984.9
1031.7
1121.1
1318.8
911.6
826.4
1106.0
1153.8
1114.4
1289.8
1077.9
806.3
ꢂ8.309
ꢂ8.308
ꢂ8.299
ꢂ8.296
ꢂ8.441
ꢂ8.353
ꢂ8.505
ꢂ8.377
ꢂ8.401
ꢂ8.338
ꢂ8.513
ꢂ8.338
ꢂ8.365
ꢂ8.762
ꢂ8.748
ꢂ8.747
ꢂ8.749
ꢂ8.753
ꢂ8.793
ꢂ8.780
ꢂ8.772
ꢂ8.760
ꢂ8.751
ꢂ8.747
ꢂ8.749
ꢂ8.800
ꢂ8.736
ꢂ8.740
ꢂ8.744
ꢂ8.740
ꢂ8.750
ꢂ8.626
ꢂ8.653
ꢂ8.877
ꢂ8.698
ꢂ8.773
ꢂ8.775
ꢂ8.699
ꢂ8.637
ꢂ8.724
ꢂ8.700
ꢂ8.692
ꢂ8.301
ꢂ8.673
ꢂ8.687
ꢂ8.827
ꢂ8.724
ꢂ8.805
ꢂ8.675
ꢂ8.782
ꢂ8.720
ꢂ0.302
ꢂ0.296
ꢂ0.288
ꢂ0.295
ꢂ0.467
ꢂ0.373
ꢂ0.693
ꢂ0.414
ꢂ0.384
ꢂ0.324
ꢂ0.695
ꢂ0.329
ꢂ0.344
ꢂ1.055
ꢂ1.038
ꢂ1.035
ꢂ1.039
ꢂ1.044
ꢂ1.088
ꢂ1.382
ꢂ1.070
ꢂ1.052
ꢂ1.033
ꢂ1.032
ꢂ1.028
ꢂ1.085
ꢂ1.041
ꢂ1.042
ꢂ1.043
ꢂ1.049
ꢂ1.045
ꢂ0.532
ꢂ0.552
ꢂ0.760
ꢂ0.899
ꢂ0.699
ꢂ0.701
ꢂ0.901
ꢂ0.596
ꢂ0.623
ꢂ0.900
ꢂ0.590
ꢂ0.545
ꢂ0.579
ꢂ0.589
ꢂ0.045
0.090
8.007
8.012
8.011
8.001
7.974
7.980
7.812
7.963
8.017
8.014
7.818
8.010
8.021
7.707
7.710
7.712
7.710
7.709
7.705
7.398
7.702
7.708
7.718
7.715
7.721
7.715
7.695
7.698
7.701
7.691
7.705
8.094
8.101
8.117
7.799
8.074
8.074
7.798
8.041
8.101
7.800
8.102
7.756
8.094
8.098
8.782
8.814
8.518
8.729
8.928
8.955
6g
6h
6i
6j
6k
6l
6m
6n
6o
6p
6q
6s
6u
6v
6x
6y
6z
6aa
6ab
6ac
6ad
6ae
6af
6ag
6ah
6ai
8a
8b
8c
The dipole moment (m) was calculated and this was taken as an
expression, from the asymmetry, of the overall molecular charge
distribution and a quantitative measurement of the molecular
polarity. The mean value (expressed in Debye units) for each
trajectory is given in Table 2. Generally, derivatives 3 have lower
m
values than derivatives 6. On analyzing all of the dipole moment
values it is possible to emphasize the greater asymmetry in the
charge distribution for derivatives 6, a situation that is not unex-
pected considering the modification of the central nucleus from
quinazoline to pyrido[2,3-d]pyrimidine.
The molecular electrostatic potential (MEP) values for the
analyzed compounds were calculated for the selected lowest
energy conformations using the DS 2.5v Electrostatic Potential
Protocol (EPP). The molecular electrostatic potential (MEP, in kT/e
units) values were considered for the comparative analysis (see
Supplementary data for details). The distribution of MEP values,
taken as the potential energy of a proton at a particular location
near a molecule, was calculated in order to evaluate the molecular
zones that are able to participate as hydrogen bond donors or
acceptors, with particular attention paid to the nitrogen and sulfur
atoms present in the nuclear scaffolds and in the lateral branch
linkages. Thus, a negative EP corresponds to an attraction of the
proton by the concentrated electron density, whereas a positive EP
corresponds to repulsion of the proton by the atomic nuclei in
regions where low electron density exists and the nuclear charge is
not completely shielded. As a general pattern (see Fig. 6 as an
illustrative example), and with respect to the quinazolin-2,4-
ꢂ0.287
8d
8e
8f
0.054
0.146
0.235
a
Mean values obtained from conformational trajectory.
Debyes.
b
c
kT/e units (1 kT/e ¼ 25.6 mV eV).
d
In eV calculated on the lowest energy representative conformation.
compound 6ac, taken as extreme values for derivatives that show
the target activity and selectivity), although in this case, as for the
active pyrimidine derivatives, the tolerance range is also wide.
The conformations obtained for each compound were distributed
in families, taking as group criteria the overall value of the dihedrals

3894
M. Font et al. / European Journal of Medicinal Chemistry 46 (2011) 3887e3899
Fig. 4. Overlay of low energy conformations (left), taking the nitrogen atoms as tethers, and representative lowest energy conformation (right) for active compounds 3e and 6ac,
and inactive compound 8g.
diamine derivatives 3, the negative EP values are preferentially
located on the N atoms of the pyrimidine ring (N2 and N4) with
a smaller contribution of nitrogens in positions 2 and 4 (N1 and N3).
The MEP values for compounds 6aes are lower than those
obtained for compound 3 and the value for N1 is slightly positive,
whereas that for S is slightly negative. The presence of N5 contrib-
utes an additional negative center and this is able to act as an HBA.
For compounds 6aaeai and for compound 8, the MEP values are
quite similar to those obtained for compound 3 with regard to
nitrogens in the pyrimidine ring. However, the presence of N5 gives
rise to a zone with very negative average values and nitrogen N1 is
more negative than that in compound 3. As a general pattern, the
negative EP values are preferentially located on N atoms in the
central scaffold (N2, N4 and N5 for compound 6, and N2 and N4 for

M. Font et al. / European Journal of Medicinal Chemistry 46 (2011) 3887e3899
3895
Fig. 5. Intramolecular H-bond (dotted line) pattern for some representative compounds.
compound 8), with a smaller contribution from nitrogens in posi-
tions 2 and 4 (N1 and N3).
although N2 and N4 continue to have negative charge values, N1 and
N3 in the branch linkage have a partial positive charge.
Finally, as quantum parameters (Table 2) the energy and
distribution of HOMO 0 and LUMO 0 orbitals, the difference
between the molecular orbital energies (gap), and the net atomic
charge over the target heteroatoms N and S were calculated by
application of MOPAC2009 software [37]. The calculations were
carried out on the lowest energy conformation, in vacuum, using
the PM3 or PM6 semi-empirical approach and with the eigenvector
used as the optimization algorithm.
With respect tothe charge distributionpattern in the quinazoline
derivatives 3, highly negative values were obtained e especially for
the nitrogens in the pyrimidine ring. The exceptions to this trend are
compounds 3g and 3k, in which the presence of a halogen (chlorine
and bromine, respectively) in the aromatic nucleus at the end of the
lateral branches leads to a significant distortion in the pattern, since
The pattern for the pyrido[2,3-d]pyrimidine derivatives was quite
different. All of the 2-(alkylsulfanyl)-N-alkylpyrido[2,3-d]pyrimidin-
4-amine derivatives 6aes display a parallel distribution in which the
negative charges (which are generally less negative than those
obtained for compound 3) are located over the scaffold nitrogens,
whereas the heteroatoms pertaining to the linkage present values
near to 0 (slightly positive for N1 and slightly negative for S).
With respect to the pyrido[2,3-d]pyrimidin-2,4-diamine deriv-
atives, the charge distribution pattern on the target heteroatoms is
quite similar to that found for compound 3, although the presence
of N5 (the pyridine ring on the central heteroaromatic nucleus)
constitutes a new negative point that could modify the binding
mode to the active site. In a similar way to compound 3 (with
halogens as substituents), N1 is positive in derivatives 6ab and 6ae,

3896
M. Font et al. / European Journal of Medicinal Chemistry 46 (2011) 3887e3899
Fig. 6. Schematic HOMO (left), LUMO (middle), MEP (right) distribution for representative active and inactive compounds, calculated from lowest energy conformation.
which contain chlorine and bromine, respectively. The behavior of
the heteroatomic charges in derivative 6y is remarkable in that
nitrogens N1 and N4, which are located in linkages, are both
positive. As previously noted in the conformational analysis, the
preferred conformation for the derivatives bearing a hydroxyl
group at the end of the lateral branches is the one in which these
side chains are folded towards the central ring; in this way intra-
molecular hydrogen bonds can be formed that can alter the charge
distribution of these derivatives.
excited. As far as the gap is concerned, it can be observed that
compound 8 generally have a higher gap than compounds 3 and 6.
Thus, for the analyzed 8 derivatives the mean gap value is greater
than 8.5 eV, whereas for 2-(alkylsulfanyl)-N-alkylpyrido[2,3-d]
pyrimidin-4-amine derivatives the mean gap is near to 7.7 eV. With
respect to the pyrido[2,3-d]pyrimidin-2,4-diamine derivatives, the
mean gap is near to 8.0 eV, which is similar to the values obtained
for the quinazoline derivatives.
The disparity in the data does not allow us to draw firm
conclusions with respect to this parameter and its possible influ-
ence on the target activity, although generally the compounds with
The gap is sometimes considered as a measure of the excitability
of the molecule: i.e., the lower the energy, the more easily it will be

M. Font et al. / European Journal of Medicinal Chemistry 46 (2011) 3887e3899
3897
the most interesting biological profiles (3e, 6c, 6e, 6j, 6p, 6z, 6ac)
have gap values in the range 7.7e8.0 eV.
derivatives (see Supplementary data for details). All compounds
were characterized by 1H NMR (Bruker 400 UltrashieldÔ, Bruker,
Rheinstetten, Germany; TMS as the internal standard), FT-IR
(Thermo Nicolet FT-IR Nexus spectrophotometer, Thermo Electron
Corporation, Waltham, MA, USA; samples as KBr pellets) and EI-MS
(HewlettePackard MSD 5973 N spectrometer, GC 6890plus/DIP;
Agilent Technologies, Inc., Santa Clara, CA, USA) techniques. The
spectroscopic data obtained are consistent with the proposed
structures.
As far as the distribution of the HOMO and LUMO orbitals is
concerned, all of the compounds display a parallel distribution,
with a strong contribution to the HOMO 0 orbital from the S located
in the 2-position of the pyrido[2,3-d]pyrimidine ring in 2-(alkyl-
sulfanyl)-N-alkylpyrido[2,3-d]pyrimidin-4-amine derivatives 6aes.
3. Conclusions
The melting points were determined with Mettler FP82 þ FP80
apparatus (MettlereToledo, Greifensee, Switzerland) and are not
corrected. Elemental microanalyses were carried out on vacuum-
dried samples using an elemental analyzer (LECO, CHN-900
Elemental Analyzer; LECO Corporation, St. Joseph, MI, USA) and
were in an acceptable range of ꢁ0.4% for all compounds.
The experimental and spectroscopic data for the synthesized
compounds are provided in Supplementary data.
The data obtained in this study allowed us to validate the design
process carried out for the compounds under investigation. The
preliminary structureeactivity relationship proposed can be
summarized as follows.
With respect to the evaluated scaffold, the preferred central
nucleus proved to be the pyrido[2,3-d]pyrimidine unit, in which the
nitrogens can act as HBAs.
The selected linkage with the lateral branches is the amino
group, which in the case with a nitrogen at position 4 can act as an
HBA or HBD, whereas the nitrogen at position 2 acts as an HBD.
The preferred length for the lateral branches is three or four
methylene groups or one or two methylene groups with an
aromatic nucleus at the end, and the best activity profile was shown
by derivatives that bear the same chain in positions 2 and 4 of the
central ring. The presence of polar groups at the ends of the lateral
amino branches leads to loss of activity. Although the tolerance for
the lipophilic values was broad, the best activity data were found
for compounds with ALogP 98 values between 4.5 and 5.5.
The lowest energy conformation for the majority of compounds
was characterized as type D, in which the NH at position 4 points
towards the pyridine ring and the NH at position 2 points towards
nitrogen 3 of the pyrimidine ring.
4.1.1. 1,2,3,4-tetrahydro-2,4-dioxopyrido[2,3-d]pyrimidine (4)
From 2-aminonicotinic acid and urea, according to previously
reported methods [11].
4.1.2. 2,4-dichloropyrido[2,3-d]pyrimidine (5)
A mixture 4 (6.0 mmol), POCl3 (5 mL) and N,N-DMF (catalytic
amount) was stirred and heated under reflux for 48 h. The solvents
were removed under vacuum and cold water (0 ꢀC, 25 mL) and
chloroform (25 mL) were added. The organic layer was washed
with water (3 ꢃ 20 mL) and dried over anhydrous sodium sulfate.
The solvent was removed under vacuum and the product was used
immediately without further purification.
4.1.3. General procedure for compound 6
As discussed in our previous reports, the structural profile of
the selected compounds (see 3e and 6ac structures as an example)
is quite different from that deduced on considering the related
compounds taken as structural references for our design, partic-
ularly in terms of the elements that act as hydrogen bond donor
(HBDs) or acceptor (HBAs), the alteration of the substitution
pattern in the central nucleus, and the lipophilic profile. Thus,
taking Piritrexim as an example of a structurally related anti-
tumoral (Fig. 2), it is worth highlighting the presence of free eNH2
groups at the 2- and 4-positions of the pyrimidine, which
presumably act as HBDs, while the nitrogens of the pyrido[2,3-d]
pyrimidine ring can act as HBAs, and the presence of substituents
(e.g., 2,5-dimethoxybenzyl for Piritrexim) at the 6-position of the
pyridopyrimidine ring. Concerning compounds 3e and 6ac, the
nitrogens at positions 2 and 4 act as carriers for benzylic groups,
whereas the benzene ring of 3e or the pyridine ring of 6ac are
unsubstituted (Fig. 2).
A mixture of 5 (5.0 mmol), the respective amine (12 mmol),
equimolecular amounts of triethylamine, and ethanol (15 mL) was
heated at 70 ꢀC for 5 h with stirring. The solvent was removed
under vacuum and the residue was dissolved in water (30 mL); the
mixture was extracted with chloroform (3 ꢃ 25 mL) and the organic
extracts were dried over anhydrous sodium sulfate and the solvent
was removed in vacuum.
4.1.3.1. N,N0-dihexylaminopyrido[2,3-d]pyrimidin-2,4-diamine (6v).
From hexylamine.
4.1.3.2. N,N0-bis(3-hydroxypropyl)pyrido[2,3-d]pyrimidin-2,4-diamine
(6x). From 3-amino-1-propanol.
4.1.3.3. N,N0-bis(4-hydroxybutyl)pyrido[2,3-d]pyrimidin-2,4-diamine
(6y). From 4-amino-1-butanol.
On the basis of the analyzed structural profiles, different
mechanisms of action can be suggested for the quinazolin-2,4-
diamine, the 2-(alkylsulfanyl)-N-alkylpyrido[2,3-d]pyrimidin-4-
amine and the pyrido[2,3-d]pyrimidin-2,4-diamine derivatives.
The precise mechanism of action of some of these derivatives is
currently under investigation in our laboratory. Optimization of our
pattern from the structureeactivity relationship data, as well as the
elucidation of the mechanisms of action, could very well lead to the
development of novel drug leads in this field.
4.1.4. General procedure for compound 8
A mixture of commercially available 2,4-dichloropyrimidine 4
(10 mmol), the selected amine (30 mmol), equimolecular amounts
of triethylamine and ethanol (15 mL) was heated at 70 ꢀC for 5 h
with stirring. The solvent was removed under vacuum and the
residue was washed with 5% HCl (25 mL). The resulting solid was
isolated and purified by recrystallization.
4.1.4.1. N,N0-dibenzylaminopyrimidin-2,4-diamine
(8a). From
benzylamine.
4. Experimental protocols
4.1.4.2. N,N0-bis(4-methylphenyl)methylpyrimidin-2,4-diamine (8b).
4.1. Chemistry
From 4-methylbenzylamine.
The synthesis of the new compounds was developed according
to previously reported methods, which were adapted to the target
4.1.4.3. N,N0-bis(4-chlorophenyl)methylpyrimidin-2,4-diamine (8c).
From 4-chlorobenzylamine.

3898
M. Font et al. / European Journal of Medicinal Chemistry 46 (2011) 3887e3899
4.1.4.4. N,N0-bis(4-methoxyphenyl)methylpyrimidin-2,4-diamine
4.3. Molecular modeling
(8d). From 4-methoxybenzylamine.
The initial computational work was performed on a Dell Preci-
sion 380 workstation provided with the software package
Discovery Studio 2.5v (DS 2.5v) and with the MOPAC2009 package.
The starting atomic coordinates were obtained from the struc-
turally related Brookhaven Protein Data Bank (PDB) and CSD
structures cited above (CSD System version 5.31: search and
information retrieval with ConQuest version 1.12 [39], structure
visualization with Mercury CSD [39] version 2.3).
4.1.4.5. N,N0-bis(2-phenylethyl)pyrimidin-2,4-diamine
2-phenylethylamine.
(8e). From
4.1.4.6. N,N0-bis(3-phenylpropyl)pyrimidin-2,4-diamine (8f). From
3-phenylpropylamine.
4.1.4.7. N,N0-bis(4-phenylbutyl)pyrimidin-2,4-diamine
(8g). From
4-phenylbutylamine.
The three-dimensional models of the studied compounds were
constructed, in the vacuum phase, using as starting fragments the
PDB and CSD related structures, as well as atoms and structural
fragments from the Viewer module (DS 2.5v) and using the
Dreiding force field [40]. Once the models had been constructed,
a preliminary conformational analysis was carried out. The applied
protocol (Diverse Conformational Generation integrated DS 2.5v
protocol) can be summed up as follows: (a) Initial construction of
the model and first minimization by application of the Dreiding
minimize protocol (steepest descent algorithm with a convergence
criterion of 10e6). The AlogP 98 [24,25] descriptor (an imple-
mentation of the atom-based ALogP method) was calculated for
each compound. (b) Application of the BEST routine for confor-
mation generation (first: conjugate-gradient minimization in
torsion space; second: conjugate-gradient minimization in Carte-
sian space; third: QuasieNewton minimization in Cartesian space).
(c) Elimination of those conformations whose relative energy is
greater than 10 kcal/mol at a global minimum. (d) Analysis of
conformational trajectory and selection of representative lowest
energy conformations. Root mean square (rms) deviations of the
structures were monitored. The energy differences between the
different conformations analyzed for each trajectory were around
5 kcal.
4.2. Biological assays
The five human cell lines were obtained from the American
Tissue Culture Collection (Manassas, VA, USA): HT-29 (ATCC HTB
38) is a colon adenocarcinoma cell line, T-24 (ATCC HTB-24) was
obtained from urinary bladder cancer, MDA-MB-231 (ATCC
HTB26) was established from adenocarcinoma of mammary
gland, CRL-8799 was obtained from breast epithelium (ATCC
184B5) and CRL-11233 (ATCC THLE-3) from human liver. HT-29
and T-24 cells were cultured in McCoy’s medium (GibcoeInvi-
trogen, Barcelona, Spain), MDA-MB-231 in Leibovitz (Gibco), CRL-
8799 in MEG (Clonetics Corporation) and CRL-11233 in BEGM
(BEGM Bullet kit, Clonetics Corporation, San Diego, CA, USA).
These media were supplemented with 10% fetal bovine serum,
penicillin (50 U/mL) and streptomycin (50 mg/mL). Cells were
grown as a monolayer in 175 mL flasks (Corning) and were
incubated at 37 ꢀC in a humidified atmosphere containing 5%
CO2. The evaluated compounds were dissolved in dimethylsulf-
oxide (DMSO). The DMSO concentration was equalized in all
media. In all cases, the concentration of solvent in the culture
medium did not exceed 0.5% (v/v).
The evaluation of cytotoxic potential and selectivity against
human cell lines was determined according to previously reported
methods [9,10], by using the neutral red assay as described by
Lowik et al. [38] (see Supplementary data for details). Survival
percentage was determined at the screening concentrations of 20
For each of the compounds, 25e30 lowest energy conformations
were selected and a new minimization cycle was applied. The
volumes of the whole molecule were also calculated.
The molecular electrostatic potential (MEP) values were calcu-
lated using the DS 2.5v implemented Electrostatic Potential
Protocol (EPP). This protocol calculates the electrostatic potential of
a molecule by solving a nonlinear PoissoneBoltzmann equation
using the finite difference method [41], as implemented in the DS
Delphi 4v [42,43] release 1.1 algorithm. The potential was calcu-
lated on grid points per side (65 ꢃ 65 ꢃ 65) and the percentage of
box to be filled was set to 50%. The outer dielectric and the
dielectric in the medium were set to 80.0 and 2.0, respectively. An
ionic exclusion radius of 2.0 Å, a solvent radius of 1.4 Å, and
a solvent ionic strength of 0.145 M were applied. CHARMM [44,45]
charges and radii were used for this calculation.
The mechano-quantic analysis of the conformations obtained in
the previous step was carried out with the package Mopac2009,
PM6 [46] (or PM3 [47] for halogen and/or sulfur-containing
derivatives) semi-empirical approaches, with the geometry opti-
mized using an eigenvector following algorithm. The energy and
distribution of the HOMO and LUMO orbitals, the atomic charges,
the electronic density (ED) and the dipolar moment were obtained
for each of the conformations. The data corresponding to the mean
value of the representative low energy conformations, selected
from the conformational trajectory for each compound, was used to
establish the preliminary structureeactivity relationships.
and 100 mM, using the survival percentage obtained with the cells
treated only with the solvent (DMSO at 0.5%) as a reference.
The results are expressed as the average of triplicate assays. IC50
values were calculated for those compounds showing less than 45%
survival in the cell lines at 100 mM concentration. This value was
determined in triplicate by curvilinear-regression analysis using
the statistical program SPSS 11.0 (SPSS Inc., Chicago, IL, USA).
With regard to selectivity, cytotoxicity was determined in cell
cultures of two non-tumoral lines, CRL-7899 and CRL-11233. The
highest IC50 calculated in the three tumoral lines was selected as
the test concentration for assays on non-tumoral cells.
The apoptosis induction in human cancer cell lines was quan-
tified through the DNA fragmentation detected at 48 h after treat-
ment using the Cell Death detection ELISA Plus kit (Roche
Diagnostics, Indianapolis, IN, USA) according to previously reported
methods [9,10]. A relative value of 1 was attributed to the apoptosis
detected in the control cultures in which the test compound was
not present.
The evaluation of caspase-3 activation was carried out by means
of cytometry, using the Active-Caspase-3 FITC Mab apoptosis kit
(Pharmingen), according to previously reported methods [9,10].
Measurements were taken at 14, 24 and 48 h and the values
obtained were compared with those of control cells incubated
without the test compounds. The test concentrations correspond to
the IC50 values determined in the cytotoxicity assay. Cytometry was
performed on a FACSCAN (Becton Dickinson).
Acknowledgments
The authors wish to express their gratitude to M.V. Domínguez, J.
Ramiro, X. Sáenz, S. Leoz, M. Echeverría, H. Chozas and E. Ardaiz for
their collaboration in the synthetic processes and to J. García-

M. Font et al. / European Journal of Medicinal Chemistry 46 (2011) 3887e3899
3899
[20] T. Hida, S. Ogawa, J.C. Park, J.Y. Park, J. Shimizu, Y. Horio, K. Yoshida, Expert.
Rev. Anticancer. Ther. 9 (2009) 17e35.
[21] C. Schettino, M.A. Bareschino, V. Ricci, F. Ciardiello, Expert Rev. Respir. Med.
(2008) 167e168.
Foncillas, L. Cordeu and E. Cubedo for their collaboration in the
biological evaluation.
[22] M. Zink, H. Lanig, R. Troschütz, Eur. J. Med. Chem. 39 (2004) 1079e1088.
[23] H.J. Park, Y.S. Kim, J.S. Kim, E.J. Lee, Y.J. Yi, H.J. Hwang, M.E. Suh, C.K. Ryu,
S.L. Lee, Bioorg. Med. Chem. 14 (2004) 3385e3388.
[24] V.N. Viswanadhan, A.K. Ghose, G.R. Revankar, R.K. Robins, J. Chem. Inf. Com-
put. Sci. 29 (1989) 163e172.
[25] A.K. Ghose, V.N. Viswanadhan, J.J. Wendolowski, J. Phys. Chem. A. 102 (1998)
3762e3372.
[26] D. Barnum, J. Greene, A. Smellie, P. Sprague, J. Chem. Inf. Comput. Sci. 36
(1996) 563e571.
Appendix. Supplementary material
Supplementary material can be found, in the online version, at
doi:10.1016/j.ejmech.2011.05.060.
References
[27] S. Putta, G.A. Landrum, J.E. Penzotti, J. Med. Chem. 48 (2005) 3313e3318.
[28] F.H. Allen, Acta. Crystallogr. B58 (2002) 380e388.
[1] K.M. Hajra, J.R. Liu, Apoptosis 9 (2004) 691e704.
[2] J.C. Reed, Cancer. Cell. 3 (2003) 17e22.
[29] H.M. Berman, J. Westbrook, Z. Feng, G. Gilliland, T.N. Bhat, H. Weissig,
I.N. Shindyalov, P.E. Bourne, Nucleic. Acids. Res. 28 (2000) 235e242.
[30] H.M. Berman, K. Henrick, H. Nakamura, Nat. Struct. Biol. 10 (2003) 980.
[31] R. Tanaka, M. Hamamura, A. Tanaka, N. Hirayama, Anal. Sci. 20 (2004) 173.
[32] P.A. Sutton, V. Cody, J. Am. Chem. Soc. 10 (1988) 6219.
[33] V. Cody, J.R. Luft, W. Pangborn, A. Gangjee, Acta Crystallogr. Sect. D 59 (2003)
1603e1609.
[3] J. Baselga, S.M. Swain, Nat. Rev. Cancer. 9 (2009) 463e465.
[4] C.H.J. Brown, S. Lain, C.H.S. Verma, A.R. Fersht, D.P. Lane, Nat. Rev. Cancer.
9 (2009) 862e873.
[5] C. Sanmartín, M. Echeverría, B. Mendivil, L. Cordeu, E. Cubedo, J. García-Fon-
cillas, M. Font, J.A. Palop, Bioorg. Med. Chem. 13 (2005) 2031e2044.
[6] C. Sanmartín, E. Ardaiz, L. Cordeu, E. Cubedo, J. García-Foncillas, M. Font,
J.A. Palop, Lett. Drug Des. Discovery 2 (2005) 341e354.
[34] P.A. Sutton, V. Cody, J. Med. Chem. 30 (1987) 1843e1848.
[35] M. Liu, S. Wang, J.E. Clampit, R.J. Gum, D.L. Haasch, C.M. Rondinone,
J.M. Trevillyan, C. Abad-Zapatero, E.H. Fry, H.L. Sham, G. Liu, Bioorg. Med.
Chem. Lett. 17 (2007) 668e672.
[36] C.A. Lipinski, F. Lombardo, B.W. Dominy, P.J. Feeney, Adv. Drug. Deliv. Rev. 23
(1997) 3e25.
[37] MOPAC2009, J.P. Stewart, Stewart Computational Chemistry, version 9.062L
web, <http://OpenMOPAC.net>.
[38] C.W. Lowik, M.J. Alblas, M. van de Ruit, S.E. Papapoulos, G. van der Pluijm,
Anal. Biochem. 213 (1993) 426e433.
[39] I.J. Bruno, J.C. Cole, P.R. Edgington, M. Kessler, C.F. Macrae, P. McCabe,
J. Pearson, R. Taylor, Acta. Crystallogr. B58 (2002) 389e397.
[40] S.L. Mayo, B.D. Olafson, W.A. Goddard, J. Phys. Chem. 94 (1990) 8897e8909.
[41] W. Rocchia, S. Sridharan, A. Nicholls, E. Alexov, A. Chiabrera, B. Honig,
J. Comput. Chem. 23 (2002) 128e137.
[42] M.K. Gilson, B. Honig, Protein Struct. Funct. Genet. 4 (1988) 7e18.
[43] W. Rocchia, E. Alexov, B. Honig, J. Phys. Chem. B. 105 (2001) 6507e6514.
[44] B.R. Brooks, R.E. Bruccoleri, B.D. Olafson, D.J. States, S. Swaminathan,
M. Karplus, J. Comput. Chem. 4 (1983) 187e217.
[7] M. Font, E. Ardaiz, L. Cordeu, E. Cubedo, J. García-Foncillas, C. Sanmartín,
J.A. Palop, Bioorg. Med. Chem. 14 (2006) 1942e1948.
[8] E. Echeverría, B. Mendívil, L. Cordeu, E. Cubedo, J. García-Foncillas, M. Font,
C. Sanmartín, J.A. Palop, Arch. Pharm. Chem. Life Sci. 339 (2006) 182e192.
[9] E. Cubedo, L. Cordeu, E. Bandrés, A. Rebollo, R. Malumbres, C. Sanmartín,
M. Font, J.A. Palop, J. García-Foncillas, Cancer. Biol. Ther. 5 (2006) 850e859.
[10] L. Cordeu, E. Cubedo, E. Bandres, A. Rebollo, X. Saenz, H. Chozas,
M.V. Domínguez, M. Echeverría, B. Mendívil, C. Sanmartín, J.A. Palop, M. Font,
J. García-Foncillas, Bioorg. Med. Chem. 15 (2007) 1659e1669.
[11] C. Sanmartín, M.V. Domínguez, L. Cordeu, E. Cubedo, J. García-Foncillas,
M. Font, J.A. Palop, Arch. Pharm. Chem. Life Sci. 341 (2008) 28e41.
[12] M. Vieth, M. Siegel, Structural fragments in marketed oral drugs Ch. 6. in:
W. Jahnke, D.A. Erlanson (Eds.), Fragment Based Drug Discovery. Wiley-VCH
Verlag GmbH & Co. KGaA, 2006.
[13] M. Vieth, M.G. Siegel, R.E. Higgs, I.A. Watson, D.H. Robertson, K.A. Savin,
G.L. Durst, P.A. Hipskind, J. Med. Chem. 47 (2004) 224e232.
[14] J.J. Sutherland, R.E. Higgs, I. Watson, M. Vieth, J. Med. Chem. 51 (2008) 2689e2700.
[15] C. Sanmartín, M. Font, J.A. Palop, Mini-Rev. Med. Chem. 6 (2006) 639e650.
[16] M.Y. Cha, K.O. Lee, J.W. Kim, C.H.G. Lee, J.Y. Song, Y.H. Kim, G.S. Lee, S.B. Park,
M.S. Kim, J. Med. Chem. 52 (2009) 6880e6888.
[45] A.D. MacKerell, B. Brooks, C.L. Brooks, L. Nilsson, B. Roux, Y. Won, M. Karplus,
in: P.v.R. Schleyer (Ed.), The Encyclopedia of Computational Chemistry, Wiley,
Chichester, U.K, 1998, pp. 271e277.
[17] J.A. Bikker, N. Brooijmans, A. Wissner, T.S. Mansour, J. Med. Chem. 52 (2009)
1493e1509.
[46] J.J.P. Stewart, J. Mol. Model. 13 (2007) 1173e1213.
[47] J.J.P. Stewart, J. Comput. Chem. 10 (1989) 209e220.
[18] J.R. Pollard, M. Mortimore, J. Med. Chem. 52 (2009) 2629e2651.
[19] S. Nandi, M.C. Bagchi, Mol. Divers. 14 (2010) 27e38.