Novel 5,7-disubstituted 6-amino-5H-pyrrolo[3,2-b]pyrazine-2,3-dicarbonitriles, the promising protein kinase inhibitors with antiproliferative activity

Article abstract of DOI:10.1016/j.ejmech.2006.03.019

New derivatives of pyrrolo[2,3-b]pyrazine were synthesized and tested on a panel of cultured human tumor cell lines. It was found that 6-amino-5-(3-chlorophenylamino)-7-(1-methyl-1H-benzo[d]imidazol-2-yl)-5H-pyrrolo[3,2-b]pyrazine-2,3-dicarbonitrile (4j)

Full text of DOI:10.1016/j.ejmech.2006.03.019

European Journal of Medicinal Chemistry 41 (2006) 727–737
http://france.elsevier.com/direct/ejmech
Original article
Novel 5,7-disubstituted 6-amino-5H-pyrrolo
3,2-b]pyrazine-2,3-dicarbonitriles,
[
the promising protein kinase inhibitors with antiproliferative activity
*
G.G. Dubinina , M.O. Platonov, S.M. Golovach, P.O. Borysko
A.O. Tolmachov, Y.M. Volovenko
ChemBio Center, Kiev National University, 6 Sosury Street, 02090 Kiev, Ukraine
Received in revised form 13 March 2006; accepted 14 March 2006
Available online 03 May 2006
Abstract
New derivatives of pyrrolo[2,3-b]pyrazine were synthesized and tested on a panel of cultured human tumor cell lines. It was found that 6-
amino-5-(3-chlorophenylamino)-7-(1-methyl-1H-benzo[d]imidazol-2-yl)-5H-pyrrolo[3,2-b]pyrazine-2,3-dicarbonitrile (4j) exhibited a significant
antiproliferative activity: GI50 for cell lines RXF 393 (renal cancer) and BT-549 (breast cancer) were 14 and 82 nM, respectively. To identify
possible molecular targets, docking of the most active compounds into the active sites of cyclin-dependent kinases was performed. Molecular
modeling of the inhibitor–enzyme complexes showed the differences in the binding poses of new pyrrolo[2,3-b]pyrazine derivatives in the kinase
ATP-binding site compared with known pyrrolo[2,3-b]pyrazine inhibitors called aloisines. The patterns of drug kinase interactions correlated well
with antiproliferative activities of novel derivatives. Key interactions and binding mode of docked compounds are discussed.
©
2006 Elsevier SAS. All rights reserved.
Keywords: Antiproliferative agents; Cytostatic agents; Pyrrolo[2,3-b]pyrazine; Kinase inhibitor; Molecular modeling; Docking
1
. Introduction
Aloisines bind to the kinase ATP-binding pocket as compe-
titive inhibitors [4,6]. The structures of the active sites and in-
teraction patterns of the kinases with aloisine A and B (aloisine
B-CDK2 and aloisine A-CDK5 (1ung.pdb) complexes; [4])
have been investigated by the X-ray method. The authors
found that the key interactions were the two hydrogen bonds
with the backbone nitrogen and oxygen atoms of Leu 83
(CDK2) and Cys83 (CDK5) formed by the nitrogen atom in
position 4 and the hydrogen atom adjacent to N(5) of the pyr-
rolo[2,3-b]pyrazine cycle. The third hydrogen bond was
formed between the nitrogen atom N(1) and ε-amino group of
Lys33 (aloisine B/CDK2) [4] and the nitrogen N(1) of aloisine
A was engaged in an hydrogen bonding network involving the
side chains of Lys33, Glu51, Asn144, and two water molecules
(CDK5) [6]. Structure–activity relationship (SAR) studies of
The derivatives of pyrrolo[2,3-b]pyrazine are known to be
biologically active. In addition to having antibronchospastic
effect [1] and the ability to inhibit the activity of p38 MAP
kinase [2], the compounds of this class can also inhibit cy-
clin-dependent kinases (CDKs) and glycogen synthase kinase-
3
(GSK-3), thereby exerting an antiproliferative effect [3–6].
The most efficient pyrrolo[2,3-b]pyrazine derivatives are aloi-
sines. Fig. 1 shows the structures of aloisines A and B.
Abbreviations: CDK, cyclin-dependent kinase; EGFR, epidermal growth
factor receptor kinase; GI50, growth inhibition; GSK-3, glycogen synthase
kinase-3; LC50, lethal concentration; LC/MS, liquid chromatography/mass
spectrometry; MAP, mitogen activated protein kinase; PG, percentage
growth; RMSD, root mean square deviation; SAR, structure–activity
50 aloisine derivatives with different substituents in the posi-
R
relationship; SRB, sulforhodamine B; TGI, total growth inhibition; t ,
retention time.
tions 2, 3, 7a–e (Fig. 1) were performed [4]. The replacement
by carbon atom or an alkylation of any of the nitrogen atoms
that formed the essential hydrogen bonds dramatically de-
*
Corresponding author. Tel.: +380 44 206 4579; fax: +380 44 537 3253.
E-mail address: g.dubinina@univ.kiev.ua (G.G. Dubinina).
0223-5234/$ - see front matter © 2006 Elsevier SAS. All rights reserved.
doi:10.1016/j.ejmech.2006.03.019

728
G.G. Dubinina et al. / European Journal of Medicinal Chemistry 41 (2006) 727–737
Fig. 1. Aloisine A and B, the inhibitors of CDKs and GSK-3.
creased the inhibitory activity. Introducing one, two, three
methoxy groups at the positions a–d of the phenyl ring or its
replacement by five or six membered heterocycles and the pre-
sence of any substituents at the position 2,3 of pyrrolo[2,3-b]
pyrazine cycle decreased the inhibitory activity. In contrast, the
presence of alkyl substituent at the position 7 and chlorine
atom or hydroxy group at the c-position of the phenyl ring
increased the anti-enzymatic effect. These facts correlated well
with X-ray data on interaction of these groups with the side
chains and hydrophobic pockets of the binding sites [4,6].
described earlier in [7], and the derivatives 4f–j were obtained.
Compounds 6a, b were synthesized as salts presented by two
resonance structures, where the negative charge is delocalized
between the carbon atom of the acetonitrile moiety and the
nitrogen atom of the pyrazine ring. The salt formation can be
explained by high C–H acidity of the carbon atom that has
three electron-withdrawing groups. Thus, a strong C–H acid
and hydrochloric acid were formed in this reaction. These acids
competitively protonated triethylamine (TEA) used as a base.
Due to this reason, the highest yields were obtained with
2
equiv. of TEA. In the previous study we investigated the
formation of a salt product in the reaction of 5,6-dichloropyr-
azine-2,3-dicarbonitrile with malononitrile and TEA as a base,
2
. Chemistry
and found that the C–H acid with pK = 0.84 was obtained [8].
The interaction of compounds 6a, b with primary amines
formed pyrrolo[2,3-b]pyrazine 7a–j.
a
In this work, 20 new pyrrolo[2,3-b]pyrazine derivatives
were synthesized, namely, 5,7-disubstituted 6-aminopyrrolo
2,3-b]pyrazine-2,3-dicarbonitriles 4a–j, 7a–j (Scheme 1). A
two-stage nucleophilic substitution of chlorine atoms in 5,6-di-
chloropyrazine-2,3-dicarbonitrile 1 with C- and N-nucleophiles
was used for building the pyrrolo[2,3-b]pyrazine cycle
[
1
H NMR spectra of compounds 4a–j, 7a–j showed a singlet
of the amino group protons at 8.21–9.85 ppm (s, 2H, NH ) that
2
disappeared after the addition of D O. Singlets of the labile
2
protons of compounds 3a–e at 12.58–12.76 ppm were absent
(
Scheme 1). α-Azaheteroarylacetonitriles 2a–e and 2-(phenyl-
in the spectra of compounds 4a–j, and multiplet signals of
sulfonyl)acetonitriles 5a, b were used in the first step as C-nu-
cleophiles (Scheme 1) to produce 3a–e and 6a, b. These pro-
ducts can exist in two tautomeric forms, A and B [7]. The next
step was the nucleophilic substitution of the second chlorine
atom by N-nucleophiles, primary amines and 3-chlorophenyl-
hydrazine (for compound 4j), followed by the addition of the
secondary amine to nitrile group and formation of the pyrrolo
+
H–N (Et)
of compounds 6a, b at 10.74–11.20 ppm were ab-
3
sent in the spectra of products 7a–j. In compounds 3a–e, 6a, b
the characteristic strong band of the conjugated nitrile group of
the acetonitrile moiety was observed in IR spectra
−1
(2195–2185 cm ). This band was not present in the IR spectra
of products 4a–j, 7a–j. Low intensity bands of pyrazine nitrile
−1
[
2,3-b]pyrazine cycle. Compounds 4a–e were synthesized as
bonds were observed in the 2227–2210 cm range.
Scheme 1.

G.G. Dubinina et al. / European Journal of Medicinal Chemistry 41 (2006) 727–737
729
3
. Antiproliferative activity studies
cer) and BT-549 (breast cancer) were 14 and 82 nM, respec-
tively. Compound 7j showed selective activity against HCC-
2998 cell line, while its potency for other lines was negligible.
The antiproliferative activity of novel compounds was in-
The GI50 and LC50 values for compound 4j differed by a factor
of ~2430. The SAR analysis of substituents at position 5 of
pyrrolo[2,3-b]pyrazine cycle showed that the addition of one
carbon atom at this position from benzyl to phenethyl moiety
(Table 1, compounds 4c, e; 7a, i) decreased the antiprolifera-
tive activity. Introducing methoxy groups (4c, e; 7e) or methy-
lenedioxy moiety (4i, 7f) to the phenyl ring of the substituent at
position 5, or replacing the phenyl ring by oxygen containing
ring (furan 4a, b, tetrahydrofuran 4d), resulted in the loss of
the activity. However, the presence of hydrophobic halogen
atom in the m,p-position of phenyl ring at position 5 of pyrro-
lo[2,3-b]pyrazine cycle (compounds 4g, h, j; 7d, j) increased
the antiproliferative efficacy.
vestigated in National Cancer Institute, USA (Developmental
Therapeutics Program). First, a pre-screening was performed
using MCF-7 (breast cancer), NCI-H460 (non-small cell lung
cancer) and SF-268 (glioma) cell lines; all compounds were
tested at 100 μM concentration. Compounds that demonstrated
antiproliferative activity ≥ 68% were selected for advanced
testing in a panel of 60 cancer cell lines at five concentrations
that differed 10-fold (from 10 nM to 100 μM).
4
. In silico analysis
Molecular modeling based on docking approach was used
to find possible target(s) of the antiproliferative activity. Dock-
ing of new pyrrolo[2,3-b]pyrazine derivatives into the active
sites of CDK2, CDK5, GSK-3 and EGFR was performed. To
estimate the binding of synthesized compounds to the active
sites of these kinases and to tune docking parameters, we in-
vestigated the known X-ray ligand–kinase complexes and cre-
ated a training set of known inhibitors for each kinase. The
docking of known aloisine analogs and SAR information [4]
were used for scoring function selection for estimation of li-
gand-CDK2, 5 and GSK-3 complexes. An in-house software
package, Multi-Filter, was used for geometric filtration. The
main idea behind geometry filters was based on the fact that
molecular docking (Flo+, QXP) can predict the molecular geo-
metric stationary point with sufficient accuracy [9], proved by
a number of X-ray structural analyses. Other characteristics,
such as energy, sometimes cannot be calculated with sufficient
precision by the force-field method because force-field approx-
imations do not take into account some important factors, e.g.
specific properties of the charge density distribution. The main
purpose of geometry filtering was to define the compounds that
fit the binding pocket, meaning an efficient ligand–receptor in-
teraction (high affinity). The selected compounds had the best
fitting in the active site and the lowest values for the steric and
chemical repulsive interactions with the receptor. The contact
and hydrogen bond energy values (pI, Cntc, Hbnd) were used
as additional filtering parameters (Tables 3 and 4).
To identify possible targets, docking of the synthesized
compounds into the active sites of CDK2, CDK5, GSK-3 and
epidermal growth factor receptor (EGFR) kinases was per-
formed. Docking of known aloisine analogs and SAR informa-
tion [4] were used as a training set to select a scoring function
for estimating the interactions of new pyrrolo[2,3-b]pyrazine
derivatives with the binding sites of these kinases. The main
affinity criterion was the presence of two hydrogen bonds
formed by the ligand with Glu81 and Cys83 amino acid resi-
dues of CDK5 in the ligand–receptor complex. The 2.5 Å dis-
tance was assumed as the largest possible distance for hydro-
gen bond formation. The aloisine A/CDK5 complex (1ung.
pdb) [6] was investigated to identify the key points of interac-
tion. Fifty-one aloisine derivatives in the tested set were
docked to the CDK5 site model. Filtering criteria were devel-
oped based on the activity data of the investigated aloisines [4]
(Table 3). Next, docking of 51 aloisines to GSK-3 (1UV5) [10]
was performed and the same trends were observed. Similarly to
CDK5, the formation of two hydrogen bonds involving
Asp133 and Val135 was critical for efficient GSK-3 binding
as ATP-competitive inhibitors. To obtain the geometry and en-
ergy filtering criteria, the complexes of CDK2 (1AQ1) [11]
and EGFR (1M17) [12] were analyzed and docked using a
set of known inhibitors of different chemical classes [5,13–
15]. Similar to CDK5 and GSK-3, it was found that two hydro-
gen bonds were necessary for high calculated affinity. The
bonds were formed between the Glu81 and Leu83 amino acid
residues for CDK2, and between Ser766 and Gln767 for
EGFR. These patterns may be used as hinge region-directed
filters for ATP-competitive inhibitor searching. Nevertheless,
the general geometric filters which were applied for post-dock-
ing inhibitor selection were not so rigid: one distance was
5.5 Å from any atom of ligand to the marked atom deep inside
the receptor pocket and the second distance was 2.5 Å from the
corresponded above amino acid residues (Table 3). Com-
pounds not satisfying the requirements of the geometric filters
were classified as non-active (Table 4). The calculated energy
scoring functions for the rest of compounds which were passed
through geometric filters successfully, were compared with
such calculated parameters of the training sets of the known
5
. Results and discussion
Seventeen newly synthesized compounds 4, 7 (Table 1)
were tested for growth inhibitory activity against human cancer
cell lines (in collaboration with National Cancer Institute).
Compounds 4f–h, j and 7d, j that demonstrated antiprolifera-
tive activity ≥ 68% at a drug concentration of 100 mM were
selected for advanced testing in a panel of 60 cancer cell lines
at five concentrations that differed 10-fold (from 10 nM to
1
00 μM). Table 2 shows the data on compounds that sup-
pressed cell proliferation at nanomolar concentrations (a 50%
growth inhibition GI50 < 1 μM).
It was found that compound 4j displayed a significant anti-
proliferative activity: GI s for cell lines RXF 393 (renal can-
5
0

730
G.G. Dubinina et al. / European Journal of Medicinal Chemistry 41 (2006) 727–737
Table 1
Antiproliferative activity of compounds 4a–j, 7a–j
Compounds Substituents
Antiproliferative activity (%) at 100 μM
Selected for
further studies
a
of tested compounds*
X; R
1
R
Breast cancer
MCF-7
9
lung cancer
NCI-H460
30
CNS tumor
SF-268
1
4a [7]
4b [7]
4c [7]
X = NH
No
No
No
X=N-CH
2
Ph
25
31
8
X= N-CH
3
–2
–1
–31
4
d [7]
e [7]
X =S
5
11
–21
–6
No
No
4
X =CONH
29
–7
4f
X =N-CH
3
74
94
11
Yes
4
4
4
4
7
g
h
i
X =N-CH
X =N-CH
X =N-CH
X =N-CH
3
3
3
3
67
55
36
82
40
74
83
9
–4
Yes
Yes
No
–10
–13
55
j
87
54
Yes
No
a
R
R
1
1
= H
= H
18
7
b
43
42
12
No
7
7
c
R
R
1
1
= H
= H
ND
31
ND
83
ND
d
–1
Yes
No
7
e
f
R
R
1
1
= H
= H
16
14
36
34
2
0
7
No
(
continued)

G.G. Dubinina et al. / European Journal of Medicinal Chemistry 41 (2006) 727–737
731
Table 1 (continued)
Compounds
Substituents
Antiproliferative activity (%) at 100 μM
Selected for
further studies
a
of tested compounds*
X; R
1
R
Breast cancer
MCF-7
ND
lung cancer
NCI-H460
ND
CNS tumor
SF-268
ND
7
7
g
R
1
= H
h
R
1
= Cl
ND
ND
ND
7
7
i
R
R
1
1
= Cl
= Cl
11
31
14
85
–12
No
j
6
Yes
a
Negative value of antiproliferative activity (%) means that the tested compound stimulated cell growth in comparison with control.
inhibitors (with IC50 < 50 uM) (Tables 3 and 4). It was re-
vealed, that the most promising among potential inhibitors
were compounds 7b, j (CDK2); 4f, j (CDK5); 7d, j (GSK-3);
planar. In contrast, for GSK-3 the two hydrogen bonds were
formed by the amino group at the position 6. The interaction
of compound 4j with CDK5 was found to have high binding
energy of 74.0 kJ mol (Table 4). Spatial models of the inhi-
bitor 4j were almost planar, and the benzoimidazole ring effi-
ciently interacted with the hydrophobic pocket, providing an
additional stability to the complex. Also, three hydrogen bonds
were formed in the 4j-CDK5 complex (Fig. 3).
–
1
4
h, 7i (EGFR) which formed high energy hydrogen bonds pre-
ferably with hinge region. Unfortunately, neither of the modern
docking programs is able to predict the real affinity of the com-
pounds with high percent of the hitting. However, the applica-
tion of the structure-based approach may be very useful for the
definition of the binding poses and further directed modifica-
tion of the active molecules.
The amino group interacted with Cys83 and Asp84 amino
residues forming two bonds. An additional hydrogen bond was
formed between Cys83 and the hydrogen atom of the second-
ary amine moiety at the position 5 of the pyrrolo[2,3-b]pyra-
zine cycle. The hydrophobic interaction was observed between
the chlorine atom and the phenyl ring of the Phe80 amino re-
sidue (bond length 3.82 Å). Such hydrophobic interaction was
observed for compounds that had a halogen atom in the m,p-
position of phenyl ring at position 5 of pyrrolo[2,3-b]pyrazine
cycle (compounds 4g, h, j; 7d, j). Addition of one carbon atom
at position 5 from benzyl to phenethyl moiety (Table 1, com-
pounds 4c, e; 7a, i) led to the steric clashes in the hydrophobic
pocket of kinases. Thus, the energy of the ligand conformation
was high and, as a result, the affinity of the ligand decreased.
These patterns correlated well with antiproliferative activities
of the tested compounds as discussed above (Tables 1 and 2).
Thus, our novel antiproliferative agents, pyrrolo[2,3-b]pyra-
zine derivatives 4a–j, 7a–j were docked in the binding sites of
kinases CDK2, CDK5, GSK-3 and EGFR. The above filtering
criteria were applied (Tables 3 and 4). The essential aloisine
hydrogen bonds in the binding sites were formed by the nitro-
gen atom at position N(4) and the hydrogen atom at N(5) (de-
scribed by Standard Kinase Interaction Pattern (Fig. 2)) [16].
This binding pattern summarizes the majority of known kinase
inhibitor binding modes due to the conservative interactions of
ATP-competitive inhibitors and common features of kinase
binding sites. In our docking studies the orientation of com-
pounds depended primarily on the substituents at the positions
2
,3,5,7 of the pyrrolo[2,3-b]pyrazine cycle. The presence of
nitrile groups and a substituent at the position 5 changed the
binding mode significantly in comparison with aloisine
The best binding interactions for 6-amino-5-(4-methoxyben-
zyl)-7-(4-chlorophenylsulfonyl)-5H-pyrrolo[3,2-b]pyrazine-
(
Fig. 2). This can be explained by the fact that the crucial hy-
drogen bonds for 6-aminopyrrolo[2,3-b]pyrazine-2,3-dicarbo-
nitriles 4a–j, 7a–j were formed by the amino group and the
oxygen atom of the sulfo group or by the heterocyclic nitrogen
atom at the position 7, due to conformational reasons. In addi-
tion, being smaller than compounds 4, 7, aloisine does not fill
the hydrophobic pocket of the kinase active site (Fig. 2).
2
,3-dicarbonitrile (7j) were formed in the complex with CDK2
(Fig. 4) and GSK-3. The oxygen atom of the sulfo group in
CDK2 formed a hydrogen bond with the H–N of the Leu83
residue (bond length 2.7 Å), while the oxygen atom of the
methoxy group bound to Asp145 (1.82 Å). The stacking inter-
action was observed between the phenyl ring of the benzyl
moiety and the phenyl ring of the Phe80 residue (bond length
3.56 Å). An additional hydrophobic interaction was found be-
tween the chlorine atom and the phenyl ring of the Phe82 ami-
no residue (3.65 Å).
Analysis of 6-amino-5-(3-chlorophenylamino)-7-(1-methyl-
1
H-benzo[d]imidazol-2-yl)-5H-pyrrolo[3,2-b]pyrazine-2,3-di-
carbonitrile (4j)—CDK2 complex showed that hydrogen bonds
were not formed, nor was the conformation of the inhibitor

732
G.G. Dubinina et al. / European Journal of Medicinal Chemistry 41 (2006) 727–737
Table 2
Antiproliferative activity of pyrrolo[2,3-b]pyrazines 4f–h, j; 7d, j against selected human cancer cell lines
a
Cell lines
Parameters
Concentration (μM)
4j
4
f
4g
4h
7d
7j
HOP-92 (non-small cell lung cancer)
GI50
TGI
LC50
GI50
TGI
LC50
GI50
TGI
LC50
GI50
TGI
LC50
GI50
TGI
LC50
GI50
TGI
LC50
GI50
TGI
LC50
GI50
TGI
LC50
GI50
TGI
LC50
GI50
TGI
LC50
GI50
TGI
LC50
GI50
TGI
LC50
GI50
TGI
LC50
GI50
TGI
0.71
3.47
28.5
1.01
6.17
> 50
7.22
20.3
> 50
0.66
4.24
28.0
1.86
5.06
49.1
0.31
7.58
50.0
2.68
20.7
> 50
1.66
8.64
49.2
1.40
4.38
36.7
0.70
1.53
33.8
1.98
4.46
> 50
9.56
24.3
> 50
10.2
34.7
> 50
0.70
1.41
28.5
1.63
18.4
>50
2.35
10.1
35.8
16.5
>50
>50
12.3
46.2
>50
11.5
>50
>50
ND
b
NCI-H322M (non-small cell lung cancer)
EKVX (non-small cell lung cancer)
UACC-257 (melanoma)
0.62
4.87
> 50
4.16
> 50
> 50
4.51
> 50
> 50
0.74
2.18
11.0
2.61
> 50
> 50
0.89
ND
> 50
0.84
2.60
20.7
0.92
3.59
> 50
0.83
2.43
10.5
1.70
4.86
> 50
2.79
18.2
> 50
7.56
22.5
> 50
1.35
3.05
12.6
ND
9.62
20.4
43.2
0.84
7.45
45.5
6.18
15.9
40.7
5.25
11.5
25.1
10.5
33.9
> 50
1.27
12.6
> 50
6.96
33.5
>50
8.91
37.3
> 50
5.44
15.0
41.2
0.014
3.54
34.0
0.082
1.96
26.0
0.84
14.1
> 50
6.95
16.9
40.9
9.99
28.9
> 50
19.1
> 50
> 50
33.1
> 50
> 50
0.93
4.17
19.5
18.3
> 50
> 50
9.65
42.3
> 50
12.0
>50
7.04
46.1
> 50
13.5
> 50
> 50
1.14
5.67
36.9
4.62
> 50
> 50
1.54
7.71
> 50
1.37
6.20
30.7
7.69
17.4
39.5
1.50
6.38
22.1
0.57
4.67
21.9
6.44
15.6
38.0
2.36
10.8
41.2
1.08
3.57
13.1
> 50
> 50
> 50
> 50
> 50
> 50
8.26
38.3
> 50
33.5
> 50
>50
21.6
> 50
> 50
2.90
19.4
> 50
3.33
> 50
> 50
6.22
>50
LOX IMVI (melanoma)
PC-3 (prostate cancer)
IGROV1 (ovarian cancer)
OVCAR-3 (ovarian cancer)
>50
786-0 (renal cancer)
7.08
15.6
34.3
4.23
13.0
35.1
2.45
13.9
> 50
10.8
41.7
> 50
0.42
10.3
26.6
3.60
10.8
25.6
ACHN (renal cancer)
RXF 393 (renal cancer)
BT-549 (breast cancer)
NCI/ADR-RES (breast cancer)
HCC-2998 (colon cancer)
a
>50
1.45
5.37
> 50
34.3
> 50
> 50
1.15
6.00
20.0
0.55
1.54
ND
LC50
GI50 – concentration of compound sufficient to reduce the number of viable cells by 50%; TGI – concentration of compound that completely suppressed cell
growth; LC50 – concentration of compound that caused death of 50% of cancer cells.
b
ND, not determined.
Table 3
Geometric filters and calculated energy scoring functions (QXP/FLO+) for training set of known inhibitors
Kinase
target
PDB code
Geometric filters
Calculated energy scoring functions*
(QXP/FLO+) for the training set
of known inhibitors
Source for
training set
Amino acid
residue
Phe80
Leu83
Phe80
Marked atom
Distance (Å) pI
Cntc
Hbnd
CDK2
CDK5
GSK-3
EGFR
1AQ1
1UNG
1UV5
1M17
C (meta-position of the phenyl ring)
O (C = O group)
C (meta-position)
N
C (methyl group)
N
C (methyl group)
N
5.5
2.5
5.5
2.5
5.5
2.5
5.5
2.5
3.9 ÷ 5.2
–59.0 ÷
–78.0
–46.1 ÷
–64.4
–54.5 ÷
–60.3
–40.0 ÷
–75.4
–2.1 ÷ –3.0
–0.1 ÷ –2.3
–3.1 ÷ –3.5
[5,13–15]
[4]
3.4 ÷ 4.9
4.4 ÷ 5.0
3.2 ÷ 5.4
Cys83
Leu132
Val135
Thr766
Met769
[4]
––3.1 ÷ –4.0 [5,13–15]
–1
*
pI, kJ mol (pI = –log10(Ki)); Cntc, kJ (ligand-site contact energy + site entropy); Hbnd, kJ (hydrogen bond energy).

G.G. Dubinina et al. / European Journal of Medicinal Chemistry 41 (2006) 727–737
733
Table 4
The energy scoring functions (QXP/FLO+) for compounds which passed through geometric filtration
o
Calculated energy scoring functions (QXP/FLO+)*
Compounds
CDK2
Cntc
–62.3
–
–
–
–
–
CDK5
Cntc
–63.0
–
–
GSK-3
EGFR
Cntc
–
–
pI
5.1
–
–
–
–
–
4.5
–
–
5.6
–
3.1
–
4.6
4.0
–
–
–
4.1
5.0
Hbnd
–0.1
–
–
–
pI
4.7
–
–
4.4
–
4.8
–
–
Hbnd
–4.1
–
–
–3.0
–
–9.2
–
–
–
pI
5.3
–
–
–
–
–
–
–
–
5.7
–
–
–
4.4
–
4.7
–
–
–
Cntc
–64.4
–
–
–
–
–
–
Hbnd
–2.3
–
–
–
–
–
–
pI
–
–
4.2
–
3.8
–
–
4.7
–
–
3.9
Hbnd
–
–
4
4
4
4
4
4
4
4
4
4
7
7
7
7
7
7
7
7
7
7
a
b
c
d
e
f
g
h
i
–70.2
–69.8
–1.1
–70.4
–75.1
–
–
–
–
–
–1.7
–
–
–
–
–71.9
–0.1
–
–
–0.2
–
–5.9
–
0.0
–0.1
–
–
–
–
–
–
–
–
–
–
–
–
–64.4
–2.1
–
–
–
–
–
j
–60.5
–67.6
5.4
4.4
–
–
3.8
–
–
4.3
–
–
–
–74.0
–78.4
–4.6
–3.8
–
–65.5
–0.8
–
–
a
b
c
d
e
f
g
h
i
–
–
–
–
–
–73.7
–62.7
–0.2
–1.1
–
–
–
–
4.0
–
–
4.0
–
–
–71.3
–75.2
–69.6
–72.8
–3.7
–
–
–2.3
–
–
–73.1
–2.9
–
–2.5
–
–
–
–
–
–
–
–
–
–
–
–
–
–82.3
–63.3
–62.8
–0.1
–1.8
–
–
–
–
–
–
–
–
–
–
–65.1
–84.6
–0.1
–3.1
3.6
–
j
–
5.1
–76.3
–2.9
–
–
*
“–” indicates that compound is discarded as non-active after geometric filtration.
Fig. 2. Schematic representation of the binding mode of the aloisine B (A) and compound 4j (B) by the Standard Kinase Interaction Pattern [16].
6
. Conclusions
differences in the binding pose of these compounds in the ki-
nase ATP-binding site in comparison with aloisines (Figs. 2–
4
). Introducing the anilino moiety into the position 5 of pyrrolo
Altogether, our data demonstrate that 5,7-disubstituted 6-
[2,3-b]pyrazine cycle (compound 4j; Tables 1 and 2 and Figs. 2
aminopyrrolo[2,3-b]pyrazine-2,3-dicarbonitriles possess the
antiproliferative properties for human cancer cell lines (Tables 1
and 2) and can be considered as potential protein kinase inhi-
bitors. The most promising targets for this class of compounds
were CDK5 and GSK-3 based on higher scoring functions and
geometric criteria for the calculated complexes of pyrrolo[2,3-
b]pyrazine derivatives with these targets including the number
of the formed H-bonds (Tables 3 and 4). Presumably, this ob-
servation can be related to the fact that the binding sites of
CDK5 and GSK-3 are larger and more flexible compared to
the sites of other kinases. It was found, that the most potential
promising inhibitors for each target were compounds 7b, j
and 3) by reaction with 3-chlorophenylhydrazine increased the
antiproliferative activity. The increased energy of binding and
formation of an additional hydrogen bond were observed for
GSK-3 and CDK-5. This fact and structure—antiproliferative
activity analysis of other substituents at the 5 and 7 position of
the pyrrolo[2,3-b]pyrazine cycle correlates well with the calcu-
lated binding poses as discussed above and opens new possi-
bilities for further purposeful modification of substituents at 5
position of 6-aminopyrrolo[2,3-b]pyrazine-2,3-dicarbonitriles.
High selectivity of interaction of the compounds 4f–h, j; 7d,
j with their particular targets may be a reason for the dramatic
difference between GI50 and LC50 (Table 2, compound 4j).
Although more studies are needed to evaluate the potency of
(
CDK2); 4f, j (CDK5); 7d, j (GSK-3); 4h, 7i (EGFR). Mole-
cular modeling of the inhibitor–enzyme complexes showed the

734
G.G. Dubinina et al. / European Journal of Medicinal Chemistry 41 (2006) 727–737
culture, after a 48 hour delay without any exposure. The re-
sponse parameters GI , TGI, LC are interpolated values re-
5
0
50
presenting the concentrations at which PG was +50, 0 and –50,
respectively.
Docking procedure. All docking procedures were per-
formed with QXP/Flo+ program developed by McMartin and
Bohacek [17]. Docking with flexible ligand and rigid protein
site using systematic docking algorithm (SDOCK+) was em-
ployed, as it yielded good correlation of the energy parameters
and X-ray RMSD on the training set of compounds. The max-
imal number of search iterations (search step number) was 250
steps. The best (five) complexes (saved docking poses) were
used for final post-docking filtration for each compound. In-
trinsic QXP scoring functions (pI, Cntc, hbnd) and two dis-
tances from essential atoms (in-house software package, Mul-
ti-Filter) were applied for energetic and geometric filtration,
respectively. The minimal geometric criteria for the post-dock-
ing selection based on calculated parameters of the training sets
are shown in the Table 3. Compounds which did not satisfy the
requirements of the geometric filters were discarded as non-ac-
tive (Table 4). The training sets of known inhibitors were
scored and ranked after the docking procedure in the respective
kinases and the range of the scoring functions was obtained
and applied for the new inhibitor selection. There are intrinsic
scoring functions (QXP/FLO+) for the compounds which have
passed successfully through geometric filters, shown in Table 4.
Fig. 3. The CDK5-4 j complex (QXP/FLO+ docking software).
Identification of structures and purity tests. Structure and
purity of synthesized compounds were verified by IR spectra
recorded on UR-20, Specord 75-I and Pye Unicam, LC/MS
(
Agilent LC/MSD SL and VL mass spectrometers, column
1
13
ZORBAX SB – C18 4.6 × 15 mm). H NMR and C spectra
were obtained on a Varian 300 MHz spectrometer and Mercury
(Varian) 400 MHz, Brooker 500 MHz, respectively. Chemical
shifts given in this work are reported in ppm downfield from
tetramethylsilane.
Synthesis of 5,7-disubsituted 6-aminopyrrolo[2,3-b]pyra-
zine-2,3-dicarbonitriles 4a–j; 7a–j.
Fig. 4. The CDK2-7 j complex (QXP/FLO+ docking software).
Compounds 4a–j were synthesized similarly to the method
7].
these compounds to non-malignant tissues, the 5,7-disubsti-
tuted 6-aminopyrrolo[2,3-b]pyrazine-2,3-dicarbonitriles are
promising as cytostatic agents with little or no general toxicity.
[
Briefly, 1.99 g (10 mmol) of compound 1 and 10 mmol of
α-azaheteroarylacetonitriles 2a–e were dissolved in 10 ml of
DMF. In case of compounds 2d, e the 1.01 g (10 mmol) of
triethylamine was added to the reaction mixture after dissol-
ving in DMF. After 3–4 hours of stirring at 30–40 °C, the
reaction mixture was kept overnight at room temperature. The
obtained precipitate was filtered, washed with little amount of
DMF and acidified water, dried and recrystallized from DMF
with yields 75% (3a); 91% (3b); 87% (3c); 95% (3d) and 88%
(3e), respectively. Then, 5 mmol of either compound 3a–e was
dissolved in 5 ml DMF and 10 mmol of the appropriate amine
or 3-chlorophenylhydrazine (for compound 4j) was added by
drops under stirring at room temperature. The formation of yel-
low precipitate was observed. The reaction mixture was kept at
50–60 °C for 2 hours, then cooled, filtered the precipitate,
washed with DMF and acidified water to remove the amine
7
. Experimental protocols
Antiproliferative activity assay was carried out jointly
with the US National Cancer Institute according to Develop-
mental Therapeutics Program. Sixty human cancer cell lines
were used to determine the in vitro anticancer activity of the
tested compounds. The anticancer assay was based on the pro-
tein staining using sulforhodamine B (SRB). The calculated
measure of efficiency was percentage growth (PG). The effect
of the compound on the cell line was calculated from the
change in the optical density values of SRB-derived coloration
observed before the exposure of the cells to the tested com-
pounds and 48 hours after exposure, and, for the baseline cell

G.G. Dubinina et al. / European Journal of Medicinal Chemistry 41 (2006) 727–737
735
+
hydrochloride, dried, and recrystallized from DMF. The yields
of products 4a–j were 84–94%.
p. >300 °C; LC/MS: tR 1.143 min (99.4%), m/z (M ) 438.9;
−
1
IR (KBr): ν
3500–3200; νCN 2225; νC=N 1645 (cm ).
NH2
1
1
.99 g (10 mmol) of compound 1 and 10 mmol of 2-(phe-
nylsulfonyl)acetonitrile 5a, b were dissolved in 6 ml of DMF.
.02 g (20 mmol) of triethylamine was drip-added to the reac-
H NMR (300 MHz, δ, ppm, DMSO-d , J (Hz)): 8.98 (s,
2H, NH ), 7.65 (d, 1H, Het(4)-H, J = 7.2), 7.62 (d, 1H, Het
(7)-H, J = 6.6), 7.43 (d, 2H, Ar(3,5)-H, J = 8.4), 7.32 (m, 4H,
6
2
2
tion mixture at 5 °C under stirring and cooling. After 1 hour of
stirring at room temperature, the reaction mixture was mixed
with 100 ml of cold water and acidified by acetic acid to neu-
tral pH. The obtained precipitate was filtered, dried and recrys-
tallized from dioxane with yields 81% (6a) and 87% (6b).
Then, 5 mmol of either compound 6a or 6b was dissolved in
Ar(2,6)H + Het(5,6)-H), 5.56 (s, 2H, CH ), 4.01 (3H, c, CH ).
2
3
1
3
C NMR (300 MHz, δ ppm, DMSO-d ): 160.1 (C );
6
4a
145.7, 143.8, 141.6 (C , C , CHet(2)); 139.0 (CAr(4)); 136.7,
6
7a
1
34.2 (CHet(3a), CHet(7a)); 132.2, 129.4 (C , C ); 116.8, 116.2
2 3
(2C, CN); 122.8, 122.3, 119.9, 118.7, 115.3, 112.4, 108.8
(CHet(4), CHet(5), CHet(6), CHet(7), CAr(1), CAr(2), CAr(3), CAr(5)
CAr(6)); 84.7 (C ); 41.8 (C, CH ); 27.1 (C, NCH ).
,
3
ml DMF and 5 mmol of the appropriate amine was added by
7
2
3
drops under stirring at room temperature. The formation of pre-
cipitate was observed. The reaction mixture was kept at
6
-Amino-5-(benzo[d] [1,3]dioxol-5-ylmethyl)-7-(1-methyl-
H-benzo[d]imidazol-2-yl)-5H-pyrrolo[3,2-b]pyrazine-2,3-di-
carbonitrile (4i): m.p. > 300 °C; LC/MS: t 1.218 min (96.8%),
1
5
0–60 °C for 2 hours, then cooled, filtered to separate the pre-
R
cipitate, washed with DMF and water, dried, and recrystallized
from dioxane. The yields of products 7a–j were 72–89%.
+
m/z (M ) 448.4; IR (KBr): ν
647 (cm ).
3500–3200; νCN 2224; νC=N
NH2
−1
1
Compounds 4a–e were synthesized as described earlier in
1
H NMR (300 MHz, δ, ppm, DMSO-d , J (Hz)): 8.96 (s,
6
[
7].
2
H, NH ), 7.68 (d, 1H, Het(4)-H, J = 6.6), 7.62 (d, 1H, Het(7)-
2
6
-Amino-5-(2-diethylamino)ethyl)-7-(1-methyl-1H-benzo
H, J = 6.3), 7.31 (m, 2H, Het(5,6)-H), 6.93 (m, 2H, Ar(2,6)-H),
.87 (d, 1H, Ar(5)-H, J = 8.4), 6.04 (s, 2H, CH ), 5.49 (s, 2H,
[
(
4
d]imidazol-2-yl)-5H-pyrrolo[3,2-b]pyrazine-2,3-dicarbonitrile
4f): m.p. > 300 °C; LC/MS: t 1.297 min (98.2%), m/z (M )
6
2
+
R
CH ), 4.03 (s, 3H, CH ).
2
13
3
13.6; IR (KBr): ν
3500–3200; νCN 2218; νC=N 1638
NH2
C NMR (300 MHz, δ ppm, DMSO-d ): 158.4 (C );
6
4a
−
1
(
cm ).
1
48.0, 147.3 (CAr(3), CAr(4)); 142.7, 140.6, 139.5 (C , C
6 7a,
1
H NMR (300 MHz, δ, ppm, DMSO-d , J (Hz)): 9.27 (s,
CHet(2)); 136.1, 129.3 (CHet(3a), CHet(7a)); 126.3, 118.0 (C2,
C ); 116.9, 116.3 (2C, CN); 101.6 (C, OCH O); 122.3,
6
2
H, NH ), 7.68 (d, 1H, Het(4)-H, J = 6.9*), 7.58 (d, 1H, Het
2
3
2
(
7)-H, J = 6.9), 7.27 (m, 2H, Het(5,6)-H), 4.37 (t, 2H, CH , J
2
121.2, 118.4, 110.4, 108.8, 108.4 (CAr(1), CAr(2), CAr(5)
,
=
5.4), 4.00 (s, 3H, CH ), 2.77 (t, 2H, CH J = 5.4), 2.52 (q,
3 2
CAr(6), CHet(4), CHet(5), CHet(6), CHet(7)); 84.5 (C ); 43.8 (C,
7
4
H, CH ), 0.87 (t, 6H, CH , J = 6.6).
2 3
CH
); 32.7 (C, NCH
).
3
2
*
Het(4,5,6,7)—numbering of the benzimidazole ring.
6
-Amino-5-(3-chlorophenylamino)-7-(1-methyl-1H-benzo[d]
1
3
C NMR (300 MHz, δ, ppm, DMSO-d ): 156.9 (C );
imidazol-2-yl)-5H-pyrrolo[3,2-b]pyrazine-2,3-dicarbonitrile (4j):
m.p. > 300 °C; LC/MS: t 1.364 min (95.4%), m/z (M ) 439.9;
6
4a
+
1
41.3, 140.2, 138.7 (C ,C CHet(2)); 138.1, 136.3 (CHet(3a)
CHet(7a)); 130.3, 129.0 (C , C ); 116.8, 116.2 (2C, CN);
20.8, 119.1, 118.7, 112.1 (CHet(4), CHet(5), CHet(6), CHet(7));
8.6 (C ); 57.3, 42.2, 46.9 (4C, CH ); 28.9 (C, NCH ); 11.8
,
6
7a,
R
−1
IR (KBr): ν
3500–3200; νCN 2221; νC=N 1628 (cm ).
2
3
NH2,NH
1
1
8
H NMR (300 MHz, δ, ppm, DMSO-d , J (Hz)): 9.83 (s,
6
7
2
3
1H, NH); 9.01 (s, 2H, NH ), 7.67 (d, 1H, Het(4)-H, J = 6.9),
2
(
2C, CH3).
7.62 (d, 1H, Het(7)-H, J = 7.2), 7.29 (m, 3H, Het(5,6) + Ar(5)-
H), 6.99 (d, 1H, Ar(4)-H, J = 8.4), 6.81 (s, 1H, Ar(2)-H), 6.68
5
-(4-Fluorobenzyl)-6-amino-7-(1-methyl-1H-benzo[d]imi-
dazol-2-yl)-5H-pyrrolo[3,2-b]pyrazine-2,3-dicarbonitrile (4g):
m.p. > 300 °C; LC/MS: tR 1.102 min (97.6%), m/z (M )
(d, 1H, Ar(6)-H, J = 8.4), 4.08 (s, 3H, CH ).
3
+
1
3
C NMR (300 MHz, δ, ppm, DMSO-d
159.1 (CAr(1)); 147.9 (CAr(3)); 142.8, 139.6, 138.7 (C , C ,
CHet(2)); 136.2, 134.4 (CHet(3a), CHet(7a)); 127.1, 117.9 (C2,
): 162.8 (C4a);
6
4
23.5; IR (KBr): ν
3500–3200; νCN 2222; νC=N 1643
NH2
6
7a
−
1
(
cm ).
1
H NMR (300 MHz, δ, ppm, DMSO-d , J (Hz), 8.96 (s, 2H,
C ); 116.6, 116.1 (2C, CN); 131.4, 122.4, 121.0, 118.5,
3
6
NH ), 7.66 (d, 1H, Het (4)-H, J = 7.2), 7.60 (d, 1H, Het (7)-H
J = 7.5), 7.37 (dd, 2H, Het(5,6)-H, J = 4.8, J = 7.8), 7.22 (m,
112.8, 111.9, 110.5 (CAr(2), CAr(4), CAr(5), CAr(6), CHet(4)
CHet(5), CHet(6), CHet(7)); 82.8 (C ); 36.2 (C, NCH ).
,
2
7
3
4
H, Ar(2,3,5,6)-H*), 5.56 (s, 2H, CH ), 4.03 (s, 3H, CH ).
5-Chloro-6-[cyano(phenylsulfonyl)methyl]-2,3-pyrazinedi-
2
3
*
Ar(2,3,4,5,6)—numbering of the phenyl substituent at po-
carbonitrile N-triethylammonium salt (6a): m.p. 158–159 °C;
LC/MS: tR 0.973 min (100.0%), m/z (M ) 343.6; IR (KBr):
+
sition 5.
−
1
1
3
C NMR (300 MHz, δ ppm, DMSO-d ): 159.2 (C );
ν
CN 2195, 2227; νC=N 1645; νSO2 1290(as), 1135(s) (cm ).
6
4a
1
1
1
(
(
44.8, 142.9, 140.7 (C , C
33.3 (CHet(3a), CHet(7a)); 131.3, 128.5 (C , C ); 117.0, 116.6
2C, CN); 122.3, 121.2, 119.1, 118.0, 112.1, 110.4, 108.8
CHet(4), CHet(5), CHet(6), CHet(7), CAr(1), CAr(2), CAr(3), CAr(5)
, C ); 137.3 (CAr(4)); 135.8,
H NMR (300 MHz, δ, ppm, DMSO-d , J (Hz)): 10.74
(sbroad, 1H, N H), 7.86 (m, 2H, Ar(2′,6′)-H*), 7.55 (m, 3H,
Ar(3′,4′,5′)-H), 3.09 (m, 6H, CH ), 1.17 (t, 9H, CH , J = 9).
6
Het(2)
7a
6
+
2
3
2
3
1
3
,
C NMR (300 MHz, δ, ppm, DMSO-d ): 151.9 (C ); 143.9
6
2
CAr(6)); 86.8 (C ); 41.3 (C, CH ); 29.8(C, NCH ).
7
2
3
(C
3
); 139.4 (C
6
); 132.5(CAr(4′)); 129.7 (CAr(1′)); 128.7, 127.9
); 118.5 (C ); 116.0, 115.3,
5-(4-Chlorobenzyl)-6-amino-7-(1-methyl-1H-benzo[d]imida-
(CAr(2′), CAr(3′), CAr(5′), C
114.1 (3C, CN); 79.6 (C ); 46.3 (3C, CH ); 9.1 (3C, CH ).
Ar(6′)
5
-
zol-2-yl)-5H-pyrrolo[3,2-b]pyrazine-2,3-dicarbonitrile (4h): m.
C
2
3

736
G.G. Dubinina et al. / European Journal of Medicinal Chemistry 41 (2006) 727–737
*
Ar(2′,3′,4′,5′,6′)—numbering of the phenylsulfonyl substi-
tuent at position 7.
-Chloro-6-[[(4-chlorophenyl)sulfonyl](cyano)methyl]-2,3-
pyrazinedicarbonitrile N-triethylammonium salt (6b): m.p.
CAr(3′), CAr(4′), CAr(5′), CAr(6′)); 116.1, 115.8 (2C, CN); 87.4
(C ); 43.8 (C, CH ); 21.5 (C, CH ).
7
2
3
5
6-Amino-5-(4-chlorobenzyl)-7-(phenylsulfonyl)-5H-pyrrolo
[3,2-b]pyrazine-2,3-dicarbonitrile (7d): m.p. > 300 °C; LC/MS:
+
+
1
59–160 °C; LC/MS: t 1.054 min (98.4%), m/z (M ) 379.6;
t
R
1.186 min (97.7%), m/z (M ) 449.9; IR (KBr): νNH2 3392(as),
R
−1
IR (KBr): ν 2185, 2225; νC=N 1640; νSO2 1289(as), 1138(s)
3308(s); νCN 2218; νC=N 1633; νSO2 1308(as), 1146(s) (cm ).
CN
−
1
1
(
cm ).
H NMR (300 MHz, δ, ppm, DMSO-d
, J (Hz)): 8.56 (s,
6
1
2
H, NH ), 8.18 (d, 2H, Ar(2,6)-H, J = 6.6), 7.45 (m, 3H, Ar
H NMR (300 MHz, δ, ppm, DMSO-d , J (Hz)): 11.20
2
6
+
(
(
3,4,5)-H), 7.17 (d, 2H, Ar(3′,5′)-H, J = 7.9), 7.06 (d, 2H, Ar
2′,6′)-H, J = 7.9), 5.29 (s, 2H, CH ).
(
sbroad, 1H, N H), 7.85 (d, 2H, Ph(2′,6′)-H, J = 9), 7.61 (d,
H, Ar(3′,5′)-H, J = 9), 3.10 (m, 6H, CH ), 1.18 (t, 9H, CH ,
2
2
2
3
1
3
J = 9).
C NMR (300 MHz, δ, ppm, DMSO-d
): 158.4 (C4a);
6
1
3
1
1
1
42.7, 140.6, 139.5 (C , C CAr(1′)); 136.1 (CAr(4)); 129.3,
C NMR (300 MHz, δ, ppm, DMSO-d ): 151.9 (C ); 142.
6
7a,
6
2
26.3 (C , C ); 116.9, 116.3 (2C, CN); 148.0, 147.3, 122.3,
8
(C ); 139.6 (C ); 137.3(CAr(4′)); 129.7 (CAr(1′)); 129.8, 128.8
2
3
3
6
21.2, 118.4, 118.0, 110.4, 108.8, 108.4 (CAr(1), CAr(2)
,
(
C2Ar′, C3Ar′, C5Ar′, C6Ar′); 118.3(C ); 116.4, 115.2, 114.1 (3C
5
CN); 79.0 (C ); 46.3 (3C, CH ); 9.1(3C, CH ).
CAr(3), CAr(5), CAr(6), CAr(2′), CAr(3′), CAr(4′), CAr(5′), CAr(6′));
C
2
3
8
4.5 (C ); 43.8 (C, CH ).
7 2
6-Amino-5-phenethyl-7-(phenylsulfonyl)-5H-pyrrolo[3,2-b]
6-Amino-5-(4-methoxybenzyl)-7-(phenylsulfonyl)-5H-pyr-
pyrazine-2,3-dicarbonitrile (7a): m.p. > 300 °C; LC/MS: tR
.121 min (99.8%), m/z (M ) 429.6; IR (KBr): ν
+
rolo[3,2-b]pyrazine-2,3-dicarbonitrile (7e): m.p. > 300 °C; LC/
1
3
3398(as),
NH2
+
−
1
MS: t 1.256 min (95.9%), m/z (M ) 445.7; IR (KBr): ν
R
NH2
308(s); ν 2210; νC=N 1648; νSO2 1297(as), 1150(s) (cm ).
CN
1
3400(as), 3315(s); νCN 2221; νC=N 1628; νSO2 1270(as), 1080
H NMR (300 MHz, δ, ppm, DMSO-d , J (Hz)): 8.69 (s,
6
−1
(
s) (cm ).
2
H, NH ), 8.02 (d, 2H, Ar(2,6)-H, J = 7.8), 7.67 (m, 3H, Ar
3,4,5)-H), 7.07 (m, 5H, Ar-H), 4.41 (t, 2H, CH , J = 7.5), 2.93
2
1
H NMR (300 MHz, δ, ppm, DMSO-d , J (Hz)): 8.75 (s,
6
(
2
2
H, NH ), 8.12 (d, 2H, Ar(2,6)-H, J = 6.3), 7.72 (m, 3H, Ar
2
(
t, 2H, CH , J = 7.5).
2
(
(
3,4,5)-H), 7.23 (d, 2H, Ar(3′,5′)-H, J = 8.4), 6.97 (d, 2H, Ar
2′,6′)-H, J = 8.4), 5.43 (s, 2H, CH ), 3.78 (s, 3H, OCH ).
1
3
C NMR (300 MHz, δ, ppm, DMSO-d ): 155.4 (C );
6
4a
2
3
1
1
1
43.3, 141.6, 140.5 (C , C , C
16.1 (2C, CN); 133.6, 130.7, 126.2, 122.3, 121.2, 118.4,
18.0, 110.4, 108.8 (CAr(1), CAr(2), CAr(3), CAr(4), CAr(5)
CAr(6), CAr(2′), CAr(3′), CAr(4′), CAr(5′),CAr(6′)), 89.5 (C ); 47.2
2C, CH2).
); 126.7, 121.3 (C , C );
6
7a
Ar(1′) 2 3
13
C NMR (300 MHz, δ, ppm, DMSO-d ): 160.0 (C4Ar);
6
1
57.5 (C ); 142.2, 140.6, 140,1 (C , C CAr(1′)); 126.5, 121.6
4a 6 7a,
,
(
C , C ); 133.9, 132.8, 127.9, 127.7, 125.2, 121.1, 120.7,
2 3
7
120.2, 108.0, 107.3 (CAr(1), CAr(2), CAr(3), CAr(5), CAr(6),
(
CAr(2′), CAr(3′), CAr(4′), CAr(5′), CAr(6′)); 116.2, 115.7 (2C, CN);
5.5 (C ); 64.8 (C, OCH ); 42.2 (C, CH ).
6
-Amino-5-(2-diethylamino)ethyl)-7-(phenylsulfonyl)-5H-
8
7
3
2
pyrrolo[3,2-b]pyrazine-2,3-dicarbonitrile (7b): m.p. > 300 °C;
LC/MS: t 1.176 min (98.7%), m/z (M ) 424.5; IR (KBr): ν
404(as), 3323(s); ν 2224; νC=N 1642; νSO2 1303(as), 1140
6-Amino-5-(benzo[d] [1,3]dioxol-5-ylmethyl)-7-(phenylsul-
+
R
NH2
fonyl)-5H-pyrrolo[3,2-b]pyrazine-2,3-dicarbonitrile (7f): m.p.
3
(
+
CN
>
300 °C; LC/MS: t 1.198 min (97.6%), m/z (M ) 459.5; IR
R
−1
s) (cm ).
(
1
KBr): νNH2 3379(as), 3304(s); ν
2230; νC=N 1649; νSO2
CN
1
H NMR (300 MHz, δ, ppm, DMSO-d , J (Hz)): 8.90 (s,
−1
6
303(as), 1141(s) (cm ).
2
H, NH ), 8.04 (d, 2H, Ar(2′,6′)-H, J = 2.1), 7.65 (m, 3H, Ar
3′,4′,5′)-H), 4.25 (t, 2H, CH , J = 4.8), 2.66 (t, 2H, CH , J
2 2
1
2
H NMR (300 MHz, δ, ppm, DMSO-d , J (Hz)): 8.66 (s,
6
(
=
2
H, NH ), 8.06 (d, 2H, Ar(2,6)-H, J = 6,3), 7.65 (m, 3H, Ar
2
4.8), 2.41 (qr, 4H, CH , J = 6.9), 0.74 (t, 6H, CH , J = 6.9).
2 3
(3,4,5)-H), 6.86 (m, 2H, Ar(5′,6′)-H), 6.76 (d, 1H, Ar(2′)-H, J
1
3
C NMR (300 MHz, δ, ppm, DMSO-d ): 156.9(C ),
6
4a
= 8.1), 5.99 (s, 2H, CH ), 5.33 (s, 2H, CH ).
2
2
1
1
43.1, 141.0, 140.0 (C , C CAr(1′)); 125.9, 120.5 (C , C );
15.5 (2C, CN); 133.2, 129.3, 125.8 (CAr(2′), CAr(3′), CAr(4′)
CAr(5′), CAr(6′)), 90.4 (C ), 50.43(2C, NCH CH ), 46.3 (2C,
NCH CH N), 11.5 (2C, CH ).
13
6
7a,
2
3
C NMR (300 MHz, δ, ppm, DMSO-d ): 156.2 (C );
6
4a
,
1
47.9, 147.3 (CAr(3), CAr(4)); 143.4, 140.8, 140.5 (C , , C
6 7a,
7
2
3
CAr(1′)); 126.9, 121.5 (C , C ); 134.5 (CAr(1)); 133.8, 129.9,
2
3
2
2
3
1
1
26.4 (CAr(2′), CAr(3′), CAr(4′), CAr(5′), CAr(6′)); 121.2 (CAr(6));
16.1 (2C, CN); 108.8, 108.3 (CAr(2), CAr(5)); 101.6 (C,
6-Amino-5-(4-methybenzyl)-7-(phenylsulfonyl)-5H-pyrrolo
[
3,2-b]pyrazine-2,3-dicarbonitrile (7c): m.p. > 300 °C; LC/MS:
t 1.169 min (96.3%), m/z (M ) 429.5; IR (KBr): ν
OCH O); 90.9 (C ); 43.8 (C, NCH Ph).
2
7
2
+
3398(as),
R
NH2
6-Amino-5-(4-ethoxyphenyl)-7-(phenylsulfonyl)-5H-pyrro-
−
1
3
324(s); ν 2227; νC=N 1643; νSO2 1285(as), 1135(s) (cm ).
CN
lo[3,2-b]pyrazine-2,3-dicarbonitrile (7g): m.p. > 300 °C; LC/
1
+
H NMR (300 MHz, δ, ppm, DMSO-d , J (Hz)): 8.73 (s,
MS: t 1.168 min (96.2%), m/z (M ) 445.6; IR (KBr): ν
6
R
NH2
2
H, NH ), 8.08 (d, 2H, Ar(2,6)-H, J = 6.6), 7.33 (m,3H, Ar
3392(as), 3308(s); ν 2218; νC=N 1633; νSO2 1308(as), 1146
CN
2
−1
(
(
3,4,5)-H), 7.13 (d,2H,Ar(3′,5′)-H, J = 8.1), 7.11 (d, 2H, Ar
2′,6′)-H, J = 8.1), 5.39 (s, 2H, CH ), 2.52 (s, 3H, CH ).
(s) (cm ).
1
2
3
H NMR (300 MHz, δ, ppm, DMSO-d , J (Hz)): 8.21 (s,
6
1
3
′
C NMR (300 MHz, δ, ppm, DMSO-d ): 157.3 (C );
2H, NH ), 8.06 (dd, 2H, Ar (2,6)-H, J = 7.8, J = 1.5), 7.68 (m,
3H, Ar (3,4,5)-H), 7.43(d, 2H, Ar(3.5)-H, J = 9), 7.17(d, 2H,
6
4a
2
′
1
1
1
43.1, 140.7, 140,1 (C , , C CAr(1′)); 126.9, 121,5 (C , C );
6
7a,
2
3
34.5, 133.8, 129.9, 128.7, 126.4, 122.6, 120.0, 120.2 108.8,
08.3 (CAr(1), CAr(2), CAr(3), CAr(4), CAr(5), CAr(6), CAr(2′)
Ar(2,6)-H, J = 9), 4.12 (qr, 2H, CH , J = 6.9), 1.39 (t, 3H,
2
,
CH , J = 6.9).
3

G.G. Dubinina et al. / European Journal of Medicinal Chemistry 41 (2006) 727–737
737
1
3
C NMR (300 MHz, δ, ppm, DMSO-d ): 160.3 (C,
2H, Ar(2′,6′)-H, J = 8.7), 5.38 (s, 2H, CH ), 3.74 (s, 3H,
6
2
COC H ); 158.4 (C ); 143.4, 142.2, 140.7 (C , C CAr(1′));
OCH ).
2
5
4a
6
7a,
3
1
3
1
33.8 (CAr(1)), 122.8, 121.4 (C , C ); 116.0 (2C, CN); 130.6,
C NMR (300 MHz, δ, ppm, DMSO-d ): 159.3 (CAr(4));
6
2
3
1
29.9, 126.6, 116.1 (CAr(2), CAr(3), CAr(5), CAr(6), CAr(2′)
,
154.7 (C ); 142.6, 140.2, 139.6 (C , C
CAr(1′)); 136.4
7a,
4a
6
CAr(3′), CAr(4′), CAr(5′), CAr(6′)); 90.7 (C ); 64.0 (C, OCH );
(CAr(4′)); 126.2, 121.2 (C , C ); 133.7, 132.3, 127.0, 125.8,
2 3
7
2
1
5.1 (C, CH3).
-Amino-5-(3,4-dimethoxyphenethyl)-7-(4-chlorophenyl-
121.6, 120.8, 108.0, 107.3 (CAr(1), CAr(2), CAr(3), CAr(5),
6
CAr(6), CAr(2′), CAr(3′), CAr(5′), CAr(6′)); 116.1, 115.9 (2C, CN);
88.3 (C ); 66.1 (C, OCH ); 48.6 (C, CH ).
sulfonyl)-5H-pyrrolo[3,2-b]pyrazine-2,3-dicarbonitrile (7h): m.
p. > 300 °C; LC/MS: t 1.247 min (98.6%), m/z (M ) 524.0;
IR (KBr): νNH2 3403(as), 3319(s); ν 2215; νC=N 1638; νSO2
7
3
2
+
R
All synthesized compounds are commercially available
from Enamine Ltd. DataBase (www.enamine.net).
CN
−
1
1
305(as), 1135(s) (cm ).
1
H NMR (300 MHz, δ, ppm, DMSO-d , J (Hz)): 8.68 (s,
6
Acknowledgments
2
H, NH ), 8.02 (d, 2H, Ar(2,6)-H), J = 9), 7.76 (d, 2H, Ar
2
(
3,5)-H), J = 8.7), 6.66 (s, 1H, Ar(2′)-H), 6.57 (d, 1H, Ar(5′)-
The authors are grateful to the colleagues at National Can-
cer Institute (USA) who participated in Developmental Thera-
peutics Program, to Enamine Ltd. (www.enamine.net) for fi-
nancial support, and to Dr. A. Shtil for critical review of the
manuscript.
H, J = 6), 6.52 (d, 1H, Ar(6′)-H, J = 6), 4.40 (t, 2H, CH , J
=
3
2
6.8), 3.87 (t, 2H, CH , J = 6.8) 3.66 (s, 3H, OCH ), 3.59 (s,
2 3
H, OCH ).
3
1
3
C NMR (300 MHz, δ, ppm, DMSO-d ): 157.4(C );
6
4a
1
46.8, 145.3 (CAr(3), CAr(4)); 143.1, 141.0, 140.7 (C , C7a,
6
CAr(1′)); 136.5 (CAr(4′)); 126.6, 121.2 (C , C ); 116.4, 116.1
2
3
References
(
1
2C, CN); 133.6, 130.4, 126.7, 124.3, 121.2, 118.0, 111.1,
08.3 (CAr(1), CAr(2), CAr(5), CAr(6), CAr(2′), CAr(3′), CAr(5′)
CAr(6′)); 82.2 (C ); 65.8, 58.3 (2C, OCH ); 44.2, 38.7 (2C,
,
[
1] C. Sablayrolles, G.H. Cros, J.C. Milhavet, E. Rechenq, J.P. Chapat, M.
Boucard, J.J. Serrano, J.H. McNeill, J. Med. Chem. 27 (1984) 206–212.
2] A. Trejo, H. Arzeno, M. Browner, S. Chanda, S. Cheng, D.D. Comer,
et al., J. Med. Chem. 46 (2003) 4702–4713.
7
3
CH2).
-Amino-5-phenethyl-7-(4-chlorophenylsulfonyl)-5H-pyrro-
[
6
lo[3,2-b]pyrazine-2,3-dicarbonitrile (7i): m.p. > 300 °C; LC/
[3] P. Ge, T. Kalman, Bioorg. Med. Chem. Lett. 23 (1997) 3023–3026.
[4] Y. Mettey, M. Gompel, V. Thomas, M. Garnier, M. Leost, I. Ceballos-
Picot, et al., J. Med. Chem. 46 (2003) 222–236.
+
MS: t 1.187 min (97.9%), m/z (M ) 463.2; IR (KBr): ν
R
NH2
3
396(as), 3322(s); ν 2226; νC=N 1643; νSO2 1308(as), 1141
CN
−
1
[5] L. Meijer, M. Flajolet, Trends Pharmacol. Sci. 25 (2004) 471–480.
(
s) (cm ).
[
6] M. Mapelli, L. Massimilinao, C. Crovace, M.A. Seeliger, L.-H. Tsai, L.
1
H NMR (300 MHz, δ, ppm, DMSO-d , J (Hz)): 8.72 (s,
6
Meijer, A. Musacchio, J. Med. Chem. 48 (2005) 671–679.
2
H, NH ), 8.03 (d, 2H, Ar(2,6)-H, J = 8.7), 7.78 (d, 2H, Ar
2
[7] Y.M. Volovenko, G.G. Dubinina, Chemistry of Heterocyclic Com-
pounds. 3 (2002) 372–379.
(
3,5)-H, J = 8.7), 7.15 (m, 5H, Ph-H), 4.23 (t, 2H, CH J
2
=
6.8), 2.96 (t, 2H, CH J = 6.8).
[8] Y.M. Volovenko, G.G. Dubinina, Chemistry of Heterocyclic Com-
pounds. 2 (2002) 241–247.
2
1
3
C NMR (300 MHz, δ, ppm, DMSO-d ): 157.3(C );
6
4a
[
9] G.L. Warren, W. Andrews, A.-M. Capelli, B. Clarke, J. LaLonde, M.H.
Lambert, et al., J. Med. Chem. (2005) ASAP article; doi: 10.1021/
jm050362n.
1
1
1
43.3, 141.5, 140.4 (C , C CAr(1′)); 135.8 (CAr(4′)); 126.2,
6 7a,
21.0 (C , C ); 116.3, 116.1 (2C, CN); 130.7, 128.4, 126.2
22.3, 121.2, 118.4, 118.0, 110.4, 108.8 (CAr(1), CAr(2)
2
3
,
[10] L. Meijer, A.-L. Skaltsounis, P. Magiatis, P. Polychronopoulous, M.
CAr(3), CAr(4), CAr(5), CAr(6), CAr(2′), CAr(3′), CAr(5′), CAr(6′)), 83.2
Knockaert, M. Leost, et al., Chem. Biol. 10 (2003) 1255–1264.
11] A.M. Lawrie, M.E. Noble, P. Tunnah, N.R. Brown, L.N. Johnson, J.A.
Endicott, Nat. Struct. Biol. 4 (1997) 796–799.
12] J. Stamos, M.X. Sliwkowski, C. Eigenbrot, J. Biol. Chem. 277 (2002)
[
[
[
[
(
C ); 44.6, 37.9 (2C, CH ).
7
2
6
-Amino-5-(4-methoxybenzyl)-7-(4-chlorophenylsulfonyl)-
5
3
H-pyrrolo[3,2-b]pyrazine-2,3-dicarbonitrile (7j): m.p.
>
4
6265–46272.
00 °C;
13] M. Knockaert, P. Greengard, L. Meijer, Trends Pharmacol. Sci. 23
(2002) 417–425.
14] R. Perez-Soler, Oncologist 9 (2004) 58–67.
[15] R.D. Mass, Int. J. Radiat. Oncol. Biol. Phys. 58 (2004) 932–940.
16] J.F. Lowrie, R.K. Delisle, D.W. Hobbs, D.J. Diller, Comb. Chem. High
+
LC/MS: tR 1.47 min (98.4%), m/z (M ) 489.1; IR (KBr):
νNH2 3395(as), 3307(s); ν 2220; νC=N 1640; νSO2 1303(as),
CN
−1
1
145(s) (cm ).
[
1
H NMR (300 MHz, δ, ppm, DMSO-d , J (Hz)): 8.74 (s,
6
Throughput Screen. 7 (2004) 495–510.
2
H, NH ), 8.07 (d, 2H, Ar(2,6)-H, J = 8.4), 7.75 (d, 2H, Ar
2
[17] C. McMartin, R.S. Bohacek, J. Comput. Aided Mol. Des. 11 (1997)
(
3,5)-H, J = 8.4), 7.21 (d, 2H, Ar(3′,5′)-H, J = 8.7), 6.93 (d,
333–344.