Identification of 2-anilino-9-methoxy-5, 7-dihydro-6H-pyrimido[5, 4-d][l]benzazepin-6-ones as dual PLK1/VEGF-R2 kinase inhibitor chemotypes by structure-based lead generation

Article abstract of DOI:10.1021/jm901388c

To develop multikinase inhibitors with dual PLK1/VEGF-R2 inhibitory activity, the d-annulated l-benzazepin-2-one scaffold present in the paullone family of kinase inhibitors was investigated as a general structure template suitable for anchoring annulated heterocycles at the hinge region of the ATP binding site. For this purpose, the indole substructure of the paullones was replaced by other nitrogen containing heteroaromatics. The designed scaffolds were synthesized and tested on the indicated kinases. The 2-anilino-5, 7-dihydro-6H-pyrimido[5, 4-d][l]benzazepin-6-ones were found to be VEGF-R2 inhibitors with selectivity against the insulin receptor kinase. The attachment of a methoxy group to the 9-position of the scaffold led to additional PLK1 inhibitory activity, which was explained by an alternative binding mode of the 9-methoxy derivatives. Selected members of the compound class inhibited the VEGF-R2 autophosphorylation in human umbilical vein endothelial cells, the sprouting of human umbilical vein endothelial cell speroids, and the proliferation of diverse cancer cell lines.

Full text of DOI:10.1021/jm901388c

J. Med. Chem. 2010, 53, 2433–2442 2433
DOI: 10.1021/jm901388c
Identification of 2-Anilino-9-methoxy-5,7-dihydro-6H-pyrimido[5,4-d][1]benzazepin-6-ones as Dual
PLK1/VEGF-R2 Kinase Inhibitor Chemotypes by Structure-Based Lead Generation
Anne-Marie Egert-Schmidt,† Jan Dreher,† Ute Dunkel,† Simone Kohfeld,† Lutz Preu,† Holger Weber,‡ Jan E. Ehlert,‡
‡
Bettina Mutschler, Frank Totzke, Christoph Schachtele, Michael H. G. Kubbutat, Knut Baumann, and Conrad Kunick*
‡
‡
‡
†
,†
€
†Technische Universita€t Braunschweig, Institut fu€r Pharmazeutische Chemie, Beethovenstrasse 55, 38106 Braunschweig, Germany, and
‡ProQinase GmbH, Breisacher Strasse 117, 79106 Freiburg, Germany
Received September 18, 2009
To develop multikinase inhibitors with dual PLK1/VEGF-R2 inhibitory activity, the d-annulated
1-benzazepin-2-one scaffold present in the paullone family of kinase inhibitors was investigated as a
general structure template suitable for anchoring annulated heterocycles at the hinge region of the ATP
binding site. For this purpose, the indole substructure of the paullones was replaced by other nitrogen
containing heteroaromatics. The designed scaffolds were synthesized and tested on the indicated
kinases. The 2-anilino-5,7-dihydro-6H-pyrimido[5,4-d][1]benzazepin-6-ones were found to be VEGF-
R2 inhibitors with selectivity against the insulin receptor kinase. The attachment of a methoxy group to
the 9-position of the scaffold led to additional PLK1 inhibitory activity, which was explained by an
alternative binding mode of the 9-methoxy derivatives. Selected members of the compound class
inhibited the VEGF-R2 autophosphorylation in human umbilical vein endothelial cells, the sprouting of
human umbilical vein endothelial cell speroids, and the proliferation of diverse cancer cell lines.
Introduction
on the one hand and a lower risk of drug-resistance develop-
ment during treatment on the other hand.5 Recently multi-
kinase inhibitors have been launched as anticancer drugs, e.g.,
sunitinib,6 sorafenib,7 and dasatinib.8,9
The investigation reported here was directed to find dual
inhibitors of the tumor-related kinases polo-like kinase 1
(PLK1a) and VEGF-R2 (vascular endothelial growth factor-
receptor 2).
PLK1 is an important serine/threonine kinase necessary for
cell division and tissue proliferation of eukaryotes. During
mitosis, PLK1 regulates centrosome maturation, microtubule-
kinetochore attachment, and the onset of cytokinesis during
late mitosis. During anaphase, PLK1 initiates cleavage furrow
ingression and spindle elongation. Cells treated with PLK
inhibitors fail to divide and exit mitosis as tetraploid binucle-
ates.10 PLK1 is overexpressed in a variety of human cancer
tissues and high PLK1 levels are correlated with poor prog-
nosis.11 PLK1-depletion by siRNA is able to induce apopto-
sis,12 and inhibition of PLK1 by antisense-oligonucleo-
tides resulted in growth inhibition of cancer cells.13 Further-
more, application of short hairpin RNA against PLK1 re-
sulted in cancer inhibition in a nude mice model.14 In murine
xenograft assays, the potent ATP-competitive and selective
synthetic PLK1-inhibitor BI2536 also inhibited tumor
growth.15 Taken together, these findings lead to the suggestion
that PLK1 is a valid biological target for potential anticancer
therapeutics.11
Because overexpression and hyperactivity of protein ki-
nases is frequently observed in human tumor tissues, the
inhibition of certain protein kinases was established as a
therapeutic concept during the past decade.1-3 In the begin-
ning of the protein kinase inhibitor era, the selective inhibition
of only one target enzyme appeared desirable to limit un-
wanted side effects. However, given the high similarity of the
ATP-binding pocket architecture in the protein kinases and
the fact that besides a few exceptions4 the majority of estab-
lished kinase inhibitors compete with ATP at the ATP-
binding site of the enzyme, it is not surprising that most kinase
inhibitors show a modest degree of protein kinase selectivity.
Hitting more than one tumor-related protein kinase at a time
is not necessarily a drawback but offers the advantage of a
broader application of the drug against diverse tumor entities
*To whom correspondence should be addressed. Phone: þ49-531-
391-2754. Fax: þ49-531-391-2799. E-mail: c.kunick@tu-braunschweig.de.
a Abbreviations: ACN, acetonitrile; AcOH, acetic acid; ATP, adeno-
sine triphosphate; AUC, area under curve; CDK, cyclin dependent
kinase; CK1, casein kinase 1; DAD, diode array detector; DCM,
dichloromethane; DMF, N,N-dimethyl formamide; DMF-DMA,
dimethylformamide dimethyl acetal; DMSO, dimethyl sulfoxide;
DTT, dithiothreitol; EGF, epidermal growth factor; EtOH, ethanol;
gk, gatekeeper; GSK-3, glycogen synthase kinase-3; HEPES, N-2-
hydroxyethylpiperazine-N0-2-ethanesulfonic acid; HPLC, high perfor-
mance liquid chromatography; HUE, human umbilical endothelial
(cells); HUVEC, primary human umbilical vein endothelial cells; INS-
R, insulin receptor kinase; IVCLS, in vitro cell line screening; NCI,
National Cancer Institute; PCC, pairwise correlation coefficient; PDB,
Protein Data Bank; PEG, polyethylene glycol; PFK, perfluorokerosene;
PLK1, polo-like kinase 1; RPMI, Roswell Park Memorial Institute;
SRB, sulforhodamine B; TCA, trichloroacetic acid; TFA, trifluoroacetic
acid; TLC, thin layer chromatography; TMB, 3,30,5,50-tetramethylben-
zidine; VEGF-R, vascular endothelium growth factor receptor kinase.
The inhibition of tyrosine kinase receptors of the vascular
endothelial growth factor (VEGF) is one of the main modes of
action of the mentioned multikinase inhibitors sunitinib16,17
and sorafenib.7 To growbeyond a volumeof 1-2 mm3, a solid
tumor depends on the supply of the cancer cells with oxygen
and nutrients through blood vessels.18 For an induction of
r
2010 American Chemical Society
Published on Web 02/19/2010
pubs.acs.org/jmc

2434 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 6
Egert-Schmidt et al.
Chart 1. Structures of Paullones: Kenpaullone (1), Alsterpaullone (2), 1-Azakenpaullone (3), and 2-Azakenpaullone (4)
angiogenesis, the tumor secrets VEGF and other growth
factors which interact with VEGF receptors (VEGF-Rs) on
vascular endothelial cells (ECs). The ECs then escape from the
existing blood vessel walls and proliferate toward the source
of the VEGF. After formation of new tubular structures,
blood flows through the new capillaries. The uncontrolled
angiogenesis in tumor tissue is characterized by an irregular
and chaotic structure of the blood vessels. A high vasculariza-
tion is not only a prerequisite for growth of solid tumors but
also increases the formation of metastases. Within the VEGF
receptor family, VEGF-R1 (Flt1) is involved in the early
inflammation process by regulation of monocyte and macro-
phage chemotaxis. VEGF-R3 (Flt4) is important for lym-
phangiogenesis and thus may play a role in the dissemination
of lymphatic metastases of solid tumors. Located on endothe-
lial cells (ECs), VEGF-R2 (KDR) is responsible for tumor
angiogenesis. The inhibition of VEGF-R kinases located on
ECs bears the advantage that in contrast to the tumor cells the
ECs are genetically stable and therefore less prone to develop
resistance against antiproliferative agents. VEGF-R2 is one of
the crucial target structures of the mentioned multikinase
inhibitors sorafenib and sunitinib as well as of drugs currently
in clinical development.19-21 We hypothesized that a single
agent inhibiting both PLK1 and VEGF-R2 might interfere
with tumor growth both on the level of uncontrolled cell
division and tumor vascularization. We therefore were inter-
ested to find novel chemotypes with such dual kinase inhibi-
tion profiles as starting points for an anticancer drug
development campaign.
The paullones (Chart 1) represent a group of established
inhibitors of cyclin-dependent kinases (CDKs) and glycogen
synthase kinase-3 (GSK-3).23 Among this compound class,
kenpaullone (1) is a commercially available dual CDK/GSK-
3-inhibitor.24 Kenpaullone’s 9-nitro analogue, alsterpaullone
(2), displays enhanced CDK/GSK-3 inhibitory potency and
induces apoptosis in cancer cells by perturbation of the
mitochondrial inner membrane potential.25 The binding
mode of alsterpaullone to the ATP binding pocket of GSK-
3β was elucidated by X-ray structure analysis of a protein-
ligand complex (PDB entry 1Q3W).26 Alsterpaullone binds,
among others, by two direct hydrogen bonds emanating from
the lactam moiety directly to the gkþ3 amino acid. A third
water mediated hydrogen bond connects the lactam carbonyl
to the gkþ1 amino acid carbonyl group.26 While the commer-
cially available 1-aza-kenpaullone (3) is widely used as selec-
tive GSK-3 inhibitor,27 its 2-aza isomer 4 lacks selectivity,
showing dual CDK/GSK-3 inhibitory activity similar to
kenpaullone 1.28 In a panel of 71 protein kinases, 1 μM
concentrations of kenpaullone inhibited 9 other kinases be-
sides GSK3 and CDK2/cyclin A.29 This observation suggests
thatthe benzolactam motif of the paullones might bea general
hinge-binding element. Using this information, we designed
the potential protein kinase inhibitor chemotypes A-D
(Table 1) by replacing the indole ring of the paullones by
other heterocyclic systems. The particular structures were
chosen because a straightforward synthesis procedure from
simple common heterocyclic enaminones was envisaged for
these scaffolds.
For the rational idea-generation toward novel protein
kinase inhibitor lead structures, it has been suggested to use
the structural data of protein kinase-ligand complexes in the
Protein Data Bank (PDB).22 A recent analysis of the available
PDB files of protein kinases cocrystallized with diverse in-
hibitors identified 53 molecular inhibitor substructures that
are responsible for binding to the hinge region in the ATP-
binding pocket of kinases.22 The main binding feature of these
structure motifs is a pattern of one to three hydrogen bonds
between the inhibitor and three attachment points in the hinge
segment of the protein kinase, namely the backbone carbonyl
oxygen of the amino acid next to the “gatekeeper” residue
(gkþ1) and both backbone carbonyl and NH group of the
amino acids two positions toward the carboxy terminus, the
so-called “gkþ3” amino acid.22
A preliminary docking study was carried out to check
whether prototypes 21, 24a, and 28a of the designed inhibitors
might fit in the ATP binding pockets of the mentioned target
kinases. The insulin receptor kinase (INS-R) was included in
these studies as a counter kinase of indispensable physiological
importance. It was explicitly desired that the newly deve-
loped kinase inhibitors should not affect INS-R in order to
prevent diabetes mellitus-like side effects during a possible
therapeutic use of the compounds. The docking placed the
designed inhibitors into the ATP binding sites of VEGF-R2
and PLK1 in meaningful poses, albeit 28a was oriented
toward the hinge area of the kinases not with its benzolactam
element but by the anilinopyrimidine partial structure. 21 and
24 were only partially accommodated in the INS-R ATP
binding site suggesting that a selectivity toward this kinase
was conceivable. These preliminary results warranted the
synthesis of the designed compounds in order to evaluate
both their VEGF-R2/PLK1 inhibitory activity and their
selectivity against INS-R.
Weherereport a search for dual PLK1/VEGF-R2inhibitor
chemotypes comprising the d-annulated 1,3-dihydro-2H-1-
benzazepin-2-one scaffold present in the paullone kinase
inhibitor structure. Because the paullone basic structure was
not included in the compilation of characteristic ATP pocket-
binding motifs published by Ghose at al.,22 we intended to
generate protein kinase inhibitors with a certain degree of
novelty by this method.
Chemistry
The test compounds listed in Table 1 were synthesized from
suitable enaminones 9, 10, 11, or 12, respectively, which were

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 6 2435
Table 1. Structures and Kinase Inhibitory Activity of d-Annulated Benzazepinones
basic
structure
IC50 VEGF-R2
[μM]
IC50 PLK1
[μM]
IC50 INS-R
[μM]
code
X
Y
R1
R2
R3
R4
1a
>100a
12
>100a
>100
>100
>100
17
>100a
>100
>100
>100
11
17a
17b
18a
18b
19
20a
20b
21
A
A
A
A
A
A
A
B
CH
CH
CH
CH
N
CH
CH
CH
CH
CH
CH
CH
CH
CH
N
CH
CH
CH
CH
CH
N
H
H
H
-Ph
H
-Ph
-Ph
H
0.63
8.6
-OCH3
-OCH3
H
H
H
H
-OCH3
H
H
H
-OCH3
-OCH3
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
1.8
2.0
>100
>100
>100
NIb
>100
>100
>100
NIb
87
16
NIb
N
-Ph
CH
CH
N
CH
CH
CH
CH
CH
CH
N
22
23
B
B
>100
>100
>100
>100
>100
>100
>100
>100
66
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
15
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
99
24a
24b
25a
25b
26a
26b
27a
27b
28a
28b
28c
28d
28e
28f
28g
28h
28i
28j
28k
28l
28m
28n
28o
28p
29g
29c
29i
C
C
C
C
C
C
C
C
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
H
CH3
H
CH3
H
CH3
H
CH3
H
I
-OCH3
Cl
-NO2
-CH3
-OH
Br
H
CH3
H
CH3
H
CH3
H
CH3
H
H
H
H
H
H
H
H
H
-OH
H
H
H
N
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
N
N
>100
0.14
0.4
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
N
H
H
H
H
H
H
H
H
H
H
Cl
Br
-OH
H
H
CH3
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
0.095
0.19
0.33
0.19
0.08
0.29
0.13
0.051
6.1
47
>100
>100
>100
>100
>100
>100
19
>100
22
>100
>100
>100
>100
>100
>100
>100
>100
24
-OC2H5
H
H
H
H
73
>100
14
57
0.53
0.073
0.035
1.2
-OCH3
-OH
-OH
-OH
-OCH3
-OC2H5
H
-OCH3
-OH
-OH
-OCH3
-OC2H5
H
-OH
-Cl
H
H
H
5.7
9.3
H
H
84
20
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
-OCH3
-OCH3
-OCH3
-OCH3
-OCH3
-OCH3
H
H
H
H
H
H
H
H
H
H
0.028
0.050
0.052
0.015
0.018
0.011
0.13
0.19
0.099
0.14
0.087
2.7
0.75
0.35
H
7.3
0.84
29j
-OH
-OH
-Cl
H
H
H
-OH
-Cl
H
H
H
29n
29o
30g
30c
30i
0.14
0.36
2.6
2.5
N
N
N
N
CH
CH
CH
CH
CH
1.7
2.3
30j
30o
31g
31c
31i
31j
31o
-OH
-OH
-OCH3
-OC2H5
H
1.4
55
N
N
N
N
1.5
2.0
>100
>100
>100
>100
-OH
-Cl
2.6
0.67
-OH
H
a The IC50 values for kenpaullone were determined in the presence of 0.1 μM ATP. For all other compounds except kenpaullone the ATP
concentration was 1 μM. b NI = not inhibited (less than 30% inhibition @ 10 μM concentration of 21).

2436 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 6
Egert-Schmidt et al.
Scheme 1. Synthesis of d-Annulated Benzazepinones and Azaanaloguesa
a (i) DMF-DMA, 115 °C, 2.5 h; (ii) AcOH, 120-140 °C (microwave oven), 10 min; (iii) AcOH, reflux, 4-12 h; (iv) NaOH, propan-2-ol, reflux, 8 h or
150 °C (microwave oven), 30-40 min.
readily preparedfrom the cyclicprecursor ketones5,30 6,31 7,28
or 827 by reaction with dimethylformamide dimethylacetal
(DMF-DMA) using a method originally described for the
synthesis of 9 from the oxolactam 5.32
in position 3 of the prototype B for a carbon atom leads to
prototype A. Interestingly, all members of this subgroup of
compounds (17-19) exhibited a considerable inhibition of
VEGF-R2. At this time, we have no reasonable explanation
for the different VEGF-R2 inhibitory activity of scaffolds A
and B. None of the derivatives belonging to scaffolds A-C
was further pursued because either the PLK1 kinase remained
unaffected or, as was found for 18b, the INS-R counter kinase
was inhibited as well.
In contrast to the results with the aforementioned com-
pounds, virtually all derivatives based on the parent scaffold
D showed potent kinase inhibitory activity for VEGF-R2.
However, most of the structures 28 with the plain benzo
annulation do not show the intended dual-kinase inhibition.
Only the 30,40-disubstituted derivatives 28n and 28o exhibit at
least one-digit micromolar inhibition concentration for
PLK1. The derivatives 28k,l,m,p displaying a 20-substitution
at the phenyl group showed the lowest activity toward VEGF-
R2 within this subgroup. Interestingly, the placement of a
9-methoxy-group into the parent scaffold D (compound class
29) improved the PLK1 inhibitory activity by at least 1 order
of magnitude. Fortunately, none of these 9-methoxy deriva-
tives 29 inhibited the counter-kinase INS-R. A replacement of
the annulated benzene ring for a pyrido substructure led to
different results depending on the position of the pyrido
nitrogen. The derivatives 30, displaying the pyrido nitrogen
in 10-position, exhibited kinase inhibitory activity compar-
able to the benzo analogues 28. In contrast, upon introduction
of a nitrogen in position 11 (Y = N; compounds 31, refer to
Table 1), the VEGF-R2 inhibitory activity was strongly
Heating the enaminones 9-12 with 1H-pyrazol-5-amines
13 in glacial acetic acid in a single mode microwave device
afforded the annulated pyrazolopyrimidines 17-20 in mod-
erate to good yields. In a similar procedure, the annulated
triazolopyrimidines 21-23 were obtained using 3-amino-1H-
1,2,4-triazole (14) as dinucleophile. The annulated pyrimido-
pyridines 24-27 were obtained by refluxing the enaminones
9-12 with 6-aminouracil derivatives 15 in acetic acid. For the
synthesis of the anilinopyrimidines 28-31, the arylguanidi-
nium nitrates 16 were heated with the enaminones 9-12 in the
presence of sodium hydroxide in propan-2-ol either under
conventional conditions or in a microwave device, yielding the
products 28-31 in modest (17%; 29n) to good yields (77%,
28g) (Scheme 1).
Biological Evaluation and Discussion
The compounds listed in Table 1 were evaluated in vitro by
measuringthe IC50 values onVEGF-R2, PLK1, and INS-Rin
a radiometric protein kinase assay (33PanQinase activity
assay). The results revealed that the new d-annulated com-
pounds 21-23 (scaffold B) and 24-27 (scaffold C) were
devoid of noteworthy activity against the three kinases. Con-
sidering the results of the docking studies, these results were
not surprising with regard to INS-R but were disappointing
with regard to VEGF-R2 and PLK1. Exchange of the nitrogen

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 6 2437
Table 2. Cellular Bioactivity of Selected “Anilinopyrimidines” D
inhibition of
VEGF-R2
phosphorylation
in HUE cells
[IC50, μM]
inhibition of
HUVEC
sprouting
[IC50, μM]
code
basic structure
X
R1
R2
R3
28j
D
D
D
D
D
D
D
CH
CH
CH
CH
CH
N
H
H
H
H
-OH
-OH
-Cl
H
1.6
2.4
NTa
NTa
NTa
1.1
28n
28o
29i
-OCH3
-OH
-OC2H5
-OH
-OC2H5
-OH
9.6
-OCH3
-OCH3
H
0.29
0.43
18.6
NIb
0.07
29o
30i
-Cl
H
1.2
NTa
NTa
30o
staurosporinec
N
H
-Cl
a NT = not tested. b NI = not inhibited at 10 μM inhibitor concentration. c Staurosporine was used as a positive control.
diminished and the PLK1-inhibition was lost completely. At the
present stage, we have no satisfactory explanation for the different
kinase inhibitory activity of 30 and 31 on a molecular basis.
To investigate whether the new compounds are also able to
cross membranes and display kinase inhibitory activity in a
cellular environment, seven selected derivatives were studied
in cellular phosporylation assays. For these studies, human
umbilical endothelial (HUE) cells were treated with the
compounds and then stimulated by VEGF. As readout,
the autophosphorylation of VEGF-R2 was determined. The
important role of the positions 9 and 10 of the parent ring
system for the cellular bioactivity can be clearly demonstrated
by the comparison of 28o, 29o, and 30o. All three congeners
have the same substitution pattern at the phenyl substituent
and inhibit the isolated VEGF-R2 enzyme in two-digit nano-
molar concentrations. However, while for inhibition of
VEGF-R2 in HUE cells high concentrations (9.6 μM) of
28o are necessary and the pyrido derivative 30o is not active
up to 10 μM, the 9-methoxy congener 29o inhibits the cellular
kinase in submicromolar concentrations (0.43 μM). The
9-methoxy derivative 29i proved to be even more potent in
the cellular experiment. As a result of these findings, the
impact of29i and29o onendothelial cellsproutingwasstudied
in spheroid-based cellular angiogenesis assays. In these in-
vestigations, spheroids of human umbilical vein endothelial
cells (HUVEC) cultured in collagen were treated concomi-
tantly with VEGF and the test compounds. Both 29i and 29o
showed a clear inhibition of the sprout growths from the
spheroids after 24 h incubation in low micromolar concentra-
tions (EC50 = 1.1 and 1.2 μM, respectively, refer to Table 2).
These results are promising regarding a potential use of this
compound class to suppress angiogenesis in tumor tissue.
For a preliminary evaluation of the antiproliferative activity
of the new kinase inhibitors, seven representatives of the
structure families 28, 29, and 30 were tested in the in vitro cell
line screening (IVCLS)33 of the National Cancer Institute for
growth inhibition on 60 human tumor cell lines. While all
compounds exhibited a strong antiproliferative activity, a
characteristic selectivity pattern was observed for distinct cell
lines. For example, some cell lines from the renal cancer
subpanel (e.g., CAKI-1) were especially sensitive to the
growth inhibitory activity of the compounds. Furthermore,
a characteristic selectivity pattern was found within certain
subpanels. For instance, the ovarian cancer cell line IGROV1
typically was stronger inhibited than other cell lines from the
same subpanel. In contrast, the cell line UACC-257 proved to
be rather resistant against the tested derivatives, with concen-
trations needed to inhibit this melanoma-derived tumor cells
being 2 orders of magnitude higher than for the cell lines
mentioned before (Table 3).
A pronounced selectivity in the IVCLS is regarded as an
indicator of a specific biological mechanism addressing dis-
tinct biological targets. For the structures presented here, it is
assumed that inhibition of a set of cancer-related protein
kinases contributes to growth inhibition. If this was the case,
the IVCLS selectivity pattern of the studied compounds
should show a considerable degree of similarity. Such simi-
larity of patterns can be studied and quantified by the
COMPARE34,35 computer tool which is accessible via the
NCI’s Developmental Therapeutics Program Web site.36 For
two compounds tested in the IVCLS, COMPAREcalculates a
pairwise correlation coefficient (PCC) as an index of similar-
ity. The maximal PCC value is 1, indicating that the compared
patterns are completely identical. A number of results from

2438 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 6
Egert-Schmidt et al.
Table 3. Antiproliferative Activity of Selected Anilinopyrimidines D on Tumor Cell Lines in the in Vitro Cell Line Screening of the National Cancer
Institute
GI50 [μM]a
code
basic structure
X
R1
R2
R3
IGROV1b
CAKI-1c
UACC-257d
28c
28g
28i
28n
28o
29i
30i
D
D
D
D
D
D
D
CH
CH
CH
CH
CH
CH
N
H
H
H
H
H
-OCH3
-OH
H
0.41
3.0
0.51
2.6
61
>100
27
H
H
-OH
-OH
-Cl
H
0.38
0.25
nae
0.23f
0.40
0.75
0.15
4.7 f
0.39
0.16 f
-OCH3
-OH
-OC2H5
-OC2H5
19
39 f
55
-OCH3
H
H
>100
a GI50 = concentration for 50% cell growth inhibition. If not indicated otherwise, results are averaged values from two independent test runs carried
out at different dates. Each test run was carried out in duplicate. Maximum tested concentration was 100 μM. If the GI50 was above 100 μM, the
maximum concentration was used for calculation of averaged values; b Ovarian cancer cell line. c Renal cancer cell line. d Melanoma cell line. e na = test
value not available. f Single test run.
Table 4. Matrix COMPARE of Compounds Tested in the in Vitro Cell
Line Screening of the National Cancer Institutea,b
the literature demonstrate that COMPARE is suitable to
identify activity mechanisms of antiproliferative compounds.
While it has been suggested that correlations >0.6 indicate
similar mechanisms within cells,37 distinct structural com-
pound families can be clustered by lower thresholds down to
PCC’s > 0.35.38
The results listed in Table 4 show a remarkable correlation,
documented by high PCC’s > 0.6, for all compounds 28 that
were tested in the IVCLSP. The aza analogue 30i also
displayed a considerable correlation to this group of com-
pounds. In contrast, the 9-methoxy analogue 29i was clearly
distinguished from the other compounds because not a single
PCC value >0.6 was found. This observation is consistent
with the finding that introduction of a 9-methoxy group
produces a change in potency and kinase selectivity of the
anilinopyrimidines belonging to parent scaffold D.
An interesting result of this study is the superiority of the
compounds 29g,c,j,n,o versus their counterparts from the
series 28, 30, 31 regarding the inhibition of PLK1. While most
of the latter derivatives were either not or only poor PLK1
inhibitors, the majority of the compounds 29 exhibited in-
hibitory activity in submicromolar concentrations. A pairwise
comparison of structural analogues, e.g., of 28o and 29o,
reveals a more than 20-fold improvement in PLK1 inhibitory
activity by the introduction of the methoxy group into posi-
a (a) Color code for pairwise correlation coefficients calculated by the
tion 9 (28 f 29). Because of these large differences in
biological activity, the binding modes of the compounds 29
to PLK1 were studied by molecular docking. A high-scored
COMPARE algorithm. b (b) GI50 values were used. 49-59 cell lines were
included in the calculation of the correlation coefficient.
docking pose of 29n in PLK1 is shown in Figure 1. Obviously,
the 9-methoxy group of 29n fits closely into a hydrophobic
pocket formed by amino acids Gly62 and Ala65 at the “roof”
of the ATP binding site.
To highlight the importance of the presence of the
9-methoxy group, a study of the interaction potential between
a molecular probe and the binding pocket was carried out
using the program GRID.39 Using a methyl group as probe at

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 6 2439
Figure 1. Docking pose of 29n in the ATP-binding site of PLK1. Left, 2D sketch; right: 3D visualization. gkþ1; gkþ3, amino acid position
related to the gatekeeper position in the hinge motif of the kinase. The 9-methoxy group of 29n (red color in the 2D-drawing) fits into a pocket
€
formed by the amino acids Gly62 and Ala65 (structures calculated by the docking module of the Schrodinger 2008 Suite; structure of the host
molecule: PDB code 2OWB).
scaffold might serve as a suitable hinge-binder template for
the design of dual VEGF-R2/PLK1 inhibitors, several new
heterocyclic scaffolds were synthesized and screened. Among
the designed structure classes, only 2-anilino-9-methoxy-5,7-
dihydro-6H-pyrimido[5,4-d][1]-benzazepin-6-ones 29 display
the desired dual potent VEGF-R2/PLK1 inhibitory activity.
A docking study indicated the important role of the
9-methoxy group of 29 which fits into a pocket formed by
the amino acids Gly62 and Ala65 in the ATP binding site of
PLK1. Selected examples of the basic scaffolds 28 and 29
showed inhibition of VEGF-R2 autophosphorylation in cells,
inhibitory activity in a cellular angiogenesis model, and anti-
proliferative activity on cancer cell lines in vitro.
Experimental Section
Kinase Assays.
A radiometric protein kinase assay
(
33PanQinase Activity Assay) was used for measuring the
activity of the protein kinases. All assays were performed
with a BeckmanCoulter/Sagian robotic system and 96-well
FlashPlates from Perkin-Elmer/NEN (Boston, MA) in a
50 μL reaction volume. The reaction cocktail was pipetted
in 4 steps in the following order: 20 μL of assay buffer, 5 μL of
ATP solution (in H2O), 5 μL of test compound (in 10%
DMSO), 10 μL of substrate/10 μL of enzyme solution
(premixed). The assay for all enzymes contained 60 mM
HEPES-NaOH, pH 7.5, 3 mM MgCl2, 3 mM MnCl2, 3 μM
Na-orthovanadate, 1.2 mM DTT, 50 μg/mL PEG20000, 1 μM
[γ-33P]-ATP (approximately 5 ꢀ 105 cpm per well). The
following substrates were used [substrate amount in square
brackets given in ng/50 μL]: Poly(Glu,Tyr)4:1 [125] for VEGF-
R2, Poly(Glu,Tyr)4:1 [125] or Poly(Ala,Glu,Lys,Tyr)6:2.5:1
[125] for INS-R; RBER-CHKtide [2000], or Casein [1000]
for PLK1. The reaction cocktails were incubated at 30 °C for
80 min. The reaction was stopped with 50 μL of 2% (v/v)
H3PO4, and plates were aspirated and washed two times with
200 μL of 0.9% (w/v) NaCl. Incorporation of 33Pi was
Figure 2. Hotspot analysis within the ATP binding site of PLK1
using a methyl probe identified the indicated pocket (yellow grid
shape in the upper center of the picture) as area most contributing to
positive ligand-protein interactions. The depicted bound structure
29n is directing the 9-methoxy group into the identified hotspot area.
a contour level of -4 kcal/mol, the area accommodating the
9-methoxy group of 29n was identified as a favorable binding
region for hydrophobic parts of the ligand (see Figure 2).
A thorough literature survey revealed that recently con-
geners of the basic structure 28 have been filed as inhibitors of
Aurora A and Aurora B kinase, Checkpoint kinase 1 (Chk-1),
and PLK1.40 However, the structural basis for the kinase
inhibition by 28, the concomitant inhibition of VEGF-R2
kinase and the activity of the compounds in cellular angio-
genesis models has not yet been mentioned. Moreover, the
major impact of the 9-methoxy group present in the novel
derivatives 29 for kinase inhibitory activity has not been
reported previously.
determined with
a
microplate scintillation counter
Future studies are directed to explore comprehensively the
group of 9-methoxy-substituted derivatives 29 and to broadly
check their protein kinase profiles and their antiproliferative
activity in cancer cell lines. Further tests in additional cellular
assays are necessary to determine which of the inhibited
protein kinases are crucial for the observed antiproliferative
activity in cancer cell lines.
(Microbeta Trilux, Wallac). For calculation of IC50 values,
data were generated for comparison without enzyme in the
presence of substrate (“low control”). Moreover, data with
enzyme but without inhibitor were generated (“high con-
trol”). The difference between high and low control was taken
as 100% enzyme activity. The residual activities for each
concentration and the compound IC50 values were calculated
using Quattro Workflow V2.x (Quattro Research GmbH,
Munich, Germany; www.quattroresearch.com). The fitting
model for the IC50 determinations was “Sigmoidal response
(variable slope)” with parameters “top” fixed at 100% and
“bottom” at 0%. The fitting method used was a least-squares fit.
Conclusion
On the basis of the starting hypothesis that the seven-
membered lactam structure present in the paullone basic

2440 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 6
Egert-Schmidt et al.
In case of solubility problems, outliers in the higher concen-
tration range were removed to generate more realistic sigmoi-
dal curves for IC50 calculation. Assays were carried out in
singlicate. As parameter for assay quality, the Z0-factor41 was
calculated for the low and high controls of each plate. Assay
plates were repeated in case the Z0-factor drops below 0.4.
Cellular VEGF-R2 Autophosphorylation Assays. In brief,
spontaneously immortalized human umbilical vein endothelial
(HUE) cells were seeded in 48-well culture plates in complete
endothelial growth medium and after overnight attachment
starved in the same medium devoid of growth supplement. On
the day of the assay, cells were treated with test compounds for
90 min at a DMSO concentration of 1%, followed by stimula-
tion of VEGF-R2 autophosphorylation using 100 ng/mL of
ligand VEGF165 for 3 min at room temperature. Subsequently,
cells were lysed and lysates loaded onto a sandwich ELISA
coated with anti-VEGFR2 antibody 3G2. The tyrosine phos-
phorylation status of bound VEGF-R2 was subsequently de-
tected by biotinylated pY-99 antibody, which was then detected
by horseradish peroxidase labeled streptavidin followed by
a chromophoric reaction using TMB substrate. Absorption
values were converted into percentage phosphorylation using
the uninhibited (DMSO-treated) cells as high (= 100%) control
and cells treated with 10 μM Staurosporine as low (= 0%)
control. IC50 values were determined using the program Graph-
Pad Prism 5 (GraphPad Software, Inc., La Jolla, CA), assuming
a sigmoidal dose response.
Endothelial Cell Sprouting in the Spheroid-Based Cellular
Angiogenesis Assay. The experiments were pursued in modifica-
tion of the originally published protocol.42 Primary human
umbilical vein endothelial cells (HUVEC, PromoCell, Heidelberg,
Germany) were used in endothelial cell growth and basal
medium (ECGM/ECBM, PromoCell). In brief, spheroids were
prepared as described43 by pipetting 500 ECs (endothelial cells)
in a hanging drop on plastic dishes to allow overnight spheroid
aggregation. Then 50 EC spheroids were seeded in 0.9 mL of a
collagen solution and pipetted into individual wells of a 24-well
plate to allow polymerization. The test compounds in combina-
tion with VEGF-A [25 ng/mL final assay concentration] were
added after 30 min by pipetting 100 μL of a 10-fold concentrated
working dilution on top of the polymerized gel. Plates were
incubated at 37 °C for 24 h and fixed by adding 4% para-
formaldehyde. Sprouting intensity of ECs was quantitated by an
image analysis system determining the cumulative sprout length
per spheroid using an inverted microscope and the digital
the plates were washed five times with water and air-dried. Then
100 μL of 0.4% sulforhodamine B (SRB) solution in 1% acetic
acid was added. After 10 min, the cavities are washed five times
with 1% acetic acid and the plates were air-dried. Bound SRB is
solubilized with 10 mM trizma base, and the absorbance at 515
nm was measured on a plate reader. The percentage of growth is
calculated at each of the five concentrations levels. Using the
different absorbance measurements [time zero, (Tz), control
growth, (C), and growth in the presence of drug (Ti)], the growth
inhibition 50% (GI50) is the concentration for a Ti value
satisfying the equation [(Ti - Tz)/(C - Tz)] ꢀ 100 = 50.
Compounds were initially evaluated in a prescreening at a
single dose of 10 μM against the 60 human cancer cell lines. Only
those structures which show a considerable antiproliferative
activity in the prescreening were subsequently tested at five
concentration levels for determination of GI50 values. Further
details of the test method have been published.33,34,44
Docking. For a selection of kinase structures suitable for the
docking, high-resolution crystal structures ranging from 1.65 to
˚
2.1 A were retrieved from the PDB. Subsequently, the most
“CDK2-like” PDB files of the respective protein kinases were
searched because the paullones, the archetype of the docked
prototypes A-D, interact best with cyclin dependent kinases.
The CDK2-likeness was analyzed by aligning the structures to
the phosphorylated CDK2/magnesium ATP complex structure
(PDB code 1QMZ) as a common reference by superimposing the
CR-backbone atoms of 38 residues able to interact with ATP-
competitive inhibitors using singular value decomposition,
which allows superimposition of all ligands in a common frame-
work.45 The rms deviations for the pairwise superimposed ATP
binding pockets were calculated and act as indicator for the
active-state CDK2-likeness. The structures selected based on
this rationale were 1YWN (for VEGF-R2), 2OWB (for PLK1),
and 3BU6 (for INS-R).
€
The docking calculations were performed using the Schrodinger
software suite with default settings if not indicated otherwise.
For protein preparation, all the crystallographic water mole-
cules were removed, hydrogens were added, and bond orders
were assigned using Maestro’s Protein Preparation Wizard
(Maestro v. 8.5207).46 The added hydrogen atoms were mini-
mized with all heavy atoms fixed using the OPLS2001 force
field. For each of the three structures, energy grids were built
using the default value of protein atom scaling within a cubic
˚
˚
˚
box of dimensions 28 A ꢀ 28 A ꢀ 28 A. This box was centered
around the centroid of the ATP bound to/from the common
reference 1QMZ. The bounding box dimensions, inside which
€
imaging software Analysis 3.2 (Soft imaging system, Munster,
˚
the centroid of a docked pose was restricted, were set to 16 A ꢀ
Germany). The mean of the cumulative sprout length of 10
randomly selected spheroids was analyzed as an individual data
point. Fitting of IC50 curves and calculation of IC50 values was
performed with GraphPad Prism 5.01 (GraphPad Software,
Inc., La Jolla, CA).
˚
˚
16 A ꢀ 16 A.
47-49
€
Docking was performed using Schodingers Glide.
To
attain a variety of docking poses, a first run was performed using
the standard-precision (SP) mode with a scaling factor for the
van der Waals radii scaling of 0.7. The obtained poses were
refined using Glide’s extra-precision mode (XP, Impact v.
5.0207) with default settings.
In Vitro Cell Line Screening. The in vitro cell line screening
was performed at the American National Cancer Institute
(Bethesda, MD).36 In brief, 60 human tumor cell lines were
cultivated in RPMI 1640 medium containing 5% fetal bovine
serum and 2 mM L-glutamine. Cells were inoculated into 96-well
microtiter plates with plating densities ranging from 5000 to
40000 cells/well and were then incubated at 37 °C, 5% CO2, 95%
air, and 100% relative humidity for 24 h. After 24 h, represen-
tatives of cell lines are fixed with TCA to determine the cell
density at the time of drug addition (Tz). Stock solutions of test
compounds in DMSO (400-fold of the final maximum test
concentration) were diluted with growth medium containing
50 μg/mL gentamicin to provide five drug concentrations. Then
100 μL aliquots of these test compound dilutions were added
to the microtiter wells already containing the cancer cells and
100 μL of medium. The plates were incubated for 48 h at 37 °C,
5% CO2, 95% air, and 100% relative humidity. Cells were fixed
in situ by addition of 50 μL of cold 50% (w/v) TCA and
incubated for 60 min at 4 °C. After discarding the supernatant,
Hotspot Analysis. GRID computations were carried out using
version 22a.50 Hydrogens were added with the program GRIN
˚
(part of the GRID package). The grid spacing was set to 0.33 A;
all other GRID input parameters retained their default values.
Synthetic Chemistry. Monomode microwave device: CEM
Discover focused microwave synthesis system with ChemDriver
software. Melting points (mp) were determined on an electric
variable heater (Barnstead Electrothermal IA 9100) and were
not corrected. IR spectra were recorded as KBr discs on a
1
Thermo Nicolet FT-IR 200. H NMR spectra and 13C NMR
spectra were recorded on the following instruments: Bruker
Avance DRX-400 and Bruker Avance II-600, solvent DMSO-d6
if not stated otherwise, internal standard trimethylsilane, signals
in ppm (δ scale). Elemental analyses were determined on a CE
Instruments FlashEA 1112 elemental analyzer (Thermo Quest).
Mass spectra were recorded on a double-focused sector field

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 6 2441
mass spectrometer Finnigan-MAT 90. Accurate measurements
where conducted according to the peakmatch method using
perfluorokerosene (PFK) as an internal mass reference; (EI)-
J = 8.5/2.6 Hz, ar H), 7.65-7.69 (m, 2H, ar H), 8.00 (d, 1H,
J = 8.8 Hz, ar H), 8.39 (s, 1H, pyrimidine-H), 9.45 (s, 1H, NH),
10.14 (s, 1H, NH). Anal. (C21H20N4O3) C, H, N. HPLC (ACN:
H2O 50:50): purity 96.3% at 254 nm, 97.8% at 280 nm; tN 2.92 min;
tm(DMSO) 1.03 min.
MS: ionization energy 70 eV. TLC: Polygram Sil G/UV254
,
Macherey-Nagel, 40 mm ꢀ 80 mm, visualization by UV illumi-
nation (254 nm). Column chromatography: silica gel 60
(Merck), column width 2 cm, column height 10 cm unless stated
otherwise. Purity was determined by HPLC using the following
devices and settings: Elite LaChrom (Merck/Hitachi), pump
L-2130, autosampler L-2200, diode array detector L-2450,
organizer box L-2000; column: Merck LiChroCART 125-4,
LiChrosphere 100, RP 18, 5 μm, flow rate 1.000 mL/min,
isocratic, volume of injection 10 μL; detection (DAD) at 254
and 280 nm; AUC % method; time of detection 15 min, net
retention time (tN), dead time (tm) related to DMSO. Prepara-
tion of H2O þ (Et3NH)2SO4-buffer (pH 2.5) for HPLC: triethy-
lamine (20.0 mL) and sodium hydroxide (242 mg) were dissolved
in water (980 mL). The solution was adjusted to pH 2.5 by
addition of sulfuric acid. Preparation of H2O/TFA mixture
pH 1.5 for HPLC: water was adjusted to pH 1.5 by addition
of TFA. All compounds employed in biological tests were used
in g95% purity. The following compounds were prepared
according to literature methods: 5,30 6,31 7,28 8,27 9,32 16.51
Compounds 13, 14, 15 were purchased from commercial sup-
pliers and were used without further purification. Synthetic
procedures for the following compounds are available in the
Supporting Information: 11, 12, 17a-b, 18a-b, 19, 20a-b, 21,
22, 23, 24a-b, 25a-b, 26a-b, 27a-b, 28a-p, 29c, 29g, 29j, 29n,
30c, 30g, 30i, 30j, 30o, 31c, 31 g, 31i, 31j, 31o.
General Procedure A for the Preparation of Enaminones 10, 11,
12. A slurry of the respective azepinedione 6, 7, or 8 (1.00 mmol)
in DMF-DMA (3.50 mL, 26.0 mmol) was stirred at 115-120 °C
for 2.5 h. After cooling to room temperature, the resulting
precipitate was collected, washed with petrolether, and crystal-
lized from ethanol.
4-[(Dimethylamino)methylidene]-8-methoxy-3,4-dihydro-1H-
1-benzazepine-2,5-dione (10). Preparation according to general
procedure A yielded yellow crystals (73%); mp 245 °C. IR
(KBr): 3174 cm-1 (NH), 1684 cm-1 and 1633 cm-1 (CdO).
1H NMR (CHCl3-d1, 400 MHz): δ (ppm) = 3.26 (s, 6H, -N-
(CH3)2), 3.83 (s, 3H, -OCH3), 3.42 (s, 2H, CH2), 6.40 (d, 1H,
J = 2.4 Hz, ar H), 6.77 (dd, 1H, J = 8.8/2.5 Hz, ar H), 7.74 (s,
1H, CdH), 7.88 (s, 1H, NH), 7.93 (d, 1H, J = 8.9 Hz, ar. H).
Anal. (C14H16N2O3) C, H, N. HPLC (ACN:H2O 20:80): purity
100.0% at 254 nm, 100.0% at 280 nm; tN 2.73 min; tm(DMSO)
1.03 min.
General Procedure B for the Synthesis of 2-Anilino-5,7-dihy-
dro-6H-pyrimido[5,4-d][1]benzazepin-6-ones 28 and 29 and the
Analogues 30 and 31. Method 1. The enaminone 9, 10, 11, or 12
(1.00 mmol) was refluxed with the respective arylguanidinium
nitrate 16 (1.2 mmol) and NaOH (48 mg, 1.2 mmol) in propan-2-
ol (5 mL) for the indicated reaction time. After cooling to room
temperature, the resulting precipitate was collected and succes-
sively washed with water and petrolether.
Method 2. The enaminone 9 (1.00 mmol) was reacted with the
respective arylguanidinium nitrate 16 (1.2 mmol) and NaOH (48
mg, 1.2 mmol) in propan-2-ol (5 mL). The reaction was con-
ducted in a microwave device using a sealed microwave reaction
vessel for 30-40 min at 150 °C. After cooling to room tempera-
ture, the resulting precipitate was collected and successively
washed with water and petrolether.
2-(4-Ethoxyanilino)-9-methoxy-5,7-dihydro-6H-pyrimido[5,4-d]-
[1]benzazepin-6-one (29i). Preparation according to general proce-
dure B, method 1 from 10 (reaction time 15 h). Crystallization from
EtOH yielded 74% colorless crystals; mp 267 °C. IR (KBr): 3257
cm-1 and 3196 cm-1 (NH), 1680 cm-1 (CdO). 1H NMR (DMSO-
d6, 400 MHz): δ (ppm) = 1.31 (t, 3H, J = 7.0 Hz, -OCH2CH3),
3.33 (s, 2H, CH2; superimposed by H2O peak), 3.83 (s, 3H,
-OCH3), 3.98 (q, 2H, J = 7.1 Hz, -OCH2CH3), 6.76 (d, 1H,
J = 2.3 Hz, ar H), 6.84-6.88 (m, 2H, ar H), 6.95 (dd, 1H,
2-(3-Chloro-4-hydroxyanilino)-9-methoxy-5,7-dihydro-6H-
pyrimido[5,4-d][1]benzazepin-6-one (29o). Preparation accord-
ing to general procedure B, method 1 (reaction time 24 h).
Crystallization from EtOH yielded 35% beige crystals; mp
275 °C. IR (KBr): 3185 cm-1 (NH), ca. 2800-3200 (OH), 1655
cm-1 (CdO). 1H NMR (DMSO-d6, 400 MHz): δ (ppm) = 3.35
(s, 2H, CH2 superimposed by H2O peak), 3.84 (s, 3H,
-OCH3), 6.78 (d, 1H, J = 2.6 Hz, ar H), 6.91 (d, 1H, J = 8.8,
ar H), 6.96 (dd, 1H, J = 8.7/2.5 Hz, ar H), 7.50 (dd, 1H, J =
8.9/2.6, ar H), 7.90 (d, 1H, J = 2.6 Hz, ar H), 8.00 (d, 1H, J =
8.8 Hz, ar H), 8.42 (s, 1H, pyrimidine-H), 9.52 (s, 1H, NH),
9.68 (s, 1H, OH), 10.16 (s, 1H, NH). (C19H15ClN4O3) HRMS
(EI) (m/z): [M - H]þ calcd 381.07544, found 381.07470. HPLC
(ACN:H2O 35:65): purity: 98.4% at 254 nm, 98.4% at 280 nm;
tN 3.84 min, tm (DMSO) 1.03 min.
Acknowledgment. We are grateful to the National Cancer
Institute for performing the in vitro cell line screening. The
work was funded in part by the European Commission
(contract no. LSHB-CT-2004-503467, to A.-M. E.-S., U. D.,
H. W., J.E.E., B. M., F. T., C. S., M. H. G. K., andC. K.) andthe
Deutsche Forschungsgemeinschaft (DFG, to J. D. and K. B.).
Supporting Information Available: Two-dimensional sketches
of the docking results for 1, 21, 24a, and 28a in VEGF-R2,
PLK1, and INS-R, respectively; 13C NMR data of 10, 29i, and
29o; details for the synthesis, spectroscopic data, and HPLC
purity data of 11, 12, 17a-b, 18a-b, 19, 20a-b, 21, 22, 23,
24a-b, 25a-b, 26a-b, 27a-b, 28a-p, 29c, 29g, 29j, 29n, 30c,
30g, 30i, 30j, 30o, 31c, 31g, 31i, 31j, and 31o. This material is
available free of charge via the Internet at http://pubs.acs.org.
References
(1) Arora, A.; Scholar, E. Role of tyrosine kinase inhibitors in cancer
therapy. J. Pharmacol. Exp. Ther. 2005, 315, 971–979.
(2) Kunick, C.; Egert-Schmidt, A. The short history of protein kinase
inhibitors. New, competitive, successful. Pharm. Unserer Z. 2008,
37, 360–368.
(3) Zhang, J.; Yang, P.; Gray, N. Targeting cancer with small molecule
kinase inhibitors. Nat. Rev. Cancer 2009, 9, 28–39.
(4) Spicer, J.; Rewcastle, G.; Kaufman, M.; Black, S.; Plummer, M.;
Denny, W.; Quin, J. r.; Shahripour, A.; Barrett, S.; Whitehead, C.;
Milbank, J.; Ohren, J.; Gowan, R.; Omer, C.; Camp, H.; Esmaeil,
N.; Moore, K.; Sebolt-Leopold, J. P. S; Merriman, R.; Ortwine,
D.; Warmus, J.; Flamme, C.; Pavlovsky, A.; Tecle, H. 4-Anilino-5-
carboxamido-2-pyridone derivatives as noncompetitive inhibitors
of mitogen-activated protein kinase kinase. J. Med. Chem. 2007, 50,
5090–5102.
(5) Petrelli, A.; Giordano, S. From single- to multi-target drugs in
cancer therapy: when aspecificity becomes an advantage. Curr.
Med. Chem. 2008, 15, 422–432.
(6) Mendel, D. B.; Laird, A. D.; Xin, X.; Louie, S. G.; Christensen,
J. G.; Li, G.; Schreck, R. E.; Abrams, T. J.; Ngai, T. J.; Lee, L. B.;
Murray, L. J.; Carver, J.; Chan, E.; Moss, K. G.; Haznedar, J. O.;
Sukbuntherng, J.; Blake, R. A.; Sun, L.; Tang, C.; Miller, T.;
Shirazian, S.; McMahon, G.; Cherrington, J. M. In vivo antitumor
activity of SU11248, a novel tyrosine kinase inhibitor targeting
vascular endothelial growth factor and platelet-derived growth
factor receptors: Determination of a pharmacokinetic/pharmaco-
dynamic relationship. Clin. Cancer Res. 2003, 9, 327–337.
(7) Wilhelm, S.; Carter, C.; Lynch, M.; Lowinger, T.; Dumas, J.;
Smith, R.; Schwartz, B.; Simantov, R.; Kelley, S. Discovery and
development of sorafenib: a multikinase inhibitor for treating
cancer. Nat. Rev. Drug Discovery 2006, 5, 835–844.
(8) Carter, T.; Wodicka, L.; Shah, N.; Velasc, A.; Fabian, M.; Treiber,
D.; Milanov, Z.; Atteridge, C.; Biggs, W. r.; Edeen, P.; Floyd, M.;
Ford, J.; Grotzfeld, R.; Herrgard, S.; Insko, D.; Mehta, S.; Patel,
H.; Pao, W.; Sawyers, C.; Varmus, H.; Zarrinkar, P.; Lockhart, D.
Inhibition of drug-resistant mutants of ABL, KIT, and EGF

2442 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 6
Egert-Schmidt et al.
receptor kinases. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 11011–
11016.
(29) Bain, J.; Plater, L.; Elliott, M.; Shpiro, N.; Hastie, C.; McLauchlan,
H.; Klevernic, I.; Arthur, J.; Alessi, D.; Cohen, P. The selectivity of
protein kinase inhibitors: a further update. Biochem. J. 2007, 408,
297–315.
(30) Kunick, C. Synthese [b]-kondensierter Azepindione durch Deal-
koxycarbonylierung. Arch. Pharm. (Weinheim) 1991, 324, 579–581.
(31) Wieking, K.; Knockaert, M.; Leost, M.; Zaharevitz, D. W.; Meijer,
L.; Kunick, C. Synthesis of paullones with aminoalkyl side chains.
Arch. Pharm. Pharm. Med. Chem. 2002, 335, 311–317.
(32) Chen, W.-Y.; Gilman, N. W. Synthesis of 7-phenylpyrimido[5,
4-d][1]benzazepin-2-ones. J. Heterocycl. Chem. 1983, 20, 663–666.
(33) Shoemaker, R. The NCI60 human tumour cell line anticancer drug
screen. Nat. Rev. Cancer 2006, 6, 813–823.
(34) Boyd, M. R.; Paull, K. D. Some practical considerations and
applications of the National Cancer Institute in vitro anticancer
drug discovery screen. Drug Dev. Res. 1995, 34, 91–109.
(35) Paull, K. D.; Shoemaker, R. H.; Hodes, L.; Monks, A.; Scudiero,
D. A.; Rubinstein, L.; Plowman, J.; Boyd, M. R. Display and
analysis of patterns of differential activity of drugs against human
tumor cell lines: development of meangraph and COMPARE
algorithm. J. Natl. Cancer Inst. 1989, 81, 1088–1092.
(9) Shah, N.; Tran, C.; Lee, F.; Chen, P.; Norris, D.; Sawyers, C.
Overriding imatinib resistance with a novel ABL kinase inhibitor.
Science 2004, 305, 399–401.
ꢀ ꢀ
(10) Petronczki, M.; Lenart, P.; Peters, J. Polo on the rise-from mitotic
entry to cytokinesis with PLK1. Dev. Cell 2008, 14, 646–659.
(11) Strebhardt, K.; Ullrich, A. Targeting polo-like kinase 1 for cancer
therapy. Nat. Rev. Cancer 2006, 6, 321–330.
€
(12) Spankuch-Schmitt, B.; Bereiter-Hahn, J.; Kaufmann, M.;
Strebhardt, K. Effect of RNA silencing of polo-like kinase-1
(PLK1) on apoptosis and spindle formation in human cancer cells.
J. Natl. Cancer Inst. 2002, 94, 1863–1877.
€
(13) Spankuch-Schmitt, B.; Wolf, G.; Solbach, C.; Loibl, S.; Knecht,
€
R.; Stegmuller, M.; von Minckwitz, G.; Kaufmann, M.;
Strebhardt, K. Downregulation of human polo-like kinase activity
by antisense oligonucleotides induces growth inhibition in cancer
cells. Oncogene 2002, 21, 3162–3171.
€
(14) Spankuch, B.; Matthess, Y.; Knecht, R.; Zimmer, B.; Kaufmann,
M.; Strebhardt, K. Cancer inhibition in nude mice after systemic
application of U6 promoter-driven short hairpin RNAs against
PLK1. J. Natl. Cancer Inst. 2004, 96, 862–872.
(36) NCI-60 DTP Human Tumor Cell Line Screen; Developmental
Therapeutics Program NCI/NIH; http://dtp.nci.nih.gov/branches/btb/
ivclsp.html (accessed 2008).
(37) Andreani, A.; Burnelli, S.; Granaiola, M.; Leoni, A.; Locatelli, A.;
Morigi, R.; Rambaldi, M.; Varoli, L.; Landi, L.; Prata, C.;
Berridge, M.; Grasso, C.; Fiebig, H.; Kelter, G.; Burger, A.;
Kunkel, M. Antitumor activity of bis-indole derivatives. J. Med.
Chem. 2008, 51, 4563–4570.
ꢀ ꢀ
(15) Steegmaier, M.; Hoffmann, M.; Baum, A.; Lenart, P.; Petronczki,
ꢀ
€
M.; Krssak, M.; Gurtler, U.; Garin-Chesa, P.; Lieb, S.; Quant, J.;
Grauert, M.; Adolf, G.; Kraut, N.; Peters, J.; Rettig, W. BI 2536, a
potent and selective inhibitor of polo-like kinase 1, inhibits tumor
growth in vivo. Curr. Biol. 2007, 17, 316–322.
(16) McIntyre, J. Sunitinib malate. Drugs Future 2005, 30, 785–792.
(17) Press, M.; Lenz, H. EGFR, HER2 and VEGF pathways: validated
targets for cancer treatment. Drugs 2007, 67, 2045–2075.
(18) Folkman, J. What is the evidence that tumors are angiogenesis
dependent? J. Natl. Cancer Inst. 1990, 82, 4–6.
(19) Gilbert, M.; Mousa, S.; Mousa, S. Current status of anti-vascular
endothelial growth factor (VEGF) stategies and future directions.
Drugs Fut. 2008, 33, 515–525.
€
€
(38) Kunick, C.; Bleeker, C.; Pruhs, C.; Totzke, F.; Schachtele, C.;
Kubbutat, M. H. G.; Link, A. Matrix compare analysis discrimi-
nates subtle structural differences in a family of novel antiproli-
ferative agents, diaryl-3-hydroxy-2,3,3a,10a-tetrahydrobenzo[b]-
cylopenta[e]azepine-4,10(1H,5H)-diones. Bioorg. Med. Chem.
Lett. 2006, 16, 2148–2153.
(20) Underiner, T.; Ruggeri, B.; Gingrich, D. Development of vascular
endothelial growth factor receptor (VEGFR) kinase inhibitors as
anti-angiogenic agents in cancer therapy. Curr. Med. Chem. 2004,
11, 731–745.
(21) Zhong, H.; Bowen, J. Molecular design and clinical development of
VEGFR kinase inhibitors. Curr. Top. Med. Chem. 2007, 7, 379–
393.
(39) Goodford, P. J. A computational procedure for determining
energetically favorable binding sites on biologically important
macromolecules. J. Med. Chem. 1985, 28, 849–857.
(40) Blackburn, C.; Claiborne, C. F.; Cullis, C. A.; Dales, N. A.; Patane,
M. A.; Stirling, M.; Stradella, O. G.; Weatherhead, G. S. Prepara-
tion of lactam compounds useful as protein kinase inhibitors.
Patent WO 2006041773, 2006.
(22) Ghose, A.; Herbertz, T.; Pippin, D.; Salvino, J.; Mallamo, J.
Knowledge based prediction of ligand binding modes and rational
inhibitor design for kinase drug discovery. J. Med. Chem. 2008, 51,
5149–5171.
(23) Pies, T.; Schaper, K.-J.; Leost, M.; Zaharevitz, D. W.; Gussio, R.;
Meijer, L.; Kunick, C. CDK1-inhibitory activity of paullones
depends on electronic properties of 9-substituents. Arch. Pharm.
Pharm. Med. Chem. 2004, 337, 486–492.
(24) Leost, M.; Schultz, C.; Link, A.; Wu, Y.-Z.; Biernat, J.; Mandelkow,
E.-M.; Bibb, J. A.; Snyder, G. L.; Greengard, P.; Zaharevitz, D. W.;
Gussio, R.; Senderowicz, A. M.; Sausville, E. A.; Kunick, C.; Meijer,
L. Paullones are potent inhibitors of glycogen synthase kinase-3b and
cyclin-dependent kinase 5/p25. Eur. J. Biochem. 2000, 267, 5983–
5994.
(41) Zhang, J.; Chung, T.; Oldenburg, K. A simple statistical parameter
for use in evaluation and validation of high throughput screening
assays. J. Biomol. Screening 1999, 4, 67–73.
(42) Korff, T.; Augustin, H. Tensional forces in fibrillar extracellular
matrices control directional capillary sprouting. J. Cell Sci. 1999,
112, 3249–3258.
(43) Korff, T.; Augustin, H. Integration of endothelial cells in multi-
cellular spheroids prevents apoptosis and induces differentiation.
J. Cell Biol. 1998, 143, 1341–1352.
(44) Grever, M. R.; Schepartz, S. A.; Chabner, B. A. The National
Cancer Institute: cancer drug discovery and development program.
Semin. Oncol. 1992, 19, 622–638.
(45) Vulpetti, A.; Bosotti, R. Sequence and structural analysis of kinase
ATP pocket residues. Farmaco 2004, 59, 759–765.
€
(46) Maestro, version 8.5; Schrodinger, L.L.C.: New York, 2008.
(25) Lahusen, T.; De Siervi, A.; Kunick, C.; Senderowicz, A. Alster-
€
paullone, a novel cyclin-dependent kinase inhibitor, induces apop-
tosis by activation of caspase
(47) Glide, version 5.0; Schrodinger, L.L.C.: New York, 2008.
9
due to perturbation in
(48) Friesner, R. A.; Banks, J. L.; Murphy, R. B.; Halgren, T. A.; Klicic,
J. J.; Mainz, D. T.; Repasky, M. P.; Knoll, E. H.; Shelley, M.; Perry,
J. K.; Shaw, D. E.; Francis, P.; Shenkin, P. S. Glide: a new
approach for rapid, accurate docking and scoring. 1. Method
and assessment of docking accuracy. J. Med. Chem. 2004, 47,
1739–1749.
(49) Halgren, T. A.; Murphy, R. B.; Friesner, R. A.; Beard, H. S.; Frye,
L. L.; Pollard, W. T.; Banks, J. L. Glide: a new approach for rapid,
accurate docking and scoring. 2. Enrichment factors in database
screening. J. Med. Chem. 2004, 47, 1750–1759.
(50) GRID version 22a; Molecular Discovery Ltd.: Oxford, UK, 2004.
(51) Tavares, F.; Boucheron, J.; Dickerson, S.; Griffin, R.; Preugschat,
F.; Thomson, S.; Wang, T.; Zhou, H. N-Phenyl-4-pyrazolo[1,5-
b]pyridazin-3-ylpyrimidin-2-amines as potent and selective inhibi-
tors of glycogen synthase kinase 3 with good cellular efficacy.
J. Med. Chem. 2004, 47, 4716–4730.
mitochondrial membrane potential. Mol. Carcinog. 2003, 36,
183–194.
(26) Bertrand, J. A.; Thieffine, S.; Vulpetti, A.; Cristiani, C.; Valsasina,
B.; Knapp, S.; Kalisz, H. M.; Flocco, M. Structural characteriza-
tion of the GSK-3beta active site using selective and non-selective
ATP-mimetic inhibitors. J. Mol. Biol. 2003, 333, 393–407.
(27) Kunick, C.; Lauenroth, K.; Leost, M.; Meijer, L.; Lemcke, T. 1-
Azakenpaullone is a selective inhibitor of glycogen synthase
kinase-3. Bioorg. Med. Chem. Lett. 2004, 14, 413–416.
(28) Stukenbrock, H.; Mussmann, R.; Geese, M.; Ferandin, Y.; Lozach,
O.; Lemcke, T.; Kegel, S.; Lomow, A.; Burk, U.; Dohrmann, C.;
Meijer, L.; Austen, M.; Kunick, C. 9-Cyano-1-azapaullone
(cazpaullone), a glycogen synthase kinase-3 (GSK-3) inhibitor acti-
vating pancreatic beta cell protection and replication. J. Med. Chem.
2008, 51, 2196–2207.