Benzo[c][2,7]naphthyridines as inhibitors of PDK-1

Article abstract of DOI:10.1016/j.bmcl.2009.07.007

A series of substituted benzo[c][2,7]-naphthyridines were prepared and showed good potency in inhibiting PDK-1. The synthesis and SAR of this series of compounds are presented as well as the X-ray crystal structure of one of these analogs in a complex wit

Full text of DOI:10.1016/j.bmcl.2009.07.007

Bioorganic & Medicinal Chemistry Letters 19 (2009) 5225–5228
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Benzo[c][2,7]naphthyridines as inhibitors of PDK-1
a
a
a
a
Kyung-Hee Kim a, , Allan Wissner , Middleton B. Floyd Jr. , Heidi L. Fraser , Yanong D. Wang ,
*
Russell G. Dushin a, Yongbo Hu a, Andrea Olland b, Bing Guo c, Kim Arndt c
a Wyeth Research, Chemical Sciences, 401 N. Middletown Road, Pearl River, NY 10965, United States
b Wyeth Research, Chemical Sciences, 87 Cambridge Park Drive, Cambridge, MA 02140, United States
c Wyeth Research, Oncology Research, 401 N. Middletown Road, Pearl River, NY 10965, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of substituted benzo[c][2,7]-naphthyridines were prepared and showed good potency in inhib-
iting PDK-1. The synthesis and SAR of this series of compounds are presented as well as the X-ray crystal
structure of one of these analogs in a complex with PDK-1.
Received 28 May 2009
Revised 29 June 2009
Accepted 2 July 2009
Available online 8 July 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
PDK-1
PDK-1 inhibitor
Benzo[c][2,7]naphthyridines
The interactions of growth factors with their cell surface recep-
tors lead to the activation of phosphatidylinositol 3-kinase (PI3K),
which generates phosphatidylinositol 3,4,5-trisphosphate (PIP3).
PI3K and PIP3 mediate activation of a group of cAMP-dependent
protein kinase/protein kinase G/protein kinase C family (AGC fam-
ily) protein kinases. Phosphoinositide-dependent protein kinase 1
(PDK-1) phosphorylates and activates the AGC family kinase mem-
bers, which play important roles in the progression of cancer, such
as AKT, S6 kinase and protein kinase C.1 PDK-1 is recruited to the
cell membrane by binding to PIP3. This event is required for AKT
activation.
The PI3K/PDK-1/AKT kinase signaling pathway is commonly
elevated in tumors through the over expression of PI3K or AKT pro-
teins.2 The pathway is also hyperactivated through the frequent
loss of the tumor suppressor PTEN, a phosphatase that removes
the phosphate from the D3 position of PIP3.3 Cancers involving
the latter presumably have elevated PIP3 levels. Since PIP3 contrib-
utes to the activation of AGC kinases and AGC kinases require PDK-
1 for activation, an inhibitor of PDK-1 would likely inhibit the
growth and proliferation of many cancers. The important role of
PDK-1 in several key signaling pathways involved in the progres-
sion of cancer led us to investigate the use of PDK-1 inhibitors as
anticancer agents.
1 inhibitor with an IC50 of 60 nM in an ELISA assay.5 The previous
work suggested that compound 1 binds to PDK-1 via three hydro-
gen bonds: N8 nitrogen to the NH of Ala 162, C6-amino group to
the backbone carbonyl oxygen of Ser 160, and N5 nitrogen to side
chain hydroxyl group of Thr 222.
The previous work also suggested that the C3-amino group was
important for good potency, but the X-ray structure of 1 com-
plexed to the catalytic domain of PDK-1 did not show a direct
hydrogen bonding interaction between the protein and this group.
Additionally, we felt that the tetracyclic core of 1 would not be an
ideal starting point to design inhibitors with good ADME proper-
ties. With this in mind, we decided to design a series where the ter-
minal ring containing the C3-amino group was removed and
replaced with a phenyl or heterocyclic ring having various ap-
pended groups. This resulted in
a
series of 2-aryl-
benzo[c][2,7]naphthyridines. The syntheses of this class of com-
pounds are illustrated in Schemes 1 and 2. As shown in Scheme
1, the reaction of 4-chloro-3-cyanoquinolines 26 with t-butyl
NH2
3
R1
N5
N
PDK-1 appears to be a relatively unexplored kinase and not
many inhibitors of PDK-1 have been reported.4 We previously re-
ported dibenzo[c,f][2,7]-naphthyridine 1 (Fig. 1) as a potent PDK-
R2
11
O
6
NH2
NH2
R3
benzo[c][2,7]naphthyridines
N
10
O
N
8
1
* Corresponding author. Tel.: +1 845 602 2525; fax: +1 845 602 5561.
E-mail addresses: kimk@wyeth.com, khkim87@hotmail.com (K.-H. Kim).
Figure 1. PDK-1 inhibitors.
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.07.007

5226
K.-H. Kim et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5225–5228
few alternative methods were also investigated to prepare the 4-
O
methyl-3-cyanoquinoline intermediates that resulted in the prep-
aration of compounds 19b–e.8,9 The compounds 19a–e were con-
densed with an ester or imidazolide to provide intermediate
ketones 20.10 These intermediates were generally not purified,
but were directly reacted with ammonium acetate in refluxing ace-
tic acid to provide benzonaphthyridines inhibitors 21–31. The
reaction sequence presumably proceeds via initial imine formation
followed by cyclization onto the appended cyano group, and tauto-
merization to give the desired heterocycle.
NC
Cl
N
O
CN
O
CN
O
b
a
R3
R3
N
3
2
R1
CN
CN
N
O
O
NH2
c
R3
N
R3
N
Our initial efforts were focused on finding an R1 substituent
which would result in a PDK-1 inhibitor that was equipotent to
1. As shown in Table 1, compounds 5–8, which contain cyclic
4
via c
e
Cl
5-11, R3=O(CH2)2OCH3
12, R3=OCH3
d
N
O
NH2
via d and e
Table 1
14, R3=O(CH2)2OCH3
15-16, R3=OCH3
R3
N
PDK-1 potency of benzonaphthyridine compounds 5–12, 15, and 21–24
R1
N
13
Scheme 1. Reagents and conditions: (a) CNCH2CO2Bu-t, NaH, DMF; (b) 1,2-
dichlorobenzene, reflux; (c) amines or imidazoles, 135 °C, THF; (d) HCl (g), CHCl3;
(e) aryl or heteroaryl boronic acid, PPh3, Pd2dba3, DMF–toluene, K2CO3 (aq.).
O
NH2
R3
N
Compounds
R1
R3
IC50 (nM)
a
O
N
5
–OCH2CH2OCH3
–OCH2CH2OCH3
–OCH2CH2OCH3
739
O
O
CN
OH
O
O
b, c
a
N
N
NH2
H
O
6
7
7490
1864
N
18
17
R1
O
CN
R2
R3
R2
R3
CN
N
d
e
N
OH
N
8
–OCH2CH2OCH3
2064
20
19a, R2, R3=OCH3
19b, R2, R3=OCH2CH3
19c, R2=OCH3, R3=OCH2Ph
N
19d, R2=OCH3, R3=O(CH2)3 -(N-piperidinyl)
9
–OCH2CH2OCH3
–OCH2CH2OCH3
70
19e, R2=OCH3, R3=OCH2 -(4-N-Me-piperidinyl)
N
R1
N
10
108
N
N
R2
NH2
R3
N
11
–OCH2CH2OCH3
2033
N
21-31
N
Scheme 2. Reagents and conditions: (a) EtO2CCH2CN, NH4OAc, 165 °C (neat); (b)
POCl3; (c) H2, Pd/C, THF, K2CO3; (d) R1–CO2CH3 or R1–imidazolide, LiHMDS, THF
(ꢀ78 to 0 °C); (e) NH4OAc, HOAc, reflux.
N
12
15
21
22
23
–OCH3
100
N
cyanoacetate provided the cyanoquinolines 3 which were ther-
mally decarboxylated to give the dicyano compounds 4. Reaction
of these dicyano compounds with various amines or imidazoles
provided the inhibitors 5–12. The reaction of compounds 4 with
hydrogen chloride gas in ethanol-free chloroform provided
chloro-substituted intermediates 13, which were reacted with aryl
or heteroaryl boronic acids under Suzuki conditions to provide the
arylated compounds 14–16.7 An alternative synthetic route for the
preparation of benzonaphthyridine derivatives where R1 is at-
tached to the core by a carbon–carbon bond is outlined in Scheme
2. As shown in Scheme 2, the reaction of the 2-amino-acetophe-
none derivative 17 with ethyl cyanoacetate and excess ammonium
acetate provided compound 18. The hydroxyl group of compound
18 was converted to the chloride using POCl3 and then the chlorine
atom was removed by hydrogenolysis over a palladium catalyst,
resulting in the formation of 4-methyl-3-cyanoquinoline 19a. A
O
–OCH3
>1000
66
N
–OCH3
N
–OCH3
>100
6923
35
N
N
–OCH2-(4-N-Me-piperidinyl)
N
24
–OCH3
a
Values are averages of two or more experiments.

K.-H. Kim et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5225–5228
5227
and acyclic aliphatic amine substituents at R1, were 10–100-fold
Table 2
PDK-1 potency of benzonaphthyridine compounds: optimization of R2 and R3
less potent than compound 1. Molecular modeling, based on our
earlier X-ray crystal structure of 1 complexed to PDK-1,5 suggested
that it might be possible to set up a new interaction with the pro-
tein by incorporating a hydrogen bond acceptor on the R1 substitu-
ent which might favorably interact with Lys 111. Accordingly,
compounds 9 and 12, both containing an imidazole substituent
at R1, were prepared and found to be of comparable potency (70
and 100 nM, respectively) to 1. Compound 10, which contains a
methyl group at the 40-position of the imidazole substituent, was
also of comparable potency (108 nM), while 11, which contains a
phenyl substituent on the imidazole ring, was considerably less po-
1'N
5'
2
N
R2
R3
4
NH2
N
R2
R3
IC50 (nM)
a
Compounds
24
25
26
14
27
–OCH3
–OCH2CH3
–OCH3
–OCH3
–OCH3
–OCH3
–OCH2CH3
–OCH2Ph
–OCH2CH2OCH3
–OCH2CH2CH2(N-piperidinyl)
35
74
146
43
tent (2 lM). Compound 21, where the imidazole ring was linked
via a carbon atom to core structure, showed essentially equivalent
potency compared to 9. Other heterocycles were installed at R1 po-
sition. Compounds 15, 22 and 23 were significantly less potent, but
compound 24 with 30-pyridyl group showed excellent potency
with an IC50 of 35 nM. Significantly, molecular modeling of 24 pre-
dicted that the N-atom on the pyridine ring would make the same
hydrogen bonding interaction with Lys 111 as predicted for the
imidazole derivatives. Clearly, the SAR of this series indicates the
importance of having a hydrogen bonding acceptor located at the
correct position on the R1 substituent.
144
a
Values are averages of two or more experiments.
ular modeling experiments. We believe that the 30-nitrogen on the
imidazole ring of 9 and 21 or the 10-nitrogen on the pyridine ring of
14 and 24–31 (Tables 2 and 3) are essential for the improved po-
tency exhibited by these analogs.
As suggested by the X-ray crystal structure of 9 complexed to
PDK-1, the structural requirements at the R2 and R3 positions were
found to be much less rigid since these substituents pointed to-
wards the solvent (Table 2). The dimethoxy groups of 24 could
be replaced with diethoxy groups (25) with little loss in potency.
Even replacing the R3 methoxy by benzyl group (26) did not greatly
diminish activity. Compounds with a longer chain at the R3 posi-
tion, such as 14 and 27, also retained good activity. However, the
compound with the ether substituent (14) was about 3 times more
potent compared to the compound having the terminal piperidino
group (27). Although these extended groups at the R2 and R3 posi-
tions did not improve potency, such modifications have the poten-
tial to improve the physicochemical properties of the molecules in
this series.
Finally, we examined the effect of substituents on the pyridine
ring of compound 24 in hopes of further improving the potency
and selectivity of this series. Examination of the X-ray crystal
structures of 9 and 1 with PDK-1 suggested that an additional
favorable interaction could be established between functionality
placed on a pyridine R1 substituent with a non-conserved Ser 94
residue in PDK-1 that could potentially improve selectivity against
other AGC kinases.
Our molecular modeling predictions were largely confirmed by
the X-ray crystal structure of compound 9 complexed to PDK-1 at
2.30 Å resolution (Fig. 2).11 The crystal structure showed that com-
pound 9 binds at the active site by establishing a number of hydro-
gen bonding interactions: N6 to the NH of Ala 162 at the hinge
region, the C4-NH2 group to the backbone carbonyl oxygen of Ser
160, and N3 to the side chain of the gatekeeper residue Thr 222.
These interactions are consistent with the interactions observed
previously with 1.5 Additionally, this crystal structure showed that
the 30-imidazole nitrogen on the head-piece makes a new hydro-
gen bond with the side chain of Lys 111, as predicted by our molec-
3' N
1'
N
2
N3
NH2
9
8
O
O
4
O
N
6
9
Table 3
PDK-1 potency of benzonaphthyridine compounds: optimization of pyridyl ring
1'N
5'
R4
2
N
O
NH2
4
O
N
a
Compounds
R4
H
IC50 (nM)
24
28
29
30
31
16
35
26
57
65
50-NH2
50-OH
50-Br
20-CH3
60-NH2
532
>1000
a
Values are averages of two or more experiments.
Figure 2. X-ray structure of 9 in complex with PDK-1 in the ATP binding site.

5228
K.-H. Kim et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5225–5228
ents on the phenyl ring. Thus, it was found that a methoxy group
H
N
Br
improved the activity of 33 relative to 32 when placed at the 2-po-
sition of the aminophenyl pyridine substituent, while potency was
lost when a single methoxy group was placed on the 3- or 4-posi-
tion (34 or 35). Introduction of an hydroxymethyl group on the 2-,
3-, or 4-positions of the phenyl ring increases potency (36–38)
compared to 32. Similarly, compounds 39 and 40, which contain
hydroxyethyl groups on the phenyl ring, displayed improved po-
tency. In general, the ortho substituted compounds showed better
activity compared to the meta or para substituted compounds (36
vs 37/38 or 39 vs 40). Within the ortho substituted compounds,
small substituents (41 and 42) showed better activity compared
to compounds with bulky groups (43 and 44). The results pre-
sented in Table 4 indicate a preference for ortho substituents on
the nitrogen linked aryl group. Further investigations of modifica-
tions in this region of the molecule will be the subject of future
reports.
In summary, we have described the synthesis and SAR of a novel
series of benzo[c][2,7]naphthyridines as potent PDK-1 inhibitors.
The incorporation of an imidazole or pyridyl moiety on the 2-posi-
tion of the benzo[c][2,7]naphthyridine core resulted in the identi-
fication of compounds with improved potency. These compounds
bind at the active site of PDK-1 kinase as determined by the X-
ray crystal structure of the protein–drug complex.
N
N
R5
N
N
O
O
O
O
a
NH2
NH2
N
N
32-44
30
Scheme 3. Reagents: (a) aniline, 2-dicyclohexylphosphino-20-(N,N-dimethyl-
amino)biphenyl, K3PO4, Pd2dba3, DMSO, 100 °C.
Table 4
PDK-1 potency of benzonaphthyridine compounds: optimization at R5 position
H
N
N
R5
N
O
NH2
O
N
Acknowledgments
R5
IC50 (nM)
a
Compounds
32
33
34
35
36
37
38
39
40
41
42
43
44
H
283
74
434
383
47
67
142
24
91
51
30
505
>1000
We thank Drs. Tarek Mansour and Robert Abraham for their
support of this work. We also thank the members of the Wyeth
Chemical Technologies group for analytical and spectral
determination.
2-OCH3
3-OCH3
4-OCH3
2-CH2OH
3-CH2OH
4-CH2OH
2-CH2CH2OH
4-CH2CH2OH
2-NH2
References and notes
1. (a) Mora, A.; Komander, D.; Van Aalten, D. M. F.; Alessi, D. R. Semin. Cell Dev.
Biol. 2004, 15, 161; (b) Storz, P.; Toker, A. Front. Biosci. 2002, 7, 886.
2. (a) Mitsiades, C. S.; Mitsiades, N.; Koutsilieris, M. Curr. Cancer Drug Target 2004,
4, 235; (b) Hanada, M.; Feng, J.; Hemmings, B. A. Biochim. Biophys. Acta 2004,
1697, 3.
2-F
2-NHPh
2-OCH2Ph
a
3. Sansal, I.; Sellers, W. R. J. Clin. Oncol. 2004, 22, 2954.
Values are averages of two or more experiments.
4. (a) Islam, I.; Bryant, J.; Chou, Y.; Kochanny, M. J.; Lee, W.; Phillips, G. B.; Yu, H.;
Adler, M.; Whitlow, M.; Ho, E.; Lentz, D.; Polokoff, M. A.; Subramanyam, B.; Wu,
J. M.; Zhu, D.; Feldman, R. I.; Arnaiz, D. O. Bioorg. Med. Chem. Lett. 2007, 17,
3814; (b) Islam, I.; Brown, G.; Bryant, J.; Hrvatin, P.; Kochanny, M. J.; Phillips, G.
B.; Yuan, S.; Adler, M.; Whitlow, M.; Lentz, D.; Polokoff, M. A.; Wu, J. M.; Shen,
J.; Walters, J.; Ho, E.; Subramanyam, B.; Zhu, D.; Feldman, R. I.; Arnaiz, D. O.
Bioorg. Med. Chem. Lett. 2007, 17, 3819.
5. Gopalsamy, A.; Shi, M.; Boschelli, D. H.; Williamson, R.; Olland, A.; Hu, Y.;
Krishnamurthy, G.; Han, X.; Arndt, K.; Guo, B. J. Med. Chem. 2007, 50, 5547.
6. Wissner, A.; Berger, D. M.; Boschelli, D. H.; Floyd, M. B., Jr.; Greenberger, L. M.;
Gruber, B. C.; Johnson, B. D.; Mamuya, N.; Nilakantan, R.; Reich, M. F.; Shen, R.;
Tsou, H.-R.; Upeslacis, E.; Wang, Y. F.; Wu, B.; Ye, F.; Zhang, N. J. Med. Chem.
2000, 43, 3244.
7. (a) Itoh, T.; Mase, T. Tetrahedron Lett. 2005, 46, 3573; (b) Billingsley, K. L.;
Anderson, K. W.; Buchwald, S. L. Angew. Chem., Int. Ed. 2006, 45, 3484.
8. Boschelli, D. H.; Wang, Y. D.; Johnson, S.; Wu, B.; Ye, F.; Sosa, A.; Golas, J. M.;
Boschelli, F. J. Med. Chem. 2004, 47, 1599.
9. Wissner, A.; Floyd, M. B., Jr.; Dushin, R.; Fraser, H. L.; Hu, Y.; Maderna, A.;
Nittoli, T.; Wang, Y. D. WO 2008109613, 2008; Chem. Abstr. 2008, 149, 355880.
10. Compounds 28 and 31 were prepared with imidazolides.
11. PDB code: 3H9O.
As shown in Table 3, various small R4 substituents were intro-
duced on the pyridyl group. Compared to 24, compound 28 with
an amino group on the 50-position showed slightly better activity
(26 nM) while compound 29 with a hydroxy group on the 50-posi-
tion showed a slight decrease in potency. Even the 50-bromo-
substituted compound (30) retained moderate activity. However,
compounds 31 and 16, which have substituents on positions other
than the 50-position, lost significant activity. Thus, the position of
substitution on the pyridyl ring does influence potency.
Based on the SAR results presented in Tables 2 and 3, the most
potent compound 28 was chosen for further exploration. A series of
compounds with various substituted aminoaryl groups at the 50-
position of the pyridine ring were synthesized using Buchwald
conditions (Scheme 3).12 As shown in Table 4, compound 32 with
a simple phenyl group was 10-fold less potent than compound
28. We also looked at some compounds with small polar substitu-
12. Charles, M. D.; Schultz, P.; Buchwald, S. L. Org. Lett. 2005, 7, 3965.