Discovery of novel 3,5-disubstituted indole derivatives as potent inhibitors of Pim-1, Pim-2, and Pim-3 protein kinases

Article abstract of DOI:10.1016/j.bmcl.2011.08.105

A series of novel 3,5-disubstituted indole derivatives as potent and selective inhibitors of all three members of the Pim kinase family is described. High throughput screen identified a pan-Pim kinase inhibitor with a promiscuous scaffold. Guided by struc

Full text of DOI:10.1016/j.bmcl.2011.08.105

Bioorganic & Medicinal Chemistry Letters 21 (2011) 6366–6369
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of novel 3,5-disubstituted indole derivatives as potent inhibitors of
Pim-1, Pim-2, and Pim-3 protein kinases
Gisele A. Nishiguchi a, , Gordana Atallah a, Cornelia Bellamacina a, Matthew T. Burger a, Yu Ding a,
⇑
Paul H. Feucht b, Pablo D. Garcia b, Wooseok Han a, Liana Klivansky a,b,c, Mika Lindvall a
a Global Discovery Chemistry/Oncology and Exploratory Chemistry, Novartis Institutes of BioMedical Research, Emeryville, CA 94608, USA
b Oncology Research, Novartis Institutes of BioMedical Research, Emeryville, CA 94608, USA
c The Molecular Foundry, Lawrence Berkeley National Laboratory, Berkeley, CA 94720, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of novel 3,5-disubstituted indole derivatives as potent and selective inhibitors of all three
members of the Pim kinase family is described. High throughput screen identified a pan-Pim kinase
inhibitor with a promiscuous scaffold. Guided by structure-based drug design, SAR of the series afforded
a highly selective indole chemotype that was further developed into a potent set of compounds against
Pim-1, 2, and 3 (Pim-1 and Pim-3: IC50 6 2 nM and Pim-2: IC50 6 100 nM).
Received 17 June 2011
Revised 19 August 2011
Accepted 25 August 2011
Available online 10 September 2011
Ó 2011 Elsevier Ltd. All rights reserved.
Keywords:
Pim-kinase inhibitors
Indole
Kinase selectivity
Oncology
Pim-1, 2, and 3 (Proviral Insertion Site in Moloney Murine Leuke-
mia Virus) are serine/threonine kinases with functions in the sur-
vival and proliferation of normal and cancer hematopoietic cells.1
In normal cells, cytokine signaling through the JAK/STAT pathway
leads to increase transcription of the Pim genes with subsequent
translation of constitutively active proteins.1b No further post-trans-
lational modifications are required for the activity of Pim kinases.1b
Indeed, the first crystal structure of Pim-1 kinase supports a model
where the unphosphorylated protein confers an active conforma-
tion.2 Thus, signaling downstream is primarily controlled at the
transcriptional/translational and protein turnover level.1b The Pim
kinases share a high level of sequence homology within the family
(>61%) and several proteins, such as: BAD, FOXO3a, p21/WAF1,
p27/KIP1 and SOCS1/3 have been identified as biological substrates
for the different isoforms.1b Studies have also shown some evidence
of functional redundancies among the three members of the family
in mice. Despite the structural and functional similarities among
Pims, triple knockout mice are viable and fertile.3 Taken together
the data indicates that the Pim family is important for growth factor
signaling but not essential for development.
implicated in the progression of hematopoetic, gastric and prostate
cancers,4 while over-expression of Pim-2 is associated with chronic
lymphocytic leukemia and non-Hodgkin lymphoma.5 In addition,
aberrant expression of Pim-3 has been seen in gastric adenoma
and hepatocellular carcinoma.6 Therefore, over-expression of Pim
in cancer, particularly hematopoietic malignancies is thought to
play a role in promoting survival and proliferation while suppress-
ing apoptosis7 and their inhibition may be an effective way of
treating cancers in which there is a Pim dependence.8 Not surpris-
ingly, Pim kinases have emerged as attractive therapeutic targets
and have elicited numerous groups to investigate and report novel
inhibitors of Pim(s), generally targeting one or two members of the
family.9 Herein, we describe our efforts towards finding potent
inhibitors of Pim kinases. Of special consideration was the design
of an inhibitor that would be potent in all three members of the
Pim kinase family, due to their overlapping implication in
hematopoietic cancers and potential functional redundancy.
Our internal high-throughput screen identified compound 1
(Fig. 1) as a potent pan-Pim inhibitor containing the 1H-pyrazol-
o[3,4-b]pyridine core and a phenol substitution at C6 of the scaffold.
The selection of this specific screening hit for further SAR analysis
was based on its attractive potency against all three Pim isoforms,
despite the fact that it showed low nanomolar activity against
several other kinases tested (Table 1) and contained a possible
metabolically-liable phenolic group.
Pim kinase expression is deregulated by several commonly
occurring oncogenic genetic alterations and it is thought that their
pro-survival and pro-proliferative activities contribute to cancer
progression and maintenance.1 Over-expression of Pim-1 has been
The X-ray co-crystal structure of a similar analogue (2) bound to
the ATP pocket of Pim-1 kinase (Fig. 2) was obtained in order to
guide the development of our structure–activity relationship
⇑
Corresponding author. Tel.: +1 510 923 2515; fax: +1 510 923 3360.
E-mail address: gisele.nishiguchi@novartis.com (G.A. Nishiguchi).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.08.105

G. A. Nishiguchi et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6366–6369
6367
H
N
favorable drug-like properties while maintaining or improving
upon the enzymatic potency of the compounds.
To initiate our studies toward attaining selectivity against other
kinases, systematic nitrogen deletion of the 1H-pyrazolo[3,4-b]pyr-
idine scaffold was pursued. Hence the corresponding 1H-pyrrol-
o[2,3-b]pyridine (3) and 1H-indole (4) scaffolds were synthe
sized12 while maintaining the substitutions at C3 and C6 constant.
The compounds were evaluated in biological assays against the tar-
get and the other kinases which 1 was not selective against (Table 1).
As shown, although the pyrrolo[2,3-b]pyridine core (3) lost potency
against all three Pims (3–7-fold), it gained remarkable selectivity
against Flt-3 (28-fold), c-Kit (158-fold) and kdr (264-fold). More
encouraging, the indole core compound(4) was able to retain the po-
N
Pim-1 IC50 = 0.009 µM
Pim-2 IC50 = 0.039 µM
Pim-3 IC50 = 0.012 µM
3
N
6
N
N
H
1
HO
Figure 1. High-throughput screening hit and pan-Pim potencies.
tency against the target while being completely inactive (>25 lM)
Table 1
Modifications of the core to attain selectivity (IC50
against the other three kinases tested. This finding is in agreement
with the original assumption that Pim inhibitors can take advantage
of the unique hinge of the protein to obtain selectivity against other
kinases and an inhibitor with a hydrogen-bond acceptor is not
essential for binding.
, lM)
R1
R2
=
HO
R1
=
N
NH
Y
N
R2
X
H
In parallel to the above approach, a survey of phenol replace-
ments was investigated on the 1H-pyrazolo[3,4-b]pyridine core.
For this purpose, intermediate 7 (Scheme 1) was synthesized with
the intention of rapidly constructing analogues substituted at the
6-position of the scaffold. The known literature procedure to 713
was followed, which commenced by treating 2,6-dichloropyridine
(5) with LDA, followed by addition of the aryl aldehyde. Upon oxida-
tion of the resulting alcohol with MnO2 to the diaryl-ketone 6, the
pyrazolo[3,4-b]pyridine core was formed via reaction with hydra-
zine in EtOH and THF. With substrate 7 in place, Suzuki reactions
were performed in the microwave in the presence of the desired
boronic ester (or boronic acid) followed by treatment with TFA in
DCM to remove the protecting group.
As depicted in Table 2, it soon became apparent that a phenol
replacement was a particularly challenging effort. Upon replacing
the phenol with phenyl (9), a 200-fold potency loss was observed
for Pim-1 and -2. The 4-methoxy phenyl analogue (10) was inac-
tive in Pim-2 and 500–1000 fold less potent in Pim-1 and Pim-3.
Maintaining only a hydrogen-bond acceptor as in the case of pyri-
dine (11) also led to significant decrease in activity. Substitutions
containing both a hydrogen-bond donor and acceptor (12, 13 and
14) were not tolerated. The indole (15) substitution was moder-
ately effective (Pim-1 IC50 = 110 nM), but none of these replace-
ments were as potent as the phenol. These results indicate that
the network of hydrogen bond interactions provided by the phenol
is an essential part of the pharmacophore and any disturbance of
Cmpd X, Y
Pim-1 Pim-2 Pim-3 Flt-3
c-Kit
PDGFR Kdr
1
3
4
N, N
N, CH
CH, CH 0.004
0.009
0.03
0.039
0.283
0.042
0.012
0.058
0.022
0.004 0.007 0.006
0.111 1.107 0.987
>25
0.018
4.76
>25
>25
nd
studies.10 The structure revealed that the N–H of the 1H-pyrazol-
o[3,4-b]pyridine core makes hydrogen bond to the hinge
a
(2.7 Å). The two other nitrogens on the scaffold do not appear to
be making any direct interaction with the protein. It has been
well-documented that while most kinases contain a conserved
hydrogen bond donor and acceptor pair at the hinge, Pim-1 con-
tains the amino-cyclized residue proline (Pro123) in place of the
typical hydrogen bond donor residue.11 It was anticipated that
we could take advantage of this unique structural feature to mod-
ulate the selectivity of our compounds against other kinases. The
X-ray structure also revealed that the hydroxyl group of the phenol
acts as a hydrogen bond donor to Glu89 (2.4 Å) and a hydrogen
bond acceptor from the backbone amide of Phe187 (3.0 Å).
The strategy towards optimization of the screening hit focused
on two main areas: (1) address the kinase selectivity associated
with the scaffold by taking advantage of the unique hinge of
Pim-1 protein, and: (2) investigate groups capable of replacing
the phenolic moiety with the intention of developing a series with
3
N
6
O
N
N
H
2
HO
Figure 2. X-ray co-crystal structure of 2 bound to Pim-1 enzyme.

6368
G. A. Nishiguchi et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6366–6369
Table 3
Indole C5 SAR (IC50, lM)
O
R1
i
ii
R1
N
R1
R2
N
H
Cl
N
5
Cl
Cl
N
Cl
N
6
Cl
R1=
N
NH
7
N
H
R1
iii
Cmpd
R2
Pim-1
Pim-2
Pim-3
R1
=
N
NBoc
N
N
H
23
24
25
26
27
5-Pyrimidine
3-Pyridine
4-Pyridine
2-Pyridine
2-Pyrazine
2.6
0.192
2.8
2.1
0.53
>25
15
>25
>25
13.6
3.3
0.49
5.7
14
0.96
R2
N
8
Scheme 1. Reagents and conditions: (i) (a) LDA, THF, À78 °C, then R1CHO; (b)
MnO2, toluene, reflux (51% two steps); (ii) hydrazine, EtOH/THF, Hunig’s base, 70 °C
(72%); (iii) (a) Pd(dppf)Cl2ÁDCM, R2B(OH)2, DME/2M aq Na2CO3, 120 °C, microwave;
(b) TFA/DCM (10–44% two steps).
R1
I
R1
R2
Table 2
Br
i
Br
ii
SAR of the 1H-pyrazolo[3,4-b]pyridine core 6-position (IC50, lM)
N
N
R1
N
H
Boc
Boc
18 (23-26)
N
NH
R1=
16
17
N
N
H
R2
N
R3
iii
Cmpd
R2
Pim-1
0.009
Pim-2
0.039
Pim-3
0.012
Y
O
B
R1
R1
N
iv
O
1
N
N
HO
Boc
H
19
20 (27-37)
9
1.7
4.9
2.5
15
7.2
7.4
12.8
5.8
Cl
N
R3
Cl
N
N
Cl
vi
10
11
12
>25
>25
>25
N
22
O
N
21
Scheme 2. Reagents and conditions: (i) R1-boronic ester, Pd(dppf)Cl2-DCM, DMF/
2M aq Na2CO3, 90 °C (75%); (ii) (a) R2B(OH)2, Pd(dppf)Cl2ÁDCM, DME/2M aq Na2CO3,
120 °C; (b) TFA/DCM (24–34% two steps); (iii) bis(pinacolato)diboron,
Pd(dppf)Cl2ÁDCM, KOAc, dioxane, 100 °C (91%); (iv) (a) Pd(dppf)Cl2ÁDCM, 22,
DME/2M aq Na2CO3, 100 °C; (b) TFA/DCM (6–44% two steps), (vi) NaH, aminoal-
cohol, THF (56–95%) or amine, DIEA, EtOH, 80 °C (>95%).
12
H2N
N
N
13
0.17
20
1.8
NH2
To assess the potential of this strategy, heteroaromatics were se-
lected as the C5 substituents and they were placed on the more ki-
nase selective indole scaffold (Table 3). These groups could
provide H-bond acceptors to interact with the Lys residue as well
as offer vectors to target the acidic residue later. Such compounds
were synthesized according to Scheme 2. From commercially avail-
able tert-butyl 5-bromo-3-iodo-1H-indole-1-carboxylate (16), pal-
ladium-catalyzed coupling reaction conditions at 90 °C allowed for
the selective introduction of a group at the C3 position (17). Subse-
quent Suzuki reaction with the corresponding boronic ester or boro-
nic acid at 120 °C or transformation of intermediate 17 to the
corresponding boronic ester (19) followed by Suzuki coupling reac-
tion, and BOC deprotection afforded a representative subset of com-
pounds depicted in Table 3. From this series, the 2-pyrazine (27) and
3-pyridine (24) analogues were encouraging, whereas other pyri-
dine isomers (25 and 26) and pyrimidine (23) were at least 10-fold
less active. The potency of compound 27 could be improved (twofold
in Pim-1) by replacing the C3-phenyl piperazine with the 2-fluoroa-
minopyridine moiety (28, Table 4), a group identified from our C3
SAR studies.16 Hence, compound 28 became our starting point to
investigate possible interactions with the nearby acidic residue with
the expectation to further improve the enzymatic activity.
H2N
14
15
11
>25
10
S
O
O
0.11
0.65
0.26
HN
this network is detrimental to the activity of the compound. A
similar SAR trend was observed when the phenol replacements
were placed on the more kinase selective indole core.14
The lack of success with the replacement of the phenolic motif
led us to reconsider our initial strategy and as a result, efforts
moved toward investigation of alternative substitution patterns
on the scaffold. Based on the X-ray structure of 2 (Fig. 2), the 5-po-
sition of the azaindazole core points towards Lys67 and a favorable
interaction could be attained with a hydrogen-bond acceptor sub-
stituent. Several X-ray crystal structures of inhibitors bound to
Pim-1 protein have been published that effectively take advantage
of this interaction.15 In addition, the acidic residue Asp186 is with-
in reasonable distance (3.8 Å) to Lys67 so that contacts with a
hydrogen bond donor group could also be explored, to create a net-
work of hydrogen bonds.
Towards this end, analogues were designed that contained either
rigid or flexible basic amine substitutions at C6 of the pyrazine. The
synthesis of these compounds first required reaction of a nucleophile

G. A. Nishiguchi et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6366–6369
6369
Table 4
maintained the same binding mode as compound 2 (unpublished
results).
Inhibitory activities against Pim-1, 2 and 3 (IC50
,
lM)
In summary, we describe a novel class of indole-based mole-
cules as potent and selective inhibitors against the family of Pim
kinases. Utilizing structure-based drug design we were able to take
advantage of unique Pim protein structural features and transition
from a promiscuous pyrazolo[3,4-b]pyridine HTS screening hit (1)
to the identification of more kinase selective 4-(3-aryl-1H-indol-6-
yl)phenol (4) scaffold. This scaffold further evolved to generate 3,5-
disubstituted indole compounds with high affinity and lacking the
potentially metabolic-liable phenolic moiety. An attractive feature
of the series is the low nanomolar Pim-2 enzymatic inhibition, a
challenging undertaking for this family of enzymes.
NH2
N
R3
N
N
F
N
H
Cmpd
R3
H
Pim-1
0.25
Pim-2
5.0
Pim-3
0.31
28
O
29
30
<0.002
0.083
0.005
Supplementary data
NH2
H
N
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2011.08.105.
<0.002
0.061
0.004
N
References and notes
NH2
31
<0.002
0.081
0.003
1. (a) Nawijin, M. C.; Alendar, A.; Berns, A. Nat. Rev. Cancer 2011, 11, 23; (b) Brault,
L.; Gasser, C.; Bracher, F.; Huber, K.; Knapp, S.; Schwaller, J. Haematologica 2010,
95, 1004.
2. (a) Qian, K. C.; Wang, L.; Hickey, E. R.; Studts, J.; Barringer, K.; Peng, C.;
Kronkaitis, A.; Li, J.; White, A.; Mische, S.; Farmer, B. J. Biol. Chem. 2005, 280,
6130; (b) Bullock, A. N.; Debreczeni, J.; Amos, A. L.; Knapp, S.; Turk, B. E. J. Biol.
Chem. 2005, 280, 41675.
N
O
32
33
<0.002
<0.002
0.056
0.026
<0.002
0.005
NH2
Et
3. Mikkers, H.; Nawijn, M.; Allen, J.; Brouwers, C.; Verhoeven, E.; Jonkers, J.; Berns,
A. Mol. Cell. Biol. 2004, 24, 6104.
O
O
NH2
4. (a) Amson, R.; Sigaux, F.; Przedborski, S.; Flandrin, G.; Givol, D.; Telerman, A.
Proc. Natl. Acad. Sci. U.S.A. 1989, 86, 8857; (b) Warnecke-Eberz, U.;
Bollschweiler, E.; Drebber, U.; Metzger, R.; Baldus, S.; Holscher, A. H.; Monig,
S. Anticancer Res. 2009, 29(11), 4451; (c) Mumenthaler, S. M.; Ng, P. Y.; Hodge,
A.; Bearss, D.; Berk, G.; Kanekal, S.; Redkar, S.; Taverna, P.; Agus, D. B.; Jain, A.
Mol. Cancer Ther. 2009, 8(10), 2882.
5. (a) Cohen, A. M.; Grinblat, B.; Bessler, H.; Kristt, D. A.; Kremer, A.; Shalom, S.;
Schwartz, A.; Halperin, M.; Merkel, D.; Don, J. Leuk. Lymphoma 2004, 45, 951;
(b) Breuer, M. L.; Cuypers, H. T.; Berns, A. EMBO J. 1989, 8, 743.
6. (a) Zheng, H. C.; Tsuneyama, K.; Takahashi, H.; Miwa, S.; Sugiyama, T.;
Popivanova, B. K.; Fujii, C.; Nomoto, K.; Mukaida, N.; Takano, Y. J. Cancer Res.
Clin. Oncol. 2008, 134, 481; (b) Li, Y.; Popivanova, B. K.; Nagai, Y.; Ishikura, H.;
Fujii, C.; Mukaida, N. Cancer Res. 2006, 66, 6741.
34
35
<0.002
<0.002
0.072
0.122
0.006
0.010
NH2
O
O
O
NH2
Ph
Ph
36
37
<0.002
<0.002
0.010
0.020
<0.002
0.005
NH2
NH2
7. (a) Morishita, D.; Katayama, R.; Sekimizu, K.; Tsuruo, T.; Fujita, N. Cancer Res.
2008, 68, 5076; (b) Gong, J.; Wang, J.; Ren, K.; Liu, C.; Li, B.; Shi, Y. J. Surg. Res.
2009, 153, 17.
8. Amaravadi, R.; Thompson, C. B. J. Clin. Invest. 2005, 115, 2618.
9. Morwick, T. Expert Opin. Ther. Patents 2010, 20, 193. and references therein.
10. The Pim1 complexed structure of 2.6 Ang and 99% complete has been
deposited into the RSCB as PBD ID 3T9I. The final R and R-Free values were
0.165 and 0.227, respectively. Additional information can be found in the
header file of the submitted pdb file.
11. (a) Jacobs, M. D.; Black, J.; Futer, O.; Swenson, L.; Hare, B.; Fleming, M.; Saxena,
K. J. Biol. Chem. 2005, 280, 13728; (b) Qian, K. C.; Wang, L.; Hickey, E. R.; Studts,
J.; Barringer, K.; Peng, C.; Kronkaitis, A.; Li, J.; White, A.; Mische, S.; Farmer, B. J.
Biol. Chem. 2005, 280, 6130; (c) Kumar, A.; Mandiyan, V.; Suzuki, Y.; Zhang, C.;
Rice, J.; Tsai, J.; Artis, D. R.; Ibrahim, P.; Bremer, R. J. Mol. Biol. 2005, 348, 183.
12. See Supplementary data.
with 2,6-dichloropyrazine to give 22, followed by Suzuki coupling of
this intermediate with 19 and protecting group removal (Scheme 2).
From this investigation, cyclic analogues (Table 4, entries 29, 30
and 31) were found to be very potent in Pim-1 and 3, reaching the
limits of detection of the biochemical assay and moderately potent
in Pim-2 (50-fold improvement over 28). Upon introduction of a
flexible side-chain, compound 32 displayed similar potency to
the cyclic analogues, and it led to a survey of the size of the hydro-
phobic group adjacent to the basic amine. We observed a twofold
improvement (Pim-2) for the ethyl group (33), however as the sub-
stitution got larger as in the case of the iso-propyl and t-butyl
groups (34 and 35), the Pim-2 potency tended to suffer. Notably,
the phenyl moiety (36 and 37) was well tolerated, similar in po-
tency as the ethyl (33) and with a slight preference for the S-enan-
tiomer. To assess the kinase selectivity of these compounds,
inhibitor 35 was further profiled in a panel of 30 kinases and it
13. Arora, N.; Billedeau, R. J.; Dewdney, N. J.; Gabriel, T.; Goldstein, D. M.; Soth, M.;
Trejo-Martin, T. A. WO 2005 085248A1.
14. Results not shown.
15. (a) Tao, Z.; Hasvold, L. A.; Leverson, J. D.; Han, E. K.; Guan, R.; Johnson, E. F.;
Stoll, V. S.; Steward, K. D.; Stamper, G.; Soni, N.; Bouska, J. J.; Luo, Y.; Sowin, T.
J.; Lin, N.; Giranda, V. S.; Rosenberg, S. H.; Penning, T. D. J. Med. Chem. 2009, 52,
6621; (b) Grey, R.; Pierce, A. C.; Bemis, G. W.; Jacobs, M. D.; Moody, C. S.; Jajoo,
R.; Mohal, N.; Green, J. Bioorg. Med. Chem. Lett. 2009, 19, 3019; (c) Qian, K.;
Wang, L.; Cywin, C. L.; Farmer, B. T.; Hickey, E.; Homon, C.; Jakes, S.; Kashem, M.
A.; Lee, G.; Leonard, S.; Li, J.; Magboo, R.; Mao, W.; Pack, E.; Peng, C.;
Prokopowicz, A.; Welzel, M.; Wolak, J.; Morwick, T. J. Med. Chem. 2009, 52,
1814; (d) Tong, Y.; Stewart, K. D.; Thomas, S.; Przytulinska, M.; Johnson, E. F.;
Klinghofer, V.; Leverson, J.; McCall, O.; Soni, N. B.; Luo, Y.; Lin, N.; Sowin, T. J.;
Giranda, V. L.; Penning, T. D. Bioorg. Med. Chem. Lett. 2008, 18, 5206; (e) Pierce,
A. C.; Jacobs, M.; Stuver-Moody, C. J. Med. Chem. 2008, 51, 1972.
16. Studies to be reported in a separate communication.
was found to display >10 lM potency against all 30 kinases. This
data supports the identification of the indole core as a selective
scaffold for this family of proteins. It is also noteworthy to mention
that in-house X-ray crystal structures of compounds 3, 4, and 34 all