Structure-based design of 2,6,7-trisubstituted-7H-pyrrolo[2,3-d]pyrimidines as Aurora kinases inhibitors

Article abstract of DOI:10.1016/j.bmcl.2012.04.085

This Letter reports the optimization of a pyrrolopyrimidine series as dual inhibitors of Aurora A/B kinases. This series derived from a pyrazolopyrimidine series previously reported as inhibitors of aurora kinases and CDKs. In an effort to improve the selectivity of this chemotype, we switched to the pyrrolopyrimidine core which allowed functionalization on C-2. In addition, the modeling rationale was based on superimposing the structures of Aurora-A kinase and CDK2 which revealed enough differences leading to a path for selectivity improvement. The synthesis of the new series of pyrrolopyrimidine analogs relied on the development of a different route for the two key intermediates 7 and 19 which led to analogs with both tunable activity against CDK1 and maintained cell potency.

Full text of DOI:10.1016/j.bmcl.2012.04.085

Bioorganic & Medicinal Chemistry Letters 22 (2012) 4033–4037
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Structure-based design of 2,6,7-trisubstituted-7H-pyrrolo[2,3-d]pyrimidines
as Aurora kinases inhibitors
Jean-Yves Le Brazidec a, , Angela Pasis a, Betty Tam a, Christina Boykin a, Deping Wang b,
⇑
Douglas J. Marcotte b, Gisela Claassen a, Jer-Hong Chong a, Jianhua Chao a, Junhua Fan a, Khanh Nguyen a,
Laura Silvian a,b, Leona Ling b, Lin Zhang a, Michael Choi a, Min Teng a, Nuzhat Pathan a, Shuo Zhao a,
Tony Li a, Art Taveras b
a Biogen Idec, 5200 Research Place, San Diego, CA 92122, USA
b Biogen Idec, 12 Cambridge Center, Cambridge, MA 02142, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
This Letter reports the optimization of a pyrrolopyrimidine series as dual inhibitors of Aurora A/B kinases.
This series derived from a pyrazolopyrimidine series previously reported as inhibitors of aurora kinases
and CDKs. In an effort to improve the selectivity of this chemotype, we switched to the pyrrolopyrimidine
core which allowed functionalization on C-2. In addition, the modeling rationale was based on superim-
posing the structures of Aurora-A kinase and CDK2 which revealed enough differences leading to a path
for selectivity improvement. The synthesis of the new series of pyrrolopyrimidine analogs relied on the
development of a different route for the two key intermediates 7 and 19 which led to analogs with both
tunable activity against CDK1 and maintained cell potency.
Received 17 February 2012
Revised 16 April 2012
Accepted 17 April 2012
Available online 25 April 2012
Keywords:
Aurora kinase
Cyclin dependent kinase
Cell cycle arrest
Pyrrolopyrimidine
Cytotoxicity
Ó 2012 Elsevier Ltd. All rights reserved.
The Aurora proteins form a small family of serine/threonine ki-
nases which have become prominent targets for the modulation of
cell cycles. The three types of Aurora kinases expressed in humans
(Aurora-A, -B and -C) play a critical role during mitosis especially
in chromosome segregation and cytokinesis.1 Aurora-A and -B ki-
nases are overexpressed in human tumors including breast, lung,
colon, ovarian, and pancreatic cancers. Overexpressed early in
mitosis and localized to the centrosome and proximal spindles,
Aurora-A kinase phosphorylates a variety of proteins such as
TPX22 TACC3,3 and others. Inhibition of Aurora-A kinase in cancer
cells delays the mitotic entry and results in accumulation of cells in
the G2/M cell cycle phase. On the other hand, localized to the kine-
tochores during mitosis and to the midbody during cytokinesis,
Aurora-B kinase phosphorylates several proteins such as INCEMP4
(inner centromere protein), histone H3 and RacGAP. Inhibition of
Aurora-B kinase prevents the proper alignment of chromosomes
to the spindle plate, halts cytokinesis, and results in the formation
of multinucleated cells.
of CDKs.6 Comparison of crystal structures of Aurora A and CDK2
reveals differences between Aurora and CDK around the binding
pocket (Fig. 1). First, Aurora A has an amino acid insertion (G216)
compared to CDK2 in the hinge region, creating a bulge and giving
a more open conformation to the inter-domain loop. Second, the
gatekeeper, which is part of the back pocket, is Leu for Aurora A
and Phe for CDK2. In this study, the design efforts were based on
K162
Gatekeeper
L209(F80 CDK2)
As previously reported,5 the pyrazolopyrimidine analogs of type
I were non-selective inhibitors of CDKs, AKA and AKB. In order to
increase their therapeutic window, we envisaged to improve their
kinase selectivity profile with a focus on dialing out the inhibition
Hinge
⇑
Corresponding author.
E-mail address: jean@chempartner.com (J.-Y.Le Brazidec).
Figure. 1. Superimposed structure of Aurora A (pink, pdb code: 3R21) and CDK2
(purple, pdb code: 1OIT).
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2012.04.085

4034
J.-Y. Le Brazidec et al. / Bioorg. Med. Chem. Lett. 22 (2012) 4033–4037
treated with cyclopentylamine followed by a Sonogashira coupling
N
N
N
R2
with 3,3-diethoxy-propyne to give the compound 2 with moder-
ated overall yield. After cyclization of 2 with TBAF,8 3 reacted with
the aminopyrrole 4 to give 5 which, after hydrolysis of the ester,
unmasking of the aldehyde and coupling with the N-methylpiper-
azine, produced the key amidoaldehyde intermediate 7 (Scheme
1). This latter was derivatized to the final compounds 8–15 using
standard chemistry procedures (Schemes 2 and 3).
N
N
R1
HN
Ar
N
HN
Ar
N
R1
Pyrazolopyrimidine (I)
Pyrrolopyrimidine (II)
Figure. 2. Conversion of pyrazolopyrimidine to pyrrolopyrimidine core.
As mentioned in the introduction, the purpose of this study was
to optimize the potency against AKA and AKB while dialing out
activity against CDKs with a special focus on CDK1 by exploring
R groups (Table 1). Without any substituent on C2 (R = H), the com-
pound 16 exhibited moderate activity against CDK1 and subopti-
mum potency against AKA and AKB. Adding a CN or an oxime
group (13 or 14) increased the potency against the three kinases
by strengthening the interaction with the catalytic lysine suppos-
edly without adding steric repulsion. A primary amide and a pri-
mary or secondary alcohol in this position (15 and 8 or 12)
the previously reported crystal structure of pyrazolopyrimidine.
Switching from the pyrazolopyrimidine to the pyrrolopyrimidine
core (Fig. 2) allowed us to add a substituent at the C2 position.
The substituents can potentially interact with the catalytic lysine7
of aurora kinases for better potency, while reducing the potency for
the CDKs due to the differences mentioned above.
Our initial efforts were focused on the small substituents off C2
position. The synthesis of the first set of pyrrolopyrimidine analogs
started from the 5-bromo-2,4-dichloro-pyrimidine 1 which was
NH3
Cl
O
O
O
N
Br
O
O
O
N
4
N
N
c
a, b
N
Cl
N
NH
d
Cl
N
Cl
Cl
N
1
2
3
O
H
O
H
N
N
O
O
N
g
e, f
N
N
HN
N
HN
N
N
HN
N
N
N
N
N
HO
N
O
O
O
5
7
6
O
Scheme 1. Reagents and conditions: (a) Cyclopentylamine, DIEA, MeOH, rt, 80%; (b) 3,3-diethoxypropyne, CuI, dppfPdCl2, DIEA, DMF, rt, 41%; (c) TBAF, THF,
D, 66%; (d) 4,
Pd(OAc)2, Xantphos, Cs2CO3, dioxane, 80 °C, 64%; (e) HCl concn, dioxane, rt, 71%; (f) KOH 5N, MeOH, DMSO, rt, 82%; (g) N-methylpiperidine, HATU, DIEA, DMF, 50 °C, 58%.
OH
N
N
N
HN
N
N
N
O
12
OH
HN
N
N
e
N
N
HN
N
HN
N
a
d
7
N
N
N
N
N
c
N
N
b
O
O
11
8
OH
O
O
N
N
N
N
N
HN
N
HN
N
N
N
N
N
N
O
O
10
9
Scheme 2. Reagents and conditions: (a) MeMgBr, THF, À30 to 5 °C, 30%; (b) TosMIC, K2CO3, MeOH,
D, 52%; (c) NaClO2, NH3SO3, THF, water, rt, 13%; (d) MeNH2, NaCNBH4, rt,
20%; (e) NaBH4, MeOH, 60%.

J.-Y. Le Brazidec et al. / Bioorg. Med. Chem. Lett. 22 (2012) 4033–4037
4035
OH
N
N
H
N
N
N
N
HN
N
HN
N
a
c
b
7
N
N
N
N
N
N
O
O
13
14
O
N
N
NH2
N
HN
N
N
N
O
15
Scheme 3. Reagents and conditions: (a) NH2OH–HCl, NaOAc, EtOH,
D
, 68%; (b) CDI, DCM, 5 °C, 71%; (c) KOH 5N, MeOH, dioxane, rt, 50%.
Table 1
In vitro SAR of 2-substituted pyrrolopyrimidinesa
N
R
N
HN
N
N
N
N
O
Compound
R
AKA IC50
(
lM)
AKB IC50
(lM)
CDK1 IC50
(l
M)
G2/M arrest (
lM)
HCT116 IC50 (lM)
13
14
16
15
11
8
CHNOH
CN
H
CONH2
CH2NHCH3
CH(OH)Me
CH2OH
COOH
0.011
0.0075
0.038
0.028
0.04
0.011
0.02
0.35
0.03
0.004
0.018
0.011
0.06
0.03
0.04
1
0.003
0.006
0.24
0.5
0.7
1.1
0.002
0.065
0.3
0.1
0.3
0.1
0.1
10
0.004
0.04
0.29
0.07
0.21
0.08
0.13
12.3
12
10
1.3
10
a
The IC50s are determined based on a curve with 6 data points in duplicate. The G2/M arrest corresponds to the minimum concentration for a G2/M versus G1 peak ratio
over 2.
K162
K162
Acceptor
Link2
Link
K162
4.4 Å
A213
a
b
Figure. 4. (a) A pharmacophore for fragment search targeting K162; (b) one of the
hits with pyrazole moiety.
Finally, the negatively charge carboxylic acid (10) was not toler-
ated by any of the 3 kinases.
Figure. 3. Crystal structure of compound 15 (pdb code: 4DHF).
The binding mode of this class of pyrrolopyrimidine inhibitors
was elucidated by co-crystallization of Aurora-A with compound
15 (Fig. 3). The compound forms two hydrogen bonds with hinge
residue A213 and one hydrogen bond with R137. However, the
oxygen of amide off C-2 position is 4.4 Å away from Nzeta of
retained the high potency against AKA and AKB while increasing
the IC50s against CDK1 to the micromolar range. A positively
charge secondary amine (11) reduced slightly the potency against
AKA and AKB but had a more pronounced effect against CDK1.

4036
J.-Y. Le Brazidec et al. / Bioorg. Med. Chem. Lett. 22 (2012) 4033–4037
TMS
b
I
N
N
N
I
a
c
2
N
Cl
N
Cl
N
NH
Cl
N
NH
17
18
19
NH2
O
Ar
B
N
N
N
N
N
O
Ar
Ar
20a-f
22
N
O
N
Cl
N
HN
O
N
d
e
N
N
N
21a-f
23a-f
Scheme 4. Reagents and conditions: (a) Trimethylsilylacetylene, CuI, dppfPdCl2, DIEA, DMF, 40%; (b) NIS, AgNO3, DMF, 68%; (c) TBAF 1M in THF, THF, 75%; (d) 20, Pd(PPh3)4,
K2CO3 2M, DMF, 80 °C, 65%; (e) 22, TsOH, DMF, 140 °C, MW, 48%.
Table 2
In vitro SAR of 2-heteroaryl-pyrrolopyrimidinesa
N
Ar
N
HN
N
N
N
N
O
Compound
Ar
AKA IC50
0.002
(lM)
AKB IC50
0.004
(
lM)
CDK1 IC50
0.028
(lM)
G2/M arrest (
l
M)
HCT116 IC50
0.06
(lM)
O
N
9
0.03
N
23a
23b
23c
0.002
0.002
0.002
0.004
0.005
0.004
0.18
0.2
0.03
0.003
0.03
N
N
0.005
0.11
N
N
0.33
N
N
N
23d
0.002
0.004
1.4
0.1
0.11
23e
23f
0.005
0.002
0.004
0.004
1.7
2
1
0.59
0.09
N
N
N
N
0.1
Bn
a
The IC50s are determined based on a curve with 6 data points. The G2/M arrest corresponds to the minimum concentration for G2/M versus G1 peak ratio over 2.
K162. In order to design more potent compounds through form-
ing hydrogen bond with K162, a pharmacophore based ap-
proached was employed using MOE.9 As shown in Figure 4a,
the query pharmacophore consists of two link atoms off C2 posi-
tion and one hydrogen bond acceptor, which can hydrogen bond
with K162. One of the best hits is pyrazole (Fig. 4b). Thus, a set of
5-membered-ring heterocycle were proposed for synthesis.
For synthesizing the second set of pyrrolopyrimidine analogs,
the intermediate 2 underwent a Sonogashira coupling with tri-
methylsilylacetylene followed by the treatment with NIS in pres-
ence of silver nitrate to give the iodoalkyne 18. Cyclization of the
latter and chemoselective Suzuki coupling with arylboronates
20a–f led to 21a–f which were coupled with the aminopyrazole
22 to produce the final compounds 23a–f (Scheme 4).
The introduction of a small 5-membered heterocyclic ring on C2
maintained a strong potency against AKA and AKB independently
of its substitution pattern (Table 2). On the other hand, only the
oxazole analog 9 exhibited strong CDK1 potency. The other analogs
with methyl substituents (compounds 23a–c) had activities rang-
ing from 0.18 to 0.33 lM; increasing further the size of the substit-
uents (compounds 23d–f) decreased further the CDK1 potencies to
the micrometer range.

J.-Y. Le Brazidec et al. / Bioorg. Med. Chem. Lett. 22 (2012) 4033–4037
4037
For this series, the cell potencies correlate better with the inhi-
References and notes
bition of Aurora A/B kinases than with the inhibition of CDK1
explaining why dialing out CDK1 activity does not reduce the cell
potency dramatically.
In conclusion, this study demonstrated that the pyrazolopyrim-
idine to pyrrolopyrimidine core conversion gave rise to a new ser-
ies of dual AKA/AKB inhibitors with tunable potency against CDK1.
The weakest compounds against CDK1 remain to be screened
against a large panel of kinases to assess their overall kinase selec-
tivity profile.
1. Dar, A. A.; Goff, L. W.; Majid, S.; Berlin, J.; El-Rifai, w. Mol. Cancer Ther. 2010, 9, 268.
2. Bayliss, R.; Sardon, T.; Vernos, I.; Conti, E. Mol. Cell 2003, 12, 851.
3. Ulisse, S.; Baldini, E.; Toller, M.; Delcros, J. G.; Guého, A.; Curcio, F.; De Antoni, E.;
Giacomelli, L.; Ambesi-Impiombato, F. S.; Bocchini, S.; D’Armiento, M.; Arlot-
Bonnemains, Y. Endocr. Relat. Cancer 2007, 14, 827.
4. Sessa, F.; Mapelli, M.; Ciferri, C.; Tarricone, C.; Areces, L. B.; Schneider, T. R.;
Stukenberg, P. D.; Musacchio, A. Mol. Cell 2005, 18, 379.
5. Le Brazidec, J.-Y.; Pasis, A.; Tam, B.; Boykin, C.; Black, C.; Wang, D.; Claassen, G.;
Chong, J.-H.; Chao, J.; Fan, J.; Nguyen, K.; Silvian, L.; Ling, L.; Zhang, L.; Choi, M.;
Teng, M.; Pathan, N.; Zhao, S.; Li, T.; Taveras, A. Bioorg. Med. Chem. Lett. 2012, 22,
2070.
6. Malumbres, M.; Barbacid, M. Nat. Rev. Cancer 2009, 9, 153.
7. Cheetham, G. M. T.; Knegtel, R. M. A.; Coll, J. T.; Renwick, S. B.; Swenson, L.;
Weber, P.; Lippke, J. A.; Austen, D. A. J. Biol. Chem. 2002, 277(45), 42419.
8. Xiong, X.; Pirrung, M. C. Org. Lett. 2008, 10, 1151.
9. Molecular Operating Environment (MOE), 2010.10; Chemical Computing Group
Inc., 1010 Sherbooke St. West, Suite #910, Montreal, QC, Canada, H3A 2R7, 2010.
Acknowledgments
Special thanks to Dr. Adeela Kamal and Dr. Srinivas R. Kas-
ibhatla for their contributions to this program.