Discovery of EGFR selective 4,6-disubstituted pyrimidines from a combinatorial kinase-directed heterocycle library

Article abstract of DOI:10.1021/ja0567485

The epidermal growth factor receptor (EGFR) tyrosine kinase was one of the first receptor tyrosine kinases to be targeted for drug development by the pharmaceutical industry due to its ubiquitous overexpression in a variety of tumors. Despite the validati

Full text of DOI:10.1021/ja0567485

Published on Web 01/31/2006
Discovery of EGFR Selective 4,6-Disubstituted Pyrimidines from a
Combinatorial Kinase-Directed Heterocycle Library
Qiong Zhang,† Yi Liu,† Feng Gao,† Qiang Ding,† Charles Cho,‡ Wooyoung Hur,‡ Yunho Jin,†
Tetsuo Uno,† Claudio A. P. Joazeiro,† and Nathanael Gray*,†
Departments of Chemistry and Cell Biology, Genomics Institute of the NoVartis Research Foundation, 10675 John
Jay Hopkins DriVe, San Diego, California 92121, and Department of Chemistry, The Scripps Research Institute,
10550 North Torrey Pines Road, La Jolla, California 92037
Received October 2, 2005; E-mail: gray@gnf.org
High-throughput screening of small molecule gene-family tar-
geted libraries has been the most efficient method to discover new
lead compounds from combinatorial libraries.1 We recently de-
scribed an efficient method to prepare libraries of kinase-directed
heterocycles.2 Our approach consists of the synthesis of libraries
where the heterocyclic core is considered as a combinatorial input,
thereby allowing the parallel synthesis of multiple kinase-privileged
scaffolds (2,4- and 4,6-pyrimidines, triazines, purines, quinazolines,
pyrazines, etc.) in a single combinatorial synthesis. These libraries
were plated in a 1536-well format and screened in a variety of
cellular and biochemical assays using a robotic HTS platform. Here
we wish to describe the identification of a class of 4,6-disubstituted
Figure 1. EGFR enzyme IC50 values for lead compound 1 and its
analogues: (A) 4,6-pyrimidine scaffold, (B) R6 fixed, (C) R4 fixed, (D)
additional inhibitors.
pyrimidines (represented by compound 1) that are highly selective
inhibitors of the epidermal growth factor receptor (Her-1, also erbB-
1, EGFR). Inhibitor 1 possesses an enzymatic IC50 ) 21 nM against
on the substitution pattern on the two pendant aniline rings. In
contrast, the 4,6-disubstituted pyrimidines represent a significantly
less explored class of kinase inhibitors because they are generally
not as potent as their corresponding 2,4-regioisomers.8 This may
be due to a greater energetic penalty required to attain the s-cis
conformation required for the hinge region interaction.
EGFR kinase in vitro and blocks receptor autophosphorylation in
cells. Inhibitor 1 was serendipitously discovered using a cell-based
reporter gene assay (RGA) for modulators of protein stability (to
be described elsewhere). Structure-activity relationship (SAR) with
respect to inhibition of EGFR kinase activity allowed the essential
pharmacophore to be identified, a binding mode to EGFR proposed,
and the high degree of kinase selectivity rationalized.
Despite the low molecular weight of inhibitor 1, it exhibited
exclusive selectivity against EGFR when tested against a panel of
55 recombinant kinases at a concentration of 10 µM (data in
Supporting Information). Inhibitor 1 also potently inhibits two
EGFR mutants associated with clinical response to gefitinib: L858R
(IC50 ) 63 nM) and L861Q (IC50 ) 4 nM), but displays much
weaker activity against Her 4 (IC50 ) 7640 nM). To understand
the SAR with respect to EGFR kinase inhibition, several derivatives
based on the structure of compound 1 were made (Figure 1).
Removal of the trifluoromethyl from the 4-aniline and introduction
of an ortho methyl (2) or chloro (3) substituent resulted in
approximately equipotent compounds. The NH at the 4-position
was not essential and could be replaced with an N-methyl (4) or
oxygen (5), suggesting that this group is not involved in an
H-bonding interaction with the enzyme. Addition of a methylene
group to form the benzylamino compound 6 resulted in an 88-fold
loss of activity. Removal of the aniline to form carbon-linked
bicyclic compounds 7 and 8 resulted in a complete loss of activity.
Replacement of the 4,6-pyrimidine core with the two possible
regioisomeric 2,4-pyrimidines (9, 10) or with a pyridine (16) also
resulted in loss of EGFR inhibitory activity, suggesting that the
correct positioning of both pyrimidine nitrogens is essential (Figure
1). A hydrogen bond donor was essential at the 6-position as the
N-methylated compound was completely inactive (11). Introduction
of a 3-trifluoromethylaniline (12) resulted in a compound equipotent
to 1, while introduction of a methoxy (13) or a larger cyclohexyl-
amide (14) resulted in a significant loss of activity.
EGFR was one of the first receptor tyrosine kinases to be targeted
for inhibitor development by the pharmaceutical industry due to
its overexpression in a variety of tumors.3 This research has
culminated in the recent approval of two highly related anilino-
quinazolines, gefitinib (Iressa) and erlotinib (Tarceva), which target
the intracellular tyrosine kinase domain, as well a chimeric
monoclonal antibody, erbitux, which targets the extracellular portion
of the receptor. In addition to the anilinoquinazolines, a variety of
EGFR small molecule inhibitors from other scaffold classes are
undergoing clinical evaluation.4 Curiously, despite years of intensive
research on EGFR inhibitors, there is a surprising dearth of
chemically distinct small molecule inhibitors with a high degree
of selectivity. There is also a need for new scaffolds due to the
recent finding of EGFR mutations which render the kinase resistant
to gefitinib and erlotinib.5
Pyrimidines substituted at the 2 and 4 positions are an extremely
well studied class of inhibitors for diverse kinases, including Cdks,
p38, Aurora, KDR, and Gsk3.6 Crystallography has revealed that
2,4-dianilinopyrimidines typically form a bidentate hydrogen bond-
ing interaction with the hinge amino acid using the pyrimidine N1
and the aniline NH at the C2 position.7 Despite their fairly generic
mode of recognizing the kinase active site, pyrimidines are capable
of displaying a wide range of kinase selectivity profiles depending
† Genomics Institute of the Novartis Research Foundation.
‡ The Scripps Research Institute.
9
2182
J. AM. CHEM. SOC. 2006, 128, 2182-2183
10.1021/ja0567485 CCC: $33.50 © 2006 American Chemical Society

C O M M U N I C A T I O N S
Figure 3. EGF-induced autophosphorylation of wild-type (WT) EGFR,
but not the T766M mutant, is sensitive to inhibition by compound 1 (10
µM) and gefitinib (1 µM, GEF). MT, mock transfected; NT, not treated
with inhibitor.
Figure 2. Chemical structure version of docking model of 1 bound to the
ATP site of the EGFR kinase domain (left). Superimposition of model of
1 (gold) with erlotinib (green) in the ATP site of EGFR kinase domain
(right).
mutation of its gatekeeper Thr residue to Met. To investigate
whether the 4,6-substitution pattern was a key component of kinase
selectivity, both 2,4-pyrimidine regioisomers (9, 10) were tested
against a panel of 55 kinases at a concentration of 10 µM. As
expected, both 2,4-pyrimidine inhibitors were considerable less
selective with compound 9 inhibiting Aurora A (IC50 ) 931 nM)
and compound 10 inhibiting Aurora A (IC50 ) 42 nM), Bmx (IC50
) 386 nM), Btk (IC50 ) 3550 nM), Lck (IC50 ) 131 nM), IGF-1R
(IC50 ) 591 nM), cSrc (IC50 ) 1980 nM), TrkB (IC50 ) 2510 nM),
and Syk (IC50 ) 887 nM). Neither inhibitor 9 nor 10 inhibited
EGFR significantly at a concentration of 10 µM.
In conclusion, we have demonstrated that screening a combina-
torial library based on a privileged class of 4,6-dianilinopyrimidines
allowed the efficient identification of potent and highly selective
inhibitors of both enzymatic and cellular EGFR kinase activity.
Docking, SAR, and mutagenesis studies suggest that a key
H-bonding interaction is required to the gatekeeper residue T766.
In an effort to rationalize the high degree of kinase selectivity
exhibited by the 4,6-pyrimidines against the EGFR kinase family,
we employed the SAR data in conjunction with comparisons to
known EGFR co-crystal structures to propose a binding mode
(Figure 2). We propose that the pyrimidine N1 and 6-NH form a
pair of hydrogen bonds with the “hinge” amino acid (Met 769).
The corresponding hydrogen bonding interaction is formed by the
quinazoline N1 of erlotinib.9 The trifluoromethyl phenyl is proposed
to be situated in a hydrophobic binding pocket which is also
occupied by the ethynyl-substituted aniline group of erlotinib. The
pyrimidine N3 appears to mimic the quinazoline N3 by forming a
hydrogen bonding interaction to the side chain hydroxyl of the
“gatekeeper” threonine 766. The pyrimidine 6-anilino substitution
partially overlaps the site occupied by the quinazoline phenyl, and
the amide substitution is directed toward solvent. The proposed
binding mode is fully consistent with the observed SAR: the
requirement for correct positioning of the pyrimidine N1, N3, and
6-NH, the tolerance for substitution of the 6-NH, and the ability to
tolerate a variety of the substitutions to the 4-aniline.
Supporting Information Available: Experimental details and
characterization data of all the reported compounds. This material is
available free of charge via the Internet at http://pubs.acs.org.
References
We hypothesized that the selectivity of compound 1 is derived
from the ability to form three hydrogen bonding interactions while
occupying a hydrophobic cavity made accessible due to the small
gatekeeper threonine 766. To test this idea and the proposed binding
mode, we replaced the gatekeeper threonine with a methionine since
primary sequence alignment revealed that Met is normally present
in several kinases in the equivalent position of EGFR T766, such
as in Jak, Syk, Fak, and Csk. Moreover, Csk and Syk are resistant
to inhibition by compound 1. Thus, the T766M substitution was
anticipated to cause resistance to both compound 1 and gefitinib
due to steric blockage of the hydrophobic binding pocket and
removal of a key hydrogen bond interaction without affecting kinase
activity. Indeed, a recent report shows that T766M mutation to
EGFR induces resistance to gefitinib.10 Treatment of U-2OS cells
transfected with either WT or T766M EGFR with epidermal growth
factor (EGF) resulted in receptor autophosphorylation (Figure 3,
lanes 2 and 5). Pretreatment of cells with either compound 1 (10
µM) or gefitinib (1 µM) resulted in complete inhibition of WT
receptor autophosphorylation (lanes 3 and 4). The ability of both
the compounds to inactivate WT EGFR is also evident from the
observation that greater levels of receptor were immunoprecipitated
from compound-treated compared to untreated cells, consistent with
reports that EGFR activation leads to its rapid degradation.11 As
expected, the T766M mutant receptor was completely resistant to
inhibition by either compound (lanes 6 and 7). Thus, we were able
to successfully design an inhibitor-resistant allele of EGFR by
(1) Bredel, M.; Jacoby, E. Nat. ReV. Genet. 2004, 5, 262-275.
(2) Ding, S.; Gray, N. S.; Wu, X.; Ding, Q.; Schultz, P. J. Am. Chem. Soc.
2002, 124, 1594-1596.
(3) Hynes, N.; Lane, H. Nat. ReV. Can. 2005, 5, 341-354.
(4) El-Rayes, B. F.; LoRusso, P. M. Brit. J. Can. 2004, 91, 418-424.
(5) Kobayashi, S.; Boggon, T. J.; Dayaram, T.; Ja¨nne, P. A.; Kocher, O.;
Meyerson, M.; Johnson, B. E.; Eck, M. J.; Tenen, D. G.; Halmos, B. N.
Engl. J. Med. 2005, 352, 786-792.
(6) (a) Anderson, M.; Beattie, J. F.; Breault, G. A.; Breed, J.; Byth, K. F.;
Culshaw, J. D.; Ellston, R. P.; Green, S.; Minshull, C. A.; Norman, R.
A.; Pauptit, R. A.; Stanway J.; Thomas A. P.; Jewsbury P. J. Bioorg.
Med. Chem. Lett. 2003, 13, 3021-3026. (b) Cumming, J. G.; McKenzie,
C. L.; Bowden, S. G.; Campbell, D.; Masters, D. J.; Breed, J.; Jewsbury,
P. J. Bioorg. Med. Chem. Lett. 2004, 14, 5389-5394. (c) Tavares, F. X.;
Boucheron, J. A.; Dickerson, S. H.; Griffin, R. J.; Preugschat, F.; Thomson,
S. A.; Wang, T. Y.; Zhou, H.-Q. J. Med. Chem. 2004, 47, 4716-4730.
(7) Wang, S.; Meades, C.; Wood, G.; Osnowski, A.; Anderson, A.; Yuill,
R.; Thomas, M.; Mezna, M.; Jackson, W.; Midgley, C.; Griffiths, G.;
Fleming, I.; Green, S.; McNae, I.; Wu, S.-Y.; McInnes, C.; Zheleva, D.;
Walkinshaw, M. D.; Fischer, P. M. J. Med. Chem. 2004, 47, 1662-1675.
(8) (a) Beattie, J.; Breault, G.; Ellston, R.; Green, S.; Jewsbury, P.; Midgley,
C.; Naven, R.; Minshull, C.; Pauptit, R.; Tucker, J.; Pease, J. Bioorg.
Med. Chem. Lett. 2003, 13, 2955-2960. (b) Thomas, A. 1995, WO 95/
15952.
(9) Stamos, J.; Sliwkowski, M. X.; Eigenbrot, C. J. Biol. Chem. 2002, 277,
46265-46272.
(10) Carter, T. A.; Wodicka, L. M.; Shah, N. P.; Velasco, A. M.; Fabian, M.
A.; Treiber, D. K.; Milanov, Z. V.; Atteridge, C. E.; Biggs, W. H., III;
Edeen, P. T.; Floyd, M.; Ford, J. M.; Grotzfeld, R. M.; Herrgard, S.; Insko,
D. E.; Mehta, S. A.; Patel, H. K.; Pao, W.; Sawyers, C. L.; Varmus, H.;
Zarrinkar, P. P.; Lockhart, D. J. Proc. Natl. Acad. Sci. U.S.A. 2005, 102,
11011-11016.
(11) Levkowitz, G.; Waterman, H.; Zamir, E.; Kam, Z.; Oved, S.; Langdon,
Y.; Beguinot, L.; Geiger, B.; Yarden, Y. Genes DeV. 1998, 12, 3663-3674.
JA0567485
9
J. AM. CHEM. SOC. VOL. 128, NO. 7, 2006 2183