Inhibitors of epidermal growth factor receptor tyrosine kinase: Novel C-5 substituted anilinoquinazolines designed to target the ribose pocket

Article abstract of DOI:10.1016/j.bmcl.2005.12.028

A series of novel C-5 substituted anilinoquinazolines, selected on the basis of docking experiments and overlays with ATP in the active site of EGFR tyrosine kinase, have been prepared and found to be potent inhibitors. In vivo pharmacokinetics and diseas

Full text of DOI:10.1016/j.bmcl.2005.12.028

Bioorganic & Medicinal Chemistry Letters 16 (2006) 1633–1637
Inhibitors of epidermal growth factor receptor tyrosine
kinase: Novel C-5 substituted anilinoquinazolines designed
to target the ribose pocket
Peter Ballard,a Robert H. Bradbury,a Craig S. Harris,b Laurent F. A. Hennequin,b
Mark Hickinson,a Paul D. Johnson,a Jason G. Kettle,a,* Teresa Klinowska,a
Andrew G. Leach,a Remy Morgentin,b Martin Pass,a Donald J. Ogilvie,a Annie Olivier,b
Nicolas Warinb and Emma J. Williamsa
aAstraZeneca, Mereside, Alderley Park, Macclesfield, Cheshire, SK10 4TG, UK
bAstraZeneca, Centre de Recherches, Z.I. la Pompelle, BP1050, 51689 Reims Cedex 2, France
Received 23 November 2005; revised 2 December 2005; accepted 6 December 2005
Available online 27 December 2005
Abstract—A series of novel C-5 substituted anilinoquinazolines, selected on the basis of docking experiments and overlays with ATP
in the active site of EGFR tyrosine kinase, have been prepared and found to be potent inhibitors. In vivo pharmacokinetics and
disease model activity are discussed.
ꢀ 2005 Elsevier Ltd. All rights reserved.
The receptor tyrosine kinases play a fundamental role
in the aberrant signalling that is characteristic of the
uncontrolled proliferation of a variety of human solid
tumours. The most widely studied of all these, the epi-
dermal growth factor receptor (EGFR) tyrosine kinase,1
features a trans-membrane receptor with an extracellu-
lar ligand-binding domain and a helical trans-membrane
region linked to an intracellular tyrosine kinase domain.
EGFR is activated by homo- and heterodimerisation
events following binding of epidermal growth factor
(EGF) and other ligands. EGFR is part of a subfamily
of four closely related receptors, three of which are cat-
alytically active: EGFR (also known as erbB1/Her 1),
erbB2 (Her 2/neu) and erbB4 (Her 4), and a fourth that
is catalytically inactive, erbB3 (Her 3), which neverthe-
less can enter into signalling through heterodimerisa-
tion. EGFR misregulation, through, for example,
overexpression and/or activating mutations, has been
implicated in the development and progression of many
solid tumour types2 including lung, breast, ovary, colon,
prostate and head and neck.
Inhibition of EGFR signalling has thus been identified
as an attractive strategy in control of tumour prolifera-
tion, and over a decade of intense activity in the field has
culminated in the discoveries and subsequent approvals
of gefitinib 13 and erlotinib 24 for the treatment of non-
small cell lung cancer following prior chemotherapeutic
intervention. Both of these agents belong to the 4-anili-
noquinazoline class of inhibitors, and their discovery
has been extensively reviewed.5 This chemical class of
inhibitors acts via competition with ATP and the tem-
plate has been heavily exploited in the search for new
inhibitors of signal transduction mechanisms, both
against EGFR, and other diverse kinases. The features
F
O
HN
N
Cl
N
O
O
N
1
HN
N
O
O
O
N
O
Keywords: EGF; Kinase; Inhibitor; Quinazoline.
*
Corresponding author. Tel.: +44 1625517920; fax: +44
1625586707; e-mail: jason.kettle@astrazeneca.com
2
0960-894X/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.12.028

1634
P. Ballard et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1633–1637
required for effective binding to the hinge region have
been well documented.6 By varying the aniline substitu-
tion pattern, fine-tuning of the kinase selectivity profile
can be achieved. For EGFR in particular, a small, lipo-
philic meta-substituent is key to potent inhibition.
MET 769
GLU 738
ASP 831
A second, general feature of the anilinoquinazoline tem-
plate is the positioning of a solubilising side chain at C-6
and/or C-7 of the quinazoline core. These side chains act
to improve physical properties and impart a more
favourable pharmacokinetic profile, though they may
also contribute to potency and selectivity. For potent
inhibition of the erbB family, these side chains are often
positioned at C-6, although side chains at both C-6 and
C-7 are noted in erlotinib 2, and the irreversible inhibitor
canertinib.7 The structures of both erlotinib8 and lapati-
nib9 bound to the catalytic domain of EGFR tyrosine ki-
nase have been solved, confirming the positioning of
these side chains at the solvent interface, and additional-
ly, prototype 4-anilinoquinazolines have been crystal-
lized in both CDK2 and the MAP kinase p38.10
ARG 817
ASP 776
GLU 780
MET 769
GLU 738
ASP 831
ARG 817
Despite extensive SAR studies aimed at modifying the
quinazoline core, little information was known about
possible beneficial effects of substitution at C-5. Indeed,
early reports with simple substituents had indicated that
C-5 substitution was detrimental to activity when com-
pared with analogous substitution at C-6 or C-7,6 and
work leading to the discovery of gefitinib had highlight-
ed the perceived importance of a free CH at this posi-
tion.11 Furthermore, detailed 3D QSAR and molecular
field analysis techniques have been used to suggest that
the positioning of bulky groups at C-5 should not be
considered.12 However, by examining an overlay of
ATP with the known binding mode of anilinoquinazo-
lines, we postulated that substitution at C-5 with an
appropriate group could lead to potent inhibitors which
utilized the ribose binding pocket.13 It was envisaged
that suitably placed cyclic groups, for example, though
containing heteroatoms, could mimic the interactions
observed for the endogenous ligand. Figure 1 shows
one such example, compound 12, modelled into the cat-
alytic domain of EGFR tyrosine kinase. In this model,
based upon the crystal structure of EGFR tyrosine
kinase in complex with erlotinib (PDB code 1M17), the
quinazoline N-1 binds to the hinge region at Met769,
with the aniline buried deep in the selectivity pocket.
Also indicated is an overlay with ANP (an unreactive
ATP mimic), created by overlaying the hinge binding
backbone N–H and carbonyl of the EGFR tyrosine ki-
nase structure with those of the Src kinase in complex
with ANP (PDB code 2src). The quinazoline and aden-
osine rings are co-planar, thus positioning the ribose
moiety and quinazoline C-5 substituent similarly. Based
on these assumptions, synthesis of this novel class of
quinazolines was undertaken, and activity with these
agents subsequently confirmed.
ASP 776
GLU 780
Figure 1. Compounds 12 (upper panel) and 24 (lower panel)
positioned in the EGFR tyrosine kinase structure solved with erlotinib
bound (PDB code 1m17). The molecules have the anilinoquinazoline
moieties overlaid directly onto erlotinib, the remaining parts of the
molecule shown in a representative conformation that is well ranked
according to Poisson–Boltzmann-,based scoring. The protein surface is
coloured according to electronic surface potential. Areas in red
indicate increasing negative potential, areas in blue increasing positive
surface potential. Carbon atoms are shown in grey, nitrogen in blue,
oxygen atoms in red, and fluorine in pale blue. Aniline meta-chloro
substituent is partially obscured and buried in the protein. Overlay
with ATP mimic ANP (shown in yellow) is based on overlaying the
hinge residues of the protein of the EGFR kinase structure with that of
the Src kinase (in PDB structure 2src).
could be selectively de-alkylated at C-5 by treatment
with magnesium bromide in pyridine13,14 to give 5. Pre-
sumably, coordination of magnesium between the C-4/5
oxygen atoms facilitates activation of the C-5 alkoxy
group towards nucleophilic de-alkylation. Selective
protection of the quinazoline N-3 is achieved with
chloromethyl pivalate, yielding substrates 6 suitable
for Mitsunobu alkylation. Treatment with DTAD/tri-
phenyl phosphine and an appropriate alcohol furnished,
after deprotection with ammonia in methanol, the corre-
sponding quinazolones 7. The requisite 3-chloro-4-fluo-
roaniline (present in both gefitinib and canertinib) was
introduced via chlorination and nucleophilic aromatic
substitution to give bis-alkoxy anilinoquinazolines 8.
Additionally, as outlined in Scheme 2, homologation
at C-7 could be achieved by deprotection of the C-7 ben-
zyl group in 9 with TFA, followed by alkylation of
7-hydroxyquinazoline 10 with 2-methoxyethyl bromide
to yield 23, or a basic side chain introduced by a two-
step sequence to yield piperazine analogue 24, via 11.
The key C-5 bis-alkoxyquinazolines 8 were synthesised
from bis-alkoxy anthranilic esters 3 according to the
general procedures outlined in Scheme 1. Quinazolone
ring formation was effected by treatment of the esters
with formamidine acetate to give quinazolones 4, which

P. Ballard et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1633–1637
Table 1. EGFR inhibition data for compounds 12–24
1635
R1
R1
O
O
O
O
F
a
O
NH
R5
R1
R1
O
HN
N
Cl
O
NH2
O
N
R6
R7
3: R1 = Me or Bn
N
4
b
Compound R5
R6
R7
EGFR IC50
(nM)a
OH
O
N
O
OH
O
N
N
O
NH
c
R1
R1
N
O
O
12
13
H
H
MeO
MeO
21
3
6
5
d, e
O
O
F
1384 150
R2
R2
O
Cl
O
HN
O
NH
f, g
14
15
H
H
MeO
MeO
187 77
8068
N
R1
R1
O
N
O
N
7
8
Scheme 1. Synthesis of bis-alkoxy anilinoquinazolines 8. Reagents and
conditions: (a) formamidine acetate, 2-methoxyethanol, 120 ꢁC, 64–
88%; (b) MgBr2, pyridine, reflux, 93–99%; (c) NaH, chloromethyl
pivalate, DMF, 0–25 ꢁC, 67–93%; (d) DTAD, R2-OH, Ph3P, DCM,
70–90%; (e) 7 N NH3 in MeOH, 70–90%; (f) POCl3, DIPEA, 1,2-
dichloroethane, reflux then; (g) 3-chloro-4-fluoroaniline, DIPEA, IPA,
reflux, 20–70% two steps.
N
16
17
18
19
20
21
MeO
H
H
225 50
N
H
3
2
HN
H
MeO
MeO
MeO
MeO
10
1
F
F
O
O
N
N
H
<2b
<2b
873
HN
N
Cl
O
HN
N
Cl
O
a
N
N
O
HO
H
9
10
O
b
H
N
F
F
O
O
R
HN
Cl
HN
Cl
O
O
O
S
c
N
N
22
23
24
H
H
H
MeO
9601
N
O
N
O
N
O
N
24
23: R = OCH2CH2OMe
11: R = OCH2CH2CH2Cl
N
O
O
671 35
66 22
O
N
Scheme 2. Synthesis of 5,7-bisalkoxy anilinoquinazolines 23 and 24.
Reagents and conditions: (a) TFA, 70 ꢁC, 82%; (b) K2CO3, DMF,
R-Br, 77–78%; (c) 1-methyl piperazine, NMP, 80 ꢁC.
O
O
N
a For determinations where n P 2, standard deviation is given.
b n = 2, lower than assay limit.
Biological data for representative compounds is high-
lighted in Table 1.15 Positioning of a cyclic, basic group
at C-5 such as the N-methylpiperidine 12 resulted in
potent enzyme inhibition. Cyclic groups that lacked this
charge tended to show lower affinity. Tetrahydropyran
13, for example, shows significantly decreased inhibi-
tion, although tetrahydrofuran 14 demonstrates a more
modest drop compared to 13, perhaps reflecting the
greater structural similarity to the sugar it is designed
to mimic. Figure 1 (upper panel) shows the anticipated
binding mode for 12 docked into the EGFR tyrosine
kinase structure. A conformational ensemble for 12
generated in OMEGA had the anilinoquinazoline moie-
ty overlaid with that of erlotinib and the resulting poses
scored using the Poisson–Boltzmann equation based
ZAP.16 The protein surface is coloured according to
electrostatic potential.17 The ribose pocket is character-
ised as having substantially negative surface potential,
and it is speculated that the high affinity seen for basic
groups in this region may be due to electrostatic comple-
mentarities with the cationic side chain. The reduced
activity observed with open-chain iso-propyl analogue
15 serves to highlight the importance of cyclic groups
in this region.

1636
P. Ballard et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1633–1637
Table 2. Selected DMPK data for compound 12
PPB free %a
Vdss (L/kg)
Cl (ml/min/kg)
t1/2 (h)
Bio-availability (%)
Mouse
Rat
3.11
2.51
8.2
12.3
20
66
5.2
2.9
103
63
a Measured at 37 ꢁC (mouse) and 25 ꢁC (rat).
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
Compound 16 illustrates that moving the methoxy
group to C-6 results in an order of magnitude drop in
affinity when compared with 12, and this may be due
to a significant conformational change in the C-5 side
chain. Removal of the methoxy group as in 17 also leads
to a potent inhibitor, although in general terms, removal
of the C-7 methoxy was often seen to be detrimental to
EGFR inhibition (when a broader range of anilines was
examined, the mean change in activity from C-7 hydro-
gen to C-7 methoxy is a potency shift of 1.2 log units, 14
matched pairs, data not shown). We next examined the
effect of extension from the basic nitrogen in 12. Consis-
tent with the binding model, the N–H piperidine ana-
logue 18 showed similar affinity to the N-methyl
analogue 12. Introduction of ethyl (compound 19) and
propyl (compound 20) also resulted in potent enzyme
inhibition, indicating a degree of tolerance for steric
bulk in this region. Consistent with the hypothesised
charge interaction around the ribose binding pocket,
neutral piperidines such as 21 and 22 showed markedly
reduced activity. Finally, we observed that by use of a
suitably positioned basic side chain at C-7, activity
could be regained with a previously less active C-5 sub-
stituent. Tetrahydropyran 13 demonstrates modest
affinity as discussed above. By appending the pipera-
zine-based side chain in 24, activity is restored and po-
tent inhibition seen. Figure 1 (lower panel) shows this
compound docked into EGFR tyrosine kinase. In this
model, the C-5 tetrahydropyran occupies the ribose
binding pocket, and the basic C-7 side chain extends
towards the solvent exposed region and may pick up
favourable charge interactions in the area towards
Glu780. That this is likely due to charge complementa-
rities is supported by compound 23 that positions a
neutral side chain at C-7 yet shows a similarly low
affinity to tetrahydropyran 13.
Control
Cpd 12 (100 mg/kg)
Cpd 12 (50 mg/kg)
60% inhibition
P < 0.05
0.0
0
2
4
6
8
10
12
14
Days of treatment
Figure 2. LoVo xenograft study with compound 12 showing degree of
inhibition of tumour volume after 13 days of treatment. Terminal
blood samples were taken 6 h after the final dose and analysis showed a
dose related increase in total exposure: 2.3 lM (50 mg/kg) and 4.1 lM
(100 mg/kg).
po with compound suspended in 1% (v/v) polysorbate
80. Tumour growth was measured twice-weekly using
callipers. Gratifyingly at the highest dose tested,
100 mg/kg, compound 12 demonstrated 60% inhibition
of tumour volume. A clear dose response was observed,
with no significant change from control seen at the lower
dose of 50 mg/kg.
In summary, chemistry has been developed to a series of
novel C-5 substituted anilinoquinazolines, selected on
the basis of docking experiments and overlays with
ATP in the active site of EGFR tyrosine kinase. Clear
SAR emerged concerning the requirements for good
affinity with these structures, in particular the require-
ments for basic moieties at positions C-5 and/or C-7.
Selected examples exhibit both good cellular activity
and pharmacokinetics, and resulted in significant inhibi-
tion in an in vivo model of anti-cancer activity.
Compound 12 was selected for further study. In an
EGF-driven KB cell proliferation assay,15 12 was shown
to inhibit proliferation with an IC50 of 87( 4) nM.
Selected DMPK data are shown in Table 2 following
an oral dose of 5 mg/kg and iv dose of 2 mg/kg. A good
pharmacokinetic profile were observed in both mouse
and rat, with data superior in the former. Slightly higher
clearance and a shorter half-life was observed in the rat
compared with mouse, and consistent with the presence
of a basic side chain, Vdss was seen to be high in both
species. Solubility in pH 7.4 buffer was measured at
2.1 lM. Based on the observed potency in cell-based
assays, and the favourable pharmacokinetic profile,
compound 12 was progressed to a LoVo Xenograft
study (Fig. 2). Female nude mice were implanted subcu-
taneously with 1 · 107 LoVo colon tumour cells. Once
tumours were established, animals were randomised
between groups (n = 7 per group) and dosed once daily
References and notes
1. Yarden, Y.; Sliwkowski, M. X. Nat. Rev. Mol. Cell Biol.
2001, 2, 127.
2. Salomon, D. S.; Brandt, R.; Ciardiello, F.; Normanno, N.
Crit. Rev. Oncol. Haematol. 1995, 19, 183.
3. Herbst, R. S.; Fukuoka, M.; Baselga, J. Nat. Rev. Cancer
2004, 4, 956.
4. Bonomi, P. Expert Opin. Investig. Drugs 2003, 12, 1395.
5. Douglas Laird, A.; Cherrington, J. M. Expert Opin.
Investig. Drugs 2003, 12, 51; Ciardiello, F.; De Vita, F.;
Orditura, M.; De Placido, S.; Tortora, G. Expert Opin.
Emerg. Drugs 2003, 8, 501.
6. (a) Rewcastle, G. W.; Denny, W. A.; Bridges, A. J.; Zhou,
H.; Cody, D. R.; McMichael, A.; Fry, D. W. J. Med.
Chem. 1995, 38, 3482; Bridges, A. J.; Zhou, H.; Cody, D.
R.; Rewcastle, G. W.; McMichael, A.; Showalter, H. D.
H.; Fry, D. W.; Kraker, A. J.; Denny, W. A. J. Med.
Chem. 1996, 39, 267.

P. Ballard et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1633–1637
1637
7. Maria, G. C. IDrugs 2004, 7, 58.
8. Stamos, J.; Sliwkowski, M. X.; Eigenbrot, C. J. Biol.
Chem. 2002, 277, 46265.
9. Wood, E. R.; Truesdale, A. T.; McDonald, O. B.; Yuan,
D.; Hassell, A.; Dickerson, S. H.; Ellis, B.; Pennisi, C.;
Horne, E.; Lackey, K.; Alligood, K. J.; Rusnak, D. W.;
Gilmer, T. M.; Shewchuk, L. Cancer Res. 2004, 64, 6652.
10. Shewchuk, L.; Hassell, A.; Wisely, B.; Rocque, W.;
Holmes, W.; Veal, J.; Kuyper, L. F. J. Med. Chem.
2000, 43, 133.
11. Barker, A. J.; Gibson, K. H.; Grundy, W.; Godfrey, A. A.;
Barlow, J. J.; Healy, M. P.; Woodburn, J. R.; Ashton, S.
E.; Curry, B. J.; Scarlett, L.; Henthorn, L.; Richards, L.
Bioorg. Med. Chem. Lett. 2001, 11, 1911.
12. Hou, T.; Zhu, L.; Chen, L.; Xu, X. J. Chem. Inf. Comput.
Sci. 2003, 43, 273.
13. For the first disclosure of C-5 substituted anilinoquinaz-
olines see Hennequin, L.F.A.; Ple, P. PCT Int. Appl.
WO2001094341; For a related series targeting erbB2
tyrosine kinase see: Ballard, P.; Bradbury, R. H.; Henne-
quin, L. F. A.; Hickinson, D. M.; Johnson, P. D.; Kettle,
J. G.; Klinowska, T.; Morgentin, R.; Ogilvie, D. J.;
Olivier, A. Bioorg. Med. Chem. Lett. 2005, 15, 4226.
14. Hennequin, L. F.; Allen, J.; Costello, G. C.; Curwen, J.;
Fennell, M.; Green, T. P.; Jacobs, V.; Lambert-van der
Brempt, C.; Morgentin, R.; Olivier, A.; Ple, P. A.;
Whittaker, R. Structure–activity relationship and in vivo
activity of a novel series of C5-substituted anilinoquinaz-
olines with highly potent and selective inhibition of c-Src
tyrosine kinase activity. Proc. Am. Assoc. Cancer Res.
2003. Abstract B193.
15. Compounds were evaluated in an EGFR kinase assay
measuring inhibition of phosphorylation of a synthetic
peptide substrate at Km ATP concentration, and their
ability to inhibit growth of the KB cell line when
stimulated with EGF assessed. For details of assay
conditions, see Hennequin, L. F. A.; Kettle, J. G.; Pass,
M.; Bradbury, R. H. PCT Int. Appl. WO2003040109.
16. OMEGA and ZAP are available from OpenEye Scientific
Software, 3600 Cerrillos Rd., Suite 1107, Santa Fe, NM
87507.
17. Images created in Pymol: Delano, W. L. The Pymol
Molecular Graphics System; Delano Scientific LLC, San
Carlos, CA, USA Baker, N. A.; Sept, D.; Joseph, S.;
Holst, M. J.; McCammon, J. A. Proc. Natl. Acad. Sci.
U.S.A. 2001, 98, 10037.