Structural insights into how irreversible inhibitors can overcome drug resistance in EGFR

Article abstract of DOI:10.1016/j.bmc.2008.02.053

Resistance to kinase- targeted cancer drugs has recently been linked to a single point mutation in the ATP binding site of the kinase. In EGFR, the crucial Thr790 gatekeeper residue is mutated to a Met and prevents reversible ATP competitive inhibitors from binding. Irreversible 4-(phenylamino)quinazolines have been shown to overcome this drug resistance and are currently in clinical trials. In order to obtain a detailed structural understanding of how irreversible inhibitors overcome drug resistance, we used Src kinase as a model system for drug resistant EGFR-T790M. We report the first crystal structure of a drug resistant kinase in complex with an irreversible inhibitor. This 4-(phenylamino)quinazoline inhibits wild type and drug resistant EGFR in vitro at low nM concentrations. The co-crystal structure of drug resistant cSrc-T338M kinase domain provides the structural basis of this activity.

Full text of DOI:10.1016/j.bmc.2008.02.053

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry 16 (2008) 3482–3488
Structural insights into how irreversible inhibitors can
overcome drug resistance in EGFR
Anja Michalczyk, Sabine Kluter, Haridas B. Rode, Jeffrey R. Simard,
¨
Christian Grutter, Matthias Rabiller and Daniel Rauh*
¨
Chemical Genomics Centre of the Max-Planck-Society, Otto-Hahn-Strasse 15, 44227 Dortmund, Germany
Received 14 February 2008; accepted 15 February 2008
Available online 20 February 2008
Abstract—Resistance to kinase- targeted cancer drugs has recently been linked to a single point mutation in the ATP binding site of
the kinase. In EGFR, the crucial Thr790 gatekeeper residue is mutated to a Met and prevents reversible ATP competitive inhibitors
from binding. Irreversible 4-(phenylamino)quinazolines have been shown to overcome this drug resistance and are currently in clin-
ical trials. In order to obtain a detailed structural understanding of how irreversible inhibitors overcome drug resistance, we used Src
kinase as a model system for drug resistant EGFR-T790M. We report the first crystal structure of a drug resistant kinase in complex
with an irreversible inhibitor. This 4-(phenylamino)quinazoline inhibits wild type and drug resistant EGFR in vitro at low nM con-
centrations. The co-crystal structure of drug resistant cSrc-T338M kinase domain provides the structural basis of this activity.
ꢀ 2008 Elsevier Ltd. All rights reserved.
1. Introduction
cules have since been developed to inhibit kinase func-
tion by targeting the ATP binding pocket of the kinase
domain.8–10 4-(phenylamino)quinazolines such as Erl-
otinib (Tarcevaꢁ; Genentech) and Gefitinib (Iressaꢁ;
Astra Zeneca) are a class of potent and selective ATP
competitive kinase inhibitors currently available for
the treatment of EGFR-associated cancer types. How-
ever, a single amino acid mutation (T790M) in the
ATP binding pocket of EGFR has been shown to lead
to the development of drug resistance and relapse of
the disease within months of the initiation of therapy.11
Aberrantly regulated kinases have been shown to play
causative roles in diseases such as cancer, diabetes and
neurological and autoimmune disorders, making this
class of enzymes an important set of therapeutic tar-
gets.1 The tyrosine kinase epidermal growth factor
receptor (EGFR) is particularly important since it has
been implicated in the development of a variety of hu-
man cancers.2 In addition to EGFR (erbB1 (HER1)),
activation of all members of the HER receptor family
(erbB2 (HER2), erbB3 (HER3) and erbB4 (HER4)) via
ligand binding or by point mutations results in cell
growth, proliferation, differentiation and migration3,4
and are associated with more aggressive disease and
poorer clinical outcome.5
The bulkier side chain of the mutated methionine resi-
due is thought to sterically impede binding of these
reversible inhibitors and disrupt the formation of a cru-
cial water-mediated hydrogen bond between the inhibi-
tor and T790 of wild type EGFR (Fig. 1).12
Inspired by the success of Imatinib (Gleevecꢁ; Novartis)
as the first targeted cancer therapeutic introduced into
the clinic at the start of the millenium for chronic mye-
logenous leukemia (CML) and gastrointestinal stromal
tumour (GISTS),6,7 a plethora of small organic mole-
The T790M mutation in EGFR is structurally analo-
gous to drug resistance mutations in other kinases such
as BCR-ABL (T315I), c-KIT (T670I) and platelet-de-
rived growth factor receptor-a (PDGFRa) (T674I)13
and was predicted to be a hot-spot for drug resistance
mutations in EGFR against ATP competitive inhibitors
even before the first clinical incidences were reported.14
More recently, a second class of 4-(phenylamino)quina-
zolines, which carry a Michael acceptor functional
Keywords: Anticancer agents; Proteins; Crystallography; Drug design;
Protein–ligand interactions.
*
Corresponding author. E-mail: daniel.rauh@cgc.mpg.de
0968-0896/$ - see front matter ꢀ 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2008.02.053

A. Michalczyk et al. / Bioorg. Med. Chem. 16 (2008) 3482–3488
3483
Figure 1. Crystal structure of wild type EGFR complexed with the reversible ATP competitive drug Erlotinib (PDB entry 1M17).23 (a) Key hydrogen
bonds formed between the quinazoline core of the inhibitor and T790 and M793 are indicated by dotted lines. (b) Drug resistance mutation T790M is
modelled and displayed in a sphere representation (magenta) and highlights the steric clash with the acetylene moiety of Erlotinib. Van der Waals
radii of the inhibitor are shown as mesh surface. Due to the T to M mutation, the water-mediated hydrogen bond between N3 of the quinazoline core
and the side chain of T790 is lost.
group at the 6-position and irreversibly alkylate a un-
ique cysteine (C797) in the ATP binding site of EGFR,
have been shown to overcome this drug resistance and
are currently in phase II/III clinical trials.15–17 Despite
intensive studies, a detailed understanding of how such
irreversible inhibitors overcome drug resistance in
EGFR-T790M at the atomic level remains limited. In
order to address this question, we used organic synthe-
sis, kinetic analysis and protein X-ray crystallography
to understand how irreversible inhibitors overcome this
mutation-associated drug resistance.
Table 1. Inhibition data of compounds 1–3 and Erlotinib against
EGFR and cSrc mutant variants
HN
N
Br
H
N
N
R =
R
N
O
1
3
2
a
IC50 (nM)
1
2
3
Erlotinib
EGFR
0.08
1.2
0.22
3.0
0.43
360
3.6
260
EGFR-T790M
cSrc-SM
cSrc-DM
2. Results and discussion
2.1. Chemistry
340
2810
236
5710
6400
>50,000
990
15,000
a Reported values are medians from independent experiments (Sup-
plementary Materials).
We synthesized a subset of irreversible 4-(phenylami-
no)quinazoline-based EGFR inhibitors (1 and 2) that
vary in the reactivities of their Michael acceptor group
(Table 1). Additionally, we synthesized a reversible
counterpart of 1, compound 3. The synthesis of quina-
zolines 1–3 was done according to published proce-
2.2. Kinase inhibition
In order to obtain structural information about the
binding modes of irreversible inhibitors 1 and 2 to drug
resistant kinases, we used the tyrosine kinase cSrc as a
model system of EGFR. The kinase domains of EGFR
and chicken cSrc share high identity in sequence (37%
overall) and structure and the ATP binding sites are
highly conserved (70%). We introduced a reactive cys-
teine (C345) via site-directed mutagenesis into a position
structurally homologous to C797 in EGFR to create a
single mutant cSrc variant (cSrc-SM). This particular
cSrc variant has recently been validated as a model sys-
tem for the structural investigation of irreversible EGFR
inhibitors.20 The high yield bacterial expression of re-
combinant cSrc,21 its readiness to crystallize as well as
the availability of soaking and co-crystallization systems
allows for rapid turnover in the study of protein-ligand
dures.18,19
A
synthesis scheme and details on
compound characterization can be found in the Supple-
mentary Materials. Briefly, N4-(3-bromophenyl)-4,6-
quinazolinediamine was treated with acryloyl chloride
and propionyl chloride in the presence of diisopropyl
ethyl amine to yield compounds 1 and 3, respectively.
Compound 2 was synthesized from (E)-4-(dimethyl-
amino)but-2-enoic acid hydrochloride18 and N4-(3-
bromophenyl)-4,6-quinazolinediamine (Scheme 1, Sup-
plementary Materials).
During preparation of the manuscript, the crystal structure of the
irreversible inhibitor HKI-272 in complex with EGFR-T790M was
reported.33

3484
A. Michalczyk et al. / Bioorg. Med. Chem. 16 (2008) 3482–3488
Figure 2. Electron density maps (2FoꢀFc contoured at 1r) of cSrc-DM in (a) the absence and (b) the presence of 2.
complexes by X-ray crystallography. Additionally, the
cSrc model system is advantageous for the study of
homologous protein kinases such as EGFR, which are
difficult to handle and often compromised by low dif-
fracting crystals. Similarly, we generated a model system
for drug resistant EGFR-T790M by introducing a struc-
turally homologous methionine (M338) to create a drug
resistant cSrc double mutant (cSrc-DM). We first per-
formed enzyme activity assays and verified that each
cSrc variant was active. Inhibition studies were then per-
formed to demonstrate the different binding specificities
of compounds 1–3 and Erlotinib in wild type and drug
resistant EGFR and cSrc (Table 1).
the cSrc-SM/DM in complex with 2, respectively
(Table 2).
In the non-ligand bound structure of cSrc-DM, the
mutation of T338 into the bulkier and more hydropho-
bic M338 does not cause major conformational changes
within the ATP binding site. However, the side chain
conformation of M338 creates a steric impediment that
would be expected to interfere with binding of reversible
ATP competitive inhibitors such as Erlotinib and 3, as
reflected in our IC50 measurements. In the structure of
cSrc-SM and cSrc-DM complexed with 2, the inhibitor
adopts binding modes iso-structural to those observed
for other irreversible and reversible 4-(phenylami-
no)quinazolines bound to EGFR and cSrc.20,22–25 The
quinazoline core of 2 is oriented with its N1 forming a
key hydrogen bond to the backbone amide NH of
M341. Notably, in the presence of 2, the Ce of M338
is flipped by 180ꢂ now pointing towards the back of
the ATP binding pocket to avoid the steric clash with
the m-bromoaniline moiety of the inhibitor (Fig. 3d).
The covalent modification of the c-sulfur of C345 by
the electrophile of the inhibitor is clearly indicated by
positive difference density and has been confirmed by
mass spectroscopy (see Section 4.4 and Supplementary
Materials).
Although potencies of the inhibitors vary between the
tested kinases, the measured IC50 values show the ex-
pected trend for reversible and irreversible inhibitors.
In both EGFR and cSrc, introduction of the drug resis-
tant threonine to methionine mutation causes a decrease
in the potency of both reversible (3, Erlotinib) and irre-
versible (1, 2) inhibitors. However, this effect is more
pronounced for the reversible inhibitors 3 and Erlotinib.
These results are in line with previously published data
which showed that irreversible inhibitors overcome
mutation-associated drug resistance in EGFR.13,15,16
2.3. Structural biology
To gain a deeper structural understanding of how irre-
versible inhibitors overcome the drug resistance muta-
tion in EGFR-T790M, we determined the crystal
structure of apo cSrc-DM (Fig. 2) as well as the struc-
tures of 2 co-crystallized with cSrc-SM and cSrc-DM
(Fig. 3). Single crystals of the cSrc in complex with 2
were grown in the presence of 1.1-fold molar excess of
2. From these crystals we could collect X-ray diffraction
data to 2.3 A resolution for cSrc-SM and 2.4 A resolu-
tion for the cSrc-DM. Molecular replacement revealed
two molecules in the asymmetric unit, including clear
additional electron density at the ligand binding posi-
tions. The models were refined to R and Rfree values of
0.21 and 0.28 for the apo form and 0.21 and 0.27 for
3. Conclusions
Application of our structural information gained from
cSrc-SM and cSrc-DM in complex with 2 to the problem
of drug resistance in EGFR-T790M suggests that a) loss
of the water-mediated hydrogen bond between N3 of the
quinazoline core and the side chain of wild type T790 of
EGFR and b) displacement of Ce of the introduced
methionine by the m-bromoaniline moiety of the inhib-
itor to avoid the steric clash, will likely weaken the initial
binding interactions for both reversible and irreversible
inhibitors, resulting in the observed lower potencies
(Table 1). Thus, despite the ability of reversible inhibi-
tors 3 and Erlotinib to inhibit EGFR-T790M at high
˚
˚

A. Michalczyk et al. / Bioorg. Med. Chem. 16 (2008) 3482–3488
3485
Figure 3. Comparison of the spatial orientation of residues T338 (cSrc-SM, (green)) and M338 (cSrc-DM) in the presence and absence (blue) of 2. A
key hydrogen bond formed between the quinazoline core of the inhibitor and M341 is indicated by a dotted line. (a) Binding of 2 to cSrc-SM. (b) A
rotation of about 90ꢂ about the y-axis reveals the covalent bond formed between 2 and C345 of cSrc-SM. The m-bromoaniline moiety of the inhibitor
adopts two conformations. (c) Binding of 2 to cSrc-DM induces a conformational change of M338 thereby extending the back of the ATP-binding
pocket in order to avoid a steric clash with the m-bromoaniline moiety of the inhibitor. (d) A rotation of 90ꢂ about the y-axis reveals the key
interactions between 2 and cSrc-DM and highlights movement of Ce of Met338 (depicted by red arrow). Figures prepared with PyMol.32
nanomolar concentrations in vitro, they are unlikely to
effectively compete with the high intracellular concentra-
tions of ATP to block EGFR-T790M enzyme activity in
cancer cells. In order for irreversible inhibitors to bind,
ligand-induced movement of the methionine and forma-
tion of the essential hydrogen bond between N1 of the
quinazoline and the backbone amide NH of M793 of
EGFR must occur to increase the residence time of
the inhibitor in the vicinity of the nucleophilic cysteine
to result in the Michael addition. In drug resistant
EGFR-T790M, these interactions may be partially dis-
rupted. However, the subsequent covalent modification
is the event which ultimately shifts the equilibrium be-
tween the free and ligand-bound form towards the inac-
tivated state of the kinase, thereby shutting down EGFR
enzyme activity.
4-(phenylamino)-quinazolines overcome the clinically
relevant T790M drug resistance mutation in EGFR.
The structural details gained from our studies show that
the side chain of the mutant methionine gatekeeper
interacts with both, reversible and irreversible 4-(phe-
nylamino)-quinazoline-based EGFR inhibitors and
interferes with ligand binding. Our results should prove
useful for the design of kinase inhibitors that account
for the prominent problem of drug resistance in ki-
nase-targeted cancer therapy.
4. Experimental
Anhydrous solvents were purchased from Acros Organ-
ics and Fluka. Other chemical materials were purchased
from Alfa Aesar, Fluka and Sigma Aldrich and were
used as received. Site-directed mutagenesis of cSrc was
performed using a standard site-directed mutagenesis
The use of cSrc as a model system has allowed us to
unravel the structural basis by which irreversible

3486
A. Michalczyk et al. / Bioorg. Med. Chem. 16 (2008) 3482–3488
Table 2. Statistics for data collection and refinement of apo cSrc-DM, cSrc-SM/DM + 2
Data collection
Space group
˚
Resolution (A)
cSrc-SM + 2
cSrc-DM
cSrc-DM + 2
P1
2.33–72.94
2.33–2.46
P1
2.32–72.55
2.32–2.45
P1
2.38–73.57
2.38–2.51
˚
Highest Shell (A)
Cell dimensions
˚
a, b, c (A)
˚
41.94; 63.53; 74.50
78.28; 90.37; 89.45
42.00; 63.02; 73.95
78.78; 88.80; 90.14
42.22; 63.36; 75.06
78.58; 89.08; 89.87
a, b, c (A)
Rsym
6.0 (40.7)
29165
7.5 (37.0)
30188
6.9 (12.6)
29935
Unique reflections
Completeness (%)
Average redundancy
90.7 (85.9)
3.6
93.8 (91.2)
2.2
98.0 (96.9)
2.1
Refinement
Rwork/Rfree
21.2/26.7
20.5/28.0
21.4/26.6
No. of atoms
Protein A,B
Ligand A, B
Water
1923, 1970
54, 27
99
2049, 1983
No ligand
172
1988, 1950
27, 27
168
2
˚
Average B-factors (A )
Protein A, B
Ligand A, B
Water
32.1, 32.7
39.5, 62.5
27.2
34.4, 34.0
No ligand
28.6
33.7, 33.6
64.9, 59.9
29.9
R.m.s. deviations
˚
Bond lengths (A)
0.016
1.770
0.016
1.782
0.018
1.841
Angles (ꢂ)
Ramachandran outliers (%)
PDB accession code
0
2QLQ
0
2QI8
0
2QQ7
kit (Qiagen). Enzyme kinetics were measured using a
commercial FRET-based assay system (Z’lyteTM) avail-
able from Invitrogen (#PV3191 and #PV3193). Wild
type EGFR was purchased from Invitrogen (#PV3872)
and EGFR-T790M was purchased from Upstate (#
14–725).
with UV-detector (Model 340) and a VP 25-21 Nucleo-
dur (C18 Gravity, 5 l) column (Serial no. 2105150).
4.2. Protein expression and purification
Prior to expression, mutations of T338 to methionine
and S345 to cysteine were introduced by site-directed
mutagenesis and confirmed by sequence analysis. The
T338M mutation was introduced using the following
primer sequences: 50-AGCCCATCTACATCGTCATG
GAGTACATGAGCAAGGGG-30 and 50-CCCCTTG
CTCATGTACTCCATGACGATGTAGATGGGCT-30.
The S345C mutation was introduced using the following
primer sequences: 50-GTACATGAGCAAGGGGGTC
TGCCTCCTGGATTTCCTG-30 and 50-CAGGAAA
TCCAGGAGGCAGACCCCCTTGCTCATGTAC-30.
All chicken cSrc (residues 251–533) variants were co-ex-
pressed in E. coli and purified as previously
described.20,21
4.1. Synthesis of compounds
All 4-(phenylamino)quinazolines were prepared accord-
ing to the literature procedures (Supplementary Mate-
rial). 1H and 13C NMR spectra were recorded on a
Varian Mercury 400, Bruker DRX 500 or Varian Inova
600 spectrometer using dimethylsulfoxide-d6 as an inter-
nal reference. Chemical shifts (d) are given in ppm and
coupling constants (J) were measured in Hz. The follow-
ing abbreviations are used; s-singlet, d-doublet, dd-dou-
blet of doublet, t-triplet, td-triplet of doublet, q-quartet,
qd-quartet of doublet, br-broad. LC–MS spectra were
obtained on a LTQ Orbitrap (high resolution mass spec-
trometer from Thermo Electron) coupled to an ‘Accela’
HPLC System (consisting of Accela pump, Accela auto-
sampler and Accela PDA detector) supplied with a
‘Hypersil GOLD’ column (50 mm · 1 mm, 1.9 lm parti-
cle size) from Thermo Electron. Melting points were
4.3. IC50 determination
IC50 values were determined with the Z’lyteTM assay sys-
tem. The reactions were performed with a 10 ll total
volume in 384-well small volume plates from Greiner
(#784076). The kinase reaction for cSrc consisted of ki-
nase buffer (50 mM HEPES pH 7.5, 0.01% BRIJ-35,
10 mM MgCl2, 1 mM EGTA), 1–15 nM kinase, 2 lM
peptide and either 25 lM ATP for cSrc-SM or 15 lM
ATP for cSrc-DM. For each IC50 determination, 16 dif-
ferent concentrations of inhibitor (range from 250,000
to 10 nM) were used in triplicate and at least three inde-
determined using open glass capillaries on a Buchi melt-
¨
ing point apparatus (B-540) and are uncorrected. Ana-
lytical TLC was carried out on Merck 60 F245
aluminium-backed silica gel plates. Compounds were
purified by column chromatography using Baker silica
gel (40–70 lm particle size). Preparative HPLC was car-
ried out on the final compounds using a Varian Prostar

A. Michalczyk et al. / Bioorg. Med. Chem. 16 (2008) 3482–3488
3487
pendent experiments were performed. Before starting
the kinase reaction, enzyme and inhibitor were incu-
bated for 40 min. The kinase reaction for EGFR con-
sisted of BSA-supplemented kinase buffer (25 mM
HEPES pH 7.5, 0.005% BRIJ-35, 2 mM MnCl2,
10 mM MgCl2, 1 mM DTT, 2.5 mM CaCl2 and
0.1 mg/ml BSA), 1–5 nM kinase, 0.5 lM peptide, 7 lM
ATP for wild type EGFR and 2 lM ATP for EGFR-
T790M. For IC50 determinations, 10 different concen-
trations (ranging from 0.1 nM to 10 lM) of inhibitor
were used in duplicate. Each experiment was repeated
at least twice. Following the kinase reaction, 5 ll of
Development Solution (included with the kit) was added
to cleave the remaining unphosphorylated peptide. Fol-
lowing a 1 h incubation time with Development Solu-
tion, 5 ll of Stop Solution (included with the kit) was
added to the reaction mixture to bring the final volume
to 20 ll. Fluorescence was then measured with a Spec-
tramaxꢁ M5 plate reader from Molecular Devices (cSrc
experiments) or a Safire2TM from Tecan. Upon excita-
tion of coumarin at 400 nm, fluorescence emission was
measured at 445 nm (coumarin) and 520 nm (fluores-
cein). The ratio of coumarin to fluorescein emission
(R = k445/k520) was used to calculate the percentage
of phosphorylation of the peptide by the kinase.
ment and electron density maps calculation were carried
out using REFMAC5.28 Compound 2 was constructed
and minimized with DS ViewerPro 6.0 (Accelrys Soft-
ware Inc.). Topology files were generated using the Dun-
dee PRODRG229 server. Model building was done using
COOT.30 Detailed data and refinement statistics are gi-
ven in Table 2. Atomic coordinates for apo cSrc-DM
and cSrc-DM-2 and cSrc-SM-2 complex structures have
been deposited to the Protein Data Bank (Accession
codes 2QI8, 2QQ7 and 2QLQ). Refined structures were
validated with PROCHECK.31 Figures were produced
using PyMol 2002 (DeLano Scientific, San Carlos, CA,
USA).
Acknowledgment
J.R.S. was funded by the Alexander von Humboldt
Foundation. Organon/Schering Plough, Bayer-Schering
Pharma, Merck-Serono and BayerCrop Science are
thanked for financial support. We thank Roman Tho-
mas for Erlotinib.
Supplementary data
4.4. Mass spectroscopy experiments
Supplementary data associated with the chemical syn-
thesis of 4-(phenylamino)quinazolines and ESI-MS
analysis can be found online. Supplementary data asso-
ciated with this article can be found, in the online ver-
sion, at doi:10.1016/j.bmc.2008.02.053.
Mass spectroscopy analysis was carried out by first incu-
bating cSrc-DM (12.5 lM) with an equimolar amount
of each irreversible inhibitor for 40 min at RT. After
dialyzing against deionized water, the protein-inhibitor
complex was prepared for analysis by mixing 4 ll of
the dialyzed protein with 16 ll deionized water and
20 ll of an acetonitrile/formic acid (2% v/v) solution.
Aliquots (20 ll) were analyzed by ESI-MS using a Finn-
igan LCQ Advantage Max mass spectrometer from
Thermo. Deconvolution and visualization were per-
formed using the BioWorks 3.1 (SR1) from Thermo
Scientific.
References and notes
1. Gschwind, A.; Fischer, O. M.; Ullrich, A. Nat. Rev.
Cancer 2004, 4, 361–370.
2. Schlessinger, J. Science 2004, 306, 1506–1507.
3. Arteaga, C. L. J. Clin. Oncol. 2001, 19, 32S–40S.
4. Dowell, J. E.; Minna, J. D. Nat. Clin. Pract. Oncol. 2004,
1, 2–3.
5. Temam, S.; Kawaguchi, H.; El-Naggar, A. K.; Jelinek, J.;
Tang, H.; Liu, D. D.; Lang, W.; Issa, J. P.; Lee, J. J.; Mao,
L. J. Clin. Oncol. 2007, 25, 2164–2170.
6. Holtz, M. S.; Slovak, M. L.; Zhang, F.; Sawyers, C. L.;
Forman, S. J.; Bhatia, R. Blood 2002, 99, 3792–3800.
7. Zhang, P.; Gao, W. Y.; Turner, S.; Ducatman, B. S. Mol.
Cancer 2003, 2, 1.
8. Baselga, J.; Arteaga, C. L. J. Clin. Oncol. 2005, 23, 2445–
2459.
9. El-Obeid, A.; Hesselager, G.; Westermark, B.; Nister, M.
Biochem. Biophys. Res. Commun. 2002, 290, 349–358.
10. Hynes, N. E.; Lane, H. A. Nat. Rev. Cancer 2005, 5, 341–
354.
4.5. Crystallization
Crystals of suitable size were grown by vapour diffusion
(using the hanging-drop method) at 298 K against well
solutions containing 15–20% (w/v) ethylene glycol. For
co-crystallization experiments, preincubations were car-
ried out with a ratio of 1:1.1 protein to inhibitor (10 mM
in DMSO). Plate-like crystals of the triclinic space group
P1 were obtained after one week.
4.6. Data collection and refinement
11. Sharma, S. V.; Bell, D. W.; Settleman, J.; Haber, D. A.
Nat. Rev. Cancer 2007, 7, 169–181.
12. Kobayashi, S.; Boggon, T. J.; Dayaram, T.; Janne, 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.
13. 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., 3rd; 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.;
Datasets for apo cSrc-DM and cSrc-SM complexed with
2 were measured in-house (Rigaku MicroMax-007, MPI
Dortmund). A dataset of cSrc-DM in complex with 2
was measured at the X10SA beamline of the Swiss Light
Source (PSI, Villingen, Switzerland). Datasets were inte-
grated using MOSFLM and scaled with SCALA
(CCP426 package 5.0.2). The structures were solved by
molecular replacement with PHASER.27 Starting coor-
dinates were taken from cSrc-SM in complex with 1
(PDB accession code 2HWP20). Crystallographic refine-

3488
A. Michalczyk et al. / Bioorg. Med. Chem. 16 (2008) 3482–3488
Varmus, H.; Zarrinkar, P. P.; Lockhart, D. J. Proc. Natl.
Acad. Sci. U.S.A. 2005, 102, 11011–11016.
14. Blencke, S.; Ullrich, A.; Daub, H. J. Biol. Chem. 2003,
278, 15435–15440.
15. Kobayashi, S.; Ji, H.; Yuza, Y.; Meyerson, M.; Wong, K.
K.; Tenen, D. G.; Halmos, B. Cancer Res. 2005, 65, 7096–
7101.
22. Hennequin, L. F.; Allen, J.; Breed, J.; Curwen, J.; Fennell,
M.; Green, T. P.; Lambert-van der Brempt, C.; Morgen-
tin, R.; Norman, R. A.; Olivier, A.; Otterbein, L.; Ple, P.
A.; Warin, N.; Costello, G. J. Med. Chem. 2006, 49, 6465–
6488.
23. Stamos, J.; Sliwkowski, M. X.; Eigenbrot, C. J. Biol.
Chem. 2002, 277, 46265–46272.
16. Kwak, E. L.; Sordella, R.; Bell, D. W.; Godin-Heymann,
N.; Okimoto, R. A.; Brannigan, B. W.; Harris, P. L.;
Driscoll, D. R.; Fidias, P.; Lynch, T. J.; Rabindran, S. K.;
McGinnis, J. P.; Wissner, A.; Sharma, S. V.; Isselbacher,
K. J.; Settleman, J.; Haber, D. A. Proc. Natl. Acad. Sci.
U.S.A. 2005, 102, 7665–7670.
17. Pao, W.; Miller, V. A.; Politi, K. A.; Riely, G. J.; Somwar,
R.; Zakowski, M. F.; Kris, M. G.; Varmus, H. PLoS Med.
2005, 2, e73.
18. Considine, J. L.; Daigneault, S.; Chew, W.; Iera, S.;
Duncan, S. M.; Ren, J. World Patent 2004066919, CAN
141:17382, 2004.
19. Tsou, H. R.; Mamuya, N.; Johnson, B. D.; Reich, M. F.;
Gruber, B. C.; Ye, F.; Nilakantan, R.; Shen, R.; Disca-
fani, C.; DeBlanc, R.; Davis, R.; Koehn, F. E.; Green-
berger, L. M.; Wang, Y. F.; Wissner, A. J. Med. Chem.
2001, 44, 2719–2734.
24. 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–
6659.
25. Yun, C. H.; Boggon, T. J.; Li, Y.; Woo, M. S.; Greulich,
H.; Meyerson, M.; Eck, M. J. Cancer Cell 2007, 11, 217–
227.
26. Collaborative Computational Project, N. Acta Crystal-
logr. D 1994, 50, 760–763.
27. Read, R. J. Acta Crystallogr. D 2001, 57, 1373–1382.
28. Murshudov, G. N.; Vagin, A. A.; Dodson, E. J. Acta
Crystallogr. D 1997, 53, 231–354.
29. Schuttelkopf, A. W.; Aalten, D. M. F. v. Acta Crystallogr.
¨
D 2004, 60, 1355–1363.
30. Emsley, P.; Cowtan, K. Acta Crystallogr. D 2004, 60,
2126–2132.
20. Blair, J. A.; Rauh, D.; Kung, C.; Yun, C. H.; Fan, Q. W.;
Rode, H.; Zhang, C.; Eck, M. J.; Weiss, W. A.; Shokat, K.
M. Nat. Chem. Biol. 2007, 3, 229–238.
21. Seeliger, M. A.; Young, M.; Henderson, M. N.; Pellicena,
P.; King, D. S.; Falick, A. M.; Kuriyan, J. Protein Sci.
2005, 14, 3135–3139.
31. Laskowski, R. A.; MacArthur, M. W.; Moss, D. S.;
Thornton, J. M. J. Appl. Crystallogr. 1993, 26, 283–291.
32. DeLano, W. L. http://www.pymol.org bf 2002.
33. Yun, C. H.; Mengwasser, K. E.; Toms, A. V.; Woo, M. S.;
Greulich, H.; Wong, K. K.; Meyerson, M.; Eck, M. J.
Proc. Natl. Acad. Sci. U.S.A. 2008.