Development of Specific, Irreversible Inhibitors for a Receptor Tyrosine Kinase EphB3

Article abstract of DOI:10.1021/jacs.6b05483

Erythropoietin-producing human hepatocellular carcinoma (Eph) receptor tyrosine kinases (RTKs) regulate a variety of dynamic cellular events, including cell protrusion, migration, proliferation, and cell-fate determination. Small-molecule inhibitors of Eph kinases are valuable tools for dissecting the physiological and pathological roles of Eph. However, there is a lack of small-molecule inhibitors that are selective for individual Eph isoforms due to the high homology within the family. Herein, we report the development of the first potent and specific inhibitors of a single Eph isoform, EphB3. Through structural bioinformatic analysis, we identified a cysteine in the hinge region of the EphB3 kinase domain, a feature that is not shared with any other human kinases. We synthesized and characterized a series of electrophilic quinazolines to target this unique, reactive feature in EphB3. Some of the electrophilic quinazolines selectively and potently inhibited EphB3 both in vitro and in cells. Cocrystal structures of EphB3 in complex with two quinazolines confirmed the covalent linkage between the protein and the inhibitors. A "clickable" version of an optimized inhibitor was created and employed to verify specific target engagement in the whole proteome and to probe the extent and kinetics of target engagement of existing EphB3 inhibitors. Furthermore, we demonstrate that the autophosphorylation of EphB3 within the juxtamembrane region occurs in trans using a specific inhibitor. These exquisitely specific inhibitors will facilitate the dissection of EphB3's role in various biological processes and disease contribution.

Full text of DOI:10.1021/jacs.6b05483

Subscriber access provided by Northern Illinois University
Article
Development of Specific, Irreversible Inhibitors
for a Receptor Tyrosine Kinase EphB3
Alvin Kung, Ying-Chu Chen, Marianne Schimpl, Feng Ni, Jianfa Zhu,
Maurice Turner, Henrik Molina, Ross Overman, and Chao Zhang
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.6b05483 • Publication Date (Web): 01 Aug 2016
Downloaded from http://pubs.acs.org on August 2, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 8
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Development of Specific, Irreversible Inhibitors for a Recep-
tor Tyrosine Kinase EphB3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Alvin Kung,† Ying-Chu Chen,† Marianne Schimpl,# Feng Ni,† Jianfa Zhu,† Maurice Turner,† Henrik
Molina,¶ Ross Overman,∞ and Chao Zhang*,†,||
†Department of Chemistry and Loker Hydrocarbon Research Institute, ||Department of Biological Sciences, University of
Southern California, Los Angeles, California 90089, United States
#Discovery Sciences, Innovative Medicines and Early Development Biotech Unit, AstraZeneca, Building 310, Cambridge
Science Park, Milton Road, Cambridge, CB4 0WG, UK
∞Discovery Sciences, Innovative Medicines and Early Development Biotech Unit, AstraZeneca, Alderley Park, Macclesfield,
Cheshire, SK10 4TG, UK
¶Proteomic Resource Center, The Rockefeller University, New York, NY 10065, United States
ABSTRACT: Erythropoietin-producing human hepatocellular carcinoma (Eph) receptor tyrosine kinases (RTKs) regulate a variety
of dynamic cellular events including cell protrusion, migration, proliferation and cell-fate determination. Small-molecule inhibitors
of Eph kinases are valuable tools for dissecting the physiological and pathological roles of Eph. However, there is a lack of small-
molecule inhibitors that are selective for individual Eph isoforms due to the high homology within the family. Herein, we report the
development of the first potent and specific inhibitors of a single Eph isoform, EphB3. Through structural bioinformatic analysis, we
identified a cysteine in the hinge region of the EphB3 kinase domain, a feature that is not shared with any other human kinases. We
synthesized and characterized a series of electrophilic quinazolines to target this unique, reactive feature in EphB3. Some of the
electrophilic quinazolines selectively and potently inhibited EphB3 both in vitro and in cells. Co-crystal structures of EphB3 in com-
plex with two quinazolines confirmed the covalent linkage between the protein and the inhibitors. A “clickable” version of an opti-
mized inhibitor was created and employed to verify specific target engagement in the whole proteome and to probe the extent and
kinetics of target engagement of existing EphB3 inhibitors. Furthermore, we demonstrate that the autophosphorylation of EphB3
within the juxtamembrane region occurs in trans using a specific inhibitor. These exquisitely specific inhibitors will facilitate the dis-
section of EphB3’s role in various biological processes and disease contribution.
INTRODUCTION
the kinase domain, and the SAM domain.13 In principle, RTK
autophosphorylation can operate in cis within one receptor, or in
trans between two neighboring receptors. Studies with prototypical
RTKs such as the epidermal growth factor receptor (EGFR) and
the insulin receptor elucidated that the autophosphorylation
mechanism varies depending on the position of the
phosphosites.14-16 However, the activation mechanism for Eph
receptors has not been fully investigated to our knowledge.
Erythropoietin-producing human hepatocellular (Eph) receptors
and their ligands ephrins orchestrate various dynamic cellular
events including cell protrusion, migration, proliferation and cell-
fate determination.1,2 There is a strong association between
dysregulation of Eph receptors and cell proliferation, anti-
apoptotic
repercussions
and
ultimately
oncogenic
transformation.3-8 For example, EphA4 and EphB4 have been
found to contribute to proliferation or metastasis of multiple forms
of cancer.6 However, for another family member EphB3, recent
studies have documented conflicting roles in cancerogenesis.9-12
While earlier studies suggested that overexpression of EphB3 in
non-small-cell lung cancer promoted metastasis by enhancing the
tumor’s survival and migratory capabilities,10 a recent report im-
plicates EphB3 as a tumor suppressor.11 The exact role of EphB3
in cancer remains controversial at the moment.
The Eph receptors have been studied using various genetic
methods including protein overexpression, gene knockout and
knockdown.1,5 These studies provided valuable information on the
cellular functions and disease relevance of Ephs. However, the
dramatic change in protein levels and slow genetic perturbation
can cause either overexpression phenotype or cellular compensa-
tions, which confounds interpretation of the experimental
results.17 Use of pharmacological modulators of Eph can minimize
these complications as they take effect rapidly without affecting
protein levels.18 Although multiple inhibitors of Eph kinases have
been reported in the literature, they invariably suffer from a lack
of specificity for individual Eph isoforms due to the high conserva-
tion within Eph kinase domains.7,19 The lack of isoform-selective
Binding of ephrin to Eph is thought to promote receptor oli-
gomerization resulting in autophosphorylation at multiple sites in
the cytoplasmic domains, including the juxtamembrane region,
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 8
inhibitors has prevented pharmacological mapping of functions of
individual Eph kinases.
quinazoline, which constitutes the core of the anticancer drug
gefitinib, erlotinib and lapatinib,29 was selected for its good fit into
the EphB3 active site and the proximity of its 7 position to the
unique cysteine residue based on the docking mode (Figure 1B). A
panel of quinazolines that contain various electrophilic groups at
the C7 position were thus designed to irreversibly target C717 in
EphB3 (Figure 1C). These electrophiles include halomethylke-
tones, acrylate, chloroacetamide, vinylsulfonamide and acryla-
mide that exhibit varying distance from the quinazoline core to
the electrophilic carbons and distinct stereochemical requirements
for reaction with the cysteine thiol group. The quinazolines were
chemically synthesized using two different routes (Schemes S1 and
S2).
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1. IC50 (nM) Values of Quinazolines Against Various Forms of
Eph Kinases In Vitroα
Com-
pound
EphB3
WT
EphB3
C717S
EphB4
WT
EphA4
WT
1
2
3
6
>10,000
415
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
6
55
Figure 1. Design of specific, electrophilic inhibitors of EphB3 guided
by structural bioinformatics. (A) Partial sequence alignment of EphB3
with ten other protein kinases within the subdomain V. The position
where EphB3 contains a cysteine (C717) is highlighted. (B) Binding
model of 4-anilinoquinazoline to the active site of EphB3 with C717
highlighted. (C) Quinazolines containing various electrophiles at posi-
tion 7 designed to covalently target C717 in EphB3.
αDisk assay based on phosphorylation of a peptide substrate with [γ-32P] ATP was used.
IC50 values were calculated using GraphPad Prism.
Selective Inhibition of EphB3 in vitro. Having synthe-
sized the panel of electrophilic quinazolines, we first characterized
their inhibition of the EphB3 kinase in vitro. Recombinant EphB3
kinase was expressed and purified from E. coli as previously de-
scribed.28 The electrophilic quinazolines were screened for inhibi-
tion against the EphB3 kinase using an in vitro kinase assay based
on the phosphorylation of a peptide substrate with [γ-32P] ATP.30
The quinazolines caused different levels of inhibition against the
EphB3 kinase (Figure 2A). Three quinazolines that contain halo-
methylketone (2 and 3) or chloroacetamide (6) inhibited EphB3
kinase activity by more than 80% when used at 2 µM, while the
other quinazolines in the panel caused little or moderate inhibi-
tion against EphB3 under the same conditions (Figure 2A). The
varying degrees of EphB3 inhibition by these quinazolines is con-
sistent with the notion that only certain electrophiles are poised to
form covalent interaction with C717. We next determined the
IC50 values of the three most potent quinazolines along with the
non-electrophilic methylketone 1 against EphB3 (Table 1). The
halomethylketones 2 and 3 inhibited EphB3 with apparent IC50
values of 415 nM and 6 nM respectively. The potency difference
is consistent with the reactivity difference between these two hal-
omethylketone groups as chlorine is a much better leaving group
than fluorine for the nucleophilic substitution reaction. The chlo-
roacetamide-containing 6 had an intermediate potency at inhibit-
ing EphB3 (IC50 = 55 nM) compared to the two halomethylke-
tones. In contrast, the non-electrophilic quinazoline 1 that differs
from 2 and 3 by a single atom substitution inhibited EphB3 poor-
A chemical-genetic approach that generates potent and specific
inhibitors of protein kinases by targeting non-conserved cysteine
residues has been previously described.17,20,21 The approach has
been employed to generate potent irreversible inhibitors of a
number of protein kinases including Rsk, Nek2, c-Src, a drug-
resistant mutant (T790M) of EGFR, JNK and Cdk7.22-27 Our
kinome-wide sequence analysis revealed additional cysteine-
containing kinases that would be amenable to this approach. In
particular, we noticed that EphB3 contains a cysteine residue near
the end of the hinge region, a feature that is shared by only two
other kinases, LKB1 and PINK1, in the entire human kinome
(Figure 1A and data not shown). Importantly, the cysteine residue
in LKB1 and PINK1 is expected to point away from the active
site and not as accessible as that of EphB3 because of a deletion in
the hinge region of these two serine/threonine kinases compared
to tyrosine kinases (Figure 1A). Thus EphB3 contains a unique
reactive feature that can be exploited for the development of spe-
cific irreversible inhibitors of EphB3. Herein, we report the de-
sign, synthesis, optimization and biological characterization of a
series of electrophilic quinazolines as irreversible inhibitors of
EphB3. Our optimized inhibitors exhibit exquisite specificity at
inhibiting EphB3 not only within the human kinome but also in
the whole proteome. These specific inhibitors were employed to
measure the engagement of various kinase inhibitors to EphB3
and probe the activation mechanism of EphB3 in cells.
ly (IC50 > 10 µM).
To determine the importance of the cysteine (C717) for inhibi-
tion of EphB3, we measured the electrophilic quinazolines’ inhibi-
tion against a mutant form of EphB3 in which this cysteine was
mutated to serine. After being expressed and purified from E. coli,
EphB3 C717S was found to have comparable activity to that of
the wild-type, and was not significantly inhibited by any
quinazolines in the panel at 10 µM (Figure S1).28 Finally, none of
the electrophilic quinazolines potently inhibit EphA4 and EphB4,
two other Eph kinases that are missing a cysteine at the homolo-
gous position (Table 1). Together, these results suggest that C717
is essential for potent inhibition of EphB3 by the electrophilic
quinazolines 2, 3 and 6.
RESULTS AND DISCUSSION
Synthesis of Electrophilic EphB3 Inhibitors. To identi-
fy an appropriate scaffold for covalently targeting the unique cys-
teine C717 in EphB3, a dozen inhibitors with different pharma-
cophores such as pyrazolopyrimidine, bisanilinopyrimidine and
quinazoline were docked into the EphB3 active site based on a
crystal structure that has been recently reported.28 A combination
of automated docking using AutoDock Vina and manual docking
in Pymol was employed in generating and analyzing binding poses
of the inhibitor scaffolds. The docking score (or potential steric
clash), distance from the C717 thiol group, and ease of derivatiza-
tion were weighed in evaluating the various scaffolds. The
2
ACS Paragon Plus Environment

Page 3 of 8
Journal of the American Chemical Society
Covalent Inhibition of EphB3 by the Quinazolines. We
nM respectively (Figure S4 and Figure 3B). This coincides with
the non-electrophilic control compound 1 incapable of inhibiting
EphB3 (Figure S5). These results suggest that the electrophilic
quinazolines can readily cross plasma membrane and reach the
cytoplasmic kinase domain of EphB3 in live cells.
1
2
3
4
5
6
7
8
examined the reversibility of EphB3 inhibition by the electrophilic
quinazolines by conducting washout experiments. Upon washout,
1 lost its inhibition against EphB3, consistent with it being a re-
versible inhibitor (Figure 2B). In contrast, 3 and 6 retained potent
inhibition of EphB3 after washing out the inhibitors, indicating
covalent inhibition. Furthermore, 1-hr pre-incubation with EphB3
kinase decreased the apparent IC50 value of 2 by 5 fold, showing
that the inhibition is time-dependent as expected for a covalent
inhibitor (Figure S2). To verify the mode of covalent inhibition,
we used mass spectrometry to measure the modification of EphB3
by the electrophilic quinazolines. The EphB3 protein was incu-
bated with 3 or 6 and then subjected to ESI-MS analysis to de-
termine the whole protein mass. We observed mass increases of
259 and 275 Da of the EphB3 kinase upon treatment with 3 and
6 respectively (Figure S3). These mass shifts are consistent with
covalent modification of EphB3 by these two electrophilic
quinazolines. Furthermore, incubation of the EphB3 protein with
6 followed by trypsinization and LC-MS/MS analysis revealed
selective covalent modification of the C717-containing tryptic
peptide (Figure 2C and Figure S3). Together, these results provide
definitive proofs that the electrophilic quinazolines irreversibly
inhibit EphB3 by covalently modifying C717.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Modification of 6 led to the identification of 9, a much
more potent inhibitor of EphB3 in cells. (A) The activity of EphB3
WT is inhibited by the electrophilic quinazolines (2, 3 and 6) but not
by the non-electrophilic analog 1 in cells. (B) Compound 6 inhibited
the autophosphorylation of EphB3 WT in a dose-dependent manner
with an EC50 of 170 nM. (C) Chemical structure of 9, a derivative of
6. (D) 9 exhibited potent inhibition of EphB3 activity with an EC50 of
3 nM in cells. Cell lysate were resolved by SDS-PAGE, transferred to
nitrocellulose membrane, and immunoblotted with antibodies for
phospho-Eph (pEphB3), EphB3, and β-actin.
Importantly, 3 and 6 do not significantly affect autophosphory-
lation of the C717S mutant of full-length EphB3 receptor at con-
centrations up to 10 µM, supporting the notion that potent inhibi-
tion of EphB3 by electrophilic quinazolines depends on C717 not
only in vitro but also in cells (Figure S6). Moreover, none of the
studied quinazolines inhibited autophosphorylation of EphB1, a
close homolog of EphB3 that does not contain a cysteine at the
position corresponding to C717 of EphB3, in cells at 10 µM (Fig-
ure S7). Taken together, these results suggest that 3 and 6 are
potent and selective inhibitors of EphB3 in cells.
Figure 2. Electrophilic quinazolines cause covalent and irreversible
inhibition of EphB3 in vitro. (A) Screening of the quinazolines (at 2
μM) for inhibition against the EphB3 kinase. Kinase activity was nor-
malized to the DMSO control. (B) Inhibition of EphB3 by 3 (10 μM)
and 6 (10 μM) could not be reversed by washout as that of 1 (100
μM). Activity was normalized to the DMSO control (mean ± S.D.
from triplicates). (C) Tandem-mass spectrum of the 6-modified C717-
containing triply charged tryptic peptide SRPVMILTEFMENCcom-
pound6ALDSFLR measured at m/z 916.77374 (mass accuracy: 1.0
ppm). Fragment ions (y and b) are annotated in the spectrum.
Potent Inhibition of EphB3 in Cells. Having established
the potent and selective inhibition of the EphB3 kinase by the
electrophilic quinazolines in vitro, we went on to investigate the
inhibition of the full-length receptor by these compounds in cells.
After transfecting EphB3 into HEK293 cells, we verified that the
resulting cells expressed EphB3 and responded to the treatment
with ephrin-B2 by inducing autophosphorylation of the receptor
(Figure 3A). Screening the quinazolines at a concentration of 10
μM revealed that 2, 3 and 6 inhibited EphB3 autophosphoryla-
tion to different degrees (Figure 3A). Specifically, 3 and 6 abol-
ished the EphB3 phosphorylation while 2 caused partial blockade.
The SAR of the quinazoline series at inhibiting EphB3 activity in
cells is consistent with that from the in vitro studies. Furthermore, 3
and 6 inhibited the autophosphorylation of EphB3 in a dose-
dependent manner with apparent EC50 values of 300 nM and 170
Figure 4. Crystal structures of EphB3 in complex with electrophilic
quinazolines verify the formation of a covalent bond to C717. (A)
Crystal structure of the EphB3-3 complex. (B) Crystal structure of the
EphB3-6 complex. (C) Overlay of the two structures. |2Fo-Fc| elec-
tron density is shown as blue mesh (contoured at 1 σ) in (A) and (B).
The protein is shown in cartoon representation while the ligand and
key residues in the hinge region and the DFG motif are shown as
sticks.
We next sought to improve the potency of EphB3 inhibition by
modifying the inhibitor structure, particularly the anilino group at
the 4 position. On the basis of reported SAR for a series of EphB4
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 8
inhibitors and docking analysis,31 we synthesized a few analogs of
contain a quinazoline core.29 In contrast, the optimized inhibitor
1
2
3
4
5
6
7
8
6 (Scheme S2). Among these, one derivative 9 that contains chlo-
ro and hydroxyl substituents on the phenyl ring improved potency
dramatically (Figure 3C and 3D). Dose-response study indicated
that 9 had an EC50 value of 3 nM at inhibiting EphB3 autophos-
phorylation in cells. Thus the subtle modifications to the aniline
group dramatically improved the potency of 7-chloroacetamido-
quinazoline’s inhibition against EphB3.
9 exhibited more specific and potent inhibition of EphB3 (Figure
5). At 50 nM, 9 caused 98% inhibition of EphB3 with only two
other kinases (RIPK2, and PHKG2) being inhibited by over 20%
in the panel. Sequence and structure analyses revealed an absence
of cysteine residues near the end of the hinge region in RIPK2
and PHKG2, suggesting that their binding to 9 is driven by non-
covalent interactions. These results indicate that our chemical
optimization led to not only improved potency but also exquisite
specificity at inhibiting EphB3 within the human kinome.
Crystal Structures of EphB3 Bound to Inhibitors. To
elucidate the binding interactions between our inhibitors and
EphB3 kinase, we solved the co-crystal structures of the EphB3
kinase domain with 3 and 6. After being expressed and purified to
homogeneity as previously described,28 the EphB3 kinase was
incubated with a 3-fold molar excess of 3 and 6 before crystalliza-
tion. EphB3-3 and EphB3-6 crystals diffracted to resolution of 1.9
Å and 2.3 Å respectively, and belonged to space group P212121
(Table S2). The structures were solved by molecular replacement
using the apo structure of EphB3 as the model.28 The protein
conformations are similar to the apo structure (RMSD for the α
carbon of 260 amino acids of EphB3-3 and EphB3-6 compared to
apo kinase are 0.63 Å and 0.50 Å, respectively). As observed for
the apo structure, EphB3 is in the inactive DFG-out confor-
mation. The co-crystal structures revealed strong electron density
between C717 and the electrophilic carbon in 3 and 6, and thus
unambiguously verified the formation of covalent bonds (Figure
4A and 4B). Interestingly, despite the different sizes of the electro-
philes in the two inhibitors, their quinazoline cores are virtually
superimposable to each other (Figure 4C), supporting the notion
that the binding of the quinazoline to the EphB3 active site is
crucial for directing the electrophile for reaction with the thiol
group of C717.
9
PINK1 and LBK1, the two kinases that appear to contain a
cysteine at the position homologous to C717 in EphB3, are not
predicted to be inhibited by the electrophilic quinazolines because
of a deletion in their hinge region from EphB3 (Figure 1A). These
two kinases were not among the profiled panel of kinases however.
We thus measured the inhibition of LBK1 by 6 and 9 in vitro.
Dose-response study indicate that 6 and 9 both have IC50 values
over 10 μM against LKB1 thus confirming that our inhibitors do
not potently inhibit kinases that contain a shorter hinge region
despite the presence of a homologous cysteine (Figure S8).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6. Inhibition and specific labeling of EphB3 by the clickable
probe 10 in cells. (A) 10, whose structure is shown at the top, inhibit-
ed the autophosphorylation of EphB3 WT in a dose-dependent man-
ner with an approximate IC50 of 10 nM. (B) Dose-dependent labeling
of EphB3 in HEK293 cells as revealed by in-gel fluorescence imaging.
(C) Time-dependent labeling with 10 at 100 nM of EphB3 in
HEK293 cells as revealed by in-gel fluorescence imaging. Coomassie
blue staining shows equal protein loading in (B) and (C).
Figure 5. Graphical representations of the kinase inhibition profiles
for 6 and 9. (A) The two major targets of 6 at 300 nM are EphB3 and
EGFR with the latter being inhibited more. (B) 9 at 50 nM causes
foremost inhibition of EphB3 among the tested kinases. Only two
other kinases, RIPK2 and PHKG2, are significantly inhibited by 9,
but to a much lesser extent. Percentage inhibition of a diverse panel of
98 kinases is mapped to the human kinome tree to visualize inhibitor
selectivity. The larger the circles are, the stronger the inhibition. Ki-
nase dendrogram was adapted and reproduced with permission from
Cell Signaling Technology Inc. (http://www.cellsignal.com)
A Clickable Probe to Monitor Target Engagement.
When used at high concentrations, electrophilic compounds may
nonspecifically label cellular proteins, especially the abundant
ones.32,33 To confirm target engagement and determine proteomic
selectivity of our inhibitors, we chose to derivatize 9 with a termi-
nal alkyne group, which is amenable to the Copper-Catalyzed
Azide-Alkyne Cycloaddition (CuAAC) click reaction. Based on
the co-crystal structure (Figure 4), we opted to attach the alkyne
tag at the 6 position of the quinazoline that is solvent exposed and
predicted to be accessible for the click reaction. The resulting
quinazoline probe 10 was synthesized in 7 steps (Scheme S3). 10
still caused potent inhibition (IC50 ~ 10 nM) of EphB3 activity in
cells suggesting that the alkyne attachment caused only a slight
decrease in potency (Figure 6A). We first employed 10 to monitor
its engagement of the target protein EphB3 in cells. After the
treatment of EphB3-expressing cells with 10 at various concentra-
tions, the cells were lysed, clicked with TAMRA-azide, resolved
by SDS-PAGE, and scanned for fluorescence image.33 We ob-
served only one major fluorescent band of ~110 kDa in the gel
when cells were treated with up to 100 nM of 10 (Figure 6B). The
fact that this band is absent in the parental HEK293 cells that do
not express EphB3 strongly suggests that the probe-labeled pro-
Kinome Inhibition Profiling. To determine the inhibitors’
selectivity profiles, we tested 6 and 9 for inhibition against a panel
of 98 diverse human kinases that cover all major branches of the
kinome phylogenetic tree using the SelectScreen Kinase Profiling
Services from ThermoFisher. The results indicate that 6 at 300
nM only significantly inhibited two kinases, EphB3 and EGFR,
out of the 98 (Figure 5). However, inhibition of EGFR by 6 ap-
pears to be stronger than that of EphB3, the target of our interest.
This off-target inhibition is probably due to the extremely high
affinity of quinazoline for EGFR, reflected by the fact that the
majority of clinically used small-molecule inhibitors of EGFR
4
ACS Paragon Plus Environment

Page 5 of 8
Journal of the American Chemical Society
tein is EphB3 (Figure S9). The sensitivity of this small-molecule
with a previous study using EGFR inhibitors.24
1
2
3
4
5
6
7
8
probe is highlighted by the fact that 10 detected low levels of en-
dogenous EphB3 proteins in two cell lines while an EphB3 anti-
body failed to do so (Figure S10). The low levels of EphB3 protein
that we detected in the adult cell lines are consistent with the ex-
pression pattern of EphB3 that is believed to peak during the em-
bryonic development, particularly in the brain.34,35 Higher con-
centrations (> 100 nM) of the probe caused nonspecific labeling of
other cellular proteins as expected for electrophilic compounds
(Figure 6B). Time-course experiment demonstrated that 10 could
engage EphB3 in cells within 5 min (Figure 6C). Together these
results suggest that 10 is a probe that can engage EphB3 potently,
specifically and rapidly in cells.
We further used 10 to measure the target engagement kinetics
of two types of kinase inhibitors. Type I kinase inhibitors are
known to bind exclusively to the ATP pocket while type II inhibi-
tors additionally bind to an allosteric pocket.36 Binding of type I
inhibitors to kinases is often rapid because the ATP pocket tends
to be readily accessible in the native state of the kinase. In con-
trast, type II inhibitors are often associated with slow binding
kinetics since the allosteric pocket becomes accessible only upon a
conformational change involving an outward movement of the
conserved DFG motif.37 Pre-treatment of EphB3-expressing cells
with 30 nM of a type I inhibitor dasatinib for various lengths of
time followed by addition of 10 showed that inhibition of EphB3
labeling was insensitive to preincubation duration indicating in-
stantaneous target engagement by dasatinib in cells (Figure 7D).
Analogous experiments with BHG-712 indicated that a preincu-
bation of over 10 min was required to effectively block EphB3
labeling by 10. A similar pre-incubation is also required for BHG-
712 to effectively compete off the labeling of EphB3 by 10 in cell
lysate (Figure S11). These results suggest that it takes over 10 min
for BHG-712 to engage the target of EphB3 in both cell lysate and
live cells consistent with the prediction that BHG-712 is a type II
kinase inhibitor based on its structural features. Furthermore, they
highlight that our probe can be used to monitor the target en-
gagement kinetics of different types of kinase inhibitors in cells,
which is difficult to achieve with other methods.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 7. Competition experiments reveal the extent and kinetics of
EphB3 engagement by different types of inhibitors in HEK293 cells.
(A) Dose-dependent occupancy of EphB3 by 9 in cells as revealed by
pulse-chase experiments and in-gel fluorescence imaging. 10 was used
at 100 nM. (B) Increasing concentrations of BHG-712 could compete
off EphB3 labeling by 10 (10 nM) in cells with an approximate IC50 of
260 nM. (C) BHG-712 blocked EphB3 autophosphorylation in a
dose-dependent manner in cells with an approximate IC50 of 320 nM.
Cell lysate were resolved by SDS-PAGE, transferred to nitrocellulose
membrane, and immunoblotted with antibodies for pEphB3, EphB3,
and β-actin. (D) Competition experiments revealed different kinetics
of target engagement by two kinase inhibitors, dasatinib and BHG-
712, in cells. 10 nM of 10 was used. Fluorescence image from 1 repre-
sentative experiment is shown. Fluorescence intensity was quantified
for the 3 repeats, averaged and normalized to the DMSO treatment.
Coomassie blue staining shows equal protein loading in (A), (B) and
(D).
Figure 8. Use of a specific inhibitor 9 to elucidate the mechanism
underlying EphB3 autophosphorylation. HEK293 cells transfected
with two EphB3 variants were treated with 100 nM of 9 for 1 hr be-
fore lysis. Cell lysate were resolved by SDS-PAGE, transferred to
nitrocellulose membrane, and blotted with antibodies for pEphB3 and
total EphB3. Relative phosphorylation levels are shown underneath
the western blots. Band intensities for phospho- and total EphB3 were
quantified with their ratios normalized to that of the control (DMSO
treatment).
We next used 10 as a probe to measure the occupancy of
EphB3 by various inhibitors in live cells. Treating EphB3-
expressing HEK293 cells with various concentrations of an irre-
versible inhibitor 9 led to the occupancy of a fraction of the target
kinase. Subsequent addition of 100 nM of 10 would completely
label the remaining fraction of EphB3. After click reaction and gel
electrophoresis, the fluorescence intensity of the target band was
normalized to the mock treatment to read out the fraction of
EphB3 that was not occupied by 9. From the dose-response study,
we were able to determine the concentration at which 50% of the
target is occupied in cells (Figure 7A). This value of 6 nM is simi-
lar to the IC50 value (3 nM) measured for 9 at inhibiting EphB3
autophosphorylation in cells. Probe 10 was also used to measure
the occupancy of EphB3 by a reversible inhibitor BHG-712 in
cells. Since BHG-712 and 10 compete for binding to the same site
in EphB3, increasing concentrations of BHG-712 should diminish
EphB3 labeling by 10 with the extent of inhibition reflecting tar-
get occupancy by BHG-712. Indeed, we observed dose-dependent
inhibition of probe labeling of EphB3 by BHG-712. The IC50
value determined for probe labeling is 260 nM, which agrees well
with the IC50 value of BHG-712 at inhibiting EphB3 autophos-
phorylation (Figure 7B and 7C). Together, these results imply that
that extent of EphB3 occupancy by an inhibitor correlates well
with the extent of inhibition of its autophosphorylation, consistent
Elucidation of EphB3 Autophosphorylation Mecha-
nism. We set out to investigate the activation mechanism of
EphB3 using our specific inhibitors. We first demonstrated that
fusion of a stable cytochrome b562RIL (BRIL) domain to EphB3
did not significantly affect the ability of the receptor to autophos-
phoryate (Figure S12).38 Furthermore, a phospho-specific anti-
body that recognizes EphB3 phosphorylated at Y608 and Y614 in
the juxtamembrane segment revealed that 100 nM of 9 abolished
the autophosphorylation at those two sites of wild-type EphB3 but
had little effects on the C717S mutant. When two copies of wild-
type EphB3 (BRIL-tagged or not) were co-expressed in HEK293
cells, the autophosphorylation of both receptors at Y608 and
Y614 was abolished by 100 nM of 9 (Figure 8). As expected, the
autophosphorylation of two C717S mutants were not affected at
all by 100 nM of 9. With cells expressing one WT and one C717S
form of EphB3 (BRIL-fused or non-fused), addition of 100 nM 9
caused partial inhibition of both receptors (Figure 8). Our results
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 8
(8) Faruqi, M. Nat Rev Drug Discov 2012, 11, 747.
(9) Chiu, S.-T.; Chang, K.-J.; Ting, C.-H.; Shen, H.-C.; Li, H.; Hsieh, F.-
J. Carcinogenesis 2009, 30, 1475.
(10) Ji, X.-D.; Li, G.; Feng, Y.-X.; Zhao, J.-S.; Li, J.-J.; Sun, Z.-J.; Shi, S.;
Deng, Y.-Z.; Xu, J.-F.; Zhu, Y.-Q.; Koeffler, H. P.; Tong, X.-J.; Xie, D.
Cancer Research 2011, 71, 1156.
are consistent with the trans autophosphorylation model because
the cis model predicts selective blockade of autophosphorylation of
the WT receptor.
1
2
3
CONCLUSION
4
5
6
7
8
9
In summary, by covalently targeting a unique reactive feature
in a receptor tyrosine kinase EphB3, we were able to identify in-
hibitors that exhibited high selectivity toward EphB3 not only
among protein kinases but also in the entire proteome. Our rigor-
ous and thorough characterizations verified the exquisite specifici-
ty of these inhibitors, assuring their application in biological stud-
ies. We solved the crystal structures of EphB3 in complex with
covalent inhibitors for the first time. In addition to verifying the
covalent linkage, these covalent complex structures can guide the
design of analogous electrophilic inhibitors for targeting cysteine
in proteins. Beyond generating and characterizing these covalent
EphB3 inhibitors, we employed them to probe drug-target en-
gagement and investigate the mechanism of receptor activation in
an innovative way. Using a clickable probe, we were able to
measure the extent and kinetics of EphB3 engagement by different
types of kinase inhibitors in cells. One specific inhibitor of EphB3
allowed us to reveal that the autophosphorylation of the receptor
in the juxtamembrane segment occurred in trans. These exquisite-
ly specific inhibitors will accelerate pharmacological characteriza-
tion of the roles of EphB3 in various biological processes such as
neural development and tumorigenesis.
(11) Li, G.; Ji, X.-D.; Gao, H.; Zhao, J.-S.; Xu, J.-F.; Sun, Z.-J.; Deng, Y.-
Z.; Shi, S.; Feng, Y.-X.; Zhu, Y.-Q.; Wang, T.; Li, J.-J.; Xie, D. Nat
Commun 2012, 3, 667.
(12) Jägle, S.; Rönsch, K.; Timme, S.; Andrlová, H.; Bertrand, M.; Jäger,
M.; Proske, A.; Schrempp, M.; Yousaf, A.; Michoel, T.; Zeiser, R.;
Werner, M.; Lassmann, S.; Hecht, A. Proceedings of the National Academy of
Sciences 2014, 111, 4886.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(13) Binns, K. L.; Taylor, P. P.; Sicheri, F.; Pawson, T.; Holland, S. J.
Molecular and Cellular Biology 2000, 20, 4791.
(14) Downward, J.; Parker, P.; Waterfield, M. D. Nature 1984, 311, 483.
(15) Walton, G. M.; Chen, W. S.; Rosenfeld, M. G.; Gill, G. N. Journal of
Biological Chemistry 1990, 265, 1750.
(16) Ballotti, R.; Lammers, R.; Scimeca, J. C.; Dull, T.; Schlessinger, J.;
Ullrich, A.; Van Obberghen, E. The EMBO Journal 1989, 8, 3303.
(17) Islam, K. ACS Chemical Biology 2015, 10, 343.
(18) Lampson, M. A.; Kapoor, T. M. Cell 2006, 126, 827.
(19) Qiao, L.; Choi, S.; Case, A.; Gainer, T. G.; Seyb, K.; Glicksman, M.
A.; Lo, D. C.; Stein, R. L.; Cuny, G. D. Bioorganic & medicinal chemistry letters
2009, 19, 6122.
(20) Cohen, M. S.; Zhang, C.; Shokat, K. M.; Taunton, J. Science 2005,
308, 1318.
(21) Knight, Z. A.; Shokat, K. M. Cell 2007, 128, 425.
(22) Cohen, M. S.; Hadjivassiliou, H.; Taunton, J. Nat Chem Biol 2007, 3,
156.
(23) Henise, J. C.; Taunton, J. Journal of Medicinal Chemistry 2011, 54, 4133.
(24) 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.
(25) Zhou, W.; Ercan, D.; Chen, L.; Yun, C.-H.; Li, D.; Capelletti, M.;
Cortot, A. B.; Chirieac, L.; Iacob, R. E.; Padera, R.; Engen, J. R.; Wong,
K.-K.; Eck, M. J.; Gray, N. S.; Janne, P. A. Nature 2009, 462, 1070.
(26) Zhang, T.; Inesta-Vaquera, F.; Niepel, M.; Zhang, J.; Ficarro,
Scott B.; Machleidt, T.; Xie, T.; Marto, Jarrod A.; Kim, N.; Sim, T.;
Laughlin, John D.; Park, H.; LoGrasso, Philip V.; Patricelli, M.;
Nomanbhoy, Tyzoon K.; Sorger, Peter K.; Alessi, Dario R.; Gray,
Nathanael S. Chemistry & Biology 2012, 19, 140.
(27) Kwiatkowski, N.; Zhang, T.; Rahl, P. B.; Abraham, B. J.; Reddy, J.;
Ficarro, S. B.; Dastur, A.; Amzallag, A.; Ramaswamy, S.; Tesar, B.;
Jenkins, C. E.; Hannett, N. M.; McMillin, D.; Sanda, T.; Sim, T.; Kim, N.
D.; Look, T.; Mitsiades, C. S.; Weng, A. P.; Brown, J. R.; Benes, C. H.;
Marto, J. A.; Young, R. A.; Gray, N. S. Nature 2014, 511, 616.
(28) Overman, R. C.; Debreczeni, J. E.; Truman, C. M.; McAlister, M. S.;
Attwood, T. K. Protein Science 2014, 23, 627.
(29) Scheffler, M.; Di Gion, P.; Doroshyenko, O.; Wolf, J.; Fuhr, U. Clin
Pharmacokinet 2011, 50, 371.
(30) Witt, J. J.; Roskoski Jr, R. Analytical Biochemistry 1975, 66, 253.
(31) Bardelle, C.; Barlaam, B.; Brooks, N.; Coleman, T.; Cross, D.;
Ducray, R.; Green, I.; Brempt, C. L.-v. d.; Olivier, A.; Read, J. Bioorganic
& medicinal chemistry letters 2010, 20, 6242.
(32) Weerapana, E.; Wang, C.; Simon, G. M.; Richter, F.; Khare, S.;
Dillon, M. B. D.; Bachovchin, D. A.; Mowen, K.; Baker, D.; Cravatt, B. F.
Nature 2010, 468, 790.
(33) Lanning, B. R.; Whitby, L. R.; Dix, M. M.; Douhan, J.; Gilbert, A.
M.; Hett, E. C.; Johnson, T. O.; Joslyn, C.; Kath, J. C.; Niessen, S.;
Roberts, L. R.; Schnute, M. E.; Wang, C.; Hulce, J. J.; Wei, B.; Whiteley,
L. O.; Hayward, M. M.; Cravatt, B. F. Nat Chem Biol 2014, 10, 760.
(34) Birgbauer, E.; Cowan, C. A.; Sretavan, D. W.; Henkemeyer, M.
Development 2000, 127, 1231.
(35) Holder, N.; Klein, R. Development 1999, 126, 2033.
(36) Liu, Y.; Gray, N. S. Nat Chem Biol 2006, 2, 358.
(37) Pargellis, C.; Tong, L.; Churchill, L.; Cirillo, P. F.; Gilmore, T.;
Graham, A. G.; Grob, P. M.; Hickey, E. R.; Moss, N.; Pav, S.; Regan, J.
Nat Struct Mol Biol 2002, 9, 268.
(38) Chun, E.; Thompson, Aaron A.; Liu, W.; Roth, Christopher B.;
Griffith, Mark T.; Katritch, V.; Kunken, J.; Xu, F.; Cherezov, V.;
Hanson, Michael A.; Stevens, Raymond C. Structure 2012, 20, 967.
ASSOCIATED CONTENT
Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.xxxxxxx.
Experimental methods, additional figures and analyses.
AUTHOR INFORMATION
Corresponding Author
*E-mail: zhang.chao@usc.edu
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENT
We thank Prof. B. Wang (Case Western University) and F. Sicheri
(University of Toronto) for providing the constructs of EphB3 and
EphA4 respectively. We thank Prof. G. Parkash (University of
Southern California) for providing ephrin-B2 and M. Greenberg
(Harvard University) for anti-phosphoEphB3. We thank Prof. K.
M. Shokat and Ms. B. Levin (University of California, San Fran-
cisco) for help with mass spectrometry data collection. This work
was supported in part by National Science Foundation (CHE-
1455306), American Cancer Society (IRG-58-007-51), and Uni-
versity of Southern California.
REFERENCES
(1) Pasquale, E. B. Nat Rev Mol Cell Biol 2005, 6, 462.
(2) Kullander, K.; Klein, R. Nat Rev Mol Cell Biol 2002, 3, 475.
(3) Coulthard, M. G.; Morgan, M.; Woodruff, T. M.; Arumugam, T. V.;
Taylor, S. M.; Carpenter, T. C.; Lackmann, M.; Boyd, A. W. The American
Journal of Pathology 2012, 181, 1493.
(4) Kandouz, M. Cancer Metastasis Rev 2012, 31, 353.
(5) Xi, H.-Q.; Wu, X.-S.; Wei, B.; Chen, L. Journal of Cellular and Molecular
Medicine 2012, 16, 2894.
(6) Pasquale, E. B. Nat Rev Cancer 2010, 10, 165.
(7) Boyd, A. W.; Bartlett, P. F.; Lackmann, M. Nat Rev Drug Discov 2014,
13, 39.
6
ACS Paragon Plus Environment

Page 7 of 8
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 8
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
82x44mm (300 x 300 DPI)
ACS Paragon Plus Environment