Discovery and characterization of a cell-permeable, small-molecule c-Abl kinase activator that binds to the myristoyl binding site

Article abstract of DOI:10.1016/j.chembiol.2010.12.013

c-Abl kinase activity is regulated by a unique mechanism involving the formation of an autoinhibited conformation in which the N-terminal myristoyl group binds intramolecularly to the myristoyl binding site on the kinase domain and induces the bending of the αI helix that creates a docking surface for the SH2 domain. Here, we report a small-molecule c-Abl activator, DPH, that displays potent enzymatic and cellular activity in stimulating c-Abl activation. Structural analyses indicate that DPH binds to the myristoyl binding site and prevents the formation of the bent conformation of the αI helix through steric hindrance, a mode of action distinct from the previously identified allosteric c-Abl inhibitor, GNF-2, that also binds to the myristoyl binding site. DPH represents the first cell-permeable, small-molecule tool compound for c-Abl activation.

Full text of DOI:10.1016/j.chembiol.2010.12.013

Chemistry & Biology
Article
Discovery and Characterization of a Cell-
Permeable, Small-Molecule c-Abl Kinase Activator
that Binds to the Myristoyl Binding Site
2
1
1
1
2
1
Jingsong Yang,1, Nino Campobasso, Mangatt P. Biju, Kelly Fisher, Xiao-Qing Pan, Josh Cottom, Sarah Galbraith,
Thau Ho,2 Hong Zhang,2 Xuan Hong,2 Paris Ward,2 Glenn Hofmann,2 Brett Siegfried,2 Francesca Zappacosta,2
Yoshiaki Washio,2 Ping Cao,2 Junya Qu,2 Sophie Bertrand,2 Da-Yuan Wang,2 Martha S Head,2 Hu Li,2 Sheri Moores,1
Zhihong Lai,1 Kyung Johanson,2 George Burton,2 Connie Erickson-Miller,1 Graham Simpson,2 Peter Tummino,1
Robert A. Copeland,1 and Allen Oliff1,2
*
1Oncology Research and Development
2Platform Technology and Science
GlaxoSmithKline, 1250 South Collegeville Road, Collegeville, PA 19426, USA
*Correspondence: jingsong.2.yang@gsk.com
DOI 10.1016/j.chembiol.2010.12.013
SUMMARY
phosphorylation, modulate its kinase activity. c-Abl is a nonre-
ceptor tyrosine kinase that shares sequence homology with
c-Abl kinase activity is regulated by a unique mecha- the Src kinase family (Wang, 1993). Like other members of the
Src family, the N-terminal portion of c-Abl is composed of three
Src homology domains (SH1/catalytic, SH2, and SH3 domains).
nism involving the formation of an autoinhibited
conformation in which the N-terminal myristoyl
group binds intramolecularly to the myristoyl binding
site on the kinase domain and induces the bending of
the aI helix that creates a docking surface for the SH2
domain. Here, we report a small-molecule c-Abl
activator, DPH, that displays potent enzymatic and
cellular activity in stimulating c-Abl activation.
c-Abl also contains a unique N-terminal cap (ꢀ80 amino acids)
that plays an important role in the regulation of its kinase activity.
The C-terminal portion of c-Abl is composed of an array of
signaling and recognition domains that are involved in specific
cellular functions of c-Abl, including a DNA-binding domain,
an actin-binding domain, several docking sites for partner
proteins, and nuclear localization/export signals. c-Abl localizes
Structural analyses indicate that DPH binds to the
myristoyl binding site and prevents the formation of
to both the nucleus and cytoplasm. Evidence suggests that
nuclear c-Abl is involved in cell-cycle regulation and stress
the bent conformation of the aI helix through steric responses to DNA damage, oxidation, and ionization whereas
cytoplasmic c-Abl plays a role in morphogenesis, F-actin
dynamics, and signal transduction induced by extracellular
hindrance, a mode of action distinct from the previ-
ously identified allosteric c-Abl inhibitor, GNF-2,
that also binds to the myristoyl binding site. DPH
represents the first cell-permeable, small-molecule
stimuli (Taagepera et al., 1998; Wang, 1993; Van Etten, 1999;
Sirvent et al., 2008). There are two human c-Abl isoforms,
c-Abl 1a and 1b, derived from alternative splicing events (Shtivel-
tool compound for c-Abl activation.
man et al., 1986). They differ only in the N-terminal cap region.
c-Abl 1b is 19 amino acids longer and is myristoylated on the
N-terminal glycine, whereas c-Abl 1a is not myristoylated. This
difference has important implications in the regulation of the
INTRODUCTION
Protein kinases are a key group of regulators of biological c-Abl kinase activity.
processes, particularly the signal transduction pathways.
c-Abl is normally maintained in its inactive state in the cell.
Consequently, the regulation of kinase activities constitutes an Over the years, its mechanism of regulation remained largely
essential element in maintaining normal cellular functions. The unknown. Although its N-terminal region shares significant
activity of protein kinases can be modulated through several homology with the Src family kinases, it does not contain the
mechanisms such as subcellular localization, interaction with C-terminal tyrosine that plays an essential role in the regulation
accessory proteins and ligands, and phosphorylation/dephos- of Src kinase activity (Xu et al., 1997; Williams et al., 1997).
phorylation. One common mechanism of kinase activation is Recently, crystal structures of the N-terminal portion of c-Abl
the conformational reorganization of the activation loop located 1b have been resolved that shed light on the potential regulatory
between the N- and C-lobes of the kinase domain. This confor- mechanism of this enzyme (Nagar et al., 2003, 2006). This
mational change is prompted by phosphorylation of key portion of c-Abl contains minimally required components for
residue(s) on/near the activation loop. Additionally, global the proper regulation of the kinase activity. Based on the struc-
conformational changes and/or reorganization of the domain tural data, it was proposed that the c-Abl kinase activity is
structures can also contribute to kinase activation.
inhibited through the formation of an assembled conformation
In the area of kinase regulation, c-Abl kinase serves as an (Figure 1). In this state, the N-terminal myristoyl group binds
example of how conformational changes, coupled with protein to a deep and largely hydrophobic pocket, often called the
Chemistry & Biology 18, 177–186, February 25, 2011 ª2011 Elsevier Ltd All rights reserved 177

Chemistry & Biology
Cell-Permeable Small-Molecule c-Abl Activator
1994; Brasher et al., 2001; Hantschel et al., 2003). Furthermore,
mutations inside the myristoyl binding site that introduce bulky or
polar groups lead to c-Abl activation, presumably by affecting
the binding of the N-terminal myristoyl group to this site (Hant-
schel et al., 2003). N-terminal deletion through gene fusion
results in the constitutively active Bcr-Abl protein which is the
driver oncoprotein for the development of chronic myelogenous
leukemia (CML).
Since aberrant kinase activation is involved in many disease
states, especially in the onset and development of cancer,
kinase inhibition has been an active area of research for devel-
oping new therapeutics. Although less effort has been devoted
to the development of kinase activators compared to ATP-
binding inhibitors, some examples can be found in a recent
review (Simpson et al., 2009). It is becoming clearer that
specific kinase activators can not only serve as important tools
for enhancing our understanding of kinase regulation, but also
have therapeutic applications in a growing number of diseases.
Recent publications suggest a role of c-Abl activation in inhib-
iting mammary tumorigenesis and breast cancer cell mobility
and invasiveness (Allington et al., 2009; Noren et al., 2006).
c-Abl activation also has potential applications in treating neu-
tropenia since it has been suggested that c-Abl is critical in
maintaining normal myelopoiesis (Caracciolo et al., 1989; Rosti
et al., 1995). Therefore, we have selected c-Abl 1b (henceforth
referred to as c-Abl) as a model system for a kinase activator
screening effort. The goal of this exercise was to identify cell-
permeable, small-molecule tool compounds that potently
Figure 1. Autoinhibited, Assembled c-Abl Structure (PDB code:
1OPK) with a Myristoyl Group
Helix aI0 is highlighted in smudge green. An overlay of the c-Abl kinase domain
only structure (PDB code: 1M52) is colored yellow to highlight difference in
respective aI helixes with and without SH2 and N-terminal myristoyl.
myristoyl binding site, on the C-lobe of the kinase domain. activate c-Abl kinase activity both in vitro and in the cell. These
This binding helps form a compact conformation in which the specific c-Abl kinase activators will enable further analyses
SH2 domain docks onto the kinase domain and the SH3 domain of the complex c-Abl biology, especially of how activation of
interacts with both the kinase domain and the linker between the c-Abl differentiates functionally from the constitutively active
SH2 and the kinase domain. In this compact conformation, the Bcr-Abl.
structural plasticity of the N- and C-lobes required for catalysis
is restricted and the enzyme is in an autoinhibited state.
Through this work, we have identified a small-molecule, DPH,
that binds to the myristoyl binding site and leads to activation of
One key feature of this mechanism is the bending of the aI helix c-Abl kinase activity. This molecule also demonstrates robust
on the C-lobe of the kinase domain induced by N-terminal cellular activity in stimulating the phosphorylation of c-Abl
myristoyl binding to the myristoyl binding site. The bending of and its downstream substrate, Crk. The crystal structure of this
the aI helix starts at Phe516 and is followed by a short loop activator with the c-Abl kinase domain suggests a potential
and concluded by a new three-turn helix, called the aI’ helix (Fig- mechanism of activation that is consistent with the proposed
ure 1). This sharp bend of the aI helix is required to form a docking model. The significance and implications of this finding, along
surface for the SH2 domain. When the myristoyl group is dis- with
a recently published small-molecule c-Abl inhibitor
lodged from the myristoyl binding site, the aI helix then adopts (GNF-2) that binds to the same myristoyl binding site (Zhang
an extended and straight conformation (Figure 1) which clashes et al., 2010; Adrian et al., 2006), are discussed.
with the SH2 domain and prevents its docking onto the kinase
domain (Nagar et al., 2003). The resulting open conformation RESULTS
has an elongated shape with the SH2 domain residing on top
of the N-lobe of the kinase domain and the SH3 domain Discovery of c-Abl Activator, DPH
completely dissociated from the kinase domain (Nagar et al.,
A high-throughput screening (HTS) campaign against the
2006). c-Abl kinase in this open conformation has elevated GlaxoSmithKline (GSK) internal compound library was carried
kinase activity (partial activation) which triggers the autoactiva- out to identify c-Abl activators (Cottom et al., 2011). Unlike
tion cycle by phosphorylating Y412 on the activation loop and in vivo, c-Abl can be activated in the presence of ATP alone,
Y245 in the linker between the SH2 and the kinase domains. although at a very slow rate, in vitro. Therefore, the screening
Full activation of c-Abl is achieved following the phosphorylation was designed to look for small molecules that could further
of these two key tyrosine residues (Brasher and Van Etten, 2000). enhance the rate of c-Abl activation in the presence of ATP.
Various pieces of evidence support this model of c-Abl activa- The N-terminal catalytic domain of c-Abl enzyme used in this
tion. For example, mutations in the SH2 and SH3 domains that screening (c-Abl1-531) has been shown to be in its inactive state
could potentially disrupt domain interactions lead to c-Abl acti- and is fully activatable (see Figure S1A available online), hence
vation (Brasher and Van Etten, 2000; Mayer and Baltimore, suitable for activator screening.
178 Chemistry & Biology 18, 177–186, February 25, 2011 ª2011 Elsevier Ltd All rights reserved

Chemistry & Biology
Cell-Permeable Small-Molecule c-Abl Activator
Figure 2. Effect of DPH on c-Abl Activation
and Kinase Activity
(A) Structure of DPH. The chiral center is repre-
sented by an asterisk (*).
(B) Effect of DPH on c-Abl1-531 activation in the end
point c-Abl activation assay (Cottom et al., 2011).
A 3-fold dilution of DPH was carried out in DMSO
with a top concentration of 50 mM. The c-Abl
activation reactions were carried out as described
(Cottom et al., 2011) in the presence of various
concentrations of DPH. A pEC50 of 6.1 0.03 is
obtained from data fitting which corresponds to
an EC50 of 794 nM. Data are represented as
mean SD.
(C) Kinase activity of fully activated c-Abl was
measured in the presence of various concentra-
tions of DPH (50, 10, 2, 0.4, 0.08, and 0.016 mM)
in a filter binding activity assay (Cottom et al.,
2011). % activity = percentage of kinase activity
in the presence of DPH relative to the control
activity that was measured in the absence of
DPH which was set as 100%.
See also Figure S1.
Two in vitro c-Abl activation assays were employed for the the true potency of DPH, the continuous activation assay
characterization of c-Abl activators: an end-point activation (Omnia) was employed. This assay utilizes an Omnia peptide
assay (IMAP) used in HTS, and a continuous activation assay that emits fluorescence upon phosphorylation (Shults et al.,
(Omnia) described in the next section. The end-point IMAP assay 2006), allowing the continuous monitoring of c-Abl kinase activity
is a two-step assay. In this assay, c-Abl is first activated in the during its activation. This analysis takes into consideration the
presence of ATP and a small-molecule compound for 10 min. entire activation progress curve, therefore avoiding errors intro-
A peptide substrate is then added to measure the c-Abl kinase duced by taking arbitrary assay end-points. As expected, when
activity acquired in the first step (Cottom et al., 2011).
DPH (5-[3-(4-fluorophenyl)-1-phenyl-1H-pyrazol-4-yl]-2,4- reaction progress curve was observed since the reaction rate
preactivated c-Abl1-531 kinase was used in this assay, a linear
imidazolidinedione or 5-(1,3-diaryl-1H-pyrazol-4-yl)hydantoin) did not change over time (Figure 3A). However, unactivated
(Figure 2A) was identified as a c-Abl activator from HTS. A c-Abl1-531, when used in this reaction, was gradually activated
dose-response analysis of DPH in the end-point activation assay whichresulted inacurvilinear reaction progress curve (Figure 3A).
(IMAP) using the inactive c-Abl1-531 gave a pEC50 of 6.1 0.03 This progress curve contains the initial phase (inactivated), the
(EC50 = 794 nM) (Figure 2B). DPH also stimulates in vitro c-Abl transition phase (partially activated) and the final phase (fully acti-
phosphorylation (Figure S1B).
vated). The relaxation time (t) is defined by the rate at which the
reaction transitions from the initial to the final phase. When
DPH was titrated in this reaction, an activation effect was
Further Characterization of the Activation Effect of DPH
Since the end-point activation assay contains both c-Abl activa- observed by a dose-dependent shortening of the relaxation
tion and activity steps, activation by DPH could be a result of the time, indicating a faster rate of c-Abl activation (Figures 3A and
compound effect on c-Abl activation, its kinase activity, or both. 3B). The relaxation time at infinite DPH concentration (t min) was
To further dissect the mechanism of activation by DPH, high obtained by plotting the relaxation time against the DPH concen-
concentrations of c-Abl1-531 (2 mM) and ATP (500 mM) were tration and fitting the data to Equation 3 (Figure 3B). Since the t min
incubated for 1 hr to preactivate c-Abl1-531 and its activity was value was fitted by extrapolating [DPH] to infinity in which the full
then measured in the presence of DPH. Since c-Abl was fully potency of DPH is realized, the 4-fold activation determined in
activated, this reaction only measured the effect of DPH on the this analysis [(t max – t min)/ t min] represents the intrinsic activation
kinase activity of c-Abl. As shown in Figure 2C, DPH has no effect capacity of DPH.
on the kinase activity of c-Abl within the same concentration
range used in Figure 2B. This study indicates that DPH acts DPH Binds to the Myristoyl Binding Site of c-Abl
through stimulating c-Abl activation, rather than its kinase Based on the crystal structure, the N-terminal myristoyl group
activity.
of c-Abl binds to the myristoyl binding site on the C-lobe of the
One limitation of the end-point activation assay is that the kinase domain and this binding helps maintain the inactive
selection of assay end-point could affect the activator potency conformation of c-Abl (Figure 1). To determine if DPH induces
measured from the assay (Kenakin, 2006). In order to determine c-Abl activation by binding to the myristoyl binding site to
Chemistry & Biology 18, 177–186, February 25, 2011 ª2011 Elsevier Ltd All rights reserved 179

Chemistry & Biology
Cell-Permeable Small-Molecule c-Abl Activator
Figure 4. DPH Displaces the Peptide Ligand from the Myristoyl
Binding Site in the Fluorescence Polarization Assay
The reactions were carried out in the presence of 160 nM c-Abl46-534 (D382N),
30 nM peptide ligand and various concentrations of DPH (25 mM top concen-
tration, 3-fold dilution). The binding Kd from this analysis is 137 7 nM. Data
are represented as mean SD. See also Figure S2.
protein phosphorylation and enables the generation of a more
homogeneous protein prep. Binding of the peptide ligand to
the myristoyl site of c-Abl can compete with small molecules
that also bind to this site. The displacement of the peptide ligand
leads to a change in fluorescence anisotropy which can be used
to calculate the binding affinity of the small molecules.
The binding of the peptide ligand to c-Abl was confirmed in the
FP binding experiment with a binding Kd of 38 nM (Figure S2A).
When a control peptide with an identical amino acid sequence to
that of the ligand peptide, but lacking the myristoyl group, was
used in the binding assay, no anisotropy change was observed
(Figure S2B), indicating that the binding is specific and requires
the presence of the myristoyl group. When ATP was used in
the assay, no competition with the peptide ligand was observed,
indicating that the peptide ligand does not bind to the ATP
pocket (Figure S2C). These data, taken together, strongly
suggest that the peptide ligand binds specifically to the myristoyl
binding site on c-Abl.
Figure 3. Determination of the Intrinsic Activation Potency of DPH
in the Continuous c-Abl Activation Assay
(A) Progress curve analysis of c-Abl1-531 activation in the presence of DPH. In
this analysis, DPH and unactivated c-Abl were mixed in the presence of 50 mM
ATP and 26 mM Omnia peptide. The reaction progress curves were monitored
at various DPH concentrations (x, 0; +, 2.5; A, 7.6; >, 23; ;, 69; 7, 206; :,
617; D, 1852; -, 5556; ,, 16,667; C, 50,000 nM). Solid lines represent fitting
of the data to Equation 1. A preactivated c-Abl was included as a control (B)
which gave a linear reaction progress curve. From this analysis, the relaxation
time (t) of c-Abl activation was determined by Equation 2 at each DPH concen-
tration.
When DPH was tested in the FP competition binding assay,
a clear dose-response effect was observed with a fitted Kd of
137 7 nM (Figure 4). These data indicate that DPH interacts
with c-Abl in the myristoyl binding site. DPH is a racemic mixture
of two enantiomers. When these two enantiomers were sepa-
rated and tested individually in the FP binding assay and the
end-point c-Abl activation assay, they displayed different activ-
ities. One enantiomer was active in both assays with a binding
pIC50 of 6.4 (IC50 = 398 nM) and activation pEC50 of 6.6
(EC50 = 251 nM), whereas the other lost ꢀ20-fold activity in
both assays (a binding pIC50 of 5.1 and activation pEC50 of
5.3). These results suggest that a specific configuration of DPH
is required for its binding to the myristoyl binding pocket.
(B) The relaxation time was plotted against DPH concentration and fitted to
Equation 3 to generate a 4-fold intrinsic activation activity of DPH. The EC50
determined in this assay does not reflect the true binding affinity between
the activator and the enzyme since this parameter is also affected by the
kinetics of the c-Abl autoactivation reaction.
displace the terminal myristoyl group, a fluorescence polariza-
tion (FP) competition binding assay was utilized.
This assay used
a kinase-dead, N-terminal truncated
c-Abl46-534 (D382N) and a TAMRA-labeled peptide ligand that
bears the same sequence as c-Abl 1b N terminus (underlined)
with a myristoyl group (Myr-GQQPGKVLGDQRRPSL(K-TAMRA) Cocrystal Structure of DPH with the c-Abl
G-amide; Myr: myristoyl group). Truncation of the N terminus of Kinase Domain
c-Abl produces an empty myristoyl pocket for peptide ligand and The above FP binding data strongly suggest that DPH binds
small-molecule binding. The kinase-dead mutation prevents to the myristoyl binding site of c-Abl. To gain a better
180 Chemistry & Biology 18, 177–186, February 25, 2011 ª2011 Elsevier Ltd All rights reserved

Chemistry & Biology
Cell-Permeable Small-Molecule c-Abl Activator
understanding of the molecular interactions that govern the
binding, a cocrystal structure of the more active enantiomer of
DPH with the c-Abl kinase domain spanning residues 248 to
531 (c-Abl248-531) was obtained. Crystals diffracted to 1.85 A
˚
resolution with space group P212121 and cell parameters of a =
˚
˚
˚
74.39 A, b = 95.43 A, and c = 115.50 A and two molecules in
the asymmetric unit (Figure 5A and Table 1). The crystal structure
(Figure 5B) shows that the more active enantiomer has R config-
uration and thus is referred to as R-DPH. R-DPH binds to c-Abl in
the myristoyl binding site and straightens the aI’ helix to elongate
the aI helix (Figure 5A). In chain A, the aI helix extends to residue
519 and in Chain B, the aI helix extends to residue 529. The
remaining C-terminal residues of the respective molecules in
the asymmetric unit are disordered. In contrast, the autoinhibited
c-Abl has aI helix ending at residue 515, with a loop between
residues 515–521 and with aI’ helix defined by residues
521–529 (Figure 1). The elongated aI helix of Chain B with
R-DPH shows significant crystal packing and is an artifact of
crystallization. However, the C terminus of Chain A with R-DPH
shows no crystal packing and will be the focus of subsequent
discussion.
The fluorophenyl ring of R-DPH occupies a hydrophobic
pocket defined by Leu359, Leu360, Ala363, Val487, Leu448,
and Phe512 (Figure 5B). The 2-NH of the hydantoin ring of
R-DPH is hydrogen-bonded to the side chain of Tyr454 via
a mediating water molecule, while the 4-NH of the hydantoin
forms a direct hydrogen bond with the backbone carbonyl group
of Glu481. The pyrazole core of R-DPH has Van der Waals
contacts with Pro484. The phenyl ring extends toward the elon-
gated aI helix and sits in a lipophilic groove with Ala356 and
Pro484 defining respective walls of the groove. One edge of
the phenyl ring is solvent exposed.
Contrary to what is seen with the R-DPH bound structure
where the aI helix is in the extended conformation, c-Abl with
GNF-2 (Zhang et al., 2010) shows that the aI helix stops at
Phe516 and the aI’ helix defined by residues 521–529 packs
against GNF-2, similar to what is seen in the myristoyl-bound
autoinhibited c-Abl structure (Figure 1). An overlay of the
R-DPH structure with the GNF-2 bound c-Abl structure illus-
trates the significant conformational differences in the aI and
aI’ helices (Figure 5C). The phenyl ring of R-DPH occupies space
where aI’ helix would reside in the inhibited form of c-Abl with
GNF-2. These structural differences provide the basis for the
opposing effects of GNF-2 and DPH on c-Abl activity.
DPH Is Cell Permeable and Stimulates Full-Length
c-Abl Activation in HEK-MSRII Cells
So far, our data have demonstrated that DPH is a potent
activator of N-terminal kinase domain (1-531) of c-Abl in vitro.
(B) Interactions of R-DPH (yellow) and the c-Abl myristoyl site. Key amino acid
residues are rendered as sticks with labels. Key hydrogen bonding interactions
˚
are indicated with dashed lines and are annotated with distances (A). 2-NH
hydrogen bonds to a water molecule and 4-NH hydrogen bonds to E481. Fig-
ure inset (upper left) shows the electron density for R-DPH.
(C) Overlay of the cocrystal structures of R-DPH (yellow) and GNF-2 (cyan,
PDB code 3K5V) bound to the c-Abl kinase C-lobe domains (green and
orange, respectively). The dots represent the Van der Waals surface of I521
and V525 to illustrate the overlap with the space occupied by the phenyl group
of DPH.
Figure 5. The Cocrystal Structure of R-DPH and the c-Abl248-531
(A) Packing of crystal asymmetric unit. Chain B (cyan) shows a longer aI helix
ending at residue 529 and packing against chain A (green). The aI helix for
chain A is only visible to residue 519. The remaining residues, 520–531, are
disordered and not visible. DPH is rendered as yellow sticks for carbon, red
for oxygen, blue for nitrogen, and white for fluorine.
Chemistry & Biology 18, 177–186, February 25, 2011 ª2011 Elsevier Ltd All rights reserved 181

Chemistry & Biology
Cell-Permeable Small-Molecule c-Abl Activator
further confirm the effect of DPH on Y412 phosphorylation, we
immunoprecipitated c-Abl from cells treated with DPH or
DMSO and determined the change in Y412 phosphorylation
upon treatment with DPH by mass spectrometric analysis. A
2.5- to 3-fold increase in Y412 phosphorylation by DPH
(10 mM) treatment (apparent phosphorylation stoichiometry
ꢀ55%), as compared with cells treated with DMSO (apparent
phosphorylation stoichiometry ꢀ20%) was observed. The
cellular EC50 of DPH on Y245 phosphorylation was determined
to be 2.5 mM (pEC50 = 5.6 0.2) in HEK-MSRII cells by InCell
western assay (Figure 6C). Taken together, these data strongly
indicate that DPH is cell-permeable and can activate full-length
c-Abl in cells.
Table 1. Data Collection and Refinement Statistics for the
Cocrystal Structure of R-DPH and c-Abl248-531
R-DPH w/ c-Abl Kinase Domain
Data collection
Space group
P212121
Cell dimensions
˚
a, b, c (A)
74.39, 95.43, 115.50
90, 90, 90
a, b, g (ꢁ)
30–1.85 (1.9–1.85)a
0.048 (0.502)
41 (2.7)
˚
Resolution (A)
Rsym or Rmerge
I / sI
Completeness (%)
Redundancy
Refinement
99 (98.3)
5 (4.9)
DPH Induces the Phosphorylation of c-Abl Substrate Crk
In the above work, we demonstrated that Y245 and Y412 phos-
phorylation was induced upon treating the cells with DPH. To
test if DPH was able to activate endogenous c-Abl in cells, we
measured the phosphorylation of c-Abl substrate, Crk, in cells
that were treated with DPH. Crk is a well-characterized cellular
substrate of c-Abl. It is phosphorylated by active c-Abl. The
pCrk antibody we utilized recognizes pY221 of Crk and pY207
of Crk-L, a Crk isoform. We first treated KG-1, a human myeloid
cell line, with DPH and detected Crk-L pY207 by immunoblotting
as a readout for enhanced c-Abl kinase activity. In accordance to
our hypothesis, DPH induction of Crk-L pY207 in KG-1 cells was
dose dependent. Moreover, these inductions were completely
abolished by imatinib treatment, suggesting a dependence on
the c-Abl kinase activity (Figure 6D). We also developed
a western blot pCrk (pY221) assay for the detection of endoge-
nous c-Abl activation in HepG2 cell line. Consistent with c-Abl
pY245 InCell western analysis, a cellular EC50 of DPH of
2.5 mM (pEC50 = 5.6 0.2) was also observed in the HepG2
pCrk assay (Figure 6E). These results confirmed DPH as a cell-
permeable small-molecule activator of c-Abl.
˚
Resolution (A)
30–1.85
66950
No. reflections
Rwork / Rfree
No. atoms
Protein
0.189/0.204
4250
160
Ligand/ion
Water
215
B factors
Protein
33.9
35.6
33.7
Ligand/ion
Water
Rmsd
˚
Bond lengths (A)
0.007
1.076
Bond angles (ꢁ)
a Values in parentheses are for highest resolution shell.
To determine if DPH is cell-permeable and if it can activate full-
length c-Abl in the cell, the cellular effect of DPH was determined DISCUSSION
by measuring the phosphorylation status of Y245 and Y412 of
c-Abl upon DPH treatment as a surrogate for c-Abl activation In this paper, we describe a small-molecule activator (DPH) of
in cells. Such c-Abl phosphorylation can be induced by pervana- the c-Abl kinase that was identified through HTS of a ꢀ1.3 million
date or growth factor treatments. However, in our hands, compound library. We show that this molecule demonstrates
commercially available pY412 and pY245 antibodies were both enzymatic and cellular c-Abl activation activities. It stimu-
insensitive in detecting DPH-stimulated endogenous c-Abl lates phosphorylation of c-Abl and only affects the activation
phosphorylation by western blot (Figure 6A, lanes 1–4, and Fig- process of c-Abl with no effect on its kinase activity. Structural
ure 6B, lane 1). We hypothesize that the c-Abl phosphorylation is studies demonstrate the binding of this small-molecule to the
milder with DPH, than with more strenuous treatments, such as myristoyl binding site and suggest a potential mechanism of
pervanadate or growth factor treatments, that may be inducing c-Abl activation by DPH.
multiple pathways of activation of c-Abl. To overcome this diffi-
c-Abl activation is an extremely dynamic process marked by
culty, we overexpressed human c-Abl 1b (full length) in HEK- the disassembly of the compact autoinhibited conformation, fol-
MSRII cells by BacMam transduction (Cottom et al., 2011), lowed by the phosphorylation of two key Tyr residues (Y245 and
which improved western blot detection of pY245 and pY412 Y412). Based on structural data, the autoinhibited conformation
c-Abl in cell lysates. As shown in Figures 6A and 6B, DPH of c-Abl is maintained through the binding of the N-terminal
induced both pY245 and pY412 c-Abl phosphorylation. In addi- myristoyl group to the myristoyl binding site on the kinase
tion to pY245 and pY412 bands, these antibodies detected many domain. This event provides specific intramolecular interactions
nonspecific bands which were also induced by DPH treatment. for the formation of the bent conformation of the aI helix which is
Interestingly, the inductions of specific, as well as nonspecific, required for the docking of the SH2 domain onto the kinase
bands in response to DPH treatment were completely abrogated domain. These interactions appear to be mainly provided by
by imatinib (Figure 6A), suggesting that the nonspecific bands the myristoyl group with minimal contributions from the adjacent
may be either substrates or breakdown products of c-Abl. To tether peptide sequence, as a bent conformation of the aI helix is
182 Chemistry & Biology 18, 177–186, February 25, 2011 ª2011 Elsevier Ltd All rights reserved

Chemistry & Biology
Cell-Permeable Small-Molecule c-Abl Activator
Figure 6. DPH Activates c-Abl in Cells
(A) (B) DPH-induced phosphorylation of c-Abl Y245 (A) and Y412 (B). Immunoblots show the inductions in phosphorylation (pY245 and pY412) by DPH in
HEK-MSRII cells transduced with c-Abl. Imatinib inhibits Y245 phosphorylation induced by DPH.
(C) Dose response analysis of DPH on Y245 phosphorylation in HEK-MSRII cells transduced with c-Abl by InCell western assay (Cottom et al., 2011) gave a cellular
EC50 of 2.5 mM (pEC50 = 5.6 0.2). The integrated intensity (800/700 nm) was calculated as described elsewhere (Cottom et al., 2011). Data are represented as
mean SD.
(D) Imatinib-sensitive dose-dependent induction of pY207 Crk-L in KG-1 cells treated with DPH.
(E) Dose-dependent induction in pY221 Crk in DPH-treated HepG2 cells by western blot analysis. Ratios of pixel intensities of pY221 Crk/total Crk in DPH treated
to DMSO control derived from the western blot was fitted in a nonlinear curve graph to give a cellular EC50 of 2.5 mM (pEC50 = 5.6 0.2). Data are represented as
mean SD.
induced in the crystal structure of c-Abl46-534 with myristic acid identifies steric clashes between the left hand side phenyl moiety
added in trans (Nagar et al., 2003).
of R-DPH and the aI’ helix. These clashes can be clearly seen as
In principle, it is possible to find small molecules that bind to the phenyl moiety extending into the Van der Waals surface of
the myristoyl binding site and provide molecular interactions the aI’ helix, indicating that R-DPH is too bulky for aI helix to
that either stabilize or destabilize the bent conformation of the adopt the bent conformation (Figure 5C). As described above,
aI helix. Such molecules are expected to significantly affect the the bent conformation of the C-terminal helix to form aI’ provides
c-Abl activation process. The identification of DPH from HTS a necessary platform for the docking of the SH2 domain and
provided a unique opportunity for the study of c-Abl activation formation of the autoinhibited c-Abl conformation. Inability to
mechanism. The site of interaction of this molecule with c-Abl form this bent conformation therefore leads to c-Abl activation.
was mapped to the myristoyl binding site by both competition During the preparation of this manuscript, we encountered one
binding and crystallographic studies. Furthermore, detailed recent publication from a different group on the identification
structural analyses suggest a potential mechanism of activation of small-molecule myristoyl site-binding c-Abl activators through
by this molecule. The overlay of the structure of inactive c-Abl a NMR-based fragment screening (Jahnke et al., 2010). Further
with the structure of the c-Abl kinase domain bound to R-DPH structural analysis indicated that the binding of these molecules
Chemistry & Biology 18, 177–186, February 25, 2011 ª2011 Elsevier Ltd All rights reserved 183

Chemistry & Biology
Cell-Permeable Small-Molecule c-Abl Activator
to the myristoyl site prevented the aI helix from forming the bent SIGNIFICANCE
conformation. These results are in full agreement with our data
and further validate the mechanism of action of this type of Although some progress has been made in understanding
c-Abl activators. However, DPH differs from these activators in the biochemical mechanism of c-Abl kinase activation
two important ways. The published c-Abl activators have weak and identification of its cellular substrates, phenotypic
(micromolar) potency and show no cellular activity. In contrast, outcomes of c-Abl kinase activation in vivo remain elusive
DPH is more potent (with nanomolar potency), cell-permeable due to the lack of a viable method to activate endogenous
and demonstrates robust cellular c-Abl activation activity. To c-Abl function. Recently, it was suggested that the activa-
our knowledge, there are currently no cell-permeable tool tion of c-Abl/Crk pathway inhibits mammary tumorigenesis,
compounds available for c-Abl activation. Therefore, DPH repre- invasion, and metastasis (Allington et al., 2009; Noren et al.,
sents the first such tool compound with immediate applications 2006). It was also suggested that c-Abl is critical for the
in the study of in vivo c-Abl biology.
normal myelopoiesis process (Caracciolo et al., 1989;
While compounds binding to the enzyme active site almost Rosti et al., 1995). Therefore, development of small-molecule
invariably act as inhibitors simply by competing with the c-Abl activators allows further investigation of the
substrate, allosteric binders can elicit differential responses physiological as well as pathophysiological functions of
depending on the specific molecular interactions they induce. endogenous c-Abl. In this report, we identified a potent
Recently, a small-molecule inhibitor of c-Abl, GNF-2, was iden- cell-permeable c-Abl activator, DPH, that binds to the
tified that also binds to the myristoyl binding site (Adrian et al., myristoyl binding pocket of c-Abl, prevents the bending of
2006; Zhang et al., 2010). This molecule is structurally distinct the aI helix and docking of the SH2 domain, and induces
from DPH. Analysis based on the c-Abl/GNF-2 cocrystal struc- c-Abl activation. Hence, the mode of action of DPH confirms
ture suggests that GNF-2 interacts closely with the aI’ helix the previously hypothesized mechanism for c-Abl activation
and stabilizes its bent conformation (Figure 5C). Therefore, one (Nagar et al., 2003, 2006). We showed that, in addition to its
potential mechanism of GNF-2 inhibition could be its stabiliza- in vitro activity, DPH also demonstrates robust cellular
tion of the autoinhibited c-Abl conformation. Such interaction activity in c-Abl activation. Therefore, the identification of
complements the function of the myristoyl group, especially in DPH provides a valuable tool molecule for the study of the
the event that the N-terminal myristoyl group spontaneously functions of endogenous c-Abl in a multitude of biological
dissociate from the myristoyl binding pocket (Zhou, 2003).
Based on these observations, the following model can be
proposed to predict the effect of small-molecule modulators
that bind to the myristoyl pocket of c-Abl. The compound is
inhibitory if it stabilizes the bent conformation of the aI helix. It
is activating if it disrupts such bent conformation. Certain
compounds may bind in a way that neither disrupt nor reinforce
the bent conformation of the aI helix. These compounds act
merely as displacers of the myristoyl group and by doing so,
eliminate the stabilizing interactions on the bent conformation
of the aI helix induced by myristoyl group binding. Therefore,
such compounds are predicted to also act as c-Abl activators.
We have observed compounds of this nature in our lead optimi-
zation efforts and the results will be presented elsewhere.
In vivo, protein conformational states are stabilized/destabi-
lized not only by intramolecular forces, but also by interactions
with various ligands and protein partners. For example, strong
experimental evidence suggests that apart from the autoinhibi-
tion mechanism described so far in this paper, c-Abl inactive
state is likely also maintained by cellular binding proteins
that stabilize this autoinhibited conformation (Wang, 2004).
This complicates the study on c-Abl regulation because for
the most part, the nature of these proteins and their interac-
tions with c-Abl are unknown. However, it is likely that
a small-molecule activator identified in the in vitro setting could,
given sufficient potency, tip the balance between the active/
inactive states of c-Abl in vivo, trigger its autoactivation cycle,
and eventually lead to c-Abl activation. In support of this
hypothesis, we saw robust cellular c-Abl activation and Crk
phosphorylation upon DPH treatment. This potent cellular
activity of DPH enables further investigation of the roles of
c-Abl in processes such as myelopoiesis, tumorigenesis and
DNA damage responses.
processes.
EXPERIMENTAL PROCEDURES
Peptides used in the FP assay were custom synthesized by 21st Century
Biochemicals, LLC. Omnia peptide (Ac-KKGEAIYAAdP(Sox)G-NH2) was
purchased from Invitrogen (Omnia Y Peptide 6, Catalog # KNZ3061) (Shults
et al., 2006). All peptides were greater than 95% pure. Their identity was
confirmed by mass spectrometric analysis. All other reagents are of analytical
grade or higher.
Cloning, Expression, and Purification of c-Abl1-531, c-Abl46-534
(D382N) and c-Abl248-531 (kinase domain)
The full-length c-Abl lb cDNA (GenBank NM_007313) was generated by PCR
and cloned into pENTR/D-TOPO cloning vector as described by manufacturer
(Invitrogen Inc.). Cloning, expression, and purification of c-Abl1-531 has been
described elsewhere (Cottom et al., 2011). Cloning, expression, and purifica-
tion of c-Abl46-534 (D382N) and c-Abl248-531 were described in the Supple-
mental Experimental Procedures.
Continuous c-Abl Activation Assay
c-Abl activation was monitored continuously in a reaction mix containing ATP
(50 mM), the Omnia peptide and various concentrations of DPH (concentrations
indicated in text and/or figure legend) in 50 mM MOPS (pH 7.2), 2 mM MgCl2,
1 mM DTT, 0.1 mg/ml BSA, 0.01% Tween-20, and 25 mM NaCl. For the control
experiment, c-Abl1-531 (2 mM) was preactivated by incubating with 500 mM ATP
for 1 hr at room temperature. To initiate the activation reactions, c-Abl1-531 was
added at a concentration of 0.5 nM with a final reaction volume of 20 ml. The
reactions were run in a black 384 well plate (Corning Costar #3658) and the
product formation was monitored by a continuous fluorescence detection
assay (Omnia assay, Invitrogen). Phosphorylation of the Omnia peptide was
quantified using an Envision 2102 fluorescence plate reader (PerkinElmer)
with an excitation and emission wavelength of 380 and 486 nm, respectively.
The c-Abl activation progress curves were fitted in Equation 1 describing
intermolecular enzyme activation during catalysis (Wu and Wang, 2003) for
the determination of the activation rate constant, kobs, which was further con-
verted to the relaxation time, t, by Equation 2 (Wu and Wang, 2003).
184 Chemistry & Biology 18, 177–186, February 25, 2011 ª2011 Elsevier Ltd All rights reserved

Chemistry & Biology
Cell-Permeable Small-Molecule c-Abl Activator
È
À
ÁÉ
C
Cocrystal Structure of DPH with c-Abl Kinase Domain (c-Abl248-531
)
obs Et
P = ln 1 ꢂ a 1 ꢂ ek
+ b
(1)
(2)
–
531
k
c-Abl248
was incubated overnight with 1 mM DPH and concentrated
to 20 mg/ml in 20 mM Tris-HCl [pH 8.0], 100 mM NaCl, 3 mM DTT, and
5% (v/v) glycerol. Crystals were grown by hanging drop vapor diffusion with
200 nl protein and 200 nl reservoir solution at 4ꢁC. The reservoir solution
contained 0.4 M ammonium phosphate. Crystals grew in 1–5 days, were
transferred to reservoir solution with 15%–30% glycerol and flash-frozen in
liquid nitrogen.
ln a
t =
ða ꢂ 1Þkobs
E
In above equations, P is the amount of product, C is the specific activity of
fully activated c-Abl determined experimentally, kobs is the transition rate
constant, a is the fraction of active c-Abl at t = 0 (calculated as a ratio between
the basal activity and the full activity), E is total c-Abl concentration, t is time
and b is background.
˚
Diffraction data were collected at 100 K using 0.97872 A radiation at the
Advanced Photon Source at LifeScience CAT, beamline 21IDf using
a Mar225 CCD detector and then integrated and scaled with HKL2000 (Otwi-
nowski and Minor, 1997). Molecular Replacement solutions were found with
PHASER (McCoy et al., 2007) and refinement done with Phenix (Adams
et al., 2002). Model building was done with COOT (Emsley and Cowtan,
2004) and figures were made with PyMol (http://pymol.sourceforge.net/).
The t values thus determined were then plotted against DPH concentrations
and the data were fitted to a general 4-parameter EC50 equation (Equation 3) to
generate t max, t min, Hill coefficient (h) and the EC50 values. In this equation,
t
and t
represent maximum and minimum value of t, respectively,
min
max
and [A] is DPH concentration.
The fold of activation was calculated, fold of activation = (t max – t min)/ t min
.
Stimulation of c-Abl Y245/Y412 Phosphorylation
ꢀ
ꢀ
ꢁꢁ
in HEK-MSRII Cells by DPH
h
t = tmin
+
ðtmax ꢂ tminÞ= 1 + ð½Aꢃ=EC50
Þ
(3)
HEK-MSRII cells were transduced with BacMam c-Abl 1b virus. 48 hr after
transduction the cells were treated with DPH at 1–30 mM concentrations for
10 min. Cells pretreated with 5 mM of Imatinib for 10 min was used to identify
whether pY245 induction was depended on c-Abl kinase activity.
c-Abl Phosphorylation Gel Analysis
Total phosphorylation of c-Abl during activation was monitored by the incorpo-
ration of the 33P radiolabel onto the protein during the activation reactions with
[g-33P]ATP (10 mCi/ml, PerkinElmer). The c-Abl1-531 activation reactions were
carried out as described above in the presence or absence of DPH and
quenched at various times by the addition of the SDS-PAGE sample loading
buffer. Quenched samples were analyzed by SDS-PAGE and the amount of
33P incorporation in c-Abl was quantified by Phosphoimager (Storm,
MolecularDynamics).
Stimulation of Crk-L Y207 Phosphorylation in KG-1 Cells by DPH
KG-1 cells (4 3 106) were seeded in 1 ml of culture medium in 1.5 ml microcen-
trifuge tubes. Cells were incubated in a 37ꢁC water bath for 1 hr and treated
with DPH for 4 hr. After incubation, cells were pelleted by centrifugation at
1000 3 g for 5 min, washed twice in ice cold PBS, and subjected to cell lysis.
Stimulation of Crk Y221 Phosphorylation in HepG2 Cells by DPH
Cells (1.5 3 106) were plated in 60 mm dishes. One day after plating, the cells
were treated with DPH for 4 hr. The cells were then washed with ice cold PBS
and subjected to cell lysis.
Fluorescence Polarization Binding Assay
The FP binding assay was carried out in pH 7.2, 20 mM MOPS buffer contain-
ing 10 mM MgCl2, 2 mM CHAPS, 25 mM NaCl, 0.1% pluronic acid, 0.05 mg/ml
BSA, and 1 mM DTT. The final concentration of c-Abl46-534 (D382N) and the
peptide ligand (Myr-GQQPGKVLGDQRRPSL(K-TAMRA)G-amide) is 160 and
30 nM, respectively. The compound dilution was carried out in DMSO in
a 96-well, round-bottom polypropylene plate (Costar Catalog # 3359) and
was transferred (0.5 ml/well) to a black, round-bottom 384-well assay plate
(Costar catalog # 3658). DMSO (0.5 ml/well) was added for low control (peptide
ligand only) and high control (peptide ligand + c-Abl, no compound). Peptide
ligand in the above buffer was then added (5 ml/well) to the assay plate, fol-
lowed by the c-Abl mix (5 ml/well). Plate was vortexed slightly to mix and
was allowed to equilibrate for 30 min. Fluorescence anisotropy readout was
taken on an EnVision (PerkinElmer) 2102 fluorescence plate reader. The disso-
ciation constant (Kd) was determined using a method described previously
(Wang, 1995).
ACCESSION NUMBERS
Coordinates have been deposited in the Protein Data Bank with the accession
code 3PYY.
SUPPLEMENTAL INFORMATION
Supplemental Information includes Supplemental Experimental Procedures
and two figures and can be found with this article online at doi:10.1016/j.
chembiol.2010.12.013.
ACKNOWLEDGMENTS
Synthesis and Characterization of DPH
All authors are current or former employees of GlaxoSmithKline. This work was
funded by GlaxoSmithKline.
To a solution of 3-(4-fluorophenyl)-1-phenyl-1H-pyrazole-4-carbaldehyde (3 g,
11.27 mmol) and ammonium carbonate (4.33 g, 45.1 mmol) in ethanol (45 ml)
and water (45 ml), potassium cyanide (1.467 g, 22.53 mmol) in water (approx-
imately 5 ml) was added and stirred at 70ꢁC under nitrogen overnight. The
reaction was cooled in an ice bath, and then 2 M hydrochloric acid (30 ml,
60.0 mmol) was added dropwise to the reaction mixture. The solution was
left to stir for 3.5 hr. The mixture was added to EtOAc (200 ml) and extracted
with EtOAc. The EtOAc phase was washed with brine (6 3 200 ml). The solvent
was evaporated to give the crude material of 5-[3-(4-fluorophenyl)-1-phenyl-
1H-pyrazol-4-yl]-2,4-imidazolidinedione as a yellow solid (3.3 g, 87%). Fifty
milligrams of the crude material was purified by reverse-phase preparative
LCMS to give the title compound (DPH) as a colorless solid (35.6 mg).
1H NMR (400 MHz, DMSO-d6): d 10.85 (br. s., 1H), 8.68 (s, 1H), 8.29
(d, J = 0.8 Hz, 1H), 7.93 (dd, J = 0.9, 8.7 Hz, 2H), 7.79–7.85 (m, 2H), 7.48–
7.55 (m, J = 7.6, 8.7 Hz, 2H), 7.28–7.37 (m, 3H), 5.26 (d, J = 0.8 Hz, 1H); 13C
NMR (101 MHz, DMSO-d6): d 174.6, 162.3 (d, J = 245.1 Hz), 157.0, 150.5,
139.2, 130.2 (d, J = 8.8 Hz), 129.5, 128.7 (d, J = 3.2 Hz), 128.3, 126.6, 118.3,
116.8, 115.5 (d, J = 21.6 Hz), 53.5; 19F NMR (376 MHz, DMSO-d6): d ꢂ114.1
(proton decoupled); HRMS (m/z): [MH]+ calcd. for C18H14FN4O2,
337.1101; found, 337.1100.
Received: September 20, 2010
Revised: November 19, 2010
Accepted: December 6, 2010
Published: February 24, 2011
REFERENCES
Adams, P.D., Grosse-Kunstleve, R.W., Hung, L.W., Ioerger, T.R., McCoy, A.J.,
Moriarty, N.W., Read, R.J., Sacchettini, J.C., Sauter, N.K., and Terwilliger, T.C.
(2002). PHENIX: building new software for automated crystallographic struc-
ture determination. Acta Crystallogr. D Biol. Crystallogr. 58, 1948–1954.
Adrian, F.J., Ding, Q., Sim, T., Velentza, A., Sloan, C., Liu, Y., Zhang, G., Hur,
W., Ding, S., Manley, P., et al. (2006). Allosteric inhibitors of Bcr-abl-dependent
cell proliferation. Nat. Chem. Biol. 2, 95–102.
Allington, T.M., Galliher-Beckley, A.J., and Schiemann, W.P. (2009). Activated
Abl kinase inhibits oncogenic transforming growth factor-beta signaling and
tumorigenesis in mammary tumors. FASEB J. 23, 4231–4243.
Chemistry & Biology 18, 177–186, February 25, 2011 ª2011 Elsevier Ltd All rights reserved 185

Chemistry & Biology
Cell-Permeable Small-Molecule c-Abl Activator
Brasher, B.B., and Van Etten, R.A. (2000). c-Abl has high intrinsic tyrosine
kinase activity that is stimulated by mutation of the Src homology 3 domain
and by autophosphorylation at two distinct regulatory tyrosines. J. Biol.
Chem. 275, 35631–35637.
Rosti, V., Bergamaschi, G., Lucotti, C., Danova, M., Carlo-Stella, C., Locatelli,
F., Tonon, L., Mazzini, G., and Cazzola, M. (1995). Oligodeoxynucleotides anti-
sense to c-abl specifically inhibit entry into S-phase of CD34+ hematopoietic
cells and their differentiation to granulocyte-macrophage progenitors. Blood
86, 3387–3393.
Brasher, B.B., Roumiantsev, S., and Van Etten, R.A. (2001). Mutational anal-
ysis of the regulatory function of the c-Abl Src homology
3 domain.
Shtivelman, E., Lifshitz, B., Gale, R.P., Roe, B.A., and Canaani, E. (1986).
Alternative splicing of RNAs transcribed from the human abl gene and from
the bcr-abl fused gene. Cell 47, 277–284.
Oncogene 20, 7744–7752.
Caracciolo, D., Valtieri, M., Venturelli, D., Peschle, C., Gewirtz, A.M., and
Calabretta, B. (1989). Lineage-specific requirement of c-abl function in normal
hematopoiesis. Science 245, 1107–1110.
Shults, M.D., Carrico-Moniz, D., and Imperiali, B. (2006). Optimal Sox-based
fluorescent chemosensor design for serine/threonine protein kinases. Anal.
Biochem. 352, 198–207.
Cottom, J., Hofmann, G., Siegfried, B., Yang, J., Zhang, H., Yi, T., Ho, T.F.,
Quinn, C., Wang, D., Johanson, K., et al. (2011). Assay development and
high throughput screening of small molecular c-Abl kinase activators.
J. Biomol. Screen 16, 53–64.
Simpson, G.L., Hughes, J.A., Washio, Y., and Bertrand, S.M. (2009). Direct
small-molecule kinase activation: Novel approaches for a new era of drug
discovery. Curr. Opin. Drug Discov. Dev. 12, 585–596.
Emsley, P., and Cowtan, K. (2004). Coot: Model-building tools for molecular
Sirvent, A., Benistant, C., and Roche, S. (2008). Cytoplasmic signalling by the
graphics. Acta Crystallogr. D Biol. Crystallogr. 60, 2126–2132.
c-Abl tyrosine kinase in normal and cancer cells. Biol. Cell 100, 617–631.
Hantschel, O., Nagar, B., Guettler, S., Kretzschmar, J., Dorey, K., Kuriyan, J.,
and Superti-Furga, G. (2003). A myristoyl/phosphotyrosine switch regulates
c-Abl. Cell 112, 845–857.
Taagepera, S., McDonald, D., Loeb, J.E., Whitaker, L.L., McElroy, A.K., Wang,
J.Y., and Hope, T.J. (1998). Nuclear-cytoplasmic shuttling of C-ABL tyrosine
kinase. Proc. Natl. Acad. Sci. USA 95, 7457–7462.
Jahnke, W., Grotzfeld, R.M., Pelle, X., Strauss, A., Fendrich, G., Cowan-Jacob,
S.W., Cotesta, S., Fabbro, D., Furet, P., Mestan, J., and Marzinzik, A.L. (2010).
Binding or bending: distinction of allosteric Abl kinase agonists from antago-
nists by an NMR-based conformational assay. J. Am. Chem. Soc. 132,
7043–7048.
Van Etten, R.A. (1999). Cycling, stressed-out and nervous: cellular functions of
c-Abl. Trends Cell Biol. 9, 179–186.
Wang, J.Y. (1993). Abl tyrosine kinase in signal transduction and cell-cycle
regulation. Curr. Opin. Genet. Dev. 3, 35–43.
Wang, J.Y.J. (2004). Controlling Abl: auto-inhibition and co-inhibition? Nat.
Kenakin, T.P. (2006). A Pharmacology Primer (New York: Academic Press).
Cell Biol. 6, 3–7.
Mayer, B.J., and Baltimore, D. (1994). Mutagenic analysis of the roles of SH2
and SH3 domains in regulation of the Abl tyrosine kinase. Mol. Cell. Biol. 14,
2883–2894.
Wang, Z.X. (1995). An exact mathematical expression for describing compet-
itive binding of two different ligands to a protein molecule. FEBS Lett. 360,
111–114.
McCoy, A.J., Grosse-Kunstleve, R.W., Adams, P.D., Winn, M.D., Storoni, L.C.,
and Read, R.J. (2007). Phaser crystallographic software. J. Appl. Crystallogr.
40, 658–674.
Williams, J.C., Weijland, A., Gonfloni, S., Thompson, A., Courtneidge, S.A.,
Superti-Furga, G., and Wierenga, R.K. (1997). The 2.35 A crystal structure of
the inactivated form of chicken Src: a dynamic molecule with multiple regula-
tory interactions. J. Mol. Biol. 274, 757–775.
Nagar, B., Hantschel, O., Seeliger, M., Davies, J.M., Weis, W.I., Superti-Furga,
G., and Kuriyan, J. (2006). Organization of the SH3-SH2 unit in active and inac-
tive forms of the c-Abl tyrosine kinase. Mol. Cell 21, 787–798.
Wu, H., and Wang, Z.X. (2003). The mechanism of p21-activated kinase 2
autoactivation. J. Biol. Chem. 278, 41768–41778.
Nagar, B., Hantschel, O., Young, M.A., Scheffzek, K., Veach, D., Bornmann,
W., Clarkson, B., Superti-Furga, G., and Kuriyan, J. (2003). Structural basis
for the autoinhibition of c-Abl tyrosine kinase. Cell 112, 859–871.
Xu, W., Harrison, S.C., and Eck, M.J. (1997). Three-dimensional structure of
the tyrosine kinase c-Src. Nature 385, 595–602.
Zhang, J., Adrian, F.J., Jahnke, W., Cowan-Jacob, S.W., Li, A.G., Iacob, R.E.,
Sim, T., Powers, J., Dierks, C., Sun, F., et al. (2010). Targeting Bcr-Abl by
combining allosteric with ATP-binding-site inhibitors. Nature 463, 501–506.
Noren, N.K., Foos, G., Hauser, C.A., and Pasquale, E.B. (2006). The EphB4
receptor suppresses breast cancer cell tumorigenicity through an Abl-Crk
pathway. Nat. Cell Biol. 8, 815–825.
Otwinowski, Z., and Minor, W. (1997). Processing of X-ray diffraction data
Zhou, H.X. (2003). How often does the myristoylated N-terminal latch of c-Abl
collected in oscillation mode. Methods Enzymol. 276, 307–326.
come off? FEBS Lett. 552, 160–162.
186 Chemistry & Biology 18, 177–186, February 25, 2011 ª2011 Elsevier Ltd All rights reserved