Download Binding Studies of Type I, II, and III Kinase Inhibitors against Bcr

MOLECULAR
SENSING
AUTHORS
Fei Shen
Robert R. Lavieri
Richard J. Isaacs
Nathan Gilbert
Stephen Dotson
*Scot R. Weinberger
POSTER NOTE
Binding Studies of Type I, II, and III
Kinase Inhibitors against Bcr-Abl Kinase
using Back-Scattering Interferometry:
A Tale of Allostery
[email protected]
Molecular Sensing, Inc.
Nashville, Tennessee
* Corresponding Author
Figure 1. A model showing where
various inhibitors bind Bcr-Abl
kinase is shown. Note the large
distance between the ATP binding
pocket and site in which GNF-2/
GNF-5 binds. Resistance to imatinib
results from the emergence of point
mutations within the kinase domain
of the Bcr-Abl protein that reduce
the binding affinity of imatinib.
The most resistant mutations are
found in the P-loop of the protein
near residues that are in direct
contact with the drug. The degree
of resistance varies from a few fold
for some of the A-loop mutants to
complete resistance for the T315I
mutation which precludes imatinib
from binding. The steady rate of
developing resistance to imatinib
has suggested that new kinase
inhibitors could be of clinical value,
if they could override imatinib
resistance and bind with higher
affinity to Bcr-Abl.
INTRODUCTION
In addition to playing an essential role in cellular energetics, kinases
and their associated signaling pathways are principally responsible for
the regulation of intracellular processes. When abnormally expressed or
controlled, kinase activity can cause cellular dysregulation and contribute to the onset of several diseases, including cancer. Based on the
fundamental understanding of kinase malfunction in cancer biology, the
discovery of small organic molecules to alter kinase function has culminated in the development of targeted cancer therapy. However, limited
selectivity and the emergence of drug resistance remain fundamental
challenges for current modern medicinal chemistry research for the development of kinase inhibitors that are effective in long-term treatments.
Most known kinase inhibitors are Type I inhibitors, ATP-competitive
compounds such as staurosporine, erlotinib (Tarceva®) and dasatinib
(Sprycel®), that bind to the ATP binding site and hydrogen bond with the
hinge region of the kinase . Type II inhibitors are compounds which bind
partially in the ATP binding site and extend past the gatekeeper and into
an adjacent allosteric site that is present only in the inactive kinase conformation. Compared to Type I inhibitors, Type II inhibitors have been shown
to possess advantageous pharmacological properties, including improved
target specificity 2. As such, many Type II inhibitors currently on the
market, such as imatinib (Gleevec®), are very effective anti-cancer drugs.
Mutations resistant to classical ATP-competitive (Type I/II) inhibitors are
emerging at a rapid pace and often limit the success of newly available targeted cancer therapies. Such mutations often result in a steric
hindrance that obstructs inhibitor binding to the hinge
region of the ATP pocket. At present, there are more
than 50 mutation sites and more than 70 individual
mutations conferring different levels of imatinib resistance found within CML patients. More recently, a
number of Type III inhibitors that function via allosteric
modulation have demonstrated promise towards
addressing mutation dependent drug resistance. As
such, the need to identify and develop reversible
inhibitors that are resistant to such mutations and bind
with a high affinity is the focus of many academic and
industrial research projects.
MOLECULAR
SENSING
Binding Studies of Type I, II, and III Kinase Inhibitors against Bcr-Abl
Kinase using Back-Scattering Interferometry: A Tale of Allostery
Figure adapted from: Conn PJ, Christopoulos A, and Lindsley CW “Allosteric modulators of GPCRs: a novel
approach for the treatment of CNS disorders” Nat Rev Drug Discov. 2009 Jan;8(1):41-54
Figure 2. An allosteric modulator compound may exert its action by altering affinity or by altering efficacy. In the
case of affinity driven modulation one would expect to see a leftward (or rightward in the case of negative modulation) shift for the binding isotherm of the endogenous ligand. If no change in affinity of the endogenous ligand is
detected in the presence of the allosteric modulator, then it is likely an efficacy driven mechanism.
MATERIALS AND METHODS
Back-Scattering Interferometry
The Back-Scattering device is a micro-scale interferometer (see figure 3). The BSI device consists of
a HeNe laser, a microchip, and a CCD camera. The
microchip receives light from the laser and illuminates
the sample containing channel. As light passes into
the channel, interference fringe patterns arise. A
camera images the fringes as shown in Figure 3.
When molecules combine, the resultant complex
causes a change in molecular mean polarizability that
is measured as a fringe pattern shift. Monitoring the
change in fringe phase as a function of ligand concentration allows equilibrium dissociation constant, Kd,
measurements to be performed.
Setting up the BSI Assay
BSI Kd determinations are performed in end-point
fashion, with target and ligand pre-incubated to
establish equilibrium. Target and ligand concentrations are chosen to initiate pseudo-first order binding
conditions, for which the target is typically held at
sparing concentration and the ligand in excess to
insure against depletion during the binding process.
Page 2
Figure 3: Back-Scattering Interferometer schematic.
POSTER NOTE
Figure 4: BSI sample preparation.
Figure 4 illustrates the RI change for a constant
concentration of target A (light blue trace, constant RI),
the increase in RI as ligand B is increased (red trace)
as a control and finally, the binding isotherm RI curve
for the AB complex after mixing and equilibration of
A and B (dark blue trace). In practice, control B is run
simultaneously in a reference channel with complex AB
probed in the analytical channel. The difference is plotted as AB-B. A. A single site binding isotherm model
is fitted to AB to determine the binding maximum or
Bmax. Kd is then established as ½ Bmax.
Preparation of Bcr-Abl Kinase Target
Solutions and Kinase Inhibitors
Wild-type Bcr-Abl Kinase as well as H396P, M351T,
Q252H, and T315I mutants were sourced from Millipore (EMD Millipore, Darmstadt, Germany). All Bcr-Abl
kinases were expressed via baculovirus in Sf21 insect
cells and provided in aliquots of 10 ug of enzyme (27%
purity) in 100 uL of 50 mM Tris/HCl pH 7.5, 150 mM
NaCl, 270 mM sucrose, 1 mM benzamide, 0.2 mm
PMSF, 0.1 mm EGTA, 0.1% 2-mercaptoethanol, 0.03%
Brij 35 and kept frozen at -70o C until used.
Assay and diluent buffer was 8 mM MOPS, pH 7,
10 mM Mg Acetate, 0.2 mM EDTA, and 1% DMSO.
Kinase working solutions were created by diluting the
above noted stock solutions to concentrations that
approximated 1/50 of the target Kd for each system
(concentration range: 0.5 nM – 10 nM). See figure 8 for
kinase inhibitor preparation
Assay Sample Preparation and BSI
Analysis
BSI binding affinity determinations were performed
under equilibrium based conditions. For these Bcr-Abl
kinase – kinase inhibitor assays, enzyme concentration
range was varied from 0.5 nM to 10 nM, while kinase
concentration ranged from 50 pM – 125 mM. Samples
were prepared using 250 uL Eppendorf tubes, and
allowed to incubate for four hours at room temperature (typically 22o C).
BSI measurements were performed using a dual-channel BSI prototype system (Molecular Sensing, Inc.,
Nashville, TN, USA). Each sample was measured in
triplicate. Difference plots of assay minus control for
BSI response at each inhibitor concentration were
constructed using GraphPad Prism® (San Diego, CA,
USA). Single-site binding model fits were performed to
determine the binding maximum (Bmax) and Kd determined as ½ Bmax.
Page 3
MOLECULAR
SENSING
Binding Studies of Type I, II, and III Kinase Inhibitors against Bcr-Abl
Kinase using Back-Scattering Interferometry: A Tale of Allostery
Figure 5: Preparation of Kinase Inhibitors. Imatinib, disatanib, and nilotinib were purchased from LC Laboratories
(Woburn, MA, USA) and were first brought up as 50 uM working stocks in 100% DMSO. Dose response series were
created by diluting each working stock with MOPS buffer to establish the appropriate target concentration range for
each binding system using a 12-point doubling dilution approach (range: 50 pM – 125 uM).
Figure 6: GNF-2 and GNF-5 are fully allosteric inhibitors
of BCR-Abl; GNF-5 has improved pharmacokinetic
properties compared to GNF-2. Interestingly, combining
GNF-5 with ATP-competitive type inhibitors seems to be
a viable approach for targeting mutations such as T315I.
Note, an ATP-competitive type compound alone is not
effective against T315I.
DATA ANALYSIS
Kinase Inhibitor Binding Curve Analysis
The results for measurements of dasatinib, nilotinib,
and imatinib binding affinity to wild-type and mutant
Bcr-Abl Kinase are illustrated in figure 7. The overall
binding affinity for these systems is summarized in
Table 1. For each of these systems, resultant assays
produced a high degree of concordance between
replicate measurements (avg Kd % RSD < 25 %). Table
1 also lists the final concentration of Bcr-Abl Kinase
used in each assay. For most of these assay systems,
the total protein consumption was quite low [range 72
picomole ( 9 mg) – 0.156 picomole (0.02 mg)].
The binding affinity of Type I, II and III Bcr-Abl kinase
inhibitors with wild type and four mutant Bcr-Abl
kinases (H396P, M351T, Q252H, and T315I) were
measured using the label-free, free solution biosensor
Page 4
technology known as Back-Scattering Interferometry
(BSI). BSI successfully demonstrated facile determination of equilibrium dissociation constants (Kd) for
all systems, with a high degree of concordance with
competition assay derived IC50 results. Binding assays
were optimized within a couple of days of experimentation, and for the entire study, less than twenty
micrograms of total enzyme was consumed. These
results indicate that BSI binding studies both class I,
II, and III kinase inhibitors can easily be performed,
allowing for confirmation of target engagement as well
as direct binding assessment of type II and III kinase
inhibitors against inactive Bcr-Abl kinase. The latter
makes BSI an attractive biophysical technique for the
study of second and third generation kinase inhibitors
to address the challenges of kinase inhibitor drug
resistance.
S ignal (pixels )
POSTER NOTE
0.65
0.60
0.55
0.50
0.45
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00
-0.05
0.01
Wild Type
M351T
Q252H
H396P
T3151
0.1
10
100
1000 10000 100000
Log[Dasatinib], nM
7A
Signal (pixels)
1
0.65
0.60
0.55
0.50
0.45
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00
-0.05
0.1
[Bcr-Abl
Kinase] (nM)
Imatinib-WT
472 +/- 83
5
Imatinib- H396P
1228 +/- 92
Imatinib- T315I
4050 +/- 1001
Nilotinib-WT
12 +/- 1.8
Nilotinib- H396P
32 +/- 6.5
Nilotinib- T315I
761 +/- 100
Dasatinib-WT
Wild Type
M351T
Q252H
H396P
T3151
1
10
100
1000
0.5
1 +/- 0.24
Dasatinib-H396P
0.7 +/- 0.10
Dasatinib-T315I
87 +/- 6.7
.025
.
Table 1: BSI determined binding Kd of desatinib, nilotinib,
and imatinib against wild-type and mutant Bcr-Abl kinase.
Final Bcr-Abl kinase enzyme concentrations for each assay
system are also indicated. See text for further details.
10000 100000
Log[Nilotinib], nM
7B
0.35
Wild Type
M351T
Q252H
H396P
T3151
0.30
Signal (pixels)
Kd (nM)
Assay System
0.25
0.20
0.15
0.10
0.05
0.00
-0.05
100
1000
10000
100000
1000000
Log[Imatinib], nM
7C
Figure 7: BSI analysis of Dasatnib (5a), Nilotinib (5b) and
Imatinib (5c) binding against Bcr-Abl kinase Wt and H396P,
M351T, Q252H, and T315I mutants. All binding systems
achieved saturation and appropriate Kd determination
easily ensued. See text for further details.
Dasatinib
(nM)
Imatinib
(nM)
Nilotinib
(nM)
IC50
KD
IC50
KD
IC50
KD
Wild
Type
1.83
1.08
527
472
17.69
13.3
M351T
1.61
0.42
926 1086
7.8
2.7
Q252H
5.6
5.49
733
961
46.7
47.6
T315I
137
86.5
9221 4050
696
761
H3696P
1.95
0.7
1280 1228
42.6
32.6
Bcr-Abl
Kinase
Table 2: Comparison of BSI determined binding affinity for
desatinib, nilotinib, and imatinib for wild-type and mutant
Bcr-Abl kinase as compared to determined IC50 values
obtained by radio-labeled abl-tide assays.
Page 5
MOLECULAR
SENSING
Binding Studies of Type I, II, and III Kinase Inhibitors against Bcr-Abl
Kinase using Back-Scattering Interferometry: A Tale of Allostery
Comparison with Known IC50 Data
Table 2 compares the obtained binding equilibrium
constants for each system with previously reported
IC50 data as compiled by O’Hare (8). Figure 8 depicts
the correlation between BSI obtained Kd and IC50
values for the studied systems (linear fit R2 = 0.9744).
As is clearly demonstrated, BSI binding results highly
correlated with previously compiled kinase activity
inhibition assays.
Figure 8: Correlation of BSI determined Kd vs literature IC50
values. Determined Kd values for each kinase are in high
agreement with various kinase inhibitor assays. See text for
further details.
Negative allosteric modulation of Bcr-Abl by GNF-5
One of BSI’s strengths is that the
technique is binding site agnostic,
which allows for direct studies of
allosteric modulation.
The application of BSI to the study
of a target, such as Bcr-Abl, with
so much structural and functional
information available has yielded
a powerful, compelling case study
that shows both how BSI can be
utilized to study allosteric systems
of increasing levels of complexity
in vitro, and how BSI can provide
evidence in support of the
mechanism(s) of action of a small
molecule.
(Reference 1). Data from both
a crystal structure and hydrogen-deuterium exchange mass
spectrometry support a mechanism
in which GNF-5 causes an allosteric
perturbation of the amino acids
near T315 that should allow for
binding of an ATP-competitive
kinase inhibitor, even in the
presence of an isoleucine residue
(Reference 1). This model explains
how GNF-5 is able to sensitize
the T315I Bcr-Abl mutant to an
ATP-competitive kinase inhibitor,
but does not fully explain how
GNF-5 is able to inhibit the WT
Bcr-Abl kinase.
The mechanism of action of GNF-5
is particularly interesting from both
a biophysical point of view and
from a pharmacological point of
view. Unlike all other compounds
mentioned in this document GNF-5
does not bind anywhere near the
ATP-binding site of Bcr-Abl.
The crystal structure shown on
page 1 suggests that some of
structural changes caused by
GNF-5 binding to Bcr-Abl lock
the kinase into an inactive conformation and this may explain how
GNF-5 inhibits the WT kinase
(Reference 1 and Figure 1).
In fact, GNF-5 binds to a myristate
site approximately 30 Angstroms
away from the ATP-binding site
BSI was utilized to measure Kd
values for ATP and GNF-5, both
alone and in combination, for both
Page 6
WT Bcr-Abl and the T315I mutant
construct of Bcr-Abl (Figure 9).
Interestingly, in the presence of
a saturating amount of GNF-5
the binding curve for ATP on
WT Bcr-Abl is substantially right
shifted (Figure 9). This suggests
that GNF-5’s mechanism of action
against the WT enzyme may be to
substantially decrease WT Bcr-Abl’s
affinity for ATP. In contrast to the
WT enzyme, GNF-5 only slightly
right shifts the binding curve of the
T315I mutant for ATP (Figure 9).
The structural and functional
data in the literature indicate
that GNF-5 sensitizes the T315I
mutant to ATP-competitive kinase
inhibitors, therefore the addition of
an ATP-competitive inhibitor, such
as imatinib, to the Bcr-Abl-GNF-5
complex may be able to substantially right shift the ATP binding
curve. These BSI experiments are
currently in progress.
0.12
0.10
0.10
0.08
0.08
0.06
T315I_Without GNF-5
0.04
T315I_With 50 uM GNF-5
0.02
0.00
-0.02
1
9A
Si gnal (pixels)
Si gnal (pixels)
0.12
10
L og[ATP] , uM
100
0.06
0.02
0.00
1000
-0.02
0.12
0.10
0.10
0.08
0.08
0.06
Wild Type_Without GNF-5
Wild Type_With 50 uM GNF-5
0.02
0.00
-0.02
0.1
1
10
Log[ATP] , uM
1
10
Log[ATP] , uM
100
1000
9B
0.12
0.04
T315I
Wild Type
0.04
100
Si gnal (pixels)
Si gnal (pixels)
POSTER NOTE
Figure 9:
A) The Kd of ATP for both WT and T315I BCR-Abl is about
4-5 micromolar.
B) The Kd values of GNF-5 for WT and T315I BCR-Abl are
about 510 nM and 130 nM respectively.
C) Interestingly, in the presence of a saturating amount
of GNF-5 the binding curve for ATP is right-shifted
substantially. Without GNF-5 present ATP has a Kd of
about 3 micromolar, but the presence of GNF-5 shifts
the ATP binding curve about 25-fold to the right. This
may explain how GNF-5 functions. The compounds
mode of action may be to substantially decrease WT
BCR-Abl’s affinity for ATP.
D) In contrast to the WT BCR-Abl protein’s behavior, the
presence of GNF-5 only shifts the binding curve for
the T315I mutant about 3-fold to the right. Given that
GNF-5 seems to sensitize the T315I to previous generation inhibitors it is likely also necessary to include
a compound such as imatinib in this experiment to
observe a large rightward shift in the binding isotherm.
These experiments are currently in progress.
T315I
Wild Type
0.04
0.02
0.00
1000
9C
0.06
-0.02
0.01
0.1
1
Log[GNF-5], uM
10
100
9D
CONCLUSIONS
This study illustrates the utility of BSI in the analysis
of type I, type II, and type III inhibitors of wild-type
Bcr-Abl kinase as well as four different imatinib
resistant mutants. The overall assay development
process was quite facile, requiring only a couple of
days and consuming minimal amounts of precious
target. Assay fidelity was quite high, with substantial
agreement of results for three independent replicate
analyses for each system. Obtained BSI kinase inhibitor affinities agreed well with previously reported
IC50 values and are consistent with theoretical and
x-ray crystallographic studies of the same systems,
lending credence to the overall approach as a viable
means to study potency for kinase inhibitors. GNF-5
was observed to decrease ATP affinity for WT kinase,
but only marginally affected T315I. This may account
for GNF-5 efficacy against WT and failure to arrest
T315I kinase activity. Because BSI evaluates direct
target engagement independent of downstream
activity, BSI is aptly suited to advance efforts in the
discovery of new inhibitors to Bcr-Abl Kinase as well
as other valuable kinase targets of medical import.
Page 7
MOLECULAR
SENSING
POSTER NOTE
REFERENCES
1. Strebhardt K & Ullrich A (2008) Paul Ehrlich’s Magic Bullet Concept: 100 Years of Progress. Nature Reviews Cancer
8(6):473-480.
2. Backes A, Zech B, Felber B, Klebl B, & Muller G (2008) Small-molecule inhibitors binding to protein kinases. Part
I: exceptions from the traditional pharmacophore approach of type I inhibition. Expert Opinion on Drug Discovery
3(12):1409-1425.
3. van der Plas DC, Soekarman D, van Gent AM, Grosveld G, & Hagemeijer A (1991) bcr-abl mRNA lacking abl exon a2
detected by polymerase chain reaction in a chronic myelogeneous leukemia patient. Leukemia 5(6):457-461.
4. O’Dwyer ME & Druker BJ (2000) STI571: an inhibitor of the BCR-ABL tyrosine kinase for the treatment of chronic
myelogenous leukaemia. The Lancet Oncology 1:207-211.
5. Daub H, Specht K, & Ulrich A (2004) Strategies to overcome resistance ot targeted protein kinase inhibitors. Nature
Reviews Drug Discovery 3(12):1409-1425.
6. Shah NP, et al. (2002) Multiple BCR-ABL kinase domain mutations confer polyclonal resistance to the tyrosine kinase
inhibitor imatinib (STI571) in chronic phase and blast crisis chronic myeloid leukemia. Cancer Cell 2(2):117-125.
7. Corbin AS, et al. (2004) Sensitivity of oncogenic KIT mutants to the kinase inhibitors MLN518 and PD180970. Blood
104(12):3754-3757.
8. O’Hare T, Eide CA, & Deininger MW (2007) Bcr-Abl kinase domain mutations, drug resistance, and the road to a cure for
chronic myeloid leukemia. Blood 110(7):2242-2249.
9. Redaelli S, et al. (2009) Activity of bosutinib, dasatinib, and nilotinib against 18 imatinib-resistant BCR/ABL mutants.
Journal of clinical oncology: Official Journal of the American Society of Clinical Oncology 27(3):469-471.
Company
Headquarters
North America Contacts
European Office & Contact
Molecular Sensing, Inc. (MSI),
is a commercial stage drug
discovery tools and contract
research services company
with headquarters and drug
discovery services laboratories
in Nashville, Tennessee and
an R&D center in Los Gatos,
California, along with a European operations center near
Frankfurt, Germany.
Molecular Sensing, Inc.
111 10th Ave, South
Suite 110
Nashville, TN 37203
Technical Support
Jake Isaacs
+1 615 938-7049
[email protected]
ph +1 615-938-7050
fax +1 615-255-0094
Technical Sales
Julian Abery
+1 919-724-0946
[email protected]
Molecular Sensing GmbH
Am Frauwald 10
DE 65510 Idstein
Germany
+49 6126 229050
[email protected]
[email protected]
www.molsense.com
© Copyright 2015 by Molecular Sensing, Inc. All rights reserved.
European Operations
Wilt Peters
+49 171 7604450
[email protected]