Label-free and real-time cell-based kinase assay for screening selective and potent receptor tyrosine kinase inhibitors using microelectronic sensor array.

Abstract

Kinases are the 2nd largest group of therapeutic targets in the human genome. In this article, a label-free and real-time cell-based receptor tyrosine kinase (RTK) assay that addresses limitation of existing kinase assays and can be used for high-throughput screening and lead optimization studies was validated and characterized. Using impedance, growth factor-induced morphological changes were quantitatively assessed in real time and used as a measure of RTK activity. COS7 cells treated with epidermal growth factor (EGF) and insulin results in a rapid increase in cell impedance. Assessment of these growth factor-induced morphological changes and levels of receptor autophosphorylation using fluorescent microscopy and enzyme-linked immunosorbent assay, respectively, demonstrates that these changes correlate with changes in impedance. This assay was used to screen, identify, and characterize a potent EGF receptor inhibitor from a compound library. This report describes an assay that is simple in that it does not require intensive optimization or special reagents such as peptides, antibodies, or probes. More important, because the assay is cell based, the studies are done in a physiologically relevant environment, allowing for concurrent assessment of a compound's solubility, stability, membrane permeability, cytotoxicity, and off-target interaction effects.

Full text

Page 1

http://jbx.sagepub.com
Journal of Biomolecular Screening
DOI: 10.1177/1087057106289334
2006; 11; 634 originally published online Jul 20, 2006; J Biomol Screen
Josephine M. Atienza, Naichen Yu, Xiaobo Wang, Xiao Xu and Yama Abassi

Tyrosine Kinase Inhibitors Using Microelectronic Sensor Array
Label-Free and Real-Time Cell-Based Kinase Assay for Screening Selective and Potent Receptor
http://jbx.sagepub.com/cgi/content/abstract/11/6/634

The online version of this article can be found at:

Published by:
http://www.sagepublications.com

On behalf of:

Society for Biomolecular Sciences

can be found at:
Journal of Biomolecular Screening
Additional services and information for

http://jbx.sagepub.com/cgi/alerts
Email Alerts:

http://jbx.sagepub.com/subscriptions
Subscriptions:

http://www.sagepub.com/journalsReprints.nav
Reprints:

http://www.sagepub.com/journalsPermissions.nav
Permissions:
http://jbx.sagepub.com/cgi/content/refs/11/6/634
SAGE Journals Online and HighWire Press platforms):
(this article cites 39 articles hosted on the
Citations

© 2006 Society for Biomolecular Sciences. All rights reserved. Not for commercial use or unauthorized distribution.
at Semmelweis Univ of Medicine on March 6, 2008 http://jbx.sagepub.comDownloaded from

Page 2

634www.sbsonline.org© 2006 Society for Biomolecular Sciences
Label-Free and Real-Time Cell-Based Kinase Assay
for Screening Selective and Potent Receptor Tyrosine Kinase
Inhibitors Using Microelectronic Sensor Array
JOSEPHINE M. ATIENZA, NAICHEN YU, XIAOBO WANG, XIAO XU, and YAMAABASSI
Kinases are the 2nd largest group of therapeutic targets in the human genome. In this article, a label-free and real-time
cell-based receptor tyrosine kinase (RTK) assay that addresses limitation of existing kinase assays and can be used for high-
throughput screening and lead optimization studies was validated and characterized. Using impedance, growth
factor–induced morphological changes were quantitatively assessed in real time and used as a measure of RTK activity.
COS7 cells treated with epidermal growth factor (EGF) and insulin results in a rapid increase in cell impedance. Assessment
of these growth factor–induced morphological changes and levels of receptor autophosphorylation using fluorescent
microscopy and enzyme-linked immunosorbent assay, respectively, demonstrates that these changes correlate with changes
in impedance. This assay was used to screen, identify, and characterize a potent EGF receptor inhibitor from a compound
library. This report describes an assay that is simple in that it does not require intensive optimization or special reagents such
as peptides, antibodies, or probes. More important, because the assay is cell based, the studies are done in a physiologically
relevant environment, allowing for concurrent assessment of a compound’s solubility, stability, membrane permeability, cyto-
toxicity, and off-target interaction effects. (Journal of Biomolecular Screening 2006:634-643)
Key words:
impedance, cell-based assay, receptor tyrosine kinase, kinase, growth factor
INTRODUCTION
K
group. More than 500 different protein kinases have been iden-
tified, constituting ∼1.7% of the human genome; of these,
17% are known to be tyrosine kinases (TKs).1The receptor
tyrosine kinase (RTK) and nonreceptor forms (non-RTK) of
TK mediate important cellular processes including prolifera-
tion, survival, differentiation, metabolism, motility, and gene
expression.2It is therefore not surprising that deregulation of
TK activity as a result of loss of control in its regulatory pro-
teins or of mutations that increase expression of its transcript,
protein, or ligand is implicated in numerous pathological con-
ditions including cancer, inflammation, metabolism, and car-
diovascular diseases.
INASES ARE ENZYMES that catalyze the transfer of a phos-
phate group from nucleoside triphosphates to an acceptor
RTKs are type I transmembrane receptor proteins that trans-
mit extracellular signals intracellularly through the transfer of
phosphate groups from adenosine triphosphate (ATP) to tyro-
sine residues on protein substrates. Among the well-studied
RTKs are the epidermal growth factor receptor (EGFR/HER1)
group that also includes HER2,
Disregulation of EGFR is implicated in the pathogenesis of
many solid tumors, including breast, head-and-neck, non–
small-cell lung, renal, ovarian, and colon cancer, in which it
acts by promoting the proliferation and progression of malig-
nant cells.4,5Increased expression of HER2 or its amplification
correlates with breast carcinoma progression and poor prog-
nosis.6,7Similar to other RTKs, the binding of its ligand,
EGF, leads to activation and autophosphorylation of EGFR.
These phosphorylated tyrosine residues serve as docking sites
for intracellular adaptor proteins and kinases including Shc,
Grb2, PLCγ, and PI3K.8,9The binding of these proteins to
EGFR couples the extracellular signal to a network of intra-
cellular signaling pathways that control cell proliferation, sur-
vival, cytoskeletal changes, adhesion, motility, migration, and
invasion.
The central role of RTKs in these cellular processes and
their drugability has made them important and viable therapeu-
tic targets. Numerous strategies including the use of peptides,
HER3, and HER4.3
ACEA Biosciences, San Diego, California.
Received Oct 27, 2005, and in revised form Mar 28, 2006. Accepted for publi-
cation Apr 3, 2006.
Journal of Biomolecular Screening 11(6); 2006
DOI: 10.1177/1087057106289334

© 2006 Society for Biomolecular Sciences. All rights reserved. Not for commercial use or unauthorized distribution.
at Semmelweis Univ of Medicine on March 6, 2008 http://jbx.sagepub.comDownloaded from

Page 3

Label-Free, Cell-Based Kinase Assay
Journal of Biomolecular Screening 11(6); 2006www.sbsonline.org 635
antibodies, small molecules, antisense, or small interfering
RNA (siRNA) are being adapted to inhibit receptor tyrosine
kinase activity. This has resulted in the development and
approval of several inhibitors for kinases that include imatinib
for Bcr-Abl/PDGFR/c-Kit, trastuzumab for HER2, erlotinib for
EGFR, and bevacizumab for VEGFR. The importance of RTKs
as therapeutic targets also led to the development of numerous
screening platforms for the identification of inhibitors for RTK.
These include antibody-based technologies such as AlphaScreen,
time resolved-fluorescence resonance energy transfer (TR-
FRET), fluorescence polarization, time-resolved fluorescence
(TRF), scintillation proximity assay, Luminex, and enzyme-
linked immunosorbent assay (ELISA) and non–antibody-based
technologies such as incorporation of radioactivity, ATP con-
sumption, and measurement of change of substrate size and
charge. Although these technologies offer some advantages,
they are limited by one or more of the following factors: com-
plicated and tedious optimization steps, limited substrate
capacity, assay component interference, and expensive assay
components, all of which can affect the signal, throughput,
time, and utility of the assay. Therefore, there is a need for an
improved assay that addresses most of these limitations and
concurrently provides an additional ability to assess a com-
pound’s toxicity, specificity, selectivity, potency, and effective-
ness within a cellular context.
One of the intracellular pathways activated by RTK is the
induction of immediate morphological changes, as exemplified
by membrane ruffling, filopodia, and lamellipodia formations.10-13
EGF binding to EGFR induces these morphological changes
through the activation of Rac and Rho, members of the
Rho family of small GTPases.10,11Using scanning electron
microscopy (SEM), time-lapse imaging, or FRET-based imag-
ing, several studies have assessed cell morphology as a sensitive
assay for the integrated biological response elicited by EGF.14-17
However, the use of cell morphology as an output for a screen
to identify inhibitors of growth factor receptors or of its down-
stream signaling pathways is hampered by the intractability of
existing techniques used to measure morphological changes.
Taking advantage of the immediate morphological changes
induced by RTKs, we report on the development and utility of a
novel cell-based RTK assay that employs a sensor-based technol-
ogy to quantitatively measure RTK-induced morphological
changes using impedance. This method is based on the process of
monitoring a cell behavior, referred to as electric cell-substrate
impedance sensing.18This has been used to assess cytotoxic-
ity,19,20cell attachment and spreading,21,22migration,23and recep-
tor activity.24Using this quantifiable parameter, we show that
EGF-induced morphological changes in COS7 cells result in
changes in impedance that correlate with levels of RTK activa-
tion. We also show that these responses are unique to each growth
factor and are mediated by signaling pathways that regulate
growth factor–induced morphological changes. Finally, we
demonstrate the 1st use of impedance as a quantitative measure
of cell morphology to functionally screen, identify, and charac-
terize RTK inhibitors and downstream signaling regulators.
MATERIALS AND METHODS
Cell culture and reagents
COS7 cells were acquired from ATCC (Manassas, VA).
They were maintained in DMEM supplemented with 10%
fetal bovine serum and incubated at 37 °C with 5% CO2. EGF
(Sigma, St. Louis, MO), EGF inhibitors including 4557W
(4-(4-benzyloxyanilino)-6,7-dimethoxyquinazoline), compound 56
(4-[(3-bromophenyl)amino]-6,7-diethoxyquinazoline), and BPIQII
(8-[(3-bromophenyl)amino]-1H-imidazo[4,5-g]-quinazoline)
and signaling inhibitors including U73122, bisindolylmaleimide I
and wortmannin (Calbiochem, San Diego, CA), and LOPAC
enzyme inhibitor ligand set (Sigma) were resuspended and stored
according to the manufacturers’ instructions. The LOPAC library
contains compounds that inhibit a variety of enzymes including
oxidases, phosphodiesterases, phosphatases, and kinases. Among
the kinase inhibitors, the library contains inhibitors for TKs: tyr-
phostin AG879 (well number 20), genistein (well number 64),
tyrphostin A9 (well number 10); protein kinase C: chelerythine
chloride (well number 62), H-7 (well number 35), NPC-15437
(well number 37); MAP kinases: U0126 (well number 40),
PD098059 (well number 48), SP600125 (well number 15); PI3K:
LY-294002 (well number 16); and protein kinase A: HA-1004
(well number 54).
ELISA
Cells were plated on sensor plates at 1 × 104cells per well
and incubated overnight. During the day of the assay, cells were
serum starved in DMEM supplemented with 0.25% bovine
serum albumin (BSA) for a total of 4 h. If pretreated with
inhibitors, cells were preincubated with inhibitors during the
last hour of serum starvation and stimulated with growth fac-
tors for 15 min. After growth factor stimulation, cells were
washed 2 times with cold phosphate-buffered saline (PBS) and
lysed. ELISA (Biosource, Carlsbad, CA) assay to detect total
EGFR, and phospho-EGFR (P-1068) was performed according
to the manufacturer’s instructions and read at 450 nm.
Immunocytochemistry
Cells were manipulated similar to ELISA except for the
following changes. Cells were pretreated with 10 µM EGFR
inhibitor, 4557W, or vehicle for 1 h and stimulated with 25
ng/mL EGF. After 0, 5, and 20 min of EGF treatment, cells were
washed once with PBS and fixed in 4% paraformaldehyde for 15
min at 37 °C. Fixed cells were washed with PBS twice and per-
meabilized with 0.1% Triton X-100 for 10 min. Permeabilized

© 2006 Society for Biomolecular Sciences. All rights reserved. Not for commercial use or unauthorized distribution.
at Semmelweis Univ of Medicine on March 6, 2008 http://jbx.sagepub.com Downloaded from

Page 4

cells were washed with PBS, blocked with 3% BSA in PBS-T
(PBS with 0.1% Triton X-100) for 1 h at room temperature. To
probe for Rac, cells were treated with mouse anti-Rac antibody
(Sigma) and probed with tetramethylrhodamine isothiocyanate–
conjugated antimouse secondary antibody (Sigma). Cells were
then probed with fluorescein isothiocyanate–conjugated phal-
loidin (Sigma) to probe for actin. For each experimental well,
4 individual fields were measured using a digital camera attached
to a Nikon fluorescent microscope.
Data analysis
All dose-response curves were generated by plotting the aver-
age of percent of controls with standard deviations versus ligand
or inhibitor concentrations. The average percentage control was
calculated either relative to the positive control, samples treated
with growth factor plus vehicle, or to the negative control, sam-
ples untreated with growth factor, of quadruplicate samples and
fitted using a 4-parameter logistic nonlinear regression equation.
The EC50for ligands and IC50for inhibitors were determined
from the fitted curve generated by XLfit 4.0.
Impedance measurement of cell behavior
Throughout the experimental process, cells were continually
monitored, and changes in impedance were acquired with the
ACEA (San Diego, CA) real-time cell electronic sensing (RT-
CES) system. Cells were manipulated similar to ELISA except for
the following changes. COS7 cells were pretreated with vehicle,
DMSO,or the indicated inhibitor and then treated with 25 ng/mL
of growth factor. After ligand treatment, data acquisition of
changes in impedance was done every minute for up to 5 h.
The principle of impedance measurement using the RT-CES
has been previously described.19Briefly, gold sensors or elec-
trodes are arrayed on the bottom of culture wells. The elec-
tronic impedance of these sensors is determined primarily by
the ion environment at the electrode/solution interface and
the bulk solution. When cells are cultured on these sensors, a
change in the local ionic environment at the electrode/solution
interface occurs, leading to an increase in the sensor imped-
ance. Furthermore, a change in the amount of coverage of these
sensors as a result of changes in cell morphology or cell
number can also lead to changes in the local ionic environment
at the electrode/solution interface resulting in impedance
changes. The RT-CES system measures these changes in
impedance and displays it as a parameter called cell index. Cell
index was calculated according to the formula
,
where the number of the frequency points at which the imped-
ance is measured (e.g., N = 3 for 10 kHz, 25 kHz, and 50 kHz),
and R0(f) and Rcell(f) are the frequency electrode resistance
without cells or with cells present in the wells, respectively.
RESULTS
Growth factors induce specific morphological
changes that can be measured by impedance
The microsensors arrayed on the bottom of the plates used
in these experiments acutely measure changes in impedance.
This is determined by changes in the ionic composition in the
solution/sensor interface resulting from changes in cellular
adhesion or sensor area covered by cells due to increased cell
number or morphological changes.18To determine if imped-
ance can be used to measure growth factor–induced changes in
cell behavior, COS7 cells, which endogenously express several
RTKs, were treated with EGF and insulin, and cell index,
which is a measure of impedance change, was continuously
recorded. Addition of 25 ng/mL of EGF or insulin produced a
distinctive pattern of cell index change (Fig. 1A). This is char-
acterized by an immediate and very rapid increase in cell index,
with the maximum cell index change (peak cell index) occur-
ring within 20 min. For EGF, this is followed by a relatively
rapid decrease in cell index within an hour. However, for
insulin, the rate of decrease occurs even slower and returns to
baseline after several hours. To determine the specificity of
these responses, COS7 cells were pretreated for 1 h with 10 µM
of EGFR inhibitor, 4557W.25The rapid and immediate rise in
cell index as a result of EGF, but not of insulin, was blocked by
4557W (Fig. 1B). The absence of the 4557W effect on the
insulin response demonstrates that the unique changes in
impedance are a result of the specific response to EGF.
Growth factor treatment of cells can induce several molecular
and cellular responses including the induction of morphological
changes characterized by lamellipodia, membrane ruffle, and
filopodia formations.10,12,13Growth factor–induced lamellipodia
formation is mediated by small GTPases including Rac11and is
characterized by rapid actin polymerization leading to dynamic
reorganization of the cytoskeleton, which is manifested as thin
membrane protrusions that extend and retract continuously. To
demonstrate that the change in cell index, which is a measure of
change in impedance, correlates with growth factor–induced
morphological change, cells were stained with phalloidin and
an antibody against Rac. Cells treated with 25 ng/mL of EGF
showed an increase in cell area as a result of lamellipodia forma-
tion as early as 5 min and persists for at least 20 min (Fig. 2A-F).
The EGF-induced membrane protrusions stained positively with
phalloidin, confirming actin polymerization and remodeling
dynamically occurring in these structures. Intense staining of
Rac was also observed at the edge of these membrane protru-
sions. These morphological changes correlate temporally with
the onset and peak of cell index after EGF treatment (Fig. 1A).
CI = max
i=1,...,N
?Rcell(fi)
R0(fi)
− 1
?
Atienza et al.
636 www.sbsonline.orgJournal of Biomolecular Screening 11(6); 2006

© 2006 Society for Biomolecular Sciences. All rights reserved. Not for commercial use or unauthorized distribution.
at Semmelweis Univ of Medicine on March 6, 2008 http://jbx.sagepub.comDownloaded from

Page 5

The specificity of this response is supported by the significant
inhibition of morphological changes and Rac translocation in
cells pretreated with the EGFR inhibitor, 4557W (Fig. 2G-H).
This is consistent with the absence of changes in cell index mea-
sured in these samples (Fig. 1B). These observations demon-
strate that the growth factor–induced increase in impedance
correlates with growth factor–induced morphological changes,
which are characterized by an increase in cell surface area cov-
ering the sensors.
Changes in impedance as a result of growth
factor–induced morphological changes are
a quantitative measure of kinase activity
To determine if the EGF-induced impedance changes corre-
late with EGFR activation, cells were treated with a range of
EGF concentrations while impedance and levels of receptor
tyrosine autophosphorylation were concurrently measured.
Treatment of COS7 cells with increasing concentrations of
EGF resulted in a dose-dependent increase in measured imped-
ance change as reflected by the increase in normalized cell
index over the time course of measurement (Fig. 3A).
Concurrently, cells were treated for 15 min with EGF lysed,
and levels of phosphorylation on Tyr1068 (pTyr1068), mea-
sured by ELISA. Similar to the dose-dependent increase in cell
index, an increase in levels of EGFR phosphorylation were also
observed with treatment of increasing EGF concentrations
(Fig. 3B). The correlation of EGFR activation and changes in
impedance is supported by the similar EC50values calculated
Label-Free, Cell-Based Kinase Assay
Journal of Biomolecular Screening 11(6); 2006www.sbsonline.org637
B
0.8
0.9
1
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
-120.0 -60.00.0 60.0 120.0 180.0 240.0 300.0 360.0
Time (min)
Normalized Cell Index
+GF
Insulin + EGFR Inhibitor EGF + EGFR Inhibitor
0.8
0.9
1
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
-120.0 -60.0 0.060.0 120.0 180.0 240.0 300.0 360.0
Time (min)
Normalized Cell Index
Peak Cell
Index
+GF
A
EGFInsulin No Treatment
FIG. 1
insulin or epidermal growth factor (EGF). (A) COS7 cells treated with
insulin or EGF (arrow, +GF) showed unique responses characterized
by a sharp and transient increase in cell index, with peak cell index
(arrow) occurring within 30 min, and followed by a slow decrease in
cell index. Insulin-treated cells showed a slower rate of decrease in
cell index relative to EGF-treated cells. (B) COS7 cells were pre-
treated for 1 h with 10 µM EGF receptor (EGFR) inhibitor, 4557W, or
vehicle prior to growth factor treatment. Cells pretreated with 4557W
and stimulated with EGF did not a show the transient increase in cell
index, whereas the insulin response remained intact. Each curve is an
average of n = 4 ± SD.
Cell index curves of COS7 cells treated with growth factors,
FIG. 2.
EGF and ±4557W. (A, B) Cells pretreated with vehicle alone and
before treatment with epidermal growth factor (EGF) appeared rela-
tively small and exhibited low Rac staining. (C, D) Cells treated with
25 ng/mL EGF showed rapid formation of lamellipodia (arrows)
within 5 min as observed by an increase in the cell size. Intense stain-
ing of actin and Rac is observed at the lamellipodia and its edge
(arrows), respectively. (E, F) These changes continue to be observed
at 20 min. (G, H) Cells pretreated with 10 µM 4557W showed a phe-
notype similar to that of cells before treatment with EGF (A, B). They
appeared smaller, and a decreased staining for Rac is observed.
Fluorescent microscopy images of COS7 cells treated with

© 2006 Society for Biomolecular Sciences. All rights reserved. Not for commercial use or unauthorized distribution.
at Semmelweis Univ of Medicine on March 6, 2008 http://jbx.sagepub.com Downloaded from