ArticlePDF Available

Multiparameter In Vitro Assessment of Compound Effects on Cardiomyocyte Physiology Using iPSC Cells

Authors:

Abstract and Figures

A large percentage of drugs fail in clinical studies due to cardiac toxicity; thus, development of sensitive in vitro assays that can evaluate potential adverse effects on cardiomyocytes is extremely important for drug development. Human cardiomyocytes derived from stem cell sources offer more clinically relevant cell-based models than those presently available. Human-induced pluripotent stem cell-derived cardiomyocytes are especially attractive because they express ion channels and demonstrate spontaneous mechanical and electrical activity similar to adult cardiomyocytes. Here we demonstrate techniques for measuring the impact of pharmacologic compounds on the beating rate of cardiomyocytes with ImageXpress Micro and FLIPR Tetra systems. The assays employ calcium-sensitive dyes to monitor changes in Ca(2+) fluxes synchronous with cell beating, which allows monitoring of the beat rate, amplitude, and other parameters. We demonstrate here that the system is able to detect concentration-dependent atypical patterns caused by hERG inhibitors and other ion channel blockers. We also show that both positive and negative chronotropic effects on cardiac rate can be observed and IC(50) values determined. This methodology is well suited for safety testing and can be used to estimate efficacy and dosing of drug candidates prior to clinical studies.
Content may be subject to copyright.
http://jbx.sagepub.com/
Journal of Biomolecular Screening
http://jbx.sagepub.com/content/18/1/39
The online version of this article can be found at:
DOI: 10.1177/1087057112457590
2013 18: 39 originally published online 12 September 2012J Biomol Screen Anson and Evan F. Cromwell
Oksana Sirenko, Carole Crittenden, Nick Callamaras, Jayne Hesley, Yen-Wen Chen, Carlos Funes, Ivan Rusyn, Blake
Cells
Multiparameter In Vitro Assessment of Compound Effects on Cardiomyocyte Physiology Using iPSC
Published by:
http://www.sagepublications.com
On behalf of:
Journal of Biomolecular Screening
can be found at:Journal of Biomolecular ScreeningAdditional services and information for
http://jbx.sagepub.com/cgi/alertsEmail Alerts:
http://jbx.sagepub.com/subscriptionsSubscriptions:
http://www.sagepub.com/journalsReprints.navReprints:
http://www.sagepub.com/journalsPermissions.navPermissions:
What is This?
- Sep 12, 2012OnlineFirst Version of Record
- Dec 20, 2012Version of Record >>
by guest on October 11, 2013jbx.sagepub.comDownloaded from by guest on October 11, 2013jbx.sagepub.comDownloaded from by guest on October 11, 2013jbx.sagepub.comDownloaded from by guest on October 11, 2013jbx.sagepub.comDownloaded from by guest on October 11, 2013jbx.sagepub.comDownloaded from by guest on October 11, 2013jbx.sagepub.comDownloaded from by guest on October 11, 2013jbx.sagepub.comDownloaded from by guest on October 11, 2013jbx.sagepub.comDownloaded from by guest on October 11, 2013jbx.sagepub.comDownloaded from by guest on October 11, 2013jbx.sagepub.comDownloaded from by guest on October 11, 2013jbx.sagepub.comDownloaded from by guest on October 11, 2013jbx.sagepub.comDownloaded from by guest on October 11, 2013jbx.sagepub.comDownloaded from by guest on October 11, 2013jbx.sagepub.comDownloaded from by guest on October 11, 2013jbx.sagepub.comDownloaded from by guest on October 11, 2013jbx.sagepub.comDownloaded from
Journal of Biomolecular Screening
18(1) 39 –53
© 2013 Society for Laboratory
Automation and Screening
DOI: 10.1177/1087057112457590
http://jbx.sagepub.com
Introduction
A primary cause of compound attrition in preclinical drug
development is cardiotoxicity.1 Early detection of cardio-
vascular side effects is critical to avoid late-stage preclini-
cal termination, adverse cardiac-related events during
clinical trials, or recall of a drug from the market. Therefore,
development of sensitive in vitro assays suitable for safety
and efficacy testing is extremely important for drug devel-
opment. Currently, electrophysiology-based assays for
interactions of compounds with hERG channels are the
most widely used in vitro methods to assess potential
adverse cardiac affects. These assays use patch clamp
methodology and immortalized mammalian cell lines over-
expressing hERG channels (ICH; S7B). Although these
assays are predictive of ion channel block, they are typi-
cally low to medium throughput and are not suitable for
assessing potential adverse interactions with biochemical
or contractile processes, which are important to maintain
proper cardiac function.
Human cardiomyocytes overcome the limitations of
immortalized cell lines overexpressing single, or relatively
few, proteins of interest. Human-induced pluripotent stem cell
(iPSC)–derived cardiomyocytes are especially attractive as
457590JBXXXX10.1177/1087057112457590Jo
urnal of Biomolecular ScreeningSirenko et al.
2012
1Molecular Devices, LLC, Sunnyvale, California, USA
2Department of Environmental Sciences & Engineering, University of
North Carolina, Chapel Hill, North Carolina, USA
3Cellular Dynamics International, Madison, Wisconsin, USA
*Currently at Vertex Pharmaceuticals, Inc., San Diego, California, USA
Received Apr 9, 2012, and in revised form May 21, 2012. Accepted for
publication Jul 8, 2012.
Supplementary material for this article is available on the Journal of
Biomolecular Screening Web site at http://jbx.sagepub.com/supplemental.
Corresponding Author:
Oksana Sirenko, Molecular Devices, LLC, 1311 Orleans Drive, Sunnyvale,
CA 94089, USA
Email: oksana.sirenko@moldev.com
Multiparameter In Vitro Assessment of
Compound Effects on Cardiomyocyte
Physiology Using iPSC Cells
Oksana Sirenko1, Carole Crittenden1, Nick Callamaras1,*, Jayne Hesley1,
Yen-Wen Chen1, Carlos Funes1, Ivan Rusyn2, Blake Anson3, and
Evan F. Cromwell1
Abstract
A large percentage of drugs fail in clinical studies due to cardiac toxicity; thus, development of sensitive in vitro assays
that can evaluate potential adverse effects on cardiomyocytes is extremely important for drug development. Human
cardiomyocytes derived from stem cell sources offer more clinically relevant cell-based models than those presently
available. Human-induced pluripotent stem cell–derived cardiomyocytes are especially attractive because they express
ion channels and demonstrate spontaneous mechanical and electrical activity similar to adult cardiomyocytes. Here we
demonstrate techniques for measuring the impact of pharmacologic compounds on the beating rate of cardiomyocytes
with ImageXpress Micro and FLIPR Tetra systems. The assays employ calcium-sensitive dyes to monitor changes in Ca2+
fluxes synchronous with cell beating, which allows monitoring of the beat rate, amplitude, and other parameters. We
demonstrate here that the system is able to detect concentration-dependent atypical patterns caused by hERG inhibitors
and other ion channel blockers. We also show that both positive and negative chronotropic effects on cardiac rate can be
observed and IC50 values determined. This methodology is well suited for safety testing and can be used to estimate efficacy
and dosing of drug candidates prior to clinical studies.
Keywords
cardio safety, iPSC cardiomyocytes, predictive toxicology, stem cell, calcium flux, ion channel, hERG, fast kinetic fluorescence
imaging, HTS
40 Journal of Biomolecular Screening 18(1)
they recapitulate the expected genomic, biochemical, mechan-
ical, and electrophysiological behaviors of human cardiomyo-
cytes.2–5 Furthermore, human iPSC-derived cardiomyocytes
are derived from a single, infinitely expandable iPSC
source and can be reproducibly differentiated in large quanti-
ties at high purity and cryopreserved until use, thus making
them especially useful for large-scale and longitudinal
investigations.6
As human iPSC-derived cardiomyocytes express full car-
diomyocyte functionality, the need has arisen for more robust
and higher-throughput assay platforms able to assess multiple
endpoints during preclinical drug development and safety test-
ing. To this end, we have developed automated cell-based
assays for measuring the impact of pharmacologic compounds
on cardiomyocyte beat rate, mechanical activity, and intracel-
lular calcium handling. One method employs high-content
imaging with automated, high-resolution image acquisition of
fluorescently stained live cardiomyocytes to capture the dual
endpoints of beat rate and mechanical movement while leav-
ing the preparation in a relatively undisturbed state. A second
method uses fast kinetic fluorescence imaging to monitor
intracellular Ca2+ levels and simultaneously provide a direct
assessment of Ca2+ handling with surrogate assessments of
beat rate and sarcolemmal electrophysiological activity. The
FLIPR Tetra system (Molecular Devices, Sunnyvale, CA)
with whole-plate imaging allows this measurement to be
accomplished in less than 2 min. When coupled with auto-
mated compound addition, the system is well suited for gener-
ating data in a quantitative high-throughput screening (HTS)
mode. Both methods enable interrogation in 96- or 384-well
formats with well-by-well or whole-plate analyses, video
archiving of data, automated data analysis, and generation of
concentration-response relations.
We have demonstrated how these newly developed
assays can be applied in both safety testing and drug discov-
ery applications. As mentioned previously, ion channel
block can lead to drug-induced arrhythmias and must be
assessed during preclinical development. Concentration-
dependent changes in beating pattern were generated by
application of a number of proarrhythmic cardiotoxic com-
pounds known to block hERG channels as well as com-
pounds affecting Na+ and Ca2+ channels. Positive and
negative inotropes are used in clinics to treat heart failure,
tachycardia, arrhythmia, and other cardiac diseases.
Therefore, the beat rate provides the basis for phenotypic
screens in drug discovery. Similar to its use in toxicity test-
ing, the fluorescence-based applications described here
detected concentration-dependent effects of several positive
and negative inotropes and chronotropes on cardiac rates
and determined EC50s at the expected ranges.
Development of new, more potent, and safer drugs
requires an in vitro system in which functional outcomes
can be quantitatively tested. The assays described here are
ideal for that task and enable determination of efficacy and
approximate dose ranges prior to clinical studies and hold
the potential to accelerate the drug discovery process.
Methods
Human iPSC-Derived Cardiomyocytes and
Cell Culture
The methods for reprogramming the source material into
iPSCs and subsequent differentiation into human iPSC-
derived cardiomyocytes (iCell Cardiomyocytes) have been
described previously.4 iCell Cardiomyocytes, plating
medium, and maintenance medium were received frozen
from Cellular Dynamics International. Cardiomyocytes
were thawed and plated according to the User’s Guide.
Briefly, cardiomyocytes were thawed and plated into
gelatin-coated wells at 20K or 4K cardiomyocytes per well
of 96 or 384 multiwell plates, respectively. Wells were
overseeded with cardiomyocytes to account for the plating
efficiency of the cardiomyocytes per the Certificate of
Analysis. Cardiomyocytes were incubated in a humidified
cell culture incubator and maintained at 37 °C in 5% CO2.
Cell contractions were observed visually under a light
microscope after ~3 days in culture. The presence of strong
synchronous contractions of cells in the wells under the
light microscope was confirmed prior to running experi-
ments on the FLIPR Tetra system. With few exceptions,
experiments were done after 5 to 7 days of cell culture.
High Content Imaging Data Acquisition and
Analysis
For imaging of mechanical contractions, cells were incu-
bated with Calcein AM (Invitrogen, Carlsbad, CA) for 10
min to provide a relatively homogenous cellular label and
then treated with different concentrations of compounds.
Image acquisition was performed on an environmentally
controlled ImageXpress Micro system (Molecular Devices,
Sunnyvale, CA) using time-lapse imaging with a 20× or 10×
objective, FITC excitation and emission filter setup, and 40
images per field at up to 10 frames per second (fps). For
imaging of calcium flux, acquisition was done on an envi-
ronmentally controlled ImageXpress Micro XL system
using time-lapse imaging with a 10× objective, FITC filter
cube, and 500 images per field at up to 100 fps. Environmental
control for both systems was 37 °C and 5% CO2.
For analysis of mechanical contractions, a time series of
images was combined into a stack and then processed using
a custom protocol, or journal. The journal consisted of
opening the stack, subtracting each image from the previous
one, and then thresholding the resulting differential image.
The intensity values of pixels above the threshold were inte-
grated, and the data were exported into a spreadsheet pro-
gram (Microsoft Excel) as percentage of threshold (%Th)
Sirenko et al. 41
for each time point. A plot of %Th versus time provided
a curve in which peaks in the %Th value corresponded to
cell contractions or beats. The number of peaks was manu-
ally calculated and a beats per minute (beat/min) value
determined.
FLIPR Tetra Data Acquisition
In the FLIPR Tetra instrument, the entire bottom of the plate
is illuminated with excitation light and then fluorescence
emission is imaged onto a fast CCD camera. The signal from
individual wells is discerned by appropriate binning of pixels
associated with each well, and the average intensity is
recorded as a function of time. This allows kinetic measure-
ments to be performed on a well-by-well basis with excellent
signal to noise. Another feature of the system is on-board
liquid handling for dispensing simultaneously into either 96-
or 384-well plates. This allows dosing to be controlled very
accurately from both a volume and timing perspective.
Components of a FLIPR Calcium 5 Assay Kit (Molecular
Devices, Sunnyvale, CA) were added to the plates according
to standard protocol and incubated for 60 min in a cell cul-
ture incubator. Compound plates were prewarmed to 37 °C
inside the FLIPR Tetra instrument, and compound addition
was done simultaneously to all wells. Filter sets were appro-
priate for the FLIPR Calcium 5 assay kit (ex 485 nm, em 530
nm); data were acquired at 8 fps and up to 800 image frames.
The stage of the instrument was kept at a constant tempera-
ture of 37 °C. Data were acquired before compound addition,
during compound addition, and at prescribed times after
compound addition. The total time for image acquisition and
analysis was approximately 2 min per plate.
Peak Analysis
For automated microscopy measurements with an
ImageXpress Micro system, beats were detected as peaks in
the fluorescence intensity versus time plots, and beat fre-
quency was determined from the ratio of the number of
beats over the acquisition time (beats/min). For measure-
ments using the FLIPR Tetra system, an automated algo-
rithm in ScreenWorks Peak Pro software processes the data
and provides the user-selectable outputs. In brief, the soft-
ware detects the signal peaks within a data trace from a
single well using a dynamic thresholding and derivative
analysis. The peaks are then fit with a binomial function to
calculate the peak amplitude, peak temporal position, peak
count, and peak full width at half of maximum of amplitude
(FWHM). The peak rise time (10% to 90% intensity), peak
decay time (90% to 10% intensity), and peak width at 10%
of the maximum intensity (peak width at 10% amplitude)
are measured separately. The time points for these measure-
ments are determined using a linear interpolation between
the two closest data points. The beat frequency is deter-
mined by the inverse of the average temporal spacing
between beats. Peaks located at the start or end of a trace
that are less than half present are filtered by the algorithm
and are not included in calculation of the results. The user
can select as an output the average or standard deviation of
those parameters on a well-by-well basis.
Results
Assessment of Mechanical Cell Movement
Using Time-Lapse Imaging
A method was developed to allow analysis of cardiomyo-
cyte contractions without using any indirect readouts. The
mechanical movement of cells is captured through auto-
mated image analysis using a live cell stain that allows cell
movement to be characterized without interference with
cell metabolism. The beating rate of contracting cardio-
myocytes was determined from a series of time-lapse
images acquired on an ImageXpress Micro system fol-
lowed by a derivate analysis. An outline of the analysis
workflow is presented in Figure 1. iPSC-derived cardio-
myocytes were cultured in a monolayer in 96-well or 384-
well plates for four days and then stained with Calcein AM
for 10 min. Then media was replaced, cells were treated
with compounds for 10 min, and images were acquired.
The impact of several pharmacologic agents on the beat
rate was tested using this method. Two of the agents, isoproter-
enol and epinephrine, are β-adrenergic receptor agonists and
known stimulants of cardiac beat rate. Compounds were tested
in duplicates in 96-well plates and characterized by measuring
the number of contraction peaks in each well and then
calculating the average beats/min for each concentration.
Concentration-response curves for epinephrine and isoproter-
enol are shown in Figure 1. The tested reagents modulated the
frequency of beating in line with their mode of action showing
the functionality of a- and β-adrenergic and acetylcholine
receptors and demonstrating IC50s of compounds in the
expected range (epinephrine ~50 nM, isoproterenol ~6 nM).
Monitoring Calcium Fluxes Using
Time-Lapse Imaging
A second method was employed to measure cardiomyocyte
contractions by using automated microscopy to monitor
changes in concentration of Ca2+ synchronous with contrac-
tions. A time-lapse series of images of beating cardiomyocytes
loaded with FLIPR Calcium 5 assay kit is shown in Figure 2.
The images were acquired using the ImageXpress Micro XL
automated microscope using a 10× objective. A significant
increase in intracellular intensity associated with synchronous
contraction events can be observed. The signal from all cells
in the images can be integrated and used as a monitor of
the contraction event. Traces of integrated intensity from a
42 Journal of Biomolecular Screening 18(1)
time-lapse series of images acquired at 33 Hz for 5 s for differ-
ent treatments are shown in Figure 2.
This method was used to evaluate the effects of two
compounds: epinephrine and verapamil, a known negative
chronotrope that is an L-type calcium channel blocker of
the phenylalkylamine class.7 Wells were analyzed on an
individual basis over a ~10 min period beginning approxi-
mately 10 min after drug addition. Time-lapse images were
acquired from each well for 23 s. The data were viewed as
a plot of integrated fluorescence intensity versus time and
manually analyzed. The number of peaks, or beats, per well
were counted, and a plot of beat rate versus concentration is
shown in Figure 2. The measured responses as well as IC50s
were consistent with the expected behavior. Additional
examples are presented in Supplemental Figures 1 and 2.
HTS Cardiomyocyte Beating Assay
on FLIPR Tetra System
There are several challenges associated with analyzing
beating cardiomyocytes with a standard microscope-based
system, including lower throughput and time difference
Automatic Method for
Beat Rate Analysis
0
1
2
3
4
5
6
7
8
9
13579111315171921232527293133353739
Images
% Treshold
Differential
Algorithm
Threshold
Analysis
Automatic Method for
Beat Rate Analysis
0
1
2
3
4
5
6
7
8
9
13579111315171921232527293133353739
Differential
Algorithm
Threshold
Analysis
Isoproterenol
-5 -3 -2 -1 0 1
10
20
30
40
50
Control
Log [Agonist] µM
Beats/min
Epinephrine
-5 -3 -2 -1 0 1
10
20
30
40
50
Control
Log [Agonist] µM
Beats/min
Figure 1. Method for determination of cardiomyocyte mechanical contractions from time-lapsed images and concentration-response
curves for two positive chronotropes. Top left: Series of images acquired of beating cells. Center left: Differential image created by a
MetaXpress software journal in which pixel intensity corresponds to %Threshold (%Th) value. Bottom left: Plot of integrated %Th versus
image number for unstimulated cells. Right: Beat rate modulation of cardiac cells as a function of concentration of isoproterenol and
epinephrine. Beat frequency was determined approximately 10 min after compound addition. Error bars represent ±1 standard deviation
of the measurements.
Sirenko et al. 43
between data acquisition from different wells. An improved
method for this assay is to automatically collect kinetic
fluorescence data from all wells simultaneously. A system
that has that capability is the FLIPR Tetra instrument.8
Intracellular transient Ca2+ fluxes underlying cardiomyo-
cyte contractions were studied on this system. Time-
dependent signals were measured on a well-by-well basis
from both 96- and 384-multiwell plates with excellent sig-
nal to noise. Representative traces of a control well and
wells dosed with epinephrine and verapamil as measured
by this system are shown in Figure 3 (top). An expanded
trace of a single contraction event is also shown in the same
figure (bottom right). It has been found that the cell-beating
pattern can be disrupted by compound addition and that the
best stability of cell beating occurs approximately after 5
min after addition. An example trace of cell-beating disrup-
tion by addition of isoproterenol is also shown in Figure 3
(bottom left). Concentration responses were measured by
adding compounds in half-log serial dilutions using on-
board liquid handling. Results from a concentration-
response study of eight different compounds are shown in
Figure 4.
Even though the assay optimization for HTS was not
in the scope of present studies, assay variability was
acceptable for most experiments. For example, the aver-
age and standard deviation of the controls wells in rows
B and L for the experiment shown in Figure 4 were
found to be 15.0 ± 1.0 and 15.0 ± 1.5 beats/min, respec-
tively. Typical well-to-well coefficients of variation are
15% or less for most parameters. The baseline rate of
beating was consistent between plates within a given
experiment or between experiments using the same pro-
tocol. We have observed some dependence of initial
beating rate on the time of cell culture (data not shown).
Further assay optimization should increase reproducibil-
ity and precision of the assay.
Figure 2. Top: Time-lapse series of images of spontaneously contracting cardiomyocytes loaded with FLIPR Calcium 5 dye. Bottom left:
Integrated intensity from calcium-sensitive dye fluorescence as a function of time (20 s per trace) for beating cardiomyocytes under
three conditions. Distinct peaks can be observed associated with each contraction event. Bottom right: Response of cardiomyocyte beat
rate to increasing doses of epinephrine and verapamil. Beats per minute were determined by manual counting of peaks. IC50 values were
determined to be 8 nm and 9 nm, respectively.
44 Journal of Biomolecular Screening 18(1)
Automated Data Analysis
Automated data analysis is required to make the system
practical for running large-scale assays. The automated
algorithm in ScreenWorks Peak Pro was used to analyze the
data. A typical beat trace and measured parameters are
shown in Figure 3. The data analysis occurs in real time,
and results are presented on a well-by-well basis to the user.
The results can be exported to a standard comma-separated-
variable (*.csv) file for further analysis.
The software was validated by comparing automated
results to manual measurements. The number of peaks in
each well were measured manually for three representative
plates and compared with automated peak count values.
Full concordance was observed for more than 96% of the
wells. In addition, several representative wells were manu-
ally characterized by measuring shape parameters (widths,
rise, and decay times) for each peak and then calculating
average and standard deviation of those values for each
well. In general, the manual shape values agreed to within
10% of the automated peak measurements. In some cases,
the peak width, or peak FWHM values, showed greater dis-
crepancy. This is attributed to the fact that the automated
method uses a second-order polynomial fit to each peak and
calculates a width based on that fit. This is less accurate for
more complex peak shapes such as seen with cisapride (see
Fig. 5). However, the peak width at 10% amplitude values,
which are not based on the peak fit, showed much better
agreement between the manual and automated methods.
Therefore, this parameter is recommended for analysis of
peaks with irregular shape.
Cardiomyocyte Beating Assay
Results
Assessment of Positive and Negative
Chronotrope Effects
Beating rate is a potentially valuable phenotypic parameter
for use in drug discovery and development. The automated
cardiac beating assay was used to assess effects of several
known positive and negative chronotropes. Positive chrono-
tropes included epinephrine and isoproterenol. In addition,
dopamine catecholamine neurotransmitter, a precursor of
epinephrine, was used. Negative chronotropes included
EpinephrineControl Verapamil
Detail Graph
Relative Light Units
3000
4000
Detail Graph
Relative Light Units
3000
4000
Detail Graph
Relative Light Units
2000
2500
3000
Time (Seconds)
0 50 100 150
2000
Time (Seconds)
0 50 100 150
2000
Time (Seconds)
0 50 100 150
4000
Units
Isoproterenol
400
500
DecayTime
Rise
Time
Peak Width
(FWHM)
1000
2000
3000
Relative Light U
0
0
0.51 1.52 2.
53
100
200
300 Decay
Peak Width
10% amplitude
Amplitude
Time (seconds) Time (seconds)
020406010080
Figure 3. Top: Fluorescence intensity versus time for three representative wells of beating cardiomyocytes loaded with FLIPR Calcium 5
dye. Measurement was done on a FLIPR Tetra system. Bottom left: Fluorescence intensity versus time for a single control well of beating
cardiomyocytes loaded with FLIPR Calcium 5 dye during addition of 0.1 µM isoproterenol. Bottom right: Fluorescence intensity profile
of a single cardiomyocyte contraction event. Data for this trace were acquired at 120 frames per second. Calculated analysis parameters
are shown on the profile.
Sirenko et al. 45
propranolol, a nonselective β-blocker; doxazosin, an alpha-1
adrenergic receptor blocker that inhibits the binding of nor-
epinephrine (released from sympathetic nerve terminals) to
the alpha-1 receptors; verapamil; and acetylcholine, an ace-
tylcholine receptor agonist. We were able to observe concen-
tration-dependent changes in the beat rate, agreeing with the
mode of action of each compound. Results from a represen-
tative experiment are shown in Figure 6A. The IC50s of
compounds tested were in good correlation with data
obtained by other types of assays9–16 (Table 1). For example,
IC50s for isoproterenol10 and verapamil13 tested in other mod-
els were 13 nM (isolated rabbit cardiomyocytes) and 167 nM
(isolated guinea pig hearts), respectively. However, we were
not able to detect the effect of acetylcholine at present assay
conditions.
Cardiotoxic Compounds
The FLIPR Tetra system also allows for detection of
concentration-dependent atypical patterns and changes in
cell-beating rate caused by several known cardiotoxic com-
pounds including hERG, Ca2+ and Na+ channel blockers. In
each case, the baseline was recorded immediately prior to
compound addition, and then measurements were taken at
regular intervals over a 60 min period. We found that high
concentrations of many compounds caused effect immedi-
ately (within seconds) after addition (see, for example, the
effect of isoproterenol addition shown in Figure 3, bottom
left).
Potentially cadiotoxic compounds can be detected in
the assay by their effect on the beat rate pattern. We have
tested several known blockers of different ion channels.
Those included blockers of Na+ channels (lidocaine, tetro-
dotoxin), Ca2+ channels (nifedipine, isradipine, verapamil),
and hERG channels (cisapride, terfenadine, astemizole,
pimozide). Presented in Figure 6B are selected concentra-
tion-dependent inhibition curves for the beating rates as
well as IC50s determined in the assay. These correlate well
with reported IC50s for cardiac toxicity of those compounds
(Table 1).
Na+ channels are responsible for the inward Na+ current
and depolarization phase of cardiac action potential.
Treatment with the Na+ blocker lidocaine resulted in
concentration-dependent inhibition of beating rate as well
as irregularities in the beating profile (see Figs. 5A, 6B, and
Table 1). The compound tetrodotoxin had a similar effect.
The IC50 for lidocaine was 11.2 µM, which is in the same
order as published values of 60 µM (rat cardiac myo-
cytes17) and 20 µM (mouse ESC-derived cells18). The IC50
for tetrodotoxin of 3.2 µM was relatively higher than the
0.28 µM value observed in impedance studies (mouse
ESC-derived cardiomyocytes)19 but close to automated
patch clamp values (1.3 µM for mouse18 and 1.5 µM for
human20 cardiomyocytes).
All
A
B
C
D
E
F
G
H
I
J
K
L
123456789101112131415161718192021222324
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
10 9 9 8 10 9 9 9 9 8 9 8 9 9 9 8 9 8 9 9 8 7 9 10
19 18 19 19 18 18 0 0 0 17 15 17 0 0 0 8 8 8 2 1 4 2 3 1
19 18 18 18 17 18 15 15 14 15 15 16 0 0 0 8 7 8 3 4 3 4 5 3
19 18 19 19 17 18 10 11 10 14 14 14 0 0 0 9 8 8 6 5 4 6 6 6
19 18 18 17 17 16 9 8 9 13 13 14 0 0 1 9 8 7 7 5 6 7 4 7
19 17 18 16 15 16 8 9 9 11 12 11 0 1 2 8 9 9 7 7 7 7 7 8
18 16 18 15 16 14 8 9 8 10 10 10 2 4 3 8 8 8 9 8 8 8 8 9
19 17 16 16 15 14 8 8 8 9 10 8 2 6 4 9 8 8 9 9 10 8 8 8
15 14 14 13 13 12 8 8 8 9 8 8 4 6 4 9 8 8 9 10 8 8 9 9
17 13 16 13 13 11 8 9 9 8 9 9 6 9 6 9 8 8 9 9 9 9 9 8
10 10 8 9 8 9 8 9 8 8 8 9 9 9 8 9 8 9 9 10 8 9 10 11
isoproterenol epinephrinedigoxinverapamil
dopaminepropranolol doxazosinac-choline
All
A
B
C
D
E
F
G
H
I
J
K
L
123456789101112131415161718192021222324
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
10 9 9 8 10 9 9 9 9 8 9 8 9 9 9 8 9 8 9 9 8 7 9 10
19 18 19 19 18 18 0 0 0 17 15 17 0 0 0 8 8 8 2 1 4 2 3 1
19 18 18 18 17 18 15 15 14 15 15 16 0 0 0 8 7 8 3 4 3 4 5 3
19 18 19 19 17 18 10 11 10 14 14 14 0 0 0 9 8 8 6 5 4 6 6 6
19 18 18 17 17 16 9 8 9 13 13 14 0 0 1 9 8 7 7 5 6 7 4 7
19 17 18 16 15 16 8 9 9 11 12 11 0 1 2 8 9 9 7 7 7 7 7 8
18 16 18 15 16 14 8 9 8 10 10 10 2 4 3 8 8 8 9 8 8 8 8 9
19 17 16 16 15 14 8 8 8 9 10 8 2 6 4 9 8 8 9 9 10 8 8 8
15 14 14 13 13 12 8 8 8 9 8 8 4 6 4 9 8 8 9 10 8 8 9 9
17 13 16 13 13 11 89 9 8 9 9 6 9 6 9 8 8 9 9 9 9 9 8
10 10 8 9 8 9 8 9 8 8 8 9 9 9 8 9 8 9 9 10 8 9 10 11
control
isoproterenol epinephrinedigoxinverapamil
dopaminepropranolol doxazosinac-choline
isoproterenol epinephrinedigoxinverapamil
dopaminepropranolol doxazosinac-choline
concentration
Figure 4. Screenshot from Screenworks Peak Pro software showing cardiomyocyte intensity fluctuation traces for positive and negative
chronotropes. Control wells were located in rows B and L. Compounds were added in half-log serial dilutions from rows C through K in
replicates of three. Numbers in each well indicate beat counts per recorded time (30 s). Note: Row A contained no cells.
46 Journal of Biomolecular Screening 18(1)
C
A
Relative Light Units
Time (Seconds)
050 100 150
2000
3000
4000
050 100150
2000
3000
4000
050 100150
2000
3000
4000
050 100 150
2000
3000
4000
050 100150
2000
3000
4000
050 100150
2000
3000
4000
050 100 150
2000
3000
4000
5000
050 100 150
2000
3000
4000
5000
050 100 150
2000
3000
4000
5000
Isoproterenol
Propranolol
Lidocaine
Increasing Concentration
B
050 100150
3000
4000
5000
050 100150
3000
4000
5000
050 100 150
3000
4000
5000
050 100150
3000
4000
5000
050 100150
3000
4000
5000
050 100150
3000
4000
5000
050 10
0
150
3000
4000
5000
050 100150
3000
4000
5000
050 100150
3000
4000
5000
Cisapride
Astemizole
Pimozide
Relative Light Units
Time (Seconds)
Astemizole 1µMCisapride 0.3µM
020 40 60 80 100 120140
2000
3000
4000
5000
020 40 60 80 100 120 140
3000
4000
5000
Relative Light Units
Time (Seconds)
Figure 5. (A) Effects of increasing concentrations of representative compounds on cardiomyocyte contractions. Data collected 10 min
after compound addition. (B) Effect of increasing concentrations of hERG inhibitors on cardiomyocyte contractions. (C) Prolongation of
repolarization stage by hERG blockers. Note overlap of beating patterns leading to high-frequency low-amplitude oscillations.
Sirenko et al. 47
The effect of Ca2+ channel blockers was tested with nife-
dipine, a dihydropyridine class of L-type calcium ion channel
blocker. We have observed concentration-dependent inhibi-
tion of the beating rate (IC50 120 nM) for this compound,
which is consistent with 79 nM observed in Langendorff-
perfused murine hearts14,21 with no dramatic change of the
beating pattern. Isradipine, another example of L-type cal-
cium channel blocker, also decreased the beating rate with an
IC50 of 80 nM (Table 1), which is on the same range as the
value of 20 nM reported for the impedance assay using
mouse stem cell–derived cardiomyocytes.19 The compound
Bay K 8644 acts as an agonist of calcium channels.22,23
Figure 6. (A) Concentration-response curves for positive and negative chronotropes from automated cardiac beating assay. FLIPR
Calcium 5 dye was added 60 min prior to compound addition, and reads were acquired 10 min after addition of compounds. Three
replicates for each concentration were used for 384-well plates. EC50 and IC50 values were determined from a four-parameter fit.
(B) Concentration-response curves for five different cardiotoxic compounds that affect ion channels. The IC50 value for propranolol
determined in 5B is believed more accurate due to the extended concentration ranged used in that experiment. Error bars represent ±1
standard deviation of the measurements.
48 Journal of Biomolecular Screening 18(1)
Treatment with this compound increased the beating rate in a
concentration-dependent manner with an EC50 of 100 nM,
which is consistent with Abassi’s observation (77 nM for
mouse ESC-derived cells) as well as other reports (33 nM for
rat primary myocytes).19
The outward K+ current is part of the repolarization phase
of cardiac action potential, and blocking the K+ channel
(hERG channel) results in a prolonged repolarization phase.
hERG inhibitors are an important class of compounds that
are considered to be potentially dangerous, and screening for
blockers of this channel is required by the Food and Drug
Administration. We have tested several known inhibitors of
the hERG channel (cisapride, astemizole, terfenadine, and
pimozide) and observed concentration-dependent inhibition
of beating rate by the compounds (see Figs. 5B, 6B, and
Table 1). The IC50s observed for cisapride and astemizole are
32 nM and 6 nM, respectively, which are close to published
values obtained by patch clamp and other methods.12,24–26 In
addition, there were dramatic changes of the beating pattern:
the repolarization phase of beating patterns become increased
as well as durations of each beat. The most significant effect
we have observed is for cisapride (Figs. 5 and 7), where
there was a very prominent plateau phase on the repolariza-
tion shoulder of each beat. We have also noted a brief oscilla-
tion part at the end of each plateau (Fig. 5). Using ScreenWorks
Peak Pro software, we were able to characterize this beating
pattern by measuring a very significant increase of peak
width at 10% of amplitude as well as duration of the depolar-
ization phase. We were able also to observe the overlap of the
next beat to the previous one, leading to irregular extra beats
flagged by the software. Interestingly, at higher concentra-
tions of compounds, we have observed a very high frequency
of Ca2+ oscillations
Finally, we have compared IC50s obtained using the car-
diac beating assay on the FLIPR Tetra system with published
data from studies using other cardiac models (Table 1). There
was a significant rank correlation (rSpearman = 0.57,
p = 0.045) of data from this system with the IC50 values from
published reports (Supplemental Figure 3). In addition, we
have compared IC50 values from the assay to reported human
Cmax values, as well as unbound plasma concentrations of
drugs shown to be associated with clinical QT interval prolon-
gation (10%–20%; Table 1; Supplemental Figure 1;
Supplemental Table 1). Accordingly, although the drugs
tested in the assay had a wide range (more than five orders of
magnitude) of reported Cmax values (Supplemental Figure 4,
“+” symbols), the IC50 values from the assay on the FLIPR
Tetra system were at the same concentration range or no more
than two orders of magnitude different from Cmax values for
most drugs (Supplemental Figure 4, “O” symbols). There
Table 1. Beating Cardiac Assay Concentration Response Results
IC50/EC50 Values (µM)
Compound Molecular Target Cardiac Beating Patch Clamp Other Assays Cmax
Isoproterenol β-adrenergetic receptor (AR) agonist 0.0075 0.01310
Epinephrine β-AR agonist 0.053 0.02–0.049
Bay K 8644 L-type Ca2+ channel agonist 0.1 0.0822,23,19
Dopamine Agonist of dopamine receptor 0.27 1211
Propranolol a-1 AR antagonist 3.6 716 0.201
Doxazosin a-1 AR blocker 1.48 0.612 0.133
Digoxin Inhibitor of Na-K ATPase 0.57 0.07315 0.0028
Tetrodotoxin Na+ channel blocker 3.24 1.318 1.520
Lidocaine Na+ channel blocker 11.2 2018 6017 36.3
Quinidine Blocks the fast inward Na+ current 0.38 1.112 4017 8.66
Astemizole hERG inhibitor 0.006 0.00112 0.0011
Cisapride hERG inhibitor 0.032 0.02712 0.04424 0.129
Terfenadine hERG inhibitor 0.001 0.00812 0.05624 0.0032
Pentamidine Effects hERG channel transport 0.5 0.1728 0.191
Pimozide hERG and Ca2+ channel blocker 0.0096 0.00112 0.225,26 0.0037
Nifedipine L-type Ca2+ channel blocker 0.12 0.518 0.0814,21 0.271
Isradipine L-type Ca2+ channel blocker 0.08 0.0219 0.080
Verapamil L-type Ca2+ channel blocker 0.055 0.13612 0.16713 0.598
EC50s and IC50s for selected compounds determined by modulation of beating frequency (peak count) in the cardiac beating assay on the FLIPR Tetra
system using human-induced pluripotent stem cell–derived human cardiomyocytes. Representative values shown from one of several independent ex-
periments. For comparison, patch clamp values as well as data from other assays are presented in the labeled columns. The last column contains values
for Cmax, maximum plasma concentrations for corresponding drugs in humans. Please refer to Supplementary Table 1 for detailed information and
references.
Sirenko et al. 49
also was a positive significant correlation (r2 =0.45) between
Cmax and the cardiac beating assay IC50. Finally, for a few of
the drugs tested, data are available on plasma concentrations
of free drug that has been shown to be associated with clini-
cally observed QT interval prolongation.27 Of four drugs with
these data, two had very good concordance with the FLIPR
Tetra system data (Supplemental Figure 4, triangles).
Measuring Delayed Effects of Compounds
It is also advantageous to asses longer-term time-dependent
effects of compounds that may indirectly effect expression of
ion channels (e.g., pentamidine, which effects transport of the
hERG channel and therefore has a delayed effect on beating
pattern).28 In this case, continuous recording using label-free
methods is an advantage. The FLIPR Tetra system is ideal for
assessment of immediate or short-term effects; however, mea-
suring long-term effects would require a modified experimen-
tal protocol. Cells need to be pretreated with compounds for
the desired length of time (e.g., for 24 h or 48 h) as for typical
toxicity studies and then loaded with FLIPR Calcium 5 dye 30
to 60 min prior to read out on the FLIPR Tetra system. We
have used this protocol while testing effects of pentamidine28
and quinidine,17 two compounds demonstrated previously to
have a delayed effect on beating pattern.19 The contraction
patterns were measured 24 h after treatment. Both compounds
had inhibitory effects on beating rate with IC50s of 0.5 µM and
0.38 µM, respectively (Table 1).
Cardiac toxicity can be also caused by compounds that
are not ion channel inhibitors (e.g., anthracycline drugs,
kinase inhibitors, or other factors that may compromise
cell metabolic activity or viability). Although those effects
can be detected by using traditional viability assays, the
cardiac beating assay is also suitable. We have tested
effects of doxorubicin and the kinase inhibitors imatinib
and staurosporine after 24 h of treatment. We observed
expected cell death in 24 h (confirmed by imaging meth-
ods) and, accordingly, inhibition of cardiac contractions.
IC50s at 24 h were 15 µM (imatinib), 20 µM (doxorubicin),
and 3 µM (staurosporine).
Figure 7. Characteristic behaviors of selected compounds on elongation of repolarization time and time between contractions
(peak spacing). Bar graphs represent an average change in peak width at 10% amplitude and peak temporal spacing caused by selected
compounds at high dose. Error bars represent ±1 standard deviation of the measurements. Asterisks indicate results that show significant
difference from control by one-way analysis of variance test.
50 Journal of Biomolecular Screening 18(1)
Surrogate Markers for Cardiac Safety
The assay allows further beating characterization by mea-
suring the irregularity of peak width, peak spacing, peak
rise, and peak decay times and allows characterizing devia-
tions from regular patterns. We have demonstrated a modi-
fied beating pattern for several hERG inhibitors, including
cisapride, which elongates the repolarization time, reduces
beat rate, and causes QT prolongation and arrhythmia in
humans. The method allows one to observe an increase of
repolarization time for the beat that leads to overlapping
beats (Figures 5 and 7). Prolongation of repolarization time
can be detected as prolongation of peak width at 10% of
amplitude or beat decay time using algorithms of
ScreenWorks Peak Pro software (Figure 7). An increase in
spacing between peaks was observed after treatment with
lidocaine, cisapride, and propranolol (Figure 7). We have
also observed shortening of peak width and spacing
between peaks accompanied by increased beating rate with
the positive chronotropes epinephrine and isoproterenol.
Beating irregularities were observed after treatment of cells
with lidocaine, astemizole, and a number of other com-
pounds. Irregularities were flagged with software by
increased standard deviation of the distance between peaks
or by outliers in the distribution of peak spacings.
Discussion
An important cause of drug withdrawals from the market in
the past several decades has been cardiotoxicity.29,30 Often,
these drugs were associated with a potentially fatal form of
ventricular arrhythmia, referred to as Torsades de Pointes
(TdP).31 A significant number of these drugs were termi-
nated in the late preclinical studies where it is very costly.
This leads to the need for predictive assays that allow for
assessment of potential cardiotoxic side effects of lead
compounds early in the drug discovery process. The appli-
cation of Ca2+ flux measurements to spontaneously con-
tracting cardiomyocytes addresses several critical needs for
development of in vitro assays suitable for safety and effi-
cacy testing. Kinetic fluorescence recording is used to
monitor cell beating rates and other temporal parameters to
assess compound-induced changes in the phenotypic
behavior of the cardiomyocytes. The automated data acqui-
sition and analysis methods are well suited for high-
throughput environments. The assay precision is also
sufficient for screening of compounds and can potentially
allow early determination of their suitability for drug devel-
opment. The ability to measure phenotypic response of
human-derived cell models provides a system that shows
good initial concordance with clinical data.
A number of limitations of currently employed methods
are overcome with this technique. Electrophysiology meth-
ods use mammalian cell lines over expressing ion channels,
which might not be representative of native human cardio-
myocytes. Current systems for impedance-based measure-
ments of contracting cardiomyocytes are lower throughput.32
In addition, the relationship between changes in impedance
and cell biology are not well understood, and potential
dielectric properties of compounds can cause changes in
impedance that further complicate such measurements.19
Observation of Ca2+ flux in smooth muscle cell contraction
was reported in 1986.33 Intracellular calcium oscillations asso-
ciated with spontaneous cardiomyocyte contractions34 are
monitored with Ca2+-sensitive fluorescent dyes.35 This phe-
nomenon was originally termed calcium sparks because of the
fast, transient nature of the response.36 The Ca2+ fluxes associ-
ated with contractions of cardiac cells have been established as
physiologically relevant and a predictive indicator of myocar-
dial performance.37–39
iPSC-derived cardiomyocytes recapitulate the expected
genomic, biochemical, mechanical, and electrophysiologi-
cal behaviors of native human cardiomyocytes and are avail-
able in large quantities required for high-throughput
environments. The assays described here can be set up in
96- or 384-well formats, and extension to 1536-well formats
is feasible with optimization of cardiomyocyte cell-plating
protocols. The FLIPR Tetra system reads all wells simulta-
neously, thereby improving the precision of the assay by
removing time-dependent well-to-well variation. The read
and analysis time is less than 2 min per plate. Automation
and on-board liquid handling further increase the flexibility
of this platform and provide a system that is amenable to
high-throughput screening. Real-time automated analysis of
beating rate, EC50 determination for compound effects, and
characterizing deviations from typical beating patterns fur-
ther decreases time to results. The data outputs, which
include beat rate, rise and decay times, beat durations, and
assessment of irregularities in the beat pattern, provide tools
not previously available from other methods.
The assay allows further peak characterization by detec-
tion of atypical patterns caused by compounds known to be
associated with long QT syndrome (e.g., cisapride and doxa-
zosin) and Na+ channel blockers (e.g., lidocaine). Measuring
irregularity of peak width, peak spacing, peak rise, and peak
decay times would allow characterizing deviations from a
regular pattern and prediction of drugs inducing long QT syn-
drome, arrhythmia, and other potentially dangerous effects.
These approaches are in their infancy; therefore, the rele-
vance of this in vitro phenotype to in vivo disease models is
yet to be confirmed. However, characterization of increased
repolarization times caused by known blockers of hERG
channels is the first step. It is too preliminary to relate these
observations directly to the known in vivo patterns such as
long QT syndrome or TdP; however, this type of signature
(i.e., prolonged peak width and increased peak decay time)
might be considered as a potentially predicative pattern for
hERG blockers and similar compounds. These data suggest
Sirenko et al. 51
that this assay can be used to assess pharmacologic modula-
tors of ion channels and adrenergic receptors and other tar-
gets that effect cardiac contraction.
An important consideration is whether the temporal
response of the calcium-sensitive dye signal, or calcium
sparks as initially called,36 is an indicator of myocardial per-
formance and hence can be used for predictive toxicology.
Electrophysiology measurements of cellular action poten-
tials (AP) have been accepted in this role. The response
times of the various ion channels that contribute to the
action potential vary from a few milliseconds for Na+ chan-
nels to several hundreds of milliseconds for the Ca2+ and K+
channels. The temporal shape of a cardiomyocyte calcium
spark is similar to that of an AP, but there are significant
differences to the underlying processes. The AP is an elec-
trochemical response typically measured on a single cell,
whereas the calcium spark stems from mechanical move-
ment of intracellular biochemicals and, in the case here, an
averaged response over a large population of cells. This
means the AP is based on inherently faster responses and
does not suffer from effects of synchronicity between cells.
Using fast optical sampling rates (up to 100 Hz), we can
observe components of the calcium spark that correspond to
the different ion channels on a time scale of the right order
of magnitude. The initial rise times, which are attributed to
the Na+ channels, are on the order of 100 ms or less. The
plateau regions, attributed to Ca2+ and K+ channels, are typi-
cally several hundreds of milliseconds. Indeed, elongation
of the plateau by K+ channel blockers has been observed to
be several seconds. The calcium spark is similar to the AP
but more physiologically like a true cardio contraction.
There are other dyes and dye systems that can be employed
to monitor cardiomyocyte contractions including Di-8-
ANEPPS,40 the Mermaid fluorescence protein FRET sys-
tem,41 and DiBAC4(3).42 The first two dye systems use ratio
metric readouts, making the instrumentation and analysis
required more complicated. The third dye requires transpor-
tation through a cell membrane to become fluorescent. This
physical movement limits the response time of the dye.
Therefore, it is felt that calcium-sensitive dyes are best
suited for high-throughput assays that measure kinetic
responses of large cell populations.
It is important to discuss the question of assay predictiv-
ity and potential use for assessment of potency or off-target
effects of developing cardiac and noncardiac drugs. To date,
the assay has shown good concordance with clinical data
for drugs with known chronotropic and/or inotropic effects.
Of 19 drugs tested, 4 expressed expected positive chrono-
trope effect, and 14 of 15 known ion channel blockers or
receptor antagonists demonstrated expected inhibition of
contractile activity. Furthermore, two of four drugs associ-
ated with QT interval prolongation were found to have good
quantitative agreement between Cmax and IC50 values.
Therefore, despite the fact that accurate assessment of assay
specificity, sensitivity, and positive and negative predictive
values would require a larger number of test compounds to
be evaluated, the utility of the assay for detecting known
effects on cardiomyocyte physiology is encouraging.
Although the use of stem cell–derived cardiomyocytes
for evaluation of pharmacologic effects is relatively novel,
there are several recent studies using this model for predic-
tive toxicology.3,19,32,43 These studies, as well as data pre-
sented here, show that contraction rates of mESC-derived
cardiac cells are clearly affected by many compounds that
are known positive or negative chronotropes, ion channel
blockers, or hERG blockers. The use of human iPSC-
derived cardiomyocytes provides a number of additional
advantages. First is that they are human cells. Murine or rat
cells are known to have a higher beating rate, different
expression levels of ion channels, and in some cases differ-
ent pharmacologic response. Second, iPSC-derived cardio-
myocytes are developed using noninvasive methods, are
scalable, and are extensively characterized for developmen-
tal stage and expression of important signaling molecules.6
The limitations of this system include relative immaturity
of cells in comparison with primary cells and also more
complex techniques for cell maintenance in comparison
with more common immortalized cell lines. The well-to-
well precision obtained is more than sufficient to support an
assay that is suitable for a screening environment.
Further developments are planned for this assay system. It
is important to expand the range of compounds assessed to
establish this assay’s capabilities and understand any limita-
tions to using this assay for predictive cardiotoxicity. These
will include known cardioactive and cardiotoxic drug panels
and, specifically, compounds that have not been identified by
other techniques as toxic but have been withdrawn from the
market. We consider additional optimization of assay proto-
cols necessary to ensure the highest sensitivity and reproduc-
ibility of the assay. Extensions of this assay to 1536-well or
other smaller-volume formats to improve throughput and
decrease the amount of cells and reagents required to perform
the assay will remain an active area of pursuit in the near
future. There is also a desire to increase the camera acquisi-
tion speed to enable faster kinetic measurements and see if it
improves the information provided by the assay. This could
be important for compounds such as Na+ blockers that have
an effect on the contraction rise time. Last, it is important to
understand whether this assay can be used to predict mecha-
nism of action. This would expand its usefulness beyond phe-
notypic drug safety and toxicology assessment to drug
discovery and development.
In summary, we describe a method and protocols for
monitoring spontaneous contractions of cardiomyocytes
that produce highly reproducible results and as such is an
excellent assay candidate for primary screening and predic-
tive toxicology. This system is likely to provide the capabil-
ity to get information about cardiac beating in HTS format,
52 Journal of Biomolecular Screening 18(1)
prior to preclinical models or studies in human subjects.
The assay overcomes limitations of other cardiac toxicity
studies (e.g., electrophysiology screening for potential
blockers of the hERG channel). It provides measurements
based on a more physiologically relevant cell system that is
closer to native human cardiomyocytes. It provides a
higher-throughput, automated platform that includes on-
board liquid addition; simultaneous, whole-plate read in
less than 2 min; and real-time data analysis. It directly mea-
sures the cardiomyocyte contractions using a Ca2+ flux, a
key component in the contractile process. These attributes
make it attractive for screening of drug candidates for car-
diac efficacy and other compounds for cardio safety and
gives scientists new tools for predicting adverse effects of
compounds on myocardial performance.
Declaration of Conflicting Interests
O. Sirenko, C. Crittenden, Y-W. Chen, J. Hesley, and E.F.
Cromwell are employed by Molecular Devices, LLC, which sells
the ImageXpress Micro and FLIPR Tetra systems. B. Anson is
employed by Cellular Dynamics International which sells the
iCell Cardiomyocytes.
Funding
The authors received no financial support for the research, author-
ship, and/or publication of this article.
References
1. Kola, I; Landis, J. Can the Pharmaceutical Industry Reduce
Attrition Rates? Nat. Rev. Drug Discov. 2004, 3, 711–715.
2. Cohen, J. D.; Babiarz, J. E.; Abrams, R. M.; Guo, L.; Kameoka,
S.; Chiao, E.; Taunton, J.; Kolaja, K. L. Use of Human Stem
Cell Derived Cardiomyocytes to Examine Sunitinib Mediated
Cardiotoxicity and Electrophysiological Alterations. Toxicol.
Appl. Pharma. 2011, 257, 74–83.
3. Guo, L.; Qian, J.-Y.; Abrams, R.; Tang, H.-M.; Weiser, T.;
Sanders, M. J.; Kolaja, K. L. The Electrophysiological Effects
of Cardiac Glycosides in Human iPSC-Derived Cardiomyo-
cytes and in Guinea Pig Isolated Hearts. Cell Physiol. Bio-
chem. 2011, 27, 453–462.
4. Ma, J.; Guo, L.; Fiene, S. J.; Anson, B. D.; Thomson, J. A.;
Kamp, T. J.; Kolaja, K. L.; Swanson, B. J.; January, C. T. High
Purity Human-Induced Pluripotent Stem Cell Derived Cardio-
myocytes: Electrophysiological Properties of Action Poten-
tials and Ionic Currents. Am. J. Physiol. Heart Circ. Physiol.
2011, 301, H2006–H2017.
5. Babiarz, J. E.; Ravon, M.; Sridhar, S.; Ravindran, P.; Swan-
son, B.; Bitter, H.; Weiser, T.; Chiao, E.; Certa, U.; Kolaja,
K. L. Determination of the Human Cardiomyocyte mRNA
and miRNA Differentiation Network by Fine-Scale Profiling.
Stem Cells Dev. 2012, 21, 1956–1965.
6. Anson, B. D.; Kolaja, K. L.; Kamp, T. J. Opportunities for Use
of Human iPS Cells in Predictive Toxicology. Clin. Pharm.
Ther. 2011, 89, 754–758.
7. Elferink, J. G.; Deierkauf, M. The Effect of Verapamil and
Other Calcium Antagonists on Chemotaxis of Polymorpho-
nuclear Leukocytes. Biochem. Pharmacol. 1984, 33, 35–39.
8. Schroeder, K.; Neagle, B. Method and Apparatus for Quanti-
tative Fluorescence Measurements. US Patent No. 5355215,
1994.
9. Tran, T. M.; Friedman, J.; Qunaibi, E.; Baameur, F.; Moore,
R. H.; Clark, R. B. Characterization of Agonist Stimulation
of cAMP-Dependent Protein Kinase and G Protein-Coupled
Receptor Kinase Phosphorylation of the Beta2-Adrenergic
Receptor Using Phosphoserine-Specific Antibodies. Mol.
Pharmacol. 2004, 65, 196–206.
10. Harding, S. E.; Vescovo, G.; Kirby, M.; Jones, S. M.; Gur-
den, J.; Poole-Wilson, P. A. Contractile Responses of Isolated
Adult Rat and Rabbit Cardiac Myocytes to Isoproterenol and
Calcium. J. Mol. Cell Cardiol. 1988, 20, 635–647.
11. Habuchi, Y.; Tanaka, H.; Nishio, M.; Yamamoto, T.; Komori,
T.; Morikawa, J.; Yoshimura, M. Dopamine Stimulation of
Cardiac β-adrenoceptors: The Involvement of Sympathetic
Amine Transporters and the Effect of SKF38393. Br. J.
Pharm. 1997, 122, 1669–1678.
12. Wible, B. A.; Hawryluk, P.; Ficker, E.; Kuryshev, Y. A.;
Kirsch, G.; Brown, A. M. HERG-Lite: A Novel Comprehen-
sive High-Throughput Screen for Drug-Induced hERG Risk.
J. Pharmacol. Toxicol. Methods 2005, 52, 136–145.
13. Kang, J.; Chen, X. L.; Ji, J.; Lei, Q.; Rampe, D. Ca2+ Channel
Activators Reveal Differential L-type Ca2+ Channel Pharma-
cology between Native and Stem Cell-Derived Cardiomyo-
cytes. J. Pharmacol. Exp. Ther. 2012, 341, 510–517.
14. Thomas, T.; Gurung, I. S.; Killeen, M. J.; Hakim, P.; Goddard,
C. A.; Mahaut-Smith, M. P.; Colledge, W. H.; Grace, A. A.;
Huang, C. L. Effects of L-type Ca2+ Channel Antagonism on
Ventricular Arrhythmogenesis in DKPQ Scn5a (Long QT3)
Murine Hearts. J. Physiol. 2007, 578, 85–97.
15. Wang, L.; Wible, B. A.; Wan, X.; Ficker, E. Cardiac Glycosides as
Novel Inhibitors of Human Ether-a-Go-Go-Related Gene Chan-
nel Trafficking. J. Pharmacol. Exp. Ther. 2007, 320, 525–534.
16. Brandts, B.; Dirkmann, D.; Borchard, R.; Wickenbrock, I.;
van Bracht, M.; Prull, M. W.; Trappe, H. J. Propranolol Inhib-
its IK(Ado) by Competitive A1-Receptor Interaction. Cardiol-
ogy 2000, 93, 533–539.
17. Hill, R. J.; Duff, H. J.; Sheldon, R. S. Class I Antiarrhythmic
Drug Receptor: Biochemical Evidence for State-Dependent
Interaction with Quinidine and Lidocaine. Mol. Pharm. 1989,
36, 150–159.
18. Stoelzle, S.; Haythornthwaite, A.; Kettenhofen, R.; Kolossov, E.;
Bohlen, H.; George, M.; Brüggemann, A.; Fertig, N. Automated
Patch Clamp on mESC-Derived Cardiomyocytes for Cardiotox-
icity Prediction. J. Biomol. Screen. 2011, 16, 910–916.
19. Abassi, Y. A.; Xi, B.; Li, N.; Ouyang, W.; Seiler, A.; Watzele,
M.; Kettenhofen, R.; Bohlen, H.; Ehlich, A.; Kolossov, E.; Wang,
X.; Xu, X. Dynamic Monitoring of Beating Periodicity of Stem
Cell Derived Cardiomyocytes as a Predictive Tool for Preclinical
Safety Assessment. B r. J. Pharma. 2012, 165, 1424–1441.
Sirenko et al. 53
20. Victor, M. A.; Hani, N.; Sabbah, H. N.; Higgins, R. S. D.; Sil-
verman, N.; Lesch, M.; Undrovinas, A. I. Novel, Ultraslow
Inactivating Sodium Current in Human Ventricular Cardio-
myocytes. Circulation. 1998, 98, 2545–2552.
21. Yamamoto, M.; Gotoh, Y.; Imaizumi, Y.; Watanabe, M. Mech-
anisms of Long-Lasting Effects of Benidipine on Ca Current
in Guinea-Pig Ventricular Cells. Br. J. Pharma. 1990, 100,
669–676.
22. Preuss, K. C.; Gross, G. J.; Brooks, H. L.; Warltier, D. C. Slow
Channel Calcium Activators, a New Group of Pharmacologi-
cal Agents. Life Sci. 1985, 37, 1271–1278.
23. Hwang, K. S.; Van Breemen, C. Effects of the Ca Agonist Bay
K8644 on 45Ca Influx and Net Ca Uptake into Rabbit Aortic
Smooth Muscle. Eur. J. Pharma. 1985, 22, 299–305.
24. Rampe, D.; Roy, M. L.; Dennis, A.; Brown, A. M. A Mecha-
nism for the Proarrhythmic Effects of Cisapride (Propulsid):
High Affinity Blockade of the Human Cardiac Potassium
Channel HERG. FEBS Lett. 1997, 417, 28–32.
25. Kang, J.; Wang, L.; Cai, F.; Rampe, D. High Affinity Blockade
of the HERG cardiac K+ Channel by the Neuroleptic Pimo-
zide. Eur. J. Pharm. 2000, 392, 137–140.
26. Enyeart, J. J.; Dirksen, R. T.; Sharma, V. K.; Williford, D. J.;
Sheu, S. S. Antipsychotic Pimozide Is a Potent Ca2+ Channel
Blocker in Heart. Mol. Pharmacol. 1990, 37, 752–757.
27. Redfern, W. S.; Carlsson, L.; Davis, A. S.; Lynch, W. G.;
MacKenzie, I.; Palethorpe, S.; Siegl, P. K.; Strang, I.; Sulli-
van, A. T.; Wallis, R.; Camm, A. J.; Hammond, T. G. Relation-
ships between Preclinical Cardiac Electrophysiology, Clinical
QT Interval Prolongation and Torsade de Pointes for a Broad
Range Of Drugs: Evidence for a Provisional Safety Margin in
Drug Development. Cardiovasc. Res. 2003, 58, 32–45.
28. de Boer, T. P.; Nalos, L.; Stary, A.; Kok, B.; Houtman, M.
J.; Antoons, G.; van Veen, T. A.; Beekman, J. D.; de Groot,
B. L.; Opthof, T.; Rook, M. B.; Vos, M. A.; van der Heyden,
M. A. The Anti-protozoal Drug Pentamidine Blocks KIR2.x-
mediated Inward Rectifier Current by Entering the Cytoplas-
mic Pore Region of the Channel. Br. J. Pharm. 2010, 159,
1532–1541.
29. Shah, R. R. Drugs, QT Interval Prolongation and ICH E14:
The Need to Get It Right. Drug Saf. 2005, 28, 115–125.
30. Laverty, H. G.; Benson, C.; Cartwright, E. J.; Cross, M. J.;
Garland, C.; Hammond, T.; Holloway, C.; McMahon, N.;
Milligan, J.; Park, B. K.; Pirmohamed, M.; Pollard, C.; Rad-
ford, J.; Roome, N.; Sager, P.; Singh, S.; Suter, T.; Suter, W.;
Trafford, A.; Volders, P. G. A.; Wallis, R.; Weaver, R.; York,
M.; Valentin, J. P. How Can We Improve Our Understanding
of Cardiovascular Safety Liabilities to Develop Safer Medi-
cines? Br. J. Pharma. 2011, 163, 675–693.
31. Fermini, B.; Fossa, A. A. The Impact of Drug-Induced QT
Interval Prolongation on Drug Discovery and Development.
Nat. Rev. Drug Discov. 2003, 2, 439–447.
32. Xi, B.; Wang, T.; Li, N.; Ouyang, W.; Zhang, W.; Wu, J.; Xu,
X.; Wang, X.; Abassi, Y. A. Functional Cardiotoxicity Profiling
and Screening Using the xCELLigence RTCA Cardio System.
J Lab Autom. 2011, 16, 415–421.
33. Benham, C. D.; Bolton, T. B. Spontaneous Transient Outward
Currents in Single Visceral and Vascular Smooth Muscle Cells
of the Rabbit. J. Physiol. 1986, 381, 385–406.
34. Sachinidis, A.; Fleischmann, B. K.; Kolossov, E.; Wartenberg,
M.; Sauer, H.; Hescheler, J. Cardiac Specific Differentiation
of Mouse Embryonic Stem Cells. Cardiovas. Res. 2003, 58,
278–291.
35. Grynkiewicz, G.; Poenie, M.; Tsien, R. Y. A New Generation
of Ca2+ Indicators with Greatly Improved Fluorescence Prop-
erties. J. Biol. Chem. 1985, 260, 3440–3450.
36. Cheng, H.; Lederer, W. J.; Cannell, M. B. Calcium Sparks:
Elementary Events Underlying Excitation-Contraction Cou-
pling in Heart Muscle. Science 1993, 262, 740–744.
37. Malgaroli, A.; Milani, D.; Meldolesi, J.; Pozzanr, T. Fura-2
Measurement of Cytosolic Free Ca2+ in Monolayers and Sus-
pensions of Various Types of Animal Cells. J. Cell Biol. 1987,
105, 2145–2155.
38. Guatimosim, S.; Dilly, K.; Santana, L. F.; Saleet Jafri, M.;
Sobie, E. A.; Lederer, W. J. Local Ca(2+) Signaling and EC
Coupling in Heart: Ca(2+) Sparks and the Regulation of the
[Ca(2+)](i) Transient. J. Mol. Cell Cardiol. 2002, 34, 941–
950.
39. Bolton, T. B. Calcium Events in Smooth Muscles and Their
Interstitial Cells: Physiological Roles of Sparks. J. Physiol.
2006, 570, 5–11.
40. Tian, Q.; Oberhofer, M.; Ruppenthal, S.; Scholz, A.; Bus-
chmann, V.; Tsutsui, H.; Miyawaki, A.; Zeug, A.; Lipp, P.;
Kaestner, L. Optical Action Potential Screening on Adult Ven-
tricular Myocytes as an Alternative QT-Screen. Cell Physiol.
Biochem. 2011, 27, 281–290.
41. Tsutsui, H.; Karasawa, S.; Okamura, Y.; Miyawaki, A. Improv-
ing Membrane Voltage Measurements Using FRET with New
Fluorescent Proteins. Nat. Methods 2008, 5, 683–685.
42. Baxter, D. F.; Kirk, M.; Garcia, A. F.; Raimondi, A.; Hol-
mquist, M. H.; Flint, K. K.; Bojanic, D.; Distefano, P. F.;
Curtis, R.; Xie, Y. A Novel Membrane Potential-Sensitive
Fluorescent Dye Improves Cell-Based Assays for Ion Chan-
nels. J. Biomol. Screen. 2002, 7, 79–85.
43. Guo, L.; Abrams, R. M. C.; Babiarz, J. E.; Cohen, J. D.;
Kameoka, S.; Sanders, M. J.; Chiao, E.; Kolaja, K. L. Estimat-
ing the Risk of Drug-Induced Proarrhythmia Using Human
Induced Pluripotent Stem Cell–Derived Cardiomyocytes.
Toxicol. Sci. 2011, 123, 281–289.
... Human iPSC-derived cardiomyocyte cell lines were divided into 6 testing batches, sex and ancestral backgrounds were balanced among batches. The cell culture conditions were performed as described in previous publications [41,42]. Briefly, the tissue-culture treated 384well plates were coated with 25 µL/well of 0.1% (w/v) gelatin solution (gelatin from porcine skin diluted in cell culture grade water), the plates were then incubated for 2 h at 37 °C and 5% CO 2 . ...
... The EarlyTox Cardiotoxicity Ca 2+ flux assay kit was used to evaluate the functional effects of the PFAS on the iPSC-derived cardiomyocytes as demonstrated previously [41,42]. Intracellular Ca 2+ flux is measured using a series of time-resolved images (8 frames per s) using FLIPR Tetra Cellular Screening System (Molecular Devices) as a quantitative functional readout based on a fluorescent Ca 2+ probe. ...
... Then, 12.5 µL of the 5× working solution was added simultaneously to each well with cells already containing 50 µL (25 µL maintenance media and 25 µL of calcium flux dye) using the automated liquid handler in the FLIPR Tetra (Molecular Devices) to yield the final concentrations of 0.1, 1, 10, 100 µM (each in 0.5% DMSO) for each test substance. The concentration of 0.5% DMSO in assay wells was consistent with previous studies and it itself has no effect on the viability of cardiomyocytes [42,43]. FLIPR Tetra settings were set to mix and then transfer test chemicals from the 5× plate to the plate with cells at height = 40 µL, speed = 1 µL/s, and removal speed = 6 mm/s. ...
Article
Full-text available
Per- and poly-fluoroalkyl substances (PFAS) are emerging contaminants of concern because of their wide use, persistence, and potential to be hazardous to both humans and the environment. Several PFAS have been designated as substances of concern; however, most PFAS in commerce lack toxicology and exposure data to evaluate their potential hazards and risks. Cardiotoxicity has been identified as a likely human health concern, and cell-based assays are the most sensible approach for screening and prioritization of PFAS. Human-induced pluripotent stem cell (iPSC)-derived cardiomyocytes are a widely used method to test for cardiotoxicity, and recent studies showed that many PFAS affect these cells. Because iPSC-derived cardiomyocytes are available from different donors, they also can be used to quantify human variability in responses to PFAS. The primary objective of this study was to characterize potential human cardiotoxic hazard, risk, and inter-individual variability in responses to PFAS. A total of 56 PFAS from different subclasses were tested in concentration-response using human iPSC-derived cardiomyocytes from 16 donors without known heart disease. Kinetic calcium flux and high-content imaging were used to evaluate biologically-relevant phenotypes such as beat frequency, repolarization, and cytotoxicity. Of the tested PFAS, 46 showed concentration-response effects in at least one phenotype and donor; however, a wide range of sensitivities were observed across donors. Inter-individual variability in the effects could be quantified for 19 PFAS, and risk characterization could be performed for 20 PFAS based on available exposure information. For most tested PFAS, toxicodynamic variability was within a factor of 10 and the margins of exposure were above 100. This study identified PFAS that may pose cardiotoxicity risk and have high inter-individual variability. It also demonstrated the feasibility of using a population-based human in vitro method to quantify population variability and identify cardiotoxicity risks of emerging contaminants.
... To overcome this issue, high-throughput pre-clinical cardiac safety assessment methods, which yield CM characterization without hologram reconstruction, are needed. While there have been attempts for a fast transformation, they still take non-negligible time for CM characterization, including patch clamping [27,28], calcium imaging and/or calcium transients [29][30][31], and image processing-based contraction-relaxation studies [32,33]. All the abovementioned methods require expensive equipment or specific expertise, which demonstrates the shortcomings in simplified methods for CM motion characterization. ...
Article
Full-text available
Traditional cell analysis approaches based on quantitative phase imaging (QPI) necessitate a reconstruction stage, which utilizes digital holography. However, phase retrieval processing can be complicated and time-consuming since it needs numerical reconstruction and then phase unwrapping. For analysis of cardiomyocyte (CM) dynamics, it was reported that by estimating the spatial variance of the optical path difference from QPI, the spatial displacement of CMs can be quantified, thereby enabling monitoring of the excitation-contraction activity of CMs. Also, it was reported that the Farnebäck optical flow method could be combined with the holographic imaging information from QPI to characterize the contractile motion of single CMs, enabling monitoring of the mechanical beating activity of CMs for cardiotoxicity screening. However, no studies have analyzed the contractile dynamics of CMs based on raw holograms. In this paper, we present a fast, label-free, and high throughput method for contractile dynamic analysis of human-induced pluripotent stem cell-derived CMs using raw holograms or the filtered holograms, which are obtained by filtering only The proposed approach obviates the need for time-consuming numerical reconstruction and phase unwrapping for CM’s dynamic analysis while still having performance comparable to that of the previous methods. Accordingly, we developed a computational algorithm to characterize the CM’s functional behaviors from contractile motion waveform obtained from raw or filtered holograms, which allows the calculation of various temporal metrics related to beating activity from contraction-relaxation motion-speed profile. To the best of our knowledge, this approach is the first to analyze drug-treated CM’s dynamics from raw or filtered holograms without the need for numerical phase image reconstruction. For one hologram, the reconstruction process itself in the existing methods takes at least three times longer than the process of tracking the contraction-relaxation motion-speed profile using optical flow in the proposed method. Furthermore, our proposed methodology was validated in the toxicity screening of two drugs (E-4031 and isoprenaline) with various concentrations. The findings provide information on CM contractile motion and kinetics for cardiotoxicity screening.
... To overcome this issue, high-throughput preclinical cardiac safety assessment methods, which yield CM characterization without hologram reconstruction, are needed. While there have been attempts for a fast transformation, they still take non-negligible time for CM characterization, including patch clamping [26,27], calcium imaging and/or calcium transients [28][29][30], and image processing-based contraction-relaxation studies [31,32]. All the above-mentioned methods require expensive equipment or specific expertise, which demonstrates the shortcomings in simplified methods for CM motion characterization. ...
Preprint
Traditional cell analysis approaches based on quantitative phase imaging (QPI) necessitate a reconstruction stage, which utilize digital holography. However, phase retrieval processing can be complicated and time-consuming since it needs numerical reconstruction and then phase unwrapping. For analysis of cardiomyocyte dynamics, it was reported that by estimating the spatial variance of the optical path difference from QPI, the spatial displacement of cardiomyocyte (CM) can be quantified, thereby enabling monitoring of the excitation-contraction activity of CMs. Also, it was reported that the Farnebäck optical flow method could be combined with the holographic imaging information from QPI to characterize the contractile motion of single CMs, enabling monitoring of the mechanical beating activity of CMs for cardiotoxicity screening. However, no studies have analyzed the contractile dynamics of cardiomyocytes based on raw holograms. In this paper, we present a fast, label-free, and high throughput method for contractile dynamic analysis of human-induced pluripotent stem cell-derived CMs using raw holograms or the filtered holograms, which are obtained by filtering only the spectrum corresponding to the real image from the raw hologram, by applying the Farnebäck optical flow method for cardiotoxicity screening. The proposed approach obviates the need for time-consuming numerical reconstruction and phase unwrapping for CM’s dynamic analysis while it still has performance comparable to that of the previous methods. Accordingly, we developed a computational algorithm to characterize the CM’s functional behaviors from contractile motion waveform obtained from raw or filtered holograms, which allows the calculation of various temporal metrics related to beating activity from contraction-relaxation motion-speed profile. To the best of our knowledge, this approach is the first to analyze drug-treated CM’s dynamics from raw or filtered holograms without the need for numerical phase image reconstruction. For one hologram, the reconstruction process itself in the existing methods takes at least three times longer than the process of tracking the contraction-relaxation motion-speed profile using optical flow in the proposed method. Furthermore, our proposed methodology was validated in the toxicity screening of two drugs (E-4031 and Isoprenaline) with various concentrations. The findings provide information on CM contractile motion and kinetics for cardiotoxicity screening.
... Meanwhile, the fluctuating signals of the calcium transient stemming from each cell can be extracted and traced (Fig. 3d), and the key parameters, such as the rise time to peak (t r ), period (t p ), and maximum fluorescence (F max ), were calculated using electrophysiology software 57 further described in the Methods section. Our results yielded the mean values of t p = 2.71 ± 0.05 s, F max = 1.80 ± 0.41, and t r = 119.5 ± 27.2 ms (n = 3 cells), aligning well with previously reported studies 58,59 and thereby demonstrating the consistency and reliability of sHAPR for studying excitation-contraction transients. ...
Article
Full-text available
Fluorescence microscopy has undergone rapid advancements, offering unprecedented visualization of biological events and shedding light on the intricate mechanisms governing living organisms. However, the exploration of rapid biological dynamics still poses a significant challenge due to the limitations of current digital camera architectures and the inherent compromise between imaging speed and other capabilities. Here, we introduce sHAPR, a high-speed acquisition technique that leverages the operating principles of sCMOS cameras to capture fast cellular and subcellular processes. sHAPR harnesses custom fiber optics to convert microscopy images into one-dimensional recordings, enabling acquisition at the maximum camera readout rate, typically between 25 and 250 kHz. We have demonstrated the utility of sHAPR with a variety of phantom and dynamic systems, including high-throughput flow cytometry, cardiomyocyte contraction, and neuronal calcium waves, using a standard epi-fluorescence microscope. sHAPR is highly adaptable and can be integrated into existing microscopy systems without requiring extensive platform modifications. This method pushes the boundaries of current fluorescence imaging capabilities, opening up new avenues for investigating high-speed biological phenomena.
... μM) for calcium transients (Fig. 6E). These results demonstrate that lidocaine's inhibitory effects are of similar magnitude as compared to the values in the literature [45,[47][48][49]. Overall, these results validate the utility of the new multiwell plate device with nanocomposite microwires and carbon electrodes in drug testing. ...
Article
Full-text available
The successful translation of organ-on-a-chip devices requires the development of an automated workflow for device fabrication, which is challenged by the need for precise deposition of multiple classes of materials in micro-meter scaled configurations. Many current heart-on-a-chip devices are produced manually, requiring the expertise and dexterity of skilled operators. Here, we devised an automated and scalable fabrication method to engineer a Biowire II multiwell platform to generate human iPSC-derived cardiac tissues. This high-throughput heart-on-a-chip platform incorporated fluorescent nanocomposite microwires as force sensors, produced from quantum dots and thermoplastic elastomer, and 3D printed on top of a polystyrene tissue culture base patterned by hot embossing. An array of built-in carbon electrodes was embedded in a single step into the base, flanking the microwells on both sides. The facile and rapid 3D printing approach efficiently and seamlessly scaled up the Biowire II system from an 8-well chip to a 24-well and a 96-well format, resulting in an increase of platform fabrication efficiency by 17,5000–69,000% per well. The device's compatibility with long-term electrical stimulation in each well facilitated the targeted generation of mature human iPSC-derived cardiac tissues, evident through a positive force-frequency relationship, post-rest potentiation, and well-aligned sarcomeric apparatus. This system's ease of use and its capacity to gauge drug responses in matured cardiac tissue make it a powerful and reliable platform for rapid preclinical drug screening and development.
... The generation and modulation of the concentration of cytoplasmic calcium [Ca 2+ ] i oscillations in neurons involve inotropic GABA(A)R, NMDAR, and CP-AMPAR receptors, as well as signaling systems that ensure the release of calcium from the ER [264,265,[274][275][276][277]. Oscillations in calcium and other signaling molecules in electronically non-excitable cells are currently being actively studied: endocrine cells, glia, cardiomyocytes, adipocytes, and endotheliocytes [12,262,264,266,270,274,288]. ...
Article
Full-text available
The presence of biological rhythms is a characteristic of all living organisms. Over the past 60 years, scientists around the world have accumulated a huge amount of data on rhythmic processes in living systems at various levels. The acquired knowledge has found applications in human economic activity and medicine. The ultradian (less than a day) rhythms at the organismal, organ, and cellular levels are characterized by high diversity. Unfortunately, biorhythms in different systems are considered, most often, in isolation from each other. Much knowledge about biorhythms was obtained using expert evaluation methods, and later methods of spectral analysis were used to describe biorhythms. Ultradian rhythms have a relatively short duration; therefore, they can be characterized by spectral analysis methods. More and more researchers believe that in order to further expand the understanding of the nature and purpose of biorhythms, the use of more advanced methods of mathematical processing is required, and rhythms in different organs, tissues, and cells should be considered parts of a single system. This review is intended to provide the reader with the variety of ultradian rhythms in living systems (organismal, organ, cellular, molecular levels), the mechanisms of their generation, and their functions to give the reader a picture of the possible relationships between these rhythms. Further, the reader will be able to get acquainted with the variety of mathematical methods for analyzing biorhythms, including bispectral and cross-correlation analyses.
Article
Full-text available
In our current study, we developed a focused series of original ((benzyloxy)benzyl)propanamide derivatives that demonstrated potent activity across in vivo mouse seizure models, specifically, maximal electroshock (MES) and 6 Hz (32 mA) seizures. Among these derivatives, compound 5 emerged as a lead molecule, exhibiting robust protection following intraperitoneal (i.p.) injection, as follows: ED50 = 48.0 mg/kg in the MES test, ED50 = 45.2 mg/kg in the 6 Hz (32 mA) test, and ED50 = 201.3 mg/kg in the 6 Hz (44 mA) model. Additionally, compound 5 displayed low potential for inducing motor impairment in the rotarod test (TD50 > 300 mg/kg), indicating a potentially favorable therapeutic window. In vitro toxicity assays further supported its promising safety profile. We also attempted to identify a plausible mechanism of action of compound 5 by applying both binding and functional in vitro studies. Overall, the data obtained for this lead molecule justifies the more comprehensive preclinical development of compound 5 as a candidate for a potentially broad-spectrum and safe anticonvulsant.
Article
Full-text available
Engineered human cardiac tissues have been utilized for various biomedical applications, including drug testing, disease modeling, and regenerative medicine. However, the applications of cardiac tissues derived from human pluripotent stem cells are often limited due to their immaturity and lack of functionality. Therefore, in this study, we establish a perfusable culture system based on in vivo-like heart microenvironments to improve human cardiac tissue fabrication. The integrated culture platform of a microfluidic chip and a three-dimensional heart extracellular matrix enhances human cardiac tissue development and their structural and functional maturation. These tissues are comprised of cardiovascular lineage cells, including cardiomyocytes and cardiac fibroblasts derived from human induced pluripotent stem cells, as well as vascular endothelial cells. The resultant macroscale human cardiac tissues exhibit improved efficacy in drug testing (small molecules with various levels of arrhythmia risk), disease modeling (Long QT Syndrome and cardiac fibrosis), and regenerative therapy (myocardial infarction treatment). Therefore, our culture system can serve as a highly effective tissue-engineering platform to provide human cardiac tissues for versatile biomedical applications.
Article
Full-text available
Cellular phenotypes and functional responses are modulated by the signals present in their microenvironment, including extracellular matrix (ECM) proteins, tissue mechanical properties, soluble signals and nutrients, and cell–cell interactions. To better recapitulate and analyze these complex signals within the framework of more physiologically relevant culture models, high throughput culture platforms can be transformative. High throughput methodologies enable scientists to extract increasingly robust and broad datasets from individual experiments, screen large numbers of conditions for potential hits, better qualify and predict responses for preclinical applications, and reduce reliance on animal studies. High throughput cell culture systems require uniformity, assay miniaturization, specific target identification, and process simplification. In this review, we detail the various techniques that researchers have used to face these challenges and explore cellular responses in a high throughput manner. We highlight several common approaches including two‐dimensional multiwell microplates, microarrays, and microfluidic cell culture systems as well as unencapsulated and encapsulated three‐dimensional high throughput cell culture systems, featuring multiwell microplates, micromolds, microwells, microarrays, granular hydrogels, and cell‐encapsulated microgels. We also discuss current applications of these high throughput technologies, namely stem cell sourcing, drug discovery and predictive toxicology, and personalized medicine, along with emerging opportunities and future impact areas.
Article
Full-text available
A new family of highly fluorescent indicators has been synthesized for biochemical studies of the physiological role of cytosolic free Ca2+. The compounds combine an 8-coordinate tetracarboxylate chelating site with stilbene chromophores. Incorporation of the ethylenic linkage of the stilbene into a heterocyclic ring enhances the quantum efficiency and photochemical stability of the fluorophore. Compared to their widely used predecessor, “quin2”, the new dyes offer up to 30-fold brighter fluorescence, major changes in wavelength not just intensity upon Ca2+ binding, slightly lower affinities for Ca2+, slightly longer wavelengths of excitation, and considerably improved selectivity for Ca2+ over other divalent cations. These properties, particularly the wavelength sensitivity to Ca2+, should make these dyes the preferred fluorescent indicators for many intracellular applications, especially in single cells, adherent cell layers, or bulk tissues.
Article
The fluorescent indicator fura-2 has been applied to a variety of cell types in order to set up appropriate conditions for measurements of the cytosolic concentration of free ionized Ca2+ [( Ca2+]i) in both cell suspensions and single cells analyzed in a conventional fluorimeter or in a fluorescence microscope equipped for quantitative analyses (with or without computerized image analyses), respectively. When the usual procedure for fluorescence dye loading (i.e., incubation at 37 degrees C with fura-2 acetoxy-methyl ester) was used, cells often exhibited a nonhomogeneous distribution of the dye, with marked concentration in multiple small spots located preferentially in the perinuclear area. These spots (studied in detail in human skin fibroblasts), were much more frequent in attached than in suspended cells, and were due to the accumulation (most probably by endocytosis) of the dye within acidic organelles after hydrolysis by lysosomal enzyme(s). When loading with fura-2 was performed at low (15 degrees C) temperature, no spots appeared, and cells remained diffusely labeled even after subsequent incubation at 32-37 degrees C for up to 2 h. Homogeneous distribution of the dye is a prerequisite for appropriate [Ca2+]i measurement. In fact, comparison of the results obtained in human skin fibroblasts labeled at either 37 or 15 degrees C demonstrated in spotty cells a marked apparent blunting of Ca2+ transients evoked by application of bradykinin. Additional problems were encountered when using fura-2. Leakage of the dye from loaded cells to the extracellular medium markedly affected the measurements in cell suspensions. This phenomenon was found to depend on the cell type, and to markedly decrease when temperature was lowered, suggesting the involvement of a facilitated transport. Calibration of fluorescence signals in terms of absolute [Ca2+]i was complicated by the increased fluorescence of fura-2 in the intracellular environment. To solve this problem we propose an in situ calibration procedure based on measurements carried out on cells in which [Ca2+]i was massively lowered (by loading the probe in a Ca2+-free medium) or increased (by treatment with the Ca2+ ionophore ionomycin, applied in a medium containing 3 mM Ca2+). These results provide explanations and, at least partial, solutions to the major problems encountered when using fura-2, and should thus be of help in clarifying the proper usage of the dye in [Ca2+]i measurements.
Article
S. Guatimosim, K. Dilly, L. F. Santana, M. S. Jafri, E. A. Sobie and W. J. Lederer. Local Ca2+ Signaling and EC Coupling in Heart: Ca2+ Sparks and the Regulation of the [Ca2+]i Transient. Journal of Molecular and Cellular cardiology (2002) 34, 941–950. The elementary event of Ca2+ release in heart is the Ca2+ spark. It occurs at a low rate during diastole, activated only by the low cytosolic [Ca2+]i. Synchronized activation of many sparks is due to the high local [Ca2+]i in the region surrounding the sarcoplasmic reticulum (SR) Ca2+ release channels and is responsible for the systolic [Ca2+]i transient. The biophysical basis of this calcium signaling is discussed. Attention is placed on the local organization of the ryanodine receptors (SR Ca2+ release channels, RyRs) and the other proteins that underlie and modulate excitation–contraction (EC) coupling. A brief review of specific elements that regulate SR Ca2+ release (including SR lumenal Ca2+ and coupled gating of RyRs) is presented. Finally integrative calcium signaling in heart is presented in the context of normal heart function and heart failure.
Article
Mechanisms underlying β‐adrenoceptor stimulation by dopamine were examined on guinea‐pig Langendorff‐perfused hearts and isolated cells from the right atrium, by using the chronotropic effects and the enhancement of L‐type Ca ²⁺ current ( I Ca,L ) in the presence of prazosin as indicators of β‐adrenoceptor stimulation. Dopamine‐induced overflow of noradrenaline (NA) concentrations was measured by high‐performance liquid chromatography. Dopamine caused positive chronotropic effects with an EC 50 of 2.5 μ m and induced NA overflow with a similar EC 50 (1.3 μ m ). The chronotropic effect of dopamine was abolished by bisoprolol (1 μ m ). The effects of dopamine were maintained during prolonged application, whereas the effects of tyramine faded with time. Dopamine (3 μ m ) restored the chronotropic effects and the NA release suppressed by pretreatment with tyramine, suggesting a de novo synthesis of NA during the exposure to dopamine. Dopamine (3 μ m )‐induced NA release was not affected by tetrodotoxin, ω‐conotoxin, rauwolscine, ICI118551 or sulpiride, but was inhibited by desipramine, a NA uptake inhibitor (IC 50 ∼1 μ m ). It was also not affected by GBR12909 and bupropion, dopamine uptake inhibitors in the central nervous system. SKF38393, a D 1 receptor partial agonist, potently inhibited the 3 μ m dopamine‐induced release of NA (IC 50 ∼0.1 μ m ). D 1 receptors are not involved in the DA‐induced release of NA, since SCH23390 (3 μ m ), a potent D 1 antagonist, inhibited the NA release only slightly, and dihydrexidine (1 μ m ) and chloro‐APB (1 μ m ), full D 1 agonists, caused no significant NA release. SKF38393 inhibited tyramine‐induced overflow of NA, and potentiated the field stimulation‐induced NA release. SKF38393 and desipramine retarded the decay of the stimulation‐induced tachycardia in a similar manner. These results indicate that SKF38393 is a potent monoamine transport inhibitor and a useful tool for the functional evaluation of indirectly‐acting sympathomimetic agonists in the heart. In the presence of SKF38393 (10 μ m ), dopamine at 1 μ m showed no chronotropic effect. Voltage clamp experiments with isolated atrial cells revealed that dopamine is a weak partial agonist. The EC 50 for I Ca,L stimulation by dopamine was high (13 μ m ). As a result, dopamine at 1 μ m did not affect I Ca,L . Bisoprolol abolished the stimulation of I Ca,L by dopamine (30 μ m ), and dihydrexidine (1 μ m ) did not affect I Ca,L . It was concluded that the cardiac effects of dopamine at clinically relevant concentrations (<1 μ m ) result almost exclusively from the indirect effect of β adrenoceptor stimulation, involving the release of NA from sympathetic nerve terminals. The roles of the direct stimulation of β adrenoceptors by dopamine at these concentrations and the stimulation of postjunctional D 1 receptors seem negligible. The desipramine‐ and SKF38393‐sensitive monoamine transporter mediates the release of NA.
Article
Specific calcium channels in myocardium and vascular smooth muscle and pharmacologic agents which possess the ability to block them have been the subject of intense research over the past several years. Many studies have utilized dihydropyridine derivatives (e.g. nifedipine, nitrendipine, nisoldipine) which have been shown to be efficacious inhibitors of calcium influx through voltage sensitive slow channels. Administration of these agents results in vascular smooth muscle relaxation and negative inotropic effects. Recently, novel dihydropyridines such as Bay k 8644, CGP 28 392 and YC-170, with actions diametrically opposed to those of compounds typified by nifedipine have been synthesized. These agents demonstrate vasoconstrictor and positive inotropic effects — actions which might be expected of compounds capable of stimulating the transmembrane influx of calcium into vascular smooth muscle and myocardium. Actions of Bay k 8644 and CGP 28 392 studied and have also shown that pharmacological blockade of beta or alpha adrenergic receptors does not influence the direct effects of these agents. Future analogs, with similar but more selective actions on myocardial calcium channels, may prove useful in the management of pathologic states characterized by insufficient contractile function of the heart.
Article
Pimozide is an antipsychotic agent also used to treat facial tics. Pimozide can cause acquired long QT syndrome and ventricular arrhythmias. To elucidate the mechanism behind these clinical findings, we examined the effects of pimozide on the cloned human cardiac K+ channels HERG (human ether-a-go-go-related gene; rapid component of delayed rectifier), Kv1.5 (ultra-rapid delayed rectifier) and KvLQT1/minK (slow component of delayed rectifier). Using patch clamp electrophysiology, we found that pimozide was a potent inhibitor of HERG displaying an IC50 value of 18 nM. In contrast, pimozide (10 μM) was a weak inhibitor of KvLQT1/minK and Kv1.5. We conclude that pimozide is a specific, high affinity antagonist of HERG, and that this interaction leads to prolongation of cardiac repolarization.
Article
Human stem cell-derived cardiomyocytes provide new models for studying the ion channel pharmacology of human cardiac cells for both drug discovery and safety pharmacology purposes. However, detailed pharmacological characterization of ion channels in stem cell-derived cardiomyocytes is lacking. Therefore, we used patch-clamp electrophysiology to perform a pharmacological survey of the L-type Ca²⁺ channel in induced pluripotent and embryonic stem cell-derived cardiomyocytes and compared the results with native guinea pig ventricular cells. Six structurally distinct antagonists [nifedipine, verapamil, diltiazem, lidoflazine, bepridil, and 2-[(cis-2-phenylcyclopentyl)imino]-azacyclotridecane hydrochloride (MDL 12330)] and two structurally distinct activators [methyl 2,6-dimethyl-5-nitro-4-[2-(trifluoromethyl)phenyl]-1,4-dihydropyridine-3-carboxylate (Bay K8644) and 2,5-dimethyl-4-[2-(phenylmethyl)benzoyl]-1H-pyrrole-3-carboxylic acid methyl ester (FPL 64176)] were used. The IC₅₀ values for the six antagonists showed little variability between the three cell types. However, whereas Bay K8644 produced robust increases in Ca²⁺ channel current in guinea pig myocytes, it failed to enhance current in the two stem cell lines. Furthermore, Ca²⁺ channel current kinetics after addition of Bay K8644 differed in the stem cell-derived cardiomyocytes compared with native cells. FPL 64176 produced consistently large increases in Ca²⁺ channel current in guinea pig myocytes but had a variable effect on current amplitude in the stem cell-derived myocytes. The effects of FPL 64176 on current kinetics were similar in all three cell types. We conclude that, in the stem cell-derived myocytes tested, L-type Ca²⁺ channel antagonist pharmacology is preserved, but the pharmacology of activators is altered. The results highlight the need for extensive pharmacological characterization of ion channels in stem cell-derived cardiomyocytes because these complex proteins contain multiple sites of drug action.