Download Predictive Assays for High Throughput Assessment of Cardiac

Predictive Assays for High Throughput Assessment of Cardiac Toxicity and Drug Safety
Oksana Sirenko, Carole Crittenden, Jayne Hesley, Yen-Wen Chen, Carlos Funes, Debra Gallant, and Evan F. Cromwell
Molecular Devices, LLC, 1311 Orleans Drive, Sunnyvale, CA 94089
High Throughput FLIPR® Tetra System
Cardiac Beating Assay
Cardiac toxicity is a serious drug safety concern because it can cause
arrhythmias or heart failure. We have developed methods for the ImageXpress
Micro system that enables image acquisition and determination of beating rate
of live cardiomyocytes from a series of time-lapse images. Beating
cardiomyocytes were loaded with the Calcium 5 reagents and imaged using a
10X objective at up to 100 frames per second (fps) to allow visualization of Ca2+
fluxes that occur synchronous with beating. The total fluorescence intensity was
integrated from each frame and then plotted as a function of time. The number
of peaks within each plot were counted and used to determine a beat rate. The
system allows saving data as a video, presenting intensity curves, and
automatic analysis of beat rates.
A method complementary to imaging uses the FLIPR Tetra system to
monitor changes in intracellular Ca2+ fluxes associated with cardiomyocyte
contractions using the FLIPR Calcium 5 Assay Kit. The FLIPR system allows
automatic addition of reagents and compounds simultaneous with reading
from 96, 384, or 1536 wells. This has been found to be advantageous for
the cardiac beating assays because it reduces well-to-well variability caused
by reading at different time points. The absolute beat rates were found to
be similar to those measured by imaging methods. Temporal response
curves for analysis and visualization of beating can be acquired in ~ 2 min
per plate making this assay suitable for high throughput screening of
compound libraries.
Visualizing beating cardiomyocytes by Ca2+ fluxes
A
We developed methods for the ImageXpress® Micro and the FLIPR® Tetra
systems that enable determination of beating rate from changes in
intracellular Ca2+ that are synchronous with beating. The protocols use
induced pluripotent stem cell (iPSC) derived cardiomyocytes loaded with a
calcium sensitive dye and allow monitoring of drug impact on the beat rate and
amplitude in 96 or 384 well formats. Peak parameters are automatically
measured using the Screenworks® Peak Pro™ software and proprietary
analysis algorithms. We have demonstrated use of both systems for two
important applications: screening compounds for cardiac toxicity, and preclinical testing of potential cardiac drug candidates. These methods are well
suited for safety testing and can be used to estimate efficacy and dosing of
drug candidates prior to clinical studies.
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
B
10
9
9
8
10
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
D
19
18
18
18
17
18
15
15
14
15
15
16
0
0
0
8
7
8
3
4
3
4
5
3
(figure courtesy of Cellular Dynamics International)
400
Intensity (arbitrary units)
380
ImageXpress® Micro XL
Automated Imaging System
References
High Content Imaging
•Image acquisition was
done on the
ImageXpress Micro XL
System using time-lapse
images at 20x or 10x
magnification using FITC
excitation and emission
filters.
•Data was acquired at
up to 100 frames per
second (fps)
•Cells were kept under
environmental control
(37C, 5% CO2).
19
18
19
19
17
18
10
11
10
14
14
14
0
0
0
9
8
8
6
5
4
6
6
6
F
19
18
18
17
17
16
9
8
9
13
13
14
0
0
1
9
8
7
7
5
6
7
4
7
G
19
17
18
16
15
16
8
9
9
11
12
11
0
1
2
8
9
9
7
7
7
7
7
8
H
Control
Isoproterenol
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
360
I
340
320
control
J
15
14
14
13
13
12
8
8
8
9
8
8
4
6
4
9
8
8
9
10
8
8
9
9
K
17
13
16
13
13
11
8
9
9
8
9
9
6
9
6
9
8
8
9
9
9
9
9
8
L
10
10
8
9
8
9
8
9
8
8
8
9
9
9
8
9
8
9
9
10
8
9
10
11
IC50
(µM)
Frequency
60
50
Epinephrine – β-AR agonist
0.005
40
Isoproterenol – β-AR agonist
0.55
30
Propranolol – β-AR antagonist
7.6
20
Lidocaine – Na+ channel blocker
11.3
10
Cisapride – hERG Channel Blocker
0.023
Astemizole – hERG Channel Blocker
0.006
0
300
280
1e-4
0.001
0.01
0.1
1
10
100
1000
1e4
Concentration, µM
260
220
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
Output Parameters
5
Time (23 sec)
epinephrine
Beats/ 23 sec
16
14
12
control
10
8
6
4
2
0
1e-4
0.001
0.01
0.1
Concentration, uM
1
Concentration, µM
10
verapamil
Figure 2. Top: Images of cardiomyocytes showing low
and high Calcium 5 dye signal through a beat cycle.
Middle Left: Temporal response curves of signals from
cells dosed with positive and negative chronotropes.
Bottom Left: Dose response of epinephrine (pos.
chronotrope) and verapamil (neg. chronotrope) as
determined by ImageXpress Micro XL system. Center:
Traces of single contraction events acquired at 33 fps for
an untreated well and one dosed with 10mM
isoproterenol. Middle Right: Action potential schematic
showing contributions of different ion channels.
Similarity is observed between components of the Ca2+
trace and those of the action potential.
Screenworks® Peak Pro™ Software
Automated Peak Analysis
• Configure Kinetic Reduction
• Show Kinetic Values
• Auto-Export Data
Comparison of Beat Rates for Control Wells
A
1
2
3
4
5
6
20
B
C
D
1
1
10
9
9
8
10
9
19
1
17
2
18
5
18
18
18
1
19
18
18
18
17
18
18
18
19
17
17
18
…
16
14
12
10
Peaks = 19
Rise = 0.18 sec
Freq = 21.5 bpm Decay = 0.38 sec
4
Control wells located
in rows B and L
2
0
Row B
(n=24)
Control
Exp1 24 wells
Statistics
30000
20000
Exp2 200 wells
Row L
Exp3 24 wells
average
stdev
% stdev
average
stdev
% stdev
average
Peak Count
8.8
0.7
8.2
8.5
1.4
16.8
15.9
Average Peak Width
1.0
0.2
20.9
2.3
0.3
12.0
3.0
0.4
14.5
Average Peak Amplitude
18.0
5.0
27.7
2249.3
530.3
23.6
195.1
stdev
31.8
16.3
Average Peak Baseline
59.4
5.5
9.2
2805.6
199.7
7.1
265.1
23.4
8.8
Average Peak Spacing
4.1
0.3
8.3
6.9
1.0
15.1
6.4
1.0
16.4
2.5
% stdev
12
12.5
13
Rise = 0.39 sec
Peaks = 12
Freq = 12.0 bpm Decay = 0.76 sec
Average Peak Rise Time
0.4
0.1
12.7
0.4
0.1
20.2
0.8
0.1
15.6
Average Peak Decay Time
1.0
0.1
13.5
1.4
0.2
16.6
1.0
0.1
13.5
Average Peak Width at 10%
Amplitude
1.7
0.2
9.1
2.7
0.4
15.0
3.6
0.5
14.7
Propranolol
Figure 4. Left: Cardiomyocyte contraction parameters calculated by
Screenworks Peak Pro software for three different wells. Top:
Typical reproducibility of Control Wells (CVs <= 10%) Middle:
Statistics for various parameters measured for control wells for three
different experiments.
13.5
Time (sec)
FOR RESEARCH USE ONLY. NOT FOR USE IN DIAGNOSTIC PROCEDURES. Molecular Devices, the Molecular Devices logo, and all other trademarks, unless otherwise indicated, are the property of Molecular Devices, LLC. ©2012 Molecular Devices, LLC. iCell is a registered trademark of Cellular Dynamics International.
5
4
3
2
0.001
0.01
0.1
1
10
100
1000
Peaks = 6
Rise = 0.53 sec
Freq = 7.2 bpm Decay = 1.81 sec
Summary
• We demonstrate live-cell assays for measuring the impact of
pharmacological compounds on the rate and magnitude of beating
cardiomyocytes using ImageXpress Micro Automated Microscope for
high content imaging and FLIPR Tetra Cellular Screening System.
Intracellular Ca2+ transient fluxes underlying cell contractions were
monitored by using FLIPR Calcium 5 Assay Kit readout.
• Peak parameters were automatically measured on the FLIPR Tetra
System using the Screenworks Peak Pro software and proprietary
analysis algorithms. The easy-to-use interface and intuitive setup
provide fast and reproducible results that agree well with manual
measurements.
15.8
-10000
11.5
6
8
6
E
40000
11
7
Figure 5. Dose response of six different
compounds as measured by the FLIPR Tetra
system in Calcium 5 loaded iPSC derived
cardiomyocytes. Top: Change in frequency of
contractions with dose. Bottom: Change in
average peak width with dose. Positive
chronotropes were found to increase beat rate
and decrease peak widths as expected. Ion
channel blockers were found to decrease beat
rate. The hERG channel blocker cisapride
was found to have a significant effect on
repolarization times as indicated by an
increase in peak width.
Figure 3. Top: ScreenWorks Peak Pro Software Output for 384 Well Plate 10 Minutes After Dose.
Bottom: Output parameters available to the user.
18
-20000
10.5
8
Concentration, µM
All
0
9
0
1e-4
Isoproterenol
10000
10
1
FLIPR Tetra System Assay Capabilities
Automated data analysis is required in order to make the FLIPR Tetra system
practical for running assays on a large numbers of compounds. Manual
counting of 384 traces is time consuming and subject to human error.
Exporting the data to a separate program in order to analyze the traces
removes subjectivity, but adds additional steps that compromises the
usefulness for environments that require true automation and high
throughput such as screening labs.
Output Parameters Available from
An automated peak detection and analysis
Screenworks Peak Pro Software
algorithm was added to the Screenworks
software to process the data and provide
user selectable outputs.
Peak Width
Rise
(FWHM)
First, signal peaks are detected based on a
Time
Decay Time
dynamic thresholding and derivative
Peak Width
10% amplitude
analysis. The peaks are then fit with a
Amplitude
& Time
binomial function to provide amplitude,
time, and width values. The data analysis
occurs immediately after acquisition and
• Peak Count & Frequency
result are presented on a well-by-well
• Peak Position (time) and Amplitude
basis to the user. The results can be
• Peak Width (FWHM)
• Rise Time (10% to 90%)
exported to a standard comma-separated• Decay Time (90% to 10%)
variable (csv) file for further analysis.
Peak Width
11
10% Peak Width (sec)
Time (sec)
18
1. Functional cardiomyocytes derived from human induced pluripotent stem cells.
Zhang J, Wilson G, F Soerens AG, Koonce CH, Yu J, Palecek SP, Thomson JA, Kamp TJ. Circ Res. 2009 Feb 27;104(4):e30-41.
Department of Medicine, University of Wisconsin, WiCell Research Institute, Madison, WI 53792-3248, USA.
2. In vitro reprogramming of fibroblasts into a pluripotent ES-cell-like state.
Wernig M, Meissner A, Foreman R, Brambrink T, Ku M, Hochedlinger K, Bernstein BE, Jaenisch R.
Nature. 2007 Jul 19;448(7151):318-24. Whitehead Institute for Biomedical Research, Massachusetts Institute of Technology, Cambridge, Massachusetts 02142, USA.
9
240
20
FLIPR Tetra System
Flux
Assay
• Reagents from FLIPR® Calcium 5
Assay Kit (Molecular Devices) were
added to the plates and incubated 1
hour at 37C in 5% CO2.
• Compound plates were pre-warmed
to 37C inside the FLIPR Tetra
instrument and compound addition
was done simultaneously to all wells
• Experimental plates were loaded
and a pre-drug read was acquired of
changes in intracellular Ca2+
concentration.
• Acquisition was set up for
Calcium 5 (Ex 485nm, Em
530nm)
• Data was acquired at ~8 fps
• FLIPR system reads were acquired
during compound addition and at
prescribed times after addition (read
time ~ 2 min)
9
Development of new, more potent and safer drugs requires an in vitro
system where efficacy and safety can be tested. Positive and negative
inotropes are used in clinics to treat heart failure, tachycardia,
arrhythmia or other cardiac diseases. We have demonstrated effects of
several positive (isoproterenol, dopamine, etc.) and negative (α− and β−
blockers) chronotropes on cardiac rates and determined EC50s at the
expected ranges. Image based assays using calcium flux and iPSC
derived cardiomyocytes are suitable for this task and could be used to
estimate efficacy and approximate dosing prior to pre-clinical studies.
We have also examined effects of known ion channel blockers (Na+ and
hERG) and found that they have a negative effect on beat rate and
prolong repolarization times. Results from assays run on the FLIPR Tetra
system are shown below.
Effects of β-Adrenergic Receptor (AR)
Agonists/Antagonists and Ion Channel Blockers
doxasozin
4
0
Ca2+
Figure 1. Steps in creation of cardiomyocytes and
other cell types from iPS cells.
ac-choline propranolol
verapamil
3
0
Single Cardiomyocyte Contractions
Cell Preparation
• iCell® Cardiomyocytes were received frozen from
FLIPR® Tetra Cellular Screening
System
dopamine
2
0
C
E
Methods
Cellular Dynamics International (CDI). Cells were
thawed and plated according to recommended
protocol.
• Cardiomyocytes were plated 20K/96well plate or 4
K/384 well on gelatin coated plates and incubated for
3-5 days. The presence of spontaneous contractions
indicated that the cells had appropriately matured.
digoxin
1
Predictive HTS Cell-Based Assay
Beats per Minute
isoproterenol epinephrine
All
concentration
A large percentage of new drugs fail in clinical studies due to cardiac toxicity.
Therefore development of highly predicative in vitro assays suitable for high
throughput screening (HTS) is extremely important for drug development.
Human cardiomyocytes derived from stem cell sources can greatly accelerate
the development of new chemical entities and improve drug safety by offering
more clinically relevant cell-based models than those presently available. Stem
cell derived cardiomyocytes are especially attractive because they express ion
channels and demonstrate spontaneous mechanical and electrical activity
similar to native cardiac cells. Here we demonstrate cell based assays for
measuring the impact of pharmacological compounds on the rate of beating
cardiomyocytes with different assay platforms. Cardioactive compounds are
used in clinical treatment of heart failure, arrhythmia or other cardiac diseases.
Cardiac toxicity can cause arrhythmias or heart failure. We have shown dosedependent atypical patterns caused by several cardiotoxic compounds and ion
channel blockers. Also we have demonstrated effects of several positive
(epinephrine, etc.) and negative (α and β blockers) chronotropic agents on
cardiac rates and determined EC50s.
Imaging & Analysis of Beating
Cardiomyocytes
Beats per Minute
Introduction
• We demonstrate applications of these assays for prospective
toxicity screening using iPS-derived cardiomyocytes by measuring
the impact of various chronotropes and ion channel blockers on the
beating rate and Ca2+ transient fluxes. Tested reagents modulated
the frequency of beating in line with their mode-of-action showing
the functional expression of ß-adrenergic and acetylcholine
receptors.