High-Throughput Cardiotoxicity Screening Using Mature Human Induced Pluripotent Stem Cell-Derived Cardiomyocyte Monolayers

Summary

Human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) offer an alternative to using animals for preclinical cardiotoxicity screening. A limitation to the widespread adoption of hiPSC-CMs in preclinical toxicity screening is the immature, fetal-like phenotype of the cells. Presented here are protocols for robust and rapid maturation of hiPSC-CMs.

Abstract

Human induced stem cell-derived cardiomyocytes (hiPSC-CMs) are used to replace and reduce the dependence on animals and animal cells for preclinical cardiotoxicity testing. In two-dimensional monolayer formats, hiPSC-CMs recapitulate the structure and function of the adult human heart muscle cells when cultured on an optimal extracellular matrix (ECM). A human perinatal stem cell-derived ECM (maturation-inducing extracellular matrix-MECM) matures the hiPSC-CM structure, function, and metabolic state in 7 days after plating.

Mature hiPSC-CM monolayers also respond as expected to clinically relevant medications, with a known risk of causing arrhythmias and cardiotoxicity. The maturation of hiPSC-CM monolayers was an obstacle to the widespread adoption of these valuable cells for regulatory science and safety screening, until now. This article presents validated methods for the plating, maturation, and high-throughput, functional phenotyping of hiPSC-CM electrophysiological and contractile function. These methods apply to commercially available purified cardiomyocytes, as well as stem cell-derived cardiomyocytes generated in-house using highly efficient, chamber-specific differentiation protocols.

High-throughput electrophysiological function is measured using either voltage-sensitive dyes (VSDs; emission: 488 nm), calcium-sensitive fluorophores (CSFs), or genetically encoded calcium sensors (GCaMP6). A high-throughput optical mapping device is used for optical recordings of each functional parameter, and custom dedicated software is used for electrophysiological data analysis. MECM protocols are applied for medication screening using a positive inotrope (isoprenaline) and human Ether-a-go-go-related gene (hERG) channel-specific blockers. These resources will enable other investigators to successfully utilize mature hiPSC-CMs for high-throughput, preclinical cardiotoxicity screening, cardiac medication efficacy testing, and cardiovascular research.

Introduction

Human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) have been validated on an international scale, and are available for in vitro cardiotoxicity screening¹. Highly pure hiPSC-CMs can be generated in virtually unlimited numbers, cryopreserved, and thawed. Upon replating, they also reanimate and begin contracting with a rhythm reminiscent of the human heart^2,3. Remarkably, individual hiPSC-CMs couple to each other and form functional syncytia that beat as a single tissue. Nowadays, hiPSCs are routinely derived from patient blood samples, so any person can be represented using in vitro hiPSC-CM cardiotoxicity screening assays^4,5. This creates the opportunity to perform "Clinical Trials in a Dish", with significant representation from diverse populations⁶.

One critical advantage over existing animal and animal cell cardiotoxicity screening approaches is that hiPSC-CMs utilize the full human genome and offer an in vitro system with genetic similarities to the human heart. This is especially attractive for pharmacogenomics and personalized medicine - the use of hiPSC-CMs for medication and other therapy development is predicted to provide more accurate, precise, and safe medication prescriptions. Indeed, two-dimensional (2D) hiPSC-CM monolayer assays have proven to be predictive of medication cardiotoxicity, using a panel of clinically used medications with a known risk of causing arrhythmias^1,7,8,9. Despite the vast potential of hiPSC-CMs and the promise to streamline and make drug development cheaper, there has been a reluctance to use these novel assays^10,11,12.

Until now, one major limitation of widespread adoption and acceptance of hiPSC-CM screening assays is their immature, fetal-like appearance, as well as their function. The critical issue of hiPSC-CM maturation has been reviewed and debated in the scientific literature ad nauseum^13,14,15,16. Likewise, many approaches have been employed to promote hiPSC-CM maturation, including extracellular matrix (ECM) manipulations in 2D monolayers and the development of 3D engineered heart tissues (EHTs)^17,18. At the moment, there is a widely held belief that the use of 3D EHTs will provide superior maturation relative to 2D monolayer-based approaches. However, 2D monolayers provide a higher efficiency of cell utilization and increased success in plating compared to 3D EHTs; 3D EHTs utilize greater numbers of cells, and often require the inclusion of other cell types that can confound results. Therefore, in this article, the focus is on using a simple method to mature hiPSC-CMs cultured as 2D monolayers of electrically and mechanically coupled cells.

Advanced hiPSC-CM maturation can be achieved in 2D monolayers using an ECM. The 2D monolayers of hiPSC-CMs can be matured using a soft, flexible polydimethylsiloxane coverslip, coated with basement membrane matrix secreted by an Engelbreth-Holm-Swarm mouse sarcoma cell (mouse ECM). In 2016, reports showed that hiPSC-CMs cultured on this soft ECM condition matured functionally, displaying action potential conduction velocities near adult heart values (~50 cm/s)¹⁸. Further, these mature hiPSC-CMs displayed many other electrophysiological characteristics reminiscent of the adult heart, including hyperpolarized resting membrane potential and expression of K_ir2.1. More recently, reports identified a human perinatal stem cell-derived ECM coating that promotes the structural maturation of 2D hiPSC-CMs¹⁹. Here, easy-to-use methods are presented to structurally mature 2D hiPSC-CM monolayers for use in high-throughput electrophysiological screens. Further, we provide validation of an optical mapping instrument for the automated acquisition and analysis of 2D hiPSC-CM monolayer electrophysiological function, using voltage-sensitive dyes (VSDs) and calcium-sensitive probes and proteins.

Protocol

hiPSC usage in this protocol was approved by the University of Michigan HPSCRO Committee (Human Pluripotent Stem Cell Oversight Committee). See the Table of Materials for a list of materials and equipment. See Table 1 for media and their compositions.

1. Thawing and plating commercially available cryopreserved hiPSC-CMs for maturation on a maturation-inducing extracellular matrix (MECM)

1. Warm all the reagents to room temperature and rehydrate the MECM plates with Hank's balanced salt solution (HBSS) or phosphate-buffered saline (PBS) containing calcium and magnesium, for 1 h prior to cardiomyocyte plating (200 µL of buffer per well of a 96-well plate).
2. Wash the MECM plates 2x with HBSS or PBS containing calcium and magnesium for 1 h prior to cardiomyocyte plating (200 µL of buffer per well of a 96-well plate) and keep the wells hydrated.
3. Prepare a 37 °C water bath.
4. Remove the cardiomyocyte tubes from the liquid nitrogen tank, transfer the tubes to dry ice, and slightly open the tube caps to release pressure.
   NOTE: Releasing pressure in the tubes is extremely important! If too much pressure builds inside the tubes, they could explode.
5. Reseal the tube caps and place them in the water bath to thaw for 4 min.
   NOTE: Allow them to thaw completely, to avoid cell damage due to partial thawing.
6. After the cells thaw, spray the tubes with 70% ethanol before opening. Transfer the cells into 15 mL conical tubes with a 1 mL pipette. Slowly drip 8 mL of plating medium, agitating the tube each time 1 mL is added, to allow the cells to adjust to changes in osmolarity.
   1. Wash the cryovial with 1 mL of plating medium using a 1 mL glass pipette. Then, slowly drip the wash into the 15 mL conical tube.
7. Centrifuge the tubes at ~300 × g for 5 min. Aspirate the supernatant and resuspend the pellet in 1 mL of plating medium. Remove an aliquot and perform live cell counting with a hemocytometer. Add additional plating medium to obtain 7.5 × 10⁵ cells/mL.
   NOTE: Approximately 10 mL of cell suspension is required to prepare 96 wells.
8. Dispense 100 µL of cell suspension per well of an MECM-coated 96-well plate using a multichannel pipette.
   NOTE: Make sure to avoid cell precipitation and obtain uniform cell density in all wells while plating.
9. Incubate the cells at 37 °C, 5% CO₂ for 2 days prior to changing the medium to maintenance medium (200 µL/well). Change the maintenance medium on day 5 after thawing. Perform EP assays on day 7 or later, as described previously^8,9. Change the medium every other day when opting for extending the cell culture.

2. hiPSC cardiac-directed differentiation and hiPSC-CM purification

1. Warm 1x commercially available ethylenediaminetetraacetic acid (EDTA) solution, HBSS without calcium and magnesium (HBSS--), and 6-well plates coated with solubilized basement membrane matrix secreted by an Engelbreth-Holm-Swarm mouse sarcoma cell (mouse ECM) to room temperature.
2. Mark the differentiated colonies by phase-contrast microscopy and aspirate/ablate. Wash each well with 1 mL of HBSS--. Perform two washes in wells containing >10 differentiation spots.
3. Aspirate the HBSS-- and add 1 mL of EDTA solution to each well. Incubate the plates for up to 5 min at 37 °C. Check the plates after 3 min and look for translucent white, visible colonies.
4. Aspirate the EDTA solution and add 1 mL to a single well. Dislodge the cells with 2 mL of hiPSC medium by pipetting the suspension up and down repeatedly, using a 10 mL glass pipette to detach all stem cells from the well, and transfer the cell suspension to a collection tube. Repeat the aspiration and dislodging with subsequent wells.
   NOTE: Dislodge hard to lift colonies with the tip of the glass pipette.
5. Count the stem cells and adjust the volume to plate 8.0 × 10⁵ cells/well. Culture the cells on hiPSC medium (2 mL/well) until the stem cells reach 90% confluence (this time is referred from now on as D₀).
6. Prepare 2 mL of basal differentiation medium supplemented with 4 µM of CHIR99021.
7. On D₀, wash each well of a 6-well plate of stem cells with 1 mL of HBSS per well. Replace the HBSS with basal differentiation medium supplemented with 4 µM of CHIR99021.
8. On D₁, do nothing.
9. On D₂, prepare basal differentiation medium supplemented with 4 µM of IWP4.
10. Replace the medium with 2 mL of IWP4-supplemented basal differentiation medium per well.
11. On D₃, do nothing for a ventricular-specific differentiation. For an atrial-specific differentiation, aspirate the medium and add 2 mL of basal medium supplemented with 4 µM of IWP4 and 1 µM retinoic acid (RA) solution per well.
12. On D₄, aspirate the medium and add 2 mL of basal medium per well for ventricular differentiation. For an atrial differentiation, aspirate the medium and add 2 mL of basal medium supplemented with 1 µM RA solution per well.
13. On D₅, do nothing.
14. On D₆, aspirate the medium and add 2 mL of basal medium per well (for both atrial and ventricular differentiation).
15. On D₇, do nothing.
16. On D₈, aspirate the medium and add 2 mL of cardiomyocyte maintenance medium. Change the medium every other day until cell separation, or follow a chronic drug exposure plan.

3. hiPSC-CM purification via MACS (magnetic-activated cell sorting)

1. Aspirate the cell culture medium and wash each well with 1 mL of HBSS--. Dissociate the cells by adding 1 mL of 0.25% trypsin/EDTA and incubating at 37 °C, 5% CO₂ for 10 min. Resuspend and singularize the cells in each well with 2 mL of plating medium to inactivate the trypsin/EDTA.
2. Collect the cells from the six wells into a 50 mL conical tube with a 70 µm strainer. Then, wash the strainer with 3 mL of plating medium. Count the cells.
3. Centrifuge the suspension at ~300 × g for 5 min. Aspirate the supernatant and wash the cells with 20 mL of ice-cold MACS separation buffer. Then, centrifuge again at ~300 × g for 5 min.
4. Resuspend the pellet in 80 µL of cold MACS separation buffer per 5 × 10⁶ cells. Add 20 µL of cold non-cardiomyocyte depletion cocktail (human) per 5 × 10⁶ cells. Gently mix the cell suspension and incubate on ice for 10 min.
5. Wash the sample by adding 4 mL of cold MACS separation buffer per 5 × 10⁶ cells. Centrifuge the sample at ~300 × g for 5 min and aspirate the supernatant.
6. Resuspend the pellet in 80 µL of cold MACS separation buffer per 5 × 10⁶ cells. Add 20 µL of cold anti-biotin microbeads per 5 × 10⁶ cells. Gently mix the cell suspension and incubate for 10 min on ice.
7. While the samples are incubating, place the positive depletion columns (fitted with the 30 µm pre-separation filters) on the MACS separator, and place the labeled 15 mL collection tubes below the columns. One column is needed for every 5 × 10⁶ cells.
8. Prime each column with 3 mL of cold MACS separation buffer. Mix the antibody-treated cell suspension with 2 mL of MACS separation buffer per 5 × 10⁶ cells, and add to the column.
   NOTE: Do not centrifuge! Centrifugation at this step has detrimental effects on cardiomyocyte yield.
9. Add 2 mL of MACS separation buffer to each column, and collect the flowthrough until 12 mL of flowthrough cardiomyocyte suspension is collected.
   NOTE: Never allow the columns to fully dry.
10. Centrifuge the cardiomyocytes at ~300 × g for 5 min, discard the supernatant, and suspend the cardiomyocytes in 1 mL of plating medium.
11. Count the cells to determine the concentration, adjust the volume to the desired seeding density, and plate the cells. Plate the purified cardiomyocytes on the MECM 96-well plates, as described above in steps 1.9-1.11 (7.5 × 10⁵ cells/well).

4. Optical mapping using voltage-sensitive dyes (VSDs) and calcium-sensitive fluorophores (CSFs)

1. Prepare the appropriate amount of VSD in HBSS with calcium and magnesium, by adding 1 µL of VSD dye per mL of HBSS and 10 µL of loading adjuvant per mL of HBSS.
   NOTE: Typically, a 96-well plate requires 10 mL of VSD solution.
2. Alternatively, prepare HBSS with calcium and magnesium supplemented with 5 µM of CSF. Aspirate the cardiomyocyte maintenance medium and replace with 100 µL of VSD or CSF per well of a 96-well plate. Incubate the cells for 30 min in the cell culture incubator.
3. Remove the dyes and replace with assay medium or HBSS. Equilibrate at 37 °C for the acquisition of baseline data optical mapping with a high-throughput optical mapping device.
4. Treat the cells with drugs for acute exposure testing, or map the cells that were chronically exposed to drugs of interest.
5. For cardiotoxicity testing in 96-well plates, use four doses of a compound with at least six wells per dose. Use doses ranging from below to above the effective therapeutic plasma concentration, including a dose of the clinical effective therapeutic plasma concentration.
6. Dilute the drugs in dimethyl sulfoxide, store them as stock solutions at -20 °C, and then dilute them in HBSS to the desired concentrations.
7. Make baseline electrophysiology measurements prior to drug application, as described in section 5. Once the drugs have been applied, make electrophysiology recordings at least 30 min later for chronic studies. See the following sections for optical mapping data acquisition and analysis procedures.

5. Optical mapping using genetically encoded calcium indicator (GECI)

1. Plate commercially available or MACS-purified hiPSC-CMs, as described above, using MECM-coated 96-well plates to make mature hiPSC-CMs. To form confluent monolayers, plate 7.5 × 10⁴ CMs per well of each 96-well plate. Use plating medium.
2. After 48 h in plating medium, switch to cardiomyocyte maintenance medium.
3. On day 4 after thawing and replating, add recombinant adenovirus for expressing GCaMP6m (AdGCaMP6m) to cells at a multiplicity of infection (MOI) = 5. Add the virus using CM assay medium.
   NOTE: Here, experiments using GCaMP6m were performed in iCell² cardiomyocytes from a commercial vendor.
4. On day 5, remove the adGCaMP6m medium and replace with fresh RPMI+B27 (cardiomyocyte maintenance medium).
5. On day 7, observe the CMs using microscopy or the optical mapping imager, to visualize spontaneous contractions and corresponding calcium transients.
6. On day 7 or later, for medication screening, directly transfer 96-well plates of mature hiPSC-CM monolayers expressing GCaMP6m to the optical mapping imager from the incubator for baseline data acquisition.
7. Following electrophysiology data acquisition, return the plates of the CMs to the tissue culture incubator, for measurements on a subsequent time point.
8. Following baseline recordings, apply the medications, using at least four doses of each medication and at least six wells per dose. Equilibrate the medications on the cells for at least 30 min prior to data collection. Warm the well temperature to ~37 °C prior to and during the acquisition of data.
9. Following baseline recordings of an entire plate, add isoproterenol (500 nM) to every well to enable robust drug response data. Quantify the effects of isoproterenol on monolayer beat rate, contraction amplitude (calcium transient amplitude), and calcium transient duration (see Figure 6), as described in section 5.

6. Acquisition of optical mapping data and analysis

1. Make sure that the optical mapping device camera, transilluminator, and plate heater are turned on.
2. Open the acquisition software and determine the file saving location.
3. Open the front drawer and position the plate on the plate heater.
4. Acquire a dark frame by clicking on the Dark Frame button.
5. Select Duration (10-30 s) and Frame Rate of Acquisition (e.g., 100 fps; 250 fps for higher temporal resolution), and click Start Acquisition.
6. Open the analysis software and, in the Import/Filter tab, select either Browse for a single file or Tile Multiple to reconstruct a plate.
7. Select Parameter Mode (APD or CaTD), enter Distance per Pixel, and use the Well Wizard to determine the wells' location in the image. Click Process Save to move to the next tab.
8. Open the ROIs (regions of interest) tab and choose to draw the ROIs Manually, Automatically, or Not Use ROIs at all, which would then consider the whole well for analysis. After the ROIs have been selected, click the Process/Save button to move to the next step. Check the Hide Wells and Only Show Filtered boxes to visualize the ROIs.
9. Open the Analysis tab and, on the top right side of the screen, select each Well or ROI to confirm the Accuracy of Automatic Beat Detection. Add or Remove Beats from the traces by pressing Add Beat/ Save Beats, or by selecting an Individual Beat and pressing Delete on the keyboard.
10. Optionally, use the Average Heat Maps feature to create heatmaps of selected parameters for the plate.
11. Click on the Time-Space Plot button in the Analysis tab to visualize data for each well or the ROI. After making sure that the beat detection is accurate, proceed to the Export tab and select File Format. Press Export and create a folder for the data to be exported to. Proceed to open data files and run the chosen statistical analysis routine.
    NOTE: .xlxs files are preferred as all the parameters are exported in a single file; other formats (.csv or .tsv) generate one file per parameter.

Representative Results

hiPSC-CM maturation characterized by phase contrast and immunofluorescent confocal imaging
The timeline for ECM-mediated maturation of commercially available hiPSC-CMs using MECM coated 96-well plates is presented in Figure 1A. These data are collected using commercially available cardiomyocytes that arrive in the laboratory as cryopreserved vials of cells. Each vial contains >5 × 10⁶ viable cardiomyocytes. The cells are ~98% pure and rigorously tested for quality control (certificate of analysis is provided with each vial). The high number of CMs enables thawing and plating the CMs onto different ECM combinations using the same batch of cells. In Figure 1, hiPSC-CMs are plated on either mouse ECM- or MECM-coated plates. The hiPSC-CMs plated on the MECM mature and become structurally distinct from the same batch of hiPSC-CMs replated on the mouse ECM. Namely, mature cells become rod-shaped, while immature cells retain a circular shape. This can be seen in phase contrast imaging and upon staining the cardiac myofilaments (Figure 1B; troponin I [TnI], red). A more extensive validation of the structural maturation of hiPSC-CMs is presented in Figure 2. A large field of view (20x objective) of CMs stained with α-actinin antibody shows the typical shape of the cells cultured on each ECM condition. α-actinin is a critical structural protein arranged with regular spacing in the cardiac myofilaments. Consistent with the TnI staining in Figure 1, the α-actinin staining further indicates the maturation of hiPSC-CMs cultured on the MECM. Besides promoting a rod-shaped mature phenotype, the MECM also induces greater sarcomere organization (60x images). Mitochondrial content and activity are also distinct between cells cultured on the mouse ECM and the MECM (Figure 2B). Fetal-immature hiPSC-CM mitochondrial content is limited to the perinuclear space, with little mitochondria being found in the cytosol. In contrast, mature hiPSC-CMs mitochondrial content is distributed throughout the cell. Mitochondrial assessment uses an established protocol¹⁹.

hiPSC cardiac-directed differentiation and cardiac chamber specification
Provided here is a protocol for the in-house production and maturation of purified, chamber-specific hiPSC-CMs (Figure 3A). This is based on a previously published report²⁰. Presented are detailed procedures for hiPSC-CM purification using a commercially available, magnet-activated cell sorting (MACS) kit. We recently validated the use of MACS purification and showed the benefits of using MACS compared to metabolic-based hiPSC-CM purification; typically, hiPSC-CM purity above 95% is anticipated²¹. It is important to point out that if the initial CM content is <50%, MACS purification may reach only ~85%. In these cases, CM enrichment may be necessary following the depletion of non-CMs. If the initial CM content from differentiation is >50%, depletion of non-CMs from the cell population using the MACS kit can achieve purity >95%; in this case, the further enrichment or positive selection of CMs is not necessary. The chamber-specific hiPSC-CMs can also be matured using MECM-coated 96-well plates, as outlined above and shown in Figure 1 and Figure 2. It should be expected that the atrial-specific cells (hiPSC-ACM) have a significantly faster spontaneous beat rate and shorter action potential duration 80 (APD80) than the ventricular-specific cells (hiPSC-VCM). These are typical electrophysiological data for action potentials recorded using VSDs and the optical mapping system (Figure 3B-D).

High-throughput cardiac electrophysiological optical mapping
Scientific rigor is dramatically increased for any assay if it can be carried out in a high-throughput way. Cardiotoxicity screening data are presented in Figure 4, Figure 5, Figure 6, and Figure 7, showing high-throughput electrophysiological screening using mature hiPSC-CM monolayers in a 96-well plate. Whole plate heatmaps for parameters such as APD80 (Figure 4A) reveal the reproducibility of a given parameter within a plate from well to well. Further, whole-plate heatmaps provide a quick examination of any outliers in the data set. For example, in well E4 of the plate presented in Figure 4A, it is clear that this well has a much greater APD80 value, indicated by the well appearing yellow, while the other wells are indigo-blue. Typical action potentials of mature 2D hiPSC-CM monolayers (Figure 4B) are reminiscent of the action potential morphology of adult cardiomyocytes isolated and tested in culture. Moreover, a typical action potential spontaneous rhythm is displayed in Figure 4C. The data in Figure 4C is a time-space plot of row A, columns 1-12. The white line across the plate map in Figure 4A depict this. Each bright fluorescent flash over time in each well represents a single spontaneous activation. Figure 5 and Figure 6 show the utility of using GCaMP6m genetically encoded calcium indicator (GECI) to measure intracellular calcium transients; Figure 6 also shows the expected response to isoproterenol-the classical cardiac positive inotrope. In response to isoproterenol, activation of the β1-adrenergic receptors cause positive chronotropy (Figure 6A), positive inotropy (Figure 6B), and positive lusitropy (Figure 6C). These responses to isoproterenol indicate the significant maturation of hiPSC-CM β1-adrenergic receptors and intracellular signaling cascades.

In Figure 7, hiPSC-CM response to human Ether-a-go-go-related gene (hERG) channel blockers is shown, using GCaMP6m calcium fluorescence to monitor rhythm and to serve as a surrogate marker for contractility. E-4031 is a hERG-specific channel blocker, that slows the spontaneous beat rate and increases the calcium transient duration (CaTD80) and triangulation (CaT triangulation). Figure 7A shows the detection of early after-depolarizations caused by the E-4031 hERG channel blockade. Other hERG channel blockers, including domperidone, vandetanib, and sotalol, were also tested, and results are shown in Figure 7E-G. These compounds and doses were selected based on the recent hiPSC-CM validation study^1,7,9.

Figure 1: Timeline for fast maturation of commercially available or other-source cryopreserved hiPSC-CMs. (A) Thawed cardiomyocytes suspended in plating medium are applied to the MECM on day 0. On day 2, the medium is replaced with maintenance medium, and the spent medium is changed on day 5. Cells are cultured for an additional 2 days, and on day 7, the mature syncytia of hiPSC-CMs may be loaded with recording solution for downstream applications or cultured for longer periods. (B) Contrast phase of syncytia of cardiomyocytes plated on the mouse ECM or the MECM show that cardiomyocytes plated on the mouse ECM have greater circularity in comparison to cardiomyocytes plated on the MECM; furthermore, immunostaining for TnI indicates that cardiomyocytes plated on the mouse ECM retain a radial symmetry morphology and disorganized sarcomeres in contrast to dolichomorphic and well-structured hiPSC-CMs plated on the MECM. Scale bars = 100 µm (B, upper); 50 µm (B, lower). Abbreviations: hiPSC-CMs = human induced pluripotent stem cell-derived cardiomyocytes; ECM = extracellular matrix; MECM = maturation-inducing ECM; 96wp = 96-well plate; DAPI = 4',6-diamidino-2-phenylindole; TnI = troponin I. Please click here to view a larger version of this figure.

Figure 2: Comparison of sarcomere organization of hiPSC-CMs plated on a mouse ECM or an MECM. (A) Mouse ECM-cultured hiPSC-CMs immunostained against α-actinin indicate radial morphology, with a lower density of sarcomeres dispersed through the cardiomyocyte in contrast to hiPSC-CMs from the same batch plated on the matrix and presenting rod-shaped morphology (20x). (60x) Observation of hiPSC-CMs with confocal microscopy shows that hiPSC-CMs cultured on a mouse ECM have radial symmetry morphology, with a denser perimetral distribution of sarcomeres and a low density of radial sarcomeres in contrast to hiPSC-CMs from the same batch that were cultured on the MECM. They present a homogeneous distribution of sarcomeres organized along the longer axis of the cells. Scale bars = 100 µm (upper); 50 µm (lower). (B) Staining of hiPSC-CMs cultured on the mouse ECM or the MECM with a mitochondrial dye that stains mitochondria with high transmembrane potential show a lower intensity of staining in cardiomyocytes cultured on the mouse ECM in comparison to the MECM. Furthermore, hiPSC-CMs cultured on the MECM have mitochondria homogenously distributed in the cardiomyocytes, in contrast to hiPSC-CMs cultured on the mouse ECM that present a perinuclear accumulation of mitochondria. Scale bars = 200 µm. Abbreviations: hiPSC-CMs = human induced pluripotent stem cell-derived cardiomyocytes; ECM = extracellular matrix; MECM = maturation-inducing ECM; DAPI = 4',6-diamidino-2-phenylindole. Please click here to view a larger version of this figure.

Figure 3: Production of chamber specific cardiomyocytes. (A) Protocols for production of chamber-specific cardiomyocytes share identical Wnt signalizing pathway manipulation, with thr stimulation of Wnt signaling by inhibition of GSK3 from day 0 to 2, and the inhibition of this pathway between day 2 and 4. Chamber specification is achieved with activation of the retinoic acid pathway and Wnt signaling manipulation between days 3 and 6. (B) As a result of chamber specification, atrial cardiomyocytes present a faster rate of spontaneous depolarization in comparison to ventricular cardiomyocytes. (C) Ventricular hiPSC-CMs have slower beat rates in comparison to atrial hiPSC-CMs; therefore, the action potential duration at 80% of repolarization is shorter in hiPSC-ACMs in comparison to hiPSC-VCMs. Abbreviations: hiPSC-CMs = human induced pluripotent stem cell-derived cardiomyocytes; hiPSC-ACM = atrial human induced pluripotent stem cell-derived cardiomyocytes; hiPSC-VCM = ventricular human induced pluripotent stem cell-derived cardiomyocytes. Please click here to view a larger version of this figure.

Figure 4: Optical mapping acquired with an optical mapping device and analyzed with Pulse. (A) Example of a heatmap for the holistic observation of parameters assessed in a 96-well plate after movie filtration and determination of regions of interest in 96-well plates, mapped with the optical mapping device. In this example, a APD80% heatmap that indicates an outlier well (E4) and wells that failed to produce data (wells H3, 4, and 5). (B) Furthermore, the user-friendly interface allows easy plotting of average action potential morphology from the selected wells. (C) Additional data visualization tools are available; in this example, a time-space plot generated from the horizontal line crossing the wells on row A (panel A) shows activation across a horizontal section of each well (white line across row A) over a period of 10 s. Please click here to view a larger version of this figure.

Figure 5: Timeline for mapping of intracellular calcium transient changes with a genetically encoded calcium indicator. On day 4 after plating commercially available cardiomyocytes in an MECM-coated 96-well plate, as indicated in Figure 1A, the cells should be transduced with 5 MOI of virus in CM assay medium overnight. The medium is replaced with CDI maintenance medium until day 6, and changed to CDI maintenance medium without phenol red between days 7 and 11 to allow for prompt or continuous monitoring of intracellular calcium transient changes with Nautilus. (B) hiPSC-CMstransduced with AdGCaMP6f assessed with optical mapping on days 7, 9, and 10 post-thaw indicate the presence of stable, intracellular calcium-mediated fluorescence changes, that allow for daily optical mapping of the same plate over extended period of times without the need for reapplication of calcium-sensitive dyes. Abbreviations: GECI = genetically encoded calcium indicator; CM = cardiomyocyte; MOI = multiplicity of infection; 96wp = 96-well plate; BSA = bovine serum albumin. Please click here to view a larger version of this figure.

Figure 6: Quick and easy utilization of heatmaps for visual comparison of data acquired from mature functional syncytia of hiPSC-CMs with Nautilus and analyzed with Pulse. (A) hiPSC-CMs treated with isoproterenol show an increase in beat rate, as observed by the inspection of heatmaps and confirmed with paired t-test. (B) Similarly, mapping of cells before and after isoproterenol treatment shows the inotropic effect of β-adrenergic stimulation by visual comparison of heatmaps and with paired t-test. (C) Lastly, utilization of heatmaps for visual comparison of data show lusitropy, another canonical effect of β-adrenergic stimulation, confirmed with paired t-test (p < 0.0001). Absence of a circle indicates failure in data acquisition/analysis for that specific well. Abbreviations: hiPSC-CMs = human induced pluripotent stem cell-derived cardiomyocytes; ISO = isoproterenol. Please click here to view a larger version of this figure.

Figure 7: Validation of GECI cardiotoxicity screening assay using hERG channel blockers. (A) Representative spontaneous calcium flux traces from baseline wells in HBSS and in the presence of 500 nM E-4031. (B-D) Quantification of baseline and +E-4031 effects on beat rate, calcium transient duration 80 (CaTD80), and calcium transient triangulation (CaT triangulation) respectively. *,** denotes significant difference; unpaired t-test; p < 0.01; n = 8 in each group. (E) GECI detection of another hERG blocker, domperidone. (F) GECI detection of hERG block induced by vandetanib. (G) GECI detection of hERG block by high dose of sotalol. Abbreviations: GECI = genetically encoded calcium indicator; hERG = human Ether-a-go-go-related gene. Please click here to view a larger version of this figure.

Table 1: Media and their compositions Please click here to download this Table.

Discussion

There are several different approaches to in vitro cardiotoxicity screening using hiPSC-CMs. A recent "Best Practices" paper on the use of hiPSC-CMs presented the various in vitro assays, their primary readouts, and importantly, each assay's granularity to quantify human cardiac electrophysiological function²⁰. In addition to using membrane-piercing electrodes, the most direct measure of human cardiac electrophysiological function is provided by VSDs. VSD assay readouts enable direct visualization and quantification of critical electrophysiological parameters, including action potential duration, action potential propagation velocity, action potential upstroke, action potential triangulation, beat rate, beat regularity, and action potential duration heterogeneities. Similarly, calcium-sensitive probes offer information on hiPSC-CM monolayer rhythm, rate, and event durations. hiPSC-CM calcium transient measurements made using fluorescent probes also provide information on the contractility and contractile strength of each contraction. This paper provides methods for the use of mature hiPSC-CMs (96-well plates) in high-throughput VSD and calcium transient measurement assays. In addition to methods for optical mapping, software for high-throughput EP data analysis is presented.

The methods outlined here are a significant advance for the cardiotoxicity screening and regulatory science fields. Here, we have presented methods for the rapid maturation and electrophysiological recording of 2D hiPSC-CM monolayers in high-throughput screening plates (96-well plates). Rapid maturation using an MECM (7 days) shown here is a major advance over previous approaches requiring maturation to occur over 30-100 days^22,23. Compared to other ECMs, which require manual application for each experiment, MECM plates are precoated with ECM and arrive in the laboratory ready to use. This aspect of the MECM makes it easier to use, less variable, and more efficient than using other ECM coatings. Importantly, this approach can be used for both cryopreserved, commercially available hiPSC-CMs and "homemade" hiPSC-CMs, which can be chamber-specific. Owing to the distinct, rod-shaped structure of mature hiPSC-CMs (Figure 1 and Figure 2), it is important to point out that a greater number of cells is required for confluent monolayer formation here, compared to protocols that utilize other ECMs. Notably, when using mouse ECM (cells have a continuously spreading pancake phenotype), we plate 50,000 hiPSC-CMs per well, but when using MECM, we plate 75,000 hiPSC-CMs per well. The number of CMs per well can also be reduced if the cells are to be used for single-cell analysis, such as patch clamp or any other imaging technique requiring single cells.

Commercially available hiPSC-CMs offer advantages for regulatory science due to the extensive characterization of these cells by Food and Drug Administration (FDA)-led international efforts. However, the commercially available cells are a mixture of nodal, atrial, and ventricular hiPSC-CMs, which provide highly relevant toxicity information but lack the chamber-specific features that mimic the human heart. Chamber-specific hiPSC-CMs recreate the well-known electrophysiological differences between atrial and ventricular cardiomyocytes, and provide an in vitro assay for the development of chamber-specific anti-arrhythmic therapies (Figure 3B-D). For example, atrial fibrillation-specific medications can now be tested and developed using hiPSC-ACM monolayers plated in 96-well plates for robust and rigorous data collection. Likewise, when screening to determine if a compound causes Torsades de Pointes (TdP), a high-risk ventricular arrhythmia-, it is optimal not to have atrial and nodal cells "contaminating" ventricular cardiac monolayers. Therefore, although the FDA-led hiPSC-CM validation efforts to date have focused on using commercially available CMs, it is likely that future regulatory science recommendations will turn to the use of chamber-specific cells to make the toxicity screening even more predictive of the human cardiac condition. The methods outlined here are based on previous reports, and provide a robust approach for the generation of chamber-specific cardiomyocytes derived from pluripotent stem cells²¹.

A major difference between these protocols and those typically used in the field is the purification approach used. The majority of laboratories that generate hiPSC-CMs in their own labs rely on metabolic-mediated selection of cardiomyocytes²². Here, we rely on using MACS to purify hiPSC-CMs, using a clinically approved cell processing approach that produces CMs with healthier phenotypes²³. The metabolic challenge approach is effective, but uses a media formulation that simulates myocardial ischemia²⁴. When utilizing MACS purification of hiPSC-CMs, it is important to utilize the non-CM depletion cocktail, which targets non-CMs for magnetic depletion from the cell population. Use of the non-CM depletion approach minimizes the shear stress that the CMs experience in the magnetic column, and is preferred over direct magnetic labeling of the CM population. Using MACS purification of chamber-specific cells will enable other laboratories to generate healthy CMs for research and toxicity testing.

Materials

Name	Company	Catalog Number	Comments
0.25% Trypsin EDTA	Gibco	25200-056
0.5 mg/mL BSA (7.5 µmol/L)	Millipore Sigma	A3294
2.9788 g/500 mL HEPES (25 mmol/L)	Millipore Sigma	H4034
AdGCaMP6m	Vector biolabs	1909
Albumin human	Sigma	A9731-1G
alpha actinin antibody	ThermoFisher	MA1-22863
B27	Gibco	17504-044
Blebbistatin	Sigma	B0560
CalBryte 520AM	AAT Bioquest	20650
CELLvo MatrixPlus 96wp	StemBiosys	N/A	https://www.stembiosys.com/products/cellvo-matrix-plus
CHIR99021	LC Laboratories	c-6556
Clear Assay medium (fluorobrite)	ThermoFisher	A1896701	For adenovirus transduction
DAPI	ThermoFisher	62248
DMEM:F12	Gibco	11330-032
FBS (Fetal Bovine Serum)	Sigma	F4135-500ML
FluoVolt	ThermoFisher	F10488
HBSS	Gibco	14025-092
iCell CM maintenance media	FUJIFILM/Cellular Dynamics	M1003
iCell2 CMs	FUJIFILM	1434
Incucyte Zoom	Sartorius
iPS DF19-9-11T.H	WiCell
Isoproterenol	MilliporeSigma	CAS-51-30-9
IWP4	Tocris	5214
L-ascorbic acid 2-phosphate sesquimagnesium salt hydrate	Sigma	A8960-5g
L-glutamine	Gibco	A2916801
LS columns	Miltenyii Biotec	130-042-401
MACS Buffer (autoMACS Running Buffer)	Miltenyii Biotec	130-091-221
Matrigel	Corning	354234
MitoTracker Red	ThermoFisher	M7512
Nautilus HTS Optical Mapping	CuriBio		https://www.curibio.com/products-overview
Nikon A1R Confocal Microscope	Nikon
nonessential amino acids	Gibco	11140-050
pre-separation filter	Miltenyii Biotec	130-041-407
PSC-Derived Cardiomyocyte Isolation Kit, human	Miltenyii Biotec	130-110-188
Pulse	CuriBio		https://www.curibio.com/products-overview
Quadro MACS separator (Magnet)	Miltenyii Biotec	130-091-051
Retinoic acid	Sigma	R2625
RPMI 1640	Gibco	11875-093
RPMI 1640 (+HEPES, +L-Glutamine)	Gibco	22400-089
StemMACS iPS-Brew XF	Miltenyii Biotec	130-107-086
TnI antibody (pan TnI)	Millipore Sigma	MAB1691
Versene (ethylenediaminetetraacetic acid - EDTA solution)	Gibco	15040-066
Y-27632 dihydrochloride	Tocris	1254
β-mercaptoethanol	Gibco	21985023