Drug screening platform using human induced pluripotent stem cell‐derived atrial cardiomyocytes and optical mapping

Abstract Current drug development efforts for the treatment of atrial fibrillation are hampered by the fact that many preclinical models have been unsuccessful in reproducing human cardiac physiology and its response to medications. In this study, we demonstrated an approach using human induced pluripotent stem cell‐derived atrial and ventricular cardiomyocytes (hiPSC‐aCMs and hiPSC‐vCMs, respectively) coupled with a sophisticated optical mapping system for drug screening of atrial‐selective compounds in vitro. We optimized differentiation of hiPSC‐aCMs by modulating the WNT and retinoid signaling pathways. Characterization of the transcriptome and proteome revealed that retinoic acid pushes the differentiation process into the atrial lineage and generated hiPSC‐aCMs. Functional characterization using optical mapping showed that hiPSC‐aCMs have shorter action potential durations and faster Ca2+ handling dynamics compared with hiPSC‐vCMs. Furthermore, pharmacological investigation of hiPSC‐aCMs captured atrial‐selective effects by displaying greater sensitivity to atrial‐selective compounds 4‐aminopyridine, AVE0118, UCL1684, and vernakalant when compared with hiPSC‐vCMs. These results established that a model system incorporating hiPSC‐aCMs combined with optical mapping is well‐suited for preclinical drug screening of novel and targeted atrial selective compounds.

unique opportunity for clinical translation. 1 Furthermore, the ability to differentiate chamber-specific cardiomyocytes allows for a more precise study of cardiac disease physiology and pharmacology.
The cardiomyocytes of the lower (ventricles) and upper (atria) chambers have distinct characteristics that arise from differential developmental pathways. Previous work in vivo has shown that the expression patterns of retinoic acid and retinaldehyde dehydrogenase 2 (RALDH2) are important determinants of the atrial fate. [2][3][4][5] These results were later recapitulated in a pivotal study by Lee and Protze et al 6 who determined that atrial cardiomyocytes (aCMs) differentiated from human embryonic stem cells (hESCs) originate from a unique mesoderm characterized by robust RALDH2 expression. This study established an atrial differentiation protocol that included the addition of retinoic acid. Retinoic acid has also been utilized to selectively differentiate hESCs and hiPSCs into aCMs in other studies. [6][7][8][9][10] The distinct properties of the atrial and ventricular cardiomyocytes are determined by the differential expression of unique sets of ion channels and other proteins that optimize their specific function. Drugs that target atrial ion channels selectively can therefore produce differences in pharmacological function in the two chambers. This atrial-selective pharmacology is of utmost interest in the study and treatment of atrial-specific diseases such atrial fibrillation (AF), which is the most common heart rhythm disorder. Investigating atrial-selective pharmacology can assist and guide novel cardiac drug development as well as improving both safety and efficacy by avoiding potential toxic electrophysiologic effects on the ventricular chambers.
The differential pharmacology of stem cell-derived aCMs was studied previously by Laksman et al 7 who showed that flecainide can rescue the AF phenotype in a dish. Other studies have also studied the selective pharmacological effects of agents on hiPSC-derived aCMs but have largely focused on proof-of-concepts using limited number of test compounds and standard measurement systems that are low in throughput. 9,10 With a focus on translation, a preclinical model platform that characterizes pharmacological activity must capture the main cardiac functional signatures that most closely mimic and predict human cardiac physiology and drug responses. As such, we established in this study an in vitro assay platform by combining hiPSC-derived atrial cardiomyocytes (hiPSC-aCMs) and high-content optical mapping, a noninvasive all-optical system that simultaneously measures membrane potential (V m ) and Ca 2+ transients at a high-resolution in a monolayer tissue format.
We first demonstrate a selective hiPSC-aCM differentiation protocol by modifying the well characterized GiWi protocol 11

| MATERIALS AND METHODS
A detailed methods section is available in the Supplemental Information.

| Maintenance and expansion of hiPSCs
hiPSCs (WiCell, IMR90-1) were maintained and expanded in mTeSR1 medium and feeder-free culture using 6-well plates coated with Matrigel. Using Versene (EDTA), hiPSCs were passaged every 4 days or 85% confluency at 1:15 ratio. Passaged hiPSCs were cultured with mTeSR1 supplemented with 10 μM Y27632 for the first 24 hours and the mTeSR1 was exchanged daily during cell culture maintenance.

| Directed differentiation of hiPSCs into atrial and ventricular subtypes
hiPSC-derived ventricular cardiomyocytes were differentiated by employing a modified GiWi protocol 11 that we previously published. 12 In brief, hiPSCs were seeded at a density of 87 500 cells/cm 2 . At day 0, differentiation was initiated using 12 μM CHIR99021. At day 3, the cells were incubated with 5 μM IWP-4.

| mRNA expression profiling
Gene expression profiling was conducted using multiplexed NanoString and real time quantitative PCR (qPCR). Pooled total RNA was used in both assays. The extracted RNA was reverse transcribed into cDNA which was used in the qPCR assay. Oligonucleotide sequences are described in Table S7. The multiplexed mRNA profiling was conducted using NanoString Technologies (Seattle, Washington) platform with a custom Codeset containing 250 gene probes. Analysis was performed on the Sprint instrument and nSolver analysis software with the Advanced Analysis module.

| Atrial natriuretic peptide measurement
The levels of atrial natriuretic peptide (ANP) of hiPSC-aCMs and -vCMs were measured by a competitive enzyme-linked immunosorbent assay (ELISA) using a commercially available kit (Invitrogen, California). The assay was conducted according to the manufacturer's protocol and was measured using a spectrophotometric plate reader. Enriched hiPSC-CMs were seeded on Matrigel-coated 24-well plates at a seeding density of 600 000 cells per well.

| Patch-clamp recordings
Single hiPSC-aCMs and hiPSC-vCMs were plated on gelatin (0.1%) and Geltrex (1:10) at 30 000 cells per well. After 48 hours in culture, glass electrodes were used to achieve the whole-cell configuration with single hiPSC-CMs and only cells with gigaohm seals were used for further analysis. The formulation for internal and external recordings solutions are outlined in the Supplemental Information.
Current recordings were performed using the Axon Instruments 700B amplifier and digitized at 20 kHz. All recordings were performed at 33-35 C as maintained. For pacing at 1 Hz, gradually increasing amounts of current were injected with a 1 ms pulse width until reliable action potentials (APs) were triggered. The maximal upstroke velocity was determined by calculating the maximum derivative and the resting membrane potential was measured during a 5 second epoch without spontaneous activity 1 minute after break-in. Further details on data analysis are found in the Supplemental Information.

| Optical mapping
Optical mapping recordings were performed on enriched monolayers of hiPSC-aCMs and -vCMs cultured in a 24-well plate format at day 45 to 60 postdifferentiation. Imaging experiments were conducted using Ca 2+ Tyrode's solution (formulation found in Supplemental Information). The hiPSC-CMs were loaded with RH-237, blebbistatin, and Rhod-2AM sequentially before imaging as described. 12,13 Both RH-237 and Rhod-2AM were excited by 530 nm LEDs. Images were acquired at a frame rate of 100 frames/second by a sCMOS camera (Orca Flash 4.0V2, Hamamatsu Photonics, Japan) equipped with an optical splitter. The cells were paced using programmable stimulation.
Data collection, image processing, and initial data analysis were accomplished using custom software. The multiwell optical mapping system was custom engineered in the lab based on a system as described previously. 12,13 Further details are found in the Supplemental Information.

| Pharmacological analyses
The drugs used in this study are listed Table S8. Drug stocks were further diluted in Ca 2+ Tyrode's solution prior to pharmacological testing with the final DMSO concentration in the experimental solution not exceeding 0.03% (v/v). Drug effects were studied in serum-free conditions (ie, Ca 2+ Tyrode's and drug only) at four doses by sequentially increasing the drug concentration in the same well with recordings at 20-minute intervals.

| Statistical analysis
Further details on data and statistical analysis can be found in the Supplemental Information. Unpaired t tests were conducted to compare two groups (ie, hiPSC-aCMs vs hiPSC-vCMs) in the analysis of qPCR, ELISA, patch clamp recordings, and optical mapping (baseline condition and normalized drug effects). Analysis of dose-dependent effects was performed using one-way ANOVA and Dunnett's post hoc test. All data are presented as mean ± SEM unless noted otherwise. Significance level for all statistical analysis was set at P < .05 with the following notation: *P < .05, **P < .01, ***P < .001.

| RA treatment drives cardiac differentiation into atrial phenotype
We first optimized the atrial differentiation protocol by altering the concentration and timing of retinoic acid (RA) based on the molecular signatures of atrial phenotype as measured by qPCR and flow cytometry ( Figures S2 and S3). Higher dose of RA reduced cardiac differentiation efficacy defined by the decrease in the cTnT + proportion of the total cell population as measured by flow cytometry ( Figure S2A).
The finalized protocol to generate hiPSC-aCMs included RA addition at 0.75 μM every 24 hours on days 4, 5, and 6 ( Figure 1A Compared with hiPSC-vCMs, hiPSC-aCMs were found to have no significant difference in pan cardiac phenotype. Expression of the pan cardiac transcript NKX 2.5 measured by qPCR was similar between hiPSC-aCMs and -vCMs ( Figure 1B), as was cardiac troponin T (cTnT) protein expression measured by flow cytometry (Figures 1C and S1).

| Gene expression analysis of hiPSC-aCMs
We performed an extensive gene expression analysis of hiPSC-aCMs and hiPSC-vCMs using NanoString technology in which each mRNA copy was digitally counted for accurate and sensitive detection of gene expression. 16 Five independent differentiation batches of each cardiac subtype were included in the analysis. The unsupervised hierarchical clustering analysis showed clear grouping of hiPSC-aCM samples that were segregated relative to hiPSC-vCMs ( Figure 2A).
The gene expression profile of the hiPSC-vCM samples were more variable with 2 samples closer in distance to the hiPSC-aCMs while three samples displayed clear segregation (Figure 2A). The overall difference in global gene expression and lineage between hiPSC-aCMs and -vCMs was also captured in the principal component analysis (PCA, Figure S4A). Out of the 250 transcripts analyzed, 200 genes were detected above background noise defined by a threshold of 50 raw digital counts as determined by the negative controls of the assay. In the hiPSC-aCMs, 14 and 27 genes were significantly upregulated and downregulated, respectively ( Figure 2C). As expected, hiPSC-aCMs displayed significantly higher expression profiles of atrial-specific markers including atrial-specific K + channel K v 1.5 (KCNA5) and transcription factors (NR2F2 and TBX18) ( Figure 2C). Meanwhile, hiPSC-vCMs displayed higher expression of ventricular-specific genes such as those encoding for contractile proteins MYL2, MYH7, and the L-type Ca 2+ channel isoform Ca v 1.2 (CACNA1C) ( Figure 2C). The genes encoding for the proteins in the sarcoplasmic reticulum complex such as TRDN, CASQ2, and RYR2 were expressed in significantly lower amounts in the hiPSC-aCMs samples ( Figure 2C). Meanwhile, pan-cardiac markers NKX2-5 and TNNT2 were expressed at similar levels in both hiPSC-aCMs and -vCMs, further corroborating the efficiency of the differentiation protocol ( Figure S4B).

| Functional phenotyping of hiPSC-derived atrial cardiomyocytes
We compared the electrophysiological characteristics of the differentiated hiPSC-aCMs and -vCMs using whole-cell patch clamp.
F I G U R E 2 Gene expression analysis of hiPSC-aCMs and -vCMs using NanoString. Global gene expression pattern of hiPSC-aCMs and -vCMs shown in A, heat map of the expression of the 250 genes across samples of hiPSC-aCMs and -vCMs. The cluster dendrogram shows the unsupervised hierarchical clustering that was conducted using the agglomerative algorithm and the Euclidian distance criterion. B, Differentially expressed genes between hiPSC-aCMs and -vCMs expressed in volcano plot shows 14 upregulated (red) and 27 downregulated (blue) genes in hiPSC-aCMs. Solid horizontal line represents the Benjamini-Hochberg false discovery rate (FDR) adjusted P-value <.05 (−log10 = 1.3). Dashed vertical lines represent the arbitrary log2 fold change cut-off of −0.5 and 0.5. C, Forty-two differentially expressed genes identified from the statistical criteria of FDR adjusted P-value <.05 and log2 fold change of <−0.5 and >0.5. Data are presented as mean ± SEM. n = 5 independent differentiation batches Rate-dependent properties are critical in cardiac function. A variable rate protocol ( Figure S6) in which the hiPSC-CMs were electrically paced with increasing frequency at every cycle was used to investigate the electrical restitution dynamics. The electrical restitution curve reflects the ability of the cardiac system to accommodate a higher pacing rate by progressive shortening of APD 80 and is described as APD 80 in relation to the diastolic interval (DI). Compared with hiPSC-vCMs, the electrical restitution curve of the hiPSC-aCMs displayed a flatter portion and did not show APD 80 shortening at longer diastolic intervals ( Figure 4F). The extensive shortening in APD 80 started at shorter diastolic intervals for hiPSC-aCMs (<275 ms) compared with hiPSC-vCMs (<500 ms). The maximum slope of the restitution curve was higher in hiPSC-vCMs compared with hiPSC-aCMs (1.26 ± 0.08 vs 0.91 ± 0.04, P < .05; Figure 4G) indicating faster kinetics of APD in response to higher pacing rate.

| In vitro screening for atrial-selective pharmacology
We first established the utility of optical mapping to detect a pancardiac pharmacological response by using dofetilide, a strong blocker F I G U R E 3 hiPSC-aCMs and -vCMs have distinct electrophysiological characteristics. Single differentiated hiPSC-aCMs and -vCMs were plated on gelatin and Geltrex after 30 days in culture. A, Whole cell current clamp recordings from a spontaneously beating hiPSC-vCM. B, Recorded action potential (APs) demonstrates typical prolonged plateau phase in both spontaneous (left) and/or paced at 1 Hz (right). C, Current clamp recording from a spontaneously beating hiPSC-aCM. D, Single AP from hiPSC-aCM demonstrates shortened action potential duration (APD) and lack of prolonged plateau phase, spontaneous (left), paced at 1 Hz (right). E, The first differential of voltage recordings from hiPSC-aCMs and -vCMs were used to calculate the maximal upstroke velocities. F, One minute after achieving the whole-cell configuration, the average resting membrane potential was measured. G, Spontaneously beating and 1 Hz paced APs were assessed for duration at 50% of peak (APD 50 ), and H, 90% of peak (APD 90 ). Statistics were performed by unpaired t test. *P < .05, ***P < .005. Data are presented as mean ± SEM. Two differentiation batches were included in this analysis of the rapid delayed rectifier K + current (I Kr ), 17 an ionic current expected to be present in both hiPSC-aCMs and -vCMs. 18 Dofetilide elicited a dose-dependent response in both hiPSC-aCMs and -vCMs.
Next, we demonstrated the functional differences in the ion channel profiles of hiPSC-aCMs and -vCMs. We aimed to show that the ultrarapid outward current (I Kur ) produced by the channel K v 1.5 (KCNA5) was functional and specific to hiPSC-aCMs, while the inward Ca 2+ current (I Ca,L ) produced the voltage-dependent L-type Ca 2+ channel Ca V 1.2 (CACNA1C) was functional and specific to hiPSC-vCMs.
We used two relatively selective compounds, 4-aminopyridine (4AP) and nifedipine, to dissect the presence of functional I Kur and I CaL , F I G U R E 4 Functional phenotyping of hiPSC-derived atrial and ventricular CMs using optical mapping. Representative average traces of A, action potential and B, Ca 2+ transients of hiPSC-aCMs and -vCMs electrically paced at 1 Hz. C, Electrical restitution curve measured at APD 80 relative to the diastolic interval (DI). D, Quantification of early-(APD 20 ), mid-(APD 50 ), and late-(APD 80 ) repolarization, unpaired t test, *P < .05, **P < .01. E, Quantification of early-(CaTD 20 ), mid-(CaTD 50 ), and late-(CaTD 80 ) Ca 2+ transient decay, unpaired t test, ***P < .001. F, Time to peak (TTP) of the Ca 2+ transient, unpaired t test, ***P < .001. G, Time constant (τ) of Ca 2+ decay, unpaired t test *P < .05. H, Maximum slope of the electrical restitution as shown in panel C, unpaired t test, *P < .05. Electrical restitution curves were measured under a variable rate pacing protocol (60-200 bpm) as described in the Supplemental Information. n = 4 (four independent differentiation batches) and cardiac enriched hiPSC-aCMs and -vCMs were analyzed in these set of experiments. Data are presented as mean ± SEM respectively. While nifedipine is also known to block Cav1.3, it is expected to have a preferential effect at lower concentrations on Ca V 1.2 based on the literature which indicates 13-fold higher block on Ca V 1.2 than Ca V 1.3. 19 At the highest tested dose (300 nM), nifedipine significantly decreased APD 50 of hiPSC-vCMs from 170 ± 14 to 121 ± 16 ms (28% ± 4% shortening) and decreased CaTD 50 from 357 ± 10 to 333 ± 23 ms (30% ± 3% shortening) ( Figure 5D; Table S2). We observed a trend in APD 50 shortening of hiPSC-aCMs in response to increasing the nifedipine dose, but the drug elicited a significantly stronger dose-dependent shortening in both APD and CaTD of hiPSC-vCMs compared with hiPSC-aCMs ( Figures S8 and S9).
Observing the percent change from predrug control, nifedipine induced differential response in overall APD and CaTD between hiPSC-aCMs and -vCMs at 10, 100, and 300 nM ( Figure 5D).
We then demonstrated the effectiveness of our drug screening platform in assessing the effects of experimental compounds designed to have targeted effects on atrial-specific ion channels using AVE0118 and UCL1684.
AVE0118 is an experimental drug that blocks I Kur , the G-proteinactivated K + current (I KAch ) , and the transient outward K + current (I to ) at a similar dose range. 20 Both I Kur and I KAch are atrial-specific ionic currents. AVE0118 prolonged mid-and late-repolarization (APD 50 and APD 80 ) of both hiPSC-aCMs and -vCMs at the two highest tested doses (3 and 10 μM; Table S4). Similarly, AVE0118 had significant effects on CaTD 50 and CaTD 80 of hiPSC-aCMs and -vCMs at all F I G U R E 6 The effects of 4-aminopyridine (4AP) and AVE0118 on action potential and Ca 2+ transient of hiPSC-aCMs and -vCMs. Representative traces of action potential and Ca 2+ transients illustrating the effects of A 4-aminopyridine (4AP) and B, AVE0118 on hiPSC-aCMs and -vCMs. Higher drug dose is presented by a progressively darker shade. The effects of C dofetilide and D, vernakalant on normalized (percent change from predrug baseline) action potential duration (APD), and B, Ca 2+ transient duration (CaTD); both parameters being measured at 20%, 50%, and 80%. Dashed line is the normalized predrug control presented as 0% change. n = 6 from six independent differentiation batches. hiPSCderived atrial cardiomyocytes (aCMs) are shown in red while hiPSC-derived ventricular cardiomyocytes (vCMs) are presented in blue. Data are presented as mean ± SEM. Drug effects were compared between hiPSC-aCMs and -vCMs at each dose using unpaired t test, *P < .05, **P < .001, ***P < .001. NS stands for not significant tested doses (Table S4). However, the APD 50 and APD 80 of hiPSC-aCMs were significantly prolonged at a lower dose of 1 μM (control: 200 ± 14 ms, 1 μM: 244 ± 16 ms; Table S4). Furthermore, the atrialselective effects of the drug were demonstrated by a larger proportional prolongation in APD 50 and APD 80 of hiPSC-aCMs compared with hiPSC-vCMs at 1, 3, and 10 μM (APD; Figure 6D). Furthermore, AVE0118 induced a larger proportional prolongation in CaTD of hiPSC-aCMs compared with hiPSC-vCMs at all tested doses ( Figure 6D). Early repolarization (APD 20 ) of hiPSC-aCMs also displayed a large dose-dependent response ( Figure S8) with a proportionally larger prolongation at 10 μM (63% ± 2% vs 43% ± 5%, P < .05; Figure 6D).
UCL1684 prolonged CaTD 80 of hiPSC-aCMs at all tested doses F I G U R E 7 The effects of UCL1684 and vernakalant on action potential and Ca 2+ transient of hiPSC-aCMs and -vCMs. Representative V m and Ca 2+ transients illustrating the effects of A, UCL1684 and B, vernakalant on hiPSC-aCMs and -vCMs. Higher drug doses are presented by a progressively darker shade. The effects of C, AVE0118 and D, UCL1684 on normalized (percent change from predrug baseline) action potential duration (APD) and Ca 2+ transient duration (CaTD); both parameters being measured at 20%, 50%, and 80%. Dashed line is the normalized predrug control presented as 0% change. n = 6 from six independent differentiation batches. hiPSC-derived atrial cardiomyocytes (aCMs) are shown in red while hiPSC-derived ventricular cardiomyocytes (vCMs) are presented in blue. Data are presented as mean ± SEM. Drug effects were compared between hiPSC-aCMs and -vCMs using unpaired t test at each dose, *P < .05, **P < .001, ***P < .001. NS stands for not significant (baseline: 300 ± 15 ms, at 0.3 μM: 372 ± 23 ms, at 1 μM: 387 ± 33 ms, at 3 μM: 413 ± 24 ms, at 10 μM: 416 ± 39 ms, P < .05; Table S5). In contrast, UCL1684 exposure showed no statistically significant effect on overall APD and CaTD of hiPSC-vCMs. The sensitivity of hiPSC-aCMs to UCL1684 was also reflected in the doseresponse relationship showing a prolongation APD 80 , in contrast to the minimal prolongation in APD 80 of hiPSC-vCMs ( Figure S8).
Finally, we tested the effects of vernakalant which is a multi-ion channel blocker that blocks the fast and late inward Na + current (I Na , I NaL , respectively), the I Kur , and the I KAch. 22 The drug is used clinically for intravenous cardioversion of patients in AF 23 and was expected to induce an atrial-specific effect due to its I Kur and I KAch blocking properties.

| DISCUSSION
In this study, we were successful in efficiently differentiating hiPSCs into a monolayer of cardiomyocytes with an atrial phenotype by modifying the GiWi protocol. 11 We used multiple phenotypic approaches such as qPCR, digital multiplexed gene expression analysis with Namely, the AP of hiPSC-aCMs was significantly shorter, along with a lack of a prolonged plateau phase as opposed to the AP of hiPSC-vCMs, an observation that is aligned with native cardiomyocyte electrophysiology. 28 Similarly, the CaT of hiPSC-aCMs had faster kinetics with a faster decay time as reflected by the differential expression of Ca 2+ channel isoforms, further demonstrating the differential physiology between hiPSC-aCMs and -vCMs.
In terms of APD measurements, we observed a good correlation between the patch clamp and optical mapping recordings for hiPSC-aCMs. In hiPSC-vCMs, however, the optical AP measurements were shorter overall than patch clamp recordings. This discrepancy may be attributed to the heterogeneity of our current ventricular differentiation protocol which generated predominantly ventricular cardiomyocytes but also contain a small proportion of nonventricular phenotypes (ie, atrial myocytes and nodal cells). Thus, the optical AP signals represents an average from about 300 000 cells in each 1 cm 2 region of interest.
Another hallmark of cardiomyocyte function is rate-dependence, as described by the electrical restitution curve. 29 We observed that the electrical restitution properties were different between hiPSC-aCMs and -vCMs. Compared with hiPSC-vCMs, hiPSC-aCMs The presence of specific ion channel currents (ie, I Kur , I KAch , and I CaL ) explain, in part, the functional differences between the two cardiac chamber subtypes, the expressions of which were already shown in our qPCR and NanoString assays. We used a series of compounds The compound 4AP has been shown to selectively block K v 1.4 (I to ) and K v 1.5 (I Kur ) 33 and is therefore expected to elicit a response in hiPSC-aCMs at lower doses than in hiPSC-vCMs as I Kur  that was predicted to demonstrate targeted effects in hiPSC-aCMs.
However, only a nuanced atrial specificity was observed in our assay.
Although the effects were proportionally larger in hiPSC-aCMs, 3.4 μM) and I Kr (IC 50 : 9.6 μM) 35 which prolongs APD at the tested doses of 3 and 10 μM as genes encoding the channels producing the I to (KCNA4) and I Kr (KCNH2) were expressed in our hiPSC-vCMs. The drug was also shown to be effective in terminating certain ventricular arrhythmias 36 which was predicted based on our results of prolongation in the APD of hiPSC-vCMs.
Next, we used UCL1684, a highly specific SK channel pore blocker, to assess the presence of functional SK channels in hiPSC-aCMs. The SK channel has three paralogs but the SK3 channel variant (KCNN3) has been shown to be atrial-specific and has been implicated in AF pathogenesis in several studies. 37,38 In this study, UCL1684 displayed high specificity toward hiPSC-aCMs with a strong dosedependent response. The drug confirmed the presence of functional SK channels in hiPSC-aCMs at 3 μM with a positive dose-dependent response while having no effect on hiPSC-vCMs at all tested doses (0.3, 1, 3, and 10 μM).
Vernakalant is touted as an atrial-selective compound clinically approved for intravenous cardioversion of AF. 39  This study has several limitations. One limitation in our findings is that we cannot directly compare the results from qPCR and NanoString as both assays have fundamental differences in technical principles and statistical methodologies. Taken together, however, both assays show the global changes in cell type specific gene markers and further validate the role of retinoic acid in directing the cardiac differentiation process toward an atrial lineage. The main limitation in this field is the maturation state of the hiPSC-CMs as they have an overall immature phenotype with some crucial differences compared with adult cardiomyocytes. 40 Nonetheless, we were able to observe the stark differences in genetic, protein, as well as functional signatures of AP and CaT in the two generated chamber-specific cell types.
Additionally, maturation stage does not explain the differences in chamber-specific phenotype as parallel batch differentiation and timein-culture were incorporated in our study design. Most importantly, we were able to capture effects of drugs that were expected to have atrial-specific properties in hiPSC-aCMs.

| CONCLUSION
The ability to differentiate hiPSC-aCMs provides a unique opportunity to study atrial physiology and its pharmacologic responses in a human-relevant in vitro model. We demonstrated an hiPSC-based in vitro model that recapitulates the molecular and functional characteristics of the phenotype of native atrial tissue. Our platform adds to the repertoire of cardiac drug screening and can be readily applied in future efforts of atrial-specific drug discovery.

CONFLICT OF INTEREST
The authors declared no potential conflicts of interest.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available on request from the corresponding author.