The generation of functional pacemaker cells from PSC has been described previously (20, 35). Briefly, programming murine ESC with TBX3 combined with antibiotic selection using the αMhc promoter led to small aggregates that we termed “induced sinoatrial bodies” (iSABs).

Double-transfected PS cells (TBX3 and αMhc promoter) were cultured at 37°C and 5% CO₂ for at least 7 days before differentiation was initiated. The culture medium consisted of Dulbecco's Modified Eagle's Medium—high glucose with stable L-Glutamine (4 mM) (GIBCO, Life Technologies, Carlsbad, USA), 10% FCS superior (Biochrome AG, Berlin, Germany), 1% Penicillin-Streptomycin (GIBCO, Life Technologies, Carlsbad, USA), 100 µM MEM Non-essential amino acids (GIBCO, Life Technologies, Carlsbad, USA), 1,000 U/ml Leukemia inhibitor factor (diluted in ddH₂O with 0.1% BSA) (Phoenix Europe, Mannheim, Germany) and 100 µM ß-mercaptoethanol (Sigma, St.Louis, USA). Subsequent selection of double-transfected cells was performed by adding 10 µg/ml blasticidin (InvivoGen, San Diego, USA) and 250 µg/ml hygromycin (InvivoGen, San Diego, USA).

To obtain iSABs, the first step was to induce differentiation using Embryoid Bodies (EBs). Consequently, Hanging Drops (HD) were generated, each containing approximately 400 cells. After 2 or 3 days in HD, the EBs were transferred to 10 cm petri dishes and maintained as a suspension culture for a further 3 days at 37°C and 5% CO₂. On day 7 of differentiation, the EBs were plated on 0.1% gelatin-coated petri dishes with medium changes every 2–3 days. The EBs began to adhere, grow, and contract spontaneously. On day 15 of differentiation, the antibiotic-based selection of αMhc-expressing cells was started using medium containing 400 µg/ml of G418 (Biochrome, AG, Berlin, Germany) and continued until treatment with collagenase IV (6,000 U/ml, GIBCO, Life Technologies, Carlsbad, USA). The latter was performed on day 20 of differentiation by incubating the cell layer with the enzyme solution for 8 min at 37°C. Mechanical agitation during pipetting separated the cell clusters from the cell layer. After centrifugation, the obtained iSABs were resuspended in a differentiation medium without ascorbic acid and then transferred to new petri dishes. The iSABs were subsequently separated from dead cells by filtration using a 40 µm cell strainer (EASystrainer, Greiner-BioOne, Frickenhausen, Germany).

The differentiation medium consisted of IMDM medium (PAN-Biotech GmbH, Aidenbach, Germany/Biochrom), 10% FCS superior (Biochrome AG, Berlin, Germany), 1% Penicillin-Streptomycin, 100 µM MEM Non-essential amino acids, 450 µM 1-thioglycerol and 1.21 µM ascorbic acid.

In order to reduce the time and space required for differentiation and to increase the yield of differentiated cells, we introduced a new method for EB formation into our existing protocol and analyzed the comparability of the generated pacemaker cells. Therefore, we compared the standard HD method with the spherical multiwell plate (Sphericalplate 5D, Kugelmeiers, Switzerland) (SP). The latter cell culture plate contains 12 out of 24 wells with 750 conical cavities each. Initially, the wells underwent incubation with pre-warmed PBS to minimize air bubbles when adding the cell suspension. For the sphericalplate, a cell suspension of 400 cells * 750 microwells was formulated for each individual well.

The EB diameter was measured after different generation times using the two mentioned EB-generating methods. Four independent experiments were conducted with 10 EBs measured per condition. We analyzed the diameter of the EBs after one, two, or three days of generation. After the appropriate time, the EBs were harvested and collected by gently washing them through the wells (SP) or plates (HD). The diameter of the EBs was then measured bidirectionally using a Zeiss Elyra PS1 microscope.

Moreover, the average number of beating foci per EB was analyzed by calculating data from 10 adherent EBs. Briefly, after formation using HD or SP, 10 EBs were either plated directly onto gelatin-coated wells (24 well plates) or cultured as “swimming” non-adherent EBs in uncoated petri dishes for an additional 3–4 days and then plated. At days 16/17 of differentiation, the number of beating areas per EB was microscopically detected and measured using the Zeiss Elyra PS1.

Following various induction conditions, cells underwent differentiation, pacemaker cells were selected, collagease treatment and filtration were performed. Afterwards, gained iSABs clusters were moved to 6 cm petri dishes using differentiation medium without ascorbic acid. The beating frequency was evaluated 3–6 days after collagenase treatment utilizing an ELYRA PS.1 LSM 780 microscope. Video recordings of the beating areas were captured using Zen Black software (Zeiss, Jena, Germany) at a sampling rate of 20 frames per second. The exposure time was set to 50 ms and the recording duration was 30 s for each 512 × 512 pixel frame. The microscope incubator temperature was maintained at 37°C during imaging.

Considering the previous results of EB diameter and beating frequency, hanging drops, hanging for 3 days, and a spherical plate, lasting for 2 days, were chosen for all further experiments. The final pacemaker cell differentiation process and a comparison of the two methods for generating EBs are provided in Figure 1. Additionally, the use of the spherical plate leads to a greatly increased efficiency in terms of yield, in addition to the incubator space, labor, and time required.

In extensive pilot experiments, we analyzed various dissociation enzymes and kits according to the manufactureŕs instructions (Supplementary Figure S1A). Thereby, Pierce™ Primary Cardiomyocyte Isolation Enzymes (hereafter referred to as PCIE) (Thermo Fisher Scientific; Waltham, USA) provided the most efficient dissociation, leading to the single cells required to generate a synchronized cell layer on MEA (Supplementary Figure S1B). Consequently, dissociation was performed using 50% PCIE dissolved in PBS for 30 min for all MEA analyses. Following this step, we optimized the number of cells seeded on the MEA, which ranged from 50,000–400,000 cells to finally ∼130.000–150.000 cells covering the electrode field (Supplementary Figure S1C).

Based on our previous investigations comparing different concentrations of Matrigel and gelatin (data not shown), Matrigel (1:50) was consistently used to coat the MEA for all experiments described herein.

Comparison of EB generation methods HD and SP in terms of drug-induced cell responses. In light of previous findings, hanging drops, with 3 days of hanging, and spherical plates with a duration of 2 days were directly compared utilizing substance testing on a microelectrode array. Simultaneously started differentiations were cultured until day 20/21, after which they underwent collagenase IV treatment followed by separation into iSABs. The following day after filtration (two days after collagenase treatment), iSABs were dissociated using PCIE. The enzyme solution was diluted by 50% with PBS and incubated for 30 min. The final cell suspension was then placed onto a pre-coated MEA array (Matrigel 1:50, 30 min) at a density of 130,000 cells per MEA in a fresh differentiation medium without ascorbic acid. The MEA arrays used were purchased from Multi-Channel-Systems (60MEA100/10iR-Ti and 60MEA200/30iR-Ti). The arrays were cultured under standard conditions for 10 days, with daily medium changes to facilitate the synchronization of the pacemaker cells. The last medium change was performed at least three hours before the measurement. On days 2, 5, and 10, the pacemaker cells generated by HD or SP were measured on MEA arrays to detect synchronization over time. For this purpose, MEA arrays were measured using the MEA2100 system from Multi-Channel Systems, with a 10 kHz sampling rate, temperature control set to 37°C, and ambient atmosphere. Following adaptation on the pre-warmed headstage, the field potential was measured. We calculated the mean heart rate, conduction velocity, amplitude, slope, and duration. We evaluated five electrodes per MEA for all parameters, except for duration (for which we evaluated ten electrodes) and velocity (which had only one value measured over the entire electrode field). Measurement and analysis were performed using the dedicated cardio-specific software, Cardio2D/ Cardi2D + software (Multi-Channel-Systems).

The effects of test substances/drugs were analyzed using 1 M stock solutions of isoprenaline and carbachol that were prepared with ddH₂O. ZD7288, which is also known as (4-(N-ethyl-N-phenylamino)-1,2-dimethyl-6-(methyl-amino) pyrimidinium chloride) was dissolved in DMSO to obtain a stock solution of 5 mM and 50 mM, respectively. For the measurement of concentrations between 10⁻⁹ and 10⁻⁶ (10⁻⁴), further dilutions were made in the differentiation medium. All substances and DMSO were purchased from Sigma Aldrich. After adaptation and a control measurement without substrate addition, increasing substrate concentrations were added gradually and washed in by gentle distribution using a pipette for 1 min. Substrate concentrations were added in small volumes to prevent the wash-in process from affecting the results. Subsequent measurements of the field potential was made for one minute.

To further investigate the time dependence of ZD7288-induced effects, cells were measured under baseline conditions and at intervals of 1, 2, 5, 10, and 20 min after the addition of 5 µM ZD7288, as used in Jung et al. (20). Finally, MEAs were then cleaned using TergazymA (Alconox Inc., White Plains, USA) (1% in ddH2O) overnight on a shaker and subjected to water rinsing and autoclaving.

Separated pacemaker cells were seeded on Matrigel-coated coverslips. After initial fixation of the cells with 4% PFA for 20 min, the cells were washed three times with PBS for 5 min each, followed by permeabilization (0.1% TritonX and 0.1 sodium citrate in PBS) for 3 min. After a first blocking was performed using a blocking solution (0.2% gelatin and 10% FCS in PBS), the cells were incubated in the dark for 1 h with the following primary antibodies diluted in antibody solution (0,01% Saponin, 0,2% gelatin, 10% FCS in PBS): αMHC [mouse anti- heavy chain cardiac myosin antibody (3–48)](Abcam Ab15) 1:200, α-sarcomeric actinin (rabbit anti—sarcomeric alpha actinin antibody)(Abcam ab68167) 1:100, HCN4 (rabbit anti-hyperpolarization-activated cyclic nucleotide-gated potassium channel 4 antibody)(Alomone Labs APC-052) 1:100, cardiac troponin (mouse anti-cardiac troponin t antibody)(Abcam ab8295) 1:150, Cx30.2 (rabbit anti-connexin 30.2 antibody)(Invitrogen 40–7,400) 1:50. Cells were washed with blocking solution, after which they were incubated for 45 min with the appropriate secondary antibody diluted in antibody solution: Alexa Fluor 647 (Donkey anti-mouse IgG H&L, Abcam 150111) 1:500, Alexa Fluor 594 (Goat anti-rabbit IgG H&L, Abcam 150080) 1:500, Alexa Fluor-488 Phalloidin (Invitrogen, A12379) 1:500, DAPI (1:2,000). Finally, the cells were washed with PBS for 5 min and embedded in FluorSave mounting medium (Merck, Burlington, USA). Microscopic imaging was performed using the ELYRA PS.1 LSM 780 microscope with a 40× or 63× oil objective.

Statistical analysis was performed using GraphPad Prism. The diameter of the EBs was studied using a One-way ANOVA with Tukey's Multiple Comparison test. The number of beating foci was analyzed by One-way ANOVA with Bonferroni's Multiple Comparison Test, while the optically determined beating frequencies were analyzed by One-Way ANOVA with Tukey's Multiple Comparison Test. The MEA measurements of controls between HD and SP were compared using an unpaired t-test with Welch's correction. The effects of substrate administration were assessed using One-way ANOVA RM or Friedman's test with Dunnett's Multiple Comparison Test, after normality testing (Shapiro–Wilk Normality test). Data are presented as the mean ± SD. Single values are plotted, while Box plots indicate the mean and minimum/maximum whiskers. A p-value of *p ≤ 0.05, **p ≤ 0.01, and ***p ≤ 0.001 was considered significant.

The formation of EBs is a commonly used initial stage in the differentiation of pluripotent stem cells into tissue-specific cell subtypes. Three-dimensional aggregation and organization of cells is a crucial step in the induction of differentiation. The HD technique, a frequently used approach for manual, small-scale production of EBs, is a technique that is time-consuming, requires significant space, and is cost-intensive, producing low output. To achieve large-scale production, a technique that generates a high number of EBs in a simplified manner is preferable. Recently, spherical multiwell plates (SP) (Sphericalplate 5D, Kugelmeiers, Switzerland) have emerged as a time- and space-saving alternative, that allows the production of size-standardized EBs on a large scale. Consequently, we evaluated and compared both methods, maintaining a slightly modified version of our established iSAB production protocol as a reference (20, 35).

In four independent experiments (each 10 EBs per condition), we analyzed the diameter of EBs after one, two, or three days of formation time (see Figure 2). Exemplary images of EBs are shown in Figure 2A. We found that EBs formed in SP exhibited greater consistency in size across multiple experiments than those formed in HD. Moreover, the collection of EBs in HD on day one was challenging due to their small size, with approximately 400 cells used per EB. The larger diameter of EBs, observed three days after the formation of HD can be attributed to the comparatively greater available space in comparison to SP. However, excessively large diameters are associated with longer diffusion paths, therefore potentially resulting in inferior nutrient supply to the centered cells (44). Even in individual experiments, there was considerable variation in the EB diameter in HD depending on the position on the plate. This could potentially lead to further inhomogeneity in subsequent experiments (Figure 2B).

Moreover, the number of beating areas per EB was optically analyzed (Figure 3A). The EBs were generated using either the HD or SP method and then either directly plated onto gelatin-coated petri dishes or cultured as non-adherent EBs for an additional 3–4 days before plating. On days 16/17 of differentiation, the average number of beating foci per initial EB was determined. The use of HD for two days followed by floating culture resulted in a significant increase in the number of beating foci per EB, as compared to SP-generated EBs under seeding/floating culture conditions. Thereby, various differentiation time points were analyzed; Figure 3A displays the outcomes on days 16/17. Longer EB formation periods could lead to a lower number of beating foci, indicating further maturation and synchronization of the cells.

The beating frequency is a decisive parameter for functional cell characterization and therefore a crucial control for successful differentiation. In this regard, Figure 4B illustrates the beating frequencies of iSABs at days 25/28 of differentiation, depending on the type of EB formation method, duration, and subsequent culture technique. As the beating frequency of iSABs derived from 2-day SP-EB is comparable to or even slightly higher than that of those derived from HD-EB, resulting in approximately 400 bpm, which reflects the physiological range of the murine pacemaker, we integrated the SP method into our differentiation protocol. Furthermore, it is more convenient for the user and saves time, space, and plastic resources. For the subsequent drug testing experiments, we utilized SP-EB formation as it is a highly standardized method for generating EBs. This method was implemented for 2 days, and it yielded more promising results in terms of beating frequency and consistency when directly seeded onto gelatin-coated wells.

After establishing an optimized dissociation and re-seeding protocol, our objective was to determine a suitable time frame for drug testing experiments. This required the creation of an electrically coupled and synchronized cell layer with a stable conduction velocity. Therefore, cells were cultured for up to 10 days on the MEA chip. Microscopic control and field potential measurements were carried out on days 0, 2, 5, and 10 after seeding of iSAB-derived single cells to study the synchronization and to compare pacemaker cell layers from HD or SP-induced differentiation (Figure 4).

The color map, shown in Figure 4A), represents the cardiac activity at a distinct time point. The signal originates from a signal-generating area (pacemaker; red area) before propagating to the surrounding cells with a decent latency. The propagation time corresponds to the longest time span between the minimum and maximum points, while the lines illustrate areas of equal (isochronous) activation time. A video of the cell layer with the corresponding signal propagation can be found in the (Supplementary Movie S1). Intercellular connections enable the development of a cell layer, which can be seen in an exemplary bright-field microscopic picture of an MEA chip (Figure 4A). Conduction velocity, a key parameter in MEA analysis, was measured over a period of 10 days following cell seeding. No significant changes were detected with increasing culture time, as shown in Figure 3B. Single values demonstrate a smaller number of data points two days after seeding, compared to the later time points. Shortly after seeding the cells, the conduction velocity could only be determined in a small area because the cell layer showed different pacemaker foci and was not yet synchronized (Figure 4C). This reveals the need for maturation on the MEA chip. By prolonging the cultivation time, the measurable electrode regions can be expanded to cover the entire area for precise conduction velocity calculation. This results in a more accurate detection of the conduction velocity, increasing the amount of data points (days 5 and 10). In HD cell layers the velocity remained almost stable compared to SP, where a slight rise was observed between day 2 and day 5 of cultivation on MEA chips with a subsequent stabilization between 0.02 and 0.03 m/sec. Exemplary traces of measured field potentials on day 0, the day of cell seeding, and after 9 further days of culture on the MEA are shown in Figure 4C. The traces became more balanced between the maximum and minimum with increasing culture times. Signal propagation matures as indicated by shorter latencies, and the number of pacemaker foci decreases with processing synchrony, as shown in the corresponding color maps. From several pacemaker foci, one takes over the main activity, resulting in a transition from undirected to more directed signal propagation.

Finally, we aimed to validate our established protocols with respect to their suitability for testing novel drugs. We pursued this goal by verifying the physiological response of the cells to test substances. Electrophysiological measurements were performed on 60-electrode MEAs, including one reference electrode, after a 10-day maturation period on the MEA.

Comparing the baseline parameters without substrate administration, the measurements revealed similar values for beating frequency, amplitude, and conduction velocity between the two EB generation strategies (Figures 5A–C). The data were obtained from all baseline measurements of the 9 MEAs before the drug wash-in. For beating frequency and amplitude, a total of 45 electrodes from 9 MEAs were analyzed. The displayed single values indicate low MEA to MEA variability. The mean contraction frequencies range from 450 to 550 bpm, while the conduction velocity reaches values around 003 m/s. Additionally, there is a relatively small average FP amplitude of approximately 80 µV. Figure 5

Concentration-dependent response to substances in iSAB pacemaker cells comparing the spherical plate (SP) and hanging drop (HD) techniques. Changes in the physiological parameters contraction frequency, amplitude, and conduction velocity before drug wash-in (baseline) (A–C) and after the administration of isoprenaline, a non-selective beta-sympathomimetic drug (D–F), carbachol, a parasympathomimetic drug (G–I) and ZD7288, a funny channel blocker (J–L) is illustrated. Red dots indicate HD-generated pacemaker cells and blue dots indicate SP-generated pacemaker cells. For control (baseline) measurements, a set of n = 9 MEAs (45 electrodes for frequency and amplitude) (A–C) was analyzed, whereas drug administration was measured at n = 15 electrodes from 3 MEAs for frequency and amplitude, and n = 3 MEAs each for velocity. Data were normalized to the respective untreated control (baseline). Box plots show single values and whiskers at minimum/maximum. Treatment graphs show mean ± SD as well as single values. For normality analysis, the Shapiro–Wilk Test was used. Significance was calculated using an unpaired t-test with Welch's correction for control (A–C). The significance of the respective untreated control was calculated using a one-way ANOVA RM or Friedman Test with Dunnett's Multiple Comparison Test (drug administration D–L). *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001.

Additionally, the non-selective funny channel inhibitor ZD7288 was analyzed for specific pacemaker cell responses (Figures 5J–L). iSAB cells showed a significant dose- and time-dependent reduction in beating frequency for SP, as expected according to our previous work (20). FP measurements revealed a stable, but slightly decreasing beating frequency. The amplitude remained nearly constant after the addition of ZD7288, showing a high level of agreement between the SP and HD techniques. Conduction velocities did not differ significantly between the HD and SP techniques. Exemplary field potential traces before and after the addition of the funny channel inhibitor are shown in Supplementary Figure S3C.

Although there were slight differences observed when adding ZD7288 at different concentrations, the anticipated significant reduction in frequency was not seen. This may be due to either the concentration being too low or the measurement period being too short. Therefore, we performed the MEA measurements with the SP-generated cells and the previously published concentration of 5 µM [Jung et al. (20), Ca Imaging] to obtain a detailed view of the time-dependent response of pacemaker cells to the administration of ZD7288 (Figure 6). The field potentials of MEAs were measured at 1, 2, 5, 10 and 20 min after adding 5 µM ZD7288. A significant decrease in normalized beating frequency was observed at 5, 10, and 20 min compared to baseline conditions. Conduction velocity did not change significantly, but the peak slope increased significantly right after the addition of ZD7288 compared to baseline field potentials. Remarkably, we found that the second upstroke of the field potential, representing repolarization, was shortened with increasing application time, reaching the t_0µV point significantly faster. This could be seen as a pendant to the field potential duration usually measured in iPSC-CMs.