An established hiPSC control line (hiPSCreg code: UKEi001-A, ERC01) and patient-derived PLN-R14del hiPSCs, along with the respective isogenic control (Cuello et al., 2021), were differentiated into cardiomyocytes. In brief, hiPSCs from the master/working cell bank (Shibamiya et al., 2020) were expanded on Geltrex-coated cell culture vessels in FTDA media (Frank et al., 2012). Embryoid bodies were generated in spinner flasks. Ventricular cardiomyocytes were differentiated in suspension/EB format by growth factor/small molecule cocktails into mesodermal progenitors and subsequently into cardiomyocytes. Collagenase-dissociated hiPSC-CMs were either cryo-preserved or directly used for EHT generation. Differentiation efficiency was determined by FACS analysis for troponin T. All procedures were performed as described by Breckwoldt et al. (2017).

The production and purification of adeno-associated virus serotype 6 (AAV6) vector particles encoding the calcium sensor G-CEPIA1er were adapted from a previous publication (Saleem et al., 2020). Recombinant adeno-associated virus serotype 6 (rAAV6) particles were produced using the baculovirus expression system techniques. To introduce the calcium sensor into the baculoviral transfer plasmid pFBGR-Ultra, a PCR fragment encoding G-CEPIAer, flanked by ER signaling and retention sequences (Suzuki et al., 2014), was generated using primers (5′-AACATCGATTGAATTCgccaccatgggatggagc and 5′- tatagggcgaattgggtaccctacagctcgtccttctcgct), with pCMV G-CEPIA1er (Addgene #58215) as a template. The fragment was inserted into pFBGR-Ultra TnT-GFP, previously digested with EcoRI and KpnI, using In-Fusion Cloning (PrimeSTAR GXL Polymerase, TaKaRa Bio Europe SAS, Cat. No. R050B). Final baculovirus transfer plasmids were confirmed by restriction digesting and whole plasmid sequencing (MWG Eurofins). The genomic titers of DNAse-resistant recombinant AAV6 particles were determined by quantitative PCR using T7/SV40 specific primers (5′-cctatagtgagtcgtattacgcgc and 5′-gctgcaataaacaagttgggccat).

Control hiPSC-CMs were plated on Geltrex-coated 96-well plates in EHT media. HiPSC-CMs were incubated for 14–21 days, with media being changed every 2–3 days. For immunofluorescent staining, the media was removed. HiPSC-CMs were briefly washed with PBS and fixed with 4% paraformaldehyde for 15 min at 4 °C. Then, fixative was removed, and hiPSC-CMs were permeabilized with either 0.2% Triton X-100 (Roth 3051.3) or 0.2% saponin (Sigma, S7900, from Quillaja bark) in PBS for 5 min at room temperature (RT), followed by washing in 200 µL PBS/well for 5 min. Subsequently, non-specific binding sites were blocked with 5% normal goat serum (NGS) diluted in antibody buffer (10 mM Tris, 155 mM NaCl, 2 mM EGTA, 2 mM MgCl₂, and 1% (w/v) BSA; pH 7.5) for 20 min at RT. For primary antibody incubation, cells were incubated overnight with 50 µL/well of the primary antibodies against calnexin (Novus, AF18), anti-alpha-actinin (Sigma-Aldrich, A7811), or SERCA2 (Invitrogen, MA3-919) diluted in antibody buffer (10 mM Tris, 155 mM NaCl, 2 mM EGTA, 2 mM MgCl₂, and 1% (w/v) BSA; pH 7.5) at 4 °C in a humid chamber under shaking. After washing three times with 200 µL PBS/well, the cells were incubated with 50 µL of secondary antibody solution consisting of the secondary antibodies Alexa Fluor 647 goat anti-mouse (1:100, Invitrogen, A21236), DyLight 550 Goat Anti-Rabbit (1:100 DyLight 550, Abcam, ab96884), and 4′,6-diamidino-2-phenylindole (DAPI, 1:100, Sigma, D9542) in a humid chamber on a shaker for 3 h at RT.

Fibrin-based strip-format engineered heart tissues were generated with 1.0 × 10⁶ hiPSC-CMs per construct, as described by Breckwoldt et al. (2017). During EHT casting, hiPSC-CMs were transfected with AAV6-CEPIAer at a multiplicity of infection of 1.0 × 10⁵. EHTs were cultivated for approximately 21 days in EHT medium (10% horse serum, 1% penicillin–streptomycin, 0.1% aprotinin, 0.1% insulin, and 0.25% tranexamic acid diluted in Dulbecco’s modified Eagle medium) in 24-well plates with media changes on Mondays, Wednesdays, and Fridays.

The EHTs were transferred to a 24-well plate prefilled with EHT measurement medium (FluoroBrite DMEM, Thermo Fisher Scientific). The plate was transferred to an EHT contractility system with a customized extension. This extension consisted of an XYZ-axis controlled fluorescence detection unit (FDU, including an objective, a filter system, and a photomultiplier) mounted under the 24-well plate and integrated into the analysis software to enable simultaneous measurement of force and GFP fluorescence. XYZ-axis coordinates were defined for contour-based video-optical analysis of contractile force and fluorescence detection. Automated force/fluorescence analysis was performed at baseline and after the addition of pharmacological modulators of ER calcium loading and release. After the experiment was completed, the EHTs were washed three times in EHT medium (10% horse serum, 1% penicillin–streptomycin, 0.1% aprotinin, 0.1% insulin, and 0.25% tranexamic acid diluted in Dulbecco’s modified Eagle medium) and transferred back to the EHT media. All media and washing solutions were pre-equilibrated in a cell culture incubator. The pharmacological compounds used were as follows: isoprenaline (Sigma-Aldrich, I-5627), thapsigargin (Sigma-Aldrich, T9033), cyclopiazonic acid (Sigma-Aldrich, 239805), istaroxime (MedChemExpress, HY-15718A), ryanodine (Sigma-Aldrich, 559276), tetracaine (Sigma, T7508), and adenosine triphosphate (ATP) (Sigma, A6419). Thapsigargin-based in vitro assays have been established to induce ER calcium dysregulation and ER stress. The stimulation was carried out for up to 24 h (Abdullahi et al., 2017). In this study, the effects of thapsigargin and other indicator compounds were analyzed until the effects on CEPIAer FI reached plateau values and for a maximum period of up to 370 min.

EHTs were incubated in EHT measurement media (FluoroBrite DMEM, Thermo Fisher Scientific) in the presence of pharmacological modulators of ER calcium loading and release for the defined period of time. EHTs were briefly washed with pre-warmed (37 °C) PBS, transferred to 2.0-mL sample tubes, snap-frozen in liquid nitrogen, and stored at −80 °C. Total RNA was isolated using TRIzol reagent (Ambion, RNA by Life Technologies, 15596–026) according to the manufacturer’s instructions. RNA concentration and quality in the eluate were evaluated using NanoDrop instruments. Reverse transcription was performed using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems 4368813) according to the manufacturer’s instructions. Quantitative real-time PCR was performed using 5x EvaGreen (Solis BioDyne) according to the manufacturer’s instructions. Primer sequences for ER stress response markers and housekeeping genes were used as described by Cuello et al. (2021).

HiPSC-CM EHTs were dissociated using a 50% (v/v) papain solution (10 U/mL papain (Sigma-Aldrich, 76220), 1 mM EDTA (Roth 8043.2), and 5 mM L-cysteine-HCl (Sigma-Aldrich, C1276) in 1x EBSS (Gibco 14155–048)) diluted in HBSS (Gibco 14175–053). The EHTs were incubated for 30–40 min until single-cell dispersal was observed. Dissociated hiPSC-CMs were stored on ice in EHT media and later transferred to EHT measurement media. CEPIAer GFP FI was determined by FACSCanto II (BD). Analysis was performed using FACSDiva software (BD).

Data were expressed as the mean ± SEM. GraphPad Prism was used to compare between groups with an unpaired, two-sided Student’s t test or one-way ANOVA, as indicated. Individual hiPSC-CM EHTs or single hiPSC-CMs from 2–3 different EHT generation batches (as indicated in the figure legends) were considered biological replicates.

hiPSC-CMs were plated as a 2D monolayer and transduced with AAV6 encoding the CEPIAer Ca²⁺ sensor, flanked by endoplasmic reticulum targeting and retention sequences under the control of a troponin T promoter. This CEPIAer construct was previously characterized and shown to have ER-specific expression (Suzuki et al., 2014). Immunofluorescence staining revealed perinuclear CEPIAer fluorescence in alpha-actinin-positive hiPSC-CMs (Figure 1, upper panel). CEPIAer fluorescence co-localized with staining for the ER resident protein calnexin (Figure 1, middle panel) and with the perinuclear fraction of SERCA2 staining (Figure 1, lower panel).

A functional assay was established to simultaneously monitor CEPIAer FI and auxotonic contraction in spontaneously beating EHTs. Figure 2A shows the average contraction peak and non-baseline-corrected CEPIAer FI. The minor increase in FI during contraction is indicative of a motion artefact. The CEPIAer FI level of non-baseline-corrected recordings without taking the increase during contraction into account (e.g., 2.4 AU, Figure 2A) was quantified as a surrogate for ER Ca²⁺ loading in this study. The plot of average force and CEPIAer FI with baseline-correction is shown in Figure 2B, and an original recording is shown in Figure 2D. Pre-treatment with the myosin inhibitor butanedione monoxime (BDM, 10 mM), which suppresses mechanical shortening during electromechanical coupling, was performed in electrically paced EHTs. Here, force was reduced to the minimal values, and CEPIAer FI with baseline-correction did not show motion artifacts but instead showed a minimal decrease during contraction (Figures 2C, E).

In order to validate the assay, the effects of indicator compounds on CEPIAer FI, force, and spontaneous beating frequency were analyzed. Time-dependent effects were studied in EHTs of a control hiPSC line (control) and the isogenic (PLNic) and PLN-R14del hiPSC line. This isogenic hiPSC pair was previously characterized and revealed cytoplasmic Ca²⁺ irregularities and lower force associated with the impairment of the ER/mitochondria compartment for PLN-R14del (Cuello et al., 2021). Cyclopiazonic acid and thapsigargin were studied as inhibitors of ER Ca²⁺ uptake. Cyclopiazonic acid is a competitive antagonist at the SERCA ATP-binding site (IC₅₀, 0.6 µM (Seidler et al., 1989)). Thapsigargin is a non-competitive antagonist at the SERCA ATP-binding site (IC₅₀, 1.0 nM (Lytton et al., 1991)). CPA (10 µM) and thapsigargin (1 µM) led to a time-dependent reduction in CEPIAer FI in all three cell lines (CPA: control: 76% ± 1.9%; PLNic: 86% ± 3.8%; and PLN-R14del: 84% ± 1.7% and thapsigargin: control: 82% ± 3.3%; PLNic: 69% ± 2.4%; and PLN-R14del: 73% ± 5.0%). No effect on force and frequency was detected. In PLNic, CPA led to a transient reduction in frequency (at 130 min, 86% ± 3.5%; Figures 3A, B, Supplementary Figures 1A, B).

Ryanodine, tetracaine, and ATP modulate ER Ca²⁺ release. Ryanodine acts on ryanodine receptors and mediates an agonistic effect at nanomolar–low micromolar concentrations and an antagonistic effect at higher micromolar concentrations (Alexander et al., 2023a). At 10 μM, ryanodine resulted in a time-dependent decrease in CEPIAer FI (control: 77% ± 1.4%; PLNic: 84% ± 3.0%; and PLN-R14del: 75% ± 5.5%). A reduction in force (89% ± 3.1%) and frequency (78% ± 4.1%) observed in the control was not detected either in PLNic or PLN-R14del EHTs (Figure 3C, Supplementary Figure 2A). At 200 μM, ryanodine resulted in a biphasic effect on CEPIAer FI in all three cell lines. An initial decrease after 10 min (control: 68% ± 3.7%; PLNic: 81% ± 3.9%; and PLN-R14del: 84% ± 2.1%) was followed by an increase after 90–210 min (control: 136% ± 5.7%; PLNic: 130% ± 6.0%; and PLN-R14del: 119% ± 2.3%). The initial decrease in CEPIAer FI is compatible with a transient agonistic effect at early time points when the ryanodine concentration at the receptor is building up. In control and PLNic, 200 µM ryanodine had no effect on force and frequency. In contrast, in PLN-R14del, a transient reduction in force and increase in frequency (56% ± 3.4% and 172% ± 9.0%, respectively) were detected (Figure 3D, Supplementary Figure 2B). ATP activates P2Y receptors and mediates a release of ER Ca²⁺ via IP₃R activation (Reddish et al., 2021). In EHT, ATP (10 mM) led to a time-dependent decrease in CEPIAer FI (control: 70% ± 2.1%; PLNic: 68% ± 4.1%; and PLN-R14del: 76% ± 2.1%). This was accompanied by a discontinuation of contraction (Figure 4A, Supplementary Figure 3A). The latter effect is compatible with the inhibition of L-type Ca²⁺ channels, as previously described for ferret and guinea pig cardiomyocytes, while an opposite effect was described for rats (Vassort, 2001).

Two additional compound effects were studied in the control hiPSC line. The sodium channel blocker tetracaine inhibits ryanodine receptors (Laver and van Helden, 2011) and reduces Ca²⁺ release from the ER. Tetracaine (50 µM) led to an initial increase in CEPIAer FI (122% ± 3.0%). This was partially reverted by an increase in extracellular Ca²⁺ concentration from 1.8 mM to 5 mM. The effects on CEPIAer FI were accompanied by a cessation of contraction due to the inhibition of sodium channels (Grima et al., 1986) (Figure 4B Supplementary Figure 3B). In hiPSC-CM EHTs, this effect has been demonstrated with the sodium channel inhibitors tetrodotoxin, ajmaline, flecainide, quinidine, and mexiletine (Mannhardt et al., 2016). Istaroxime is an activator of SERCA2 and an inhibitor of the Na, K-ATPase (Rocchetti et al., 2005). At 1 and 4 µM, istaroxime led to a non-significant gradual CEPIAer FI increasing trend, which is compatible with the proposed mechanism of action. Force and frequency were not altered at 1 µM, but variability increased at later time points, and at 4 µM, istaroxime led to an abrogation of contractility, which is explained by the Na, K-ATPase inhibition (Figures 4C, D, Supplementary Figure 3B). Similar effects were described by digoxin in hiPSC-CM EHTs (Saleem et al., 2020). Increasing extracellular Ca²⁺ concentration is a well-established experimental method to increase cytosolic Ca²⁺ concentration (Wang et al., 2003), sarcoplasmic reticulum Ca²⁺ loading, and inotropy. Extracellular Ca²⁺ also activates Ca²⁺-sensing receptors and triggers Ca²⁺ release from ER via Gα _q/11/IP₃ signaling (Gorkhali et al., 2021). Moreover, the increase in cytosolic calcium triggers the opening of IP₃R in cardiomyocytes and also ryanodine receptors at higher concentrations (Alexander et al., 2023a; Mak et al., 1998). An increase in extracellular calcium concentration (5 mM) in this EHT model resulted in a decrease in CEPIAer FI in the control and PLNic EHTs (control: 82% ± 2.0%; PLNic: 82% ± 3.8%). In contrast, in PLN-R14del, it had no effect on CEPIAer FI. High extracellular calcium (5 mM) evoked a positive inotropic effect in all three cell lines (control: 113% ± 3.5%; PLNic: 134% ± 7.6%; and PLN-R14del: 128% ± 3.4%), accompanied by a reduction in spontaneous beating frequency (control: 68% ± 2.6%; PLNic: 55% ± 2.4%; and PLN-R14del: 68% ± 4.2%) (Figure 5A, Supplementary Figure 4A).

Isoprenaline-mediated beta-adrenergic activation is a well-established mechanism for PKA-mediated PLN phosphorylation, thereby reducing PLN-mediated SERCA2a inhibition and increasing re-uptake of calcium from the cytosol into the sarcoplasmic reticulum. In addition, PKA phosphorylates IP₃Rs (Taylor, 2017) and RyRs (Marx et al., 2000), two important components of calcium release from the ER, thereby increasing their open probability. In control and PLNic EHTs, isoprenaline (100 nM) had no effect (control) or a decreasing effect (PLNic: 83% ± 4.0%) on CEPIAer FI. In contrast, isoprenaline induced a transient increase in CEPIAer FI in PLN-R14del (116% ± 5.4%). In all three cell lines, isoprenaline evoked a transient positive inotropic (control: 113% ± 3.7%; PLNic: 113% ± 3.9%; and PLN-R14del: 126% ± 7.0%) and chronotropic effect (control: 140% ± 5.6%; PLNic: 130% ± 2.8%; and PLN-R14del: 165% ± 10.4%; Figure 5B, Supplementary Figure 4B). Supplementary Figure 5 shows the time vehicle controls (TVCs) for these functional analyses. Recordings of CEPIAer FI and force showed stable values with low variability. The beating frequency was less stable and increased in several TVC experiments at late recording time points.

The calcium loading experiment revealed subtle alterations in ER calcium loading in PLN-R14del, which are compatible with higher SERCA-mediated uptake. To investigate whether this finding translates into differences in ER calcium loading at baseline under conditions shown to be associated with cytoplasmic calcium irregularities for PLN-R14del (Cuello et al., 2021), hiPSC-CMs were dissociated from PLNic and PLN-R14del EHTs on day 35–40 of development and subjected to flow cytometry analysis. The analysis revealed higher CEPIAer FI in PLN-R14del (17.2 ± 1.8 AU) than in PLNic (10.3 ± 1.5 AU, mean ± SEM, Figure 6A), suggesting higher ER calcium load in PLN-R14del under baseline conditions. The same CEPIAer AAV batches were used to transduce hiPSC-CMs from both lines. Nevertheless, to account for the variability in CEPIAer transduction and expression, analyses of n = 11–12 EHTs from three different EHT generation batches were performed. Thapsigargin is a known trigger of ER calcium dysregulation and stress response (Abdullahi et al., 2017). To verify this effect in the hiPSC-CM-EHT model and investigate whether such effects were also induced by other modulators of ER calcium homeostasis used in this study, transcript levels of the ER stress response were quantified in control EHTs under the different stress conditions. This validation analysis focused on control hiPSC-CMs as PLN-R14del hiPSC-CMs showed signs of ER stress response at baseline (Cuello et al., 2021). The analysis confirmed an increase in the expression of ER stress markers by thapsigargin and also by other modulators such as CPA, calcium (5 mM), and ryanodine (Figure 6B). These results suggest that the experimental conditions in this study lead not only to changes in the ER calcium load but also to an ER stress response.