Differentiation of hiPSCs into cardiomyocytes

The hiPSC line utilized in the current study (hIH-1) was recently established in our laboratory [15] by retroviral delivery of three reprogramming factors: OCT4, SOX2, and KLF4 together with valproic acid (VPA), a histone deacetylase inhibitor potentiating the reprogramming ability of these factors [16]. This hiPSCs line was demonstrated to fulfill all the criteria defining the iPSC state including full reprogramming, pluripotency, and genetic stability [15]. In the current study we used two clones (hIH-1 clones 1&2) of this line that were derived independently during reprogramming of the human fibroblasts. In addition, we also studied a second well-characterized hiPSCs line (hFib2-iPS cells; kindly provided by G.Q Daley) [8], [17], which was established by retroviral transduction of human fibroblasts with OCT4, SOX2, c-MYC, KLF4, together with hTERT and SV40-large T [17].

Undifferentiated hiPSC colonies were cultured on a mitotically-inactivated MEF feeder layer [8]. The culture medium consisted of 80%-knockout high-glucose glutamine-free DMEM with sodium-pyruvate supplemented with 20%-serum-replacement, 1 mM L-glutamine, 0.1 mM mercaptoethanol, 4 ng/mL human recombinant basic fibroblast-growth-factor, and 1%-nonessential amino acid stock (Invitrogen).

To induce differentiation, hiPSCs were dispersed into small clumps using collagenase-IV (Life-Technologies, 1 mg/mL, 20 min) and cultivated in suspension where they aggregated to form embryoid bodies (EBs). The EBs were plated after 10 days on gelatin-coated culture dishes and observed for the appearance of spontaneous contracting areas. The beating areas within the EBs were mechanically microdissected at 30–50 days following the appearance of spontaneous beating to allow comparison with studies assessing hESC-derived cardiomyocytes at similar developmental stages [18], [19]. This was followed by enzymatic dispersion (1 mg/ml collagenase B, Roche, Mannheim, Germany) at 37°C for 30 min to derive single cardiomyocytes or small monolayered clusters. After dispersal, the cells were plated onto fibronectin-coated glass coverslips for experimentation.

Reverse transcription PCR (RT-PCR) analysis

The expression of various cardiac Ca²⁺-handling genes was tested using semi-quantitative RT-PCR. Total RNA was extracted from human embryonic kidney cells (HEK293, American Type Culture Collection (ATCC), Manassas, VA) and from beating hiPSC-CMs using RNeasy Mini Kit (Qiagen). cDNA was synthesized using Verso™-RT-PCR-Kit (Thermo-Scientific). 1.6 ng of cDNA was subjected to the following PCR program: 3 min at 93°C, 30 sec at 93°C, 30 sec at 60°C, and 30 sec at 72°C. The PCR-related primers are detailed in Table 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Primers names and sequences.

https://doi.org/10.1371/journal.pone.0018037.t001

Immunocytochemistry

Single cells and small monolayered clusters were fixed using 4% paraformaldehyde and permeabilized with 1% Triton-X-100 (Sigma). Cells were blocked with 5% horse serum and 1% bovine serum albumin and incubated overnight at 4°C with primary antibodies targeting sarcomeric α-actinin (1∶200, Sigma-Aldrich), ryanodine receptor (1∶1,000, Chemicon), and pan-IP3R (reacting with the C-terminal cytoplasmic domain of IP3R types 1, 2, and 3) (1∶10, Millipore). The secondary antibodies were Cy3-conjugated donkey anti-mouse and Cy2-conjugated goat anti-rabbit IgG antibodies at 1∶200 for 1 h (Jackson). Nuclei were counterstained with DAPI (Sigma). Preparations were examined using a laser-scanning confocal microscope (Zeiss LSM-510-PASCAL).

Ca²⁺ Imaging

Cells were loaded with 5 µM fluo-4 fluorescent Ca²⁺ indicator (Molecular Probes) in the presence of Pluronic F-127 (Molecular Probes) at a dilution of 2∶1 to allow the recording of intracellular Ca²⁺-transients (whole-cell [Ca²⁺]_i transients) as previously described [18]. Experiments were conducted in the presence of indicator-free tyrode solution containing in mM: NaCl-140, KCl-5.4, CaCl₂-1.8, MgCl₂-1, HEPES-10, glucose-10 at pH 7.4 via 1 mM NaOH. The tyrode solution was superfused at a rate of 1 ml/minute and at a temperature of 37°C. Intracellular Ca²⁺-transients were recorded using a confocal imaging system (Fluoview; Olympus) mounted on an upright BX51WI Olympus microscope equipped with a X60 water objective [20]. Data were analyzed utilizing MatLab-based custom-written software.

To investigate caffeine mobilization of store Ca²⁺, temporally-limited puffs of caffeine were applied. This specific technique [18] was chosen to overcome the technical issue of the rate of caffeine delivery to the cells. The caffeine puff (20 mM, <2s) was applied by pressure-ejection through a pipette situated approximately 100 µm away from the target cells' position. The focal puff pipette was positioned central to the plane of the line scan and in the direction of constant bulk flow of the tyrode solution.

Statistical analysis

Data are presented as mean ± standard error of mean (SEM). Student's paired t-test was used to compare between means. When the effects of multiple concentrations of the pharmacological agents were studied, then one way ANOVA was used followed by Dunnet's post-hoc comparison to baseline values. p<0.05 was considered to be statistically significant.

Expression of Ca²⁺-handling molecules in hiPSC-CMs

To evaluate whether cardiac-related Ca²⁺-handling molecular components are present in the hiPSC-CMs, we first used semi-quantitative RT-PCR analysis. We assessed for the expression of the following Ca²⁺-handling proteins: RyR2, IP3R2, SERCA2a, Ca_v1.2 (encoding the α-1C subunit of the L-type Ca²⁺ channel), Calsequestrin, and Phospholamban. As can be seen in Figure 1, all of these genes were expressed in the hiPSC-CMs while absent in control HEK293 cells.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 1. Semi-quantitative RT-PCR analysis for cardiac-specific Ca²⁺-handling proteins in hiPSC-CMs.

Gene expression pattern of proteins involved in Ca²⁺-handling in hiPSC-CMs (upper panel) and control HEK293 cells. The lower panel displays genes expression in control HEK293 cells. β-tubulin was used as a normalizing house-keeping gene. Abbreviations: RyR2, ryanodine receptor; IP3R2, inositol 1,4,5-triphosphate receptor; SERCA2a, SR Ca ATPase; Ca_V1.2, L-type Ca²⁺ channels; Cal, calsequestrin, and PLN, phospholamban.

https://doi.org/10.1371/journal.pone.0018037.g001

Spontaneous whole-cell [Ca²⁺]_i transients in hiPSC-CMs

Spontaneous whole-cell [Ca²⁺]_i transients in hiPSC-CMs were recorded from spontaneously beating dispersed single cells or small monolayered clusters, under native cytosolic conditions. These transients were monitored in Fluo-4 (a fluorescent Ca²⁺ indicator) loaded cells, examined under a laser scanning confocal microscope utilizing the line-scan mode. Line scans were adjusted to avoid the cell nuclei and were localized at mid-cell z-section depth. Under control conditions, in the presence of 1.8 mM external bath Ca²⁺, spontaneous whole-cell [Ca²⁺]_i transients were recorded as cell-wide rhythmic events in all cells tested (n = 96) (Figure 2A).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 2. Whole-cell [Ca²⁺]_i transients in hiPSC-CMs and their requirement for Ca²⁺ influx via L-type Ca²⁺ channels.

(A) Confocal line-scan images showing changes in intracellular Ca²⁺ in a fluo-4 loaded hiPSC-CM (derived from the hIH1 line). The images showing spontaneous whole-cell [Ca²⁺]_i transients are displayed either as a distance-time plot (upper panel) or as a line-scan tracing (lower panel). (B) Whole-cell [Ca²⁺]_i transients in the presence (left) and absence (right) of bath Ca²⁺. (C) Whole-cell [Ca²⁺]_i transients before (left) and after (right) application of nifedipine (1 µM). (D) Dose-response curve showing the percentage of decrease in [Ca²⁺]_i transients amplitude relative to baseline conditions as function of nifedipine concentration (0.05, 0.1 and 1 µM, n = 6). *p<0.05 and **p<0.01 when mean absolute values were compared to baseline values. Abbreviations: F/Fo, fluorescence (F) normalized to baseline fluorescence (Fo); sec, seconds.

https://doi.org/10.1371/journal.pone.0018037.g002

Ca²⁺ influx via L-type Ca²⁺ channels contributes to whole-cell [Ca²⁺]_i transients

Transmembranal Ca²⁺ influx is an important initial trigger for excitation-contraction coupling in adult cardiomyocytes [9] and in hESC-derived cardiomyocytes (hESC-CMs) [19]. Therefore, the next step was to investigate whether the development of hiPSC-CMs whole-cell [Ca²⁺]_i transients require external Ca²⁺. To this end, we recorded whole-cell [Ca²⁺]_i transients in the presence (1.8 mM bath Ca²⁺) and absence of Ca²⁺ in the bath solution (n = 6). As can be appreciated in Figure 2B, in the absence of bath Ca²⁺ the whole-cell [Ca²⁺]_i transients were completely abolished.

To test whether the L-type Ca²⁺ channel serves as an important transmembranal Ca²⁺ influx pathway in hiPSC-CMs, as documented in adult cardiomyocytes [9], we tested the effect of Nifedipine, a L-type Ca²⁺ channel blocker. Whole-cell [Ca²⁺]_i transients were recorded before and after the application of 1 µM Nifedipine (Figure 2C). Similarly to what was observed in the absence of bath Ca²⁺ (Figure 2B), 1 µM Nifedipine led to the complete elimination of whole-cell [Ca²⁺]_i transients (n = 9). Dose-response studies using lower concentrations of nifedipine (0.05–1 µM, n = 6) demonstrated that the cells were extremely sensitive to L-type channel blockade with a steep decrease in [Ca²⁺]_i transients amplitude observed at a very low concentration (Figure 2D).

To verify that the results obtained were not due to clonal or line variations, we compared the results obtained in cardiomyocytes differentiated from two different clones of the main hiPSC line studied (hIH1 - clones 1&2 derived independently from the human fibroblasts) as well as from an additional well-characterized hiPSC line (hfib2-5) derived using the traditional 4-factors method [8], [17]. The dependency of whole-cell [Ca²⁺]_i transients on the presence of functional L-type Ca²⁺ channels was found to be independent of the specific hiPSC clone or line used. Thus, Nifedipine (1 µM) application resulted in complete elimination of whole-cell [Ca²⁺]_i transients in all cases (Figure S1).

Taken together these data confirm that transmembranal Ca²⁺ influx and specifically Ca²⁺ entry via L-type Ca²⁺ channels are important requirements for the generation of whole-cell [Ca²⁺]_i transients in hiPSC-CMs.

Functional RyR-mediated intracellular Ca²⁺ stores exist and contribute to whole-cell [Ca²⁺]_i transients

We next conducted immunocytostaining studies of hiPSC-CMs probing for both RyR2 and sarcomeric α-actinin in small monolayered clusters (Figure 3A). As previously shown in hESC-CMs [18], [21] sarcomeric α-actinin staining in hiPSC-CMs displayed a relatively disorganized striated sarcomeric arrangement (Figure 3A, left). RyR2 expression was exhibited throughout the cytosol (Figures 3A, middle), with some myofilaments co-localization (Figure 3A, right). The perinuclear region displayed intense staining as was similarly observed in mouse ESC-CMs [22] and hESC-CMs [18].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 3. Localization and functionality of Ca²⁺ store ryanodine receptors.

(A) A hIH1 hiPSC-CM co-labeled with antibodies for sarcomeric α-actinin (left) and RyR2 (middle). The merged image is displayed in the right panel. (B) A line-scan presenting the effect of 20 mM caffeine puff application (noted by the arrows). (C) A dose-response curve displaying the magnitude of the caffeine-induced Ca²⁺ release (quantified by the percentage increase of the amplitude of the caffeine-induced transient relative to the preceding action-potential induced [Ca²⁺]_i transient) as function of the caffeine bolus dosage (10, 20 and 40 mM). *p<0.05 when compared to the lowest (10 mM) dosage. (D) A line-scan tracings presenting caffeine-induced Ca²⁺ transients in the presence (left) and absence (right) of bath Ca²⁺. Abbreviations: F/Fo, fluorescence (F) normalized to baseline fluorescence (Fo); sec, seconds.

https://doi.org/10.1371/journal.pone.0018037.g003

To determine whether hiPSC-CMs possess loaded SR Ca²⁺ stores that release Ca²⁺ via functional RyRs we tested for caffeine responsiveness. Caffeine mobilization of store Ca²⁺ and its effect on whole-cell [Ca²⁺]_i transients were measured by applying a pressure-ejected caffeine puff (20 mM, 2 sec) on to fluo-4 loaded hiPSC-CMs. As can be observed in Figure 3B, caffeine application elicited an instantaneous, rapid, and large release of Ca²⁺ from the intracellular stores, resulting in a high amplitude caffeine-induced Ca²⁺ transient (n = 26). This was followed by reversible quiescence of whole-cell [Ca²⁺]_i transients (Fig. 3B) postulated to be a consequence of intracellular Ca²⁺ stores depletion. This phenomenon was noted in cardiomyocytes derived from all hiPSCs clones and lines studied (Figure S2A). Finally, dose-response studies (10, 20, and 40 mM caffeine puffs, n = 6) showed an escalating effect with an increase in the relative magnitude of caffeine-induced Ca²⁺ release (Figure 3C).

Next, it was important to validate that the caffeine-induced [Ca²⁺]_i transient was indeed a consequence of RyR-mediated SR Ca²⁺ release. To exclude the plausible contribution of Ca²⁺ influx through voltage-gated Ca²⁺ channels 20 mM caffeine puffs were applied in the presence (1.8 mM Ca²⁺; Figure 3D, left) and absence (Figure 3D, right) of bath Ca²⁺. Similarly to what was observed under control conditions (Figure 3D, left), caffeine puffs applied in the absence of bath Ca²⁺ (n = 6) induced an instantaneous rapid caffeine-induced [Ca²⁺]_i transient (Figure 3D, right) displaying an amplitude similar to that observed under control conditions. To exclude the possibility that the caffeine-induced [Ca²⁺]_i transient was a result of a mechanical stimulation to the cell surface, caused by the actual pressure injected puff, control puff (tyrode solution) trials were conducted. These control puffs did not trigger any apparent intracellular Ca²⁺ response (data not shown).

Finally, we also tested the effect of ryanodine, a RyR antagonist. For this purpose we monitored whole-cell [Ca²⁺]_i transients before (Figure 4A, left) and after (Figure 4A, right) application of ryanodine (10 µM). Ryanodine administration led to a significant reduction in Ca²⁺ release, as observed by the decrease in whole-cell [Ca²⁺]_i transients amplitude (Figure 4B; n = 14; p<0.05) and also to significant slowing of whole-cell [Ca²⁺]_i transients frequency (Figure 4C; n = 6; p<0.05). The effect of ryanodine was noted in cardiomyocytes derived from all hiPSC clones and lines studied (Figure S2B) and was dose-dependent, as increasing doses of ryanodine (5, 10, 15, and 20 µM, n = 8) led to a gradual decrease in whole-cell [Ca²⁺]_i transients amplitude in both lines studied (Figures 4D and S2C). Taken together, these data demonstrate that hiPSC-CMs display caffeine-responsive and ryanodine-sensitive SR Ca²⁺ stores capable of unloading Ca²⁺ via RyR-mediated Ca²⁺ release and contributing to whole-cell [Ca²⁺]_i transients.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 4. Ryanodine-sensitive Ca²⁺ stores.

(A) Line-scan tracings of whole-cell [Ca²⁺]_i transients recorded from a representative hIH1 hiPSC-CM under control conditions (left) and in the presence of 10 µM Ryanodine (right). (B–C) Summary of the ryanodine-mediated (10 µM) diminution in whole-cell [Ca²⁺]_i transients amplitude (B) and frequency (C). *p<0.05 when compared to baseline (control) values. (D) Dose-response curve for the effects of ryanodine (5, 10, 15 and 20 µM) on whole-cell [Ca²⁺]_i transients amplitude displayed as the percentage of decrease from baseline values. **p<0.01 when mean absolute values were compared to baseline values. Abbreviations: F/Fo, fluorescence (F) normalized to baseline fluorescence (Fo); sec, seconds.

https://doi.org/10.1371/journal.pone.0018037.g004

SERCA-mediated SR Ca²⁺ uptake is required for whole-cell [Ca²⁺]_i transients

We next tested for the functionality and contribution of another important Ca²⁺-handling protein situated on the SR membrane, SERCA. To test for SERCA functionality in hiPSC-CMs we recorded whole-cell [Ca²⁺]_i transients before (Figure 5A, left) and after (Figure 5A, right panel) application of 10 µM thapsigargin, a specific SERCA inhibitor. Application of 10 µM thapsigargin (15 min incubation) resulted in a decrease in whole-cell [Ca²⁺]_i transients' amplitude in both lines studied (n = 7, Figure 5A, and Figure S3A). The effect of thapsigargin was dose-related with a greater decrease in whole-cell [Ca²⁺]_i transients amplitude (observed at 15 min following drug application) with escalating doses (1, 5, 10 and 20 µM, Figure 5B).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 5. The effects of SERCA inhibition.

(A) Line-scan tracings of whole-cell [Ca²⁺]_i transients in a hIH1 hiPSC-CM under control conditions (left) and after 15 minutes (right) of continuous 10 µM thapsigargin exposure. (B) Dose-response effects of thapsigargin (1, 5, 10 and 20 µM) on [Ca²⁺]_i transients amplitude (displayed as percentage decrease versus baseline values). *p<0.05 and **p<0.01 when mean absolute values were compared to baseline values. (C) Caffeine-induced Ca²⁺ transients (denoted by arrows) under control conditions (upper), after 30 minutes (middle), and 32 minutes (lower) of thapsigargin exposure. Abbreviations: F/Fo, fluorescence (F) normalized to baseline fluorescence (Fo); sec, seconds.

https://doi.org/10.1371/journal.pone.0018037.g005

We have shown above (Figures 3, 4, 5A) that in the case of hiPSC-CMs SR Ca²⁺ release is an important contributor of whole-cell [Ca²⁺]_i transients. Therefore, we wanted to confirm whether the inhibiting effect of thapsigargin on whole-cell [Ca²⁺]_i transients was due to a decrease in SR Ca²⁺ content, as a consequence of SERCA Ca²⁺ uptake inhibition. To this end we conducted repeated SR Ca²⁺ load measurements by applying 20 mM caffeine puffs under control conditions (Figure 5C, upper panel) and in the presence of 10 µM thapsigargin (Figure 5C, middle and lower panels). In contrast, to the enhanced caffeine-induced [Ca²⁺]_i transient observed under control condition (Figure 5C, upper panel), application of a caffeine puff, when complete succession of all whole-cell [Ca²⁺]_i transients has taken place as a result of the thapsigargin uptake inhibition, produced only a minor effect. This effect was displayed as a miniscule caffeine-induced [Ca²⁺]_i transient (Figure 5C, middle panel), that was completely omitted in the subsequent caffeine puff (n = 7; Figure 5C, lower panel). A similar phenomenon was also observed in cardiomyocytes derived from a second hiPSC line (hfib2-5, Figure S3B). The absent caffeine-induced signal at this stage is postulated to be a consequence of the inability of the SR to reload due to SERCA uptake inhibition by thapsigargin.

IP3-mediated calcium release contributes to whole-cell [Ca²⁺]_i transients

IP3-dependent signaling has been shown to play an important role during the process of cardiac development [23], [24]. It was recently demonstrated that in both mouse [25], [26] and human [18], [27] ESC-CMs an IP3-releasable Ca²⁺ pool is expressed and functional. To evaluate the potential role of an IP3-releasable Ca²⁺ pool in hiPSC-CMs, we first utilized immunocytostaining to detect the presence of the IP3R. These stainings displayed positive IP3R immunosignal, strongly distributed around the nucleus (Figure 6A), in a similar fashion to that observed in neonatal rat cardiomyocytes [28], mouse ESC-CMs [26], and hESC-CMs [18].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 6. IP3Rs localization and function.

(A) A hIH1 hiPSC-CM co-stained for pan-IP3R (middle) and sarcomeric α-actinin (left). Merged images are presented in the right panel. (B) Whole-cell [Ca²⁺]_i transients before (left) and after (right) 2-APB (2 µM) application. (C) Whole-cell [Ca²⁺]_i transients before (left) and after (right) U73122 (2 µM) application. (D) Summary of the effects of 2-APB and U73122 on whole-cell [Ca²⁺]_i transients amplitude and frequency. *p<0.05 when compared to baseline values. (E–F) Dose-response curves displaying the effects of 2-APB (0.1, 0.5, 1, 2, 5 and 10 µM) (E) and U73122 (0.5, 1, 2 and 3 µM) (F) on whole-cell [Ca²⁺]_i transients amplitude (displayed as the relative decrease (%) versus baseline values). *p<0.05 and **p<0.01 when mean absolute values were compared to baseline values. Abbreviations: F/Fo, fluorescence (F) normalized to baseline fluorescence (Fo); IP3R, inositol-1,4,5-trisphosphate receptor; sec, seconds.

https://doi.org/10.1371/journal.pone.0018037.g006

To assess the potential contribution of the IP3R to the modulation of whole-cell [Ca²⁺]_i transients in hiPSC-CMs, we tested the effect of IP3R blockade utilizing two different antagonistic approaches. First, whole-cell [Ca²⁺]_i transients were recorded before (Figure 6B, left) and after (Figure 6B, right) application of a low concentration of 2-aminoethoxyphenyl borate (2-APB, 2 µM, n = 14), a well-known cell-permeate IP3R antagonist. 2-APB application resulted in a significant decrease in whole-cell [Ca²⁺]_i transients amplitude (Figures 6B,D) and significantly slowed down whole-cell [Ca²⁺]_i transients frequency (Figure 6B,D). This effect was also dose-related (0.1, 0.5, 1, 2, 5, and 10 µM; n = 8) as displayed in Figure 6E. Strengthening the above-mentioned results, application of 2 µM of U73122 (a phosopholipase C inhibitor, n = 15) led to a significant diminution of whole-cell [Ca²⁺]_i transients amplitude (Figure 6C–D) and a slowing of whole-cell [Ca²⁺]_i transients frequency (Figure 6C–D). This effect was dose-dependent (0.5, 1, 2, and 3 µM; n = 8; Figure 6E). The effects of 2-APB and U73122 were found to be independent of the hiPSC line used (Figure 6 and Figure S4). These observations imply that in hiPSC-CMs an IP3-releasable Ca²⁺ pool is functional and contributes to the modulation of Ca²⁺-handling in these cells.

Whole-cell [Ca²⁺]_i transients in hiPSC-CMs depend on Ca²⁺ entry via L-type Ca²⁺ channels and Ca²⁺ release from RyR-mediated SR Ca²⁺ stores

In adult cardiomyocytes the development of an action potential triggers the opening of L-type Ca²⁺ channels. The resulting Ca²⁺ influx serves as a trigger for activating RyR-mediated SR Ca²⁺ release by a mechanism known as CICR [9], [10]. In addition to the well-established CICR mechanism in adult mammalian cardiomyocytes other mechanistic models were also proposed to be responsible for E-C coupling in various species and at earlier cardiomyocyte developmental stages. These include reports in frog [13] and turtle [11] adult ventricular cells as well as in primary embryonic murine myocytes [12], in which whole-cell [Ca²⁺]_i transients were shown to be derived solely from Ca²⁺ influx through membrane Ca²⁺ channels. In contrast, global whole-cell [Ca²⁺]_i transients in early cardiac cells derived from mouse ESC [14] were reported to be the result of spontaneous Ca²⁺ release from intracellular Ca²⁺ stores without the triggering of membrane Ca²⁺ currents.

The mechanism underlying E-C coupling in hESC-CMs is somewhat controversial. While some reports suggested the absence of a functional SR Ca²⁺ store in hESC-CMs and postulated that essentially all of the [Ca²⁺]_i transients in these cells were a consequence of transsarcolemmal Ca²⁺ influx via membranal Ca²⁺ channels [29], others have argued for a more mature-like CICR mechanism [18], [19], [30]. The latter studies reported the presence of a functional caffeine-responsive and ryanodine-sensitive SR Ca²⁺ store in at least a subset [30] if not all [18], [19], of the cells tested; in a similar manner to the hiPSC-CMs studied in the current study.

Our results support the contribution of both the transsarcolemmal Ca²⁺ influx and intracellular Ca²⁺ store release to whole-cell [Ca²⁺]_i transients in hiPSC-CMs. The importance of the L-type Ca²⁺ current in generating whole-cell [Ca²⁺]_i transients in these cells was manifested by the elimination of these transients in the absence of external Ca²⁺ or in the presence of nifedipine, a selective L-type Ca²⁺ channel antagonist. A similar requirement for external Ca²⁺ and the consequent transsarcolemmal Ca²⁺ influx was documented in adult cardiomyocytes [9], [31], [32], hESC-CMs [19] and mouse ESC-CMs [33], [34].

The contribution of Ca²⁺ release from intracellular SR Ca²⁺ stores to whole-cell [Ca²⁺]_i transients was demonstrated by pharmacological studies interfering either with SR Ca²⁺ release (caffeine and ryanodine) or reuptake (thapsigargin). Caffeine increases RyR2s opening, leading to a single large-amplitude caffeine-induced Ca²⁺ transient, considered to be a descriptive index of the level of SR Ca²⁺ load [35]. In hiPSC-CMs a local pressure ejected puff of caffeine elicited a local bolus release of Ca²⁺, followed by a short and reversible succession of whole-cell [Ca²⁺]_i transients. These results suggest that caffeine induced depletion of the SR Ca²⁺ store and point to a whole-cell [Ca²⁺]_i transient dependency on SR Ca²⁺ content.

The main Ca²⁺ source for the caffeine induced [Ca²⁺]i increase is RyR-mediated SR Ca²⁺ release and is not dominated by Ca²⁺ influx via voltage-gated Ca²⁺ channels. This was demonstrated by the similar caffeine-induced rise in intracellular Ca²⁺ documented in the absence of extracellular bath Ca²⁺ further confirming the presence of a caffeine-responsive intracellular Ca²⁺ store. Similar results were also acquired in hESC-CMs [18], [19].

We also applied the RyR antagonist, ryanodine, to further study hiPSC-CMs RyR-mediated SR Ca²⁺ release. Ryanodine, has been reported to reduce by approximately twofold the conductance of RyRs in the SR [36]. In hiPSC-CMs, ryanodine application led to a dose-dependent diminution in Ca²⁺ release observed as a significant decrease in the amplitude of whole-cell [Ca²⁺]i transients. A similar ryanodine induced effect was also reported in pacemaker cells isolated from rabbit sinoatrial node [37], in ESC-CMs [18], [30] and in mouse ESC-CMs [33]. In addition, in hiPSC-CMs ryanodine application also resulted in slowing of the spontaneous whole-cell [Ca²⁺]i transients firing rate. This phenomenon was also previously documented in rabbit sinoatrial node pacemaker cells where a similar slowing in firing rate was detected in the presence of ryanodine [37].

Functional SERCA pumps enable the loading of SR Ca²⁺ store content required for whole-cell [Ca²⁺]_i transients

For cellular relaxation to take place Ca²⁺ must be removed from the cytosol. In adult cardiomyocytes, one of the main Ca²⁺ removal pathways is the SR Ca²⁺ ATPase pump (SERCA) [9], [38]. These pumps decrease intracellular Ca²⁺, by sequestering Ca²⁺ back into the SR, and in this manner also regulate SR Ca²⁺ load [38]. In adult human cardiomyocytes, SERCA pumping activity is responsible for 70% of Ca²⁺ sequestration from the cytosol back into the SR [9]. To investigate the functionality and contribution of the SERCA pumps to whole-cell [Ca²⁺]i transients through their ability to reload the SR Ca²⁺ stores in hiPSC-CMs we applied the SERCA inhibitor thapsigargin. Thapsigargin acted slowly to progressively decrease the amplitude of whole-cell [Ca²⁺]i transients, eventually leading to their complete inhibition. A similar effect was observed in spontaneously beating fluo-4 loaded isolated mouse ESC-CMs³⁴. An antagonistic effect of thapsigargin on [Ca²⁺]i transients was also reported in human ESC-CMs [19], [30].

The key role of SERCA in reloading the SR, and thereby indirectly modulating hiPSC-CMs whole-cell [Ca²⁺]_i transients, was further demonstrated by the miniscule effect of caffeine in hiPSC-CMs pretreated with thapsigargin, as a result of a pronounced diminution in SR Ca²⁺ content. Interestingly, under conditions of SERCA uptake inhibition (thapsigargin) a low SR Ca²⁺ content was retained yet [Ca²⁺]_i transients were completely abolished. This can be explained by reports showing that decrease in SR Ca²⁺ content can disproportionately inhibit SR Ca²⁺ release [39], which as shown here is an important contributor to hiPSC-CMs whole-cell [Ca²⁺]_i transients. In an immediate subsequent caffeine puff the caffeine-induced [Ca²⁺]i transient was completely omitted. The absent caffeine-induced signal at this stage is postulated to be a consequence of caffeine induced depletion of the SR Ca²⁺ store and the inability of the SR to accumulate Ca²⁺ as a result of the thapsigargin treatment.

IP3R expression, function, and contribution to whole-cell [Ca²⁺]_i transients in hiPSC-CMs

IP3-mediated Ca²⁺ release presents a fundamental pathway for intracellular Ca²⁺ release in electrically non-excitable adult cells [40]. While, in adult cardiomyocytes IP3Rs contribution to cardiac physiology has remained elusive and controversial [9], [41], [42] they have been shown to play an important role during the process of cardiac development [23], [24], [42]. In fact, in the embryo the IP3R is reported to be the first expressed Ca²⁺ release channel [24], [42].

The IP3Rs have been reported to contribute to spontaneous activity in mouse ESC-CMs [25], [26] and are expressed and functional in hESC-CMs [18], [27]. To test for functionality and a potential contribution of an IP3-releasable Ca²⁺ pool to the modulation of Ca²⁺-handling in hiPSC-CMs we first examined the expression and localization of the IP3R at the protein level. Immunostainings of these hiPSC-CMs stained positive for IP3R with a strong subcellular distribution of the immunosignal around the nucleus (Figure 6A) in a similar manner to that observed in hESC-CMs [18], mouse ESC-CMs [26], and neonatal rat cardiomyocytes [28].

Next, to evaluate for IP3-releasable Ca²⁺ pool functionality and participation in the regulation of Ca²⁺-handling in hiPSC-CMs we tested the effect of IP3R blockade utilizing two different antagonistic approaches. First, to block IP3Rs we used the potent cell-permeate inhibitor 2-aminoethoxydiphenyl borate (2-APB) [25]. Application of 2-APB resulted in a significant dose-dependent diminution of whole-cell [Ca²⁺]_i transients amplitude, as was also reported in human ESC-CMs under these conditions [18]. In addition, a slowing of whole-cell [Ca²⁺]_i transients frequency was observed under the influence of 2-APB. Next we applied U73122, a phosopholipase C blocker (PLC). Blocking the activation of PLC inhibits a receptor-stimulated increase in the production of the second messenger IP3 required as a trigger for IP3R-mediated Ca²⁺ release [25], [27]. Superfusion of hiPSC-CMs with U73122 also significantly decreased whole-cell [Ca²⁺]_i transients amplitude and frequency. A U73122 PLC inhibitory effect was also reported in mouse ESC-CMs [25]. These observations imply that an IP3-releasable Ca²⁺ pool is expressed and functional in hiPSC-CMs and that the resulting IP3R-mediated Ca²⁺ release contributes to the modulation of Ca²⁺-handling of these cells.

Potential clinical and research applications

The hiPSC technology has raised significant excitement with regards to its unique potential for regenerative medicine and for the study of various genetic disorders as well as for drug discovery and screening. In the current work we focused on the characterization of the Ca²⁺-handling properties of cardiomyocytes differentiated from hiPSCs and demonstrated that they share components that are present in adult cardiomyocytes, such as functional RyR-mediated SERCA-sequestering SR Ca²⁺ stores. Importantly, the results of this study showing similar properties in cardiomyocytes derived from different differentiation batches, from different hiPSCs clones, and from different hiPSCs lines may have important implications for their potential use for the aforementioned tasks. The hiPSC-CMs may serve as attractive cell-candidates for myocardial cell replacement therapy because of their inherent cardiac-specific properties and the potential for autologous therapy. Nevertheless, since functional compatibility between donor hiPSC-CMs and host myocardium is likely to contribute to an improved functional outcome of the cell engraftment as well as a reduction in potential pro-arrhythmic risk, detailed characterization of their Ca²⁺-handling characteristics is mandatory.

Similarly, phenotypic characterization of Ca²⁺-handling in these cells may be important if one wishes to utilize the iPSC technology for establishment of personalized in-vitro models of cardiac tissue for the development and testing of pharmacological compounds targeting these functional properties. Finally, detailed understanding of the Ca²⁺-handling properties of hiPSC-CMs generated from healthy individuals may be used as a future reference when studying E-C coupling in hiPSC-CMs derived from patients with genetic disorders involving Ca²⁺-handling. Such patient/disease-specific models can be established, for example, from families suffering from catecholinergic polymorphic ventricular tachycardia (CPVT) [43], [44], a potentially lethal disorder resulting from mutations in either the ryanodine receptor or calsequestrin.

Conclusion

In the present work we investigated basic Ca²⁺-handling components of hiPSC-CMs. Our results show that hiPSC-CMs display functional and loaded RyR-regulated intracellular Ca²⁺ stores. These stores can release Ca²⁺ via RyRs and can reload their content through SR Ca²⁺ uptake utilizing functional SERCA pumps. We present evidence showing the expression and functionality of inositol-1,4,5-trisphosphate receptors (IP3Rs). Furthermore, our findings demonstrate that the observed whole-cell [Ca²⁺]_i transients in hiPSC-CMs depend on both sarcolemmal Ca²⁺ entry via L-type Ca²⁺ channels and on intracellular store Ca²⁺ release. Taken together hiPSC-CMs recapitulate functional key Ca²⁺ handling proteins that have been shown to be expressed and functional in mouse ESC-CMs [25], [26], [33], hESC-CMs [18], [19], [30], and adult cardiac tissue [9]. The results of the current study may have important implications for the potential applications of the iPSC technology in basic and translational cardiac research.