Cardiac Subtype-Specific Modeling of Kv1.5 Ion Channel Deficiency Using Human Pluripotent Stem Cells

Figures

• 

  Figure 1. Generation of KCNA5 knockout model. (A) Illustration of polycistronic self-replicating RNA used for reprogramming (Yoshioka et al., 2013). Reading frames for OCT4, KLF4, SOX2, GLIS1, and puromycin resistance are separated by 2A peptide-encoding or IRES sequences. (B) Phase contrast morphology (5 x) and alkaline phosphatase staining of the F1 wild-type hiPSC line generated using the RNA vector shown in (A). (C) Representative F1 karyogram indicating a normal male genotype. (D) The RNA vector is undetectable in established F1 hiPSCs following puromycin withdrawal, whereas freshly transfected parental fibroblasts display robust transgene expression (RT-qPCR data, n = 2). (E) Bisulfite sequencing of OCT4 and NANOG promoter regions of the indicated samples showing an hESC-like hypomethylated state in F1 cells. (F) RT-qPCR analysis of F1 hiPSCs in comparison to the original fibroblasts and a hESC control (n = 2). (G) Immunostains showing robust pluripotency factor expression in F1 hiPSCs at the protein level. (H) Illustration of the Kv1.5 channel. The voltage sensor in transmembrane helix S4 is indicated by symbols. The ion pore region is in-between segments S5 and S6. (I) KCNA5…

• 

  Figure 2. Cardiac subtype-specific differentiation of WT and KCNA5fs/fs hiPSCs. (A) Illustration of directed differentiation protocol. Retinoic acid addition (RA, 0.5 μM) during the indicated time interval was used to selectively induce an atrial differentiation fate. (B) Subtype-specific cardiac differentiation showing RA-dependent suppression of ventricular gene expression (green) and induction of atrial markers (red) both in WT and KCNA5fs/fs hiPSCs (n = 2). Note the diminished KCNA5 expression level in RA-treated mutant cardiomyocytes. (C) Immunocytochemical staining of Kv1.5 (top), atrial natruretic peptide, and ventricular-specific myosin light chain 2 (bottom). Note that the MLC2v staining appears relatively weak and diffuse due to the early time-point of analysis (2.5 week after differentiation start).

• 

  Figure 3. Subtype-specific phenotype of KCNA5fs/fs hiPSC-cardiomyocytes. (A) Left: MEA-based quantification of beat-to-beat intervals in ventricular and atrial CMs (n = 3, **p < 0.01). Right panels: Representative MEA traces showing frequency differences in spontaneous beating behavior as dependent on genotype and cardiac subtype. (B) Frequency-corrected field potential durations in ventricular and atrial WT and KCNA5fs/fs hiPSC-CMs. Quantification from three independent experiments (left, ***p < 0.001) and representative MEA traces (right). Horizontal bars indicate QTmax intervals which were more accurately determined by averaging several consecutive signals. (C) Left: Representative action potential traces elicited at 1 Hz in default and RA-treated CMs using the perforated patch clamp technique. Right: Quantification of AP-related parameters from independent measurements and cell batches. *p < 0.05 WT+RA vs. WT and KCNA5fs/fs. #p < 0.05 WT+RA vs. KCNA5fs/fs+RA, **p < 0.05 KCNA5fs/fs+RA vs. WT and KCNA5fs/fs. APD, action potential duration at 20, 50, and 90% or repolarization, RMP, resting membrane potential, APA, maxiumum AP amplitude, Plateau, potential 20 ms after the AP upstroke,…

• 

  Figure 4. Atrial and WT-specific response to a Kv1.5 channel blocker. (A) Quantification of field potential durations in response to 4-aminopyridine (4-AP, 1 mM, n = 3). Note the lack of response in atrial KCNA5fs/fs hiPSC-CMs. **p < 0.01. (B) Representative action potential traces at 1 Hz highlighting a specific 4-AP (50 μM) response exclusively in atrial WT CMs.

• 

  Figure 5. KCNA5fs/fs-specific irregularities in spontaneous beating and action potential firing at low stimulation frequency. (A) Effects of carbachol stimulation on atrial WT and KCNA5fs/fs hiPSC-CMs. Representative MEA recordings (top) and corresponding Poincaré plots (bottom). Carbachol-induced baseline irregularities (top middle panel) or chaotic beating (top right) was exclusive to atrial KCNA5fs/fs CMs and observed in 4 out of 8 experiments. (B) At single-cell level, atrial KCNA5fs/fs CMs were prone to develop EADs at low (0.2 Hz) stimulation frequencies, as observed in 5 out of 12 cells (right). WT controls (left) did not show any action potential irregularities (0 out of 9 cells).