Membrane remodelling triggers maturation of excitation-contraction coupling in 3D-shaped human-induced pluripotent stem cell-derived cardiomyocytes

Abstract

The prospective use of human-induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CM) for cardiac regenerative medicine strongly depends on the electro-mechanical properties of these cells, especially regarding the Ca²⁺-dependent excitation-contraction (EC) coupling mechanism. Currently, the immature structural and functional features of hiPSC-CM limit the progression towards clinical applications. Here, we show that a specific microarchitecture is essential for functional maturation of hiPSC-CM. Structural remodelling towards a cuboid cell shape and induction of BIN1, a facilitator of membrane invaginations, lead to transverse (t)-tubule-like structures. This transformation brings two Ca²⁺ channels critical for EC coupling in close proximity, the L-type Ca²⁺ channel at the sarcolemma and the ryanodine receptor at the sarcoplasmic reticulum. Consequently, the Ca²⁺-dependent functional interaction of these channels becomes more efficient, leading to improved spatio-temporal synchronisation of Ca²⁺ transients and higher EC coupling gain. Thus, functional maturation of hiPSC-cardiomyocytes by optimised cell microarchitecture needs to be considered for future cardiac regenerative approaches.

Keywords: 3D reshaping; BIN1; Excitation–contraction coupling; Maturation; hiPSC cardiomyocytes; t-tubules.

Fig. 1

BIN1 is successfully expressed in transduced hiPSC-CM. A Schematic timeline of cardiogenic differentiation and subsequent experimental design. B 3D scaffolds in hexagonal (left images) and cuboid (right images) shapes. C Representative images of bright field (left image) and dsRed expression (right image) of hiPSC-CM. D Quantification of mRNA expression levels relative to control, measured by real-time qRT-PCR (N = 4–5, *p < 0.05 for BIN1 vs. CTRL). The list of primers can be found in Table S1. E Immunoblot of BIN1 protein expression in transduced hiPSC-CM. Quantification of the change in protein abundance of BIN1 (N = 3, *p < 0.05). Data are presented as mean ± SE (bar graph) or mean ± SD (box blot), significance tested by Student’s t test

Fig. 2

3D reshaping and BIN1 overexpression lead to adaptations in cell morphology and microarchitecture. Immunolabeling of α-actinin (green, A), BIN1 (green, B), actin (magenta) and DNA (DAPI, blue) in BIN1-expressing, non-patterned, cuboid and hexagonally shaped hiPSC-CM. White boxes indicate the area of magnification. C Representative images of confocal live cell imaging of the sarcolemma using di-8-ANEPPS to visualise the tubular membrane network. White boxes illustrate the skeletonized images of the analysed regions of interest (ROI). D Statistical analysis of the tubule density within the ROIs in different experimental groups of hiPSC-CM. Two-way ANOVA was conducted to examine the synergistic effect of BIN1 expression and 3D reshaping on tubule density. N = 4, n = 19–31 cells; *indicates comparison between different shapes, and # between BIN1-overexpressing and shape control cells; p < 0.05. Data are presented as a box plot and whiskers show SD

Fig. 3

Membrane remodelling reorganises the expression pattern of Ca²⁺ handling proteins and induces dyad formation. A Representative confocal images of the expression pattern of BIN1 (green) and LTCC (magenta) in a BIN1-expressing hiPSC-CM (left image; centre image: magnification of the ROI) and intensity profiles of the depicted ROI in the centre image. B Immunolabelling of RYR2 (green) and LTCC (magenta) and intensity profiles from the ROIs (white boxes) demonstrating spatial alignment of both ion channels relative to each other. C Representative confocal images from PLA: green signals indicate the sites of interaction between RYR2 and LTCC in dyad microdomains (< 40 nm). Nuclei are stained in blue. D PLA signal density was measured as a fraction of cell area occupied by fluorescence signals. Two-way ANOVA was conducted to examine the effect of BIN1 expression and 3D reshaping on LTCC-RyR2 cluster density. N = 4, n = 18–30 cells; * indicates comparison between different shapes; # indicates comparison between BIN1-overexpressing and shape control cells; p < 0.05. Data are presented as a box plot and whiskers show SD

Fig. 4

Characterisation of I_CaL properties in reshaped BIN1-overexpressing hiPSC-CM. A, D, F, I Representative I_CaL traces triggered by the indicated voltage protocols. B Current–voltage relationship, E voltage-dependent steady-state activation and inactivation curves. C Analysis of peak current (I_max, N = 4–5, n = 7–13 cells). G, H Analysis of half-maximal voltage-dependent (V_1/2) activation and inactivation (N = 4–5, n = 6–13 cells). I, J Analysis of the time constant of the Ca²⁺-dependent inactivation of I_CaL (τ1) at + 10 mV fitted with a bi-exponential function (N = 7–8, n = 20–32 cells). Two-way ANOVA revealed no significant differences between the experimental groups. Data are presented as a box plot and whiskers show SD

Fig. 5

Characterisation of Ca²⁺ spark events in reshaped BIN1-overexpressing hiPSC-CM. A Representative line-scan image of a spontaneous Ca²⁺ transient and Ca²⁺ sparks. Analysis of the number of spark events per 50 µm and second (B), full duration at half maximum (FDHM, C) and full width at half maximum (FWHM, D) in the different groups of hiPSC-CM. Statistical differences were tested by two-way ANOVA (N = 3, n = 17–30 cells); *indicates comparison between different shapes; # indicates comparison between BIN1-overexpressing and shape control cells; p < 0.05. Data are presented as a box plot and whiskers show SD

Fig. 6

Maturation of Ca²⁺ transient dynamics in 3D-reshaped BIN1-overexpressing hiPSC-CM. A Representative line-scan images and plot profiles of spontaneous Ca²⁺ transients in the different experimental groups. Analysis of time-to-peak (TTP; B, D), and full duration half maximum (FDHM; C, E) of spontaneous and stimulated Ca²⁺ transients, respectively. Statistical differences were tested by two-way ANOVA (N = 3–4, n = 33–49 cells); *indicates comparison between different shapes; #indicates comparison between BIN1-overexpressing and shape control cells; p < 0.05. Data are presented as a box plot and whiskers show SD

Fig. 7

EC coupling gain is improved in cuboid BIN1-expressing hiPSC-CM. A Voltage protocol for I_CaL stimulation (upper panel), Ca²⁺ transients with corresponding line-scan image and line profile (middle panel) and recordings of I_CaL (lower traces). B EC coupling gain measured at − 25 mV (N = 7–8, n = 19–29 cells) and distribution curves. C Stimulation protocol (upper panel), Ca²⁺ transient (line-scan and line profile, middle panel), and current traces (lower traces) to measure fractional release. Evaluation of (D) SR content, E fractional release and F NCX activity in hiPSC-CM (N = 5–6, n = 9–24 cells). Statistical comparison by two-way ANOVA. Data are presented as a box plot and whiskers show SD. *indicates comparison between different shapes; # indicates comparison between BIN1-overexpressing and control cells; p < 0.05