Directed Differentiation of Zebrafish Pluripotent Embryonic Cells to Functional Cardiomyocytes

Abstract

A cardiomyocyte differentiation in vitro system from zebrafish embryos remains to be established. Here, we have determined pluripotency window of zebrafish embryos by analyzing their gene-expression patterns of pluripotency factors together with markers of three germ layers, and have found that zebrafish undergoes a very narrow period of pluripotency maintenance from zygotic genome activation to a brief moment after oblong stage. Based on the pluripotency and a combination of appropriate conditions, we established a rapid and efficient method for cardiomyocyte generation in vitro from primary embryonic cells. The induced cardiomyocytes differentiated into functional and specific cardiomyocyte subtypes. Notably, these in vitro generated cardiomyocytes exhibited typical contractile kinetics and electrophysiological features. The system provides a new paradigm of cardiomyocyte differentiation from primary embryonic cells in zebrafish. The technology provides a new platform for the study of heart development and regeneration, in addition to drug discovery, disease modeling, and assessment of cardiotoxic agents.

Keywords: cardiomyocytes; embryogenesis; in vitro differentiation; pluripotency; zebrafish.

Graphical abstract

Figure 1

Pluripotency Timing during Embryogenesis in Zebrafish (A) Representative morphology of early embryos at 256-cell, high, oblong, sphere, dome, 30%-epiboly, 50%-epiboly, and 70%-epiboly stages. hpf, hours post fertilization. Scale bar, 250 μm. (B) Morphology of isolated embryonic cells from different stages. Asymmetric spread of cytoplasm was observed in all stages (lower panel). Scale bar, 50 μm. (C) RT-PCR analysis of gene expression patterns in different stages of the embryos. Primer sequences and PCR conditions are listed in Table S1. NTC, no template control.

Figure 2

Factors and Conditions that Affect Cardiomyocyte Differentiation Efficiency from Primary Embryonic Cells (A) Selected factors affecting BCC (beating cell clusters) generation efficiency at different experimental steps, including stage of embryos, seeding density of embryonic cells, factors supplemented in the medium, and culture-ware coatings. (B) Effect of culture-ware coatings on BCC generation efficiency. Oblong-stage embryonic cells were used for cardiomyocyte differentiation. Coating materials, fibrin gel (FG), poly-L-lysine (PLL, 0.1 mg/mL), gelatin (GEL, 2 mg/mL), ZF4 cells (ZF4, 2 × 10⁴ cells/cm²), and control (None). Experiment repeated once, n = 3 wells of cells/group. (C) Effect of developmental stages of embryonic cells on BCC generation efficiency. The culture plates were coated with gelatin. The dashed line indicates the comparison between 256-cell and 50% epiboly. Five independent experiments, n = 3–20 wells of cells/group. (D) Effect of seeding density of embryonic cells on BCC generation efficiency. Embryonic cells from oblong stage were seeded in gelatin-coated plates at different cell densities during in vitro cardiomyocyte differentiation. Eight independent experiments, n = 3–10 wells of cells/group. In the following investigations (E–G), oblong-stage cells were seeded at a density of 1–2 × 10⁴ cells/cm² in gelatin-coated plates during in vitro cardiomyocyte differentiation. (E) Evaluating effect of supplementation of epidermal growth factor (EGF), zebrafish embryonic extract (ZEE), ZF4 cell-conditioned medium (ZF4 CM), INSULIN (INS), or control (complete medium) on BCC generation efficiency by removing single factor. Experiment repeated once, n = 3 wells of cells/group. (F) Effect of INSULIN supplemental doses on BCC generation efficiency at days 0, 1, and 2 of culture. INSULIN was supplemented at a concentration of 0, 10, 25, or 50 μg/mL. Experiment repeated once, n = 3 wells of cells/group. (G) Effect of INSULIN addition (50 μg/mL) at 0, 1, 2, and 4 hr post seeding cells on BCC generation efficiency. Experiment repeated once, n = 3 wells of cells/group. (H) Numbers of BCCs per embryo at different days of differentiation under optimum conditions from (B–G). Two independent experiments, n = 3–4 wells of cells/group. (I) Representative image of efficient BCC generation at day 2 of differentiation. Dashed lines, beating areas; dots, contracting sites. The image was extracted from a video (Movie S1). Scale bar, 50 μm. (J) EGFP expression driven by myl7 promoter in myl7:EGFP transgenic embryonic cells on day 3 of differentiation. Scale bar, 200 μm. (K–M) Effects of NRG1 on cardiomyocyte proliferation using in vitro cardiac differentiation system in zebrafish. (K) A dose-response evaluation of NRG1 for BCC generation (NRG1 at 0, 50, 100, 200, 500 and 1,000 ng/mL). The linear regression line was y = 0.0297x + 5.6657. Two independent experiments, n = 2 wells of cells/group. (L) Effects of NRG1 treatment (100 ng/mL) on BCC formation on days 2, 3, and 4 of differentiation. Three independent experiments, n = 3–8 wells of cells/group. CTR, 0 ng/mL of NRG1. (M) Proliferative effects of NRG1 on cardiomyocytes. Cell culture was stained with Hoechst 33342 prior to observation under an inverted fluorescent microscope. Numbers of nuclei within each BCC (0 or 100 ng/mL of NRG1 treatment) were recorded on days 2, 3, and 4 of differentiation. Two independent experiments, n = 23–66 BCCs/group. Data are shown as mean ± SEM. ^∗p < 0.05, ^∗∗p < 0.01.

Figure 3

Morphology and Beating Characteristics of BCCs Differentiated from Primary Embryonic Cells (A) Schematic diagram of in vitro differentiation procedures. (B–G) Morphology of BCCs at different days of differentiation when co-culturing with ZF4 cells. Dashed lines, contracting area; arrows, contracting directions; Red dots show the sites that were used for track analysis of BCC contractile kinetics in (H) and (I). Scale bar, 20 μm. (H and I) Contraction and relaxation velocity (H) and contraction amplitudes (I) of BCCs at different days of differentiation. Three contractions of each BCC pointed by red dots in (B–G) were tracked. Data are shown as mean ± SEM. (J) Beating rates of BCCs at day 1–20. Cells were cultured in gelatin-coated plates. Two independent experiments, n = 52–101 BCCs/group. Blue dots show the beating rates of the BCC in (B–G). (K) Gene-expression analysis by RT-PCR during cardiomyocyte differentiation of primary embryonic cells. Primer sequences and PCR conditions are listed in Table S1. NTC, no template control. (L) Cellular and molecular characteristics of Tg (myl7:EGFP) BCCs at different stages of differentiation. Early stage, day 2; middle stage, day 6; late stage, days 9 and 14. Cardiomyocytes marker, Myl7 (green); sarcomeres maker, α-skeletal actin (Acta1, red). Enlarged squares indicate the process of sarcomere formation during cardiomyocyte maturation. Nuclei were stained with DAPI (blue). Scale bars, 50 μm.

Figure 4

Electrophysiological Characterization of BCCs (A) Representative action potential (AP) patterns of three subtypes of cardiomyocytes. AP was recorded by whole-cell patch-clamp method on day 5 of differentiation. (B) Proportions of ventricular-like, nodal-like, and atrial-like BCCs. (C–E) AP amplitude (APA), AP duration at 50% repolarization (APD₅₀), and AP duration at 90% repolarization (APD₉₀) in three types of BCCs. Two independent experiments: ventricular-like, n = 22 BCCs; nodal-like, n = 4 BCCs; atrial-like, n = 2 BCCs. Data are shown as mean ± SEM.

Figure 5

Stimulation Rates Adaptation and Drug Responses of BCCs (A) APs of day-5 ventricular-like BCCs were recorded at 5-ms current pulse stimulation at 1, 2, and 3 Hz. All APs were from the same BCC stimulated at 1, 2, and 3 Hz. (B) Fractional changes of AP durations at 50% and 90% repolarization. The values were normalized to the values observed at 1-Hz stimulation. Experiment repeated once, n = 5 BCCs. (C–J) Effect of adrenergic receptor agonist and Na⁺ channel blocker on spontaneous AP of day-5 BCCs. (C) Representative APs of the BCC before (Pre-EPI) and during (EPI) perfusion of bath solution containing 10 μM adrenergic receptor agonist epinephrine (EPI). (D–F) Fractional changes of beating rates (D), AP durations at 50% repolarization (E), and AP durations at 90% repolarization (F). The values were normalized to the values obtained before EPI applied (Pre-EPI). Experiment repeated once, n = 8 BCCs. (G) Representative APs of the BCC before (Pre-TTX) and during (TTX) perfusion of bath solution containing 10 nM Na⁺ channel blocker tetrodotoxin (TTX). (H–J) Fractional changes of beating rates (H), AP durations at 50% repolarization (I), and AP durations at 90% repolarization (J). The values were normalized to the values obtained before TTX applied (Pre-TTX). Experiment repeated once, n = 4 BCCs. Data are shown as mean ± SEM.

Figure 6

A Working Model for In Vitro Cardiomyocyte Differentiation from Primary Embryonic Cells in Zebrafish A pluripotent window of embryogenesis in zebrafish, from zygotic genome activation to a short moment after the oblong stage (<2 hr), allows us to efficiently induce the oblong-stage cells to differentiate into functional cardiomyocytes under a combination of optimal conditions. The induced cardiomyocytes can differentiate into specific cardiomyocyte subtypes, including ventricular, atrial, and pacemaker cells.