Non-genetic Purification of Ventricular Cardiomyocytes from Differentiating Embryonic Stem Cells through Molecular Beacons Targeting IRX-4

Abstract

Isolation of ventricular cardiomyocytes (vCMs) has been challenging due to the lack of specific surface markers. Here we show that vCMs can be purified from differentiating mouse embryonic stem cells (mESCs) using molecular beacons (MBs) targeting specific intracellular mRNAs. We designed MBs (IRX4 MBs) to target mRNA encoding Iroquois homeobox protein 4 (Irx4), a transcription factor specific for vCMs. To purify mESC vCMs, IRX4 MBs were delivered into cardiomyogenically differentiating mESCs, and IRX4 MBs-positive cells were FACS-sorted. We found that, of the cells isolated, ~98% displayed vCM-like action potentials by electrophysiological analyses. These MB-purified vCMs continuously maintained their CM characteristics as verified by spontaneous beating, Ca(2+) transient, and expression of vCM-specific proteins. Our study shows the feasibility of isolating pure vCMs via cell sorting without modifying host genes. The homogeneous and functional ventricular CMs generated via the MB-based method can be useful for disease investigation, drug discovery, and cell-based therapies.

Keywords: IRX4; cell selection; molecular beacons; pluripotent stem cell; ventricular cardiomyocyte.

Graphical abstract

Figure 1

Selection of Optimal Ventricular Cardiomyocyte-Specific IRX4 Molecular Beacons (A) Irx4 mRNA structure was predicted using the RNAfold web server. Three unique target sequences were identified in Irx4 mRNA that maximized the number of predicted unpaired bases as well as the binding affinity of a complementary probe. (B) Stem sequences were appended to the complementary sequence and evaluated using QUIKFOLD to minimize the free energy that causes the oligonucleotide to assume a hairpin structure in solution. (C–E) Flow cytometry results after delivering various IRX4 MBs designed to identify Irx4 mRNAs, or control MB, into mouse embryonic fibroblasts (C), eonatal mouse ventricular CMs (D), and HL-1 CMs (E). The number in each panel represents the percentage of fluorescent cells. FSC indicates forward scatter. All experiments were performed on three independent biological replicates (C–E).

Figure 2

Purification of Ventricular Cardiomyocytes from Differentiating mESCs through IRX4-2 MBs (A) A schematic of the protocol to differentiate mESCs to the cardiac lineage. ESCs, mouse embryonic stem cells; EBs, embryoid bodies. (B) Flow cytometric scattergrams showing the percentages of cells expressing both TNNT2 and MYL2 at differentiation day 18. (C) A flow cytometry plot showing IRX4-2-MB-positive cells at differentiation day 18. (D) Flow cytometric scattergrams showing the percentages of cells expressing both TNNT2 and MYL2 after FACS sorting with IRX4-2 MB. All experiments were performed on three (B and D) or six (C) independent biological replicates.

Figure 3

Electrophysiological Characteristics of IRX4-2-MB-Purified Ventricular Cardiomyocytes (A) Representative action potentials of IRX4-2-MB-positive and -negative cells (upper panel) and primarily isolated mouse fetal ventricular and atrial cardiomyocytes (lower panel). (B) Half action potential duration (APD₅₀) of IRX4-2-MB-positive and -negative cells and mouse fetal ventricular and atrial cardiomyocytes. (C) The percentages of the action potential types recorded from IRX4-2-MB-positive cells. A-like, atrial-like AP; V-like, ventricular-like AP. Action potentials were measured from 50 cells in each group (A–C). (D) Representative spontaneous calcium transients in IRX4-2-MB-positive cardiomyocytes (upper panel), mouse primary fetal ventricular cardiomyocytes (middle panel), and mouse primary fetal atrial cardiomyocytes (lower panel). In each panel, calcium transients were recorded in the upper section, where increasing calcium is indicated by the change in color from dark blue to light blue, and fluorescence intensity was normalized to the baseline measured at time 0 (F_o). (E) The averages of the beating frequency (BPM) and calcium amplitude (F/F₀) of IRX4-2-MB-positive cells and mouse fetal ventricular and atrial cardiomyocytes. Calcium transient experiments were performed on 15 cells in each group (D and E).

Figure 4

Characterization of Purified Ventricular CMs through IRX4-2 MBs (A) Immunocytochemistry for ACTN2, TNNT2, and MYH6/MYH7 on IRX4-2-MB-positive cells isolated from cardiomyogenically differentiated mESCs. Scale bars, 20 μm. (B) Expression of GJA1 determined by immunocytochemistry on IRX4-2-MB-positive cells isolated from cardiomyogenically differentiated mESCs. Scale bars, 50 μm. (C) mRNA expression of cardiac (Myh6/Myh7 and Tnnt2), ventricular (Myl2 and Irx4), atrial (Myl7), and non-cardiac genes (Acta2, Ddr2, and MyoD) in IRX4-2-MB-positive and -negative cells measured by qRT-PCR. y axis represents relative mRNA expression of target genes to GAPDH. ^∗p < 0.05 compared to in IRX4-2-MB-negative cell group. Data are represented as mean ± SEM. All experiments were performed on three independent biological replicates.