Engraftment and characterization of BM cells after direct injection or cytokine mobilization in the infarcted heart, including impact on heart function. (A) Sirius red staining shows extended scar (red) formation in an infarcted heart 4 wk after LCA ligation. (B–D) Engraftment of large numbers of EGFP⁺ cells after cytokine mobilization (B and C) and direct injection of BM cells (D) 4 wk after infarction. (E) Increased engraftment of EGFP⁺ cells in the border zone and the native myocardium after mobilization. (F–H) EGFP⁺ cells were CD45⁺ (Cy5, magenta; F and H) and α-actinin⁻ (Cy3, red; F and G) within the scar (F) and in the border zone (G and H). (I–K) Large (I) and small (J and K) vessels in the infarcted area were PECAM⁺ (Cy3, red) and ASMAC⁺ (Cy5, magenta) but EGFP⁻; some small vessels lacked the smooth muscle layer (J). Note the EGFP⁺ transmigrating cell (K, inset). Nuclei were stained with Hoechst dye (blue). (L–N) Depolarizing current ramps (10 s) evoked action potentials in EGFP⁻ cardiomyocytes (L) but not in EGFP⁺ cells (N, left). (M, inset; and N, right) Voltage ramps (−80 to +50 mV, 250 ms) evoked inward currents in adult (M) and embryonic (E9.5; M, inset) cardiomyocytes but not in EGFP⁺ cells (N, right). Images show transmission and fluorescent images of a patched cell (EGFP, green; CD45, red). (O and P). Statistics of LVEF 3–4 wk after infarction using catheterization. The values are normalized to data obtained from WT C57BL/6 mice (n = 6). Original data are listed in Table S1. Note that mice that underwent cytokine mobilization (P) show reduced LVEF, which was likely caused by irradiation (P = 0.95 [O] and 0.77 [P], respectively). Bars: (A) 900 μm; (B) 550 μm; (C) 250 μm; (D) 300 μm; (E) 400 μm; (F) 15 μm; (G and H) 20 μm; (I–K) 35 μm.

Teratoma formation after injection of undifferentiated ES cells and cut-out beating areas into the infarcted heart. (A) Opened mouse chest in situ. In the apical part of the heart, a large tumor mass was seen 68 d after injecting 10⁵ cells obtained from EGFP⁺-beating areas of EBs. (B) Periodic acid Schiff–stained cross sections of this tumor showed exophytic tumor growth. (C) Mouse heart 31 d after injecting 10⁵ undifferentiated ES cells. Hematoxylin and eosin staining of cross sections in this heart revealed intracardiac tumor distribution. (D) Periodic acid Schiff staining evidenced cartilagineous (closed arrows), squamous (red arrows), and glandular (open arrows) differentiation. (E) Combined desmin (red) and toluidine blue staining proved differentiated cartilage (closed arrows), mucous glandular (open arrows), and red-stained muscle tissue. Bars: (B and C) 250 μm; (D and E) 150 μm.

Puromycin-based selection and enrichment of ES cell–derived cardiomyocytes. (A) Scheme of the bicistronic vector (top). EGFP⁺ areas in EBs at day 9 (left) and fast selection after addition of 10 μg/ml puromycin (middle and right). (B) Puromycin treatment induced a 6–10-fold increase in EGFP⁺ cells (triangles) compared with untreated EBs (circles). Data were obtained from four experiments, and the cell number at day 16 was set for each experiment to 100% for the puromycin group. (C) BrdU labeling (green) and anti-EGFP staining (Cy3, red) in untreated (left) and puromycin-treated (3 d after selection; right) EBs. (D) Statistics of BrdU labeling in untreated (circles) and treated (triangles) EBs. The number of proliferating cells was calculated after 24 h of BrdU incubation at days 0, 3, or 7 of puromycin selection. Data were obtained from two independent experiments. Bars: (A) 250 μm; (C) 50 μm.

Purity of puromycin-selected ES cell–derived cardiomyocytes. (A) Oct-4 expression in undifferentiated ES cells (lane 1), beating EBs with EGFP⁺ areas plated for 9 d (lane 2), and EBs treated for 10 d with puromycin (lane 3). (B) α-Actinin (Cy3, red) staining of puromycin-selected EGFP⁺ cardiomyocytes (green); cell nuclei were stained with Hoechst dye (blue). (C) In 9-d-old EBs, ∼2% cardiomyocytes (EGFP⁺/α-actinin⁺; n = 1334, one representative experiment out of two; left graph) were found. After 10 d of puromycin selection, over 99% of cells were cardiomyocytes (n = 168; right graph). Bars: (B, top) 21 μm; (B, bottom) 12 μm.

Characteristics of puromycin-selected ES cell–derived cardiomyocytes and interaction with fibroblasts. (A) Action potentials in an isolated EGFP⁺ cardiomyocyte exposed for 16 d to puromycin. (B) Action potential duration at 90% repolarization (APD₉₀; left y axis; P < 0.01) and maximal diastolic potential (MDP; right y axis; P < 0.01) at different time points of puromycin exposure (brief exposure [6–10 d], closed bars; long exposure [13–16 d], gray bars). (C) Morphology of EGFP⁺ cardiomyocytes after co-cultivation with embryonic fibroblasts for 6 d (top) and on gelatin-coated dishes for 6 d (bottom). (D) Co-culture of EGFP⁺ cardiomyocytes and embryonic fibroblasts on an MEA for recording field potentials (red traces). The image and the recordings were taken 4 d after starting the co-cultivation. Bars: (C) 100 μm; (D) 400 μm.

Engraftment of puromycin-selected ES cell–derived cardiomyocytes into injured hearts of syngeneic mice, including statistics of tumor incidence. (A) Engraftment of EGFP⁺ cardiomyocytes in an infarcted heart 40 d after injection of 10⁵ ES cell–derived cardiomyocytes and 10⁵ embryonic fibroblasts. (B) Cross section of a heart 60 d after injecting 8 × 10⁴ ES cell–derived cardiomyocytes and 8 × 10⁴ embryonic fibroblasts. (C) α-Actinin staining (middle) of a transplanted, engrafted EGFP⁺ (left) ES cell–derived cardiomyocyte (overlay of the two images, right). (D) Immunostaining for connexin 43 (Cy3, red) showing gap junctions between the transplanted cardiomyocytes (arrows) and nuclear staining (blue) with Hoechst dye. (E) Statistics of tumor incidence. Bars: (B) 120 μm; (C) 17 μm; (D) 29 μm.

LVEF measurements after transplantation of ES cell–derived cardiomyocytes/fibroblasts or only fibroblasts. (A) Representative pressure-volume loops 3–4 wk after injecting 10⁵ ES cell–derived cardiomyocytes and fibroblasts (left) or vehicle (right) into infarcted hearts. (B) Statistics of LVEF (P < 0.01). (C) Sirius red staining of injured myocardium (left) and engraftment of EGFP⁺ fibroblasts into the infarcted heart (right) after injection of 10⁵ embryonic fibroblasts. (D) Statistics of LVEF after injecting fibroblasts or vehicle (P = 0.42). LVEF in B (SV129) and D (C57BL/6) are normalized to data from respective congeneic WT mice. Original data are listed in Table S1. Note the difference in LVEF between the two mouse strains. Bars: (C, left) 250 μm; (C, right) 270 μm.

In vitro transplantation experiments. (A) Protocol of in vitro transplantation experiments. (B) Measurements of force generation in a stimulated (downward deflections on top bar) control slice without transplanted cells (note the different scaling). (C) Slices with ES cell–derived cardiomyocytes generate force spontaneously (left) and after electrical stimulation (6 Hz, right). (D) Statistics of spontaneous and stimulated force generation in slices without (control) and with ES cell–derived cardiomyocytes (P < 0.001).