(A–B) Partial sequence alignment of klf2a^bns11 (A) and klf2b^bns12 (B) alleles with WT and schematics of their predicted protein products. (C–D) Representative brightfield images of a WT (C) and a klf2a^bns11/bns11; klf2b^bns12/bns12 double mutant (hereafter referred to as klf2 mutant) at 96 hpf (D); lateral views, anterior to the left; arrowhead points to pericardial edema. (E–F’) Maximum intensity projections of 96 hpf klf2 WT (E–E’) and mutant (F–F’) hearts; (G–H’) Two-dimensional (2D) confocal images of 96 hpf klf2 WT (G–G’) and mutant (H–H’) hearts; ventricular outer curvature (dashed boxes) in (E, F, G and H) magnified in (E’, F’, G’ and H’), respectively. (I–J) Three-dimensional reconstructions from confocal images of 96 hpf klf2 mutant ventricular wall. Arrows point to extruding cardiomyocytes; V: ventricle; At: atrium; scale bars: 0.5 mm (C–D), 50 µm (E–H), 20 µm (I–J).

(A) Maximum intensity projections of confocal z-stacks of 96 hpf klf2 WT and mutant hearts, non-injected or injected with tnnt2a MO at the one-cell stage; part of ventricular outer curvature (dashed boxes) magnified on the right side of each panel. (B) Time-lapse 2D confocal images of a klf2 mutant heart during BDM treatment. Arrows point to an extruding cardiomyocyte returning to the compact layer upon inhibiting contraction; V: ventricle; At: atrium; scale bars: 50 µm (A), 10 µm (B).

(A–D) 2D (mid-sagittal sections) (A and C) and maximum intensity projections of confocal z-stacks (B and D) of 82 hpf klf2 WT (A–B) and mutant (C–D) hearts stained with Acridine Orange to visualize cell death; arrows point to extruding cardiomyocytes. (E–H) Confocal images of 96 hpf klf2 WT (E and G) and mutant (F and H) hearts to visualize cardiomyocyte proliferation. (I) Number of mVenus-gmnn positive ventricular and atrial cardiomyocytes in 96 hpf klf2 WT and mutant hearts; dots represent individual hearts; values represent means ±SEM; ***p≤0.001, ns (not significant), by Student’s t-test. (J–K’’) Mid-sagittal confocal sections of 96 hpf klf2 WT (J) and mutant (K) hearts. Higher magnification images of the outer curvature of the ventricular wall (white dashed boxes) in (J) and K) are shown in (J’), (J’’), (K’) and (K’’); arrows point to ectopic accumulation of Cdh2-EGFP proteins on the apical side of cardiomyocytes. (L–M’) 2D confocal views of 96 hpf klf2 WT (L) and mutant (M) hearts. Magnified images of dashed boxes in (L) and (M) are shown in (L’) and (M’), respectively. Arrows point to mislocalized Cdh2-GFP on the apical side of cardiomyocytes; V: ventricle, At: atrium; scale bars, 50 µm.

(A) Schematic representation of the experiment shown in (B–E). (B–E) Transplantation of Tg(myl7:MKATE-CAAX); klf2^+/+ donor cells into Tg(myl7:LIFEACT-GFP); klf2^+/? (B and D) or klf2^-/- (C and E) hosts shown at 96 hpf; white arrows point to klf2^-/- extruding cardiomyocytes in klf2^-/- heart, orange arrows point to klf2^+/+ extruding cardiomyocytes in klf2^-/- hearts; maximum intensity projections of confocal z-stacks of hearts in (B) and (C) are shown in (D) and (E), respectively. (F) Schematic representation of the experiment shown in (G-J). (G–J) Transplantation of Tg(myl7:LIFEACT-GFP); klf2^+/? (G and I) or klf2^-/- (H and J) donor cells into Tg(myl7:MKATE-CAAX); klf2^+/+ hosts shown at 96 hpf; maximum intensity projections of confocal z-stacks of hearts in (G) and (H) are shown in (I) and (J), respectively. V: ventricle, At: atrium; scale bars, 50 µm.

(A–P) Endothelial- and myocardial-specific overexpression of klf2a or klf2b in klf2 WT and mutant hearts. Endothelial overexpression of klf2b (C–D, G–H). Myocardial overexpression of klf2a (K–L, O–P) or klf2b (I–J, M–N); maximum intensity projections of hearts in (A–D) and (I–L) are shown in (E–H) and (M–P), respectively. (Q–R’) Immunostaining of adult klf2 WT (Q–Q’) and rescued mutant (R–R’) heart sections for Caveolin1 to label epicardial cells and phalloidin for overall myocardial structure; magnified images of dashed boxes in (Q) and (R) are shown in (Q’) and (R’), respectively; arrows point to extruding cardiomyocytes; V: ventricle, At: atrium; scale bars: 50 µm (A–P), 300 µm (Q–R’).

(A–D) Confocal images of 96 hpf hearts; WT animals treated with DMSO as a control or FGFR inhibitor (SU5402) from 75 to 96 hpf; maximum intensity projections of hearts in (A) and (D) are shown in (C) and (D) respectively; arrows point to extruding cardiomyocytes. (E–H) 75 hpf Tg(myl7: mCherry-CAAX) (E–F) or Tg(hsp70:dn-fgfr1-EGFP);Tg(myl7: mCherry-CAAX) (G–H) animals were heat-stressed at 39°C for 1 hr (F and H) and their hearts imaged at 96 hpf; arrow in (H) points to an extruding cardiomyocyte (n = 9/13 hearts). (I) klf2a^+/-; klf2b^-/- animals are more likely than WT siblings to exhibit cardiomyocyte extrusion upon Fgfr inhibition; number of treated larvae for each condition is shown above the individual columns. (J–K) Hearts of 96 hpf Tg(myl7: mCherry-CAAX); klf2 ^+/+ or klf2 ^-/- animals immunostained for pERK. (L–M) Hearts of 96 hpf Tg(fli1a:klf2b-p2A-tdTomato);Tg(myl7:EGFP-Hsa.HRAS); klf2 ^+/+ or klf2 ^-/- animals immunostained for pERK. Arrows and arrowheads point to extruding cardiomyocytes and pERK positive endocardial cells, respectively; V: ventricle, At: atrium; scale bars, 50 µm.