• 
  

• Myocardial H3K27 tri-methylation is required for zebrafish heart regeneration in vivo. (A) Schematics depicting driver and inducible transgenes used for Cre/loxP-mediated myocardial-specific expression of Tg(hsp70l:h3.3) (wild-type control) or Tg(hsp70l:h3.3^K27M)^CM (mutant) during the regenerative window. (B) Schematic depicting the experimental strategy for heart regeneration analyses. Double-transgenic embryos were exposed to 4-HT at 24 h post-fertilization (hpf) and 48 hpf to recombine the mutant transgene specifically in the myocardium. Zebrafish were raised to adulthood and their ventricles resected followed by a 24 h recovery period and daily 1 h heat-shocking for 60 days prior to analysis. (C-F) Sections of uninjured zebrafish ventricles co-immunostained for the cardiomyocyte marker, myosin heavy chain (blue) and H3K27me3 (yellow). Boxed regions in C and E are shown in D and F. Myocardial and non-myocardial cells displayed strong H3K27me3 signals in Tg(hsp70l:h3.3) hearts (C,D; n=5/5 hearts). By contrast, myocardial H3K27me3 signals failed to be detected in Tg(hsp70l:h3.3^K27M)^CM hearts, whereas non-myocardial signals were preserved (E,F; n=5/5 hearts). (G-L) Representative cardiac sections from 60 dpa heat-shocked Tg(hsp70l:h3.3) control (G,J) and Tg(hsp70l:h3.3^K27M)^CM (H,I,K,L) animals evaluated by immunofluorescence for the myocardial marker TPM (green; G-I) or AFOG staining (J-L). Whereas control animals robustly regenerated myocardium (asterisks in G,J; n=9/9 hearts), Tg(hsp70l:h3.3^K27M)^CM hearts failed to regenerate new muscle (arrows in H and I; n=9/11) with either ventricular wall deficits (arrow in K; n=4/11) or apparent collagen deposits indicative of scarring (arrow in L; n=5/11). Scale bars: 50 µm in C,E; 10 µm in D,F; 100 µm in G-L.

  
  

H3K27me3 deposition is required for sarcomere disassembly in wound edge cardiomyocytes. (A) Bar graph showing relative levels of sarcomeric transcripts in Tg(hsp70l:h3.3) and Tg(hsp70l:h3.3^K27M)^CM wound edge samples at 5 dpa, as measured by quantitative PCR (n=3 biological replicates for each cohort). Data are mean±s.d. *P<0.05; **P<0.01; ***P<0.001. (B,C) Section in situ hybridization for ventricular myosin (vmhcl) expression (blue) and antibody staining for myosin heavy chain (red). Control Tg(hsp70l:h3.3) hearts (B; n=6/6), but not Tg(hsp70l:h3.3^K27M)^CM hearts (C; n=5/5), exhibit reduced vmhcl expression in wound edge myocardium at 5 dpa. (D,E) Representative Tg(hsp70l:h3.3) (D) or Tg(hsp70l:h3.3^K27M)^CM (E) cardiac sections co-immunostained for tropomyosin (TPM; blue) and embCMHC (yellow). Tg(hsp70l:h3.3^K27M)^CM hearts showed relatively lower embCMHC signals at the wound edge compared with Tg(hsp70l:h3.3) hearts (n=8/8 and 11/11 hearts, respectively). (F) Bar graph showing the integrated signal densities of embCMHC in arbitrary units in Tg(hsp70l:h3.3) and Tg(hsp70l:h3.3^K27M)^CM hearts at 10 dpa. (G-J) Wound edge regions in cardiac sections immunostained for tropomyosin (TPM) at 10 dpa in Tg(hsp70l:h3.3) or Tg(hsp70l:h3.3^K27M)^CM hearts. Boxed regions in G and I are shown in H and J, respectively. Although wound edge cardiomyocytes show dissociated sarcomeres in Tg(hsp70l:h3.3) animals (n=5/5), sarcomere structure appears preserved in Tg(hsp70l:h3.3^K27M)^CM animals (n=4/6). (K-P) Split channel (K,L,N,O) or merged (M,P) confocal scans of dissociated cardiomyocytes immunostained for tropomyosin (TPM; red) and embryonic cardiac myosin heavy chain (embCMHC; green), and counterstained with DAPI (blue) to visualize sarcomere integrity in injury-responsive cardiomyocytes. (Q) The percentage of embCMHC⁺ cardiomyocytes with partially disassembled sarcomeres as shown in N-P in Tg(hsp70l:h3.3) (n=57 embCMHC+ CMs derived from nine apices) and Tg(hsp70l:h3.3^K27M)^CM (n=77 embCMHC+ CMs derived from nine apices) hearts. Data are mean±s.d. *P<0.05; **P<0.01; ***P<0.001. P-values were calculated using unpaired two-tailed Student's t-test. Scale bars: 100 µm in B-E; 10 µm in G-P.