Dusp6 attenuates Ras/MAPK signaling to limit zebrafish heart regeneration

Abstract

Zebrafish regenerate cardiac tissue through proliferation of pre-existing cardiomyocytes and neovascularization. Secreted growth factors such as FGFs, IGF, PDGFs and Neuregulin play essential roles in stimulating cardiomyocyte proliferation. These factors activate the Ras/MAPK pathway, which is tightly controlled by the feedback attenuator Dual specificity phosphatase 6 (Dusp6), an ERK phosphatase. Here, we show that suppressing Dusp6 function enhances cardiac regeneration. Inactivation of Dusp6 by small molecules or by gene inactivation increased cardiomyocyte proliferation, coronary angiogenesis, and reduced fibrosis after ventricular resection. Inhibition of Erbb or PDGF receptor signaling suppressed cardiac regeneration in wild-type zebrafish, but had a milder effect on regeneration in dusp6 mutants. Moreover, in rat primary cardiomyocytes, NRG1-stimulated proliferation can be enhanced upon chemical inhibition of Dusp6 with BCI. Our results suggest that Dusp6 attenuates Ras/MAPK signaling during regeneration and that suppressing Dusp6 can enhance cardiac repair.

Keywords: Cardiac repair; Cardiomyocyte proliferation; Dual specificity phosphatase 6; Heart regeneration; Ras/MAPK signaling; Zebrafish.

Fig. 1.

dusp6 is induced upon cardiac injury and expressed in cardiomyocytes and endothelial cells. (A) Q-PCR analysis of dusp6 in adult zebrafish hearts at 1 and 7 days post amputation (dpa), compared with uninjured hearts. dusp6 is induced after ventricle apex amputation. Values are normalized to β-actin and rnap expression. Data represent three independent replicates. *P<0.05, one-way ANOVA. (B) dusp6 expression in uninjured hearts and at 3 dpa using RNAscope. In uninjured hearts, dusp6 (red) is weakly detected throughout the heart including endothelial cells [Tg(fli1a:EGFP)^y1], myocardium (Myo) and compact myocardium (CM). At 3 dpa, dusp6 expression was detected in the myocardium and in endothelial cells. Dotted line demarcates the amputation plane. Sections were counterstained with DAPI (blue) to visualize nuclei. Boxed areas are magnified beneath. (C) Confocal image from Tg(dusp6:memGFP)^pt21 heart immunostained with anti-GFP and anti-Mef2C, showing that cardiomyocytes express memGFP in uninjured hearts. (D) AFOG images (left) and confocal images (center and right) of Tg(dusp6:memGFP)^pt21; Tg(kdrl:NLS:mCherry)^is5 hearts at multiple time points after ventricle apex amputation (n=3 for each time point). memGFP is expressed in the vessels of the compact myocardium in uninjured hearts and in the nascent vessels inside the clot after injury. Sections were counterstained with DAPI (blue) to visualize nuclei. Scale bars: 100 µm.

Fig. 2.

dusp6 mutant hearts exhibit mild cardiomegaly. (A) The TALENs targeting dusp6 exon 1. (B) Recovery of dusp6 mutant alleles pt30a-d. Mismatches created by the 1 bp insertion in pt30d are indicated in red. (C) Predicted amino acid sequence of dusp6^pt30 alleles. (D) Western blot showing absence of Dusp6 protein in dusp6^pt30a/pt30a embryos. β-actin was used as a loading control. (E,F) Whole-mount images of uninjured WT (E) and dusp6^pt30a/pt30a (F) hearts at 5 months of age. dusp6 mutant hearts show cardiomegaly. A, atrium; V, ventricle; BA, bulbus arteriosus. (G) Quantification of the ratio of ventricle area/body weight (VA/BW) in WT (n=5) and dusp6 mutant (n=5) fish. dusp6^pt30a/pt30a fish have a larger VA/BW ratio than WT fish. **P<0.01, Student's t-test. (H,I) AFOG staining of uninjured heart sections at 5 months of age. dusp6 mutant hearts (I) have a thicker compact myocardium (brackets) than WT hearts (H). (J) Quantification of compact myocardium thickness in uninjured hearts at 5 months of age for WT (n=4) and dusp6 mutant (n=5). ****P<0.0001, Student's t-test. (K,L) Sections of uninjured WT (n=4) and dusp6 mutant (n=4) fish with the Tg(fli1a:EGFP)^y1 background to visualize the endothelium. dusp6 mutant hearts (L) have a thicker compact myocardium (red arrows) containing more vessels than WT hearts (K). (M,N) pERK is detected in fli1a:EGFP⁺ vessels of the compact myocardium. dusp6 mutant hearts (n=4) (N) have a thicker compact myocardium with more vessels showing pERK staining than in WT uninjured hearts (n=4) (M). Scale bars: 100 µm.

Fig. 3.

Increased cardiomyocyte proliferation and angiogenesis in dusp6 mutant hearts after cardiac injury. (A) Hearts at 7 dpa, immunostained for Mef2c (green; cardiomyocyte nuclei) and Pcna (red; proliferation marker). Cardiomyocyte proliferation is increased in dusp6 mutant hearts (n=23) compared with WT hearts (n=35). Arrowheads indicate proliferating cardiomyocytes. (B) Quantification of cardiomyocyte proliferation index in WT and dusp6 mutant hearts. ****P<0.0001, Student's t-test. (C) EGFP⁺ vessels were visualized in Tg(fli1a:EGFP)^y1; dusp6^pt30a/pt30a (n=16) and Tg(fli1a:EGFP)^y1 (n=10) hearts at 8 dpa. Dashed line demarcates the resection plane. (D) Quantification of new vessels formed inside the clot area of hearts at 8 dpa. **P<0.01, Student's t-test. Scale bars: 100 µm.

Fig. 4.

dusp6 mutant hearts show accelerated regeneration. (A,B) Sections of hearts at 20 dpa stained with AFOG to visualize the deposition of fibrotic tissue (arrows). dusp6 mutant hearts (n=13) (B) have reduced fibrotic tissue area compared with WT hearts (n=17) (A). (C) Quantitation of fibrotic tissue area at 20 dpa. ***P<0.001, Student's t-test. (D) Representative ECG from WT (n=25) and dusp6 mutants (n=24) at 20 dpa. dusp6 mutant fish exhibit fewer arrhythmic events than WT. ECG are shown from four fish (#1, #10, #11, #13). (E) Quantitation of the arrhythmic events, measured as R-R intervals, in WT and dusp6^pt30a/pt30a zebrafish at 20 dpa. ****P<0.0001, Student's t-test. (F) RNA-seq analysis of uninjured hearts and hearts at 20 dpa. At 20 dpa, WT hearts express fibrosis genes, but in dusp6^pt30a hearts fibrosis genes are downregulated, suggesting faster regeneration. (G) Q-PCR analysis of selected fibrosis genes in hearts at 20 dpa. dusp6^pt30a hearts show downregulation of fibrosis genes. Values are normalized to β-actin and rnap expression. Data are from one representative experiment from three independent biological replicates.

Fig. 5.

Chemical inhibition of Dusp6 increases cardiomyocyte proliferation and angiogenesis and reduces fibrosis after cardiac injury. (A-C) Adult zebrafish hearts at 7 dpa, injected for 6 days with DMSO (n=16) (A), 0.5 mg/kg BCI (n=13) (B) or 0.5 mg/kg BCI215 (n=12) (C) and stained for Mef2c and Pcna. (D) Quantification of proliferating cardiomyocytes at 7 dpa. BCI and BCI215 increased cardiomyocyte proliferation compared with DMSO vehicle. (E-G) Tg(fli1a:EGFP)^y1 hearts at 8 dpa injected for 6 days with BCI (n=16) (F), BCI215 (n=19) (G) or DMSO (n=14) (E) and stained for Mef2c. (H) Quantification of new vessels formed inside the clot area of hearts at 8 dpa. (I-K) Sections of hearts at 25 dpa stained with AFOG to visualize the scar. Intact cardiac muscle stains orange-brown, fibrin stains red and collagen blue. Fish injected with BCI (n=12) (J) and BCI215 (n=18) (K) resolved the injury faster than DMSO-injected fish (n=14) (I). (L) Quantification of clot areas at 25 dpa. (D,H,L) ****P<0.0001, ***P<0.001, **P<0.01, *P<0.05, one-way ANOVA. Scale bars: 100 µm.

Fig. 6.

dusp6 mutant hearts are mildly affected by PDGFR inhibition. (A-D) Hearts at 7 dpa were injected for 6 days with PTKI III (C,D) or vehicle DMSO (A,B) and stained for Mef2c and Pcna to determine cardiomyocyte proliferation. WT hearts (DMSO, n=22; PTKI III, n=19) (A,C) and dusp6^pt30a/pt30a hearts (DMSO, n=21; PTKI III, n=15) (B,D). Arrowheads indicate proliferating cardiomyocytes. (E) Quantification of cardiomyocyte proliferation in WT and dusp6 mutant hearts after injection of PTKI III or DMSO. Each point represents the average proliferation index from one heart. (F-I) AFOG staining of WT or dusp6^pt30a/pt30a hearts at 20 dpa injected for 6 days with PTKI III (WT, n=8; dusp6^pt30a/pt30a, n=12) (H,I) or DMSO (WT, n=10; dusp6^pt30a/pt30a, n=7) (F,G). Dashed line demarcates injury zone. (J) Quantification of fibrotic tissue area in hearts at 20 dpa. (E,J) ****P<0.0001, ***P<0.001, **P<0.01; ns, not significant; one-way ANOVA.

Fig. 7.

EGF receptor inhibition in dusp6 mutants dampens cardiac regeneration. (A) Q-PCR analysis of errfi1b at 0 and 3 dpa. After amputation, dusp6 mutant hearts show higher errfi1b expression compared with WT. Data represent six independent replicates. (B-E) Hearts at 7 dpa injected for 6 days with AG1478 (C,E), or DMSO vehicle (B,D) and stained for Mef2c and Pcna to determine cardiomyocyte proliferation. Arrowheads indicate proliferating cardiomyocytes. For WT: DMSO, n=36; 10 µM and 25 µM AG1478, n=17 and n=14, respectively. For dusp6^pt30a/pt30a: DMSO, n=31; 10 µM and 25 µM AG1478, n=19 and n=13, respectively. (F) Quantification of cardiomyocyte proliferation in WT and dusp6 mutant hearts after injection of AG1478 or DMSO. (G-J) Tg(fli1a:EGFP)^y1 hearts at 8 dpa injected with AG1478 (H,J) or DMSO (G,I). MHC (red) marks the resection plane. AG1478 treatment blocked angiogenesis in WT hearts, but not in dusp6 mutant hearts. Dashed line demarcates resection plane. n=17 for DMSO in WT and dusp6^pt30a/pt30a; n=7 for 25 µM AG1478 in WT and dusp6^pt30a/pt30a. (K) Quantification of new vessels formed at 8 dpa. (L) Quantification of cardiomyocyte proliferation in erbb2st^61/+ hearts at 7 dpa, injected with BCI215 (n=13) or DMSO (n=14) as compared with sibling WT hearts (DMSO, n=17). erbb2st^61/+ hearts have a diminished cardiomyocyte proliferation index compared with WT hearts, and this is rescued by BCI215. (A,F,K,L) ****P<0.0001, ***P<0.001, *P<0.05; ns, not significant; one-way ANOVA. Scale bars: 100 µm.

Fig. 8.

BCI augments NRG1 activity in primary rat cardiomyocytes. Postnatal day 1 rat neonatal ventricular myocytes were isolated and cultured in 10% FBS. Cells were exposed to DMSO vehicle (A), NRG1 (50 ng/ml) (B) or NRG1 plus BCI at 5 µM (C) for 3 days. All treatments were performed in the presence of 10% FBS. Cardiomyocyte proliferation was quantified by H3P staining (n=2). (D) BCI treatment augments the effect of NRG1 on neonatal cardiomyocyte proliferation. ****P<0.0001, ***P<0.001, ANOVA followed by Bonferroni post-hoc test.