gSAIzGFFM218A directs GFP expression in zebrafish hearts and harbors an insertion in the csrp3 genomic locus. A Mosaic GFP expression in the hearts of heterozygous gSAIzGFFM218A (+ /218A) larvae and juveniles at indicated stages. Scale bars, 50 μm. B GFP in + /218A overlapped with the myocardial marker (cmlc2:mCherry reporter), but neither with the endocardial marker (flk:mCherry reporter) nor the epicardial marker (anti-Pck staining). Scale bars, 20 μm. C Schematic diagram illustrated the components of the gene-trap construct and its insertion in intron 3 of the csrp3 locus, which was predicted to result in a truncated csrp3 transcription. D Sequencing analysis of csrp3 cDNA from wild-type or homozygous 218A larvae. E RT-PCR analysis of csrp3 transcripts from wild-type, + /218A, and 218A larvae using specific primers indicated in C. M, molecular weight markers. F Western blotting revealed a dramatic reduction in Csrp3 protein level in 218A mutants. G Whole-mount in situ hybridization (WISH) showed the endogenous csrp3 expression in wild-type or 218A larval hearts. Scale bars, 50 μm

Csrp3 is involved in trabeculation, and its expression is regulated by hemodynamics and Notch signaling during heart development. A Upper: visualization of trabeculae using Tg(vmhc:mCherry-NTR) displayed increased trabeculation in 218A larval hearts. Scale bars, 20 μm. Bottom: Quantification of the number or density of trabeculae in wild-type and 218A larval hearts at indicated stages. N = 9 each. Data are presented as mean ± SD, Student’s t-test, **, p < 0.01. B Upper: HE staining showed an increased myocardial density in 218A adult hearts. Scale bars, 100 μm. Bottom: Quantification of the myocardial density in wild-type and 218A adult hearts. N = 6 each. Data are presented as mean ± SD, Student’s t-test, ***, p < 0.001. C Comparison of the fluorescence pattern in the hearts of + /218A and Tg(tp1:d2GFP) larvae. Scale bars, 20 μm. D, E WISH and confocal images showed that inhibition of Notch signaling with DATP treatment enhanced both endogenous csrp3 expression in wild-type hearts (D) and GFP fluorescence in + /218A hearts (E). Scale bars, 20 μm. F Confocal and WISH images showed that reducing blood flow via tnnt2a MO injection markedly suppresses both GFP fluorescence in + /218A hearts and endogenous csrp3 expression in wild-type hearts. Scale bars, 20 μm

csrp3 expression increases in response to zebrafish heart injury. A The temporal expression profiles of GFP in + /218A hearts (upper) and endogenous csrp3 in wild-type hearts (bottom) after ventricle ablation. Scale bars, 20 μm. B Immunostaining showed the colocalization of GFP fluorescence with MF20-marked myocardium in + /218A larval hearts after ventricle ablation. Scale bar, 20 μm. C Confocal images illuminated a significant increase of GFP fluorescence in TnnT2(CT3)-marked myocardial cells at the border zone of the cryoinjured + /218A adult ventricle. Areas of dashed boxes are magnified. Scale bars, 100 μm. D Immunostaining revealed that GFP fluorescence of + /218A was specifically expressed in myocardial (Actn1⁺) cells, while absent in the endocardial (Flk⁺) and epicardial (Pck⁺) cells. Areas of dashed boxes are magnified. Scale bars, 100 μm. E Fluorescence in situ hybridization (FISH) analysis corroborated a robust upregulation of csrp3 expression at the border zone of cryoinjured ventricle, compared to the faint and dispersive signal in the uninjured adult ventricle. Scale bars, 100 μm

Csrp3 deficiency reduces CM proliferation and impedes zebrafish heart regeneration. A Representative images of the recovered/ unrecovered heart of 218A mutant larvae at 4 dpa/7 dpf. Scale bars, 20 μm. B Quantification of the regeneration ratio of ablated wild-type, + /218A, and 218A hearts at 4 dpa/7 dpf. The numbers of larvae analyzed are indicated. Binomial test, ****P < 0.0001, NS, non-significant. C Immunostaining of the mitotic marker phospho-histone H3 (pH3) revealed a significant decrease of proliferating CMs in 218A larval hearts at indicated stages. Scale bars, 20 μm. D Quantification of the number of pH3⁺ CMs in ablated wild-type and 218A hearts. N = 18 each. Data are presented as mean ± SD, Student’s t-test, ****, p < 0.0001. E Representative AFOG staining of cryoinjured ventricles from wild-type, + /218A, and 218A fish at 30 dpci. Scale bars, 100 μm. F Quantification of ventricular scar area ratio in wild-type, + /218A, and 218A fish at 30 dpci. N = 7, 6, 7, respectively. Student’s t-test, ***, p < 0.001, NS, non-significant. G Representative confocal images of cryoinjured ventricle sections from wild-type and 218A adult fish at 7 dpci stained with anti-PCNA (green) and anti-Mef2c (red) antibodies. Box areas are amplified. Arrowheads indicate proliferating CMs. Scale bars, 100 μm. H Quantification of CM proliferation index at border zone and injury site of ventricle sections at 7 dpci. N = 6 each. Student’s t-test, ***, p < 0.001

Csrp3 deficiency results in elevated CM apoptosis. A TUNEL assay showed significantly increased apoptosis signals in the 218A larval hearts at the early stages after ablation. Scale bars, 20 μm. B Quantification of TUNEL-positive CMs in the ablated ventricle of wild-type and 218A larval at indicated stages. N = 10 each. Student’s t-test, ****, p < 0.0001. C, D WISH showed the upregulation of apoptosis-related gene tp53 and the downregulation of antiapoptotic gene fosl2 in 218A larval hearts. Scale bars, 50 μm

Csrp3 deficiency impairs injury-induced CM dedifferentiation and sarcomere reassembly. A Immunostaining of cTnT indicated that sarcomere disassembly of 218A CMs was not blocked but instead exacerbated during regeneration while sarcomere reassembly was hindered. Scale bars, 20 μm. B WISH showed reduced expression of cardiogenic factors nkx2.5 and hand2 in ablated 218A larval hearts during regeneration. Scale bars, 50 μm. C Immunostaining showed reduced embryonic Myh7 signal in the border zone of injured 218A adult ventricle at 7 dpci. Scale bars, 100 μm

csrp3 overexpression relieves the inhibitory effects of multiple signaling blockage on zebrafish heart regeneration. A, B Comparative analysis of the expression changes of GFP in ablated + /218A hearts and endogenous csrp3 in ablated wild-type hearts after inhibiting blood flow (tricaine), Notch (DAPT), and ErbB (AG1478) signaling. Scale bars, 20 μm. C Fluorescence pattern of Tg(cmcl2:Csrp3-EGFP) at 4 dpf. Scale bars, 20 μm. D Confocal images showed that Csrp3 overexpression (OE) ameliorated the impairment of heart regeneration resulting from inhibiting blood flow (tricaine), Notch (DAPT), and ErbB (AG1478) signaling. Scale bars, 20 μm. E Quantification of the regeneration ratio of ablated wild-type and Csrp3 overexpressed larvae after treatment with indicated inhibitors at 4 dpa/7 dpf. N = 481, 548, 487, 411, 689, 524, 312, 328, respectively. Data are presented as mean ± SD, Student’s t-test, **P < 0. 01, ***P < 0.001, ****P < 0.0001