Lateral view of Spen-deficient (B,D) and control-injected embryos (A,C) at 72 h post fertilization (hpf). Spen deficiency results in development of a large pericardial edema at 72 hpf (B,D black arrow), while the controls don't display cardiac abnormalities (A,C). (E) 86% (190/221) MO-spensplice- and 79% (145/184) MO-spenATG- injected embryos display the same cardiac phenotype. Embryos with pericardial edema, bradycardia, and impaired contractility were counted as affected. (F) Efficacy of morpholino-mediated knockdown experiments was confirmed by Western Blot analysis; representative immunoblots are shown. (G) Relative quantification of Spen protein levels (n = 3). Images (A,B,C,D,F) have been cropped. Assessment of significance by chi-squared test (E) or t-test (G). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

(A,B) Fractional shortening of spen morphants and controls at 48 hpf and 60 hpf. Impairment of ventricular fractional shortening in Spen-deficient embryos starts at 48 hpf (12 [10%–15%] vs. 43 [41%–53%], n = 15) and becomes more severe at 60 hpf (3 [1%–5%] vs. 52 [44%–65%], n = 15). (C,D) Heart rate (in beats per minute) is severely reduced in spen morphants compared to control-injected fish at 48 hpf (49 ± 15 vs. 150 ± 15, n = 15) and 60 hpf (58 ± 16 vs. 166 ± 12, n = 15). A–D: Assessment of significance by Wilcoxon-Mann-Whitney test. ****p < 0.0001.

(A) Overview of the heart rhythm phenotype of ten Spen-deficient embryos at different developmental time points. (hpf: hours post fertilization). (B-D) Zebrafish electrocardiograms. While in the controls every atrial excitation (p: p-wave) is followed by a ventricular potential (R: R-wave); (B), spen morphants show atrioventricular conduction abnormalities, as not every p-wave is followed by an R-wave (C). At 72 hpf only an uncoordinated atrial and ventricular electric activity can be documented, resembling atrial and ventricular fibrillation (D). V: Volt; Images (B,C,D) have been cropped.

(A) By screening for dysregulated genes via a cardiovascular disease assay we identified gap junction protein Connexin 43 (Cx43) to be downregulated (supplementary material Fig. A.3A) in Spen-deficient embryos 72 hpf, which was confirmed by cx43-specific qPCR in Spen-deficient hearts (0.47 ± 0.10, p = 0.0063, n = 3) as well as Spen-deficient embryos (0.46 ± 0.04, p = 0.0001, n = 3). (B) Western Blot analysis verifies downregulation of Cx43 in Spen-deficient embryos on the protein level. (C) Quantification of Cx43 levels in Western Blots of Spen-deficient and control embryos (n = 3). Images (B) have been cropped. A, C: Assessment of significance by t-test. *p < 0.05, **p < 0.01, ***p < 0.001.

(A,B) Lateral view of Cx43-deficient (B) and control-injected embryos (A) at 72 hpf. Cx43 deficiency results in development of a large pericardial edema at 72 hpf (B, black arrow) while the controls don't display cardiac abnormalities (A). (C) 47% (47/99) MO-cx43_ATG- injected embryos displayed a heart failure and arrhythmia phenotype. Embryos with pericardial edema, bradycardia, and impaired contractility were counted as affected. (D) Efficacy of morpholino-mediated knockdown experiments was confirmed by Western Blot analysis. Images (A, B) have been cropped. C: Assessment of significance by chi-squared test. ****p < 0.0001.

(A,B) Fractional shortening of connexin43 (cx43) morphants and controls at 48 hpf and 60 hpf. Impaired ventricular fractional shortening in Cx43-deficient embryos starts at 48 hpf (19 [13%–29%] vs. 44 [39%–52%], n = 15) and becomes more severe at 60 hpf (12 [6%–19%] vs. 51 [42%–55%], n = 15). (C,D) Heart rate (in beats per minute) is severely reduced in cx43 morphants compared to control-injected fish at 48 hpf (47 ± 14 vs. 150 ± 16, n = 15) and 60 hpf (48 ± 15 vs. 170 ± 11, n = 15). A-D: Assessment of significance by Wilcoxon-Mann-Whitney test. ****p < 0.0001.

(A) Injection of 500 ng/μg cx43 mRNA in MO-cx43ATG-injected zebrafish leads to 28% (21/78, three independent experiment) embryos displaying heart failure and arrhythmia, while MO-cx43ATG-injection results in 51% (55/108, three independent experiments; p = 0.0024) affected embryos. (C) MO-spensplice injection resulted in 62% (78/126, three independent experiments) affected embryos, while co-injection with either 250 ng/μl or 500 ng/μl cx43 mRNA significantly decreased the amount of diseased embryos to 36% (52/146, three independent experiments, p < 0.0001) and 23% (28/120, three independent experiments, p < 0.0001), respectively. Embryos with pericardial edema, bradycardia, and impaired contractility were counted as affected. (B, D) Western Blot analysis of MO-cx43ATG, MO-cx43ATG + cx43 mRNA and MO1-Ctrl and MO-spensplice + cx43 mRNA injected fish confirms reconstitution of Cx43 levels in cx43- and spen morphants, respectively. (E-G) Lateral views of embryos injected with sub-phenotypic doses of MO-cx43ATG (E), MO-spensplice (F) and a co-injected embryo (G) at 72 h post fertilization (hfp). (H) Sensitizing experiment. While injection of sub-phenotypic doses of morpholino did not cause relevant phenotypic abnormalities 72 hpf, co-injection resulted in a significant supra-additive effect (79/136, three independent experiments; p < 0.0001, respectively), suggesting molecular crosstalk between Spen and Cx43. Images (B,D,E,F,G) have been cropped. A, C, H: Assessment of significance by chi-squared test. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.