Cfdp1 Is Essential for Cardiac Development and Function

Abstract

Cardiovascular diseases (CVDs) are the prevalent cause of mortality worldwide. A combination of environmental and genetic effectors modulates the risk of developing them. Thus, it is vital to identify candidate genes and elucidate their role in the manifestation of the disease. Large-scale human studies have revealed the implication of Craniofacial Development Protein 1 (CFDP1) in Coronary Artery Disease (CAD). CFDP1 belongs to the evolutionary conserved Bucentaur (BCNT) family, and to date, its function and mechanism of action in Cardiovascular Development are still unclear. We utilized zebrafish to investigate the role of cfdp1 in the developing heart due to the high genomic homology, similarity in heart physiology, and ease of experimental manipulations. We showed that cfdp1 was expressed during development, and we tested two morpholinos and generated a cfdp1 mutant line. The cfdp1^-/- embryos developed arrhythmic hearts and exhibited defective cardiac performance, which led to a lethal phenotype. Findings from both knockdown and knockout experiments showed that abrogation of cfdp1 leads to downregulation of Wnt signaling in embryonic hearts during valve development but without affecting Notch activation in this process. The cfdp1 zebrafish mutant line provides a valuable tool for unveiling the novel mechanism of regulating cardiac physiology and function. cfdp1 is essential for cardiac development, a previously unreported phenotype most likely due to early lethality in mice. The detected phenotype of bradycardia and arrhythmias is an observation with potential clinical relevance for humans carrying heterozygous CFDP1 mutations and their risk of developing CAD.

Keywords: arrythmias; bradycardia; cardiovascular development; cfdp1; coronary artery disease; zebrafish models of human disease.

Figure 1

Expression analysis of cfdp1 in different zebrafish development stages shows that cfdp1 emerges during early development. (A) Whole-mount in situ hybridization of cfdp1 in wild-type zebrafish embryos at different development stages. Higher magnification images of the lateral and ventral plane of the 120 hpf cfdp1 ISH-stained embryos are shown in the lower panels, respectively. Arrows point at the region of the heart. The expression of the gene is apparent from 24 hpf and is restricted at the region of the head and the heart by 120 hpf. n = 30 in each of three independent experiments. Scale bar (upper panels) 150 μm; scale bar (lower panels) 200 μm. (B) Top: Illustration of frontal cutting plane of zebrafish; image panels: paraffin sections of 120 hpf ISH-stained embryos with cfdp1 antisense RNA probe (left) and cfdp1 sense RNA (negative control, right). Arrows point at the stained embryonic heart. Scale bar: 50 μm.

Figure 2

Generation of cfdp1 zebrafish mutant line. (A) Schematic representation of the zebrafish cfdp1 gene. For generation of the CRISPR/Cas9-mediated mutant line, a target site in exon 3 was selected. (B) Schematic representation of the strategy for CRISPR/Cas9-mediated zebrafish line. The injection mix of gRNA/Cas9 was injected at the one-cell-stage embryos. The crispants (F0-injected) grow until adulthood and are crossed with wild-type adults. The F1 generation is genotyped to identify possible founders of the line. After the identification, the corresponding F1 heterozygous generation is kept for further analysis. (C) Upper: nucleotide alignment between cfdp1 mutant and cfdp1 wild-type sequence. A 5 bp deletion in cfdp1 mutant is detected. Lower: chromatogram of Sanger sequencing of cfdp1 mutant and cfdp1 wild-type sequence and the corresponding aa they encode. In the cfdp1 mutant, at the point of DNA break, seven novel amino acids (aa) are inserted before a premature stop codon.

Figure 3

Impaired cardiac performance of cfdp1 mutant embryonic hearts. (A) Stereoscopic images of representative 120 hpf cfdp1-MO-injected (cfdp1 morphants) and uninjected control embryos. Black arrows point to the swim bladder, yellow arrows point to the pericardiac edema, red arrows point to the mouth opening position. Scale bar 150 μm. (A’) Quantification of phenotypic scoring via GraphPad Prism. Data are presented as mean ± SD. n = 40 in each of five independent experiments. (B) cfdp1-MO injections in sibling embryos derived from the cross between heterozygous cfdp1 adult fish. n = 154 in four independent experiments. Black arrows point to the swim bladder, yellow arrows point to the pericardiac edema, red arrows point to the mouth opening position. (B’) Quantification of phenotype scoring of cfdp1 siblings (pool of all three genotypes: cfdp1^−/−, cfdp1^−/+, cfdp1^+/+) and cfdp1-MO-injected cfdp1 sibling embryos. Data are presented as mean ± SD. Scale bar 150 μm. (C) Defective cardiac performance of 120 hpf cfdp1^−/− embryos compared to their siblings cfdp1^+/+ based on ventricular measurements after recording their heart rates. Data are presented as mean ± SD, n = 10–14/genotype. (C’) Brightfield and fluorescent image of cfdp1 mutant embryos utilizing their Tg(myl7:EGFP) (also referred to as cmlc2) background. Dashed lines indicate the long ventricular axis and the short ventricular axis, respectively.

Figure 4

Silencing of cfdp1 expression with a second splice-blocking morpholino. (A) Schematic representation of splicing outcome after microinjection of splice morpholino targeting the exon 2–intron 2 splice junction. The sequenced product shows a frameshift that results in a premature termination codon. (B) Stereoscopic images of representative 72 hpf cfdp1-splice-MO-injected and uninjected control embryos. Black arrows point to the head and heart edema and quantification of phenotypic scoring via GraphPad Prism 9. Data are presented as mean ± SD. n = 50 in each of three independent experiments. Scale bar 150 μm. (C) Max projection of fluorescent images showing that cfdp1 silencing does not affect Notch activity at 72 hpf splice-MO-injected embryos (N = 3, n = 20), while (D) Wnt/β-catenin reporter activity is reduced compared to wild-type control siblings (N = 2, n = 25). AV: Atrioventricular valve; B: bulbus arteriosus; scale bar: 50 μm.

Figure 5

Study of cfdp1^−/+ embryonic and adult hearts. (A) Expression of cfdp1 in cfdp1 siblings (pool of three genotypes: cfdp1^−/−, cfdp1^−/+, cfdp1^+/+). After performing in situ hybridization using the cfdp1 RNA probe in cfdp1 siblings (n = 69) at 120 hpf, ISH signal intensity was quantified. The cfdp1^low and cfdp1^neg ISH signal embryos were genotyped (n = 20) and it was confirmed that they corresponded to cfdp1^−/− and cfdp1^+/- embryos. (B) Brightfield images of freshly isolated whole-mount hearts (upper panels) and brightfield images of 5 μm paraffin-embedded, H&E-stained cardiac slices (lower panels) of 7 months post fertilization (mpf) cfdp1^+/+ (n = 3) and cfdp1^+/- (n = 3) adult hearts. Scale bar (upper panels): 200 μm; scale bar (lower panels): 50 μm. (C) Workflow strategy. Collection of cfdp1 siblings after cross of heterozygous cfdp1 adult fish followed by RNA isolation and genotyping of single embryos. As shown via quantitative real-time PCR, cfdp1 expression levels in cfdp1^−/− embryos are statistically significantly reduced compared to the wild-type cfdp1^+/+ siblings when normalized to act2b as a reference gene. (C) cfdp1^+/- embryos exhibit a range of gene expression (ΔCts plot) while a portion of heterozygous embryos present similar expression levels to the mutant embryo siblings. Student’s t-test (two-tailed distribution, unpaired); ** significantly different p-value < 0.01; ns: non-significant. Data are presented as mean ± SD.

Figure 6

cfdp1 abrogation shows impaired Wnt/β-catenin signaling. (A) Confocal images of cfdp1 morphants show that Wnt/β-catenin reporter activity is diminished compared to the uninjected sibling controls. Max projection of z-stack confocal images of 120 hpf cfdp1-MO embryos. Endothelial cells are labeled with green (Tg(fli1:EGFP)) and Wnt-activated cells are labeled with red (Tg(7xTCF-Xla.Siam:nlsmCherry)); n = 6 in each of three independent experiments. (A’) percentage of phenotypic scoring. AV, atrioventricular valve. B, bulbus arteriosus. Scale bar 150 μm. (B) Confocal images of 120 hpf cfdp1 mutant and wild-type siblings expressing nlsmCherry in Wnt-activated cells and Tg(cmlc2:eGFP) in all cardiomyocytes (Ν = 3, n = 13). Scale bar: 50 μm.

Figure 7

cfdp1 abrogation does not affect Notch signaling. (A) Notch signaling remains unaffected in cfdp1 morphants compared to the uninjected sibling controls. Max projection of z-stack confocal images of 72 hpf cfdp1-MO embryos. Ventricular cardiomyocytes are labeled with green (Tg(myl7:GFP)) and Notch-activated cells are labeled with red (Tg(Tp1:mCherry)). n = 10 in each of three independent experiments. Scale bar 150 μm. (B) Confocal images of 120 hpf cfdp1 mutant and wild-type siblings expressing Tg(TP1:mcherry) in Notch-activated cells and Tg(cmlc2:eGFP) in all cardiomyocytes. (N = 3, n = 12) A: atrium; V: ventricle; B: bulbus arteriosus; scale bar: 50 μm.

Figure 8

cfdp1 is required for proper cardiac trabeculation (A) Schematic representation of retrospective analysis. (B) Single confocal plane of fluorescent phalloidin staining (actin filaments) in 120 hpf cfdp1 embryos, expressing Tg(cmlc2:eGFP) in all cardiomyocytes. Asterisks: trabeculae cardiomyocytes; AV: atrioventricular; OFT: outflow tract. n = 5 in each of three independent experiments Scale bar: 50 μm.