Objectives: Pseudoxanthoma elasticum is an ectopic mineralization disease due to biallelic ABCC6 mutations. Currently, therapeutic options are minimal, necessitating the development of reliable models for high throughput compound screening. One candidate animal model is the zebrafish (Danio rerio), which has two abcc6 gene copies, abcc6a and abcc6b, of which abcc6a is considered the functional orthologue. Interestingly, previously reported morpholino (MO)-induced knockdown and mutant zebrafish (gräte) models for abcc6a show important phenotypic discrepancies. This led to concerns regarding possible aspecificity of the observed structural anomalies in morphants and the size of effect of the missense variant in mutants. To address this, we developed the first complete abcc6a knockout (KO) model using CRISPR/Cas9 and compared its phenotype to that of another naturally occurring mutant model (Sa963) and a splice junction morpholino model.
Methodology: Abcc6a KO: One-cell stage wild type embryos were injected intracellularly with 250 pg Cas9 Nickase and 25 pg abcc6a sgRNA and reared to adulthood. A breeding founder was selected based on the presence of disruptive indels in the germline. For two consecutive generations carriers (c.180delTCGG) were outbred. Incross of carriers generated wild type, heterozygotes and knockouts in F3, which were genotyped and characterized at adulthood. Similarly, F4 zebrafish were characterized until 10 dpf. Abcc6a MO: MO was injected into 1-2 cell stage wild type embryos. Optimal dosage was determined by concentration gradient (1 to 10 ng) of abcc6a or control scrambled MO, with 1.5x dose of p53 MO. Optimal dose was 3 ng/injection. Embryos were characterized until 10 dpf. Abcc6a mutants: Sa963 is a zebrafish line with a 5’ splice site mutation in intron 17 (c.2250+1G>A). Purchased carriers were outcrossed for 2 generations. Incross of carriers generated 3 genotypes (F3), which were genotyped and analyzed at adulthood. F4 offspring were characterized until 10 dpf. Morphological features were analyzed during development and at adulthood. To determine mineralization alizarin red S staining was performed at 10 dpf and at adulthood. Images were analyzed using ImageJ software.
Findings: In the abcc6a KO, we noted advanced mineralization of the vertebrae in larvae which progressed into adulthood. Vertebral bodies were fused with calcified nodular lesions. The ribs had irregular margins, were bifid and fused. As a result, adult fish developed severe scoliosis and were shorter and more fragile compared to WT. A similar phenotype was noted in the MO and Sa963 mutant model. Neither skin mineralization nor other structural anomalies or lethality were seen.
Significance: Our data show that, in contrast to the reported MO data, abcc6a is not essential for embryonic survival or morphological development but rather that it plays an essential role in controlling mineralization, comparable to the mutant model. Our results indicate a direct relation between loss of abcc6a expression and dysregulated osteogenesis. As such, our models recapitulate part of the human phenotype in which ectopic mineralization and pro-osteogenic signaling have been reported. Because of its reproducibility in three models and its ease of quantification, we consider this phenotype to be unequivocally the result of abcc6a deficiency and as such as an excellent readout for drug screening purposes and multiplex mutagene analysis.