Adult fish were raised and maintained under standard laboratory conditions. Fish experiments were performed in accordance with institutional guidelines and as approved by the Animal Experimentation Committee of the Royal Netherlands Academy of Arts and Sciences. The cof mutant was identified during a forward genetic screen performed at the Hubrecht Institute, Utrecht, The Netherlands. N-Ethyl-N-nitroso-ureum (ENU) mutagenesis was performed as previously described for the creation of the Hubrecht Institute target selected mutagenesis library ^[15]. F1 progeny of mutagenised male fish were outcrossed to wild-type fish in order to produce approximately 300 F2 families, which were then intercrossed. F3 progeny were screened for epidermal integrity defects at 2–3 dpf. Meiotic mapping of the collapse of fins mutation was performed using standard simple sequence length polymorphisms (SSLP). SSLP primer sequences can be found on www.ensembl.org. Genotyping PCR and subsequent sequencing of the ca5^T839A mutation on finclip DNA or DNA of single embryos was performed with the following primers: F: 5 -cggacagcaagacatctg-3′ and R: 5′-ttgtggatacacatccccatag-3′.

Embryos were raised in egg medium (60 μg/ml sea salt) pH 7. After 24 hpf dechorionated embryos were collected and placed in agarose-coated culture dishes with egg medium or 1x Danieau's medium (58 mM NaCl, 0.7 mM KCl, 0.4 mM MgSO₄, 0.6 mM Ca [NO₃]₂ buffered at different pH with 10 mM Hepes.

Acetazolamide (Sigma) was dissolved in DMSO to a concentration of 0.5 M and diluted to a working concentration of 2.5 mM and 5 mM in egg- or Danieau's medium. Control embryos were treated with the same amount of DMSO solvent.

Whole mount in situ hybridization (ISH) was performed as described previously ^[16]. Embryos for ISH were fixed with 4% PFA/PBS and stored in 100% methanol. After ISH, embryos were cleared in methanol and mounted in benzylbenzoate/benzylalcohol (2:1) before images were taken. The following primers were used to produce the ca5 cDNA fragment: F: 5′-tgcatccaatgtggcaggag-3′; R: 5′-ttgtgtctgactgcaggcaagg-3′ and the insulin cDNA fragment: F: 5′-ttggtcgtgtccagtgtaag-3′; R: 5′-tgcctctcttccttatcagc-3′. Fragments were cloned into the pCRII-TOPO vector (Invitrogen) and antisense dig-labelled probes were synthesised according to standard protocols. Full-length zebrafish ca5 cDNA (MGC:171653; IMAGE:7448163) was derived by PCR on cDNA with the primers: F: 5′-gcgaattcaccatggtcacactgacagccat-3′ and R: 5′-gcctcgagttattccttagaggggg-3′ and cloned into the pCS2+ vector with EcoR1/Xho1. RNA was synthesised in vitro by using the SP6 mMessage mMachine kit (Ambion). The ca5^T839A mutation was introduced using the QuickChange kit (Stratagene).

From a forward genetic screen in zebrafish we derived a mutant allele, collapse of fins (cof) that is characterized by defects of epidermal integrity and collapse of the medial fins at 2 days post-fertilization (dpf) (Figure 1A, B). During later stages of development, cardiac failure with edema and necrosis of the yolk-sac can be observed (Figure 1C, D), eventually leading to the rapid degeneration of the complete embryo at 4 dpf. The cof mutant phenotype is not fully penetrant, only 19% (instead of 25%) of the embryos can be phenotypically identified as a mutant in a batch of cof embryos (see Table 1). Meiotic mapping placed the cof allele on chromosome 25 between markers G39307 and z68140 (Figure 1E). Sequencing the open reading frames of the genes within the corresponding genomic interval revealed a T839A mutation in the coding region of the ca5 gene (Figure 1F). ca5 encodes for the zebrafish orthologue of CA5. The ca5^T839A mutation results in an amino acid substitution of residue M280 to a lysine (Figure 1F). CA5 protein comparison analyses show that M280 is highly conserved across species and other members of the CA protein family (Figure 1G). The zebrafish genome contains only one ca5 gene and comparison of the amino acid sequences reveals 31% identity between zfCA5 and huCA5A, and 40% between zfCA5 and huCA5B. In order to study the ca5 mRNA expression, whole mount in situ hybridization was performed on wild-type embryos at various stages of development. This revealed ca5 mRNA expression in the lens and in a specific part of the embryo that resembles the developing pancreas at 24 hpf (Figure 1H). Previous studies have identified human CA5B in the insulin-producing β-cells of the pancreas ^[17]. To verify the mRNA expression of ca5 in the pancreatic β-cells in zebrafish, we compared ca5 expression with the expression of insulin, a marker for the pancreatic β-cells at 24 hpf. Indeed ca5 mRNA is localized at the same position as the insulin expressing cells (Figure 1H). During later stages of development, ca5 remains expressed in the pancreas (Figure 1H). The expression of insulin mRNA in the ca5^cof mutants was indistinguishable from that in wild-type embryos (Figure 1H), suggesting that β-cell development is not impaired in ca5^cof mutants during development. This was confirmed by determining the level of insulin mRNA expression by PCR on cDNA of wild-type sibling and ca5^cof mutant embryos at 60 hpf (Figure 1H). Although we observed a clear morphological defect in the medial fins of the ca5^cof mutants, ca5 expression could not be detected in the fin epidermis by in situ hybridisation.

To examine whether the ca5^T839A mutation in ca5^cof mutant embryos causes the ca5^cof mutant phenotype, we restored CA5 expression by injecting the full-length zebrafish ca5 RNA. Injecting 100 pg ca5 RNA rescued the ca5^cof mutant phenotype completely, whereas it was not rescued after the injection of 100 pg mutant ca5^T839A RNA (Figure 2A-F and Table 1).

In order to determine whether the ca5^M280K substitution results in a reduced enzymatic activity of the CA5 protein we treated dechorionated wild-type embryos at 24 hpf with acetazolamide (AZA), a general CA inhibitor. Treatment with 5 mM AZA generated essentially a phenocopy of the ca5^cof mutant fish including collapse of the medial fins, cardiac failure and necrosis of the yolk (Figure 2I-L and Table 1), ultimately leading to degeneration of the embryo. We verified the capacity of AZA to inhibit CA5 enzymatic activity by performing synergistic interaction experiments in ca5^cof heterozygous sibling embryos. We treated wild-type embryos and a batch of cof embryos with suboptimal concentrations of AZA. The morphology of wild-type embryos treated with 2.5 mM AZA was not altered (Figure 2M, N and Table 1), however in the cof batch of embryos around 66% of the embryos showed the ca5^cof mutant phenotype (Table 1). Sequencing revealed that a suboptimal dosage of AZA could induce the cof mutant phenotype in heterozygous embryos (Figure 2O, P and Table 1), whereas all homozygous wild-type sibling embryos were not affected. All this shows that inhibition of CA5 activity by AZA treatment during embryonic development can mimic the ca5^cof mutant phenotype. Thus the ca5^T839A missense mutation results in a severe reduction or loss of CA5 activity that initially leads to a collapse of the medial fins, followed by complete degeneration of the embryo.

Carbonic anhydrases are also involved in the regulation of acid-base balance, also in fish ^[18]. Therefore we examined the effect of altered pH levels on the collapse of the medial fins in AZA-treated wild-type embryos and ca5^cof mutant embryos. First, wild-type embryos were raised from 24 hpf onwards in Danieau's medium of pH 5, pH 7.6 or pH 10, containing 5 mM AZA. These experiments show that wild-type embryos are less susceptible to AZA, when cultured in pH 5 medium, compared to embryos cultured in pH 7.6 or pH 10 medium (Figure 3A-L). Furthermore, the ca5^cof mutant phenotype can be rescued by raising mutant embryos in Danieau's medium of pH 5 (see Table1), suggesting that normally the increase in cellular pH during embryonic development is compensated by the activity of mitochondrial CA5 (Table 1). We could not observe any significant developmental defects when wild-type embryos were raised in medium of pH 5 or pH 10 (Table 1). All this shows that CA5 is involved in maintaining cellular acid-base balance during zebrafish embryonic development.