Schematic diagram of vitamin B6 metabolism. B6 vitamers, including pyridoxine (PN), pyridoxal (PL), and pyridoxamine (PM), can be phosphorylated to pyridoxine 5′-phosphate (PNP), pyridoxal 5′-phosphate (PLP), and pyridoxamine 5′-phosphate (PMP) by pyridoxal kinase. The phosphorylated form of B6 vitamers can also be hydrolyzed by pyridoxal phosphatase. The conversion of PNP/PMP to PLP requires an active pyridoxine 5′-phosphate oxidase (PNPO).

Structural and phylogenic comparison of zebrafish Pnpo with enzymes from four different sources. The amino acid sequences of PNPOs from the indicated species were analyzed and compared for functional domains and evolutionary conservation. (A) The peptide sequences of PNPOs from the indicated species were aligned. The shaded letters indicate the amino acids that are identical among compared species. Dots indicate the PNP/PLP binding sites, and arrowheads indicate the FMN binding sites. (B) A phylogenetic table shows the identity between compared enzymes, revealing the evolutionary relationships of PNPOs among the compared species. (C) The crystal structure of experimental hPNPO (Musayev et al., 2003), (D) the simulated structures of hPNPO, and (E) zPnpo are shown in ribbon diagram. (F) Superposed experimental hPNPO crystal structure (grey) and simulated zPnpo (magenta) and hPNPO (cyan) structures. The simulated structures are obtained using I-TASSER (http://zhanglab.ccmb.med.umich.edu/I-TASSER/). The compared peptide sequences include zebrafish Pnpo (NP_001243107.1); human PNPO (NP_060599.1); rabbit PNPO (XP_002719371.1); mouse PNPO (NP_598782.1); and Xenopus Pnpo (NP_001120016.1). PNPO, pyridoxine 5′-phosphate oxidase; PLP, pyridoxal 5′-phosphate.

Spatial and temporal distribution of zebrafish pnpo. The expression profiles of pnpo in zebrafish at different developmental stages and tissues in adult fish were examined by RT-PCR and WISH. Embryos at the (A) indicated stages and (B) tissues of adult female zebrafish were extracted for total mRNA and subjected to RT-PCR with zpnpo-specific primers. The amount of zpnpo transcripts was normalized with zebrafish β-actin. Results presented are the averages shown in mean ± SEM from at least three independent repeats. (C) Embryos at various developmental stages were examined with WISH by probing with the DIG-labeled riboprobes specific to zpnpo as described in the Materials and Methods. RT-PCR, reverse transcription–polymerase chain reaction; WISH, whole-mount in situ hybridization; DIG, digoxigenin.

The impact of knocking down zPnpo. (A) The protein levels of zPnpo in zebrafish embryos injected with 10 ng of scrambled control MO (CTL MO; lane 2), 10 ng of zPnpo MO (MO; lane 3), or zPnpo MO plus 800 pg of zebrafish pnpo mRNA (MO + RNA; lane 4) at 14 hpf and wild-type embryos of the same stage (WT; lane 1) were analyzed with Western blotting (left) and quantified by normalizing with β-actin (right). Shown here are the representative image and average of at least five independent repeats. (B) Dose-dependent decreased survival was observed in zPnpo morphants at 3 dpf. Data presented were the representatives and averages of at least three independent repeats with n = 40∼60 embryos for each group and expressed as mean ± SEM. The statistical significance was calculated with one-tailed Mann–Whitney nonparametric U test by comparing the experimental groups with wild type (*) or with zPnpo MO (#). *^/#p < 0.05; **p < 0.01; ***p < 0.001. MO, morpholino oligonucleotide.

The morphological characteristics of zPnpo morphants. Zebrafish embryos of wild-type and transgenic lines injected with zPnpo MO at one- to two-cell stages were grown in embryo water and recorded for embryonic organogenesis and development. Tg(sox10:eGFP) embryos with/without MO injection were imaged at 18 hpf under dissecting light and fluorescent microscopes (A–D) and also subjected to WISH analysis with riboprobes pax2.1(E–H). Migratory neural crest cells (green fluorescent signals pointed by white arrows) were absent in the closing neural tube (dashed line) of wild-type larvae but remained in that of morphants. Lack or significantly reduced signal for pax2.1 (arrows) were also observed in zPnpo morphants. Both wild-type and zPnpo morphants at 2 dpf were imaged for brain region (I–N). A crater-like opening was observed in diencephalic ventricle (arrows) and hindbrain (circled area) of zPnpo morphants. H&E-stained cryosections prepared from the head region [indicated by vertical lines in (N)] of 2-dpf morphants revealed a cavity (*), which was not observed in wild-type larvae (M’ and N’). Tg(cmlc2:eGFP) embryos with/without MO injection were imaged (O–R) and video-recorded (S and T, the representative still frames from Suppl. 3–10) under a fluorescent microscope for heart development at 3 dpf. A tubing heart with incorrectly positioned ventricle and atrium were apparent in zPnpo morphants. The somite formation (U and V) and the development of trunk and tail (W and X) were recorded at 18 hpf and 3 dpf, respectively. The somites with chevron shape (solid line in U) can be seen in wild-type embryo. The full view of larvae at the indicated stages was imaged from the lateral view to show the overall morphology and size, which also revealed apparent body curvature for zPnpo morphants (Y). For the images to be taken, embryos/larvae of the indicated stages were removed from the embryo water and placed individually on a drop of methylcellulose for photographing. All images were taken with anterior to the left for lateral view except for brain (A–D, E–L) and somites (U–V) (dorsal) and video recording (S and T) (ventral). WT, wild-type; MO, zPnpo morphants; F, forebrain ventricle; M, midbrain ventricle; H, hindbrain ventricle; nt, neural tube; V, ventricle; A, atrium; WISH, whole-mount in situ hybridization; H&E, hematoxylin and eosin.

Rescue of zPnpo morphant morphology. Embryos injected with 5 ng of zPnpo MO at one- to two-cell stage were co-injected with 800 pg of zpnpo or hPNPO mRNA (columns 4 and 5) or grown in embryo water containing 1 mM of PLP, 1 mM of PN, 0.1 mM of PM, or 0.1 mM of PL (columns 6–9). Embryos injected with 10 ng of scrambled control MO were used as control (CTL). Development of various tissues was examined under light dissecting microscope at 3 dpf for brain (A), eyes (B), and heart (C); at 4 dpf for trunk (body curvature) (D); and at 5 dpf for swim bladder (E). The larvae, displaying the morphology comparable with that of wild-type control of the same stage for the characterized tissue/organ, were categorized as normal, and the percent of larvae displaying normal morphology in each group was recorded. (F) The survival rate of zPnpo morphants was recorded at 4 dpf. The averages of data are reported from at least three independent repeats with n = 60∼80 embryos for each group and expressed as mean ± SEM. The statistical significance was calculated with one-tailed Mann–Whitney nonparametric U test by comparing the experimental groups with wild type (*) or with zPnpo MO (#). “*^/#p < 0.05; **^/##p < 0.01. MO, morpholino oligonucleotide; PLP, pyridoxal 5′-phosphate; PN, pyridoxine; PM, pyridoxamine; PL, pyridoxal.

The locomotor activity of zPnpo morphants at 24 hpf. Embryos injected with 5 ng of zPnpo MO at one- to two-cell stages were co-injected with 800 pg of zpnpo/hPNPO mRNA (columns 4 and 5) or grown in embryo water containing 1 mM of PLP, 1 mM of PN, 0.1 mM of PM, 0.1 mM of PL, or 0.5 mM of GABA (columns 6–10) and recorded for larval spontaneous contraction at 24 hpf. Embryos injected with 10 ng of scrambled control MO were used as control (CTL). Reported are the averages of data from at least three independent repeats with n = 80∼120 embryos for each group and expressed as mean ± SEM. The statistical significance was calculated with one-way ANOVA (Kruskal–Wallis test) by comparing the experimental groups with wild type (*) or with zPnpo MO (#). ^##p < 0.01; ***/^###p < 0.001. MO, morpholino oligonucleotide; PLP, pyridoxal 5′-phosphate; PN, pyridoxine; PM, pyridoxamine; PL, pyridoxal; ANOVA, analysis of variance.

The locomotor activity of zPnpo morphants at 4 dpf. Zebrafish larvae with the indicated treatment were placed individually in each well on a 48-well plate and rested in the tested chamber for 5 min before recording. (A) The swimming track, representing larval cumulative location in 1 min, was recorded. (B–E) Larval swimming behavior was recorded for 30 min and analyzed for velocity (B), maximum acceleration (C), body activity (D), and absolute turning angle (E) with the built-in analytic software in DanioVision High-throughput tracking system. Reported are the averages of data from at least three independent repeats with n = 30∼164 embryos for each group and expressed as mean ± SEM. The statistical significance was calculated with one-way ANOVA (Kruskal–Wallis test) by comparing the experimental groups with wild type (*) or with zPnpo MO (#). ^#p < 0.05; ^##p < 0.01; ***/^###p < 0.001. MO, morpholino oligonucleotide; ANOVA, analysis of variance.

Unillustrated author statements