Knockdown of pnpla6 leads to distinct morphological change. (A) 36-hpf embryo injected with control MO. (B) 36-hpf pnpla6 morphant displays a curly tail. (C-F) Distinct shorter axis is seen in zebrafish embryos injected with pnpla6 MO: four-somite control MO embryo (C), four-somite pnpla6 MO morphant (D), six-somite control MO embryo (E) and six-somite pnpla6 MO morphant (F). (G,H) Eye development in control (G) and morphant (H) embryo with unclosed eye fissures (black arrow). (I,J) Otic vesicle development in control (I) and morphant (J) embryos. (K,L) Analysis of 25-hpf MHB defects, with the arrow denoting the location of the MHB: lateral view of MHB in control (K) and pnpla6 MO (L) embryos. (M) 72-hpf embryo injected with control MO. (N) 72-hpf pnpla6 morphant displays a swelling of the pericardium. (O) The dose-dependent effect on curly tails; *P<0.01 compared with control MO group. (P) The average distance between head and tail (straight-line length between two arrowheads) of early-stage (four-somite period) zebrafish embryos; *P<0.01 compared with control MO group. Scale bars: 200 μm (A-F); 50 μm (G,H); 100 μm (I,J); 50 μm (K,L); 200 μm (M,N).

Rescue of pnpla6 MO phenotype with human PNPLA6 mRNA. (A) Clustal alignment of partial Pnpla6 protein sequence in human, mouse and zebrafish. The NEST/patatin domain (blue line) is highly conserved. Three key conserved amino acids for the catalytic activation are shown in boxes. (B-G) Rescue of curly-tail phenotype. The abnormal phenotype caused by pnpla6 knockdown can be rescued by co-injection of human PNPLA6 mRNA, but not mutant human mRNA (D960A, S966A or D1086A mRNAs). (H) Quantification of embryos with curly tail; *P<0.001 compared with control MO group. Scale bar: 200 μm.

Knockdown of pnpla6 leads to primary motor neuron defect. (A,B) Lateral views of whole-mount embryos labeled with islet1 mAb. Zebrafish pnpla6 morphants (B) showed fewer motor neurons than control animals (A). (C-F) Lateral views of whole-mount embryos labeled with mAb Znp-1: 36-hpf embryo injected with control MO (C); 36-hpf pnpla6 MO morphant (D) shows truncated (arrow) motor axons, and branching of (arrowhead) motor axons. The defects including truncation and branching can be rescued by human wild-type mRNA (E) but not by mutant mRNA (F). (G,H) Apoptosis analysis of pnpla6 morphants. Zebrafish pnpla6 morphant (H) and control (G) embryos were analyzed by TUNEL assay. (I) Quantification of motor neurons in pnpla6 morphants and controls. The total number of islet1-positive nuclei in the spinal cord spanning ten somites on one side of the spinal cord was counted. Values given are means + s.e.m. and 6–8 individual embryos were included in each group; *P<0.05 compared with control. (J) Quantification of the motor axon defects (truncation or branching) in four groups of embryos: pnpla6 MO, control MO, pnpla6 MO + human wild-type PNPLA6 mRNA and pnpla6 MO + human mutant PNPLA6 mRNA at 26 hpf. #P<0.05 compared with control. Scale bars: 50 μm (A-F); 200 μm (G,H).

Motor neurons of pnpla6 morphants exhibit axon defects in primary culture. (A-D) Primary culture of motor neurons prepared from pnpla6 MO (n=92) or control MO (n=104) were stained with Znp-1 immunofluorescence. (E) The lengths of motor neuron axons were measured. (F) Quantification of the motor neurons with branching defect. *P<0.01 compared with control MO group. Scale bar: 20 μm.

Zebrafish pnpla6 acts cell-autonomously in motor neurons. (A-H) RLDx-labeled cells from embryos injected with control MO or pnpla6 MO were transplanted into wild-type embryos. (A) Homotopic transplantation of control or pnpla6 morphant cells. The cell mass injected is encircled. (B) Number of sides with at least one of the specified motor axon defects. (C,F) Immunostaining of embryos with antibodies to Znp-1 (green). (D,G) View of RLDx-labeled (red) donor cells in recipients. (E,H) In merged images, motor neurons from pnpla6 morphants (H) exhibit aberrant truncation and branching compared with control (E). Scale bars: 200 μm (A,B); 50 μm (C-H).

Zebrafish pnpla6 acts cell-autonomously in motor neurons. (A-H) RLDx-labeled cells from embryos injected with control MO or pnpla6 MO were transplanted into wild-type embryos. (A) Homotopic transplantation of control or pnpla6 morphant cells. The cell mass injected is encircled. (B) Number of sides with at least one of the specified motor axon defects. (C,F) Immunostaining of embryos with antibodies to Znp-1 (green). (D,G) View of RLDx-labeled (red) donor cells in recipients. (E,H) In merged images, motor neurons from pnpla6 morphants (H) exhibit aberrant truncation and branching compared with control (E). Scale bars: 200 μm (A,B); 50 μm (C-H).

Upregulation of BMP signaling in pnpla6 MO morphants. (A) Western blot of phosphorylated Smad1/5/8 levels at 24 hpf in four groups of embryos, as described for Fig. 3J. β-actin was used as a loading control. (B) Dorsomorphin (DM) treatment rescued the curly tail phenotype of pnpla6 MO morphant embryos. Scale bar: 200 μm. (C) The curly tail phenotype of pnpla6 MO was rescued by inhibition of the BMP signal pathway with 6 μmol/l DM; *P<0.001 compared with control MO group.