Zebrafish fam83fa is expressed in the hatching gland (a) qRT-PCR of fam83fa expression (arbitrary units) in whole embryos across developmental time-series as shown. Data points = one biological replicate, n = 3. Error bars = s.e.m. All data normalized to 18S and calculated using the ∆∆Ct (Livak) method. (b–f) fam83fa expression by whole mount in situ hybridization in stages as labelled. Lateral views (b–f), dorsal (b) or ventral views (c′–f′) and zoomed in regions of centre image (b″–f″) shown. Numbers of embryos are denoted in first image. fam83fa is expressed in the pre-polster, the polster and the hatching gland. Scale bars = 250 μm (except 24 hpf = 500 μm). (g–i) hgg1 expression by whole mount in situ hybridization in stages as labelled. Lateral views (g–i), anterior (g’) or ventral views (h′–i’) and zoomed in regions of centre image (g″–f″) shown. Scale bars = 500 μm.

Generation of fam83fa^−/− zebrafish lines by CRISPR/Cas9 (a) Schematic representation of fam83fa transcripts 201 and 202 showing CRISPR/Cas9 gRNA targeting regions. Dark blue and grey boxes denote exons and untranslated exonic regions, respectively. Zoomed region shows exons 1 and 2 and the gRNA (1–4) target regions (red arrows). Transcript 201 is the presumed primary transcript; however, all gRNAs were designed to promote complete loss of the DUF1669 domain by introducing indels at the 5’ end of either transcript. (b) Schematic of Fam83fa showing locations of the PTCs induced in the fam83fa^-/- KO1 and KO2 lines by combined injection of gRNA1+ gRNA3 (blue triangles). Exons 1 and 2 (a) code for the first 210 aa of Fam83fa, therefore all gRNAs were predicted to induce PTCs in the DUF1669 domain upstream of the F-x-x-x-F sequence motif (for simplicity, only KO1-KO2 are shown). (c) Relative expression of fam83fa in WT and MZ-fam83fa^−/− KO1–KO4 homozygous embryos as labelled, normalized to 18S (arbitrary units). Data points represent biological replicates. Error bars = s.e.m. ****p < 0.00005, one-way ANOVA with Tukey’s post-hoc test. (d) In silico translation of the Sanger sequenced qRT-PCR products amplified in (c). PTCs are denoted by red circles. Alignment performed using MegAlign software (DNASTAR). (e) Relative expression of fam83fb in WT and MZ-fam83fa^−/− KO1–KO4 homozygous embryos as labelled, normalized to 18S. Error bars = s.e.m. **p < 0.005; *p < 0.05, one-way ANOVA with Tukey’s post-hoc test. Data points represent biological replicates. (f) As (e) except showing relative expression of fam83g.

MZ-fam83fa^−/− mutant embryos hatch earlier than WT. (a) Estimated mean hatching times across replicates (n = 3) for WT (red), MZ-fam83fa^−/− KO1 (green) and MZ-fam83fa^−/− KO2 (blue). Log-odds of hatching is plotted over time (hpf); where the curves cross 0.0 on the y-axis, 50% of the embryos have hatched. p < 0.0001, proportional odds logistic regression. (b) Common effect size across replicates shown as predicted mean hatching hour. Note that MZ-fam83fa^−/− KO1 embryos hatch ~10 h earlier than WT. (c) Examples of still images from time-lapse movies used to determine hatching times.

No difference in the p53 response was detected in MZ-fam83fa^−/− embryos compared with WT. (a) Representative western blot for p53 and β-actin (input) in protein extracts from zebrafish embryos treated with IR at 24 hpf and harvested for protein extraction 2 h (t1) and 10 h (t2) after treatment. (b) Quantification of p53 band density for three independent experiments as represented in (a), normalized to β-actin and expressed as arbitrary units. *p < 0.05, two-way ANOVA with Šídák’s multiple comparison test. n.s. = not significant. Error bars = s.d. (c) qRT-PCR of mRNA extracted from WT or MZ-fam83fa^−/− KO embryos as labelled, 2 h following treatment with IR at 24 hpf. qRT-PCR was performed for p53 target genes, including p53 itself, and nrz, the zebrafish bcl−2 homologue. Data represent mean of four biological replicates (data points), each consisting of 25 embryos per line, normalized to 18S and expressed as fold induction relative to untreated WT. (d) Same as (c) except at 10 h following IR treatment. Two-way ANOVA with Tukey’s post-hoc test showed no significant differences between IR-treated WT and MZ-fam83fa^−/− KO for any gene tested. Error bars = s.d.

Transcriptomic analysis suggests degradation pathways are impaired in MZ-fam83fa^−/− embryos (a) Venn diagram showing the DEGs in untreated MZ-fam83fa^−/− KO1 and KO2 embryos compared to WT at t1 (26 hpf). (b) and (c) As (a) except IR treated embryos at t1 (26 hpf—2 h post IR treatment) and t2 (34 hpf—10 h post IR treatment), respectively. (d) Heat map of DEGs common to both KO1 and KO2 (from Venn diagram intersections above) using Euclidean distance to cluster rlog-transformed data, at conditions/times as labelled. Data displayed as normalized variance stabilized count. (e) Volcano plot of fold change in MZ-fam83fa^−/− KO1 and KO2 (x- and y-axes, respectively) of genes with Padj < 0.05, at t1. Vertical and horizontal dashed lines represent no change in KO1 or KO2, respectively, diagonal dashed line represents identical fold change in both KO1 and KO2. Dotted lines represent log2 fold change threshold of ±0.5. Venn diagram in bottom right shows the proportion of genes in each category as shown in the legend. (f) Same as (e) except at t2. (g) MF GO terms for DEGs common to both MZ-fam83fa^−/− KO1 and KO2 at t1 (intersection of Venn diagram in (b)). (h) Same as (g) except CC GO terms. (i) MF GO terms for DEGs common to both MZ-fam83fa^−/− KO1 and KO2 at t2 (intersection of Venn diagram in (c)). (j) Same as (i) except CC GO terms. CC, cellular component; GO, gene ontology; MF, molecular function.

Fam83fa is targeted to the lysosome. (a) Representative confocal images showing HEK293T cells transfected with a plasmid containing tandem fluorescent proteins GFP/mCherry (empty vector, EV), tandem GFP/mCherry fluorescently tagged Fam83fa protein, with or without Bafilomycin A1 (Baf) and tandem GFP/mCherry fluorescently tagged truncated Fam83fa^1-500 protein as labelled. Scale bars = 20 µm. (b) Ratio of mCherry/GFP in all conditions shown in (a). Ratio of each biological repeat (independent transfection) is displayed normalized to empty vector control. Error bars = s.e.m. *¹p = 0.0203 and * ²p = 0.0330, Fisher’s LSD test, n.s. = not significant.