Ankfn1-mutant vestibular defects require loss of both ancestral and derived paralogs for penetrance in zebrafish

Abstract

How and to what degree gene duplication events create regulatory innovation, redundancy, or neofunctionalization remain important questions in animal evolution and comparative genetics. Ankfn1 genes are single copy in most invertebrates, partially duplicated in jawed vertebrates, and only the derived copy retained in most mammals. Null mutations in the single mouse homolog have vestibular and neurological abnormalities. Null mutation of the single Drosophila homolog is typically lethal with severe sensorimotor deficits in rare survivors. The functions and potential redundancy of paralogs in species with two copies are not known. Here, we define a vestibular role for Ankfn1 homologs in zebrafish based on the simultaneous disruption of each locus. Zebrafish with both paralogs disrupted showed vestibular defects and early lethality from swim bladder inflation failure. One intact copy at either locus was sufficient to prevent major phenotypes. Our results show that vertebrate Ankfn1 genes are required for vestibular-related functions, with at least partial redundancy between ancestral and derived paralogs.

Keywords: Ankfn1; Ankfn1-like; Banderuola; nmf9; wide awake; domain architecture; paralogy; penetrance.

Fig. 1.

Conserved features of Ankfn1 homologs. For each histogram a–f) a higher bar indicates higher conservation or constraint at a single residue site for an alignment of Holozoan ANKFN1 proteins. a) Jensen–Shannon divergence (JSD) scores standardized to a range of zero to one. Colored bars above the histogram indicate the positions of the two ANK repeats (ANK1, 2) and FN3 domains, the conserved GLYLGYLK peptide, and the RA domain of ancestral paralogs. A vertebrate-conserved DVLQ peptide (red line) marks the carboxyterminal extent of homology between ancestral and derived paralogs. Gray ovals indicate approximate positions of the three conserved nonmotif regions of Zhang et al. (2015). b) JSD scores at the conserved peptide, y-axis same as (a). c) ScoreCons residue conservation using Valdar’s scoring method. d) ScoreCons scores at the conserved peptide, y-axis same as (c). e) MAPP median score for substitution at each site. f) MAPP scores at the conserved peptide, y-axis same as (e). g) AlphaFold2-predicted structure for mouse ANKFN1 protein, showing positions of the ANK, FN, and GLYLGYLK (yellow highlight) regions and confidence scores. h) Detailed view of the structure in (g) showing the predicted GLYLGY beta strand in relation to flanking strands.

Fig. 2.

Targeted mutagenesis of Ankfn1 homologs. Peptide and genomic sequences for targeted sites in D. rerio Ankfn1 homologs. a) Ankfn1 on chromosome 12 is a derived paralog that includes two ANK and one FN3 domain but no RA domain. b) Ankfn1-like on chromosome 24 is an ancestral paralog including the RA domain. The targeted GLYLGYLK site is highlighted in yellow. CRISPR guide RNA sites are underlined and protospacer adjacent motifs are boxed in each reference sequence. Recovered mutations are shown with deleted base pairs as dashes and insertion of a frame-shifting duplicated sequence (gray shadow) at chromosome 24 shown in a box below the caret.

Fig. 3.

Swim bladder and posture defects in Chr12/Chr24 mutant zebrafish. By 5 days postfertilization mutant animals frequently failed to maintain dorsal orientation while swimming, with unaffected control animals maintaining a dorsal-up posture while double-mutant animals frequently held a rotated posture, with dorsoventral axis at a skew angle or fully to the side (top view). Swim bladders typically failed to inflate in mutant animals (side view, red arrows).