TALEN-induced mutation in the auts2a gene. A, Zebrafish auts2a gene locus, TALEN target sites and isolated alleles. The TALENs target a pair of binding sites (in blue) flanking a spacer with a restriction enzyme site (in green). Exonic and intronic sequences are shown in upper and lower cases, respectively. In contrast to genomic sequence annotated in the Ensembl (WT), our “in-house” zebrafish AB strain (WT*) has a polymorphism in intronic sequence adjacent to exon 8 (in red). Alleles ncb101, ncb103, ncb104, and ncb105 have the nucleotide deletions that disrupt the donor splice site (in bold) leading to a frameshift after S498 and premature stop codons. Deletion in allele ncb102 does not affect correct splicing and the Auts2a protein sequence. Three single point mutations were introduced in the intron of auts2a^ncb104 allele (in orange) leading to a stop codon creation (underlined). The alternative donor splice sites used to splice mutant auts2a^ncb104 pre-mRNA are highlighted in gray. B, top, Auts2a and Auts2a^ncb104 proteins. The ncb104 mutation causes the loss of the C-terminal portion of Auts2a, comprising PY motif, PR region PR2 and the Auts2 family domain. B, bottom left, RT-PCR analysis of auts2a mRNA, isolated from wild-type (WT), ncb104 heterozygote (HET), and homozygote (HOM) embryos. M, 100-bp DNA ladder (NEB). B, bottom right, Partial protein sequences of mutant alleles. See also Extended Data Figure 1-1.

Morphologic characterization of auts2a mutants. A, Bright field images of wild-type (WT) and mutant (MUT) larvae. Scale bar: 2 mm. B, Whole-mount immunostaining with 3A10 antibody at 30 hpf in auts2a mutants. Arrowhead points to the cell body of the Mauthner neuron and arrow points to the axon. C, Maximum intensity Z-projection of Mauther neuron (top) and homologs (bottom) of wild-type (left) and mutant (right) larvae. Scale bar: 10 μm. D, Comparison of Mauthner lateral dendrite lengths in wild-type and auts2a mutant larvae. E, Comparison of Mauthner soma volume in WT and mutant larvae. n_WT = 8 cells from 7 larvae; n_mut = 15 cells from 12 larvae; *p < 0.05; ns: not significant; Mann–Whitney U test. F, Comparison of soma volume of the M-cell homologs. n_WT = 7 cells from 5 larvae; n_mut = 14 cells from 12 larvae. ns, not significant; Mann–Whitney U test.

Onset of escape response is delayed and highly variable in auts2a mutants. A, Schematic of experimental set up. Escape response was evoked in zebrafish larvae (6–8 dpf) by directing a strong water jet at the OV. Escape response is characterized by a large angle tail deflection, contralateral to the direction of water jet. B, Time lapse of escape response. (1) Prestimulus frame. (2) Water jet makes first contact with OV. (3) First visible tail contraction (marked with asterisk). (4) Representative frame showing references used for maximum tail bend angle calculation. C, Pie chart showing percentage of contralateral, ipsilateral, and no tail bend responses observed across wild type (n = 143 trials), heterozygotes (n = 258 trials), and auts2a mutants (n = 371 trials); χ² test. D, Escape latencies across successive trials from five wild-type and mutant larvae. Color bar represents escape latencies. NR: no response. E, Comparison of escape response latencies in wild type (WT), heterozygotes (HET), and auts2a mutants (MUT). n_WT = 140 trials from 24 larvae, n_HET = 254 trials from 43 larvae and n_MUT = 292 trials from 57 larvae. F, Cumulative density function plot for short-latency escapes (latencies ≤20 ms) in wild type (n = 140 trials), heterozygotes (n = 254 trials), and auts2a mutants (n = 273 trials). G, CV of latencies across successive trials in individual larvae for wild type (n = 24), heterozygotes (n = 43), and mutants (n = 53) groups. H, Comparison of maximum tail bend angle of contralateral turns between the three groups (n = 140 trials, WT; n = 254 trials, HET; n = 281 trials, mutants). Kruskal–Wallis; Mann–Whitney test for between-groups comparisons with Bonferroni correction for multiple comparisons, *p < 0.025, **p < 0.005, ***p < 0.0005; ns, not significant.

Escape response defects in auts2a mutants persist on changing the location of sensory stimulation. A, Schematic of experimental set up. B, Time lapse of escape response evoked by tail stimulation. (1) Prestimulus frame. (2) Water jet makes first contact with the tail. (3) First visible tail contraction (marked with asterisk). (4) Representative frame for maximum tail bend angle calculation. C, Pie chart showing percentage of contralateral tail bends, ipsilateral tail bends, and failures to initiate an escape response between wild type (n = 92 trials) and mutants (n = 113 trials). D, Comparison of escape response latencies on tail stimulation between wild type (n = 71 trials,16 larvae) and auts2a mutants (n = 67 trials, 19 larvae). E, Comparison of CV of latencies across successive trials for each larva between wild-type (n = 17) and mutant (n = 18) groups. F, Maximum tail bend angle of turns for WT (n = 72 trials) and mutants (n = 72 trials); *p < 0.05, **p < 0.005, ***p < 0.0001; ns, not significant.

Mauthner neuron fails to fire reliably in auts2a mutants. A, Schematic representation of experimental set up. Mauthner neuron was retrogradely labeled with OGB-1 dextran and calcium activity was monitored on electrical stimulation (40 μA, 1 ms) of OV. B, left, Raster plot of all trials in WT (n = 68 trials; 10 larvae) showing consistent calcium activity across several trials on OV stimulation. Right, Calcium responses observed across all trials in the mutant group (80 trials; 14 larvae). White line represents the time of stimulus delivery. C, top, ΔF/F profile of a Mauthner neuron in an example wild-type larva across eight trials in response to electrical stimulation of the OV. Bottom, ΔF/F profile of a Mauthner neuron in an example auts2a mutant larva showing subthreshold response as well large calcium transients across eight trials on electrical stimulation of OV. D, Probability of calcium activity response across trials per larva (n_WT = 10 larvae, n_MUT = 14 larvae). E, Peak ΔF/F in WT and mutants (n_WT = 68 trials, n_MUT = 80 trials). F, Comparison of CV of peak ΔF/F between wild-type and mutant larvae. Mann–Whitney U test; **p < 0.01, ***p < 0.0001, ns: not significant.

Mauthner neuron in auts2a mutants have reduced excitability. A, Schematic of experimental set up. Calcium activity in the Mauthner neuron was observed on antidromic stimulation. Mauthner neuron was retrogradely labeled with OGB-1 dextran. B, ΔF/F profile for an example wild-type (WT) larva on antidromic stimulation with 10-μA (left) and 20-μA (right) stimulus intensity. Mauthner neuron fired reliably at the threshold intensity of 20 μA. C, Representative raster plot from a wild-type larva (left) and mutant larva (right). Each row represents average ΔF/F over five trials at the respective stimulus intensity. The threshold for calcium activity for wild-type larva is 20 μA, whereas for the mutant larva is 70 μA. D, Normalized histogram of calcium activity threshold for wild type (n = 9 larvae) and auts2a mutants (n = 12 larvae). E, Summary data of probability of calcium activity at 0.5×, 1×, 1.5× threshold stimulus intensity for wild-type and mutant group. F, ΔF/F profiles for a representative wild-type larva (black) and a mutant larva (red) on antidromic stimulation. Shaded regions represent SEM from five trials. G, Summary data of peak calcium signal in wild type (n = 45 trials, 9 larvae) and auts2a mutants (n = 60 trials, 12 larvae); *p < 0.05, ***p < 0.0005; ns: not significant; Mann–Whitney U test. H, Calcium activity threshold for wild-type (n = 9) and mutant (n = 7) larvae before and after bath application of 50 μm strychnine and 100 μm gabazine. Mauthner neurons were labeled with calcium green dextran for this experiment. Wilcoxon signed-rank test. I, Peak ΔF/F for wild-type (n = 9) and mutant (n = 7) larvae before and after bath application of 50 μm strychnine and 100 μm gabazine. Paired sample t test.

Summary of behavioral abnormalities in escape response in auts2a mutants. Top, In response to threatening stimuli, the ipsilateral Mauthner neuron and its homologs in the hindbrain (marked in a dashed box) fire reliably (yellow) resulting in short latency escape responses across consecutive trials (left to right) in wild-type larvae. Bottom, In auts2a mutants, Mauthner neurons fire unreliably. This means that on some trials, larvae exhibit normal short latency escapes when the Mauthner neuron is able to fire (left). On trials, where the Mauthner fails to fire, long latency escape responses may be initiated perhaps because of the activation of homologs (middle) and if neither the Mauthner, nor the homologs fire, then the larvae fail to respond (right). “?” denotes putative activity in Mauthner homologs during Mauthner-mediated and non-Mauthner-mediated escapes.