p60-katanin knockdown with MOkatna1aug1 morpholino induces similar spinal motor axon defects to MOp60Kat morpholino. (A) Immunolabelling of secondary motoneuron (sMN) axon tracts in 72 h post-fertilisation (hpf) transgenic Tg(Hb9:GFP) larvae injected with control (n = 40) or MOkatna1aug1 (n = 40) morpholinos, using Zn-5 and GFP antibodies. Lateral views of the trunk, anterior to the left. Bottom panels represent higher magnifications of the boxed region in the corresponding top panels. Dotted lines delineate lateral myosepta. Full arrowheads and full arrows point at normal rostral and dorsal nerves, respectively. Empty arrowheads show misguided rostral nerves. Asterisks indicate ectopic sorting points of sMN axons from the spinal cord. (B, C) Quantifications of sMN defects in larvae analysed in panel A and pooled from three independent experiments. Mean number of missing dorsal nerves (B) and misguided rostral nerves (C) per larva. Non-blind quantifications were performed on 24 spinal hemisegments located around the yolk tube per larva. Violin Plots; horizontal bars indicate the median ± the 1st and 3rd quartiles. Mann–Whitney test. P values are displayed on graphs. Source data are available online for this figure.

Molecular and morphological characterisation of zebrafish katna1 mutants. (A) Sequence analysis of control and katna1 mutant genomic DNA. Dotted line indicates the junction between exon 4 (pink background) and intron 4 (white background). The green arrowhead points at the nucleotide substitution (G > A) affecting the donor splice site of intron 4 (G in control, black arrowhead) in katna1 mutants. (B) Schematic representation of the RT-PCR strategy used to test the impact of the katna1 splice-site mutation on katna1 mRNA splicing. Dotted lines indicate intron splicing. Arrows represent the primers used for RT-PCR analysis. Primers were designed on exon/intron junctions to avoid contamination by genomic DNA amplification. In: intron; Ex: exon. (C) RT-PCR analysis of katna1 intron-4 splicing on total RNA extracts from katna1+/+ and katna1-/- MZ maternal zygotic mutant embryos. Homozygous katna1-/- MZ embryos lack wild-type transcript and show different populations of misspliced transcripts (1, 2 and 3). (D) Sequence analysis of katna1 misspliced transcripts. Sequences corresponding to exon 4, intron 4 and exon 5 are respectively indicated in pink, white and blue. The splice-site mutation is highlighted in green. Misspliced transcripts include various sized insertions of intron 4, which all lead to a frameshift and the occurrence of a premature stop codon at the same amino-acid position (red asterisk). (E) Gross morphology of 72-hpf control (katna1+/+) and maternal zygotic katna1 mutant (katna1-/-MZ) larvae. Arrows point at the curved-tail phenotype of some mutant larvae. Scale bar: 0.5 mm. (F) Percentage of larvae exhibiting a curved-tail phenotype. Chi2 test. (G) RT-qPCR analysis showing the differential expression levels of p60-katanin-related genes from the ATPase Meiotic Clade in control and katna1-/- MZ larvae. RNAs were extracted from 5 independent pools of 10 katna1+/+, 10 katna1-/-MZ and one pool of 10 wild-type embryos at 24 hpf. Unpaired t test. Values are shown as means ± SEM. (F, G) P values are displayed on graphs. Source data are available online for this figure.

p60-katanin and spastin transcripts are both expressed in developing spinal motor neurons. (A) Whole-mount in situ hybridisation with p60-katanin sense (upper panel) and antisense (bottom panel) riboprobes at 18 somites and 24 h post-fertilisation (hpf). Lateral views of the embryos, anterior to the left. P60-katanin is highly enriched in the developing nervous system at both 18 somites and 24 hpf, two stages at which the axons of primary (pMN) and secondary (sMN) motor neurons exit the spinal cord to navigate towards their muscle targets. Scale bars: 200 µm. (B) In toto hybridisation chain reaction (HCR) on 72-hpf Tg(Hb9:GFP) larvae using zebrafish spastin and katna1 probes. Lateral view of the trunk, anterior to the left. Right panels are higher magnifications of boxed region of the corresponding left panel. Arrows point at spinal motor neurons co-expressing katna1 and spastin transcripts. Scale bars: 20 µm. Source data are available online for this figure.

Overexpression of a catalytic-dead variant of TTLL6 or TTLL11 respectively fails to rescue the axon pathfinding errors associated with p60-Katanin or Spastin partial knockdown. (A) Immunolabelling of sMN axon tracts using Zn-5 and GFP antibodies in 72-hpf Tg(Hb9:GFP) larvae injected with MOCtl (n = 29), MOp60Kat/1.3pmol (n = 21) or co-injected with MOp60Kat/1.3pmol and the mRNA encoding a catalytic-dead variant of TTLL6 (TTLL6dead) (n = 23). Full and empty arrows respectively point at normal and misguided dorsal projections. (B) Mean number of split/misguided dorsal nerves per larva. Quantifications were conducted on the larval set analysed in (A). (C) Immunolabelling of sMN axon tracts using Zn-5 and GFP antibodies in 72-hpf Tg(Hb9:GFP) larvae injected with MOCtl (n = 27), MOspATG (n = 27) or co-injected with MOspATG and the transcript encoding a catalytic-dead variant of TTLL11 (TTLL11dead) (n = 34). Full and empty arrowheads indicate normal and misguided rostral projections, respectively. (D) Mean number of misguided rostral nerves per larva. Quantifications were conducted on the larval set analysed in (B). (A, C) Scale bars: 50 μm. (B, D) Non-blind quantifications were performed on 24 spinal hemisegments located around the yolk tube per larva. Analysed larvae were pooled from three independent experiments. Violin Plots; horizontal bars indicate the median ± the 1st and 3rd quartiles. Kruskal–Wallis ANOVA test with Dunn’s post hoc test. P values are displayed on graphs. Source data are available online for this figure.

TTLL6 also tunes p60-Katanin-driven pMN axon development. (A) Immunolabelling of pMN axons in 26-hpf embryos injected with MOCtl (n = 32), MOp60Kat/1.3pmol (n = 32) or co-injected with MOp60Kat/1.3pmol and 120 pg of human KATNA1 transcripts (MOp60Kat/1.3pmol + mRNAKATNA1, n = 32) using Znp-1 antibody. (B) Mean number of pMN axon branches per embryo analysed in the panel-A embryo set. (C) Immunodetection of pMN axons with Znp-1 antibody in 26-hpf embryos injected with MOCtl (n = 19), MOTTLL6 (n = 22), MOTTLL11 (n = 19) morpholinos or co-injected with MOTTLL6 or MOTTLL11 and mouse TTLL6 or TTLL11 mRNA (MOTTLL6 + mRNATTLL6, n = 20; MOTTLL6 + mRNATTLL11, n = 21; MOTTLL11 + mRNATTLL11, n = 19; MOTTLL11 + mRNATTLL6, n = 18). (D, E) Mean number of CaP pMN branches (D) and truncated CaP axons per embryo analysed in the panel-C embryo set. (F) Immunostaining of pMN axons with Znp-1 antibody in 26-hpf embryos injected with MOCtl, (n = 20), MOp60Kat/1.3pmol (n = 21) or co-injected with MOp60Kat/1.3pmol and mouse TTLL6 (MOp60Kat/1.3pmol + mRNATTLL6, n = 19) or TTLL11 mRNA (MOp60Kat/1.3pmol + mRNATTLL11, n = 16). (G) Mean number of CaP pMN branches per embryo analysed in the panel-F embryo set. (A, C, F) Arrows and asterisk respectively indicate hyper-branched and truncated ventrally projecting pMN CaP axons. Empty arrowheads indicate normal dorsally projecting pMN MiP axons. Scale bars: 25 μm. (B, D, E, G) Non-blind quantifications were performed on 24 spinal hemisegments located around the yolk tube per embryo. Analysed larvae were pooled from three independent experiments. Violin Plots; horizontal bars indicate the median ± the 1st and 3rd quartiles. One-way ANOVA test with Bonferroni’s post test (D, G) or Kruskal–Wallis ANOVA test with Dunn’s post test (B, E). P values are displayed on graphs. Source data are available online for this figure.

Mouse Sp + /- cortical neurons exhibit a significant number of axonal swellings. (A) Mouse Sp + /+, Sp + /- and Sp-/- cultured cortical neurons immunolabelled with a βIII-tubulin antibody at different days in vitro (DIV). Pink arrows point at axonal swellings. Scale bars: 50 µm. (B, C) Mean number of axonal swellings per 100 nuclei. At least 2500 neurons from two independent experiments were analysed in unblind manner per condition. Violin Plots; horizontal bars indicate the median ± the 1st and 3rd quartiles. Kruskal–Wallis ANOVA test with Dunn’s post hoc test. P values are displayed on graphs. (D) Primary culture of Sp + /- and Sp-/- cortical neurons immunolabelled at DIV9 with βIII-tubulin antibodies, F-actin probes (Phalloidin, pink) and DAPI (grey). Axonal swellings (pink arrows) of Sp + /- exhibit the same characteristic features as those described in Sp-/- cultures. They are (i) always located close to the growth cone (arrowheads), (ii) their diameter is at least 2 to 3 times larger than the diameter of the axon shaft, (iii) they are always strongly labelled by tubulin antibodies and (iv) are always negative for DAPI staining (asterisk). Scale bars: 25 µm. Source data are available online for this figure

p60-katanin impairs the axon pathfinding of secondary motor neurons.

(A) Immunolabelling of sMN tracts in 72-hpf Tg(Hb9:GFP) larvae injected with control morpholino (MO^Ctl; n = 32), p60-katanin morpholino (MO^{p60Kat/1.3pmol}, n = 32 and MO^{p60Kat/3.4pmol}, n = 32) or co-injected with p60-katanin morpholino and human KATNA1 mRNA (MO^{p60Kat/1.3pmol} + mRNA^KATNA1, n = 32; MO^{p60Kat/3.4pmol} + mRNA^KATNA1, n = 32) using Zn-5 and GFP antibodies. Dotted lines mark lateral myosepta. (B–D) Mean number of misguided dorsal projections (B), missing dorsal projections (C) and misguided rostral nerves (D) per larva. (E) Percentage of larvae with ectopic sorting of sMN axons from the spinal cord. (F) Immunolabelling of sMN axons in 72-hpf controls (katna1^+/+, n = 34), as well as heterozygous (katna1^+/-, n = 53), homozygous zygotic (katna1^−/−Z, n = 33) and maternal zygotic (katna1^{-/- MZ}, n = 12) mutant larvae using Zn-5 antibody. Insets are higher magnifications of the dorsal nerve. (G) Percentage of larvae with dorsal nerve defects. (H) Mean number of misguided dorsal projections. (I) Percentage of larvae with rostral nerve defects. (A, F) Lateral views of the trunk, anterior to the left. Full arrowheads and full arrows show normal rostral and dorsal nerves, respectively. Empty arrowheads and empty arrows indicate misguided rostral and split/misguided dorsal projections, respectively. Asterisks indicate missing dorsal nerves while the red arrowhead points at aberrant exit points of sMN axons from the spinal cord. Scale bars: 25 µm, inset: 10 µm (F). (B–E, G–I) Non-blind quantifications were performed on 24 spinal hemisegments around the yolk tube per larva. Analysed larvae were pooled from three independent experiments. (B–D, H) Violin Plots; horizontal bars indicate the median ± the 1st and 3rd quartiles. Kruskal–Wallis ANOVA test with Dunn’s post hoc test. (E, G, I) Chi² test. P values are displayed on graphs. Source data are available online for this figure.

Loss of p60-Katanin causes a dramatic decrease in zebrafish larval mobility.

(A) Touch-evoked escape behaviour of 72-hpf larvae injected with a control morpholino (MO^Ctl; n = 90), increasing doses of p60-katanin morpholino (MO^{p60Kat/1.3pmol}; n = 102 and MO^{p60Kat/3.4 pmol}; n = 85), or co-injected with MO^p60Kat morpholino and human KATNA1 mRNA (MO^{p60Kat/1.3pmol} + mRNA^KATNA1, n = 91; MO^{p60Kat/3.4pmol} + mRNA^KATNA1, n = 97). (B, C) Mean swimming distance (B) and speed (C) of the larvae tracked in (A). (D) Touch-evoked escape behaviour of 72-hpf control (katna1^+/+; n = 65) and katna1 mutant larvae injected or not with human KATNA1 mRNA (katna1^{-/- MZ};n = 74 and katna1^{-/- MZ} + mRNA^KATNA1; n = 78) and pooled from three independent experiments. (E, F) Mean swimming distance (E) and speed (F) of the larvae tracked in (D). (A, D) Each line represents the trajectory of one larva after touch stimulation while the distance between two dots indicates the distance covered by a larva between two consecutive frames. Scale bar: 5 mm. (B, C, E, F) Violin Plots; horizontal bars indicate the median ± the 1st and 3rd quartiles. Kruskal–Wallis ANOVA test with Dunn’s post hoc test. P‐values are displayed on graphs. Source data are available online for this figure.

Spastin and p60-katanin have non-redundant roles in motor axon guidance.

(A) Immunolabelling of sMN axons in 72-hpf Tg(Hb9:GFP) larvae injected with MO^Ctl (n = 10), 1.3 pmol of MO^p60Kat (n = 10), or co-injected with MO^Ctl or MO^p60Kat and human SPG4 mRNA (MO^Ctl + mRNA^Spast, n = 14; MO^{p60Kat/1.3 pmol} + mRNA^Spast, n = 10) using Zn-5 and GFP antibodies. (B, C) Mean number of misguided/split dorsal nerves (B) and misguided rostral nerves (C) per larva. (D) Immunolabelling of sMN axons in 72-hpf Tg(Hb9:GFP) larvae injected with MO^Ctl (n = 29), MOsp^ATG (n = 30), or co-injected with MO^Ctl or MOsp^ATG and human KATNA1 mRNA (MO^Ctl + mRNA^KATNA1, n = 32; MOsp^ATG + mRNA^KATNA1, n = 29) using Zn-5 and GFP antibodies. (E, F) Mean number of abnormal rostral nerves (i.e., caudally targeted or missing; E) and missing dorsal nerves (F) per larva. (A, D) Dotted lines indicate lateral myosepta. Full arrowheads and full arrows show normal rostral and dorsal nerves, respectively. Empty arrowheads and empty arrows indicate misguided rostral and dorsal projections, respectively. All images are lateral views of the trunk, anterior to the left. Scale bars: 25 µm. (B, C, E, F) Non-blind quantifications were performed on 24 spinal hemisegments around the yolk tube per larva. Analysed larvae were pooled from three independent experiments. Violin Plots; horizontal bars indicate the median ± the 1st and 3rd quartiles. Kruskal–Wallis ANOVA test with Dunn’s post hoc test. P values are displayed on graphs. Source data are available online for this figure.

TTLL6 and TTLL11 knockdown leads to different motor axon pathfinding defects mimicking the respective phenotypes of p60-Katanin- and Spastin-depleted larvae.

(A) Immunolabelling of polyglutamylated microtubules (MTs) in 26- (left panels) and 72-hpf (right panels) wild-type embryos using GT335 (upper panels) and polyE (lower panels) antibodies. Polyglutamylated MTs are observed in both pMN (26 hpf) and sMN (72 hpf) axons. Blue and pink arrows point at dorsally and ventrally projecting pMN axons, respectively. White arrows, yellow arrows and white arrowheads indicate dorsally, ventrally and rostrally projecting sMN axons, respectively. Scale bars: 25 µm. (B) Immunolabelling of polyglutamylated microtubules (MTs) in 26-hpf Tg(Hb9:GFP) embryos injected with MO^Ctl, MO^TTLL6 or MO^TTLL11 morpholinos using the polyE antibody. Scale bars: 25 µm. (C) Mean PolyE fluorescence intensity per axon (A.U.). Non-blind quantifications were conducted on MO^Ctl (n = 120), MO^TTLL6 (n = 74) or MO^TTLL11 (n = 68) pooled from two independent experiments. Twelve axons located around the yolk tube were analysed per embryo. (D) Upper panels: Overall morphology of 72-hpf control (MO^Ctl), TTLL6 (MO^TTLL6) and TTLL1 (MO^TTLL11) morphant larvae. Both TTLL6 and TTLL11 morphants exhibit a severe ventrally curved body axis phenotype compared to MO^Ctl-injected larvae. Scale bars: 250 μm. Middle and bottom panels: Immunolabelling of sMN axons in 72-hpf Tg(Hb9:GFP) larvae injected with MO^Ctl (n = 30), MO^TTLL6 (n = 30) or MO^TTLL11 (n = 30) larvae using Zn-5 and/or GFP antibodies. Dotted lines delineate lateral myosepta. Full arrowheads and full arrows, respectively, point at normal rostral and dorsal nerves. Empty arrowheads and empty arrows, respectively, indicate misguided rostral and dorsal projections. Brackets show defasciculated rostral nerves. Asterisks and yellow arrows, respectively, indicate the ectopic sorting of spinal motor neuron axons and somata from the spinal cord. Scale bars: 25 µm. (E–J) Quantifications of sMN defects in larvae analysed in (D). Mean number of split/misguided dorsal nerves (E), missing dorsal nerves (F), defasciculated/missing rostral nerves (G) and misrouted rostral nerves (I) per larva. (H, J) Percentage of larvae with ectopic sorting of sMN somata (H) or axons (J) from the spinal cord. Non-blind quantifications were performed on 24 spinal hemisegments located around the yolk tube per larva. Analysed larvae were pooled from three independent experiments. (C, E–G, I) Violin Plots; horizontal bars indicate the median ± the 1st and 3rd quartiles. Kruskal–Wallis ANOVA test with Dunn’s post hoc test. (H, J) Chi² test. P values are displayed on graphs. Source data are available online for this figure.

TTLL6 and TTLL11 play non-overlapping roles in spinal motor axon navigation.

(A) Rescue and cross-rescue experiments of TTLL6 morphant phenotypes by co-injection of mouse TTLL6 or TTLL11 transcripts. (B) Reciprocal rescue and cross-rescue experiments of TTLL11 morphant phenotypes by co-injection of mouse TTLL11 or TTLL6 transcripts. (A, B) Upper panels: Overall morphology of 72-hpf control, morphant and rescued larvae. Scale bars: 250 μm. bottom panels: Immunolabelling of sMN axons with Zn-5 and GFP antibodies in 72-hpf Tg(Hb9:GFP) larvae injected with control (MO^Ctl, n = 22), TTLL6 (MO^TTLL6, n = 22), TTLL11 (MO^TTLL11, n = 21) morpholinos or co-injected with MO^TTLL6 or MO^TTLL11 and mouse TTLL6 or TTLL11 mRNA (MO^TTLL6 + mRNA^TTLL6, n = 19; MO^TTLL6 + mRNA^TTLL11, n = 22; MO^TTLL11 + mRNA^TTLL11, n = 17; MO^TTLL11 + mRNA^TTLL6, n = 22). Dotted lines delineate lateral myosepta. Full arrowheads and full arrows, respectively, indicate normal rostral and dorsal nerves. Empty arrowheads and empty arrows, respectively, point at misguided rostral and dorsal projections. Scale bars: 50 μm. (C–H) Quantifications of sMN defects in larvae analysed in (A, B). Mean number of split/misguided dorsal nerves (C), misrouted rostral nerves (D), missing dorsal nerves (E) and defasciculated rostral nerves (F) per larva. (G, H) Percentage of larvae with ectopic sorting of sMN axons (G) and somata (H) from the spinal cord. (C–H) Non-blind quantifications were performed on 24 spinal hemisegments located around the yolk tube per embryo or larva. Analysed larvae were pooled from three independent experiments. (C–F) Violin Plots; horizontal bars indicate the median ± the 1st and 3rd quartiles. One-way ANOVA test with Bonferroni’s post hoc test (D) or Kruskal–Wallis ANOVA test with Dunn’s post hoc test (C, E, F). (G, H) Chi² test. P values are displayed on graphs. Source data are available online for this figure.

Selective regulation of p60-Katanin activity by TTLL6 is required for zebrafish motor axon targeting and larval locomotion.

(A) Upper panels: Immunolabelling of sMN axon tracts using Zn-5 and GFP antibodies in 72-hpf Tg(Hb9:GFP) larvae injected with MO^Ctl (n = 20), MO^{p60Kat/1.3pmol} (n = 20) or co-injected with MO^{p60Kat/1.3pmol} and mouse TTLL6 (n = 19) or TTLL11 (n = 20) mRNAs. Dotted lines delineate lateral myosepta. Lower panels: Tracking analysis of 72-hpf larvae injected with MO^Ctl (n = 30), MO^{p60Kat/1.3pmol} (n = 30) or co-injected with MO^{p60Kat/1.3pmol} and mouse TTLL6 (n = 30) or TTLL11 (n = 30) mRNAs in a touch-evoked escape response test. Each line represents the trajectory of one larva after touch stimulation while the distance between two dots indicates the distance covered by a larva between two consecutive frames. Scale bar: 5 mm. (B–E) Quantifications of the sMN and locomotor defects of larvae analysed in (A) and pooled from three independent experiments. Mean number of split/misguided dorsal nerves (B) and misrouted rostral nerves (C) per larva. (D, E) Mean swimming covered distance (D) and speed (E). (F) Immunolabelling of sMN axon tracts with Zn-5 antibody in 72-hpf katna1^+/+ (n = 57), katna1^-/-MZ (n = 61) and katna1^-/-MZ larvae injected with mouse TTLL6 (n = 56) or TTLL11 (n = 54) mRNAs. (A, F) Full arrowheads and full arrows point at normal rostral and dorsal projections while empty arrowheads and empty arrows indicate misguided rostral and dorsal tracts, respectively. Scale bar: 25 μm. (G–I) Quantifications of dorsal nerve and motility defects of larvae analysed in (F) and pooled from two independent experiments. (G) Percentage of larvae with dorsal nerve defects. (H) Mean number of split/misguided dorsal nerves per larva. (I) Mean larval swimming speed in the escape-touch response test. Swimming speed values were extracted from tracking analysis of 72-hpf katna1^+/+ (n = 65), katna1^-/-MZ (n = 74) and katna1^-/-MZ larvae injected with the mRNAs encoding mouse TTLL6 (n = 78) or TTLL11 (n = 47). (B, C, G, H) Non-blind quantifications were performed on 24 spinal hemisegments located around the yolk tube per larva. (B–E, H, I) Violin Plots; horizontal bars indicate the median ± the 1st and 3rd quartiles. Kruskal–Wallis ANOVA test with Dunn’s post hoc test. (G) Chi² test. P values are displayed on graphs. Source data are available online for this figure.

TTLL6 overexpression rescues the defects of MT dynamics caused by the partial loss of p60-Katanin MT-severing activity in navigating motor axons.

(A) Representative z-projection still images of EB3-GFP comet time-lapse recordings in sMN axons of 52-hpf Tg(Mnx1:GAL4;UAS:EB3-GFP) larvae injected with MO^Ctl, MO^{p60Kat/1.3pmol}, or co-injected with MO^{p60Kat/1.3pmol} and mouse TTLL6 mRNAs (related to Movie EV4). Arrowheads point at MT plus ends (i.e., EB3-GFP comets). Insets are higher magnifications of the boxed axonal portion. Scale bars: 5 μm and 2 μm (insets). (B) Upper panels: Representative kymograms of a 5-min EB3-GFP recording. Lower panels: Schematic representation of the corresponding kymograms illustrating the EB3-GFP traces. Horizontal bar, 5 µm; vertical bar, 2 min. (C, D) Mean number of EB3 comets per 50 μm of axon (C) and mean EB3 comet velocity (D). MT plus-end dynamics was monitored and quantified in a non-blind manner in 58, 74 and 38 dorsally projecting sMN axons of 72-hpf larvae injected with MO^Ctl, MO^{p60Kat/1.3pmol} or co-injected with MO^{p60Kat/1.3pmol} and mouse TTLL6 mRNAs, respectively. These axons were obtained from at least 6 larvae per condition and pooled from four independent experiments. Violin Plots; horizontal bars indicate the median ± the 1st and 3rd quartiles. Kruskal–Wallis ANOVA test with Dunn’s post hoc test. P values are displayed on graphs. Source data are available online for this figure.

Selective regulation of spastin activity by TTLLL11 controls zebrafish motor circuit wiring and larval motility.

(A) Upper panels: Immunolabelling of sMN axon tracts using Zn-5 antibodies in 72-hpf Tg(Hb9:GFP) larvae injected with MO^Ctl (n = 69), MOsp^ATG (n = 79) or co-injected with MOsp^ATG and mouse TTLL6 (n = 63) or TTLL11 (n = 78) mRNAs. Dotted lines delineate lateral myosepta. Lower panels: Tracking analysis of 72-hpf larvae injected with MO^Ctl (n = 71), MOsp^ATG (n = 90) or co-injected with MOsp^ATG and mouse TTLL6 (n = 98) or TTLL11 (n = 124) mRNAs in a touch-escape response test. Each line represents the trajectory of one larva after touch stimulation while the distance between two dots indicates the distance covered by a larva between two consecutive frames. Scale bar: 5 mm. (B–D) Quantifications of the sMN (B) and locomotor defects (C, D) of larvae analysed in (A) and pooled from three independent experiments. (B) Mean number of abnormal rostral nerves. (C, D) Mean swimming speed (C) and covered distance (D). (E) Immunolabelling of sMN axons in 72-hpf sp^+/+ (n = 52), sp^C68X/C68X (n = 34) and sp^C68X/C68X larvae injected with mouse TTLL6 (n = 47) or TTLL11 (n = 49) transcripts using Zn-5 antibody. (F, G) Quantifications of rostral nerve defects in larvae analysed in (E) and pooled from three independent experiments. (F) Percentage of larvae with rostral nerve defects. (G) Mean number of abnormal rostral nerves (i.e., defasciculated or missing) per larva. (A, E) Full and empty arrowheads, respectively, point at normal and defasciculated/missing rostral nerves. Full arrows in (E) indicate caudally oriented “rostral” nerves. Scale bar: 50 μm. (B, F, G) Non-blind quantifications were performed on 24 spinal hemisegments located around the yolk tube per larva. (B–D, G) Violin Plots; horizontal bars indicate the median ± the 1st and 3rd quartiles. Kruskal–Wallis ANOVA test with Dunn’s post hoc test. (F) Chi² test. P values are displayed on graphs. Source data are available online for this figure.

TTLL11, but not TTLL6, rescues the axonal swelling phenotype caused by spastin haploinsufficiency in mammalian cortical neurons.

(A) Immunolabelling of β-III tubulin and GFP on DIV9 Sp + /+, Sp + /- and Sp-/- cortical neurons transduced or not at DIV2 with lentiviruses encoding GFP-tagged wild-type or catalytically dead variants (TTLL11d and TTLL6d) of mouse TTLL11 or TTLL6. Blue arrowheads and pink arrows, respectively, indicate non-swollen and swollen axons. Insets are higher magnifications of the distal part of the axons framed in the boxed region of each corresponding image. Arrowheads indicate growth cones. Scale bars: 20 μm. (B) Percentage of swollen axons. Blind quantifications were performed on at least 400 axons per condition pooled from two or three independent experiments. The “d” stands for catalytically dead variants of the TTLL enzymes. Chi2 test. P values are displayed on graphs. Source data are available online for this figure.

Schematic representation of MT-severing enzyme selective regulation by specific TTLL enzymes in zebrafish and mammalian neurons.

(A) Upper panel: TTLL6-mediated tubulin polyglutamylation selectively tunes p60-Katanin activity in zebrafish dorsally projecting secondary motor axons to control their targeting. Lower panel: TTLL11-driven MT polyglutamylation is required for accurate axon pathfinding of rostrally projecting secondary motor nerves in zebrafish larvae. (B) Spastin haploinsufficiency induces axonal swellings in mammalian cortical neurons (middle panel), which can be selectively rescued by promoting TTLL11-mediated tubulin polyglutamylation (lower panel), most likely through the boosting of residual spastin activity upon the critical threshold (i.e., 50%).