Fignl1-mediated restriction of dynein activity is required for accurate navigation of RoP-like sMN axons. (A and B) Immunolabeling of sMN axons in 56-hpf WT larvae treated at 46 hpf with 4 µM of the dynein inhibitor Ciliobrevin D or its vehicle (DMSO) using the Zn-5 antibody. (B) Quantification of the mean number of missing rostral nerves in Ciliobrevin- and DMSO-treated WT larvae. (C) Immunolabeling of sMN in 56-hpf WT and dynactin1b^−/− mutant larvae. (D and E) Mean number of mistargeted (i.e., caudally oriented; D) or missing (E) rostral nerves in control and mutant larvae. (F) Analysis of sMN in 56-hpf control and Fignl1-overexpressing larvae. (A–F) Loss of dynein or dynactin 1b function leads to sMN axon pathfinding defects that reproduce those observed in Fignl1-overexpressing embryos. (G) Immunolabeling of sMN axons in 56-hpf Fignl1 morphant larvae treated with 1 µM Ciliobrevin D or DMSO. (H) Mean number of defasciculated/split rostral nerves in Ciliobrevin D– or DMSO-treated larvae. Pharmacological reduction of dynein activity partially rescues the axon pathfinding defects of RoP-like sMN axons in Fignl1-depleted larvae. (A, C, F, and G) Empty arrowheads indicate normal rostral nerves. Arrowheads and arrows, respectively, point at misguided and missing (A, C, and F) or defasciculated/split (G) rostral nerves. Scale bars, 40 µm. (B, D, E, and H) Quantifications were performed in 24 spinal hemisegments of 28 DMSO-treated and 27 Ciliobrevin D–treated control larvae (B), 39 WT and 37 dynactin1b^−/− mutant larvae (D and E), and 36 DMSO-treated and 36 Ciliobrevin D–treated Fignl1 morphant larvae (H) pooled from three independent experiments. **, P ≤ 0.01; ***, P ≤ 0.001; ns, nonsignificant; Mann–Whitney test (B, D, E, and H). Error bars are SEM.