DELAY identifies ybx1 as a global transcriptional regulator upstream of atoh1a.

(A) A t-SNE embedding depicts scRNA-seq data from Baek et al. (2022) containing HC progenitors, young HCs, mature HCs, central SCs, dorsoventral (DV) SCs, anteroposterior (AP) SCs, amplifying SCs and mantle cells during homeostasis and 0, 0. 5, 1, 3, 5, and 10 hpa. (B) Two pseudotime trajectories depict central SC and HC regeneration from 0-10 hpa. (C) A bubble plot shows key TFs in the combined DELAY GRN for HC and central SC regeneration. Bubble size represents outdegree centrality. (D) The expression of key TFs is depicted using generalized additive models fitted to cells’ timepoints from 0–5 hpa. The expression of ybx1 precedes that of atoh1a. (E) A dot plot shows that ybx1 is expressed across all cell types and timepoints but is relatively enriched across HC progenitors and young HCs at 3–5 hpa. The analysis in (A–E) is based on data from Baek et al., (2022). (F) Relative enrichment of bipartite motifs suggests that Ybx1 regulates transcription of its targets proximally to their transcription start sites (TSSs). (G) Two-thirds of 303 TFs in the combined GRN are predicted targets of ybx1 and possess ≥1 bipartite Ybx1 motif in their enhancers. (H) Alignment of atoh1a revealed a candidate regeneration-responsive promoter element with two Ybx1 DNA-binding sites and a 5′  UTR element with Ybx1 RNA-binding and internal ribosome entry (IRE) sites. Cyprinid species also lack proximal C-sites (C1) for Notch-mediated suppression of atoh1a transcription (Abdolazimi, Stojanova & Segil, 2016). rc, reverse complement.

ybx1 promotes rapid HC regeneration and atoh1a expression after damage.

(A) Representative images show phalloidin staining of HCs in post-ablation and control NMs from wild-type, heterozygous, and homozygous ybx1 mutant larvae (scale bar = 4 µm). (B and C) Quantification of HCs and DAPI ⁺ nonsensory cells from post-ablation and control NMs. Each data point represents a single NM. (D and E) Representative images of Ybx1 immunostaining across NMs from wild-type, heterozygous, and homozygous ybx1 mutant larvae (scale bar = 10 µm). Inset images show Ybx1 accumulation (arrowheads) in atoh1a⁺ cells (inset scale bars = 5 µm). (F) Quantification of Ybx1 immunofluorescence in atoh1a⁺ cells from wild-type (N = 4) and heterozygous ybx1 mutant (N = 3) larvae. Data points represent individual atoh1a⁺ HC progenitors or young HCs across wild-type (n = 27) and mutant (n = 21) categories. (G) Representative fluorescence images show atoh1a expression in newly regenerated HCs from wild-type and heterozygous ybx1 mutant larvae. dTomato marks HCs’ apical surfaces and the growing kinocilia (arrowhead) in older HCs (scale bar = 4 µm). (H and I) Quantification of new HCs in control and 12–24 hpa NMs from wild-type and heterozygous ybx1 mutant larvae. (J) Proposed model of atoh1a transcriptional regulation during pre-ablation, early-phase, and late-phase HC regeneration. Where shown, average quantities represent mean values across the pooled data points and the error bars represent SEMs. Complete descriptions of quantifications and statistical analyses can be found in the Methods section and Supplementary Tables (*, p < 0.05; **, p < 0.01).