Neuromast regeneration in the lateral line of tail-amputated zebrafish larvae. (a) As illustrated, we cut the tails of Et(HG7L) larvae at 3 days post-fertilization (dpf) to remove the L6–L8 neuromasts. We observed the regeneration of a new neuromast (indicated by arrowheads) by epifluorescence microscopy and categorized this process into four distinct stages. (b) Phase I: The increase in fluorescent cells was not observable in the lateral line between the L5 neuromast and the wound site (dotted line). (c) Phase II-1: Fluorescent cells increased and piled up into two layers. (d) Phase II-2: A cluster of fluorescent cells formed with more than two layers. (e) Phase III: The formation of a new neuromast. (f) We counted the larvae at each phase on the designated day post-amputation (dpa). The percentages of larvae in each phase among the 59 surviving larvae (pooled from four independent experiments) are shown. Due to low resolution, the boundary between layers is not clearly visible under epifluorescence microscopy. Confocal images for each corresponding phase are presented in Panels B′–E'.

Clodrosome effectively reduces the number of macrophages. We administered either 2 nL of the vehicle control (encapsome [En], n = 15) or clodrosomes (Clo, 2.125 mg/mL, n = 18) to Tg(mpeg1:mCherry) larvae, as illustrated in Panel (a), or left them untreated (Unt, n = 15). Subsequently, as depicted in Panel (b) (with a fluorescence image superimposed on a bright field image), we quantified the number of macrophages (expressing mCherry in red) in the tail region, located 200 μm from the turning point of the tail blood vessel. At 1 and 24 h post-injection, we employed an epifluorescence microscope with a Rhodamine filter cube to assess the macrophages. (c) Representative images illustrating a focal plane for each treatment at specified time intervals. Due to the low resolution of the original images, we globally enhanced the images by increasing brightness and contrast using Photoshop. Utilizing the z-axis, we tallied the number of macrophages. The scatter plot with standard error bars is presented in (d). The unpaired t-test was used to analyze the differences between treatments. The p-values were calculated using the online tool GraphPad by Dotmatics (https://www.graphpad.com/quickcalcs/ttest1/?Format=SEM). ***p < .0001; the remaining p-values are shown in the figure.

Inhibition of macrophages reduces cell proliferation in the lateral line. We injected encapsome (En) or clodrosomes (Clo) into the posterior cardinal vein of Et(HG7L) larvae 3 days post-fertilization, and the tails were amputated as previously described or left intact. The larvae were fixed 1 and 2 days post-amputation (dpa) and subjected to the EdU assay and immunohistochemistry against EGFP to label proliferating cells (EdU+ in red) and the lateral line (in green), respectively, as shown in a representative image of an untreated larva (a). Then, we counted the number of proliferating lateral line cells around the L5 neuromast (white arrowheads) and the lateral line (LL) behind L5 (yellow arrowheads). (b) For the tail-amputated groups, the treatment effect was significant in both L5 and the LL behind L5 (ANOVA, p < .001). The groups with significant differences as determined by post hoc t-tests are marked on the plots. (c) For the uninjured groups, the treatment effect was significant in the L5 group (ANOVA, p < .01), and only the expression of EdU+ GFP+ cells in the En+/Clo− group at 1 dpi was found to be significantly lower than in the En−/Clo− group (t-test, p < .05). No significant effects were found in the LL behind L5 group (i.e., in between L5 and L6). Error bars indicate SEM. *p < .05; **p < .01.

Inhibition of macrophages has no significant effect on the differentiation of regenerating neuromasts. (a) At 3 days post-fertilization, we performed tail amputation on Et(HG7L) larvae. At 2 days post-amputation (dpa), larvae were either left untreated (Unt) or injected with 2 nL of the control encapsome (En) or clodrosomes (Clo) into the posterior cardinal vein. The percentage of surviving larvae at each phase for each treatment was recorded at the designated times, as outlined in Figure 1. The chi-square test did not identify significant differences in the ratios of phases between groups for each time point. For each treatment, n = 56–59 samples were used, and the inclusion or exclusion of deceased larvae did not alter the results of statistical analyses. At the designated time points, larvae were labeled with Rhodamine 123 (R123) to visualize live hair cells. An illustration of an untreated larva at Phase III with R123-labeled hair cells is provided in (b). Subsequently, the hair cell-labeled larvae underwent immunohistochemistry against Sox2 to highlight mantle cells (MCs) and supporting cells (SCs). An untreated larva at Phase III with Sox2-expressing cells in the regenerating neuromast is shown in (c), with a magnified view on the right. (d, e) Grouped column scatter plots with error bars display the numbers of R123-positive (R123+) cells (d) and Sox2+ cells (e) per neuromast (the regenerating neuromast). The number of hair cells differed significantly only between different time points (ANOVA, p < .01), and no significant effects were observed for the number of Sox2+ cells. Error bars represent SEM.

Effects of cytokine inhibitors on neuromast regeneration. We cut the tails of Et(HG7L) larvae at 3 days post-fertilization and then cultured them in the absence (Unt) or presence of (a) 0–10 μM LMT-28 (n = 58–60 for each treatment group at 0 dpa), (b) 0–100 μM pentoxifylline (PTX) (n = 40–42 for each treatment group at 0 dpa), (c) 0–25 μM SB431542 (n = 31–39 for each treatment group), and (d) 0–10 μM SB505124 (n = 33–41 for each treatment group at 0 dpa). (e) Larvae from the cross of Et(HG7L) and Tg(−8.0cldnb:NTR-hkiKGR) were treated with 2 mM metronidazole (Mtz) 3 days post-fertilization for 12 h to chemical-genetically ablate neuromasts, washed, and treated without (Unt) or with 25 μM SB431542. We examined the regeneration of neuromast. The % of larvae in each neuromast regeneration phase is presented as described in Figure 1, except that Phase II-1 and Phase II-2 are combined as Phase II (panel on the left). We counted the number of larvae at each phase at designated time points (presented as days post-amputation [dpa]). The percentage of larvae in each phase is shown. n.s. not significant; *p < .05; **p < .01; ***p < .001.

Inhibiting TGF-β decreases the proliferation of neuromast progenitor cells. (a) We cut the tail fins of Et(HG7L) larvae 3 days post-fertilization and then cultured them in the absence (Unt) or presence of 25 μM SB431542. Larvae were fixed at the designated time points and subjected to the EdU assay and EGFP immunohistochemistry to probe actively proliferating cells (red) and the lateral line (green), respectively. Proliferating cells on the lateral line were counted as described in Figure 3. The data are presented in a grouped column scatter plot in (b). Post hoc t-tests revealed significant differences after a significant treatment effect was found by ANOVA (L5, p < .001; LL behind L5, p < .01). *p < .05; **p < .01; ***p < .001.

Inhibiting matrix metalloproteinases does not significantly affect neuromast formation but increases the number of hair cells. (a) We cut the tails of Et(HG7L) larvae at 3 days post-fertilization and then cultured them in the absence (Unt) or presence of 0–100 μM GM6001. We counted the number of larvae at each phase at designated time points (presented as days post-amputation [dpa]). The percentage of larvae in each phase is shown. For each treatment group, n = 42–46 samples were included at 0 dpa. Data were obtained from three independent experiments. Including or excluding the dead larvae did not change the results of statistical analyses. Therefore, we show only data from surviving larvae (n = 41–46 for each treatment). Chi-square tests did not reveal any significant differences in the ratios of phases between treatments. (b) Hair cells were labeled by Rhodamine 123 and counted (R123+) at designated time points. The data are presented in a grouped column scatter plot. Post hoc t-tests revealed significant differences after ANOVA revealed a significant treatment effect (p < .01). Error bars indicate SEM. *p < .05.