All zebrafish, including the wild type AB line and the transgenic Tg (cldnb: lynGFP) and Tg (brn3c: mGFP)^s356t lines were bred in 28.5°C constant temperature incubator in embryo medium according to the standard formula (Kimmel et al., 1995). The stage of embryonic development was marked as hours- or days- after fertilization (hpf or dpf) (Kimmel et al., 1995). In order to avoid pigmentation, the embryos should be further immersed in 1-phenyl-2-thiourea (PTU) (Sigma-Aldrich) in the culture medium from 10 hpf (Tang et al., 2019). The operations on zebrafish were discussed and permitted by the Animal Conservation and Utilization Committee of Fudan University in Shanghai.

Cingulin b-MO, sequenced in 5’-TCCTGTCCGCAGAGAGGG AACTCAT-3’, was injected at a dose of 2 ng or 3 ng at one or two cell stage of embryos to reduce the expression of cingulin b. The other siblings were considered as controls by injection with a sequence of 5’-CCTCTTACCTCAGT TACAATTTATA-3’, namely control-MO (Control-MO). For the messenger RNA (mRNA) rescue experiment, a mixture of cingulin b-MO and cingulin b mRNA (Forward primer: 5’-AT GAGTTCCCTCTCTGCGGA-3’; Reverse primer: 5’-TCAACAG CTGGTGGTCTGAA-3’) was injected at the same stage with other groups.

WISH experiment was operated as previously disclosed (He et al., 2014; Thisse and Thisse, 2014). To examine the expression pattern of cingulin b in zebrafish, we collected embryos at various stages including 3.7, 14, and 48 hpf. To verify the efficacy of cingulin b-MO in the lateral line system of zebrafish, we collected embryos at 48 hpf. After a series of gradient solutions for dehydration, the collected embryos were stored in pure methanol (100% concentration) at −20°C. Before hybridization, the embryos would be gradient rehydrated first, and then digested with 20 μg/ml protease K. The probe was added and hybridized at 65°C constant temperature overnight. After thorough washes with the SSC-series at 65°C, the embryos were blocked in 2x BBR at room temperature for at least 1 h. Anti- digoxigenin (Dig)-AP Fab fragment (Roche) was added and incubated with specimens overnight at 4°C. Primers for synthesizing the objective genes were listed in Supplementary Table 1. Color reaction was implemented with BM purple AP substrate (Roche) in the dark at 37°C, and stopped with NTMT. The embryos after three times rinses were then re-fixed in 4% PFA and treated with different gradients of glycerol/PBS. The final specimens were stored in 100% glycerol and photographed by fluorescence stereoscopic microscope. All images were prepared by Photoshop and Illustrator software (2018, Adobe).

Bromodeoxyuridine (BrdU) co-incubation was conducted to label the proliferative cells. The dechorionated embryos at 34 hpf were incubated in 10 mM BrdU (Sigma-Aldrich) for 2 hours to show the cell proliferation in the PLL primordium, while the dechorionated larvae at 2 dpf were incubated in 10 mM BrdU (Sigma-Aldrich) for 24 hours to examine the proliferative cells in PLL neuromasts of zebrafish. The corresponding embryos or larvae were collected, anesthetized in 0.02%MS-222 (Sigma-Aldrich), and fixed in 4% PFA at 4°C overnight. After washing with PBT-2 for 3 times, the collected embryos were soaked in 2 N HCl at 37°C for 30 min. After incubation with the primary anti-BrdU monoclonal antibody (1:200 dilution; Santa Cruz Biotechnology) for 1 h at 37°C following 4°C overnight, the samples were washed for several times and then incubated with the secondary Cy3 polyclonal antibody (1:300 dilution; Jackson) for 1 h at 37°C. DAPI (1:800 dilution; Invitrogen) was added and incubated with the embryos or larvae for 20 min at room temperature to label the nuclei. The fluorescence-labeled embryos were imaged by Leica confocal fluorescence microscope (TCS SP8; Leica). The images obtained were further rotated, cut, and adjusted in the brightness by Photoshop (2018, Adobe) and then the images were aligned and added with fonts or labels by Illustrator software (2018, Adobe).

Before specimen collection, the zebrafish embryos at 48 hpf accepted depletion of chorion and the yolk sac. The total RNA was extracted with TRIzol reagent (Thermo Fisher Science) and reversely transcribed into cDNA using the first strand of transcriptional cDNA synthesis kit (Roche). An Illumina HiSeq X Ten platform was used for library sequencing. Raw reads were firstly filtered out the data in low-quality, and the remaining high quality raw data were used for downstream analyses. We used the Spliced Transcripts Alignment to a Reference (STAR) software as the reference genome library. Differential expression analysis was conducted with the DESeq (2012) R package, and p-value <0.05 indicated significant difference. R package was performed for KEGG pathway enrichment analysis of DEGs on the basis of hypergeometric distribution. KEGG pathway database were the reference for further functional and pathway enrichment analysis.

In order to fully quantity the mRNA level of target genes, a quantitative real-time PCR (qRT-PCR) system (LightCycler^®480) was operated on 48 hpf larvae in the Control-MO group and the cingulin b-MO group, using the PrimeScript RT reagent Kit (RR047A, Takara Biomedical Technology) and the SYBR PreMix Ex Taq Kit (RR820A, Takara Biomedical Technology). ^ΔΔCt method was chosen for results analysis. The primer sequences used in the study were described in Supplementary Table 2. Each qPCR assay was repeated in triplicate, and GAPDH was used as the internal reference genes.

All statistics were performed with GraphPad Prism software (version, 8.0c). Comparison between two groups was conducted with double-tailed Student t test, while comparisons among multiple groups were carried out by One-way ANOVA. Statistics were recorded as mean ± SEM (standard error of mean), and the difference was considered to be of significant difference with p-value less than 0.05.

In order to detect whether cingulin b is expressed in zebrafish, we collected embryos at various stages and conducted WISH analysis for cingulin b staining. As shown in Figure 1, cingulin b was detected expressed in the oblong stage at 3.7 hpf (Figures 1A,B), the 10-somite stage at 14 hpf (Figure 1C), and the deposited PLL neuromasts at 48 hpf (Figure 1D), mainly in the central HC area (Figure 1E). To confirm the expression of cingulin b in the early development of zebrafish, we also conducted the sense control probe for cingulin b at 48 hpf, however, we didn’t detect any expression of cingulin b in the lateral line system of zebrafish compared to that using antisense mRNA probe for cingulin b (Supplementary Figure 1).

To explore the role of cingulin b in the development of zebrafish, we injected specific morpholino (MO) targeting cingulin b at one or two cell stage of embryos for knockdown of cingulin b. The control group was injected with Control-MO to eliminate the effect of injection operation. The efficacy of cingulin b-MO-injection was examined by in situ staining of cingulin b and qRT-PCR analysis, that we found significantly down-regulated expression of cingulin b in the PLL neuromasts in the cingulin b-MO morphants compared to the controls (Figures 2A–D), and the quantitative level of cingulin b was remarkably decreased after cingulin b-MO injection compared to the embryos injected with Control-MO (Figure 2E). We also examined the embryos as a whole in the Control-MO and cingulin b-MO groups, and we did not find any obvious malformation in the entire zebrafish after injection with cingulin b-MO (Supplementary Figure 2).

The Tg (cldnb: lynGFP) zebrafish were used in this study to directly observe the morphology of neuromasts (Figure 3A). We counted the number of neuromasts at 48 hpf, a time point when the PLL primordium stops migration and finishes deposition (Kimmel et al., 1995; Nechiporuk and Raible, 2008). The average number of neuromasts in the trunk was notably decreased in the cingulin b-MO-injected morphants compared to that in the Control-MO-injected embryos (Figures 3A–C,F). The average number of neuromasts was even lower in 3 ng cingulin b-MO group than that in 2 ng cingulin b-MO group (Figures 3B–C,F), showing a dose-dependent manner, thus, we chose 3 ng dose for the following experiments. To avoid the non-specific effect of morpholino technology, we co-injected p53 with cingulin b-MO, and surprisingly the average number of neuromasts in the trunk in cingulin b-MO + p53 group was equivalent to that in cingulin b-MO-only group (Figures 3C–D,F). In addition, we also carried out rescue experiment, that combined injection with cingulin b mRNA and morpholino could partially restore the reduced number of neuromasts in the trunk (Figures 3E,F). The findings suggested that loss of cingulin b would affect the normal deposition of PLL neuromasts during the embryonic development of zebrafish. These findings were further validated by the expression of eya1, a marker for the neuromast in the lateral line of zebrafish (Kozlowski et al., 2005), that the number of neuromasts in the trunk was severely reduced in the cingulin b-MO morphants compared to that in the Control-MO embryos (Figures 3G,H). Taken together, our findings indicated that cingulin b was required in the lateral line system of zebrafish.

During the development of zebrafish lateral line, the collective cells migrate and form rosette-like structure in the trailing region (Aman and Piotrowski, 2008). The deposition of neuromasts occurs after assembly of the last rosette (Nechiporuk and Raible, 2008). Here, we found that cell proliferation in the primordium was destroyed during the embryonic development of zebrafish by BrdU staining after knocking down the gene expression of cingulin b (Figures 4A–F). BrdU index was defined as the number of BrdU-positive cells divided by the number of total cells labeled by DAPI in this article, which was used to evaluate cell proliferation ability. In 36 dpf, the BrdU index in the primordium of cingulin b-MO morphants decreased significantly, that the BrdU index was 10.04% ± 0.02 (n = 17) compared to 39.74% ± 0.02 in the Control-MO group (Figure 4G).

To investigate the sustained effect of cingulin b in zebrafish embryonic development, we stained the proliferative cells with BrdU (Figures 5B,F) and collected embryos at 3 dpf. The Tg (brn3c: mGFP)^s356t zebrafish were used here because of the HCs in neuromasts were labeled with GFP (Figures 5C,G). The total cells in neuromast were labeled with DAPI (Figures 5A,E). The merged images were shown in Figures 5D,H. The number of neuromast HCs in the trunk in cingulin b-MO experimental group decreased significantly compared to that of Control-MO group (Figure 5I). The BrdU index was also decreased severely in the cingulin b-MO group compared to that in the Control-MO group (Figure 5J). We also performed ISH staining of atohla, a maeker of HC, and found the expression of atoh1a was significantly decressed after knocking down of cigulin b compared to the control group (Figure 5K). Altogether, the data showed that knocking down cingulin b inhibited cell proliferation during primordia migration and neuromasts deposition in the early development process of zebrafish PLL system.

To explore the potential mechanism of cingulin b in regulating the development of zebrafish PLL system, we conducted RNA sequencing analysis to compare the difference between the control group and the cingulin b-MO mutants. KEGG analysis figured out the 13 top enriched pathways, of which MAPK signaling pathway and cellular senescence were the most two significant pathways evaluated by p value and gene counts (Figure 6). The key KEGG pathways, namely MAPK signaling pathway and cellular senescence pathway were listed in Figure 7. Also, the location of DEGs in cingulin b-MO siblings and overlapping genes of enriched pathways were revealed.

Heatmap analysis of DEGs of MAPK pathway and cellular senescence was screened in Figure 8 for Control-MO group vs. cingulin b-MO experimental group, respectively. RT-PCR analysis for some genes from MAPK and cellular senescence signaling pathways was conducted to verify our findings in RNA-sequencing data. The primer sequences were as listed in Supplementary Table 2. As shown in Figure 9, a total 9 genes in MAPK signaling pathways, 4 genes in cellular senescence pathway, and 9 genes overlapped in the two signaling pathways were examined. The mRNA levels of mapk1, mapk3, akt2, akt3b, atf7b, ppp3cca, and ppp3r1a were significantly decreased after knockdown of cingulin b, while the expression levels of tp53, mef2ca, mapk12b, and gadd45aa were significantly increased in cingulin b-MO group. The results of qRT-PCR were in consistency with those found in KEGG analysis, indicating that MAPK signaling was inhibited whereas cellular senescence was activated by repression of cingulin b.