Knockout of mafba Causes Inner-Ear Developmental Defects in Zebrafish via the Impairment of Proliferation and Differentiation of Ionocyte Progenitor Cells

Zebrafish is an excellent model for exploring the development of the inner ear. Its inner ear has similar functions to that of humans, specifically in the maintenance of hearing and balance. Mafba is a component of the Maf transcription factor family. It participates in multiple biological processes, but its role in inner-ear development remains poorly understood. In this study, we constructed a mafba knockout (mafba−/−) zebrafish model using CRISPR/Cas9 technology. The mafba−/− mutant inner ear displayed severe impairments, such as enlarged otocysts, smaller or absent otoliths, and insensitivity to sound stimulation. The proliferation of p63+ epidermal stem cells and dlc+ ionocyte progenitors was inhibited in mafba−/− mutants. Moreover, the results showed that mafba deletion induces the apoptosis of differentiated K+-ATPase-rich (NR) cells and H+-ATPase-rich (HR) cells. The activation of p53 apoptosis and G0/G1 cell cycle arrest resulted from DNA damage in the inner-ear region, providing a mechanism to account for the inner ear deficiencies. The loss of homeostasis resulting from disorders of ionocyte progenitors resulted in structural defects in the inner ear and, consequently, loss of hearing. In conclusion, the present study elucidated the function of ionic channel homeostasis and inner-ear development using a zebrafish Mafba model and clarified the possible physiological roles.


Introduction
Approximately 360 million people are born with congenital and progressive hearing defects worldwide [1]. Genetic defects account for approximately 50% of hearing loss [2], most of which is caused by mutations in genes associated with inner-ear development [3]. The vertebrate inner ear is a conserved sensory organ responsible for vestibular and auditory functions [4]. It plays an important role in detecting body acceleration and keeping balance [5,6]. The mammalian ear is composed of inner, middle, and external components, while the zebrafish, belonging to the teleost fish, only exhibits the inner ear. Although there is considerable structural diversity of the inner ears among different species, the molecular basis of their development is relatively conserved [7]. Zebrafish has become an excellent vertebrate model for exploring the mechanism of inner-ear development and related diseases in recent years [8,9]. There is an urgent need to enhance our understanding of the molecular mechanisms of inner-ear development and to pragmatize biological strategies for restoring auditory functions.

Generation of Mafba Mutant by CRISPR/Cas9 Technology
The target sites were designed using a web tool (http://chopchop.cbu.uib.no/, accessed on 30 August 2019). The target sequences of mafba we used are listed in Table S1. The gRNA and Cas9 mRNA were synthesized with a TranscriptAid T7 High Yield Transcription Kit (Thermo Scientific, Waltham, MA, USA) and mMESSAGE mMACHINE T7 transcription synthesis kit (Invitrogen, Carlsbad, CA, USA), respectively. A mixture of 600 pg Cas9 mRNA and 200 pg gRNA were co-injected into the one-cell-stage, wild-type embryos. At 3 dpf, 18 injected zebrafish embryos were collected to identify F0 founders by genotyping in order to evaluate CRISPR/Cas9 system target efficiency. A 464-bases pair (bp) DNA product containing the mafba targeting site was amplified by PCR (primers used are listed in Table S1) and sequenced. The F1 generation was obtained through the outcrossing of F0 mutant zebrafish to wild-type zebrafish. The F2 generation of mutant zebrafish were obtained and raised to adults using the same method as for the F1 generation described above. The F2 generation mafba +/− zebrafish were crossed to produce the sibling (mafba +/+ ), mafba +/− , and mafba −/− mutant embryos.

Whole-Mount In Situ Hybridization
Whole-mount in situ hybridization (WISH) for zebrafish embryos was executed as previously described [40]. The sequences of all probes and primers are listed in Table S1. The images were acquired using an optical microscope (10 × lens, BX53, Olympus, Tokyo, Japan). After imaging, we identified the genotypes by extracting DNA from the embryos. The numbers at the bottom right of each picture indicate the number of embryos with a similar staining pattern across all embryos examined.

EDU and TUNEL Assay
Zebrafish embryos were treated with 2 mM EDU (5-ethynyl-2 -deoxyuridine) for 30 min at 4 • C in egg-water. The embryos were transferred to fresh E3 medium for 2 h and fixed in 4% paraformaldehyde at 4 • C overnight. We treated the zebrafish embryos with Cell-Light™ EdU Cell Proliferation Detection (C10310-1, Ribobio, Guangzhou, China), according to the operating protocol, to visualize the proliferating cells. For the triple labeling of EDU, p63, and dlc, the samples were continuously treated in accordance with the standard immunofluorescence procedure and assessed via a confocal microscope (FV31S-SW, Olympus), as mentioned in Section 2.4. TUNEL staining was performed as previously described [45]. The apoptosis cells were labeled with the TUNEL BrightRed Apoptosis Detection Kit (Vazyme Biotech, NanJing, China). For double labeling of the TUNEL and atp1a1a.4 and the TUNEL and atp6v1a, as well as triple labeling the TUNEL, p63, and dlc, the samples were treated in accordance with the standard immunofluorescence procedure, and assessed by a confocal microscope (FV31S-SW, Olympus), as mentioned in Section 2.4.

Bioinformatics
All relevant Maf factor structures and sequences are available at Genome Browser (http://genome.ucsc.edu/cgi-bin/hgBlat, accessed on 12 July 2020) and Ensembl (http:// asia.ensembl.org/, accessed on 12 July 2020). SWISS-MODEL (https://swissmodel.expasy. org/, accessed on 12 July 2020) provided the fully automated protein structure homology modeling [47]. The amino acid sequences of Maf family factors from the other species mentioned above were downloaded from the NCBI GenBank database (http://www. ncbi.nlm.nih.gov/protein, accessed on 12 July 2020) and used for the multiple sequence alignments performed by ClustalW software.

Comet Assay
Heads from 30 zebrafish embryos were cut off and placed in 1.5 mL of ice-cold PBS (PBS + 20 mM EDTA). The cell suspensions were prepared by mincing the tissues. Alkaline Comet Assay was carried out in accordance with the manufacturer's procedure using the Comet Assay Kit (Abcam, Ab238544) [48]. DNA damage in single-strand breaks (SSB) and double-strand breaks (DSB) can be detected by alkaline electrophoresis. Images were captured by fluorescence microscopy (Nikon Ti_i80).

Startle Response Tests
The sound-evoked C-start escape response was examined in 96-well plastic plates and recorded with a high-speed camera (Olympus, E-M10, 1000 fps) under infrared light illumination. Pure tone stimulations (10 ms, 500 Hz) with two different intensities (20 dB and 40 dB) were supplied through a plastic board mounted on a voice box (HiVi, D1080MKII). Positive C-startle response was confirmed if the response occurred within 20 ms after the application of stimuli. For each group, 32 larvae were tested. Each larva was tested 10 times and each larva's C-startle times was computed. The possibility of measuring a C-startle response in a group of larvae was defined as the average percentage of C-startle [49].

Quantitative RT-PCR Analysis
Zebrafish embryos at 4 dpf were collected for total RNA extraction. The tails were dissected and used for genotyping. The residual head parts of approximately 30 were cut off and blended together to extract RNA by using the TRIzol Reagent (Vazyme Biotech, NanJing, China). They were then reverse transcribed into cDNA using HiScript ® II Q RT SuperMix (+gDNA wiper) for qPCR (Vazyme Biotech, Nanjing, China). The qPCR was carried out with AceQ TM qPCR SYBR Green Master Mix (Vazyme Biotech, Nanjing, China) in a StepOnePlus™ real time PCR machine, and analyzed with Graphpad 6.1 software. The relative mRNA expression levels of the selected genes were calculated using the 2 − Ct method [50]. The β-actin was set as an internal control. All qPCR primers used in this research are shown in Supplementary Table S1.

Flow Cytometry Analysis for Cell Cycle
For 4-dpf fish, 100 embryos were collected and washed once in ice-cold PBS. We removed all the solution into a 50-mL Falcon tube, then passed it through a 100-µm nylon filter (BD Falcon, Cat. 352360, Franklin Lakes, NJ, USA). We aspirated the solution and resuspended the pellet in 1 mL of Dispase II (Gibco, 1U/mL in PBS, Shanghai, China) to break up tissues. The solution was transferred into a 1.5-mL tube and incubated at 37 • C for 20 min. Then 1 mL of washing buffer (Hanks buffered saline solution containing 20% FBS, 5 mM CaCl 2 , and 0.5 u/mL DNAse) was added to the samples. The samples were treated with 0.2 mL of PBT (100 µg/mL RNase A and 50 00B5g/mL PI in PBS). Setting 200 µL PBS as the control group, we commenced dark treatment at room temperature for 30 min. We resuspended the cells in 1 mL PBS (with 2-3% FBS), which was then passed through a 40-µm nylon filter (BD Falcon, Cat. 352340, Franklin Lakes, NJ, USA), and then we counted the cells. The labeled cells were detected for PI staining using a Beckman Coulter CytoFlex 5, and the percentages of cells in the G0/G1, S, and G2/M phases in each sample were determined via gating using the Flowjo 10.0 software (Beckman Coulter, Brea, CA, USA).

Statistical Analysis
Statistical analyses were performed using GraphPad Prism 6.0 software. The experiments were repeated at least three times. The numbers of samples measure in every experiment are discussed in the Section 2 and in the figure legends. The two-tailed Student's t-test was used for determining significance. The results are presented as the mean ± SD. The level of significance was set to p < 0.05, and p > 0.05, p < 0.05, p < 0.01, p < 0.001, and p < 0.0001 were labeled as ns, *, **, ***, and ****, respectively.

Generation of mafba −/− Mutant Zebrafish with CRISPR/Cas9 Technology
There are two mafba transcripts (mafba-201(NM_131015.3) and mafba-202 (AB006322.1)) listed in the ZFIN database. Each contains only one exon, encoding 397 and 356 amino acids, respectively. Both can be detected in zebrafish embryos. Zebrafish Mafba proteins are 73%/72% identical at the amino acid level to the human/mouse Mafb protein, respectively. These are significantly higher than the sequence similarities among zebrafish Mafba, Mafbb, and Mafa proteins ( Figure S1A). The functional domains of the Maf family members are highly conserved among zebrafish and mammalian Mafb proteins, including N-terminal transcriptional activation domain (TAD), the C-terminal basic domain (BR), and the leucine zipper domain (LZ) ( Figure S1A,C), which are the functional domains of Maf family members [26]. All of the tertiary structures consist of three α-helixes and two β-sheets sharing many similarities ( Figure S1B). In vertebrates, small Maf evolves into three members (MafF, MafG, and MafK), while large Maf evolves into four members (MafA, MafB, C-MAF, and Nrl). Some of the genes duplicate because of an extra copy of the genome in the teleost [51]. Zebrafish have five large Maf genes and four small Maf genes. Due to an extra copy of the genome in the teleost fish, gene duplication may occur in both MafB and MafG [52].
To elucidate the physiological role of Mafba in vivo, we generated mafba knockout zebrafish using CRISPR/Cas9 technology. The target sites were designed in the exon of mafba-201 (NM_131015.3, NP_571090.2), in a sequence common to both transcripts. The detailed design of the mafba knockout is shown in Figure 1A. A 7-bp deletion mutation (c. 298_304delACTCCTA) would generate a premature termination codon and encode a truncated Mafba protein (p. Ser100Argfs*141). The del7 mutation may abolish the Mafba protein's function, and so this zebrafish line was selected for our research ( Figure 1B).
To further ensure the 7-bp deletion mutation, we verified the expression of mafba on the mRNA and protein levels. Firstly, semi-quantitative RT-PCR analysis showed that the expression levels of both transcripts (mafba-201, mafba-202) are significantly reduced in mafba mutant zebrafish ( Figure S2). Secondly, qRT-PCR analysis showed that mafba mRNA expression decreased by 25% in mafba +/− embryos and decreased more significantly in mafba −/− mutant embryos at 4 days postfertilization (dpf) ( Figure 1C). This suggests that the truncated mafba mRNAs are subject to nonsense-mediated decay. Thirdly, qRT-PCR analysis showed that the mRNA expressions of Maf homologues, such as large Maf (mafba, mafbb, mafa, c-maf, and nrl) and small Maf (maff, mafk, mafga, and mafgb), decrease significantly in the mafba mutants ( Figure S3). We speculated that mafba mutants may have systemic non-specific effects that reduce the expressions of many genes, otherwise affecting the viability of some tissues. Additionally, the temporal and spatial expression of mafba was determined in sibling ( Figure S4A-J) and mafba −/− mutants ( Figure S4K-T) via whole-mount in situ hybridization (WISH). Mafba displayed a dynamic expression pattern throughout embryo development ( Figure S4A-J). It emerged in bud-stage embryos ( Figure S4C) and was then expressed throughout the embryos ( Figure S4D-J). Noticeably, at 36 hpf, mafba was broadly expressed in the rhombomeres 5 and 6 [36], the inner ear, the eye, and the kidney ( Figure S4F,G). Moreover, the expression pattern seen at 2 dpf was maintained until 5 dpf ( Figure S4H-J). Semi-quantitative RT-PCR analysis showed that the expression of mafba was initiated at early developmental stages (6-somites stage) and then stabilized from 12 h postfertilization (hpf) to 6 dpf ( Figure S4U). This is similar to the temporal expression pattern of siblings, shown in Figure S4A-J. WISH results showed that mafba can be detected in mafba −/− mutants, and that it displayed strong expression at the bud stage and 6-somites stage ( Figure S4M,N). However, from 24 hpf onwards, mafba mRNA levels were dramatically decreased in mafba −/− embryos and were almost undetectable from 3 dpf ( Figure S4O-T). Western blot analysis also showed that the Mafba protein's expression was markedly decreased in the mutant line at 4 dpf, but did not completely disappear, likely due to the maternal deposit ( Figure 1D). In addition, immunostaining showed that Mafba was localized in the rhombomere (r) and inner-ear region of the sibling at 2 dpf. However, almost no Mafba signals were detected in mafba −/− embryos ( Figure S4V), suggesting its potential correlation with hindbrain and inner-ear development. dpf ( Figure S4U). This is similar to the temporal expression pattern of siblings, shown in Figure S4A-J. WISH results showed that mafba can be detected in mafba −/− mutants, and that it displayed strong expression at the bud stage and 6-somites stage ( Figure S4M,N). However, from 24 hpf onwards, mafba mRNA levels were dramatically decreased in mafba −/− embryos and were almost undetectable from 3 dpf ( Figure S4O-T). Western blot analysis also showed that the Mafba protein's expression was markedly decreased in the mutant line at 4 dpf, but did not completely disappear, likely due to the maternal deposit ( Figure 1D). In addition, immunostaining showed that Mafba was localized in the rhombomere (r) and inner-ear region of the sibling at 2 dpf. However, almost no Mafba signals were detected in mafba −/− embryos ( Figure S4V), suggesting its potential correlation with hindbrain and inner-ear development.

Depletion of Zebrafish mafba Results in Inner-Ear Defects and Hearing Loss
To determine whether mafba is essential for inner-ear development in zebrafish, the phenotype of mafba −/− embryos was examined and compared with that of sibling embryos. The mafba −/− mutants were indistinguishable in gross morphology from the sibling and mafba +/-embryos in the inner ear at 4 dpf. Morphological defects of the inner ear abruptly

Depletion of Zebrafish mafba Results in Inner-Ear Defects and Hearing Loss
To determine whether mafba is essential for inner-ear development in zebrafish, the phenotype of mafba −/− embryos was examined and compared with that of sibling embryos. The mafba −/− mutants were indistinguishable in gross morphology from the sibling and mafba +/− embryos in the inner ear at 4 dpf. Morphological defects of the inner ear abruptly occurred in the mafba −/− embryos at 5 dpf (Figure 2A,B). At 5 dpf, otoliths in the mutants showed varying degrees of size reduction. The otic defects of 5 dpf mafba −/− mutants included larger otocysts, thinner otic epithelia, and smaller or even eliminated otoliths (Figure 2A,C). According to the otolith size, we categorized the mafba −/− mutants into four groups: normal otoliths (accounting for 10.3%, n = 92), with otoliths comparable in size to the sibling embryos (on average 4662 ± 238 µm 2 ); small otoliths (accounting for 24.4%), at 35-75% of the average otolith size of the siblings; tiny otoliths (30.1%), 10-34% of the average otolith size of the siblings; and absent otoliths (35.2%), at less than 10% of the otolith size of the siblings ( Figure 2B). Less than 4% of the otic vesicles in sibling embryos (n = 117) could be categorized into the small group. Both the otic vesicle lumen and otolith sizes of zebrafish embryos were measured, and their areas were compared. The results further suggested the otic defect in mutants ( Figure 2C). The inner-ear structures of mafba −/− mutants were almost destroyed. The mutation caused defects in inner-ears in a recessive mode. cessive mode.
The inner ear plays crucial roles in zebrafish hearing and balance maintenance [53]. Due to the severe defects in the inner ears of mafba −/− mutants, we further investigated hearing abilities of the mutants. We assessed the fast escape reflex (named the C-shaped startle response) using near-field pure tone stimulation at two different levels of sound intensity [8]. The sibling and mafba +/-group larvae responded within 9 ms, but the mafba −/− larvae showed either a delayed response or no response at all at 5 dpf ( Figure 3A). The inner ear plays crucial roles in zebrafish hearing and balance maintenance [53]. Due to the severe defects in the inner ears of mafba −/− mutants, we further investigated hearing abilities of the mutants. We assessed the fast escape reflex (named the C-shaped startle response) using near-field pure tone stimulation at two different levels of sound intensity [8]. The sibling and mafba +/− group larvae responded within 9 ms, but the mafba −/− larvae showed either a delayed response or no response at all at 5 dpf ( Figure 3A).
The survival rates of mafba mutants are comparable with those of sibling and mafba +/-embryos up to 5 dpf. Most of the mafba −/− larvae showed inner-ear defects at 5 dpf. The survival rate of mafba −/− mutants decreased to 65% under normal feeding conditions, while the survival rate of mafba +/-embryos at 6 dpf was 92% (data not shown). This reduction may have been caused by the dietary deficiency resulting from the defective hearing and balance system [55].  For three genotypes, the numbers of immediate responses and delayed responses were recorded and are shown in Figure 3B. As described previously [54], we eliminated delayed responses and identified the percentage of successful C-shaped startle responses, and the result was that the probabilities of a C-startle response in sibling and mafba +/− embryos were indistinguishable and markedly higher than those of mafba −/− mutants ( Figure 3C). These data support the idea that mafba −/− mutant zebrafish embryos display the hearing loss.
The survival rates of mafba −/− mutants are comparable with those of sibling and mafba +/− embryos up to 5 dpf. Most of the mafba −/− larvae showed inner-ear defects at 5 dpf. The survival rate of mafba −/− mutants decreased to 65% under normal feeding conditions, while the survival rate of mafba +/− embryos at 6 dpf was 92% (data not shown). This reduction may have been caused by the dietary deficiency resulting from the defective hearing and balance system [55].

Deficiency of mafba Does Not Affect Otic Patterning or Hair Cell Development
The early development of the inner ear appears be normal in mafba −/− mutants. The induction of the otic placode appeared as expected at 18 hpf. In order to comprehend the molecular function of mafba in ontogenesis, we assessed the expression levels of several marker genes involved in the inner-ear patterning structure. The expressions of patterning markers for medial otic vesicle (pax2a), dorsal otic vesicle (dlx3b), and ventral otic vesicle (foxj1b) showed no difference between the sibling and mafba −/− mutant embryos ( Figure S6A-F). The expressions of the semicircular canal projections' marker udgh ( Figure S7G,H), the semicircular canal sensory cristae marker bmp4 ( Figure S7A,B), the utricular and saccular maculae marker cahz ( Figure S7C,D), and the endolymphatic duct marker foxi1 ( Figure S7E,F) were also the same in the mafba −/− embryos and the sibling. In addition, live bright-field images of siblings ( Figure S5A,C,E) and mafba −/− mutants ( Figure S5B,D,F) showed that the inner-ear sizes were comparable at 54 hpf and the semicircular canal projections were forming normally. There was no obvious difference between the mutant and the sibling at 78 hpf in the otocysts, semicircular canal, utricle, and saccule. The organogenesis of the inner ear appeared to be normal until 102 hpf, suggesting the correct specification of early otic tissue.
The formation and mineralization of the otolith is a complex process that requires the orderly regulation and participation of many developmental processes, such as the secretion of matrix proteins [8], hair cell development, and endolymphatic fluid homeostasis [20], which are necessary for zebrafish otolith formation and growth. The WISH results showed that the expressions' levels of the matrix protein genes otomp and starmaker (stm) at 2 dpf and 5 dpf in the sibling group were in accordance with those in mafba −/− embryos ( Figure S8A-H). Then, we tested the otic hair cells, which play essential roles in the development of the inner ear. There was no significant difference between the mafba −/− and sibling embryos as stained by the hair cell marker atp1b2b at 24 hpf ( Figure S9A). Immunostaining using the anti-Myo7a antibody indicated that the number of hair cells in mafba −/− and sibling embryos showed no difference in the utricular and saccular maculae at 36 and 48 hpf ( Figure S9B,C). Phalloidin staining for stereociliary bundles and kinocilium growth labeled by Acetylated-α-Tubulin both suggested that the differentiation of the sensory cristae and maculae was normal in mafba −/− mutant embryos at 5 dpf ( Figure S9D). In conclusion, these results implied that the patterning and structural specification of the otic vesicles, the differentiation and maturation of hair cells, and the growth of ciliary bundles were normal in the mafba-deficient embryos.

Knockout of mafba Suppresses the Proliferation of p63+ Epidermal Stem Cells and Reduces dlc+ Ionocyte Progenitor Cell Number
Mouse mafb plays an important role in regulating epidermal differentiation and homeostasis [56]. After confirming that matrix proteins and hair cell development were normal in the mafba −/− inner ear, the balance of ionocytes' homeostasis was explored. Epidermal stem cells were marked with p63 and are known to produce both keratinocytes (dlc − p63 + cells) and ionocytes (dlc + p63 + cells). Whether mafba is expressed in the epidermis and affects the development of ionocytes in zebrafish embryos needs further investigation. The p63 expression was observed in the ventral ectoderm of 90% epiboly and bud-stage embryos [18], while the expression of Mafba occurred in the bud stage and 5-somites stage. The p63 + Mafba + colocalized cell number was high in the dorsal ectoderm region at the bud stage and 5-somites stage ( Figure S10A). In addition, colocalization of Mafba and dlc was observed at the bud stage and 5-somites stages in the epidermal ionocyte region. Almost half of the dlc + cells were positively stained with Mafba ( Figure S10B). The colocalization of p63 + Mafba + and dlc + Mafba + suggests that Mafba participates in the development of epidermal stem cells and ionocytes.
Mafb is widely known for regulating epidermal keratinocyte differentiation and epidermal homeostasis in mammals [27]. We speculated that zebrafish Mafba also modulates p63 + epidermal stem cell proliferation and dlc + ionocyte progenitor differentiation. The p63 + epidermal stem cell number was reduced by 14.4% in mafba −/− mutants compared to the sibling group at the bud stage ( Figure 4A(g,m)), and we also saw a 19.6% reduction in the proportion of p63 + EDU + epidermal stem cells in mafba −/− embryos. These results suggest that Mafba is essential to maintaining the proliferation rate of p63 + epidermal stem cells. The reduction in p63 + epidermal stem cell led to a decreased number of dlc − p63 + keratinocytes and dlc + p63 + ionocytes in mafba −/− mutant embryos at the bud stage ( Figure 4A(g,h,j,m)). The percentage of EDU + signals colocalized with dlc − p63 + and dlc + p63 + was significantly decreased in mafba −/− mutant embryos at the bud stage ( Figure 4A(k,l,n)). A previous study suggested that foxi3a and foxi3b are the main regulators of epidermal ionocytes' specification in zebrafish embryos [20]. At the bud stage, a prominent reduction in ionocyte progenitors was detected by dlc labeling (Figure 4C). A significant decrease in the densities of ionocyte progenitors expressing either foxi3a or foxi3b was observed in mafba −/− embryos at the 5-somites stage ( Figure 4D). Mafba is, thus, essential to maintaining both the proliferation of p63 + epidermal stem cells and the cell numbers of dlc − p63 + keratinocytes and dlc + p63 + ionocytes.  The cell density of foxi3a + and foxi3b + ionocyte progenitors are compared between the sibling (a,d) and mafba −/− group (b,e) at the 5-somites stage, respectively. Cell densities of foxi3a + and foxi3b + ionocyte progenitors are quantified in (c) and (f), respectively. Scale bars: 100 µm. Data are expressed as mean ± SD; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001. The significant reduction in p63 + epidermal stem cells in mafba −/− embryos suggests that these could be arrested in the cell cycle. To determine the cell cycle status of p63 + epidermal stem cells in mafba −/− embryos, EdU incorporation (labeling cells in S-phase) combined with the immunofluorescence of phospho-histone3 (labeling cells in M-phase) was performed at the bud stage. The p63 + epidermal stem cell number is also decreased in mafba −/− embryos ( Figure 4B(e,i)) in accordance with the results shown in Figure 4A(m).
Double immunostaining showed that the proportion of p63 + pPH + cells decreased significantly, by 3.6%, in mafba −/− mutant embryos ( Figure 4B(g,h,j)). We thus speculated that the epidermal stem cells lacking Mafba may be arrested in G0/G1, and this may result in the increased apoptosis of these ionocyte progenitors. However, the TUNEL assays showed only a slight increase in apoptotic signal in the p63 + cells (by 1.73‰) and dlc + p63 + cells (by 6.71‰) in mafba −/− mutants compared to the sibling group, while no significant change was observed in dlc − p63 + cells ( Figure 5). In summary, we confirmed a reduction in the proliferation of p63 + epidermal stem cells and the cell numbers of dlc − p63 + keratinocytes and dlc + p63 + ionocytes as well as the differentiation of the epidermal ionocytes' progenitor in mafba −/− embryos.

Apoptosis of the Differentiated Epidermal Ionocytes Is Increased in mafba −/− Embryos
A previous study suggested that foxi3a and foxi3b control ionocyte progenitor specification into NR and HR cells through a positive feedback loop [24]. Ionocyte progenitors' regulators, such as dlc, foxi3a, and foxi3b, were downregulated in mafba −/− mutant embryos. We, thus, wished to assess whether the formation of NR and HR cells is affected by deprivation of mafba expression. We performed WISH using relevant markers (atp1a1a.1, atp1a1a.4, and atp1b1a for NR cells, atp6v1aa for HR cells) to evaluate the differentiated epidermal ionocytes in the inner ear. At 24 hpf, the expressions of atp1a1a.1, atp1a1a.4, atp1b1a, and atp6v1aa in the mafba −/− embryos reduced markedly compared to those in the siblings ( Figure 6A). Similarly, a significant reduction in atp1a1a.4 and atp6v1aa expression in the inner ear was observed in mafba −/− mutants from 2 dpf to 5 dpf ( Figure 6B,C), respectively. Considering the dramatic decline in atp1a1a.4-and atp6v1aa-expressing cells in the inner ear at 5 dpf, we suspected that the apoptosis may also occur in mafba mutants. To validate this hypothesis, we measured the apoptosis of atp1a1a.4-and atp6v1aa-expression cells in mafba −/− embryos using TUNEL assays. There was a significant increase in apoptotic atp1a1a.4 and atp6v1aa cells' signals in the inner-ear region in mafba −/− embryos compared with those in the siblings at 4 dpf and 5 dpf ( Figure 6B,Bj,C,C-j). The percentage of apoptotic cells with atp1a1a.4 and atp6v1aa at 5 dpf was notably higher than that at 4 dpf. The defective inner ear might result from the decreased The quantitative analyses of p63 + cells, dlc − p63 + cells, and dlc + p63 + cells of sibling and mafba −/− embryos at the bud stage are shown in (a). The quantitative analysis of p63 + , dlc − p63 +, and dlc + p63 + cell numbers colocalized with TUNEL in sibling and mafba −/− embryos at the bud stage are shown in (b). The n = 12 for each panel. Scale bars: 100 µm. Data are mean ± SD; ns, p > 0.05; *, p < 0.05; **, p < 0.01; ***, p < 0.001.

Apoptosis of the Differentiated Epidermal Ionocytes Is Increased in mafba −/− Embryos
A previous study suggested that foxi3a and foxi3b control ionocyte progenitor specification into NR and HR cells through a positive feedback loop [24]. Ionocyte progenitors' regulators, such as dlc, foxi3a, and foxi3b, were downregulated in mafba −/− mutant embryos. We, thus, wished to assess whether the formation of NR and HR cells is affected by deprivation of mafba expression. We performed WISH using relevant markers (atp1a1a.1, atp1a1a.4, and atp1b1a for NR cells, atp6v1aa for HR cells) to evaluate the differentiated epidermal ionocytes in the inner ear. At 24 hpf, the expressions of atp1a1a.1, atp1a1a.4, atp1b1a, and atp6v1aa in the mafba −/− embryos reduced markedly compared to those in the siblings ( Figure 6A). Similarly, a significant reduction in atp1a1a.4 and atp6v1aa expression in the inner ear was observed in mafba −/− mutants from 2 dpf to 5 dpf ( Figure 6B,C), respectively. Considering the dramatic decline in atp1a1a.4-and atp6v1aa-expressing cells in the inner ear at 5 dpf, we suspected that the apoptosis may also occur in mafba mutants. To validate this hypothesis, we measured the apoptosis of atp1a1a.4-and atp6v1aa-expression cells in mafba −/− embryos using TUNEL assays. There was a significant increase in apoptotic atp1a1a.4 and atp6v1aa cells' signals in the inner-ear region in mafba −/− embryos compared with those in the siblings at 4 dpf and 5 dpf ( Figure 6B(j),C(j)). The percentage of apoptotic cells with atp1a1a.4 and atp6v1aa at 5 dpf was notably higher than that at 4 dpf. The defective inner ear might result from the decreased ionocyte progenitors and increased apoptosis in NR and HR cells. In summary, Mafba is essential to the proliferation of the epidermal stem cells and ionocyte progenitors, which, in turn, determine the numbers of NR and HR cells and maintain ion homeostasis in zebrafish inner ear. The differentiated ionocyte progenitors provide a stable environment for the development of statoacoustic organs such as otoliths and otocysts, which are essential to the hearing and balance systems ( Figure 6D). However, more research is needed to reveal the specific regulatory mechanism of inner-ear developmental defects.

Deprivation of mafba Activates p53 Apoptosis Pathway and Arrests the Cell Cycle in the G0/G1 Phase
In order to determine how Mafba deprivation triggers apoptosis, we quantified the expression levels of several apoptosis-related genes, including caspase8, caspase10, and p53, and its target genes in zebrafish embryos. The expressions of p53 and its target genes baxa, mdm2, and puma, along with apoptosis-related genes caspase10 and caspase8, showed

Deprivation of mafba Activates p53 Apoptosis Pathway and Arrests the Cell Cycle in the G0/G1 Phase
In order to determine how Mafba deprivation triggers apoptosis, we quantified the expression levels of several apoptosis-related genes, including caspase8, caspase10, and p53, and its target genes in zebrafish embryos. The expressions of p53 and its target genes baxa, mdm2, and puma, along with apoptosis-related genes caspase10 and caspase8, showed significant increases in mafba −/− mutants at 4 dpf. In addition, the expressions of mdm4 and bcl2a, typical inhibitors of p53, decreased significantly ( Figure 7A). Based on the observation that the proliferation of epidermal stem cells was suppressed, the p53 pathway was activated, and the apoptosis of NR and HR cells was increased, we hypothesize that G0/G1 cell cycle arrest may also occur in mafba mutants. A previous study revealed that mouse Mafb promotes cell proliferation with detectable changes in cell cycle factors [57]. Cell cycle distribution was examined via flow cytometry. The proportions of G0/G1 phase cells increased significantly, while S and G2/M phase cells decreased in the mafba −/− mutant at 4 dpf ( Figure 7B). The knockout of mafba led to G0/G1 cell cycle arrest, suggesting it affects cell cycle factors. significant increases in mafba −/− mutants at 4 dpf. In addition, the expressions of mdm4 and bcl2a, typical inhibitors of p53, decreased significantly ( Figure 7A). Based on the observation that the proliferation of epidermal stem cells was suppressed, the p53 pathway was activated, and the apoptosis of NR and HR cells was increased, we hypothesize that G0/G1 cell cycle arrest may also occur in mafba mutants. A previous study revealed that mouse Mafb promotes cell proliferation with detectable changes in cell cycle factors [57]. Cell cycle distribution was examined via flow cytometry. The proportions of G0/G1 phase cells increased significantly, while S and G2/M phase cells decreased in the mafba −/− mutant at 4 dpf ( Figure 7B). The knockout of mafba led to G0/G1 cell cycle arrest, suggesting it affects cell cycle factors.
A number of cell cycle factors that control G1/S phase transition were analyzed via RT-qPCR. The expressions of cdk2, cdk4, cdk6, cyclin d1 (ccnd1), and cyclin e1 (ccne1) were downregulated. This agrees with the results that the protein levels of CDK6, Cyclin D1, and Cyclin E1 were lower in mafba −/− mutants than in sibling embryos at 4 dpf ( Figure  7C,E). Meanwhile, the cycle regulators of the G1/S checkpoint, such as cnkn1a (p21), cdkn1b (p27), cdkn3, cul1a, and cul3, were upregulated, while mycb, dhfr, e2f1, e2f3, foxa1, rangap1, tk1, and rb1 were downregulated in mafba −/− mutants at 4 dpf ( Figure 7C,D). These results indicated that cell cycle arrest occurred between the G1 and S phases in the mutants. The reduced protein levels of Pcna also confirmed the arrest of G1/S phase transition. These results suggest that Mafba is essential for cell proliferation via the regulation of cell cycle factors levels.

Accumulation of DNA Damages in mafba −/− Inner-Ear
The apoptosis of the differentiated ionocytes may have resulted from the activation of the p53 pathway and/or cell cycle arrest at 4 dpf in mafba −/− inner ear. DNA damage may also occur in mafba −/− mutants, which would explain the apoptosis. The immunostaining of γH2AX indicated the DNA damage in sibling and mafba −/− inner ears at A number of cell cycle factors that control G1/S phase transition were analyzed via RT-qPCR. The expressions of cdk2, cdk4, cdk6, cyclin d1 (ccnd1), and cyclin e1 (ccne1) were downregulated. This agrees with the results that the protein levels of CDK6, Cyclin D1, and Cyclin E1 were lower in mafba −/− mutants than in sibling embryos at 4 dpf ( Figure 7C,E). Meanwhile, the cycle regulators of the G1/S checkpoint, such as cnkn1a (p21), cdkn1b (p27), cdkn3, cul1a, and cul3, were upregulated, while mycb, dhfr, e2f1, e2f3, foxa1, rangap1, tk1, and rb1 were downregulated in mafba −/− mutants at 4 dpf ( Figure 7C,D). These results indicated that cell cycle arrest occurred between the G1 and S phases in the mutants. The reduced protein levels of Pcna also confirmed the arrest of G1/S phase transition. These results suggest that Mafba is essential for cell proliferation via the regulation of cell cycle factors levels.

Accumulation of DNA Damages in mafba −/− Inner-Ear
The apoptosis of the differentiated ionocytes may have resulted from the activation of the p53 pathway and/or cell cycle arrest at 4 dpf in mafba −/− inner ear. DNA damage may also occur in mafba −/− mutants, which would explain the apoptosis. The immunostaining of γH2AX indicated the DNA damage in sibling and mafba −/− inner ears at 3, 4, and 5 dpf ( Figure 8A,B). The γH2AX-positive cells were observed in mafba −/− mutant inner ears at 3 dpf. The numbers of γH2AX-labeled cells in mafba −/− embryos inner ears increased dramatically at 4 and 5 dpf, compared to those in the sibling. DNA single/double-strand break levels were assessed directly via an alkaline comet assay [58]. We also detected more DNA breaking signals in mafba −/− mutants at 4 and 5 dpf ( Figure 8C,D). Meanwhile, Western blot analysis also showed that the expression levels of γH2AX, P-ATM, P-ATR, and p53 accumulated gradually in mafba −/− embryos ( Figure 8E). These results suggest that Mafba is essential for preventing DNA damage and maintaining genomic stability in differentiated ionocytes.
Biomedicines 2021, 9, x FOR PEER REVIEW 16 of 23 3, 4, and 5 dpf (Figure 8 A,B). The γH2AX-positive cells were observed in mafba −/− mutant inner ears at 3 dpf. The numbers of γH2AX-labeled cells in mafba −/− embryos inner ears increased dramatically at 4 and 5 dpf, compared to those in the sibling. DNA single/double-strand break levels were assessed directly via an alkaline comet assay [58]. We also detected more DNA breaking signals in mafba −/− mutants at 4 and 5 dpf ( Figure  8C,D). Meanwhile, Western blot analysis also showed that the expression levels of γH2AX, P-ATM, P-ATR, and p53 accumulated gradually in mafba −/− embryos ( Figure 8E). These results suggest that Mafba is essential for preventing DNA damage and maintaining genomic stability in differentiated ionocytes.

Discussion
In the present study, we characterized a zebrafish mafba knockout line generated using CRISPR/Cas9 technology. The inner ears of mutants showed specific defects, including enlarged otocysts, small or absent otoliths, and insensitivity to sound stimulation, whereas matrix protein expression and hair cells' development appeared to be normal. Genetically, mafba plays a positive role in epidermal stem cells' proliferation and can also promote the differentiation of the ionocyte progenitor in zebrafish inner ears. It also takes part in the differentiation of NR and HR cells, which are necessary for the ion homeostasis and epidermis differentiation.
The mouse kreisler/mafb is a member of the Maf transcription factor family. When mutated, it is identified as the causative gene of physiological defects resulting in segmentation abnormalities in the caudal hindbrain and defective inner-ear development [59]. In various vertebrates (including zebrafish), there are two paralogous othologues of mammalian mafb, i.e., mafba and mafbb [60,61]. Progressive syndromic deafness caused by

Discussion
In the present study, we characterized a zebrafish mafba knockout line generated using CRISPR/Cas9 technology. The inner ears of mutants showed specific defects, including enlarged otocysts, small or absent otoliths, and insensitivity to sound stimulation, whereas matrix protein expression and hair cells' development appeared to be normal. Genetically, mafba plays a positive role in epidermal stem cells' proliferation and can also promote the differentiation of the ionocyte progenitor in zebrafish inner ears. It also takes part in the differentiation of NR and HR cells, which are necessary for the ion homeostasis and epidermis differentiation.
The mouse kreisler/mafb is a member of the Maf transcription factor family. When mutated, it is identified as the causative gene of physiological defects resulting in segmentation abnormalities in the caudal hindbrain and defective inner-ear development [59]. In various vertebrates (including zebrafish), there are two paralogous othologues of mammalian mafb, i.e., mafba and mafbb [60,61]. Progressive syndromic deafness caused by heterozygous loss-of-function Mafb mutations was identified in a large family, though severity and age of onset differed among individuals. The mutation was located in the DNA-binding domain [33,34]. Researchers have been unable to characterize this inner-ear defect in mice because mutations in the mafb locus are lethal in the early postnatal period [62]. The locus of the mutation in our mafba zebrafish line is similar to that of the 0819 mutation in human MAFB [34], which explains why mafba +/− mutants have no effect on inner-ear defects. A mouse model with conditional knockout Mafb in the otic epithelial cells may help to illustrate the function of mafba in mammalian inner-ear development. Previous studies have shown that Mafb not only promoted the differentiation of otic epithelial cells, but also acted in the gene expression program that positively regulates epidermal keratinocyte differentiation [27]. Our research revealed that zebrafish mafba −/− mutant embryos exhibit inner-ear defects, with variable expressivity between individuals. The mutants with the more severe phenotypes possess bigger otocysts and smaller otoliths. Therefore, the mutant zebrafish mafba −/− line is the optimum animal model for progressive hearing loss in humans caused by mafb mutation, and can be applied to further study the mechanisms underlying mafb-associated inner-ear developmental defects.
Mafba plays crucial roles in various tissues and developmental stages. Based on a previous study and our research, zebrafish mafba is expressed in the otic vesicles, hindbrain, renal system, visual system, and other tissues during embryogenesis [36,63]. Developmental abnormalities in the inner ear were first detected in mafba −/− mutants at 5 dpf. Human progressive hearing loss caused by Mafb mutation is syndromic with other detectable defects, such as atopic dermatitis (AD), psoriasis vulgaris [27], aberrant extraocular muscle innervation, MCTO, and focal segmental glomerulosclerosis (FSGS) with DRS [33,34]. Zebrafish Mafba and human Mafb share a highly conserved amino acid sequence. The similarity in the inner-ear phenotype between zebrafish and human mafb-deficient individuals implies that mafba may be essential to inner-ear development and the maintenance of hearing capacity. In addition, mafba −/− mutants need to be examined at later stages of development to determine whether other tissues/organs are affected.
It is widely known that apical junctional complex proteins (AJC), extracellular matrix (ECM), and ion channels in the otic epithelial/epiderma cells act as barriers and are important to homeostasis maintenance, which is essential to normal primary inner-ear formation and subsequent development. The adhesion class G protein-coupled receptor (Gpr126) acts through a cAMP-mediated pathway to control the outgrowth and adhesion of canal projections in the zebrafish inner ear via the regulation of ECM gene expression [64]. The secretion of ECM molecules drives the growth of the epithelial projections and regulates the morphogenesis of the semicircular canals [65]. Regulated fluid secretion is crucial for many organs. The loss of chloride channel cystic fibrosis transmembrane conductance regulator (CFTR)-mediated fluid secretion impairs Kupffer's vesicle (KV) lumen expansion, leading to defects in organ laterality [66]. Interestingly, a previous study showed that fluid accumulation increases hydrostatic pressure in the lumen. Consequently, the stress passes into the epithelium and expands the otic vesicle. In turn, this pressure inhibits fluid transport into the lumen. This negative feedback between pressure and transport allows the otic vesicle to adjust its growth rate, resulting in otic vesicle size variations [67]. A previous study showed that mouse Kreisler/Mafb and zebrafish valentino/mafba mutants display early inner-ear defects that are related to abnormalities in the hindbrain development [36,68]. Facial motor neurons will fail to migrate through r5/r6 and complete caudal migration in zebrafish embryos with mafba deficiencies. [69]. The expression of genes in r5/r6 is regulated by RA, FGF, hnf1ba, and valentino (mafba); any losses of these factors will abolish r5/r6 gene expression [70]. Mafba is also essential to abducens' motor neuron development in zebrafish [71]. It has also been shown that the expression of fgf3 in the hindbrain rhombomere is essential to the dorsal otic vesicle pattern in valentino mutant zebrafish, and its primary role is to maintain and focus the expression of the dorsal inner-ear gene that is induced by WNT signals [72]. Both of these reports imply that the otic vesicle pattern is regulated or influenced by hindbrain development. Interestingly, other researchers have provided evidence that fgf3 is not sufficient to guide otic regionalization [35]. However, both of these studies focused on early inner-ear defects caused by cascading effects of the external environment, such as abnormal hindbrain development. Whether there is any direct internal mechanism responsible for the abnormal inner-ear structure is yet to be determined.
The transcription factor Mafb plays a crucial role in epidermal keratinocyte differentiation in the mammalian epidermis by regulating downstream key transcription factors, such as grhl3, znf750, klf4, and prdm1 [73,74] The ectopic expression of Mafb in basal keratinocytes results in excessive proliferation and disturbs epidermal homeostasis [56]. The terrestrial vertebrate epidermis is a stratified epithelium consisting of distinct layers of keratinocytes, which build a functional osmotic barrier that prevents dehydration. In contrast, the fish epidermis is constituted of keratinocytes and ionocytes, which transport ions and acid-base substances to maintain the homeostasis of bodily fluids [75]. Most, if not all, studies of Mafb have focused on the homeostasis of epidermal keratinocytes and pay no attention to the ionocytes. In zebrafish embryos, epidermal stem cells specify and differentiate into ionocytes and keratinocytes during the bud stage, which is controlled by the Dlc-Notchmediated lateral inhibition [18,75]. Our data revealed that Mafba helps to maintain the ionocyte progenitor population by regulating epidermal stem cell proliferation, resulting in fewer stem cells and a decreasing number of differentiated ionocyte progenitors, and then decreases NR and HR cells in the inner-ear region. Finally, the loss of homeostasis could impede the volume control of the otocysts and the growth of otoliths. Collectively, these defects lead to hearing and balance problems ( Figure 6D). Previous studies have shown that klf4, which serves as a downstream regulator of mafba, maintains the ionocyte progenitor population via dlc-mediated lateral inhibition [18]. Taken together, our results clarify the novel and key role of mafba in the maintenance of ionocytes and ion homeostasis during zebrafish embryogenesis.
Interestingly, unlike in mice [37], the apoptosis of NR and HR cells increases markedly in mafba −/− mutant inner ears compared with those of sibling embryos from 4 dpf. The discrepancy between mice and zebrafish phenotypes may stem from differences between species or it might also be attributed to diversities between ionocytes and other cells. The reduction in NR and HR cells was due to cell apoptosis, and this led to inner-ear ion homeostasis imbalances, which eventually resulted in the mafba −/− mutant inner-ear defects at 5 dpf. Follow-up studies identified that the activation of p53-mediated apoptosis and cell cycle arrest at the G0/G1 phase in mafba −/− embryos at 4 dpf and DNA damage appeared in the inner-ear area from 3 dpf to 5 dpf. DNA damage is a relatively common event that leads to the activation of DNA damage checkpoints, such as ATM and ATR. These signal transducers activate p53 and deactivate cyclin-dependent kinases to inhibit cell cycle progression [76]. These results imply that DNA damage is the major factor causing p53 activation in response to zebrafish mafba deletion. Our results show that mafba depletion leads to a reduction in CDK6 protein level and transcriptional activity, accompanied with an accumulation of p53 mRNA and protein expression, which may be the mechanism of cell cycle arrest. Previous research has also shown that mafb deprivation destroys direct downstream regulator cyclin-dependent kinase 6 (CDK6) transcription and impedes clinical colorectal cancer (CRC) cell proliferation, as a result of the cell cycle arrest at the G0/G1 phase [37]. The specific deletion of mafb suppresses nasopharyngeal carcinoma cell (NCC) proliferation [77]. Overexpression of mafb enhances cell foci formation and increases cyclin B1 and D2 expression [40]. These studies suggest that DNA damage, p53 pathway activation, and the impairment of cell proliferation are responsible for the ion channel homeostasis imbalances in mafba −/− mutant inner ears. Aside from the ion channels, junction proteins are important for the homeostasis maintenance and inner-ear development [8,9]. Therefore, it is expected that mafba will exert its function on inner-ear development through the regulation of the junction proteins' factor. Further studies may focus on identifying the downstream targets among these candidates in order to clarify the specific mechanism of zebrafish mafba −/− mutant inner-ear defects.
In summary, we established a mafba knockout zebrafish line displaying auditory impairment for the first time. We elucidated the roles of Mafba in maintaining ion channel homeostasis in the zebrafish inner ear via the control of the proliferation of the ionocyte progenitor and the populations of differentiated NR and HR cells. This study provides novel insights into the inner-ear pathogenic mechanisms of Mafba and offers an ideal model for identifying new therapeutic interventions for inner-ear defects.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/biomedicines9111699/s1. Figure S1: The domain features and homologous comparison of the transcription factor Mafb among mammals and zebrafish. Figure S2: The mafba mRNA levels were detected in 4-dpf sibling and mafba mutant zebrafish by qRT-PCR. Figure S3. Relative expressions of large maf (mafba, mafbb, mafa, c-maf, and nrl) and small maf (maff, mafk, mafga, and mafgb) were determined by qRT-PCR analysis in sibling, mafba +/−, and mafba −/− embryos at 4 dpf. Figure S4: Expression of the zebrafish mafba gene during embryonic development in siblings and its corresponding homozygous mutants. Figure S5: Early patterning of the inner-ear appears normal in mafba −/− mutants. Figure S6: Early patterning of the otic vesicle appears normal in mafba −/− mutants. Figure S7: The expression of otic specification-related markers was normal in mafba −/− mutant embryos. Figure S8: The expression of two otic matrix protein markers was normal in mafba −/− mutant embryos at 2 and 5 dpf. Figure S9: Deficiency of mafba does not affect hair cell development. Figure S10: Mafba protein expression pattern and colocalization of Mafba/p63 and Mafba/dlc during late-gastrulation and early-somite stages. Table S1: Primers used for Cas9 target design, WISH probes construction, genotype identification, and qRT-PCR.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.