Editor's evaluation: Kiaa1024L/Minar2 is essential for hearing by regulating cholesterol distribution in hair bundles
Year: 2022
Type: peer-review
Abstract: Article Figures and data Abstract Editor's evaluation eLife digest Introduction Results Discussion Materials and methods Appendix 1 Data availability References Decision letter Author response Article and author information Metrics Abstract Unbiased genetic screens implicated a number of uncharacterized genes in hearing loss, suggesting some biological processes required for auditory function remain unexplored. Loss of Kiaa1024L/Minar2, a previously understudied gene, caused deafness in mice, but how it functioned in the hearing was unclear. Here, we show that disruption of kiaa1024L/minar2 causes hearing loss in the zebrafish. Defects in mechanotransduction, longer and thinner hair bundles, and enlarged apical lysosomes in hair cells are observed in the kiaa1024L/minar2 mutant. In cultured cells, Kiaa1024L/Minar2 is mainly localized to lysosomes, and its overexpression recruits cholesterol and increases cholesterol labeling. Strikingly, cholesterol is highly enriched in the hair bundle membrane, and loss of kiaa1024L/minar2 reduces cholesterol localization to the hair bundles. Lowering cholesterol levels aggravates, while increasing cholesterol levels rescues the hair cell defects in the kiaa1024L/minar2 mutant. Therefore, cholesterol plays an essential role in hair bundles, and Kiaa1024L/Minar2 regulates cholesterol distribution and homeostasis to ensure normal hearing. Editor's evaluation This is an important study linking cholesterol homeostasis to sensory hair cell function. Using the knockout approach in zebrafish, the authors provided compelling evidence that Minra2, a gene known to be associated with deafness in humans, encodes a protein that functions to regulate cholesterol homeostasis in the hair bundle of sensory hair cells. https://doi.org/10.7554/eLife.80865.sa0 Decision letter Reviews on Sciety eLife's review process eLife digest Cholesterol is present in every cell of the body. While it is best known for its role in heart health, it also plays a major role in hearing, with changes in cholesterol levels negatively affecting this sense. To convert sound waves into electrical brain signals, specialised ear cells rely on hair-like structures which can move with vibrations; cholesterol is present within these hair ‘bundles’, but its exact role remains unknown. Genetic studies have identified over 120 genes essential for normal hearing. Animal data suggest there may be many more – including, potentially, some which control cholesterol. For instance, in mice, loss of the Minar2 gene causes profound deafness. Yet exactly which role the protein that Minar2 codes for plays in the ear remains unknown. This is in part because that protein does not resemble any other related proteins, making it difficult to infer its function. To find out more, Gao et al. investigated loss of minar2 in zebrafish, showing that deleting the gene induced deafness in the animals. Without minar2, the hair bundles in ear cells were longer, thinner, and less able to sense vibrations: cholesterol could not move into these structures, causing them to dysfunction. Exposing the animals to drugs that lower or raise cholesterol levels respectively worsened or improved their hearing abilities. A recent study revealed that mutations in MINAR2 also cause deafness in humans. The findings by Gao et al. highlight the need for further research which explores the role of cholesterol and MINAR2 in hair bundle function, as this may potentially uncover cholesterol-based treatments for hearing problems. Introduction Hearing loss is one of the most common disabilities in humans (World-Health-Organization, 2021). Studies of genes linked to non-syndromic deafness have identified over 120 genes essential for normal hearing (Shearer et al., 1993; Van Camp and Smith, 2021). This large collection of genes likely reflects the developmental and physiological complexity of the vertebrate auditory system. Large-scale, unbiased genetic screens in the mouse (Bowl et al., 2017; Ingham et al., 2019) and the zebrafish (Whitfield et al., 1996) model systems have additionally identified multiple hearing loss genes. Interestingly, a number of these phenotypically identified hearing loss genes are previously uncharacterized and have no demonstrated functional roles, hinting that some biological processes required for normal auditory function have not been sufficiently explored (Bowl et al., 2017; Ingham et al., 2019). The auditory hair cells are the sensory receptor cells that convert acoustic and mechanical stimuli into electrical signals initiating hearing (Fettiplace, 2017; Hudspeth, 1989; Ó Maoiléidigh and Ricci, 2019). The hair bundle, a specialized organelle situated at the apical surface of each hair cell, responds to mechanical displacement in a direction-dependent manner (Flock, 1964; Tilney et al., 1992). The hair bundle is essential for mechanoelectrical transduction (MET), and defects in the hair bundle can cause hearing loss (Belyantseva et al., 2009; Blanco-Sánchez et al., 2014; Kozlov et al., 2007; Noben-Trauth et al., 2003; Perrin et al., 2013). Several proteins, such as adhesion molecules, actin-bundling proteins, and disease-associated proteins are required for the morphogenesis and physiological regulation of the hair bundle (Barr-Gillespie, 2015; Blanco-Sánchez et al., 2017; McGrath et al., 2017; Tilney et al., 1980). In addition, specific lipid molecules may play a role in the hair bundle. For instance, phosphatidylinositol-4,5-bisphosphate (PIP2) is localized to hair bundles, and it binds to the MET channel component TMIE (Cunningham et al., 2020) and regulates the rates of fast and slow adaptation (Effertz et al., 2017; Hirono et al., 2004). Cholesterol is an important component of eukaryotic cell membranes, and it controls membrane stiffness, tension, fluidity, and other membrane properties (Maxfield and van Meer, 2010; Subczynski et al., 2017). Cholesterol also plays a regulatory function by interacting with membrane proteins (Harris, 2010). Previous studies showed that abnormally high or low cholesterol levels are detrimental to hearing (Corwin and Warchol, 1991; Crumling et al., 2012; Ding et al., 2020; Guo et al., 2005; King et al., 2014a; King et al., 2014b; Morizono and Paparella, 1978; Sikora et al., 1986; Takahashi et al., 2016; Thoenes et al., 2015; Xing et al., 2015; Yao et al., 2019). Early studies also showed that cholesterol is not uniformly distributed in the hair cell membranes. The intensities of cholesterol labeling the hair cells were higher in the apical membranes when compared to those of the lateral membranes (Nguyen and Brownell, 1998; Takahashi et al., 2016), and freeze-fracture images of hair cells suggested that the stereocilia membrane was densely covered with cholesterol (Forge et al., 1988). Nevertheless, the distribution of cholesterol in hair cells in vivo, and cholesterol’s functional role in hair cells are not well characterized. Kiaa1024L/Minar2, a previously understudied gene, was identified in a hearing loss screen in mouse knockout strains using the auditory brainstem response test. The homozygous Kiaa1024L/Minar2 knockout mice aged 14 weeks old had severely raised ABR thresholds at all frequencies tested (Bowl et al., 2017; Ingham et al., 2019), but how Kiaa1024L/Minar2 functioned in the hearing was unclear. Here we show kiaa1024L/minar2 is expressed in the zebrafish mechanosensory hair cells, and disruption of kiaa1024L/minar2 causes hearing loss in the zebrafish larvae. We next show that GFP or FLAG-tagged Kiaa1024L/Minar2 protein is distributed in the stereocilia and the apical endo-membranes. Defects in mechanotransduction, longer and thinner hair bundles, and enlarged apical lysosomes are observed in hair cells in kiaa1024L/minar2 mutant. In vitro studies in cultured cells show that the Kiaa1024L/Minar2 protein is mainly localized to lysosomes, and overexpression of Kiaa1024L/Minar2 recruits cholesterol and results in increased intracellular cholesterol levels. Strikingly, we show cholesterol is highly enriched in the hair bundle membranes, and loss of kiaa1024L/minar2 reduces cholesterol distribution in the hair bundles. Drug treatment that lowers cholesterol levels aggravates, whereas treatment that raises cholesterol levels rescues hair cell defects and hearing in the mutant kiaa1024L/minar2 larvae. Together, our results indicate cholesterol plays an essential role in the hair bundles, and Kiaa1024L/Minar2 regulates cholesterol distribution and homeostasis in auditory hair cells to ensure normal hearing. Results kiaa1024L/minar2 is required for normal hearing in the zebrafish Kiaa1024L/minar2 gene orthologs are found in vertebrate species only (Figure 1—figure supplement 1A). It belongs to the UPF0258 gene family, which also includes kiaa1024/minar1/ubtor (Ho et al., 2018; Zhang et al., 2018). Based on genome annotations and BLAST search results, there are two kiaa1024/minar1/ubtor gene orthologs (named ubtora and ubtorb in Zhang et al., 2018), and a single kiaa1024l/minar2 gene ortholog in the zebrafish genome. To conform to current human gene nomenclature, kiaa1024L/minar2 gene orthologs are referred to as minar2 hereafter. We surveyed available sequencing data (Barta et al., 2018; Elkon et al., 2015; Erickson and Nicolson, 2015; Liu et al., 2018) and found the transcripts of minar2 orthologs were highly enriched in the auditory hair cells of the mouse and the zebrafish (Figure 1—figure supplement 1B). In a human inner ear organoid model (Steinhart et al., 2022), MINAR2 is specifically expressed in differentiated hair cells, similar to known differentiated hair cell markers (Figure 1—figure supplement 1B). Consistent with these sequencing-based data, in situ hybridization results confirmed minar2 was specifically expressed by the hair cells of the inner ears and the lateral line neuromasts in the developing zebrafish (5 dpf, days post fertilization, Figure 1A and Figure 1—figure supplement 1A). Figure 1 with 2 supplements see all Download asset Open asset minar2 is required for normal hearing in the zebrafish. (A) RNA in situ hybridization results showed that minar2 was specifically expressed by the hair cells of the inner ears and the lateral line neuromasts (5 dpf). Arrows point to hair cells. An asterisk in the upper panel marks a head neuromast located next to the inner ear. AC: anterior crista; LC: lateral crista; PC: posterior crista; UM: utricular macula; SM: saccular macula. (B) C-start response rates for wild type and homozygous minar2fs139 mutants at 8 dpf (n=63 and 64, respectively. ****p<0.0001, Mann-Whitney test). (C) Auditory evoked potentials (AEP) thresholds in wild type and the minar2fs139 mutants (n=11 and 13, respectively. For 100-, 200-, and 400 Hz, **p<0.01). (D) Evaluation of mechanotransduction by AM 1–43 staining. The lateral line L3 neuromasts of 5 dpf and 8 dpf larvae were imaged and quantified (for 5 dpf, n=35 and 36, t=7.465, df = 64.84, ****p<0.0001; for 8 dpf, n=49 and 51, t=6.444, df = 86.90, ****p<0.0001). (E) Quantification of hair cell numbers by counting the myo6:Gal4FF;UAS-EGFP-positive cells in lateral line L3 neuromast (for 5 dpf, n=30 and 32, t=2.578, df = 59.93, *p=0.0124; for 8 dpf, n=58 and 58, t=4.148, df = 114, ****p<0.0001). (F) Quantification of hair cell numbers in the inner ears of zebrafish adult. Hair bundles in dissected utricles (upper panels) and saccules (lower panels) were labeled with fluorescence-conjugated phalloidin. Diagrams of a utricle and saccule on the left. Numbered boxes (1-3) in the diagrams indicate the positions of imaged and counted areas (for utricles, n=15 and 19; for saccules, n=12 and 9. *p<0.05, **p<0.01, ***p<0.001). A: anterior; L: lateral; P: posterior; V: ventral. Scale bars represent 25 μm (A), and 10 μm (D, E, and F). Figure 1—source data 1 Functional requirement and expression of minar2 in hair cells Figure 1B-FFigure 1—figure supplement 1B, D, E; Figure 1—figure supplement 2A-C. https://cdn.elifesciences.org/articles/80865/elife-80865-fig1-data1-v2.xlsx Download elife-80865-fig1-data1-v2.xlsx To study minar2 function in the zebrafish, we generated minar2 mutant alleles by CRISPR/Cas9-mediated mutagenesis. The mutation in the minar2fs139 allele was a 5 bp deletion in exon 1, and in the minar2fs140 allele was a 5 bp insertion in exon 1 (Figure 1—figure supplement 1C). Both mutations were frameshift mutations and led to premature termination of protein translation. The translatable protein sequence in either mutant was very short and lacked the transmembrane helix located at the carboxyl terminus of the protein (Figure 1—figure supplement 1C). The mutant Minar2fs139 protein was predicted to translate to the 25th amino acid, then adds 28 code-shifted residues before the reading frame stopped. The mutant Minar2fs140 protein was predicted to translate to the 26th amino acid, then adds 55 code-shifted residues. Thus, both mutants were expected to be loss of function alleles. In addition, quantitative PCR analyses showed the expression levels of minar2 in the two mutants were markedly down-regulated in the developing zebrafish, most likely because the premature translational terminations activated the nonsense-mediated mRNA decays (Figure 1—figure supplement 1D). To assess if loss of minar2 resulted in genetic compensation (El-Brolosy et al., 2019; Ma et al., 2019) by upregulating kiaa1024/ubtor/minar1 gene expression, we carried out quantitative PCR and found that expression levels of ubtora/minar1a and ubtorb/minar1b were not changed in the minar2fs139 nor the minar2fs140 mutants (Figure 1—figure supplement 1E). We subsequently focused our investigation using the minar2fs139 allele, and in some experiments we corroborated our findings using the minar2fs140 allele. We first examined the short-latency C-start (SLC) response evoked by auditory stimuli (Burgess and Granato, 2007; Wolman et al., 2011) to assess the hearing abilities of zebrafish larvae. We found the SLC response rates to a 200 Hz stimulus were significantly reduced in the minar2fs139 mutant at 8 dpf (Figure 1B, median response rates: 80% in the wild type, and 60% in the mutant. P<0.0001, Mann-Whitney test). To further analyze the hearing sensitivity of zebrafish larvae, we followed a procedure similar to the auditory brainstem response recording (Higgs et al., 2003; Higgs et al., 2002; Wang et al., 2015), and recorded auditory evoked potentials (AEP) in 7–8 dpf zebrafish larvae (Figure 1C). Two-way repeated-measures ANOVA revealed that the minar2fs139 mutant had significantly elevated AEP thresholds compared with the wild type control (For genotype factor, F(1, 22)=7.457, p=0.0122. For genotype x frequency, F(5, 110)=6.072, p<0.0001). The AEP thresholds were significantly higher at 100–400 Hz tone bursts in the minar2fs139 mutant (for 100 Hz tone, 151.0±1.2 dB for minar2fs139, 145.7±1.4 dB for wild type control; for 200 Hz tone, 152.2±1.1 dB for minar2fs139, 145.2±2.0 dB for wild type control; for 400 Hz, 143.6±1.1 dB for minar2fs139, 138.6±1.3 dB for wild type control). These results indicate that minar2 is required for normal hearing in the zebrafish larvae, and together with the loss of hearing phenotype in Minar2 knockout mice (Bowl et al., 2017; Ingham et al., 2019), strengthen the conclusion that minar2 orthologs play essential roles in the auditory functions of the vertebrates. AM1-43 labeling is reduced in minar2 mutant Because minar2 was specifically expressed by the mechanosensitive hair cells, we examined whether there were defects in the hair cells in the minar2 mutants. We first determined levels of mechanotransduction in zebrafish larvae using AM1-43 labeling (Figure 1D). AM1-43, a fixable analog of FM1-43, is a vital fluorescence dye that labels hair cells by traversing open mechanosensitive channels (Meyers et al., 2003), thus providing an estimate of active mechanotransduction. We found the labeling intensities of AM1-43 were markedly reduced in the hair cells of the lateral line neuromasts of the minar2fs139 mutant (56.8% and 51.1% of wild-type controls at 5 dpf and 8 dpf, respectively). We next counted the number of hair cells in the lateral line neuromasts by crossing the minar2fs139 mutant with a myo6:Gal4FF;UAS-EGFP transgenic line, which specifically labeled hair cells (Figure 1E). The results showed the average number of hair cells was decreased in the minar2fs139 mutant (82.3% and 82.1% of wild-type controls at 5 dpf and 8 dpf, respectively). Because the decreases in hair cell numbers (~18% reduction) are smaller than those reductions of AM1-43 labeling intensities (~50% reduction), these results suggest that loss of minar2 function mainly affects hair cell mechanotransduction in the zebrafish larvae. We also examined the number of inner ear hair cells in the minar2 mutant. In the zebrafish larvae, loss of minar2 didn’t alter the number of hair cells in the inner ears (for 5 dpf wild type: 23.9±0.6, minar2fs139 mutant: 23.6±0.4, minar2fs140 mutant: 24.9±0.6 hair cells, F(2, 82)=1.399, p=0.253; for 8 dpf wild type: 32.5±0.7, minar2fs139 mutant: 32.9±0.5, minar2fs140 mutant: 33.0±0.7 hair cells, F(2, 79)=0.2378, p=0.789, both of the lateral crista of inner ear, Figure 1—figure supplement 2A). The homozygous minar2 mutants were viable, and the body length and body weight of adults were no different from the wild-type controls (Figure 1—figure supplement 2B). Nevertheless, hematoxylin and eosin staining of head sections suggested the inner ear regions were smaller and the numbers of hair cells were decreased in the utricle, semicircular canal crista, and saccule of the minar2fs139 mutant (Figure 1—figure supplement 2C). To quantify the number of hair cells in adult zebrafish, we excised utricles and saccules from the inner ears and labeled the hair cells with fluorescence-conjugated phalloidin. The average numbers of hair cells were broadly decreased across different regions of utricles and saccules in the minar2fs139 mutant, ranging from 71.2% to 84.3% of wild-type controls (Figure 1F, for the utricles, the genotype factor, F(1, 32)=37.77, p<0.0001; for the saccules, the genotype factor, F(1, 19)=19.94, p<0.001). We conclude that minar2 is required for mechanotransduction in the zebrafish larvae, and loss of minar2 reduces the number of inner ear hair cells in the zebrafish adults. Minar2 protein localizes to the stereocilia and the apical region of the hair cells To study how Minar2 functioned in the hair cells, we first examined the subcellular localization of Minar2 protein in the hair cells. As the Minar2 antibodies were not available, we generated a GFP-Minar2 fusion construct driven by a hair cell-specific promoter, myosin 6b (Kindt et al., 2012; Maeda et al., 2017). We injected this construct into fertilized oocytes and expressed the GFP-Minar2 fusion in the hair cells. We observed that GFP-Minar2 was localized to the stereocilia and the apical region of the hair cell, apparently around and just below the cuticular plate. The GFP-Minar2 fluorescence co-localized with the phalloidin-labeled stereocilia (Figure 2A). These spatial distribution patterns of GFP-Minar2 were similar regardless of apparent expression levels. To corroborate the distribution of GFP-Minar2, we generated a FLAG-tagged Minar2 construct, myo6b:GFP-P2A-FLAG-Minar2, which allowed labeling of hair cells by GFP and localization of Minar2 by the small FLAG tag. The results showed that FLAG-Minar2 was similarly localized to the stereocilia and the apical region of hair cells (Figure 2—figure supplement 1A). We subsequently generated a stable transgenic line using the myo6b:GFP-Minar2 construct and we found that GFP-Minar2 was also localized to the stereocilia and the apical region of the hair cells in the transgenic animals (Figure 2B). Prominent vesicle-like structures were seen below the cuticular plate, while a ring-like structure was often observed at the base of the hair bundle. We co-labeled the kinocilia with anti-acetylated tubulin antibodies in the myo6b:GFP-Minar2 transgenic zebrafish and found there was no observable GFP-Minar2 signal in the kinocilia (Figure 2—figure supplement 1B). Figure 2 with 2 supplements see all Download asset Open asset Localization and function of Minar2 in the stereocilia and the apical region of the hair cells. (A) Representative images of transiently expressed GFP-Minar2 in hair cells. The dashed line marks the border of a hair cell expressing GFP-Minar2. Stereocilia were labeled with phalloidin. Nuclei were counterstained by DAPI. (B) Distribution of GFP-Minar2 in hair cells in the stable myo6:GFP-Minar2 transgenic line. Representative images of hair cells of lateral crista of the inner ear (Inner ear) and lateral line neuromast (Lateral the of hair cells. Hair cells were also imaged with a The panel an enlarged of the (C) Quantification of hair bundle of the inner ear hair cells in zebrafish larvae. Hair bundles were labeled with and the lateral crista regions of inner ears were Hair bundle were from and images of wild type, minar2fs139, and minar2fs140 larvae at 5 dpf, or 36, and images of larvae at 8 For 5 dpf, and F(2, For 8 dpf, and F(2, are on the (D) Quantification of hair bundle and of inner ear hair cells in zebrafish adult The show enlarged of the Hair bundles in the saccules were from 8 images for the wild type, and 8 images for the minar2fs139 mutant. and and distribution of lysosomes in the hair cells of the inner ear (E) and lateral line neuromast (F) in zebrafish larvae (5 dpf). For the inner ear, and images of lateral crista regions in the wild type and minar2fs139 mutant were respectively and ****p<0.0001, exact test). For the lateral line, 10 and images of lateral line L3 neuromasts were counted and ****p<0.0001, exact test). Scale bars represent 10 Figure data 1 Localization and function of Minar2 in the apical regions of hair cells Figure Figure 2—figure supplement Download To the subcellular we imaged hair cells with a The results showed that GFP-Minar2 was broadly distributed in the the stereocilia and in the apical region of the hair cell, the GFP-Minar2 signal was of multiple of (Figure 2B). Hair bundles are longer and thinner in minar2 mutant Hair bundles are essential for mechanotransduction. Because the Minar2 protein localizes to the hair bundles and the apical regions the hair bundles we next examined the hair bundles in the minar2 mutant. We labeled the hair bundles with and observed that the hair bundles of inner ear hair cells were longer in the mutant minar2fs139 and minar2fs140 larvae (Figure for 5 dpf, F(2, for 8 dpf, F(2, and the hair bundles also The and of the mutant hair bundles were more in adult minar2fs139 mutants (Figure In the the average length of mutant hair bundles was more than longer than that of wild type controls df = and the was only that of the controls df = p<0.0001). We also found that the kinocilia of lateral line hair cells were in the minar2 mutant larvae, in to the normal together in the wild-type animals (Figure 2—figure supplement for 5 dpf, for 8 dpf, exact test). The kinocilia bundle was a of the kinocilia seen in neuromasts mechanical et al., 2021). at the apical region of the hair cells in minar2 mutant Previous and studies showed that the apical region of the hair cell is with many of which are lysosomes et al., 2014; et al., et al., 2018). We generated a GFP fusion construct for a and expressed the fusion in the hair cells using the to the findings in the mouse we observed that lysosomes were distributed in the apical region of the inner ear hair cells. were also a lysosomes distributed the membrane of the hair cells. In the wild-type most of these labeled lysosomes were small or and structures were observed out of hair and these structures likely were abnormally enlarged lysosomes or In in the inner ear hair cells of the minar2fs139 mutant, the large structures were observed in out of hair and these abnormally enlarged were located at the apical region of the hair cells (Figure exact test). results showing abnormally enlarged were also observed in the hair cells of the lateral line neuromasts (Figure exact test). These data suggested that Minar2 may play a role in the apical lysosomes in the hair cells. Consistent with this GFP-Minar2 signals with the signals at the apical region of the inner ear hair cells (Figure 2—figure supplement 1C). We next examined the subcellular localization of Minar2 in cultured cells in Because the MINAR2 gene is expressed at low levels in cultured human cell in over human cell we expressed a GFP-Minar2 fusion construct in cultured and cells and and to the and respectively. We found GFP-Minar2 most co-localized with lysosomes (Figure 2—figure supplement when the and distribution of the lysosomes were treatment with GFP-Minar2 signals followed the changes of lysosomes, increased in the and the region (Figure 2—figure supplement We further the with staining of which was the treatment with and we observed that GFP-Minar2 signals as that the signals, GFP-Minar2 was localized on the membranes (Figure 2—figure supplement Minar2 increases cholesterol labeling and with cholesterol in cultured cells Because protein sequence to any functional in we sequence et al., to into function. We first highly sequence patterns from the multiple sequence of Minar2 protein orthologs (Figure 1—figure supplement then a protein search the of the sequence patterns had in the proteins, and out of the to protein from species (Figure A revealed that the sequence in the which is known to with other proteins et al., et al., and lipid membranes et al., The of also binds to cholesterol in the membranes et al., Liu et al., 2016; et al., and this may to the of cholesterol in the and 2005; et al., Figure Download asset Open asset Minar2 increases cholesterol labeling and with cholesterol in cultured cells. (A) sequence search for Minar2 The Minar2 sequence is in et al., 2002; and is by the properties of the amino (B) of Minar2 on levels and of labeling in cultured cells. fluorescence the of all of the average of signals located within the For cells, and for cells, and
Source: eLife
Author: Doris K Wu
Institution: National Institutes of Health
Language: English
Subfield: Sensory Systems
Field: Neuroscience
Domain: Life Sciences
Open Access status: gold
Estimated APC paid: $2,500
OpenAlex ID: W4312831284