Transgenic Tg(Kcnj10-ZsGreen) Fluorescent Reporter Mice Allow Visualization of Intermediate Cells in the Stria Vascularis

The stria vascularis (SV) is a stratified epithelium in the lateral wall of the mammalian cochlea, responsible for both endolymphatic ion homeostasis and generation of the endocochlear potential (EP) critical for normal hearing. The SV has three layers consisting predominantly of basal, intermediate, and marginal cells. Intermediate and marginal cells form an intricate interdigitated network of cell projections making discrimination of the cells challenging. To enable intermediate cell visualization, we engineered by BAC transgenesis, reporter mouse lines expressing ZsGreen fluorescent protein under the control of Kcnj10 promoter and regulatory sequences. Kcnj10 encodes KCNJ10 protein (also known as Kir4.1 or Kir1.2), an ATP-sensitive inwardly-rectifying potassium channel critical to EP generation, highly expressed in SV intermediate cells. In these transgenic mice, ZsGreen fluorescence mimics Kcnj10 endogenous expression in the cochlea and was detected in the intermediate cells of the SV, in the inner phalangeal cells, Hensen’s, Deiters’ and pillar cells, in a subset of spiral ganglion neurons, and in glial cells. We show that expression of the transgene in hemizygous mice does not alter auditory function, nor EP These transgenic Tg(Kcnj10-ZsGreen) mice allow live and fixed tissue visualization of ZsGreen-expressing intermediate cells and will facilitate future studies of stria vascularis cell function.


Introduction
The stria vascularis (SV) is a strati ed epithelial tissue in the lateral wall of the cochlear duct, responsible for ion transport, in particular potassium, into the endolymph-containing scala media.The processes that lead to and maintain an elevated potassium concentration in the endolymph are necessary for the generation and maintenance of an endocochlear potential (EP) of 80 to 100 mV, which is critical for normal hearing 1,2 .The SV contains three layers of main cell types: the basal cells, which are adjacent to the brocytes of the spiral ligament, the marginal cells, which directly border the scala media and are in direct contact with the endolymph, and the intermediate cells between these two cell types (Fig. 1a).While the basal cells are easily distinguishable from the other cell types, the intermediate and marginal cells form an interdigitated network of cell projections within the SV [3][4][5] (Fig. 1b, 1c).This intricate interdigitation presents a challenge to delineate intermediate cells from the surrounding cells of other types, hindering efforts to identify changes in their number, morphology, ion exchanger and channel composition for examples.Being able to readily identify these cells would facilitate their characterization and the re nement of our understanding of the roles they play in generating the EP and other mechanisms they may be involved in, in both physiological and pathological conditions.Highlighting the importance of studying the physiology of the SV, decreased hearing function has been linked to pathological changes of the SV associated with Alport syndrome 6 , Pendred syndrome 7,8 , Norrie disease 9 , neuraminidase 1-de ciency 10 and some forms of nonsyndromic deafness 11,12 .KCNJ10 also known as Kir4.1 and Kir1.2, is encoded by the gene Kcnj10 in mice.It is an ATP-sensitive inwardly rectifying potassium channel critical to EP generation, highly expressed in SV intermediate cells where it is found concentrated at the plasma membrane of their digitations in contact with the marginal cells (Fig. 1c, d) 1,13,14 .Notably, single-nucleus RNA-sequencing data previously generated by our group also demonstrated the speci city of the expression of Kcnj10 mRNA in the intermediate cells as compared to the other cell types of the SV (Fig. 1d).As previously reviewed by Chen and Zhao, Kcnj10 is also expressed in the Deiters' cells surrounding the outer hair cells in the organ of Corti, and in the satellite glial cells of the cochlear spiral ganglion 13 .
To enable live visualization and xed tissue identi cation of the intermediate cells of the SV and their digitations, we generated bacterial arti cial chromosome (BAC) transgenic mouse reporter lines expressing cytoplasmic ZsGreen uorescent protein under the control of the Kcnj10 promoter region and regulatory sequences.We show that in these reporter mice, ZsGreen uorescence indeed parallels Kcnj10 endogenous expression and does not alter auditory function nor EP.These reporter mice are a new tool to study the SV at the cellular level and further de ne its functions.
To generate Kcnj10-ZsGreen transgenic mice, a modi ed BAC (RP23-157J4) expected to contain Kcnj10 promoter and regulatory sequences was obtained.A ZsGreen expression cassette was used to replace Kcnj10 mouse coding sequence (genomic region corresponding to NM_001039484.1:exon 2 nucleotide 243 to 1393) by BAC recombineering approach (Fig. 2a).Before pronucleus injection, the modi ed BAC was sequenced across ZsGreen to verify that the intended insertion of ZsGreen occurred.The absence of gross recombination inside the BAC was checked by restriction enzyme digestion followed by pulsed eld gel.
Four founder mice were obtained.After one breeding with C57BL/6J mice, their progeny was studied for the presence of the transgene, leading to the isolation of four transgenic mouse lines.Two of these, lines 850 and 858 (Fig. 2b), were studied in detail for their ZsGreen expression in the cochlea.These lines were backcrossed for over 10 generations on a C57BL/6J background, to obtain congenic mice.
As BAC transgenes are sometimes inserted in multiple copies, often in tandem, we aimed to identify the number of copies of the ZsGreen transgene in these two mouse lines.To do so, we used digital droplet PCR (ddPCR) to study the gDNA of mice from these two lines and their wildtype littermates.We compared the number of droplets positive for ZsGreen and for the autosomal gene Actb used as a reference.As shown in Fig. 2c, hemizygous Kcnj10-ZsGreen mice from lines 850 and 858 both have one copy of the ZsGreen transgene.
Overview of Kcnj10-ZsGreen expression in the cochlea.
To evaluate ZsGreen expression in the cochlea, we examined ZsGreen uorescence in different settings commonly used for the study of the inner ear in two month old hemizygous mice.ZsGreen in vivo uorescence was readily visualized in the inner ear without ampli cation even through the bone of the cochlear duct of the mature inner ear as well as in xed cochlea cryosections (Fig. 3a, b).In low magni cation view of a mid-modiolar cross-section of the cochlea, ZsGreen signal was detectable in the SV, in the root cells, in the organ of Corti, as well as in region of the modiolus (Fig. 3b for line 850, Fig. S1 for line 858).ZsGreen uorescence was present all along the cochlear duct from base to apex.ZsGreen expression in mice from lines 850 and 858 was undistinguishable.
In the organ of Corti whole-mount imaged at the hair cell nuclei level, ZsGreen uorescence could be seen in supporting cell area, suggestive of expression in the inner phalangeal cells, pillar cells and Deiters' cells (Fig. 3d).The phalangeal processes of the Deiters' cells were recognizable by their acetylated tubulin immunoreactivity as previously described 15 .In whole-mount tissue, ZsGreen uorescence was also detected in a subset of cells of the SV, in a region with strong vascularization as indicated by the presence of endothelial cells labeled by isolectin GS-IB4 16 , consistent with the potential expression of ZsGreen in the intermediate cells (Fig. 3c).Similarly, light sheet microscopy of Kcnj10-ZsGreen cochleae demonstrated expression of the uorescent reporter in the intermediate cells of the SV, in a subset of cells of the organ of Corti, and in cells located in the modiolus (Video S2).Three-dimensional reconstruction from lateral to medial through the lateral wall further con rmed the expression of ZsGreen in SV intermediate cells which lie between the basal and marginal cell layers of the SV (Fig. 1a and Video S2).
In the stria vascularis, Kcnj10-ZsGreen transgene is expressed speci cally in intermediate cells.
As shown in Fig. 4a, ZsGreen was detected in the cells which express endogenously KCNJ10 protein in the SV.ZsGreen uorescence also overlapped with Kcnj10 mRNA detected by smFISH (Fig. 4b-c).Thus, the expression of the uorescent reporter parallels endogenous expression of Kcnj10 and KCNJ10 in the intermediate cells of the SV allowing them to be readily identi ed.ZsGreen uorescence was present in the nuclei and the cytoplasm of these intermediate cells, with some ZsGreen uorescence present in aggregates in the cytoplasm or subcellular compartments (Fig. 4b).The presence of aggregates may be suggestive of a disturbance in proteostasis as denoted by Blumenstock and colleagues 17 .
Kcnj10-ZsGreen transgene is expressed in pillar cells and Deiters' cells in the organ of Corti.
In the organ of Corti, ZsGreen uorescence was not seen in cochlear hair cells identi able by their MYO7A immunolabeling.ZsGreen uorescence was detected in a subset of supporting cells.It partially overlapped with the immunolabeling of acetylated tubulin, which is known to be present in both pillar cells and Deiters' cells where acetylated tubulin is particularly concentrated in their phalangeal processes 15 (Fig. 4d).ZsGreen uorescence overlapped with Kcnj10 mRNA expression detected by smFISH in in the inner phalangeal cells, pillar and Deiters' cells, with a higher degree of expression in the inner phalangeal cells, and in Deiters' cells (Fig. S3).The uorescent reporter appeared to be largely cytoplasmic in these cells.Of note, amongst supporting cells, KCNJ10 protein was also localized in ZsGreen-expressing Deiters' cells underlying the outer hair cells but did not appear to be present at detectable level in pillar cells unlike Kcnj10 mRNA (Fig. S4a-d).Kcnj10 mRNA expression in both pillar and Deiters' supporting cells was also identi ed in single cell transcriptome datasets from the P7 organ of Corti 18 (Fig. S5).The discrepancy between mRNA expression and protein detection may be due to a difference in Kcnj10 mRNA expression or alternatively, post-transcriptional or post-translational modi cations involving KCNJ10 protein preventing its detection by the antibodies we used.Tissue processing involving decalci cation and cryopreservation could also in uence KCNJ10 epitope accessibility to the antibodies.
Kcnj10-ZsGreen transgene is expressed in satellite glial cells and a subset of spiral ganglion neurons.
In the modiolus, ZsGreen uorescence was consistently detected in the satellite glial cells as demonstrated by its presence in the cytoplasm and to a lesser extend the nuclei of the cells which express SOX10, a nuclear marker of glial cells [19][20][21][22] (Fig. 4e, Fig. S6a-c).On the contrary, ZsGreen was not detected in most spiral ganglion neurons recognizable by their TUJ1-immunoreactivity (Fig. 4e-f).
Kcnj10 mRNA expression detected by smFISH was present in both glial cells, as well as at least in a subset of spiral ganglion neuron nuclei (Fig. S6).Corroborating these results, single cell transcriptomic analysis of Kcnj10 mRNA expression in adult spiral ganglion neuron subtypes (Types IA, IB, IC, II) and satellite glial cells demonstrated its expression in both satellite glial cells and type II spiral ganglion neurons 23 (Fig. S7).While ZsGreen was detected consistently in glial cells expressing endogenously Kcnj10 (Fig. S6b-c), its expression was more variable in spiral ganglion neurons, including those expressing endogenously Kcnj10 (Fig. S6d-f).KCNJ10 immunolabeling was also detected in SOX10expressing satellite glial cells (Fig. S8a-d) and at least a subset of TUJ1-expressing spiral ganglion neurons (Fig. S8e-h).In these experiments, ZsGreen was also detected consistently in the glial cells, but seemed to be only detectable in a subset of spiral ganglion neurons (Fig. S8).
In summary, ZsGreen transgene expression appears to re ect endogenous Kcnj10 mRNA and KCNJ10 protein expression in satellite glial cells, with a more variable expression in a subset of spiral ganglion neurons, offering a useful genetic label for the study of these cell types in the future.
ZsGreen expression does not alter the audiometric pro le nor the endocochlear potential of Kcnj10-ZsGreen mice.
To test whether ZsGreen expression affected hearing and ionic homeostasis in the transgenic mice, auditory brainstem responses (ABRs), distortion product of otoacoustic emissions (DPOAEs) and EP were measured in these mice at P30. ABR of Kcnj10-ZsGreen mice (N = 8) and wildtype littermates (N = 8) were measured at 8, 16, 32, and 40 kHz, to assess cochlear function at different positions along the cochlear duct (Fig. 5a).Two-way ANOVA analysis did not detect any signi cant differences between ABR obtained for the two genotypes at these different frequencies (p = 0.90) (Fig. 5a).As ZsGreen is expressed in the glia cells, which could in uence type I auditory neurons function and viability, we quanti ed the amplitude and latency of the wave I of these ABRs at 80 dB stimulation to test for the presence of potential hidden hearing loss and auditory dyssynchrony (Fig. S9a, b).Two-way ANOVA analysis of these results did not detect signi cant differences between ABR amplitude and latency for the two genotypes at the different frequencies tested (p = 0.98 and 0.69, respectively).DPOAEs were measured using F2 frequency at 4005,  8011, 9632, 11158, 12779, 14401, 16022, 19169, 22411, 25559, 28801, 32044, 38433, 39959, 28801,   32044, 38433, 39959, 44823 Hz (Fig. 5b).Two-way ANOVA analysis demonstrated no signi cant differences between DPOAEs for the two genotypes at the frequencies tested (p = 0.95) (Fig. 5b).EP measurements revealed similarly normal EP values between Kcnj10-ZsGreen mice (N = 8, 84.39 ± 12.66 mV) and their wildtype littermates (N = 8) (mean = 91.75 ± 7.88 mV) (p = 0.26) (Fig. 5c).These results indicate that these Kcnj10-ZsGreen hemizygous mice have normal hearing and EP and are suitable for the study of hearing function and hearing loss as it relates to cell types with endogenous Kcnj10 expression.

Discussion
While the role of the stria vascularis in maintaining the EP has been well documented 2,24 the mechanisms by which speci c SV cell types maintain ion homeostasis and EP remains incompletely de ned.To further study these mechanisms, tools that allow for the study of speci c cell types in the SV are needed.
Here we report the generation and cochlear characterization of transgenic mice expressing ZsGreen uorescent reporter driven by the promoter and regulatory sequences of a gene speci cally expressed in intermediate cells of the stria vascularis, Kcnj10 2 .We demonstrate that these mice enable the visualization of SV intermediate cells in both live and xed tissue.We also show that the expression of the reporter does not affect hearing as measured by ABRs and DPOAEs, and does nor affect the EP, providing physiological evidence that this reporter mouse is a relevant tool to study the role of the SV in hearing function.
ZsGreen uorescence in Kcnj10-ZsGreen reporter mice recapitulated the endogenous expression of Kcnj10 as shown by smFISH and single cell data.ZsGreen was expressed in the intermediate cells of the Kcnj10-ZsGreen transgenic mouse allowing their readily identi cation from the other major cell types of the SV in both live and xed tissue.ZsGreen was also expressed in the pillar cells, Deiters' cells, satellite glial cells and a subset of spiral ganglion neurons, all of which support normal hearing function.The supporting cells have various functions including formation of the tunnel of Corti by maturation of pillar cells 25 and the maintenance of hair cell homeostasis by Deiters' cells 26 .Satellite glial cells provide structural support to neurons and help facilitate neuron to neuron communication 27 , while spiral ganglion neurons transmit auditory information from hair cells to the brainstem 28 .
Other groups have established that using an open vessel-window approach combined with intravital uorescence microscopy, one can directly visualize the SV in vivo 29 .Using the Kcnj10 ZsGreen reporter mice with this imaging methodology, one could directly observe the intermediate cells of the SV.Presence of intermediate cells detectable through ZsGreen expression could be utilized as a proxy for intermediate cell viability both in vivo, ex vivo and in xed tissue.This has further applications to study intermediate cell physiology in live tissue, including their electrophysiological properties.When combined with auditory physiology measurements such as ABRs and EP, intermediate cell viability could be correlated with clinical hearing changes.For example, cell type-speci c effects of pharmacologic interventions on intermediate cells, particularly as they relate to intermediate cell morphology and viability beyond cell death could be ascertained utilizing this mouse model for agents including furosemide [30][31][32] , aminoglycosides 32 , and cisplatin [33][34][35] .This reporter mouse could also be utilized to assess the impact of known pathogenic genetic variants on intermediate cell physiology.For example, pathogenic variants affecting GJB2, are the most common genetic causes of nonsyndromic human sensorineural hearing loss 36 .GJB2 encodes connexin 26 which is expressed in SV intermediate and basal cells in mice 4,37 with slightly different distributions in these cell types in humans 38 .Early oxidative stress and metabolic dysregulation have been identi ed as downstream results of these genetic variants in the SV 36 .If bred with the appropriate mouse model of disease, this reporter mouse offers an opportunity to study the physiological impacts of these pathogenic variants at the cell type-speci c level in regions where connexins are expressed including SV intermediate cells as well as pillar and Deiters' cells 38 .
Limited mouse auditory cell lines are available for research purposes, with the HEI-OC1 cell line being the most used 39 .With respect to the SV in particular, despite SV cell behavior in explant settings having been widely studied [40][41][42][43] , SV cell lines remain even more limited.Available options include the SV-k1 cell line, which has been described as having mostly marginal cell like characteristics.While SV-k1 demonstrates the presence of markers speci c to strial marginal cells such as Na,K-ATPase α1 and β2 subunits, the line has not been shown to express the marginal cell speci c marker, voltage dependent potassium channel KCNE1 44,45 .Considering the SV contains multiple cell populations, there is a need to produce more SV cell lines that re ect the characteristics of the different cell populations to better study the physiology of each cell type.The Kcnj10-ZsGreen reporter mice's unique feature of uorescently labelled intermediate cells that can be readily seen on whole-mount preparations, opens the door for the creation of a cell line with cell speci c markers and in vitro testing applications.This includes the study of the effects of novel and repurposed pharmacologic agents for the treatment of hearing loss on the SV identi ed from other studies including computationally based repurposing studies 4,35,46,47 .While cell lines may not fully replicate the in vivo function in the de ned microenvironment of the cochlea, the possibility of organoids containing elements of the SV have been suggested by recent work by van der Valk and colleagues 48 and offer the potential organoid-based in vitro testing applications.Finally, the possibility of using this mouse model for targeted uorescence-activated cell sorting (FACS)-puri cation of ZsGreen-expressing SV intermediate cells might allow for more in-depth analyses of these cell types, including potential interactions with neighboring immune cells in the intermediate layer of the SV.

Conclusions
In summary, we report the generation of a BAC transgenic mouse which utilizes the Kcnj10 promoter and regulatory sequences to drive ZsGreen expression in the intermediate cells of the SV and to a lesser extent in the satellite glial cells of the spiral ganglion region and cochlear supporting cells, a pro le which reproduces the endogenous expression of Kcnj10.Expression of the uorescent reporter does not alter auditory physiology as measured by ABRs, DPOAEs and EP.Overall, this Kcnj10-ZsGreen transgenic reporter mouse provides a much needed tool to study, both in live and xed tissue, intermediate cell function.

Methods
Ethics declaration and approval for animal experiments.
All animal experiments and procedures were performed according to protocols approved by the Animal Care and Use Committee of the National Institute of Neurological Diseases and Stroke and the National Institute on Deafness and Other Communication Disorders, National Institutes of Health.
The bacterial arti cial chromosome (BAC) clone RP23-157J4 from RPCI − 23 Female (C57BL/6J) Mouse BAC Library (GenBank: AC074311.28)was identi ed as containing the gene Kcnj10.This BAC (obtained from the BACPAC Resource Center located at Children's Hospital Oakland Research Institute in Oakland, CA) contained 186 kb of mouse gDNA spanning from Chr1:172236604-172422466 (GRCm38/mm10).Kcnj10 mRNA (NM_001039484.1) spans from Chr1:172341210-172374085.This gene has 2 exons, with a coding sequence starting at the start of exon 2. The transgenic construct based on BAC RP23-157J4 was engineered with minor modi cations using methods described in Lee and colleagues 49 and Zeidler and colleagues 50 .Reporter mice were generated at the Transgenic Animal Model Core of the University of Michigan's Biomedical Research Core Facilities.
The synthetic donor DNA was produced by PCR ampli cation of the recombineering plasmid R6K-PGK-ZsGreen.The PCR primers contained 80 nucleotides of homologous genomic sequences that matched the DNA 5' and 3' of the desired insertion in Kcnj10 genomic DNA.The BAC clone and the synthetic donor DNA were combined in DH10B competent bacteria.Introduction of the synthetic donor into the BAC resulted in kanamycin resistance of the bacteria clones containing the recombined BAC.The kanamycin cassette was then removed by induction of FLP recombinase expression, using plasmid pE-FLP (Addgene, #45978).DNA from kanamycin sensitive BACs were analyzed to identify correctly modi ed BACs by sequencing and enzyme restriction.A nal recombination step replaced the BAC-backbone internal loxP site with ampicillin resistance cassette.
The modi ed BAC was sequenced to verify that the intended insertion of ZsGreen occurred, using 1350 bp amplicon obtained with primers located 5' and 3' of ZsGreen: 5'-CCACCACCTCCAACATGAAT-3' and 5'-CTCTCTTTCCCCCAAGCTG-3' and GoTaqGreen polymerase (Promega) with an annealing temperature of 55°C.The nal BAC showed a silent (A > G) mutation in the amino acid Lysine at position 15 of ZsGreen.The absence of gross recombination inside the BAC was checked by restriction enzyme digestion followed by pulsed eld gel.This recombinant RP23-157J4 BAC was microinjected into fertilized eggs obtained by the mating of B6SJLF1/J female mice with B6SJLF1/J male mice (stock number #100012) obtained from the Jackson Laboratory (Bar Harbor, ME, USA) at the Transgenic Core Facility of the University of Michigan.
Four transgenic founder mice were obtained by random integration of the transgene in their genome.
Genomic DNA prepared from tail or ear clips using the Maxwell 16 System (Promega, Madison, WI, USA) was used to identify the mice carrying ZsGreen (Table S1).Four independent transgenic lines were obtained.After maintaining these for 5 generations and noticing no differences in ZsGreen expression pattern between lines, two of them were cryopreserved and the two remaining, lines 850 and 858, were maintained alive and further backcrossed.Male mice hemizygous carrier of the ZsGreen transgene were mated with C57BL/6J wildtype female mice (stock number #000664) obtained from the Jackson Laboratory for over 10 generations leading to obtention of congenic mice carrier of the transgene.The mice studied here were from the 11th generation and were all hemizygous for the transgene.The resulting mouse strains was named B6.Cg-Tg(Kcnj10-ZsGreen) skMHa and B6.Cg-Tg(Kcnj10-ZsGreen) skMHb for lines 850 and 858, respectively.Location of the primers utilized to detect the Kcnj10-ZsGreen transgene in mice are shown in Fig. 2a.Primer sequences and PCR conditions used for genotyping are available in Table S10.

Single-cell and single-nucleus RNA-sequencing dataset analysis
Adult spiral ganglion neuron and satellite glia cells and P7 organ of Corti single cell RNA-Seq datasets and adult SV single-nucleus RNA-Seq datasets from the mouse were analyzed for expression of Kcnj10 3,18,23 .Violin plots of expression among cell types in the SV (marginal, intermediate, basal and spindle cells), cell types in the P7 organ of Corti (inner and outer hair cells, pillar cells, and Deiters' cells) and in the spiral ganglion region (Type IA, IB, IC and type II spiral ganglion neurons, satellite glial cells) were constructed as we have previously described 3,4 .
Reporter transgene copy number assessment.
To assess the number of transgene copies integrated in the gDNA of the transgenic mice, digital droplet PCR (ddPCR, Bio-Rad, Hercules, CA, USA) was performed.Seven animals (4 males and 3 females) from each of the two founder lines (850 and 858) were tested in Kcnj10-ZsGreen group.Five wildtype littermates were included as control.gDNA was isolated from tail snip using Maxwell 16 DNA puri cation system (Promega, RRID:SCR_020254), DNA concentration was measured using Thermo Scienti c NanoDrop One/OneC Microvolume UV Vis Spectrophotometer (Thermo Fisher Scienti c, Waltham, MA, USA, RRID:SCR_023005) and was adjusted to 50 ng/uL.The autosomal gene Actb was used as reference.In ddPCR reaction, each well contained 1 uL of ZsGreen probe (FAM labeled)/primers mix, 1 uL of the probe recognizing the gene Actb (HEX labeled)/primers mix, 1uL of gDNA (50 ng), 10 uL Bio-Rad ddPCR Supermix (Bio-Rad, Hercules, CA, USA, Cat#1863024) for Probes and 7 uL molecular grade water.The ddPCR Droplets were generated using the QX200 AutoDG Droplet Digital (Bio-Rad, Cat#1864101).PCR was performed as described in the QX200 ddPCR EvaGreen Supermix instructions.Droplets were read with a QX200 Droplet Reader (Bio-Rad, Cat#1864003) and analyzed with QuantaSoft software (Bio-Rad, Cat#1864011).Sequences of the primers and probes used for this experiment are presented in Table S11.
Tissue preparation for immuno uorescence labeling and in situ hybridization.
Hemizygous transgenic mice and their wildtype littermates were studied.Unless indicated otherwise mice were studied at P56 to P65 (summarized here at P60) for all experiments.Inner ears from these mice were dissected and placed in 4% paraformaldehyde (PFA in 1 x PBS solution) overnight at 4°C.Fixed adult mouse inner ears were then decalci ed in 0.25 M EDTA for 3 days in 4°C on orbital shaker.If the tissue was designated for immunostaining or hybridization as whole mounts, it was washed in PBS, dissected and stored in PBS in 4°C for further use.If the tissue was designated for cryosections, it was washed in PBS, transferred to 30% sucrose in PBS at 4°C overnight, followed by immersion in 50/50 mix of 30% sucrose in PBS and nally in super cryo-embedding medium (SCEM) (C-EM001, Section-Lab Co, Ltd.; Hiroshima, Japan).Tissue was ash-frozen in liquid nitrogen after the transfer and 2-hour incubation in fresh 100% SCEM in a cryomold biopsy square.
Fluorescent immunohistochemistry was performed as follows.Cryosections or whole mount tissue were washed in PBS then permeabilized and blocked (2 hours at room temperature (RT) for cryosections, overnight at 4°C for whole-mounts) in PBS with 0.2% Triton X-100 (PBS-T) with 10% fetal bovine serum (A3840001, ThermoFisher Scienti c, Waltham, MA, USA).Samples were then incubated in the appropriate primary antibodies in PBS-T with 10% fetal bovine serum for 24 hours in 4°C, followed by two 10-minute washes in PBS-T and labelling for 2 hours at RT with AlexaFluor 488, 555 and/or 647-conjugated secondary antibodies made in donkey and directed against appropriate species (Life Technologies, Waltham, MA) diluted at 1:250 in PBS-T.DAPI (4,6-diamidino-2-phenylindole,1:10,000, Life Technologies, Waltham, MA) was included with the secondary antibodies to detect nuclei.Alexa Fluor 647 phalloidin was used to label F-actin in a subset of experiments.Samples were washed in PBS four times for ve minutes and mounted in SlowFade Gold (S36937, Invitrogen, ThermoFisher).Specimens were imaged using Zeiss LSM880 confocal microscopes (Zeiss, Oberkochen, Germany) using 40x, 1.4 numerical aperture and 63x, 1.4 numerical aperture objectives.Primary antibodies utilized are detailed in Table S12.

Single molecule uorescent in situ hybridization (smFISH).
In situ hybridization on PFA xed, frozen tissue was performed with RNAscope® Multiplex Fluorescent Reagent Kit v2 (Advanced Cell Diagnostics, Hayward, CA, USA) with substantial modi cation of pretreatment process was performed as previously described 3,4,35 .Brie y, frozen sections on Superfrost Plus microscope slides were removed from − 80°C freezer, placed on heating block (37°C) for 30 minutes to thaw and dry, then moved to RT. Hydrogen peroxide solution was applied on the section and incubated for 10 minutes at RT. Slides were washed 2 times in deionized water, dried initially on heating block for 10 minutes at 37°C, then placed on heating block and baked for 30 minutes at 60°C.After the baking, slides were moved to RT and hydrophobic barrier was applied around the specimen and left to dry.After 10 minutes, Protease PLUS was applied on the sample and slide was moved to hybridization oven set to 40°C and incubated for 25 minutes.Then slides were washed twice in deionized water, appropriate probe was applied to cover the sample and slides were incubated in hybridization oven for 2 hours at 40°C.Next steps of the protocol directly followed RNAscope® Multiplex Fluorescent Reagent Kit v2 User Manual (Doc.No. 323100-USM).
Light sheet microscopy imaging of Kcnj10-ZsGreen adult mouse cochlea.

Sample processing
Mouse cochlea xed in 4% paraformaldehyde was washed in PBS at RT for half a day and then transferred into 50 mL of decalci cation solution (20% EDTA in 1X PBS, pH = 9) for 7-10 days at 40°C.Fresh decalci cation solution was supplied every 3 days.Finally, the decalci ed cochlea was washed in PBS for 1 day at 4°C.To clear the tissue, decalci ed mouse cochlea was dehydrated in a step gradient of tetrahydrofuran/water solutions (20%, 40%, 60%, 80%, 100% tetrahydrofuran) at 4°C.The duration of each step was 12 h and the cochlea was incubated in 100% tetrahydrofuran one extra time at 4°C overnight.Dehydrated cochlea was incubated in 100% dibenzyl ether at 4°C until the sample sank and was incubated one more time in fresh 100% dibenzyl ether at 4°C overnight before imaging.

Cleared tissue DISPIM
The dual-view inverted selective plane illumination microscope optimized for cleared tissue (CT-DISPIM; Applied Scienti c Instrumentation, Eugene, OR, USA), and the associated data processing pipeline have been described in detail previously 51 .
We used a pair of 0.7 N.A. multi-immersion objectives (Special Optics; Denville, NJ, USA) to acquire images of a cochlea that had been cleared using iDISCO and mounted in dibenzyl ether (DBE).The sample was excited by a digitally scanned OBIS laser (Coherent; Santa Clara, CA, USA) light sheet.
Fluorescence was ltered through an emission bandpass lter before being recorded on a Hamamatsu Flash 4 v3 sCMOS camera (Hamamatsu Photonics; Shizuoka, Japan).Kcnj10-ZsGreen was imaged using 488 nm excitation and a 525/50 bandpass emission lter.Auto uorescence was imaged using 637 nm excitation and a 676/37 bandpass emission lter (both lters from Semrock; West Henrietta, NY, USA).Image volumes were acquired in single view mode as stage-scanned tiles at full frame (2048 × 2048 pixels, FOV 520 µm, 0.254 µm per pixel) with 1 µm perpendicular inter-plane distance and 10% overlap between adjacent tiles.Image tiles were then stitched and deskewed on the NIH Biowulf supercomputer.Three-dimensional rendering was done in Imaris (version 9.9, Oxford Instruments).
Auditory brainstem responses and distortion product otoacoustic emissions.
Auditory brainstem responses (ABRs) were detected in both ears of anesthetized mouse at age P60-P65.Wildtype littermates lacking the transgene were used as a control group.Mice were anesthetized with an intraperitoneal injection of ketamine (56 mg/kg) and dexdomitor (0.375 mg/kg) and placed on a heating pad connected to a temperature controller (TC-2000, World Precision Instruments, Sarasota, FL, USA) inside a sound-attenuated booth (Acoustic Systems, ETS-Lindgren, Austin, TX, USA) to maintain animal body temperature at 37°C.Recordings were obtained using Tucker-Davis Technologies (Alachua, FL, USA) hardware (RZ6 Processor) and software (BioSigRZ, version 5.7.5).For ABR testing, subdermal electrodes (Rhythmlink, Columbia, SC, USA) were placed at the vertex, under the test ear, and under the contralateral ear (ground).Blackman-gated tone burst stimuli (3 ms, 29.9/s, alternating polarity) were presented to the test ear at 8, 16, 32, and 40 kHz via a closed-eld Tucker-Davis Technologies MF-1 speaker.Responses were ampli ed (20×), ltered (0.3-3 kHz), and digitized (25 kHz) with 512-1024 artifact-free responses per waveform.For each frequency, testing began at 80 dB SPL and decreased in 10 dB steps until the ABR waveform was no longer discernable.Once the response was lost, testing continued in 5 dB steps with a minimum of two waveforms per stimulus level to verify repeatability of ABR waves.ABR thresholds were determined by visual inspection of stacked waveforms for the lowest stimulus level that yielded repeatable waves.
Distortion-product otoacoustic emissions (DPOAEs) were measured in both ears using Tucker-Davis Technologies hardware (RZ6 Multi I/O processor, MF-1 speakers) and software (BioSigRz, version 5.7.5) in conjunction with an Etymotic ER-10B + microphone.Two tones were presented simultaneously at levels of f1 = 65 dB SPL and f2 = 55 dB SPL with the higher frequency tone (f2) set between 4-44.8 kHz (5 points per octave) and f2/f1 = 1.25.Mean noise oors were calculated from levels at six frequencies surrounding the 2f1-f2 DPOAE frequency.

Endocochlear potential measurement.
Methods for EP measurement have been described previously [52][53][54][55] .Brie y, P63-P68 mice were anesthetized with 2,2,2-tribromoethanol (T4842, Sigma-Aldrich, St. Louis, MO, USA) at a dose of 0.35 mg/g body weight.EP measurements were made using glass microelectrodes inserted into the round window and through the basilar membrane of the rst turn of the cochlea.Induction of anoxia, allowing measurement of anoxic-state EP, was accomplished by intramuscular injection of succinylcholine chloride (0.1 µg/g, NDC-0409-6629-02, P zer, New York, NY, USA) after establishment of deep anesthesia followed by additional injection of 2,2,2-tribromoethanol.Anoxic-state EP provides an indicator of the lowest EP and sensory hair cell function.In the presence of functional hair cells, the anoxic-state EP is negative, whereas the EP is zero if the hair cells are not functional.Data were recorded digitally (Digidata 1440A and AxoScope 10; Axon Instruments) and analyzed using Clamp t10 (RRID: SCR_011323, Molecular Devices, San Jose, CA, USA).Eight Kcnj10-ZsGreen transgenic mice and eight wildtype C57BL/6J littermates were evaluated.

Statistical Analysis
of cations are presented as mean ± standard deviation (SD).For pairwise comparisons of ABR and DPOAE data between Kcnj10-ZsGreen mice and wildtype littermates, two-way ANOVA was conducted.For pairwise comparison of EP between Kcnj10-ZsGreen mice and wildtype littermates, an unpaired t-test with Welch's correlation was used.All statistical analysis were performed using GraphPad Prism version 8.4.3 (GraphPad Software, San Diego, CA, USA).promoter, coding sequence and regulatory elements responsible for the endogenous expression of Kcnj10, is shown on the top of the panel.The BAC backbone contains a gene of resistance to chloramphenicol (cpl r ).The BAC and a synthetic donor DNA containing the coding sequence of ZsGreen followed by bgh-polyA(pA) sequences, and a gene conferring kanamycin/neomycin resistance (kan r /neo r ) surrounded by FRT sites, were combined in recombineering competent bacteria.The insertion of this synthetic donor in place of the coding sequence of Kcnj10 by homologous recombination was made possible by the insertion in 5' and 3' of this cassette of 80 bp fragments of DNA homologous to the target regions in the BAC (darker blue regions).Introduction of the cassette into the BAC resulted in kan r of the bacteria.DNA from kan r BACs were further analyzed to identify correctly modi ed BACs.The kan r cassette has both prokaryotic and eukaryotic (mouse Pgk 1) promoters.Black lines indicate regions of recombination between mouse gDNA in the BAC DNA and homologous sequences introduced by PCR in the synthetic DNA donor plasmid.The kan r /neo r cassette was removed by expression of ipase (FLP) recombinase.DNA from kanamycin sensitive BACs were analyzed to identify correctly modi ed BACs.A nal recombination step replaced the BAC-backbone internal loxP site with ampicillin resistance (step not shown).b, Mice carrying Kcnj10-ZsGreen transgene were identi ed by PCR.F6/R6 primer pair targets the DNA region between the promotor of Kcnj10 and the 5' sequence of ZsGreen.F1/R1 primer pair targets the 3' sequence of ZsGreen and its polyA region.Amplicons of 156 bp with primers F6/R6 and 924 bp with primers F1/R1 indicate the presence of the transgene.These amplicons were detected in Kcnj10-ZsGreen mice gDNA but not in the gDNA of their wildtype littermates.The sequence of these primers is reported in Table S10.H 2 O: negative control without gDNA template.DNA size marker (ThermoScienti c GeneRuler 100 bp Plus DNA ladder SM0323, Waltham, MA) was used.c, To evaluate Kcnj10-ZsGreen transgene copy number, ddPCR was performed on the gDNA of ZsGreen hemizygote mice from the founder lines 850 and 858 and their wildtype (WT) littermates.The gene Actb encoding b actin, located on chromosome 5 in mice, was used as a reference.ddPCR revealed the presence of ZsGreen targeted ampli cation in half as many droplets as Actb positive droplets, supporting the fact that hemizygous 850 and 858 ZsGreen mice each only carry one copy of the ZsGreen transgene.Inner hair cells (IHCs, arrowhead) and outer hair cells (OHCs, bracket) identi ed by their MYO7A immunolabeling, did not show ZsGreen uorescence.Additionally, some faint signal was seen in regions corresponding to the inner sulcus/border cells and Hensen's cells.e,Cross-section of the modiolus showing ZsGreen uorescence (green) in cells expressing SOX10, a satellite glial cell nuclear marker (blue).ZsGreen expression was not found in the spiral ganglion neurons immunolabeled with TUJ1 (red).f, Same confocal acquisition as Fig. 4e showing that ZsGreen uorescence did not overlap with TUJ1 labeling.Cell nuclei were labeled with DAPI.Imaging from P60 mice.

Declarations
ContributionsDS performed experiments, wrote and edited manuscript, RTO, SN performed experiments, SK worked on transgenic mouse development, SA wrote and edited manuscript, AJG worked on transgenic mouse development, YS worked on sample preparation and optimization of protocol for light sheet imaging, JL worked on data processing, including deskewing, stitching, channel registration for lightsheet imaging, HV performed light sheet imaging, SG performed single cell RNA-sequencing data analysis, TS generated transgenic construct and reporter mice, IR and MH worked on transgenic mouse development, wrote and edited manuscript.

Figures
Figures

Figure 2 Generation
Figure 2