A. Mouse models
Lgr5EGFP–CreERT2 (Stock 008875) mice were purchased from The Jackson Laboratory (Bar Harbor, MN, USA). Wild-type C57BL/c6 mice were purchased from Nara Biotech (Seoul, Korea). Both male and female post-natal D1–2 (P1– P2) mice were used in all experiments. Animal care and experiments were performed according to a protocol approved by the Animal Care and Use Committee of Dankook University (approval number DKU-19-026 and DKU-21-014).
B. DNA isolation and genotyping
While homozygous mice are not viable, heterozygous Lgr5-EGFP-IRES-CreERT2 mice are both viable and fertile. They harbor an Lgr5-EGFP-IRES-creERT2 “knock-in” allele that abolishes Lgr5 (Gpr49) gene function, with expression of the EGFP and CreERT2 fusion protein from the Lgr5 promoter/enhancer elements. To determine mutants against wild-type progenies after breeding, DNA was isolated from pups using the NaOH extraction method (Truett GE et al. 2000, Biotechniques 29(1):52-54) for genotyping.
Briefly, a 2-mm section of the tail of each mouse pup was removed and placed into an Eppendorf tube or 96-well plate, to which 75 µL of 25 mM NaOH/0.2 mM EDTA was added. Then, the samples were placed in a thermocycler at 98ºC for 1 hour, after which the temperature was reduced to 15°C until it was time to proceed to the next step. Next, 75 µL of 40 mM Tris HCl (pH 5.5) was added, and the samples were centrifuged at 4,000 rpm for 3 minutes. The supernatant was collected and aliquoted for PCR (2 µL undiluted, or 2 µL of a 1:100 dilution/reaction).
Qualitative PCR was performed using the following primer sequences recommended by Jackson Laboratories: mutant reverse (5′- CTG AAC TTG TGG CCG TTT AC-3′), wild-type reverse (5′- GTC TGG TCA GAA TGC CCT TG-3′), and common (5′- CTG CTC TCT GCT CCC AGT CT-3′). The primers were obtained from Bioneer (Daejeon, Korea). The PCR products were separated on a 2% agarose gel, stained with EtBR, and photographed. Two PCR products were expected: a 386-bp product from wild-type alleles and a 119-bp product from the mutated alleles.
C. LPC isolation
LPCs were isolated using either MI or MACS. For MI, cochleae from at least three P1–P2 mouse inner ears were dissected in phenol red-free Hanks' balanced salt solution (HBSS) with calcium and magnesium (Gibco, #14025134). The cartilage was opened, the stria vascularis was removed, and the sensory epithelia were detached and incubated in a 100–200-μl droplet of Matrisperse Cell Recovery Solution (Corning, Corning, NY, USA) for 1 h at room temperature. This non-enzymatic solution disintegrates the extracellular matrix of the sensory epithelia, enabling the subsequent separation of HCs and supporting cells from the mesenchyme and neurons. Separation was performed using a stripping method in which one pair of forceps was used to hold the sensory epithelium in place while another pair was used to gently remove the layer of HCs and supporting cells. The forceps holding the intact sensory epithelium were then used to remove the mesenchyme and neurons from the dish, while the HCs and supporting cells remained in the dish for subsequent collection. The cells were then collected, centrifuged for 5 min at 0.5 × g, and incubated in TrypLE (Gibco) for 20 min at 37°C. After additional centrifugation for 5 min at 0.5 × g, the dissociated cells were suspended in HBSS, triturated 50–80 times using a 200-μl pipette tip, and strained using a 40-micron cell strainer to produce a single-cell suspension.
For MACS isolation, whole cochleae from at least 40 mouse inner ears were dissected in Dulbecco's phosphate-buffered saline (DPBS) and incubated in 2 mg/mL collagenase A (Roche/Merck) for 1 h at 37°C before being centrifuged for 5 min at 300 × g. The dissociated cells were suspended in DPBS, triturated 50–80 times using a 1,000-μl pipette tip, and strained using a 40-micron cell filter to produce a single-cell suspension. The single-cell suspension was successively incubated with rabbit anti-mouse Lgr5 polyclonal antibody (Miltenyi Biotech, Bergisch Gladbach, Germany) diluted 1:100 with PBS for 45 min at room temperature. After being washed with cold MACS buffer (Miltenyi Biotech) and centrifuged, the cell pellet was resuspended in MACS buffer and incubated for 15 min at 4°C with 20 µl of anti-rabbit IgG microbeads (Miltenyi Biotech) per 107 total cells. The Lgr5-positive cells were sorted using an LS magnetic column (Miltenyi Biotech) placed in the magnetic field of a separator according to the manufacturer’s instructions.
The isolated LPCs obtained via MI and MACS were assayed against a phycoerythrin (PE)-conjugated LGR5 antibody and analyzed for purity using a CytoFLEX flow cytometer (Beckman Coulter, Brea, CA, US).
D. Sphere culture and differentiation assay
LPCs isolated via MI and MACS were resuspended in 100% Matrigel Basement Membrane Matrix (growth factor reduced, LDEV-free; Corning). Five droplets (30 μl) of Matrigel were placed in each well of a 6-well plate. The plate was incubated at 37°C for 5 min to aid Matrigel polymerization, and the droplets were then covered with spheroid-forming media. Spheroid colonies were formed by culturing the LCPs with basal DMEM/F12 medium supplemented with GlutaMAX, 2% B27, 1% N2, ampicillin (50 μg/ml), basic FGF (bFGF), epidermal growth factor (EGF), and insulin-like growth factor (IGF) (50 ng/ml each), the antioxidant 2-phospho-L-ascorbic acid (pVc; 280 μM), the histone deacetylase (HDAC) inhibitor valproic acid (VPA; 1 mM), and the Gsk-3β inhibitor CHIR99021 (CHIR; 3 μM). The medium was changed every other day for 10 days. For the Fgf2 inhibition experiment, we omitted the bFGF from the spheroid-forming media but added the potent Fgf2 receptor inhibitor AZD4547 to block any exogenous Fgf2 signaling.
Differentiation of the LPC spheroids into inner ear organoids started at D11. MI and control MACS organoids were placed in a solution containing GSK3β inhibitor, CHIR (3 μM), and the Notch inhibitor LY411575 (LY, 10 μM) to induce differentiation for 2 weeks. We also conducted a Notch inhibition experiment in which we doubled the concentration of Ly in the differentiation media to 20 μM. Furthermore, we conducted an SHH activation experiment in which we added the smoothened receptor agonist purmorphamine (1 M, 04–0009; Stemgent, Beltsville, MD, USA) on D11. In all groups, CHIR was replaced with the Wnt/beta-catenin inhibitor IWP-2 (3 mM, 3533; Tocris, Bristol, UK) on D18.
E. qRT-PCR and immunohistochemistry
To analyze the expression of inner ear progenitor-related genes, 1 µg of RNA was collected from the MI and MACS spheroids at D10 using GeneAll Hybrid-R (GeneAll Biotechnology, Songpa-gu, Korea) according to the manufacturer’s instructions. The forward and reverse primers (Oligomer; Bioneer) used to synthesize cDNA are listed in Supplementary Table S8. Detailed information of qRT-PCR is summarized in Supplementary Table S9
Individual organoids were isolated and fixed with 2% paraformaldehyde for immunostaining. The organoids were then blocked with 5% normal goat serum and 0.3% Triton X-100 in PBS for 1 h at room temperature, followed by overnight incubation at 4°C with primary antibodies (listed in Supplementary Table S10). Then, the slides were washed and incubated for 60 min at 37°C in Alexa Fluor 488-conjugated goat anti-rabbit (IgG) at 1:1,000 (#A11034; Life Technologies, Carlsbad, CA, USA) and Alexa Fluor 568-conjugated goat anti-mouse at 1:1,000 (#A21124; Life Technologies). Finally, the samples were washed with PBS and mounted using VectaShield (Vector Laboratories, Newark, CA, USA) with or without DAPI. Stained cells were carefully examined under the Flow-View 3000 confocal microscope (Olympus, Tokyo, Japan).
F. TEM and SEM
Ultra-morphological features of the D24 organoids were obtained via TEM and SEM. For TEM, the organoids were immediately fixed with 2.5% glutaraldehyde in phosphate buffer solution overnight. After being washed, the cultures were dehydrated in acetone and embedded in a mixture of Araldite and Epon 812 in flat rubber molds. The blocks were then mounted in an Ultracut Ultramicrotome (Leica, Wetzlar, Germany). Two micrometer semi-thin sections were cut with a glass knife and stained with 1% toluidine blue in 0.5% borax buffer. When a location of interest was found, 60-nm thin sections were cut with a diamond knife and collected on 300-mesh grids. After being stained with 3% uranyl acetate for 15 min and 1.5% lead citrate for 3 min, the samples were examined using a JEM-1400 transmission electron microscope (JEOL, Tokyo, Japan).
For SEM, organoids were washed for 30 minutes using cold Matrisperse Cell Recovery Solution to remove the Matrigel and fixed for another 30 minutes with 2.5% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.4). After fixation, the cells were washed three times (for 1 minute each) with 0.1 M sodium cacodylate buffer (pH 7.4), post-fixed for 10–15 minutes with 1% osmium tetroxide (OsO4) in the same sodium cacodylate buffer, and then washed. The fixed organoids were progressively dehydrated using a graded ethanol series, critical point-dried using hexamethyldisilazane (HMDS), and sputter-coated with gold/palladium (Au/Pd). The cells were then examined by field emission scanning electron microscopy (FE-SEM; Sigma 500; Zeiss, Oberkochen, Germany).
G. Bulk RNA-seq and DEG analysis
RNA was purified using the RNeasy kit (Qiagen, Hilden, Germany). Approximately 200 ng of total RNA was used to construct the library for sequencing via the NEBNext® Ultra RNA Library Prep Kit for Illumina (New England Biolabs, Ipswich, MA, USA). Next, paired-end sequencing was performed using the NovaSeq 6000 sequencing instrument (Illumina, San Digo, CA, USA) according to the manufacturer’s instructions, yielding 150-bp paired-end reads. The sequences were aligned to the mus musculus genome assembly GRCm39 using STAR version 2.7.8a. RSEM version 1.3.3. was used in combination with the STAR program. The default parameters were used in STAR and RSEM72-75. Differential expression analysis and enrichment tests were conducted using the DESeq2 v1.34.0 method to identify DEGs in the RNA-seq data. In this analysis, the expected counts were normalized, and thresholds of a log2 fold change (log2FC) ≥ 3 or ≤ -3 were applied, with a q-value < 0.01, to determine significant changes in gene expression76. The overall expression pattern was visualized in scatter, moving average, and principal component analysis plots generated using the ggplot2 R package. A volcano plot of the significant genes and heatmap of DEGs in the bulk RNA-seq were created using the EnhancedVolcano tool (available at https://github.com/kevinblighe/EnhancedVolcano) and hclust2 package v3.6.2 (available at https://github.com/SegataLab/hclust2), respectively. Functional classification analysis of the DEGs from the bulk RNA-seq was performed using the Metascape (https://metascape.org/) tool. This enabled us to predict the biological pathways that are statistically significantly related to the identified gene set77 (we used an enrichment factor of 3 and a p-value threshold of 0.01 for filtering purposes).
H. Single-cell RNA sequencing of organoids
For droplet-encapsulation scRNA-seq, > 5,000 cells were collected for analysis. A sequencing library was prepared using the Chromium Single Cell Reagent Kit v2 according to the manufacturer’s protocol (10x Genomics, Pleasanton, CA, USA). The cell suspension was diluted to 500 cells per mL and quickly processed using the high-throughput Gemcode platform from 10x Genomics with v3 chemistry reagents. This system captured each cell in a droplet alongside a gel bead containing oligo(dT) primers, a unique cell barcode, and unique molecular identifiers (UMIs) within lysis buffer. After that, the transcriptomes were converted to cDNA through reverse transcription. RT-PCR amplified the cDNA, and libraries were prepared from the 3' ends according to the manufacturer's protocol. The generated scRNA-seq libraries were sequenced on the NovaSeq 6000 Illumina platform using a high-output flow cell. We obtained sufficient paired-end reads to generate ~100–150 million fragments. After sequencing, cell barcodes and UMIs were used to demultiplex the source cells and mRNA transcripts from the PCR-amplified cDNA.
For the scRNA-seq analysis, raw sequencing data for genes detected in > 60% of the cells were processed using 10x Genomics Cell Ranger software (v6.1.2). The Seurat R package (v3.1.3) was employed to analyze gene expression data from droplet-based sequencing experiments, including variable gene selection, dimension reduction, and clustering. All genes expressed in > 1% of cells, along with cells expressing over 500 genes with < 10% mitochondrial genes, were retained for differential expression analysis. Cluster type specificity scores were defined for each gene and cluster. Correlations among shared marker genes were examined using cell clusters of organoids as reference points. Differential gene expression analysis of all cell types in the dataset was performed, and marker genes were identified based on expression differences and statistical similarities (cut-off: log2FC ≥ 2, ≤ -2, and p-value < 0.01) via the Seurat package. This allowed us to focus on the genes that were most likely to play a role in HC development. UMAP plots and expression landscapes of key cellular markers critical to HC development were visualized using Loupe software (10x Genomics). Heatmaps displaying the expression patterns of cellular markers, categorized by cell type as identified in previous studies, were generated using the ggplot2 package in R. Additionally, functional classification and gene network analysis of cluster 11, identified as an inner hair-like cell type, were conducted using STRING analysis.
I. Multielectrode assay
The multielectrode assay plate was coated with a 0.1% PEI solution the day before seeding cells. After coating, the organoids were seeded in a Matrigel dome onto the Multielectrode assay plate. Recording of the spontaneous neural activity of the organoids was performed using the Maestro Edge™ through Axion software (AxIS version, Axion BioSystems, Atlanta, GA). The cells were maintained at 37 °C with 5% CO2, regulated by the Maestro instrument. Recording of the electrical activity of the organoids was for 20 minutes. Raw data was acquired using AxIS by Navigator at a sampling frequency of 12.5 kHz, and digitally filter the data using a Kaiser Window filter with a high pass of 0.2 kHz and a low pass of 3 kHz.
J. Statistical analysis
The datasets were analyzed using Prism software 8.4.3 (GraphPad Software, La Jolla, CA, USA; RRID:SCR_002798). The stereocilia length and FM1-43-positive cell counts were compared between the MI and MACS groups using two-tailed Mann–Whitney U tests (nonparametric) and unpaired nonparametric t-tests. To analyze multiple treatment groups for MACS, a nonparametric one-way analysis of variance (ANOVA) was performed, and Tukey’s multiple comparison test was used as a post hoc test. P-values < 0.05 were considered to indicate significance.