Wild-type C57BL/6 mice and CX3CR1^GFP/GFP transgenic C57BL/6 mice were purchased from Jackson Laboratory (ME, United States) via Orient Bio (Sungnam, Republic of Korea). CX3CR1^GFP/+ mice were used to visualize monocytes and macrophages. The mice were housed and maintained according to the requirements for animal research established by Yonsei University Health System, and all procedures were approved by the Institutional Animal Care and Use Committee (approval number 2019-0182). Mice (4–12 weeks old, both sexes) were fed ad libitum and housed in cages in an environmentally controlled room under a 12-h light cycle. We used 4 mice for validating permanent hearing loss after acoustic overstimulation, 8 mice in each condition for flow cytometry experiment, 3 mice in each condition for imaging studies, 4 mice in each condition for polymerase chain reactions, and 3 mice in each group for RNA sequencing.

White noise (300–10,000 Hz) was generated using a personal computer and an amplifier (R-399; Inter M, Seoul, South Korea) and delivered through speakers (290-8L; Altec Lansing, Oklahoma City, OK, United States) in a noise booth. Mice were continuously exposed to a 120-dB peak equivalent sound pressure level for 1 h to produce permanent threshold shift and acoustic overstimulation-induced cochlear inflammation.

The auditory brainstem response (ABR) was measured after anesthetizing the mice with xylazine [20 mg/kg, intraperitoneal injection (IP)] and ketamine hydrochloride (30 mg/kg IP). The hearing level of each mouse was checked by measuring the ABR threshold with a TDT auditory evoked potential workstation (Tucker-Davis Technologies, Alachua, FL, United States). Both ears of each mouse were stimulated with an ear probe sealed in the ear canal. The ABRs to click and tone burst stimuli were recorded, and thresholds were obtained for each ear. The ABRs were measured before and at 2 and 7 dpn.

After sacrificing the mice in a CO₂ chamber, the bilateral temporal bones were dissected and fixed in 4% paraformaldehyde for 24 h at 4°C after local perfusion with a fixative through the oval and round windows. After fixation, the samples were incubated in 0.125M ethylenediaminetetraacetic acid (ETDA) solution for 24 h at 4°C for decalcification. To obtain lateral wall samples, decalcified cochleae were cut in half through the apex-oval window axis. Under optical microscopy, the cochlear lateral wall of the basal turn was carefully separated using forceps and microscissors. Other structures, including the organ of Corti, Reissner’s membrane, and modiolus, were trimmed.

The tissues were blocked with 10% donkey serum and incubated with allophycocyanin (APC)-conjugated anti-Ly6C antibodies (HK1.4; cat. no. 128016; Biolegend) at 4°C overnight. The samples were then mounted with a mounting solution (Sigma-Aldrich, St. Louis, MO, United States) and viewed under an LSM980 confocal microscope (Zeiss, Jena, Germany). All immunostaining experiments were repeated in three biological replicates per time point. For three-dimensional (3D) imaging, lateral wall samples were incubated with fluorescein isothiocyanate-conjugated anti-Ly6C antibodies (HK1.4; cat. no. 128005; Biolegend) at 4°C overnight. The samples were then directly investigated under a two-photon microscope (LSM7MP; Carl-Zeiss, Germany) using the same method as in our previous study (Bae et al., 2021b).

We calculated the outer hair cell count as described previously (Bae et al., 2020). Briefly, after fixation and decalcification as described above, the cochlea was separated into apical and basal portions under an optical microscope. Then, the organ of Corti was directly visualized under a two-photon microscope (LSM7MP; Carl-Zeiss).

We performed intravital imaging of the collecting venule as described previously (Bae et al., 2021b). Briefly, the cochlea of anesthetized CX3CR1^GFP/+ mice was exposed by surgery, and Texas-red conjugated dextran (500 μg/animal) was injected into the retro-orbital sinus. The stapedial artery was ligated, and the bony capsule of the basal turn was carefully drilled until the collecting venule could be visualized by optical microscopy. Then, a two-photon microscope (LSM7MP; Carl-Zeiss) was used to acquire imaging data. Images were obtained for 30 min (1 frame/min). The cells that appeared in four consecutive sequences (a 3-min interval) were selected manually and connected to create a track using IMARIS software (Bitplane AG, Zurich, Switzerland).

Cochleae were obtained from mice following cardiac perfusion with phosphate-buffered saline (PBS). We carefully removed the soft tissues from the cochlea in ice-cold PBS. The cochlea was then transferred to a new dish filled with ice-cold PBS, and the bony capsule was carefully removed, ensuring that the bone marrow was not exposed. We discarded the remnants after extracting the inner tissue of the cochlea, including the lateral wall, the organ of Corti, and modiolus, through the opening. The tissue from the cochlea was collected and trypsinized (0.25% trypsin/ETDA solution) for 10 min at 37°C. The trypsinized tissue was ground on a filter (pore size, 40 μm). Peripheral blood was obtained by cardiac puncture before cardiac perfusion. Samples were treated with red blood cell lysis buffer to remove red blood cells, and the remaining cells were stained with antibodies at a 1:200 dilution for 30 min. Samples were then analyzed using a FACSverse II BD flow cytometer (BD Biosciences, Sparks, MD, United States). One sample consisted of four cochleae from two mice, and four samples were included in each group. The flow cytometry data were managed using FlowJo software (Tree Star, Ashland, OR, United States). Peripheral blood was used as a reference for gating strategies for inflammatory monocytes in the cochlea because the immune cells infiltrated from the peripheral blood (Supplementary Figures 1, 2).

7-Aminoactinomycin D (7-AAD; cat. no. 420403; Biolegend) was used for live cell gating, phycoerythrin (PE)-Cy7-conjugated anti-CD11b (M1/70; cat. no. 25-0112-82; Invitrogen, Carlsbad, CA, United States) was used for myeloid cell gating, PE-conjugated anti-Ly6G (1A8; cat. no. 127608; Biolegend) was used for neutrophil gating, APC-conjugated anti-F4/80 (BM8; cat. no. 17-4801-82; Invitrogen) was used for monocyte/macrophage gating, and APC-Cy7-conjugated anti-Ly6C (HK1.4; cat. no. 128026; Biolegend) was used for inflammatory monocyte gating (Supplementary Figures 1, 2). For monocyte depletion, 0.2 mL (5 mg/mL) clodronate liposomes (Liposoma) was intraperitoneally injected into mice 1 day before acoustic overstimulation.

Bilateral temporal bones were separated from the body immediately after the sacrifice. Immediately after the cochlea was isolated from the vestibule, it was immersed in TRIzol reagent (Invitrogen). Three biological replicates (6 cochleae from 3 mice) were used in each group, untreated and 3 h after noise exposure. Transcriptome sequencing was performed using an Illumina platform. We used the HISAT2 software to map the reads to the mouse genome (mm10_NCBI_108) and generate gene expression values in the normalized form of transcripts per kilobase million values. All differentially expressed genes (DEGs) were selected based on the following cut offs: p < 0.05 and |fold change| ≥ 2. Gene-set enrichment analysis and pathway analysis were conducted using public databases. Functional enrichment with Gene Ontology (GO) was performed using the g:Profiler.¹ The Kyoto Encyclopedia of Genes and Genomes (KEGG) database was used to analyze pathway enrichment.²

Bilateral temporal bones were separated from the body immediately after the sacrifice. Immediately after the cochlea was isolated from the vestibule, it was immersed in TRIzol reagent (Invitrogen). Four biological replicates (8 cochleae from 4 mice) were used in each group that was 3 h, 1 day, 5 days after noise exposure, and untreated control. Total RNA was isolated and purified using a PureLink RNA mini kit (Ambion, Austin, TX, United States), according to the manufacturer’s instructions. RNA samples from the cochlea were reverse-transcribed using an iScript Select cDNA Synthesis Kit (Bioneer, South Korea). We performed RT-qPCR using Taqman Master Mix (Applied Biosystems, Foster City, CA, United States). The PCR mixture contained 400 μg cDNA and primers. To quantify the mRNA levels of interleukin 1 beta (Il1-β), interleukin-6 (Il-6), and C-C motif chemokine ligand 2 (Ccl2) in the cochlea, we used gene-specific primer pairs with glyceraldehyde 3-phosphate dehydrogenase (GAPDH) control primers for normalization and TaqMan probes (Applied Biosystems). Amplification and quantification were performed using an ABI 7500 RT-PCR system (Applied Biosystems). Assay ID of used TaqMan probes are listed below.

Il1-β: Mm00434228_m1

Il-6: Mm00446190_m1

Ccl2: Mm00441242_m1

GAPDH: Mm99999915_g1

To compare multiple groups, we performed one-way analysis of variance (ANOVA) and post hoc Tukey’s multiple comparisons test. To compare multiple time points in multiple groups, we performed two-way ANOVA and post hoc Dunnet’s multiple comparisons test. All data are presented as means and standard errors of the means (error bars). SPSS 25.0 (IBM, Armonk, NY, United States) and Prism 8.0 (GraphPad Software, San Diego, CA, United States) were used for statistical analyses. Results with P values less than 0.05 were considered statistically significant.

Acoustic overstimulation resulted in a permanent threshold shift in the hearing function of mice at all examined frequencies (Figure 1A). Importantly, this hearing function did not recover after 2 or 7 dpn. Outer hair cell degeneration was also identified in the noise-exposed cochlea, particularly in the basal turn, which is known to be vulnerable to acoustic overstimulation (Sha et al., 2001) (Figure 1B). In previous studies, increased macrophage infiltration was observed from 3 dpn, and blood-derived monocytes seem to transform into macrophages after infiltration (Hirose et al., 2005; Shi, 2010; He et al., 2020). Therefore, in this study, we investigated the infiltration of monocytes into the cochlea at 2 dpn. Indeed, flow cytometry analysis indicated that the 2 dpn cochlea contained CD11b⁺Ly6G^–F4/80⁺Cx3cr1⁺Ly6C⁺⁺ cells, which are inflammatory monocytes based on flow cytometry data from the peripheral blood (Figures 1C,D and Supplementary Figure 1).

Next, we treated peripheral blood with clodronate liposomes to deplete monocytes and confirm that the monocytes in the 2 dpn cochlea were recruited from the blood stream. Clodronate liposomes significantly reduced the infiltration of monocytes into the 2 dpn cochlea. The absolute cell count of the CD11b⁺Ly6G^–F4/80⁺CX3CR1⁺Ly6C^– population, representing resident macrophages, was not significantly different between groups (Figures 1E–G). Taken together, these findings suggested that the inflammatory monocytes infiltrating into the cochlea after acoustic overstimulation were mainly derived from the peripheral blood.

Next, we conducted serial flow cytometry analysis to determine the time course of the monocyte/macrophage response in the cochlea after acoustic overstimulation. Inflammatory monocytes were significantly increased at 1 and 2 dpn, followed by increases in macrophage numbers (Figures 2A,B). Interestingly, the total CD11b⁺Ly6G^–F4/80⁺CX3CR1⁺ population was not significantly changed after 2 dpn, suggesting that the infiltration of monocytes was limited within 2 dpn and that these cells may subsequently be transformed into macrophages.

Next, we evaluated CD11b⁺Ly6G⁺CX3CR1^– neutrophils using different gating strategies in the same dataset (Figures 2C,D and Supplementary Figure 2). For clear distinguishment of neutrophils, CD11b + myeloid cells were sorted by GFP and PE signals to exclude the possible spillover phenomenon resulting from PE-conjugated anti-Ly6G antibodies and green fluorescent protein (GFP)-CX3CR1 reporter mice. The Ly6G⁺CX3CR1^– neutrophils were distinct in the peripheral blood. Interestingly, neutrophils were rare (less than 5% in average from the CD11b⁺ population) at any time point after acoustic overstimulation (Figure 2E). However, although statistically insignificant, several neutrophils were detected in 1 and 2 dpn. We next performed an immunofluorescence study of the lower spiral ligament, using Ly6C and Ly6G antibodies to confirm the presence of neutrophils. Indeed, several neutrophils were identified in the cluster of inflammatory monocytes at 1 dpn (Figure 2F). However, as the flow cytometry analysis indicated, the inflammatory monocytes were a substantially large number compared to neutrophils.

The density plot of the flow cytometry data was analyzed to investigate changes in the expression of Ly6C and CX3CR1 proteins (Figure 3). In the peripheral blood, there are two major CD11b⁺Ly6G^–F4/80⁺CX3CR1⁺ populations showing differential expression of Ly6C; these populations are patrolling monocytes (Ly6C^–) and inflammatory monocytes (Ly6C⁺⁺) (Geissmann et al., 2003). When compared with patrolling monocytes in the peripheral blood, cochlear macrophages showed high expression of CX3CR1 at all time points. There was no newly observed population suspected to be infiltrated patrolling monocytes after acoustic overstimulation. By contrast, inflammatory monocytes abruptly appeared at 1 dpn. This population showed a gradual reduction in Ly6C expression and increase in CX3CR1 expression. Finally, these populations merged with the resident macrophage population at 5 dpn. We confirmed that there was no additional infiltration of inflammatory monocytes after 2 dpn; instead, the infiltrated monocytes were transformed into macrophages by 5 dpn.

According to the flow cytometry results, we performed immunofluorescence analysis of the lateral wall of the cochlea at 1 and 2 dpn (Figure 4). We observed distinct adherence of inflammatory monocytes in the collecting venule, which was located beneath the organ of Corti in the lateral wall, at 1 dpn (Figure 4B). Furthermore, in the cochlea at 2 dpn, the number of adhered cells appeared to decrease, and inflammatory monocytes were distributed in the lower part of the spiral ligament (Figure 4C). At 3 dpn and 5 dpn, CX3CR1 positive cells were increased and Ly6C signal of these cells was decreased compared to 1 dpn and 2 dpn (Figures 4D,E). These findings were consistent with the flow cytometry results and with the two-photon image using wild-type mice (Supplementary Figure 3). The crawling monocytes were also found in intravital imaging of the collecting venule in the basal turn at 1 dpn (Supplementary Video 1). Taken together, these findings suggested that the interaction between immune cells and the endothelium began at approximately 1 dpn and that infiltration mainly occurred at around 2 dpn.

Our results revealed that the infiltration of innate immune cells may occur earlier than expected. To support our results at the molecular level, we analyzed DEGs from RNA sequencing data to evaluate the expression of genes associated with monocyte infiltration at an extremely early stage (3 h) after acoustic overstimulation. In total, 585 genes were significantly different between control and noise-exposed cochleae. Among these genes, 340 were upregulated, and 245 were downregulated in noise-exposed cochleae compared with control cochleae. The GO analysis of molecular functions showed strong enrichment in protein binding function (Figure 5A). However, in the biological process category, processes associated with leukocyte migration were not within the top 20 terms (Figure 5B). In KEGG pathway analysis, the cytokine-cytokine receptor interaction pathway, which includes chemokines that attract monocytes from the blood flow, was the third most differentially expressed pathway (Table 1). Additionally, among the top 20 DEGs, four genes were chemokines (Ccl2, Ccl7, Cxcl10, and Ccl12; Table 2). The expression of Ccl2 was increased by 29.89-fold. Importantly, qPCR results confirmed the RNA sequencing data, demonstrating significant increases in Ccl2 expression at 3 h and 1 dpn (∼25 fold) and then a decrease to the baseline level at 5 dpn. Furthermore, Il-1b and Il-6 expression levels increased shortly after acoustic stimulation, although these changes were not statistically significant (Figures 5C–E).