CX3CR1^GFP/GFP C57/BL6 mice were purchased from the Jackson Laboratory (Bar Harbor, ME, USA) via Orient Bio (Seongnam, Korea). All mice were maintained in a specific pathogen free environment in the animal facility at Avison Biomedical Research Center at Yonsei University College of Medicine (Seoul, Korea). Temperature, humidity, and light/dark cycle were properly controlled, and food and water were accessible ad libitum. Less than five animals were maintained in a single cage. CX3CR1^+/GFP mice were produced by crossing wild-type C57/BL6 mice (Orient Bio) with CX3CR1^GFP/GFP mice. Randomly selected male and female mice (aged 4-8 weeks and weight 15–25 g) were used in the experiments, and all animals were carefully maintained in cages and were not exposed to noise. After general anesthesia, 5 µL of lipopolysaccharide (LPS; 5 mg/mL) was inoculated in the mouse middle ear through the tympanic membrane in the experimental group. The concentration and amount of injected LPS were based on previous studies [13,14]. The experimental group mice were sacrificed at 1, 4, 7, and 10 days after LPS inoculation. Control group mice were sacrificed without any injection or sacrificed 4 days after 5 µl of PBS inoculation in the middle ear. The right side ear was used in all experimental animals. All experiments were performed during the daytime in the animal laboratory. A heating pad (37°C) was used to maintain the body temperature of the animal during general anesthesia. The animal experiments were approved by the Institutional Animal Care and Use Committees at the Yonsei University College of Medicine (Project No. 2019-0182).

After sacrificing the mice in a CO2 chamber, the bilateral temporal bones of the mice were obtained. The mouse temporal bones were fixed in 4% paraformaldehyde for 24 hours in a 4°C room. After fixation, samples were incubated in 1 mL of 0.25 M ethylenediaminetetraacetic acid (EDTA) solution for 24 hours at 4°C for decalcification. Decalcified cochlea were cleared using Binaree Clearing Kits (Binaree, Daegu, Korea). Cleared cochlea were mounted on a silicone bond in the petri dish, with spiral ligaments were positioned at the top. Eighty percent of glycerol was then dropped on the sample, and a cover slide was positioned on it. Distilled water was filled between cover slide and water-immerse lens of the two-photon microcopy.

Two-photon microscopy (LSM7MP; Carl Zeiss, Oberkochen, Germany) was used for imaging data generation. Zen software (Carl Zeiss) was used for image acquisition and basic image analysis. To create three-dimensional images and videos, we used IMARIS software (Bitplane, Zurich, Switzerland). The Mai-Tai HP Ti:Sa Deep See laser system (Spectra-Physics, Mountain View, CA, USA) was used to generate excitation lasers with a wavelength of 810 nm. Images were acquired at a resolution of 512× 512 pixels using band-pass filters with 420–480 nm (blue), 500– 550 nm (green), and 575–610 nm (red). A 20× object lens was used in all experiments.

Quantification was performed using Zen software (Carl Zeiss), and the type III fibrocyte area, which is parallel to X-Y plane, was investigated. Regions of interest (ROIs) were drawn as a circle with area of 25,000 µm² lateral to the stria vascularis. The number of green cells were counted manually between cortical bone and the stria vascularis in each ROI. Kruskal-Wallis test with post-hoc was used to evaluate the significance (P<0.05) of differences between groups. Statistical analysis was performed by IBM SPSS ver. 25.0 (IBM Corp., Armonk, NY, USA).

We obtained three-dimensional images of intact cochleae from CX3CR1^+/GFP mice using two-photon microscopy (Fig. 1, Supplementary Video 1). The cochlear bony capsule was colored blue, and capillaries were colored white because of endogenous fluorescence. Spiral ligaments appeared as dim endogenous fluorescence, whereas the stria vascularis showed brighter fluorescence. Capillaries in the spiral ligament, including the stria vascularis vessels, were clearly seen with their characteristic branching pattern. Resident macrophages appeared in bright green color. Even though the intensity of fluorescence grew dimmer as the depth increased, macrophages in the spiral ligament were clearly distinguishable.

To reduce selection bias, we captured extended images of 100 µm in length along the cochlear turn (images were projected to the sagittal plane). The apical and basal turns showed similar patterns of macrophage distribution (Fig. 2A and B). Notably, the spiral ligament macrophages were not evenly distributed in the sagittal plane. The lateral edge of the spiral ligaments showed dense distributions of macrophages. Perivascular macrophages contacting stria vascularis vessels were also found, and several macrophages were seen at Reissner’s membrane. However, resident macrophages were scarce in the area adjacent to the stria vascularis. In particular, the wide area from the crista basilaris to beneath the stria vascularis showed a lack of macrophages, even though this area was thought to be the primary site where macrophages reside [2].

Upon inflammation, macrophages and monocytes are known to migrate robustly into the cochlea. We induced cochlear inflammation using LPS, which was inoculated in the middle ear through the tympanic membrane. At 1, 4, 7, and 10 days after LPS injection, GFP-positive cells (macrophage and monocytes) were present in higher numbers in the cochlea than in the PBS injection group (Fig. 3). Interestingly, the distribution patterns in the spiral ligament were similar to those in the untreated cochlea, even though macrophage density had significantly increased (Fig. 2). In addition, migration of green cells was identified in the basilar membrane and spiral limbus, a finding which is consistent with previous studies [2,5].