Under otoscopic view, the purulent exudate from the inside of the right ear canal was collected from the patient using a pair of sterile microforceps and obtained using a flocked ESwab transport system (480CE; Copan Diagnostics, Brescia, Italy). Obtained specimens were immediately stored in ESwab Liquid Amies preservation medium and transferred to the clinical microbiology laboratory. The specimen was loaded into the WASPLab laboratory automation system (Copan Diagnostics) for automatic inoculation on 5% sheep blood, chocolate, and MacConkey agar plates (all Autobio, Zhengzhou, China). The inoculated agar plates were moved to a digital imager via conveyor belt to obtain the first series of images for the initial plating step. The plates were moved into the WASPLab incubator in an environment with an ambient air temperature of 35°C, where they were maintained for 16 and 40 h. At each incubation time point, a second (16 h) or third series of images (40 h) was acquired and evaluated for bacterial growth by skilled laboratory personnel, according to routine microbiology laboratory protocols.

Suspicious pathogen colonies were identified by matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS). Briefly, a portion of a single bacterial colony was applied directly to a target slide (bioMérieux, Marcy l’Etoile, France) using a disposable 1-ml loop. Immediately, 1 ml of Vitek MS-CHCA matrix solution (α-cyano-4-hydroxycinnamic acid; bioMérieux) was added directly to the bacterial film and allowed to dry at room temperature, followed by mass spectrometric analysis. The mass spectrometer was calibrated using a freshly prepared Escherichia coli ATCC 8739 reference strain. Individual colonies were analyzed by the Vitek MS system (bioMérieux) in linear positive-ion mode with Vitek MS Acquisition Station software (version 2.0; bioMérieux). The similarities between each sample and the reference isolates in the MS-ID database were calculated and expressed as an optimized quantitative value (confidence value).

For confirmation of the bacterial identification obtained from MALDI-TOF MS analysis, the presumptive Vibrio strain was sub-cultured on CHROMagar™ Vibrio agar (CHROMagar, Paris, France). Concurrently, clinical strains of V. vulnificus and V. cholera were inoculated on chromogenic agar plates to serve as controls. All plates were incubated at 35°C and were observed to determine colony morphologies and coloration at 24 and 48 h, in accordance with the instructions of the manufacturer.

The confirmatory identification of isolates was also performed using two commercially available biochemical identification systems, including API 20E and 20NE strips and the Vitek-2 Compact system (bioMérieux). According to the recommendations of the manufacturer, a single bacterial colony was resuspended in 0.85% sodium chloride (NaCl; bioMérieux) using sterile swabs to prepare a 0.5-McFarland suspension. The suspension was inoculated into the eight left-sided microtubes on an API 20NE strip and all of the microtubes on an API 20E strip. Subsequently, 200 μl of suspension was pipetted into 7-ml API AUX medium (bioMérieux) and inoculated into the remaining 12 right-sided microtubes on the API 20NE strip. Supplementary investigations of cytochrome oxidase were performed using Microbact Oxidase strips (Oxoid, Basingstoke, United Kingdom). All inoculated API strips were incubated at 35°C for 24 h, and the species identification results were obtained using the API database included in APILAB Plus software (bioMérieux). Simultaneously, the 0.5-McFarland bacterial suspension was transferred onto a Vitek-2 Gram-negative identification card (GN; bioMérieux), and the card was loaded into the Vitek-2 Compact system for bacterial identification. The E. coli ATCC 25922 strain was used as quality control.

Further confirmation of bacterial identification was performed through the polymerase chain reaction (PCR)-based amplification of the 16S rRNA gene using the universal primers 27F (5′-AGAGTTTGATCCTGGCTCAG-3′) and 1492R (5′-TACGGCTACCTTGTTACGACTT-3′). PCR conditions consisted of 95°C for 5 min; 35 cycles of 95°C for 1 min and 55°C for 1 min; annealing at 72°C for 2 min; and a final extension step at 72°C for 10 min. PCR amplification was performed in an ABI 7500 thermocycler (Applied Biosystems, Foster City, CA, United States), and the resulting amplified product was verified by size separation on a 1% agarose gel. The DNA sequences of the purified PCR products were identified using an ABI Prism 3700 Genetic Analyzer (Applied Biosystems) and compared with the sequences deposited in the GenBank database from the National Center for Biotechnology Information (NCBI) by using BLAST.¹

Following identification, the Vibrio isolate was subjected to antimicrobial susceptibility testing to analyze its antibiotic-resistant phenotype. According to the recommendations of the manufacturer, 145 μl of freshly prepared 0.5-McFarland bacterial suspension was diluted into 3 ml of 0.85% NaCl and transferred onto a Vitek-2 Gram-negative antimicrobial susceptibility testing card (AST-GN13; bioMérieux). Subsequently, the card was loaded into the Vitek-2 Compact system for automatic incubation, and the minimum inhibitory concentration (MIC; μg/ml) results were reported through the software. Antimicrobial susceptibility was interpreted according to the current M45-S3 guidelines of the Clinical and Laboratory Standards Institute (CLSI). The E. coli ATCC 25922 strain was used as quality control for antimicrobial susceptibility testing using the broth microdilution method.

For the generation of bacterial supernatant, the clinical Vibrio strain was cultured in brain-heart infusion broth (BHI; BD Laboratories, Franklin Lakes, NJ, United States) supplemented with 0.85% NaCl at 35°C for 16 h. Subsequently, the culture broth was diluted with fresh BHI medium at a ratio of 1:10, allowing bacteria to grow for 5 h at 35°C, to reach the logarithmic growth phase. The bacterial culture was pelleted by centrifugation at 5,000 × g and 4°C for 10 min, and 1 × 10⁶ colony-forming units (CFU)/ml bacteria were resuspended in high-glucose Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS; both Thermo Fisher Scientific, MA, United States) or Basal Medium Eagle (Sigma-Aldrich/Merck KGaA, Darmstadt, Germany). After incubating for 8 h, the bacterial suspension was centrifuged, and the resulting supernatant was filtered through a 0.22-μm filter (Millex GP; Millipore/Merck KGaA, Darmstadt, Germany). The accuracy of the bacterial concentration was confirmed by quantitative cultures on 5% sheep blood agar plates.

Post-natal day 3 (P3) Sprague–Dawley rats were obtained from the Laboratory Animal Center of the Fourth Military Medical University. The experimental procedures for the animal study were approved by the Animal Care Committee of Fourth Military Medical University. To minimize the number of animals used and their suffering, the rat pups were sacrificed by rapid decapitation, and their skulls were opened midsagitally. For the preparation of TM keratinocyte cultures, both sides of the external ears were removed, and the tympanic bulla was isolated under a dissecting microscope (Stemi 508; Carl Zeiss, Göttingen, Germany). The TM was dissected and rinsed with ice-cold Hank’s Balanced Salt Solution (HBSS) containing 2% penicillin–streptomycin solution (both Thermo Fisher Scientific, MA, United States) prior to processing. For the preparation of cochlear organotypic cultures, rat cochleae were promptly separated from the temporal bone, rinsed with ice-cold HBSS, and collected for further use.

Harvested TMs were adhered to a single 35-mm culture dish with a glass-bottom (15-mm glass diameter; NEST Biotechnology, Wuxi, China) previously coated with Cell-Tak (Corning Life Science, Corning, NY, United States), with the lateral surface of the TM facing down. Defined keratinocyte-serum-free medium (dKSFM; Thermo Fisher Scientific) containing antibiotics was added to the TM explant cultures, and the cultures were incubated in an atmosphere of 5% CO₂ and 95% humidity at 35°C. Keratinocyte outgrowths from each explant were allowed to form a homogeneous cell monolayer. Upon reaching 80–90% confluence, cells were passaged by enzymatic digestion and fed with high-glucose DMEM containing 10% FBS. House Ear Institute-Organ of Corti 1 (HEI-OC1) cells (a gift from Dr. Renjie Chai, Southeast University, Nanjing, China) were cultured in high-glucose DMEM containing 10% FBS in an atmosphere of 5% CO₂ and 95% humidity at 33°C.

Tympanic membrane keratinocytes and HEI-OC1 cells within passages 3–5 were used for further experiments. Cells from experimental cultures were incubated with fresh medium supplemented with bacterial supernatant at a 1:1 ratio and maintained for 24 h (keratinocytes) or 12 h (HEI-OC1 cells) prior to further analysis. Cells incubated in supernatant-free medium alone were used as the baseline control.

The viabilities of TM keratinocytes and HEI-OC1 cells were quantitatively assessed using the cell-counting kit-8 (CCK-8; Beyotime, Beijing, China). Both types of cells were seeded in 96-well plates at a density of 5.0 × 10³ cells/well using three replicates per group and subjected to the previously described treatments. After cultivation, the cell culture was incubated with 100 μl of fresh medium containing 10 μl CCK-8 reagent in each well for 60 min at 37°C. Cell viability was calculated by measuring the optical densities of each well at 450 nm using a microplate reader (Bio-Rad Laboratories, Hercules, CA, United States).

Tympanic membrane keratinocytes were seeded into the reservoirs of a two-well culture insert that was adhered to a 35-mm glass-bottom dish (Ibidi, Martinsried, Germany), at a density of 5.0 × 10³ cells/insert, and cultured overnight at 37°C. Once the cells reached confluence, the insert was removed from each dish using a pair of forceps, and the dish was gently rinsed with phosphate-buffered saline (PBS) to remove cell debris. The cells were then cultured in fresh medium, either with or without bacterial supernatant supplementation, and placed into a Cytation 5 Cell Imaging Multi-Mode Reader (BioTek Instruments, Winooski, VT, United States). Phase-contrast images of the wound gaps were viewed and captured under a ×4 objective lens using Gen5 software (BioTek Instruments) at 1-h intervals over 24 h. The wound gap areas at times 0 and 24 h were analyzed with ImageJ software (version 1.46r; National Institutes of Health, Bethesda, MD, United States) to calculate repair rates (expressed in μm²/h).

Apoptosis rates for TM keratinocytes and HEI-OC1 cells treated with bacterial supernatant were detected by flow cytometry according to the protocol supplied with the FITC Annexin V/propidium iodide (PI) apoptosis detection kit (BD Pharmingen, San Diego, CA, United States). In brief, cells were plated in a 60-mm culture dish at a density of 1.0 × 10⁶ cells/dish and incubated in medium with or without bacterial supernatant supplementation at 37°C in a 5% CO₂ atmosphere. Cells from each experimental group were dissociated with 0.25% trypsin, pelleted by centrifugation, and washed twice with ice-cold PBS. For flow cytometric analysis, harvested cells were resuspended in 1× binding buffer at a concentration of 1 × 10⁶ cells/ml. A 100-μl volume of binding buffer containing 5 μl of Annexin V-FITC and 5 μl of PI solution was added to each sample and incubated for 15 min in the dark. After adding an additional 400 μl of binding buffer, all samples were evaluated using a flow cytometer (FACS Canto™ II; BD Biosciences).

For immunofluorescent analysis of cell apoptosis, TM keratinocytes and HEI-OC1 cells were plated in a 35-mm glass-bottom culture dish (NEST Biotechnology) at a density of 5.0 × 10⁴ cells/dish and treated with bacterial supernatant, as described previously. Apoptosis was determined by terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) using an in situ cell death detection kit (TMR red; Roche, Nutlet, NJ, United States), according to the instructions of the manufacturer. Briefly, cells in the culture dishes were fixed in 4% paraformaldehyde for 20 min at room temperature, and staining was performed using a TUNEL reaction mixture at 37°C for 60 min in a humidified atmosphere. After washing, the cell samples were permeabilized and blocked with a PBS solution containing 5% bovine serum albumin (BSA; Sigma-Aldrich) and 0.1% Triton X-100. Cells were then labeled at 4°C overnight using antibody solution (1% BSA and 0.1% Triton X-100 in PBS) containing the following primary antibodies: monoclonal mouse anti-cytokeratin (CK)-19 primary antibody (diluted 1:200; GeneTex, Irvine, CA, United States) for keratinocytes, or polyclonal rabbit anti-Myosin-VIIa antibody (diluted 1:1,000; Proteus Biosciences, Ramona, CA, United States) for HEI-OC1 cells. Immunolabeling was further visualized with appropriate secondary antibodies: Alexa Fluor 488-conjugated donkey anti-mouse IgG or Alexa Fluor 488-conjugated donkey anti-rabbit IgG (both diluted 1:500; Thermo Fisher Scientific) diluted in antibody solution for 2 h at room temperature. In parallel, all specimens were randomly processed without incubation with primary antibodies, which served as the negative control. After nuclear counterstaining with Hoechst 33342 (diluted 1:1,000; Thermo Fisher Scientific) for 15 min, all samples in the culture dishes were mounted with Prolong Gold anti-fade mounting medium (Thermo Fisher Scientific) and randomly visualized on a confocal microscope system (FV1000; Olympus, Tokyo, Japan). The confocal images were converted to TIFF format using FV10-ASW software (version 4.2; Olympus) and processed for optimal brightness and contrast with Adobe Photoshop software (CC 2020; Adobe Systems, San Jose, CA, United States). The TUNEL/Hoechst double-positive cells were calculated, and the number was normalized against the total number of viable cells to determine the TUNEL-positive rate.

The experimental protocol for cochlear organotypic cultures was performed as previously described (Ding et al., 2011; Prakash Krishnan Muthaiah et al., 2017). A drop (20 μl) of cool, type I rat tail collagen gel (Corning Life Science) in 10× basal medium Eagle (BME; Sigma-Aldrich) supplemented with 2% sodium carbonate (Sigma-Aldrich) at a 9:1:1 ratio was placed in a 35-mm culture dish at room temperature for 15 min to initiate gelation. Subsequently, 1.2 ml of serum-free medium, consisting of 1× BME (Sigma-Aldrich), 1% serum-free supplement (100× insulin-transferrin-selenium-X supplement; Sigma-Aldrich), 1% BSA, 2 mM glutamine, 5 mg/ml of glucose, and 1% antibiotics, was added to cover the collagen gel in each dish. Using a dissecting microscope, the membranous cochlea was revealed from the cochlear capsule of a P3 rat. After removing the spiral ligament and stria vascularis, the remaining auditory sensory epithelium containing the organ of Corti and SGNs was placed evenly on the collagen gel and maintained overnight at 37°C with 5% CO₂ before exposure to the bacterial supernatant. On the following day, the cochlear explants were incubated with fresh medium supplemented with or without bacterial supernatant at a 1:1 ratio and maintained for 4 h prior to immunofluorescent analysis. After fixation and permeabilization, HCs and SGNs from the organotypic cultures were labeled with monoclonal mouse anti-β-III Tubulin (TUJ1) antibody (diluted 1:500; GeneTex) and polyclonal rabbit anti-Myosin-VIIa antibody (diluted 1:1,000), followed by incubation with Alexa Fluor 647-conjugated donkey anti-mouse IgG and Alexa Fluor 488-conjugated donkey anti-rabbit IgG (both diluted 1:500; Thermo Fisher Scientific). In parallel, cochlear specimens were randomly processed without incubation with primary antibodies, which served as a negative control. Alexa Fluor 594-conjugated phalloidin (diluted 1:500; Thermo Fisher Scientific) was used for F-actin labeling within 15 min. After counterstaining with Hoechst 33342, all explant samples were subsequently mounted with Prolong Gold medium and photographed under a confocal microscope.

All experimental data are shown as the mean ± standard deviation (SD) and were analyzed using the Statistical Program for Social Science software (SPSS, version 26.0; IBM Inc., Armonk, NY, United States). Two-tailed, unpaired Student’s t-tests were used to assess significant differences between two groups. Values of p less than 0.05 (p < 0.05) were considered significant.

As depicted in Figure 1A, the audiometric evaluation of the patient displayed a mild sensorineural hearing loss, with a mild drop at high frequencies in both ears. Computed tomography of the temporal bone showed normal middle ear and mastoid air cells, with no signs of cholesteatoma or otomastoiditis. An otoscopic examination revealed some white purulent discharge on the deep wall of the external auditory canal and the lateral surface of the TM in the right ear (Figure 1B1). The exudate was removed with a pair of microforceps, collected using a sterile swab, and transferred to the clinical microbiology laboratory for microorganism culture. After sampling, otoscopic visualization of the right ear showed an intact, retracted, and opaque TM, with local calcification in the pars tensa and some granulated tissue on the deep posterior wall of the ear canal near the TM (Figure 1B1′). The patient was prescribed topical ciprofloxacin eardrops, applied three times per day. After 4 days, the organism recovered from the exudate sample was identified as V. alginolyticus, and the strain showed susceptibility to most antimicrobial agents, including levofloxacin. The patient was finally diagnosed with chronic otitis externa caused by V. alginolyticus. Following a successful course of antibiotics, the patient eventually recovered after a follow-up of 1 week. The otoscopic examination also demonstrated local calcification in the intact and thickened TMs on both sides of the ears (Figures 1B1″,B2), and an additional ear swab culture was negative for V. alginolyticus.

Two days after plating, cultures of the exudate specimen obtained from the infected ear and inoculated on sheep blood agar revealed a large growth of pure, gray, non-hemolytic bacterium colonies and some scattered white Staphylococcus colonies (Figure 2A). Gram staining of the isolate showed a Gram-negative, curved, rod-shaped morphology resembling Vibrio spp. (Figure 2B). The strain was identified at the species level using MALDI-TOF MS analysis, revealing a 99.9% coefficient of variation for identification as V. alginolyticus (Supplementary Figure 1). Sub-cultured strains were identified as grayish colonies on sheep blood (Figure 2C1) and chocolate (Figure 2C2) agar media and as transparent, glucose-non-fermenting colonies on MacConkey agar plate (Figure 2C3). For confirmation of the V. alginolyticus identification, the isolate was sub-cultured overnight on a CHROMagar™ Vibrio agar plate. The colonies that grew on the chromogenic media appeared creamy or colorless, as expected for V. alginolyticus (Figure 2D1), whereas the control colonies of V. vulnificus (Figure 2D2) and V. cholera (Figure 2D3) appeared green-blue to turquoise-blue.

Additionally, a biochemical investigation of the identified Vibrio strain was performed using a Vitek-2 GN ID card on a Vitek-2 Compact system, and the API 20E (Figure 2E) and 20NE (Figure 2F) system also identified the isolate as V. alginolyticus, which was positive for the following tests: growth in nutrient broth supplemented with 1 and 6% NaCl; production of acid from glucose, mannitol, and saccharose; utilization of glucose, mannitol, N-acetyl-glucosamine, gluconate, and malate; lysine decarboxylase; cytochrome oxidase; nitrate reduction; gelatin hydrolysis; indole; and motility. By contrast, the strain was negative for growth in nutrient broth without NaCl supplementation; H₂S production; production of acid from inositol, sorbitol, rhamnose, melibiose, amygdalin, arabinose, cellobiose, lactose, and salicin; utilization of arabinose, mannose, maltose, caprate, adipate, citrate, and phenylacetate; urea hydrolysis (urease); arginine dihydrolase; ornithine decarboxylase; Voges–Proskauer (VP); β-galactosidase (4-nitrophenyl-β-D-glucopyranoside, PNPG); and O-nitrophenyl-b-D-galactopyranoside (ONPG). The isolate was biochemically identical to the reactions reported for V. alginolyticus among four common pathogenic Vibrio species listed in Table 1. Molecular identification using 16S rRNA gene sequencing was also performed, indicating that the species was most closely related to V. alginolyticus (Supplementary Data), and the overall identity score was 100%, with no mismatches.

Antimicrobial susceptibility testing of the V. alginolyticus strain was performed using a Vitek-2 AST-GN13 card on a Vitek-2 Compact system and interpreted according to the CLSI M45-S3 breakpoints for Vibrio spp. The results revealed that the isolate was sensitive to piperacillin, ampicillin/sulbactam, piperacillin/tazobactam, ceftazidime, cefepime, imipenem, meropenem, gentamicin, amikacin, levofloxacin, ciprofloxacin, and trimethoprim–sulfamethoxazole but resistant to ampicillin (Table 2).

To explore the potential cytotoxic effects of V. alginolyticus on rat TM keratinocytes, we isolated keratinocytes from TM tissue explant cultures obtained from P3 rats, and the cells were treated with V. alginolyticus culture supernatants diluted 1:1 for 24 h. The keratinocyte outgrowth from the TM explant formed a monolayer of epithelial cells that displayed a tightly packed, cobblestone-like morphology (Figure 3A). When maintained in normal medium, the passaged keratinocytes proliferated promptly and formed large numbers of homogenous clones. After exposure to the bacterial supernatant for 24 h, the proliferation of keratinocytes was markedly compromised, accompanied by a cytopathic effect that manifested as the shrinkage and rounding of individual cells (Figure 3B). To quantify cell proliferation, CCK-8 assays were performed, which revealed that the cell viability of keratinocytes treated with Vibrio supernatant decreased compared with that of the control culture (Figure 3C). To assess the functionality of TM keratinocytes, a time-lapse wound-healing migration assay was performed, in which the repair rates of wound gaps were analyzed. As shown in Figure 3D, healthy epidermal cells maintained under control conditions proliferated and migrated to close the wound gap created by the two-well culture insert over 24 h (Supplementary Video 1). Conversely, the wound-healing repair rate for keratinocytes cultured in Vibrio-derived conditioned medium was significantly reduced relative to the normal control (Supplementary Video 2). To investigate the cell death mechanisms induced by Vibrio supernatant, keratinocytes incubated under different conditions were stained with Annexin V-FITC/PI or TUNEL. Flow cytometric analysis was performed with Annexin V/PI double staining, which indicated that the proportions of apoptotic and dead cells were significantly increased in the bacterial supernatant-treated culture compared with the control culture (Figure 3E). In addition, the number of TUNEL-positive apoptotic cells was significantly higher after exposure to Vibrio supernatant compared with control conditions (Figure 3F).

To evaluate the possible effects of V. alginolyticus on HEI-OC1 cells, we treated the cells with the bacterial supernatant at the same concentration for 12 h. After exposure to Vibrio-derived conditioned medium, HEI-OC1 cells appeared as grape-like aggregates of swollen, refractile cells (Figure 4A). The results from the CCK-8 assay demonstrated significantly decreased viability among HEI-OC1 cells treated with supernatant (Figure 4B). Similarly, exposure to bacterial supernatant resulted in a significantly higher ratio of apoptotic and dead HEI-OC1 cells compared with the control condition (Figure 4C). In addition, Vibrio supernatant treatment increased the number of dead cells that exhibited TUNEL/Hoechst double-positive staining and condensed nuclei (Figure 4D). These results indicated that the V. alginolyticus strain exerted cytotoxic effects against TM keratinocytes and HEI-OC1 cells by inhibiting cell proliferation and migration and inducing cell apoptosis and death.

To study the ototoxic effects of V. alginolyticus on sensory HCs and SGNs, we analyzed the cell density and morphological integrity of cochlear organotypic explants after exposure to a 1:1 dilution of bacterial culture supernatant for 4 h. As shown in Figure 5, the ototoxic damage was quantified (Figures 5D,E) at the apical (Figure 5A), middle (Figure 5B), and basal (Figure 5C) turns of the cochlear explants, and the effects on HCs and SGNs were assessed by immunostaining the whole mounts with the HC marker Myosin 7a, a phalloidin dye that stains stereocilia bundles, and the neuronal marker TUJ1. Significantly reduced HC and SGN survival was observed in the basal turn compared with the corresponding cochlear regions in control cultures. In detail, the morphology of Myosin 7a-labeled inner HCs and phalloidin-labeled stereocilia bundles were entirely absent, the density of TUJ1-labeled auditory neurons and nerve fibers were significantly reduced, and the outer spiral fiber bundles of type II SGNs were swollen and disarranged (Figure 5C). By contrast, most of the HCs and SGNs were maintained in the apical turns of the cochlear explants exposed to the bacterial supernatant (Figure 5A). These results indicated a high susceptibility of HCs and SGNs to V. alginolyticus-induced ototoxicity in vitro, with increased vulnerability in the basal and middle cochlear regions.