Development of a physiological model of human middle ear epithelium

Abstract Introduction Otitis media is an umbrella term for middle ear inflammation; ranging from acute infection to chronic mucosal disease. It is a leading cause of antimicrobial therapy prescriptions and surgery in children. Despite this, treatments have changed little in over 50 years. Research has been limited by the lack of physiological models of middle ear epithelium. Methods We develop a novel human middle ear epithelial culture using an air‐liquid interface (ALI) system; akin to the healthy ventilated middle ear in vivo. We validate this using immunohistochemistry, immunofluorescence, scanning and transmission electron microscopy, and membrane conductance studies. We also utilize this model to perform a pilot challenge of middle ear epithelial cells with SARS‐CoV‐2. Results We demonstrate that human middle ear epithelial cells cultured at an ALI undergo mucociliary differentiation to produce diverse epithelial subtypes including basal (p63+), goblet (MUC5AC+, MUC5B+), and ciliated (FOXJ1+) cells. Mature ciliagenesis is visualized and tight junction formation is shown with electron microscopy, and confirmed by membrane conductance. Together, these demonstrate this model reflects the complex epithelial cell types which exist in vivo. Following SARS‐CoV‐2 challenge, human middle ear epithelium shows positive viral uptake, as measured by polymerase chain reaction and immunohistochemistry. Conclusion We describe a novel physiological system to study the human middle ear. This can be utilized for translational research into middle ear diseases. We also demonstrate, for the first time under controlled conditions, that human middle ear epithelium is susceptible to SARS‐CoV‐2 infection, which has important clinical implications for safe otological surgery. Level of Evidence NA.

Quantitative fluorescent polymerase chain reaction (PCR) was used to exclude common aneuploidies. The temporal bone was micro-dissected to identify the middle ear cavity and middle ear mucosal tissue was extracted.

| Air-liquid interface cultures
The tissue was dissociated into a cell suspension and passed through a 100-μm cell strainer (Greiner Bio-One, Kremsmünster, Austria) and centrifuged. Cells were seeded onto a collagen-coated flask with PneumaCult-Ex Plus (Stemcell Technologies) and cultured in a humidified atmosphere containing 5% CO 2 at 37 C. Cells were expanded over 7 to 10 days with media changes every 48 to 72 hours.
Upon reaching, 70% confluence cells were passaged onto Costar transwell membranes (Sigma Aldrich, St Louis, Missouri), with 0.4 μm pore size, pre-coated with collagen. Cells underwent expansion in the transwell system using PneumaCult-Ex Plus (Stemcell Technologies) for 4 days with media supplied to both the apical and basolateral compartments. On day 4, the media was removed from the apical compartment and the cells were exposed to air on their apical surface only. After 4 weeks, cultures underwent subsequent validation experiments.

| Immunohistochemistry
Transwell membranes were fixed in 4% paraformaldehyde. The specimen was dehydrated through graded ethanol baths, infiltrated with paraffin wax, and embedded into paraffin blocks at 60 C. The tissue blocks were then sectioned at 5-μm thickness using an RM2235 Microtome (Leica). The Ventana Discovery Autostainer (Ventana, Tucson, Arizona) was used for immunohistochemistry (IHC) following the manufacturer's instructions.

| Immunofluorescence
The transwell membranes were washed and fixed in 4% paraformaldehyde, permeabilized with 0.3% Triton and blocked with 5% bovine serum albumin and incubated with the primary and secondary antibodies (Supporting Information S1) as previously described. 3 Transwell scaffolds were then fixed to glass slides and imaged on an Axioimager without immersion (Zeiss, Oberkochen, Germany).

| Transmission electron microscopy
Membrane-bound epithelial cells were fixed, dehydrated, and impregnated with increasing concentrations of epoxy resin. They were then polymerized at and sections were cut on an ultramicrotome (70 nm). Sections were stained with 1% uranyl acetate and 2% lead citrate on an EM AC20 automatic staining machine (Leica) before being viewed on a Hitachi HT7800 transmission electron microscope as described previously. 4

| Scanning electron microscopy
Membrane-bound epithelial cells were fixed in 2.5% glutaraldehyde,

| SARS-CoV-2 Infection
A clinical isolate of SARS-CoV-2 (England/2/2020) was obtained from Public Health England. The initial viral stock was propagated in Vero E6 cells as previously described. 5 Infections were performed in a containment level 3 facility. 2 Â 10 5 pfu of SARS-CoV-2 was added to the apical side of two wells of middle ear epithelial cells differentiated at an ALI and one well of confluent Vero E6 cells (passage 10). Two wells of middle ear epithelium and one well of Vero E6 cells were left untreated (mock control).
All samples were incubated (37 C/5% CO 2 ) for 2 hours. Subsequently, the samples were washed and fresh media added. The cells were returned to the incubator until cell collection.

| Quantitative Reverse Transcription (RT) PCR
Mock and SARS-CoV-2 infected cells were lysed with Trizol reagent (Invitrogen) for RNA extraction, with resultant RNA being reverse transcribed with Superscript III (ThermoFisher Scientific) according to manufacturer's instructions. cDNA templates underwent qPCR as previously described. 6 Two PCR probes against the SARS-CoV-2 nucleocapsid (N) genes (N1/N2) were tested using the 2019-nCoV RUO kit (Integrated DNA Technologies, Coralvillle, Iowa) in accordance with manufacturer's instruction. The qPCR was then run using an AriaMx real-time PCR system (Agilent Technologies, California). Expression was determined by ddCT where the nucleocapsid expression was normalized to the endogenous housekeeping gene, RNAseP, and duplicate samples averaged.

| Air-liquid interface cultures
On the first and only attempt at running these cultures middle ear epithelial cells were observed to form discrete islands in the days following initiation of submerged culture which gradually connected one another from approximately day 7. A typical epithelial "cobblestone"

| Electron microscopy
Scanning electron microscopy (SEM) reveals a surface with moderately dense cilia coverage (Figure 4). Transmission electron microscopy (TEM; Figure 4B) similarly demonstrates cilia (green arrow) and also the presence of tight junctions (red arrow).

| Ussing chamber
Tight epithelial junctions were confirmed by resistance measurements on Ussing chamber experiments. Selective ion channel inhibitors (Table 1) were sequentially added to assess individual ion channels;    indicating the production of a complex family of cell populations which more closely resemble the physiological environment in vivo. 18 We also examined other markers of known importance, such as  26 This supports the need for appropriate personal protective equipment in otological surgery. Finally, in patients with tympanic membrane perforations and chronic ear discharge who have been exposed to SARS-CoV-2, it seems probable that such discharge could be contaminated with infected middle ear epithelial cells, and therefore, such clinical specimens ought to be considered infectious until proven otherwise.