Sars-Cov-2 Infects an Upper Airway Model Derived from Induced Pluripotent Stem Cells

Abstract As one of the primary points of entry of xenobiotic substances and infectious agents into the body, the lungs are subject to a range of dysfunctions and diseases that together account for a significant number of patient deaths. In view of this, there is an outstanding need for in vitro systems in which to assess the impact of both infectious agents and xenobiotic substances of the lungs. To address this issue, we have developed a protocol to generate airway epithelial basal-like cells from induced pluripotent stem cells, which simplifies the manufacture of cellular models of the human upper airways. Basal-like cells generated in this study were cultured on transwell inserts to allow formation of a confluent monolayer and then exposed to an air-liquid interface to induce differentiation into a pseudostratified epithelial construct with a marked similarity to the upper airway epithelium in vivo. These constructs contain the component cell types required of an epithelial model system, produce mucus and functional cilia, and can support SARS-CoV-2 infection/replication and the secretion of cytokines in a manner similar to that of in vivo airways. This method offers a readily accessible and highly scalable protocol for the manufacture of upper airway models that could find applications in development of therapies for respiratory viral infections and the assessment of drug toxicity on the human lungs.

both infectious agents and xenobiotic substances of the lungs. To address this issue, we have developed a protocol to generate airway epithelial basal-like cells from induced pluripotent stem cells, which simplifies the manufacture of cellular models of the human upper airways. Basal-like cells generated in this study were cultured on transwell inserts to allow formation of a confluent monolayer and then exposed to an air-liquid interface to induce differentiation into a pseudostratified epithelial construct with a marked similarity to the upper airway epithelium in vivo.
These constructs contain the component cell types required of an epithelial model system, produce mucus and functional cilia, and can support SARS-CoV-2 infection/replication and the secretion of cytokines in a manner similar to that of in vivo airways. This method offers a readily accessible and highly scalable protocol for the manufacture of upper airway models that could find applications in development of therapies for respiratory viral infections and the assessment of drug toxicity on the human lungs.

Significance statement
Demonstration of the ability of SARS-CoV-2 to infect an airway construct generated from induced pluripotent stem cells is significant since it paves the way for broader studies of viral airway infection using a system that can be manufactured reproducibly.
[Correction added on 10 July 2021, after first online publication: Author name Shaun H. Penington has been changed to Shaun H. Pennington and Ivo Djidrovksi has been changed to Ivo Djidrovski.]

| INTRODUCTION
The outbreak of the novel coronavirus disease, COVID-19, caused by coronavirus SARS-CoV-2 has been designated as a pandemic by the World Health Organization and is currently a significant threat to human health. At the time of submission (February 2021), there were 110 958 312 million confirmed cases, of whom 2 454 901 had died. 1 The pandemic has resulted in major challenges to global healthcare systems and has severe consequences for the global economy if the spread of the virus is not effectively controlled.
The causative agent of COVID-19, SARS-CoV-2, has been shown to infect the respiratory system resulting in viral pneumonia, but it may also affect the gastrointestinal system, heart, kidney, liver, and central nervous system leading to multiple organ failure. 2,3 Previous studies have shown that SARS-CoV predominantly infects airway and alveolar epithelial cells, and macrophages 4 using the angiotensinconverting enzyme 2 (ACE2) receptor for entry. 5,6 Rapid viral replication in these cells can lead to epithelial cell apoptosis causing the release of pro-inflammatory cytokines, 7 which can potentially cause airway damage and diminished patient survival. This is exemplified by the observation that in SARS-CoV-2-infected individuals, interleukin (IL)-6, IL-10, and tumor necrosis factor α (TNFα) surge during illness and decline during recovery. 8 Severely affected patients who require intensive care treatment can be distinguished by significantly higher levels of IL-6, IL-10, and TNFα and fewer CD4+ and CD8+ T cells. 9 Although it is likely that the ingress of large numbers of cytokinesecreting inflammatory macrophages into the lung tissue accounts for a significant proportion of the cytokines detected in such cases, the initial damage to the airway epithelial cells probably contributes not only to the overall concentration of cytokines but may also be responsible for recruitment of inflammatory macrophages.
Other mechanisms besides apoptosis can lead to activation of the inflammasome. The binding of SARS-CoV-2 to the Toll-like receptor causes the release of pro-IL-1β, which is cleaved by caspase-1, followed by inflammasome activation and production of active mature IL-1β that mediate lung inflammation, fever, and fibrosis. 10 To underline this, suppression of pro-inflammatory IL-6 has been shown to have a therapeutic effect in COVID-19. 11 IL-1β can also enhance the constitutive detachment (or shedding) of an enzymatically active ectodomain fragment of ACE2 from the airway epithelial cells, an event associated with acute lung injury. 12,13 Interestingly, SARS-CoV-2 infection is associated with ACE2 downregulation and ectodomain shedding thought to be induced by the SARS-CoV-2 Spike (S) protein. 14,15 How the released form of ACE2 (the so-called soluble or sACE2) causes lung damage is not completely clear but it seems to be tightly coupled to TNFα production, so it may be involved in inflammatory response to SARS-CoV-2 infection. 16 Despite this understanding of the mechanisms by which SARS-CoV-type viruses damage the airway epithelia, there are no effective treatments for the resulting COVID-19 disease. Current management of COVID-19 is supportive, and respiratory failure from acute respiratory distress syndrome is the leading cause of mortality. In view of this, there is an urgent and currently unmet need for model systems that can function as high-throughput preclinical tools for the development of novel, effective therapies for COVID-19.
The use of in vitro models mimicking the human airways generated from primary pulmonary epithelial cells grown at an air-liquid interface (ALI) has increased in popularity over the recent years. The cells can form a pseudostratified airway epithelium composed of all in vivo relevant cell types found in the airway epithelium including rare types such as pulmonary neuroendocrine cells and ionocytes. 17 This type of model has proven to be particularly useful for toxicological assessment of aerosol particles, 18,19 drug discovery, 20 and more recently with the COVID-19 outbreak for viral infection studies. 21 Although these models offer significant advantages, their availability is limited to primary samples, which can significantly differ depending on the donor's genetic background and thus affecting the generated airway.
Induced pluripotent stem cells (iPSCs) offer the potential to complement the limitations of primary cells by generating a large supply of cells with the genetic background of the donor. The pathways involved in differentiation into proximal and distal airway lineages have already been established by previous groups, who successfully differentiated functional ciliated cells, mucus-producing cells, and alveolar cells. [22][23][24][25][26] The current methods of lung differentiation use 3D self-forming spheroids or 2D cultures of mixed population of cells, which are more challenging when performing experiments similar to the primary airway ALI models. [27][28][29] In this study, we isolated a population of basal-like cells from differentiating iPSCs and used these to Mature ALI cultures prepared in this manner were fixed directly on the membrane with 4% paraformaldehyde (PFA) for 10 minutes at 3 C and then washed with PBS (3 Â 1.0 mL). The tissues were then removed together with the membrane, placed into molds, and embedded in optical coherence tomography (OCT) matrix (Cell Path, Newtown, UK, KMA-0100-00A). The molds were placed at À20 C to solidify. Once solid, they were sectioned into 5-to 10-μm slices on slides using a cryostat. The sectioned membrane was removed with PBS washes, and the slides were then stained using the same procedure as the basal cells using the same antibodies with the addition of antibodies (Tables 1 and   2). Once the staining was finalized, a few drops of Vectashield medium containing Hoechst were added to the slides then they were covered with coverslips, sealed with nail polish, and left to dry at 4 C.

| Characterization of basal epithelial cells by immunofluorescence and flow cytometry
Basal airway-like cells generated in this manner were characterized by a combination of flow cytometry (Fortessa flow cytometer and FloJo analysis) and immunofluorescence (IF) (see Table 1    settes. Subsequently, paraffination was performed using the Excelsior AS Tissue processor (Thermo Fisher Scientific, A82300001), and the paraffinized tissues were placed into molds. Once the paraffin had solidified, 3-μm sections were made using a microtome. The slides were rehydrated using xylene and an ethanol series (100%, 96%, and 70%). Subsequently, the slides were stained using Mayer's hematoxylin (Sigma-Aldrich, MHS32) and alcoholic eosin Y (Sigma-Aldrich, HT110116) followed by dehydration using an ethanol series (80%, 96%, and 100%). Xylene-washed slides were mechanically covered using coverslips and dried at room temperature. Histology was assessed using an AxioVert 25 inverted microscope (Zeiss).

| Quantification of cilia beat frequency
Prior to imaging, the apical surface of the ALI cultures was washed using medium from the basal chamber. Subsequently, high-speed videos were captured using the Nikon Eclipse Ti2 LIPSI high content imaging microscope equipped with a phase 1 phase contrast ring and a CFI S Plan Fluor LWD Â20 objective. The middle of the Prime BSI sCMOS camera was used to capture 550 images (512 Â 512 pixels) over 5.5 seconds at a rate of 100 frames per second. During imaging, the atmosphere was constantly kept at 37 C, 5% CO 2 , and 95% humidity. At least three fields containing cilia for three inserts were imaged.
For analysis MATLAB was used to calculate cilia beat frequency.
Here, the intensity-time trace of each pixel was filtered for frequencies between 5 and 25 Hz using a band-pass filter. Subsequently, the power spectrum density was calculated. The frequency of each pixel is defined by the highest peak of the power spectrum density. The ciliated cells were visualized by plotting the frequency of each pixel into a heat map.
The frequency distribution of the ciliated cells was visualized by plotting the amount of nonzero pixels into a histogram. The average frequency of all nonzero pixels was compared using a student's t test.  Organoids were embedded in 1% low-melting agarose at 30 C containing Slowfade diamond mounting oil and imaged using a Zeiss LSM880 confocal laser scanning microscope. At Â400, sections were 3D reconstructed from a series of z-stacks (20-40-mm slices) with automatic optimal thickness and 1 Airyscan unit pinhole. Orthogonal views were generated using Zeiss Zen 3.3 software.

| Quantification of cytokine release after infection with SARS-CoV-2
Quantification of cytokines was performed using the V-PLEX Plus  Figure 2D) and the expression of mucin-1 (goblet cells, Figure 2D). The presence of putative pulmonary neuroendocrine cells is indicated by the expression of synaptophysin ( Figure 2D). The expression of all the four markers was similar between iPSC-derived and primary lung cell-derived ALI constructs ( Figure 2D), although for MUC1 and CC10, the expression in the primary lung cell-derived ALI constructs was stronger and local- In conclusion, collectively our data show that airway epithelial constructs generated from iPSCs-derived basal cells have promise as models of the response of the airway epithelia to viral infections.
F I G U R E 5 Lung airway constructs secrete inflammatory cytokines in response to SARS-CoV-2 infection. Data shown as mean ± SEM, n = 3

| DISCUSSION
In this study, we established an efficient method of isolation of iPSCsderived basal-like cells from a mixed population of lung progenitors.
Comparably with primary cells, when grown in BEGM and 3T3-J2 feeders, the iPSC-derived basal cells can self-renew and form colonies.
They express genes associated with a basal cell phenotype (KRT14, deltaNp63, NGFR, integrin alpha 6) while maintaining their multipotent capacity. One other group has reported the isolation of a basal cell-like population from lung spheroids 44

CONFLICT OF INTEREST
The authors declared no potential conflicts of interest.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.