
==== Front
Am J Respir Cell Mol Biol
Am J Respir Cell Mol Biol
ajrcmb
American Journal of Respiratory Cell and Molecular Biology
1044-1549
1535-4989
American Thoracic Society

38843491
2023-0356MA
10.1165/rcmb.2023-0356MA
Major Technical Advances
Cell Culture Differentiation and Proliferation Conditions Influence the In Vitro Regeneration of the Human Airway Epithelium
Redman Elisa 1 2
Fierville Morgane 1 2 3
Cavard Amélie 1
Plaisant Magali 1
Arguel Marie-Jeanne 1 2
Ruiz Garcia Sandra 1
McAndrew Eamon M. 1 2
Girard-Riboulleau Cédric 1
Lebrigand Kevin 1 2
Magnone Virginie 1 2
Ponzio Gilles 1 2
Gras Delphine 4
Chanez Pascal 4
Abelanet Sophie 1
https://orcid.org/0000-0001-9632-6483
Barbry Pascal 1 2 3
Marcet Brice 1 2 *
https://orcid.org/0000-0001-6747-7928
Zaragosi Laure-Emmanuelle 1 2 *
1 Institut de Pharmacologie Moléculaire et Cellulaire, Centre National de la Recherche Scientifique (CNRS), Institut National de la Santé et de la Recherche Médicale (INSERM), and Université Côte d’Azur,
2 IHU RespirERA, and
3 Interdisciplinary Institute for Artificial Intelligence (3IA Côte d’Azur), Université Côte d’Azur, Sophia Antipolis, France; and
4 Centre de Recherche en Cardiovasculaire et Nutrition, Institut National de la Santé et de la Recherche Médicale (INSERM), and Institut National de Recherche pour L’agriculture, L’alimentation et L’environnement (INRAE), Université Aix-Marseille, Marseille, France
Correspondence and requests for reprints should be addressed to Laure-Emmanuelle Zaragosi, Ph.D., Université Côte d’Azur, Centre National de la Recherche Scientifique, Institute Pharmacology Moléculaire et Cellulaire, 06560 Sophia-Antipolis, France. E-mail: zaragosi@ipmc.cnrs.fr. Brice Marcet, Ph.D., Université Côte d’Azur, Centre National de la Recherche Scientifique, Institute Pharmacology Moléculaire et Cellulaire, 06560 Sophia-Antipolis, France. E-mail: marcet@ipmc.cnrs.fr.
* These authors contributed equally to this work.

6 6 2024
1 9 2024
6 6 2024
71 3 267281
9 10 2023
6 6 2024
Copyright © 2024 by the American Thoracic Society
2024
https://creativecommons.org/licenses/by-nc-nd/4.0/ This article is open access and distributed under the terms of the Creative Commons Attribution Non-Commercial No Derivatives License 4.0. For commercial usage and reprints, please e-mail Diane Gern.

The human airway mucociliary epithelium can be recapitulated in vitro using primary cells cultured in an air–liquid interface (ALI), a reliable surrogate to perform pathophysiological studies. As tremendous variations exist among media used for ALI-cultured human airway epithelial cells, the aim of our study was to evaluate the impact of several media (BEGM, PneumaCult, Half & Half, and Clancy) on cell type distribution using single-cell RNA sequencing and imaging. Our work revealed the impact of these media on cell composition, gene expression profile, cell signaling, and epithelial morphology. We found higher proportions of multiciliated cells in PneumaCult-ALI and Half & Half, stronger EGF signaling from basal cells in BEGM-ALI, differential expression of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) entry factor ACE2, and distinct secretome transcripts depending on the media used. We also established that proliferation in PneumaCult-Ex Plus favored secretory cell fate, showing the key influence of proliferation media on late differentiation epithelial characteristics. Altogether, our data offer a comprehensive repertoire for evaluating the effects of culture conditions on airway epithelial differentiation and will aid in choosing the most relevant medium according to the processes to be investigated, such as cilia, mucus biology, or viral infection. We detail useful parameters that should be explored to document airway epithelial cell fate and morphology.

Keywords

airway epithelium
single-cell RNA sequencing
air–liquid interface cell culture
culture medium
differentiation
Fondation pour la Recherche Médicale 10.13039/501100002915 DEQ20180339158 Institut National de la Santé et de la Recherche Médicale 10.13039/501100002915 Chan Zuckerberg Initiative 10.13039/501100002915 2017-175159-5022 H2020 Health 10.13039/501100002915 Canceropôle PACA 10.13039/501100002915 Centre National de la Recherche Scientifique 10.13039/501100002915 Agence Nationale de la Recherche 10.13039/501100002915 21-ESRE-0052 HORIZON EUROPE Marie Sklodowska-Curie Actions 10.13039/501100002915 101072892
==== Body
pmcClinical Relevance

The human airway mucociliary epithelium can be recapitulated in vitro using primary cells cultured in an air–liquid interface, a reliable surrogate to perform pathophysiological studies. Our study provides a comprehensive repertoire for evaluating the effects of culture conditions on airway epithelial differentiation and will aid in choosing the most relevant medium according to the processes to be investigated.

The mammalian airways are lined by a mucociliary epithelium, composed of basal cells, club cells, goblet cells, multiciliated cells (MCCs), and rarer cell types (deuterosomal cells, ionocytes, pulmonary neuroendocrine and tuft cells, and microfold cells) (1–3). The cell composition of the epithelium varies according to the macroanatomical location, with the highest differences found between the nasal and tracheobronchial epithelia (4). Within the tracheobronchial airways, cellular distribution is relatively stable, with modifications occurring in the most distal bronchioles (4–7). The airway epithelium is frequently challenged by the inhalation of noxious compounds, chemicals, microorganisms, and viruses, which alter its integrity. Specific repair mechanisms allow full epithelial restoration so that normal physiological function can be reestablished (see Figure E1A in the data supplement). In chronic lung diseases, such as cystic fibrosis (CF), asthma, and chronic obstructive pulmonary disease, frequent injuries and chronic inflammation induce remodeling of the epithelium, often associated with a progressive loss of MCCs and an increased content of goblet cells. Studying the regeneration of the epithelium in normal or pathological conditions is necessary to identify mechanisms regulating the physiological regeneration and pathological remodeling of the airway epithelium. The development of cell cultures at the air–liquid interface (ALI) of nasal or bronchial epithelial cells has been crucial in many mechanistic studies (8, 9, 10). These cultures are not just convenient surrogates of in vivo epithelium (11–14); they were extensively used to establish the cellular roadmap of airway regeneration (11, 15) (see Figure E1A). This is illustrated by the recent approval of Kalydeco (Vertex Pharmaceuticals Incorporated) by the U.S. Food and Drug Administration, on the basis of work using primary cultures to predict clinical response in patients with CF bearing ultrarare CFTR mutations, for whom direct clinical trial studies are not feasible (16). Over the past three years, it has also become the most appropriate model for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in vitro infectivity studies. Although most studies have identified MCCs as the major entry cell type (17–22), others have shown a more widespread distribution (23, 24), some with nonciliated cells being infected preferentially over MCCs (25, 26). One striking difference between these studies was indeed the culture medium that was used.

Considering the large differences existing in cell culture conditions across the many studies using ALI cultures, it is important to define exactly the impact of the different experimental setups in term of cellular composition (8). Some studies have already performed direct comparison of cell culture media on ALI culture structure and/or function (27–34), but none has combined quantitative assessment of cell type distribution with differential gene expression profiles. The effect of using distinct commercial proliferation media during the cell expansion stage has never been reported either. We thus investigated the impact of four distinct proliferation and differentiation media on the epithelial structure and cellular composition of the reconstructed airway epithelium. We selected two commercially available and widely used media (PneumaCult and BEGM; Lonza), which we compared with Half & Half (H&H), which has been described by Susan Reynolds’s group and is considered to produce fewer biases compared with other media (35). The last medium is used by Clancy’s group to perform electrophysiological investigations (36). We used single-cell RNA sequencing (scRNA-seq) to quantify cell type distribution and gene expression profile variations across the four media. We also evaluated the impact of proliferation media on ALI differentiation and the variations caused by the use of distinct porous membranes. Given the widespread use of ALI culture in SARS-CoV-2 studies, we also evaluated the effect of these media on viral entry factor expression.

Some of the results of these studies have been previously reported in preprint form (www.biorxiv.org/content/10.1101/2024.03.16.584842v1.full).

Methods

To evaluate the impact of cell culture media on cell composition of fully differentiated ALI-regenerated airway epithelium, we first used primary cells that were freshly dissociated from human bronchi and set up proliferation and differentiation in four culture media: PneumaCult-ALI (hereafter Pneuma-ALI) and BEGM-ALI (i.e., the two most widely used media) in parallel with H&H (35) and Clancy medium (36). Freshly isolated human bronchial epithelial cells (HBECs) underwent one passage in PneumaCult-Ex Plus medium, then were split and seeded on Transwell (Corning) membranes in four distinct proliferation media: PneumaCult-Ex Plus (hereafter Pneuma-Ex+), BEGM, Wu medium, and Clancy medium. Once cells reached confluence, differentiation was induced by removing the medium in the apical chamber and adding matched differentiation medium in the basal chamber. The differentiation media were Pneuma-ALI, H&H, BEGM-ALI, and Clancy. scRNA-seq was performed at three time points: after the initial propagation step in plastic flasks, at the onset of the ALI (ALI0), and at full differentiation (ALI28) (see Figure E1B). To evaluate whether some differences in ALI differentiation might stem from the medium used at the initial cell propagation steps, we amplified human nasal epithelial cells (HNECs) for two passages on plastic, in either BEGM or Pneuma-Ex+ (Stemcell Technologies). We then seeded them on membranes, maintained cell amplification in the same medium until reaching confluence, and then set up the ALI in either BEGM-ALI or Pneuma-ALI (see Figure E1C). We evaluated epithelial composition using quantitative PCR (qPCR) and immunostaining. The effects of using alternative semiporous membranes were also evaluated. Detailed methods are described in the data supplement. scRNA-seq data are available at http://carra.ipmc.cnrs.fr:3838/CultureMedia2024/ and https://cellxgene.cziscience.com/collections/73cf6939-3caa-4105-bc57-e073ee885a28.

Results

ALI Differentiation Medium Influences Epithelial Morphology and Cell Type Distributions

After full differentiation of HBECs, epithelia were fixed for histological analysis and immunostaining. Hematoxylin and eosin staining on epithelia sections revealed large differences in tissue structures (Figure 1A). Pneuma-ALI and H&H media generated thicker epithelia with an apical surface covered with cilia. BEGM-ALI and Clancy media generated much thinner epithelia, with this effect more prevalent for Clancy medium. Cilia immunostaining with acetylated α-tubulin suggested a higher content of MCCs in Pneuma-ALI and H&H media than in BEGM-ALI and Clancy media (Figure 1A). MUC5AC+ goblet cell content was difficult to compare among culture media because staining patterns differed according to media with, for instance, smaller and more intense MUC5AC patches in Pneuma-ALI (Figure 1A).

Figure 1. Single-cell RNA sequencing analysis of human bronchial epithelial cells (HBECs) at 28 days of differentiation (ALI28) in four distinct media. (A) General characterization of HBEC differentiation in four cell culture media. (Top) Representative images of H&E staining of sections of HBECs at ALI28. Scale bars, 100 μm. For each condition and each of the two independent cell cultures, two inserts were sectioned and imaged. (Bottom) Representative images of immunostaining of HBECs at ALI28 for acetylated α-tubulin and MUC5AC. Scale bars, 30 μm. Nuclei were stained with DAPI and shown in blue on the merged images. For each condition and each of the two independent cell cultures, three inserts were stained and imaged. (B) Uniform manifold approximation and projection (UMAP) of the integrated dataset of HBECs at ALI28 containing all cells from the four distinct conditions, from two independent cultures from two healthy donors. (Top) UMAP colored by cell type. (Bottom) UMAP colored by medium. (C) UMAPs for each medium, colored by cell type, with color code identical to A. (D) Quantification of cell type proportions in each medium. (E) P value of t test using the propeller package, using the arcsine transformation, for each two-to-two medium comparison of cell type proportions. (F) F1 score derived from universal cell embeddings (UCE) label transfer against the Human Lung Cell Atlas core reference (8). The F1 score is calculated by comparing the predicted labels, obtained through the UCE label transfer method, against those manually annotated using marker genes and clustering; the maximal score is 1. Ac. Tub = acetylated α-tubulin; ALI = air–liquid interface; H&E = hematoxylin and eosin; H&H = Half & Half.

Thus, to obtain a more quantitative and detailed description of the regenerated epithelia in each culture medium, we performed 3′ scRNA-seq on fully differentiated HBEC cultures in each medium (ALI28). To reduce batch effects, we multiplexed all conditions using cell hashing (37). Datasets for all conditions and all donors were aggregated for subsequent analyses, allowing cell type quantification and differential gene expression among culture media (Figure 1B). Cell types were determined on the basis of the expression of known marker genes (see Figures E2A–E2C and Table E1). Basal cells were identified by high expression of KRT5 and TP63 and cycling basal cells by typical proliferation markers such as MKI67. Suprabasal cells were defined as cells located at the basal side of the epithelium but not directly lying on the basal lamina, expressing KRT5 but very low TP63. Secretory cells included not only club cells that were SCGB1A1+ and CYP2F1+ but also other cell types that did not necessarily express SCGB1A1 but instead markers such as AQP5 and FAM3D. Squamous cells were characterized by an intermediate signature between suprabasal and secretory cells and also through expression of specific genes such as SPRR1A, IVL, and SCEL. Goblet cells displayed a typical secretory cell gene expression program, with SCGB1A1, BPIFA1, and BPIFB1 together with the expression of MUC5B and/or MUC5AC. Of note, the length of the MUC5AC and MUC5B transcripts can affect their detection in scRNA-seq datasets and probably contributed to an underestimation of MUC5AC+ and MUC5B+ cells. MCCs were identified as FOXJ1+ and DYNLRB2+, while deuterosomal cells, which are cells amplifying centrioles, were FOXJ1+, CDC20B+, and DYNLRB2−. Rare cells formed a small but distinct cell cluster, characterized by the expression of the ionocyte-specific FOXI1, as well as HEPACAM2 and NREP, which we previously detected expressed by both ionocytes and neuroendocrine cells. We also detected STMN1, a specific marker of tuft cells, and MARCKSL1 and CRYM, expressed by both tuft and neuroendocrine cells as expected (1, 4).

Figure 1B shows the distribution of the different cell type clusters at ALI28, after aggregation of the four datasets (top) and colored according to the differentiation media (bottom). The uniform manifold approximation and projection representation is split in Figure 1C according to the four media, with the corresponding cell type distributions quantified in Figure 1D. Pneuma-ALI and H&H cultures displayed similar distribution profiles (Figures 1C and 1D) with, for instance, 15.7% and 19.5% of MCCs, respectively, and 13.1% and 17.7% of goblet cells, respectively (Figure 1D and Table E2). BEGM-ALI and Clancy media cultures displayed more diverse profiles characterized by much lower MCC content (3.8% for BEGM-ALI and 0.6% for Clancy medium) and a higher goblet cell proportion in BEGM-ALI (26.6%). We assessed the statistical significance of the differences among scRNA-seq conditions by performing pairwise comparisons using the propeller package (38, 39), which confirmed the significance of the observations for the multiciliated and goblet cell populations (Figure 1E). The Clancy medium also yielded a significant difference in basal cell content against all other media (Figure 1E). Compared with the luminal immunostaining from Figure 1A, the higher proportion of goblet cells and the absence of a significant decrease of MCCs in BEGM-ALI might seem surprising but can be explained by the greater tissue thickness in Pneuma-ALI and H&H. As the scRNA-seq dataset quantifies all cells, from the basal to the luminal compartment, the proportion of luminal cells in thick epithelia is decreased compared with thin epithelia. In addition, immunostaining detected MUC5AC+ cells only, as opposed to scRNA-seq, which identified goblet cells on the basis of the entire set of enriched genes, including MUC5B. We quantified goblet cells expressing MUC5AC alone, MUC5B alone, or both and found some significant differences between differentiation media, with BEGM and H&H producing more MUC5B-only and MUC5AC-only goblet cells, respectively (see Figures E3A and E3B).

We next sought to compare the regenerated epithelia with in vivo data to identify the in vitro conditions that best reproduced the healthy airway environment. We used a label transfer tool based on universal cell embeddings (39) to map each medium-specific dataset onto the Human Lung Cell Atlas (HLCA) core reference (1). We computed an F1 score by comparing the predicted labels, obtained through the universal cell embeddings label transfer method, against those manually annotated using marker genes and clustering. The F1 score evaluates the classification performance of the method, and the method should perform best on cells most resembling the in vivo reference. Figure 1F shows that the best F1 score was obtained by Pneuma-ALI media (0.5), followed by Clancy (0.42), H&H (0.39), and BEGM (0.37). We used label transfer to obtain prediction proportions between our in vitro and the in vivo reference, and we observed variable results depending on cell clusters (see Figure E4). Despite high mapping with HLCA basal cells, basal cells from BEGM-ALI and H&H also mapped with HLCA suprabasal cells, suggesting diverse gene expression profiles in these media. On the other hand, suprabasal cells from Pneuma-ALI and Clancy media better mapped onto their HLCA counterparts. MCCs from all media mapped almost perfectly with HLCA MCCs, but in Clancy and BEGM-ALI, MCCs also displayed high scores for deuterosomal cells, suggesting more immature MCCs in these media. Finally, the most striking difference was observed for Pneuma-ALI goblet cells, which did not map with HLCA goblet cells. Instead, cells that we classified as secretory in Pneuma-ALI mapped with goblet cells from HLCA (see Figure E4). These data indicate that secretory and goblet cells, which share very close expression profiles (see Figure E2B), might carry subtle differences between media that influence their classification.

Thus, scRNA-seq allowed the identification of differences among differentiation media, with a higher MCC content in Pneuma-ALI and H&H, as well as gene expression differences shown by differential mapping to an in vivo reference. To investigate further these findings, we next aimed to identify more specific effects of each differentiation medium on gene expression profiles within each cell cluster.

ALI Differentiation Medium Influences Gene Expression Profiles

We first performed media pairwise comparison using a pseudobulk strategy. The largest number of differentially expressed genes was obtained when comparing Pneuma-ALI with BEGM-ALI and Clancy (see Figure E5 and Table E3). In contrast, the smallest number of differentially expressed genes was observed when comparing H&H with Pneuma-ALI. We then displayed the top expressed genes in all pairwise comparisons. Some genes were specifically enriched in one medium, such as IL33 and ZBTB16 for Pneuma-ALI and Clancy, respectively (Figure 2A). IL33 is expressed in vivo in healthy lung, in basal and suprabasal cells, and in endothelial cells (1, 4). Interestingly, it is a T-helper cell type 2–oriented cytokine that is involved in asthma susceptibility (40). ZBTB16 encodes a zinc finger transcription factor that may play a role in the transcriptional memory of hormone stimulation (41). CYP26A1 was also expressed only by basal cells from Pneuma-ALI medium. This cytochrome-encoding gene contributes to retinoic acid clearing (42). As Pneuma-ALI is the only of the four media for which the composition is not publicly available, interpreting this difference concerning retinoic acid clearing was difficult. VIM was also enriched in cycling basal cells from Clancy medium, suggesting a mesenchymal-like phenotype. Pneuma-ALI and H&H shared several differentially enriched genes compared with the other media, such as the secreted peptides TFF3, ELAPOR1, and SCGB1A1. They also shared the upregulation of GLIPR2, a Golgi-associated protein that negatively regulates autophagy (43). On the other hand, BEGM-ALI and Clancy media shared enriched genes such as several anion exchangers (SLC5A5 and SLC34A2), secreted proteins (STATH, VSTM2L, and BPIFA2), and GCNT3, an N-acetylglucosaminyltransferase contributing to mucin glycosylation. Altogether, these data suggest significant differences in the secretomes by Pneuma-ALI and H&H on one hand and BEGM-ALI and Clancy on the other.

Figure 2. Genes and pathways regulated by each cell culture medium at ALI28 of bronchial epithelial cell differentiation. (A) Violin plot showing the top 30 regulated genes among all medium comparisons by pairs. Only genes expressed in at least 30% of cells of at least one cell cluster were selected. The top 30 genes were sorted by the lowest adjusted P value of each comparison by pair. For each gene, expression in the four tested media is shown. (B) Circle plots showing the inferred intercellular communication networks for EGF, NOTCH, and BMP signaling. The edge width is proportional to inferred interaction strengths considering ligand–receptor pairs. Black arrowheads show directionality of the inferred interactions. For B, the color code for cell types is identical to A. BMP = bone morphogenetic protein; EGF = epidermal growth factor; NOTCH = notch receptor.

Given the tremendous number of possible comparisons, we used CellChat (https://github.com/sqjin/CellChat) to stratify the different expressed pathways. CellChat quantifies outgoing and incoming signals per signaling pathway, at the level of each cell cluster, thus identifying cells involved in sending and receiving signals for each pathway in an autocrine and paracrine manner. When comparing the number of inferred interactions within each culture medium, we found that BEGM-ALI generated the largest number of interactions compared with all other media and that basal and suprabasal cells are the source of the major incoming signals (shown by top bars of incoming signaling patterns in Figure E6 and by Figures E7A and E7B). However, the communication probability of these interactions, represented by the interaction strength or “weight” calculated by CellChat, appeared equivalent among all media (see Figure E7A, right, and Figure E7B). Some pathways showed striking differences in communication probabilities, such as the EGF, NOTCH, BMP, and WNT pathways (Figures 2B and E7C). The EGF pathway showed stronger outgoing signals in BEGM-ALI and H&H media (see Figure E6) with basal and cycling basal cells sending EGF signals to all other cell types of the epithelium, including themselves (Figure 2B). This signal is consistent with the strong expression of AREG in cycling basal and basal cells of BEGM-ALI and H&H, together with the specific EREG and HBEGF signals in BEGM-ALI only (see Figure E8A). Contribution of each ligand–receptor pair could be inferred from the expression data and showed high contribution of AREG signaling toward EGFR or EGFR together with ERBB2 in BEGM-ALI (see Figures E8B and E8C). We also looked more closely at differences observed for the NOTCH pathway. Expression of ligands and receptors were modulated by differentiation media (see Figure E9A), which produced large differences in the inferred interactions (see Figures E9B and E9C). We found that suprabasal, deuterosomal, multiciliated, and squamous cells were predicted as receiving high Notch pathway signals in BEGM compared with Pneuma-ALI (see Figure E9C). As AREG has been reported to increase cell proliferation and mediate the upregulation of mucus-related genes in the mouse lung and airway epithelial cell lines (44–47), and as the NOTCH pathway favors the secretory fate over the multiciliated fate (48–51), the differences we report here might explain the greater number of goblet cells and smaller number of MCCs detected in BEGM-ALI.

Hence, the choice of differentiation media can strongly influence signaling pathways between the distinct cell types of the epithelium, which might affect the balance between the different cell types.

Identification of Cell Types at the Onset of Differentiation

As differentiation media influenced epithelial composition and gene expression, we next assessed whether the cell composition at the onset of differentiation was equivalent among all culture conditions.

We first analyzed HBECs from the same donors as previously, immediately after the initial amplification step in Pneuma-Ex+ (see Figure E1B) to identify the different cell identities that were seeded on the culture membranes. Among the 1,716 cells we analyzed, we found a majority of basal and cycling basal cells (see Figure E10 and Table E4). Interestingly, KRT13+ basal cell types were identified and comprised some proliferative cells, on the basis of MKI67 expression. A small fraction (3.50%) of goblet cells was also detected, identified by the expression of SCGB1A1 and MUC5AC. No proliferation was detected among goblet cells, as evidenced by the absence of MKI67 expression (see Figures E10A and E10B).

We then evaluated whether the use of the four distinct proliferation media during the HBEC propagation step on Transwell membranes could influence epithelial cell composition at ALI0, using scRNA-seq. Although basal and cycling basal cells composed the majority of the cultures, we also detected secretory cells (SCGB1A1+ and SLPI+) and suprabasal cells with a secretory signature (LY6D+ and SLPI+) (Figures 3A and 3B). No MUC5AC+ goblet cells were detected at this stage. Few cells that we named “undefined” were detected predominantly in the BEGM condition and did not match with any cell type usually found in vivo. These cells are negative for all basal cell markers and positive for DDIT3, SLC3A2, ISG15, and SQSTM1, among other specifically expressed genes (Figures 3A–3C and Table E5) that seem related to DNA damage, endoplasmic reticulum stress, negative regulation of RNA transcription, and protein ubiquitination. Cell composition was affected by proliferation medium: the BEGM medium generated a smaller fraction of proliferative cells, which is consistent with the lower cell densities observed in this medium, despite reaching confluence. Indeed, contact inhibition appeared higher in BEGM (data not shown). BEGM generated more secretory cells, with lower expression of SCGB1A1 than all other media (Figures 3B and 3C). In the absence of proliferation in these secretory cells, they probably emerged from differentiation during the expansion process. Wu and Pneuma-Ex+ generated similar cell compositions (Figure 3C). We noticed at this stage a ubiquitous expression of the goblet cell–specific transcription factor SPDEF in Pneuma-Ex+ and Wu media (Figure 3B). SPDEF was not detected in either BEGM or Clancy medium in basal, cycling basal, or suprabasal/secretory intermediates but was detected in secretory cells from Clancy medium. To explore further this finding, we set up cultures on four independent batches of HNECs, from four distinct healthy donors, either in BEGM or in Pneuma-Ex+. HNECs were used at this stage as a convenient material for experimental replication and were shown as reliable surrogates to HBECs (10). qPCR analysis confirmed strong upregulation of SPDEF in all four HNEC cultures at the proliferation stage (see Figure E10C). We then repeated this experiment with three additional HNEC cultures either in Pneuma-Ex+ or PneumaCult-Ex, which is equivalent to BEGM for proliferation, and performed scRNA-seq analysis. In this dataset, we could identify cell populations that were similar to those we found in HBECs, and we detected few multiciliated and endothelial cells, which were probably carried over from tissue isolation (Figures 3D and E10D and Table E6). We confirmed that proliferation in Pneuma-Ex+ for one passage was sufficient to induce SPDEF expression (Figure 3E). Surprisingly, we found few proliferative goblet cells at this stage of cell propagation, with a clear enrichment in Pneuma-Ex+ compared with PneumaCult-Ex (Figure 3F).

Figure 3. Impact of distinct proliferation media on cell type composition and cell expression at ALI0 of epithelial cell differentiation. (A) Experimental strategy (left) and UMAPs of the integrated dataset at the onset of differentiation of HBECs containing all cells from the four distinct conditions, colored by cell type and by medium according to the indicated color codes (right). (B) Expression of the top 10 marker genes and of SPDEF for the five detected cell populations from A, displayed for each proliferation medium. (C) Quantification of cell type proportions from A for each proliferation medium. (D) UMAP of the integrated dataset composed of human nasal epithelial cell (HNECs) analyzed after the first expansion stage in either PneumaCult-Ex (Pneuma-Ex) or PneumaCult-Ex Plus (Pneuma-Ex+), colored by cell type. (E) Expression of SPDEF by each cell type of dataset from D in either Pneuma-Ex or Pneuma-Ex+, as indicated by color code displayed in D. (F) Identification of MUC5AC+/MKI67+ cells in the integrated dataset composed of HNECs analyzed after the first expansion stage in either Pneuma-Ex or Pneuma-Ex+. Cells were selected if they had normalized expression of MUC5AC of > 1 and MKI67 of > 0.5. The percentage of MUC5AC+/MKI67+ cells for each sample is indicated on the plot. ALI0 = onset of air–liquid interface; scRNA-seq = single-cell RNA sequencing; Sec. interm. = secretory intermediate.

Altogether, these results indicate that proliferation conditions could affect cell type distribution and gene expression profiles. Thus, we next evaluated whether some differences in ALI differentiation might stem from the initial cell propagation step.

Effect of Proliferation Medium on Features of Differentiated Cultures

HNECs were amplified in either BEGM or Pneuma-Ex+, in plastic flasks, and on Transwell membranes and then induced to differentiate at the ALI in either BEGM-ALI or Pneuma-ALI (see Figure E1C). We evaluated epithelial composition by qPCR and immunostaining using the MCC-specific markers FOXJ1 and acetylated α-tubulin and the goblet cell–specific marker MUC5AC. Both qPCR and immunostaining showed that although Pneuma-ALI yielded more MCCs than BEGM as expected, the proliferation medium had no effect on the MCC content. However, when differentiation was performed in BEGM-ALI, the use of Pneuma-Ex+ for proliferation increased MUC5AC content, as shown by qPCR (Figure 4A), immunostaining (Figure 4B), western blot (Figures 4C, 4D, E11A, and E11B). When differentiation was performed in Pneuma-ALI, although qPCR did not show any difference in MUC5AC expression comparing proliferation in BEGM or Pneuma-Ex+ (Figure 4A), western blot did show a significant increase in MUC5AC content when cells were amplified in Pneuma-Ex+ (Figures 4C, 4D, E11A, and E11B). MUC5B content tended to also be increased by proliferation in Pneuma-Ex+, although not significantly (see Figures E11C–E11E).

Figure 4. Effect of distinct proliferation media on epithelial composition after ALI differentiation of nasal epithelial cultures. (A) Expression of FOXJ1 and MUC5AC (quantitative PCR) on ALI cultures after full differentiation of nasal epithelial cells in the indicated media. Data shown are the values for 2Ct(FOXJ1-TBP) and log10[2Ct(MUC5AC-TBP)] for three independent cultures from three donors. P values are the results of Mann-Whitney U tests. (B) Representative images for immunostaining of MUC5AC and acetylated α-tubulin on HNEC ALI cultures after full differentiation in the indicated media. Scale bars, 30 μm. Nuclei were stained with DAPI and are shown in blue on the merge images. For each condition, three independent cell cultures were performed, and three inserts were stained and imaged for each. (C) Western blot for MUC5AC and FOXJ1 on HNEC ALI cultures after full differentiation in the indicated media. Tubulin was used as loading control. For each condition, two lanes were loaded, each with an independent Transwell membrane from the same culture. (D) Quantification of the MUC5AC signal from C; P values are the results of Welch’s t tests. For each condition, three independent cell cultures were performed, and two inserts were used for western blot. Data from the additional cultures are shown in Figure E11. ns = not significant.

Thus, even though proliferation in Pneuma-Ex+ did not favor goblet or secretory cell content during proliferation (Figures 3 and E10D), this proliferation medium favored goblet cell differentiation or mucin production. This increase might be due to the proliferative goblet cells that we detected during the initial cell propagation of nasal cultures in Pneuma-Ex+ (Figure 3F), although there is no evidence that these cells were retained during the amplification step on Transwell membranes. For HBECs, even though we detected some goblet cells at the initial propagation step (see Figures E10A and E10B), these cells were not retained after amplification on membranes (Figure 3A). Another possibility is that Pneuma-Ex+ induced basal cells’ imprinting toward a subsequent goblet fate, as suggested by increased expression of SPDEF (Figures 3B and 3E).

Effect of Cell Culture Media on SARS-CoV-2 Entry Factors

The ALI model is appropriate for SARS-CoV-2 in vitro infectivity studies, but discrepancies in major entry cell types have been described (17–25). As various cell culture media have been used in past studies, we wondered whether SARS-CoV-2 entry factor expression varied according to differentiation and proliferation media. Although ACE2 was very poorly detected in scRNA-seq data of HBECs at ALI28 (Figure 5A), other entry factors, such as TMPRSS2, FURIN, BSG, CTSB, ANPEP, ST6GAL1, and ST3GAL4, were well detected. TMPRSS2 was well expressed in MCCs from all media but was detected only in BEGM-ALI for secretory and goblet cells and in H&H for squamous cells. Other entry factors showed some medium-specific expression, such as FURIN in BEGM-ALI and Clancy medium; ANPEP, which was restricted to Clancy medium in goblet cells; and ST3GAL4, which was detected in BEGM-ALI and H&H for goblet cells and H&H only in squamous cells (Figure 5A). ACE2 expression was then investigated using qPCR and western blot, for the two proliferation and differentiation commercial media, in HNECs. Although proliferation media did not produce significant effects (Figures 5B, 5C, and E12A), qPCR showed that ACE2 expression was significantly lower in Pneuma-ALI. The western blot analysis showed a decreased expression of the main isoform in Pneuma-ALI compared with BEGM-ALI, which appears as a ∼135-kD band, and an increased expression of a ∼100-kD differentially glycosylated isoform (52) (Figures 5C and E12B).

Figure 5. Effect of proliferation and differentiation media on ACE2 expression at ALI28 of epithelial cell differentiation. (A) Violin plot showing expression of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) entry factors in each of the four media conditions used for HBECs at ALI28 (Figure 1B). (B) qPCR for ACE2 on three independent HNEC ALI cultures from three donors after proliferation and full differentiation in the indicated media. Data shown are the values for 2Ct(ACE2-TBP) for three independent cultures from three donors. P values are the results of Mann-Whitney U tests. (C) Western blot for ACE2 on HNEC ALI cultures after full differentiation in the indicated media. Tubulin was used as loading control. For each condition, three independent cell cultures were performed, and two inserts were used for western blot. Data from two additional independent cultures are shown in Figure E12. qPCR = quantitative polymerase chain reaction.

Thus, both ACE2 expression and glycosylation displayed medium-specific effects.

Effect of Membranes and Media Alternatives

The pandemic crisis and the development of ALI models in respiratory virus studies led to reagent shortages. Alternative media or porous membranes other than Transwell were evaluated. PneumaCult-ALI-S (Pneuma-ALI-S), an additional commercial medium, designed for the differentiation of small airways, and ThinCert membranes (Greiner Bio-One), a commercial membrane, were assessed on HNECs. The morphology of multiciliated and goblet cells and content in reconstructed epithelium were compared between Pneuma-ALI-S and Pneuma-ALI. Figures 6 and E13 show that the use of the Pneuma-Ex+/Pneuma-ALI combination appeared to yield more MUC5AC+ cells than the Pneuma-Ex/Pneuma-ALI combination. In addition, when comparing epithelial layer morphology, the use of Pneuma-Ex+ during the proliferation phase generated thick epithelia, with invaginations within the layer, as shown by actin, hematoxylin–eosin, and nuclei staining (Figures 6 and E13), which is not the case when cells proliferate in PneumaCult-Ex. In contrary, despite the use of Pneuma-Ex+ for proliferation, the use of Pneuma-ALI-S restored the epithelial morphology, as shown by the absence of invaginations in the reconstructed tissue, and a flat epithelial surface. MUC5AC+ cell content appeared similar to that obtained with Pneuma-ALI (Figures 6 and E13). Although further experiments are needed to precisely evaluate the effects of the use of Pneuma-ALI-S on cell composition of ALI cultures, this medium can be used to generate apparent healthy epithelia after HNEC expansion in Pneuma-Ex+. We also evaluated the effect of using alternative polyethylene terephthalate membranes, which are very similar to Transwell filters that carry 0.4-μm pores. The difference between these products is in the pore density, which is 2.106 pores/cm2 for ThinCert membranes and 4.106 pores/cm2 for Transwell membranes. We evaluated these membranes after HNEC proliferation in Pneuma-Ex+ medium, as above, and induced differentiation with either Pneuma-ALI or PromoCell-ALI (Promo-ALI; PromoCell) (whose composition is identical to that of BEGM-ALI). As previously mentioned, thick and invaginated epithelia were obtained when using Transwell membranes with the Pneuma-Ex+/Pneuma-ALI combination (Figure 7). Promo-ALI generated thinner epithelia, but they still displayed invaginations. The basal and suprabasal cell layers, stained with KRT5, seemed expanded in both media. The use of ThinCert membranes strongly reduced the overall thickness of the epithelia and yielded epithelia with the expected morphology together with the expected basal cell, goblet cell, and MCC disposition in Pneuma-ALI. Nonetheless, the use of Promo-ALI on ThinCert membranes produced thin epithelia containing few multiciliated and MUC5AC+ cells (Figure 7).

Figure 6. Effect of an alternative differentiation medium on epithelial composition after ALI differentiation of nasal epithelial cells. (Top) Representative images of immunostaining for MUC5AC and Ac. Tub. on HNEC ALI cultures after full differentiation in the indicated media. Phalloidin was used to stain for actin. Nuclei were stained with DAPI and are shown in blue on the merge images. Scale bars, 30 μm. (Bottom) H&E staining of sections of paraffin-embedded epithelia. Scale bars, 100 μm. For each condition, one cell culture was performed, and two inserts were stained and imaged. Additional images are shown in Figure E13.

Figure 7. Effect of cell culture inserts and differentiation media on epithelial composition after ALI differentiation of nasal epithelial cultures. (Top) Representative images of immunostaining on whole inserts for MUC5AC and Ac. Tub. on HNEC ALI cultures after full differentiation in the indicated media. Scale bars, 30 μm. (Middle) Representative images of immunostaining on sections of inserts for MUC5AC and Ac. Tub. on ALI cultures after full differentiation in the indicated media. Scale bars, 30 μm. (Bottom) Representative images of H&E staining of sections of paraffin-embedded inserts. Scale bars, 100 μm. For each condition, two inserts were stained and imaged.

Together, these results show that if using Pneuma-Ex+ for basal cell expansion, restoration of the expected epithelial morphology and gross composition can be achieved by using ThinCert membranes if differentiating in Pneuma-ALI or by using Pneuma-ALI-S instead of Pneuma-ALI on Transwell membranes.

Discussion

In this study we analyzed, at the single-cell level, the effects of five distinct cell culture media and two types of insert membranes for ALI differentiation of nasal and bronchial epithelial cells. All media allowed the detection of all major cell types present in the airway epithelium. We have shown that two cell culture media, Pneuma-ALI and H&H, generated ALI cultures with high MCC content. H&H contains half PneumaCult-ALI medium, minus the 100× supplement, which has a proprietary composition. We assumed that this supplement includes retinoic acid. The other half of this medium is composed of Wu medium, which itself contains retinoic acid. According to the authors who developed this medium, it was intended to create conditions that are more similar to in vivo conditions (35). However, except for some subtle differences, we found that cell identities and distributions were highly similar between Pneuma-ALI and H&H. On the contrary, the BEGM-ALI and Clancy media yielded many fewer MCCs. Thus, the choice of the culture conditions should be conditioned by the study to be performed. For cilia biology, Pneuma-ALI and H&H should be largely preferred, as the MCC content was much higher in these conditions, with a neat deuterosomal cell cluster. However, if studying secretory cells, BEGM-ALI and Clancy media should be considered, taking into consideration the differences we found in secreted mucin expression. We also noticed the rare occurrence of ionocytes, brush cells, and pulmonary neuroendocrine cells, which were found in too small proportions to allow a comparison among media (8). Future investigations analyzing higher cell numbers should better identify the impact of differentiation media on the amounts and gene expression profiles of these cells.

We used a label transfer method that incorporates reference similarity beyond marker genes alone to compare cells from the ALI model with in vivo airway cells. Given the very wide use of this culture model, such a comparison could be very useful. We have found some cell type and medium-dependent differences, mainly in basal, suprabasal, secretory, and goblet cells, which are cell types showing continuous expression profiles, as opposed to MCCs, which have very distinct expression profiles and the largest number of marker genes. These data highlight the difficulty to annotate cell types on the basis of automatic clustering followed by manual annotation and illustrate that using automatic annotations with label transfer that allow datasets to be compared with a reference in an unbiased manner tools could be a very useful complement.

Leung and colleagues recently published a comparative study of Pneuma-ALI and BEGM-ALI (27). Their findings, in terms of multiciliated, goblet cell content and epithelial thickness, were very similar to ours. Saint-Criq and colleagues also compared Pneuma-ALI to University of North Carolina at Chapel Hill medium, which is quite similar to BEGM. They also obtained thicker epithelia with Pneuma-ALI, and their differential gene expression analysis also showed increased expression of IL33 and TFF3 in Pneuma-ALI and increased expression of ion transport genes in University of North Carolina at Chapel Hill medium, similar to our findings with BEGM and Clancy media. Importantly, they showed that differentiation media can influence CF cell cultures responses to CFTR modulators (29). We add here that the choice of differentiation medium can also modify the expression of SARS-CoV-2 entry factors: ACE2 expression appeared lower in Pneuma-ALI, with a distinct glycosylation profile, and TMPRSS2 was restricted to MCCs in this medium, whereas it showed wider expression in BEGM-ALI. Most studies have used Pneuma-ALI and shown higher SARS-CoV-2 tropism toward MCCs (17–22). However, V’kovski and colleagues (25) and Johansen and colleagues (26) used BEGM and found a higher tropism toward nonmulticiliated cells. Thus, medium selection might be a crucial parameter for viral infection studies.

We extended our investigation to the stages occurring before differentiation. With the exception of a few studies describing conditional reprogramming of epithelial cells (coculture with irradiated fibroblast feeder cells and a Rho kinase inhibitor) (53, 54), or the use of dual SMAD signaling inhibition (55), these stages have been largely ignored in previous works using ALI cultures. Yet proliferation conditions might influence cell fate by modifying the identity and/or potency of basal cells, which are the stem cells generating the entire epithelium upon setting up the ALI. Here, scRNA-seq was instrumental to identify the precise cell composition of epithelial cell cultures during the initial propagation steps. We have shown that after a first stage in plastic flasks, cycling basal and basal cells are highly dominant in both bronchial and nasal cultures, but secretory, goblet, multiciliated, and even endothelial cells were detected, with variations depending on the proliferation medium that was used. For nasal epithelial cells in Pneuma-Ex+ medium, we detected few proliferative goblet cells. These findings should be considered when using this medium, and a more accurate characterization of these cells would be required. Interestingly, we found that gene expression profiles of basal cells were affected, with, for instance, the upregulation of SPDEF by Pneuma-Ex+, which could influence the fate of these cells. In line with these findings, we found that proliferation in Pneuma-Ex+ favored the goblet cell lineage after differentiation.

Finally, we evaluated the effects of using an alternative differentiation medium (Pneuma-ALI-S) and alternative porous membranes (ThinCert) on epithelial morphology after proliferation in Pneuma-Ex+. We found that both restored the normal epithelial morphology which was lost with the use of Pneuma-Ex+ for all proliferation stages, followed by Pneuma-ALI differentiation. As the composition of PneumaCult products is not known, it is not possible to identify the differential cues that produced these effects. The use of alternative porous membranes has also been very informative and suggests that the pore density influences cell morphology. ThinCert membranes having a lower pore density, we speculate that the flow of nutrient might be slower in these conditions and might account for the difference in epithelial morphology.

Conclusions

Our study provides a reference for evaluating the influence of culture conditions on airway epithelial differentiation. Our scRNA-seq data could be further analyzed and could lead to the identification of other pathways regulating differentiation. Beyond the technical usefulness of our results, our study also identifies parameters that should be further explored to study cell fate and epithelial morphology.

Supported by Fondation pour la Recherche Médicale grant DEQ20180339158; Institut National de la Santé et de la Recherche Médicale Inserm Cross-Cutting Scientific Program HuDeCA 201; Chan Zuckerberg Initiative grant 2017-175159-5022; H2020 Health DiscovAIR; Canceropôle PACA Innovation Technologique 2018; Centre National de la Recherche Scientifique; Agence Nationale de la Recherche grant 21-ESRE-0052, ANR-19-P3IA-0002, ANR-19-CE14-0027, ANR-23-IAHU-0007, ANR-10-INBS-09-02, ANR-10-INBS-09-03; and HORIZON EUROPE Marie Sklodowska-Curie Actions grant 101072892. Conseil Départemental 06 016-294DGADSH-CV; Fondation du Souffle.

Author Contributions: E.R. designed and performed experiments and wrote the manuscript. A.C., M.P., M.-J.A., S.R.G., G.P., V.M., and D.G. performed experiments. S.A. performed some image acquisition. M.F. and E.M.M. performed transcriptomics data analysis and wrote the manuscript. C.G.-R. and K.L. performed transcriptomics data analysis. P.B. and P.C. performed data analysis and edited the manuscript. B.M. supervised the study and edited the manuscript. L.-E.Z. supervised the study, performed experiments, and wrote and edited the manuscript.

This article has a data supplement, which is accessible at the Supplements tab.

Originally Published in Press as DOI: 10.1165/rcmb.2023-0356MA on June 6, 2024

Author disclosures are available with the text of this article at www.atsjournals.org.
==== Refs
References

1. Sikkema L Ramírez-Substage C Strobl DC Gillett TE Zappia L Madissoon E et al. Lung Biological Network Consortium An integrated cell atlas of the lung in health and disease Nat Med 2023 29 1563 1577 37291214
2. Surve MV Lin B Reedy JL Crossen AJ Xu A Klein BS et al. Single-cell transcriptomes, lineage, and differentiation of functional airway microfold cells Am J Respir Cell Mol Biol 2023 69 698 701 38038398
3. Hogan B Tata PR Cellular organization and biology of the respiratory system Nat Cell Biol 2019 10.1038/s41556-019-0357-7
4. Deprez M Zaragosi L-E Truchi M Bécavin C Ruiz García S Arguel MJ et al. A single-cell atlas of the human healthy airways Am J Respir Crit Care Med 2020 202 1636 1645 32726565
5. Zepp JA Morrisey EE Cellular crosstalk in the development and regeneration of the respiratory system Nat Rev Mol Cell Biol 2019 20 551 566 31217577
6. Sun X Perl AK Li R Bell SM Sajti E Kalinichenko VV et al. NHLBI LungMAP Consortium A census of the lung: CellCards from LungMAP Dev Cell 2022 57 112 145.e2 34936882
7. Rustam S Hu Y Mahjour SB Rendeiro AF Ravichandran H Urso A et al. A unique cellular organization of human distal airways and its disarray in chronic obstructive pulmonary disease Am J Respir Crit Care Med 2023 207 1171 1182 36796082
8. Cavard A 2020 https://theses.hal.science/tel-03200434/
9. Yamaya M Finkbeiner WE Chun SY Widdicombe JH Differentiated structure and function of cultures from human tracheal epithelium Am J Physiol 1992 262 L713 L724 1616056
10. Lopez-Souza N Avila PC Widdicombe JH Polarized cultures of human airway epithelium from nasal scrapings and bronchial brushings In Vitro Cell Dev Biol Anim 2003 39 266 269 14531727
11. Ruiz García S Deprez M Lebrigand K Cavard A Paquet A Arguel M et al. Novel dynamics of human mucociliary differentiation revealed by single-cell RNA sequencing of nasal epithelial cultures Development 2019 146 dev177428 31558434
12. Greaney AM Adams TS Brickman Raredon MS Gubbins E Schupp JC Engler AJ et al. Platform effects on regeneration by pulmonary basal cells as evaluated by single-cell RNA sequencing Cell Rep 2020 30 4250 4265.e6 32209482
13. Gras D Chanez P Vachier I Petit A Bourdin A Bronchial epithelium as a target for innovative treatments in asthma Pharmacol Ther 2013 140 290 305 23880290
14. Carlini F Picard C Garulli C Piquemal D Roubertoux P Chiaroni J et al. Bronchial epithelial cells from asthmatic patients display less functional HLA-G isoform expression Front Immunol 2017 8 6 28303134
15. Zaragosi LE Deprez M Barbry P Using single-cell RNA sequencing to unravel cell lineage relationships in the respiratory tract Biochem Soc Trans 2020 48 327 336 31922198
16. Durmowicz AG Lim R Rogers H Rosebraugh CJ Chowdhury BA The US Food and Drug Administration’s experience with ivacaftor in cystic fibrosis: establishing efficacy using in vitro data in lieu of a clinical trial Ann Am Thorac Soc 2018 15 1 2 29020455
17. Ravindra NG Alfajaro MM Gasque V Huston NC Wan H Szigeti-Buck K et al. Single-cell longitudinal analysis of SARS-CoV-2 infection in human airway epithelium identifies target cells, alterations in gene expression, and cell state changes PLoS Biol 2021 19 e3001143 33730024
18. Beucher G Blondot ML Celle A Pied N Recordon-Pinson P Esteves P et al. Bronchial epithelia from adults and children: SARS-CoV-2 spread via syncytia formation and type III interferon infectivity restriction Proc Natl Acad Sci U S A 2022 119 e2202370119 35749382
19. Fiege JK Thiede JM Nanda HA Matchett WE Moore PJ Montanari NR et al. Single cell resolution of SARS-CoV-2 tropism, antiviral responses, and susceptibility to therapies in primary human airway epithelium PLoS Pathog 2021 17 e1009292 33507952
20. Hou YJ Okuda K Edwards CE Martinez DR Asakura T Dinnon KH III et al. SARS-CoV-2 reverse genetics reveals a variable infection gradient in the respiratory tract Cell 2020 182 429 446.e14 32526206
21. Zhu N Wang W Liu Z Liang C Wang W Ye F et al. Morphogenesis and cytopathic effect of SARS-CoV-2 infection in human airway epithelial cells Nat Commun 2020 11 3910 32764693
22. Morrison CB Edwards CE Shaffer KM Araba KC Wykoff JA Williams DR et al. SARS-CoV-2 infection of airway cells causes intense viral and cell shedding, two spreading mechanisms affected by IL-13 Proc Natl Acad Sci U S A 2022 119 e2119680119 35353667
23. Purkayastha A Sen C Garcia G Jr Langerman J Shia DW Meneses LK et al. Direct exposure to SARS-CoV-2 and cigarette smoke increases infection severity and alters the stem cell-derived airway repair response Cell Stem Cell 2020 27 869 875.e4 33259798
24. Hao S Ning K Kuz CA Vorhies K Yan Z Qiu J Long-term modeling of SARS-CoV-2 infection of in vitro cultured polarized human airway epithelium mBio 2020 11 e02852-20 33158999
25. V’kovski P Gultom M Kelly JN Steiner S Russeil J Mangeat B et al. Disparate temperature-dependent virus–host dynamics for SARS-CoV-2 and SARS-CoV in the human respiratory epithelium PLoS Biol 2021 19 e3001158 33780434
26. Johansen MD Mahbub RM Idrees S Nguyen DH Miemczyk S Pathinayake P et al. Increased SARS-CoV-2 infection, protease, and inflammatory responses in chronic obstructive pulmonary disease primary bronchial epithelial cells defined with single-cell RNA sequencing Am J Respir Crit Care Med 2022 206 712 729 35549656
27. Leung C Wadsworth SJ Yang SJ Dorscheid DR Structural and functional variations in human bronchial epithelial cells cultured in air-liquid interface using different growth media Am J Physiol Lung Cell Mol Physiol 2020 318 L1063 L1073 32208929
28. Rayner RE Makena P Prasad GL Cormet-Boyaka E Optimization of normal human bronchial epithelial (NHBE) cell 3D cultures for in vitro lung model studies Sci Rep 2019 9 500 30679531
29. Saint-Criq V Delpiano L Casement J Onuora JC Lin J Gray MA Choice of differentiation media significantly impacts cell lineage and response to CFTR modulators in fully differentiated primary cultures of cystic fibrosis human airway epithelial cells Cells 2020 9 2137 32967385
30. Morgan R Manfredi C Easley KF Watkins LD Hunt WR Goudy SL et al. A medium composition containing normal resting glucose that supports differentiation of primary human airway cells Sci Rep 2022 12 1540 35087167
31. Awatade NT Reid AT Nichol KS Budden KF Veerati PC Pathinayake PS et al. Comparison of commercially available differentiation media on cell morphology, function, and anti-viral responses in conditionally reprogrammed human bronchial epithelial cells Sci Rep 2023 13 11200 37433796
32. Luengen AE Kniebs C Buhl EM Cornelissen CG Schmitz-Rode T Jockenhoevel S et al. Choosing the right differentiation medium to develop mucociliary phenotype of primary nasal epithelial cells in vitro Sci Rep 2020 10 6963 32332878
33. Broadbent L Manzoor S Zarcone MC Barabas J Shields MD Saglani S et al. Comparative primary paediatric nasal epithelial cell culture differentiation and RSV-induced cytopathogenesis following culture in two commercial media PLoS One 2020 15 e0228229 32214336
34. Lee DDH Petris A Hynds RE O’Callaghan C Ciliated epithelial cell differentiation at air-liquid interface using commercially available culture media Methods Mol Biol 2020 2109 275 291 31707647
35. Malleske DT Hayes D Jr Lallier SW Hill CL Reynolds SD Regulation of human airway epithelial tissue stem cell differentiation by β-catenin, P300, and CBP Stem Cells 2018 36 1905 1916 30171668
36. Brewington JJ Filbrandt ET LaRosa FJ III Moncivaiz JD Ostmann AJ Strecker LM et al. Brushed nasal epithelial cells are a surrogate for bronchial epithelial CFTR studies JCI Insight 2018 3 e99385 29997283
37. Stoeckius M Zheng S Houck-Loomis B Hao S Yeung BZ Mauck WM III et al. Cell hashing with barcoded antibodies enables multiplexing and doublet detection for single cell genomics Genome Biol 2018 19 224 30567574
38. Phipson B Sim CB Porrello ER Hewitt AW Powell J Oshlack A propeller: testing for differences in cell type proportions in single cell data Bioinformatics 2022 38 4720 4726 36005887
39. Simmons S Cell type composition analysis: comparison of statistical methods 2022 https://www.biorxiv.org/content/10.1101/2022.02.04.479123v1
40. Chan BCL Lam CWK Tam LS Wong CK IL33: roles in allergic inflammation and therapeutic perspectives Front Immunol 2019 10 364 30886621
41. Bothe M Buschow R Meijsing SH Glucocorticoid signaling induces transcriptional memory and universally reversible chromatin changes Life Sci Alliance 2021 4 1 17
42. Osanai M Takasawa A Takasawa K Kyuno D Ono Y Magara K Retinoic acid metabolism in cancer: potential feasibility of retinoic acid metabolism blocking therapy Med Mol Morphol 2023 56 1 10 36592231
43. Zhao Y Zou Z Sun D Li Y Sinha SC Yu L et al. GLIPR2 is a negative regulator of autophagy and the BECN1-ATG14-containing phosphatidylinositol 3-kinase complex Autophagy 2021 17 2891 2904 33222586
44. Manzo ND Foster WM Stripp BR Amphiregulin-dependent mucous cell metaplasia in a model of nonallergic lung injury Am J Respir Cell Mol Biol 2012 47 349 357 22493011
45. Val S Belade E George I Boczkowski J Baeza-Squiban A Fine PM induce airway MUC5AC expression through the autocrine effect of amphiregulin Arch Toxicol 2012 86 1851 1859 22820758
46. Zuo W-L Yang J Gomi K Chao I Crystal RG Shaykhiev R EGF-amphiregulin interplay in airway stem/progenitor cells links the pathogenesis of smoking-induced lesions in the human airway epithelium Stem Cells 2017 35 824 837 27709733
47. Wang J Zhu M Wang L Chen C Song Y Amphiregulin potentiates airway inflammation and mucus hypersecretion induced by urban particulate matter via the EGFR-PI3Kα-AKT/ERK pathway Cell Signal 2019 53 122 131 30291869
48. Mori M Mahoney JE Stupnikov MR Paez-Cortez JR Szymaniak AD Varelas X et al. Notch3-Jagged signaling controls the pool of undifferentiated airway progenitors Development 2015 142 258 267 25564622
49. Rock JR Gao X Xue Y Randell SH Kong Y-Y Hogan BLM Notch-dependent differentiation of adult airway basal stem cells Cell Stem Cell 2011 8 639 648 21624809
50. Kuchibhotla VNS Heijink IH Join or leave the club: Jagged1 and Notch2 dictate the fate of airway epithelial cells Am J Respir Cell Mol Biol 2020 63 4 6 32228394
51. Lafkas D Shelton A Chiu C de Leon Boenig G Chen Y Stawicki SS et al. Therapeutic antibodies reveal Notch control of transdifferentiation in the adult lung Nature 2015 528 127 131 26580007
52. Li W Moore MJ Vasilieva N Sui J Wong SK Berne MA et al. Angiotensin-converting enzyme 2 is a functional receptor for the SARS coronavirus Nature 2003 426 450 454 14647384
53. Reynolds SD Rios C Wesolowska-Andersen A Zhuang Y Pinter M Happoldt C et al. Airway progenitor clone formation is enhanced by Y-27632-dependent changes in the transcriptome Am J Respir Cell Mol Biol 2016 55 323 336 27144410
54. Suprynowicz FA Upadhyay G Krawczyk E Kramer SC Hebert JD Liu X et al. Conditionally reprogrammed cells represent a stem-like state of adult epithelial cells Proc Natl Acad Sci U S A 2012 109 20035 20040 23169653
55. Mou H Vinarsky V Tata PR Brazauskas K Choi SH Crooke AK et al. Dual SMAD signaling inhibition enables long-term expansion of diverse epithelial basal cells Cell Stem Cell 2016 19 217 231 27320041
