key: cord-1020746-l3k9bb5m
authors: Purkayastha, Arunima; Sen, Chandani; Garcia, Gustavo; Langerman, Justin; Shia, David W.; Meneses, Luisa K.; Vijayaraj, Preethi; Durra, Abdo; Koloff, Caroline R.; Freund, Delilah R.; Chi, Justin; Rickabaugh, Tammy M.; Mulay, Apoorva; Konda, Bindu; Sim, Myung S.; Stripp, Barry R.; Plath, Kathrin; Arumugaswami, Vaithilingaraja; Gomperts, Brigitte N.
title: Direct exposure to SARS-CoV-2 and cigarette smoke increases infection severity and alters the stem cell-derived airway repair response
date: 2020-11-17
journal: Cell Stem Cell
DOI: 10.1016/j.stem.2020.11.010
sha: 98cb92147311edc4889e523e768a0854afb34b3e
doc_id: 1020746
cord_uid: l3k9bb5m

Current smoking is associated with increased risk of severe COVID-19 but it is not clear how cigarette smoke (CS) exposure affects SARS-CoV-2 airway cell infection. We directly exposed air-liquid interface (ALI) cultures derived from primary human nonsmoker airway basal stem cells (ABSCs) to short term CS and then infected them with SARS-CoV-2. We found an increase in the number of infected airway cells after CS exposure with a lack of ABSC proliferation. Single cell profiling of the cultures showed that the normal interferon response was reduced after CS exposure with infection. Treatment of CS-exposed ALI cultures with Interferon β-1 abrogated the viral infection, suggesting one potential mechanism for more severe viral infection. Our data show that acute CS exposure allows for more severe airway epithelial disease from SARS-CoV-2 by reducing the innate immune response and ABSC proliferation and has implications for disease spread and severity in people exposed to CS.

Current smoking is associated with increased risk of severe COVID-19 but it is not clear how cigarette smoke (CS) exposure affects SARS-CoV-2 airway cell infection. We directly exposed air-liquid interface (ALI) cultures derived from primary human nonsmoker airway basal stem cells (ABSCs) to short term CS and then infected them with SARS-CoV-2. We found an increase in the number of infected airway cells after CS exposure with a lack of ABSC proliferation. Single cell profiling of the cultures showed that the normal interferon response was reduced after CS exposure with infection. Treatment of CS-exposed ALI cultures with Interferon β-1 abrogated the viral infection, suggesting one potential mechanism for more severe viral infection. Our data show that acute CS exposure allows for more severe airway epithelial disease from SARS-CoV-2 by reducing the innate immune response and ABSC proliferation and has implications for disease spread and severity in people exposed to CS.

Coronavirus disease is an infectious disease caused by the newly discovered severe acute respiratory syndrome-related coronavirus 2 (SARS-CoV-2) and is responsible for the current pandemic that is endangering lives globally. SARS-CoV-2 is an enveloped positive-sense single-stranded RNA virus that enters its host cell by binding to the angiotensin converting enzyme 2 (ACE2) receptor (Ziegler et al., 2020) .

SARS-CoV-2 primarily targets the respiratory tract and ACE2 is expressed in a gradient along the airways and is present on ciliated cells and some secretory cell subtypes (Hou et al., 2020) . Severe COVID-19 lung disease has been most closely associated with older age, especially age over 65 years. Among these older hospitalized adults, underlying medical conditions have been associated with severe COVID-19, which include hypertension, obesity, diabetes mellitus and chronic lung disease (Sanyalou et al., 2020) .

The Surgeon General's Report in 1964 determined that cigarette smoke (CS) is the most important cause of chronic lung disease, including chronic bronchitis and emphysema, in addition to causing lung cancer (NCCDPHP (US) Office on Smoking and Health, 2014) . Mechanistically, CS has been shown to reduce mucosal innate immunity leading to increased viral replication. The underlying mechanisms include degradation of the Type I interferon receptor (HuangFu et al., 2008) , inhibition of type II interferon-dependent gene expression through a decrease in Stat1 phosphorylation (El-Mahdy et al., 2009) and reduction in the immediate-early, inductive, and amplification phases of the type I IFN response that were abrogated with glutathione antioxidant treatment (Bauer et al., 2008) .

Given the importance of CS in the development of chronic lung diseases, it has been suggested that CS may be a significant risk factor for severe COVID-19. The World Health Organization concluded that CS is associated with increased severity of disease and death in hospitalized COVID-19 patients, although they could not quantify the risk to smokers (Igić, 2020) . The lack of clarity on the issue is likely because there has been a lower than expected prevalence of CS reported in retrospective and observational databases because of incomplete reporting of smoking status in patients in emergency situations. Therefore, some J o u r n a l P r e -p r o o f studies reported no increase in CS-related disease whereas more in depth analyses of demographic data showed an increased risk of severe COVID-19 associated with CS (Guo, 2020) (Vardavas and Nikitara, 2020) (Zhao et al., 2020) . In addition, a recent study showed that CS is a risk factor for more severe COVID-19 among young adults (Adams et al., 2020) . Several studies have examined ACE2 expression in the airways of smokers and found that CS increases ACE2 expression , Cai et al., 2020 .

However, there have been no direct studies to examine the effect of CS on the airway epithelium in the setting of SARS-CoV-2 infection and therefore it has remained unclear as to whether CS influences SARS-CoV-2 infection. Therefore, we developed a system to expose primary human mucociliary epithelial cultures at the air-liquid interface (ALI) to CS and subsequently infected the cultures with SARS-CoV-2. CS exposure of the ALI cultures has been shown to mimic the in vivo acute CS exposure seen in patients (Gindele et al., 2020) . We found an increase in the number of infected cells after CS exposure and that acute CS exposure increases airway basal stem cells (ABSCs) while SARS-CoV-2 infection prevents the normal repair response from ABSCs. We also found that SARS-CoV-2 infection upregulates the interferon response, while short-term CS exposure reduces the interferon response, suggesting that the modulation of the interferon response by SARS-CoV-2 is causally linked to more active infection in CS-exposed cultures. Consistent with this hypothesis we found that the CS-induced increase in SARS-CoV-2 infection could be abrogated by treatment with exogenous interferonβ-1.

We used primary human ABSCs from three different healthy lung transplant donors and two other nonsmoker donors from commercial sources for ALI cultures (Gruenert et al., 1996) . The cultures were exposed to or mock exposed to short-term CS for 4 days and then infected or mock infected with SARS-CoV-2 at a MOI of 0.1 ( Figure 1A ). At one, two and three days after infection or mock treatment, we examined the cultures for evidence of SARS-CoV-2 replication. We performed quantitative real time PCR (qRT-PCR) on RNA obtained from the airway epithelial cells at these time points and performed this time course on two biological replicates. We consistently found that the highest intracellular viral genome copies (based on N gene transcripts) was at 48 hours post infection and that there was a 2-3 fold increase in viral load in samples that were first exposed to CS ( Figure 1B ). This is consistent with the fold increase in the number of infected cells seen by immunofluorescence (IF) staining for SARS-Co-V-2 Spike protein at 72 hours post infection ( Figure 1C ). However, at 72 hours post infection, we saw variability in viral genome copies in CS exposed cultures ( Figure 1B ) and this is consistent with the cells with the highest viral load becoming apoptotic and being extruded from the cultures and may also reflect the complexity of the genetic backgrounds of the different patients. Infectious virions released to the basal chamber culture media were below detectable levels at these time points. We found an increased number of infected cells in ALI cultures that had CS exposure, which was consistent across 5 patient samples and 25 technical replicates from these 5 patients (p<0.0001) ( Figure 1C ).

Next, we quantified the number of ciliated and secretory airway epithelial cell types across all conditions by IF for cell-type specific markers ( Figure 1D -G). We found that CS exposure alone showed a trend towards a decrease in the number of ciliated cells (p=0.06) marked by the presence of acetylated β-tubulin and significantly increased the number of Muc5AC-expressing mucus cells (p<0.001)(Figures 1D-G). We found that the number of ciliated or mucus cells did not significantly change with SARS-CoV-2 infection ( Figures   1D-G) . We then assessed the number of ABSCs, the key stem cell type orchestrating the repair response by proliferation, by IF for keratin 5 (KRT5). We found that CS significantly increased the number of ABSCs as part of a repair response (p<0.01) and that SARS-CoV-2 viral infection alone or viral infection with CS did not trigger the expected increase in ABSCs needed for repair ( Figure 1H ,I). The repair response is mediated by proliferation of ABSCs as well as transient amplifying cells in the ALI cultures. We therefore assessed all proliferating cells in the ALI cultures by immunostaining for Ki67 combined with PCNA. We found that J o u r n a l P r e -p r o o f proliferation was induced after CS exposure but that SARS-CoV-2 infection did not further alter the proliferation under the CS and control conditions, implying that SARS-CoV-2 inhibits the repair process in the airway epithelium ( Figure 1J ,K). IF for cleaved caspase 3 (CC3), across two biological replicates and six technical replicates at 3 days post infection, revealed that apoptosis was infrequently seen in the ALI cultures with CS exposure alone, but viral infection significantly increased the number of apoptotic cells and the combination of CS and infection further increased the number of apoptotic cells ( Figure 1L ,M). Based on these data, SARS-CoV-2 infection is promoting cell death while reducing the normal airway epithelial repair response. As ACE2 is the receptor for SARS-CoV-2, we examined ACE2 expression by IF and found that there was a trend towards increased ACE2 expression after CS exposure and that there was no change in ACE2 expression in infected cells or infected cells exposed to CS (Lee et al., 2020)(Supplemental Figure   S1A ,B). QRT-PCR revealed no change in ACE2 gene expression across all exposure groups and across our infection time course (Supplemental Figure S1C ). Overall, across five different patients, we found a significant increase in the number of SARS-CoV-2 infected cells after CS exposure and differences in cellular responses to these airway exposures.

We then sought to determine a possible mechanism for the more active cellular infection seen upon exposure of the ALI cultures to both CS and SARS-CoV-2. We applied single cell RNA-sequencing to determine the transcriptional alterations taking place which might explain the differences in SARS-CoV-2 infectivity between CS exposure and no exposure. We used the same experimental timeline as outlined in the schematic in Figure 1A before performing single cell dissociation for RNA-seq. We recovered 19361 single cell transcriptomes that passed filtration criteria. Plotting of the data showed that mock-treated cells and cells exposed to CS were largely mixed and that cells with SARS-CoV-2 infection were separated Figure 2B ), expressing typical human airway cell type marker genes (DNAAF3, ciliated cells; KRT5, basal cells; and SCGB1A1 and MUC5B, secretory cells; Figure 2C and Supplemental Figure   S2C ). We detected SARS-CoV-2 transcripts in all major cell types and found the proportion of infected cells to be highest in FOXN4+ cells (Supplementary Figure S2A) , however very few FOXN4+ cells were detected ( Figure 2A ).

We observed a global downregulation of genes in SARS-CoV-2 exposed samples, with approximately 10-15% fewer transcripts in infected cells compared to uninfected controls ( Figure 2D ). Both CS-exposed and unexposed samples yielded this pattern, suggesting that it is not a unique handling of the sample. Future studies are required to explore the mechanisms leading to this observation. Using differential expression to determine genes specific to each condition, we detected the downregulation of 2805 genes in SARS-CoV-2 exposed samples, which was consistent across all major cell types ( Figure 2E , Supplementary Figure S2B ). Among the downregulated genes were genes related to the viral immune response and various metabolic processes. 475 genes were upregulated upon exposure to the virus, including genes related to interferon signaling and chromatin organization. Although the majority of gene expression changes were controlled by virus exposure, we found another 559 genes that were altered in response to CS exposure. Even prior to virus exposure, CS downregulated the innate immune response and stimulated airway differentiation genes.

In response to SARS-CoV-2, CS-treated cells downregulated genes related to metabolic and wound healing processes and upregulated cilia-related genes. Interestingly, a small class of genes was induced in the non-CS-exposed cells in response to SARS-CoV-2 but downregulated in the CS-exposed infected ALI cultures. Table S3 ). Together these data suggest that CS exposure prevents an effective interferonbased response to the SARS-CoV-2 virus.

Because the interferon response was increased in viral infection alone but decreased in the setting of CS with viral infection, we reasoned that the greater infection seen upon treatment with CS and SARS-CoV-2 was, at least in part, due to CS-induced reduction in the innate immune response. We therefore treated the CS-exposed or unexposed and virally infected or uninfected ALI cultures with interferonβ-1 and included remdesivir (a direct acting antiviral agent) or no drug, as controls. These were added to the ALI cultures after CS exposure and just prior to viral infection and were present in the cultures for three days after which the cultures were immunostained for the apoptotic marker CC3 and Spike antigen. We found that interferonβ-1 completely abrogated the infection ( Figure 2G,H) . Remdesivir also showed an inhibitory response to viral infection and these effects were most pronounced in the ALI cultures that received CS-exposure prior to infection ( Figure 2H ). This shows that a lack of interferon response is at least one important mechanism for the higher SARS-CoV-2 infectivity seen in CS exposed ALI cultures.

There has been some controversy about whether CS exposure increases SARS-CoV-2 infectivity and whether this might lead to more severe disease. Our primary human mucociliary epithelial cell culture model with direct exposure to CS and SARS-CoV-2 infection demonstrates that short-term exposure to CS in primary human airway cells from previously healthy patients who were not chronic smokers leads to increased infection. Our data suggest two potential and synergistic mechanisms for this: reduced innate immunity of the cells leading to more active infection, lack of ABSCs to appropriately proliferate for airway repair after infection, which could lead to worse tissue damage and/or stem cell exhaustion and increased apoptosis in airway cells exposed to CS and SARS-CoV-2.

The reduction in the interferon response by smoking is one mechanism whereby SARS-CoV-2 may more easily enter and replicate in epithelial cells. Interestingly, interferon response genes were actually induced by the virus, which may explain why we only see a low percentage of infected cells in the ALI cultures and why the majority of the human population does not develop severe respiratory infections. Our data suggest that J o u r n a l P r e -p r o o f there are additional factors from CS exposure that could make the cells more vulnerable to infection. For example, the synergistic increase in apoptotic cells after CS exposure and infection suggests that the CS may act with the virus to potentiate other pathways, such as the DNA damage response pathway, to injure the epithelium. We also saw a reduction in cell division in the ABSCs which prevented the normal host response with repair and regeneration of the airway. This result was in contrast to smoking injury where ABSCs proliferate robustly to repair the airway epithelium. Despite smoking injury, the SARS-CoV-2 response was dominant and prevented the normal ABSC proliferation response.

The interferon response is reported to be reduced or delayed in serum levels in ferrets and patients with severe COVID-19 where pro-inflammatory cytokines and chemokines are more dominant (Blanco-Melo et al., 2020) . Others have found a temporal expression of interferon genes after SARS-CoV-2 infection (Yoshikawa et al., 2010) . We noted an increase in interferon response genes in ALI cultures at 3 days post infection with low MOI (0.1). This suggests that an intact, normal airway epithelium could function as a barrier to COVID-19 with an innate immune response and could explain why the majority of infected individuals have mild symptoms and even asymptomatic carriage. Abrogation of SARS-CoV-2 infection with type I interferons has been reported by others in other settings (Lokugamage et al., 2020; Mantlo et al., 2020) , including a small clinical trial .

The human primary ALI cell culture model provides a useful system for studying the direct effects of CS on the airway epithelium. Our experiments directly examined the effects of short-term CS exposure on the airway, which implies that current smokers are at risk of more severe infection. It is not clear whether former smokers will have the same risk of infection that current smokers do, and this is something that remains to be tested. Our acute CS exposure of ALI cultures did not show increased ACE2 expression, as has been previously demonstrated, but likely reflects the short time frame of exposure and the low expression of ACE2 in the distal trachea. The increased number of infected cells in smokers has implications for more severe infection in smokers resulting in increased lung disease although our cultures are of the proximal airways so J o u r n a l P r e -p r o o f we could not directly assess for effects on diffuse alveolar damage or ARDS. Overall our data provides evidence for the need for health measures to stop smoking to reduce severe COVID-19.

There are several factors that make these studies challenging. The first is that there is a wide variation in the differentiation capacity of the ALI cultures from different donors, together with variation in the number of SARS-CoV-2 infected cells in each set of cultures. Therefore, it can be challenging to assess effects across heterogeneous patient samples. In addition, the sample size is limited by the challenges and costs of obtaining human tissue. This limits the number of patient samples and number of replicates that can be performed from each patient. There are also many changes that occur in the cell after infection and the timing of these is critical. We found that time courses were very helpful in determining the point of greatest infection and that this time point was quite consistent across cultures. However, this point of maximal viral load didn't necessarily correspond with the biggest changes in cell types or apoptosis. One other limitation of our model is the lack of inflammatory cells and this is something that we are currently incorporating into our models.

We would like to thank Andrew Lund and WooSuk Choi for their input on the manuscript. This work was 

The authors declare no competing interests. Graph represents mean, n = 2 biological replicates (2 different patients), each with 2 technical replicates (2 different ALI transwell cultures derived from each of the 2 patients for all three treatment groups and four environmental exposure conditions (48 transwells)). P-values are calculated from all technical replicates across each of the biological replicates described above. ns = not significant by Student's t test. Scale bar = 50µm.

biological replicate was used for two experiments. The biological replicates were from ABSCs isolated from lung transplant donors or NHBE samples from Lonza. No demographic data was available for the normal lung donor samples. Lonza samples were obtained from lungs from donors ranging from 30-50 years and represented both males and females. No significant difference was seen between the ALI cultures from cells from these two sources.

Human ABSCs were isolated following a previously published method by our laboratory (Hegab et al., 2012b (Hegab et al., , 2012a (Hegab et al., , 2014 Paul et al., 2014) . Briefly, airways were dissected, cleaned, and incubated in 16U/mL dispase for 30 minutes at room temperature. Tissues were then incubated in 0.5mg/mL DNase for another 30 minutes at room temperature. Epithelium was stripped and incubated in 0.1% Trypsin-EDTA for 30 minutes shaking at 37°C to generate a single cell suspension. Isolated cells were passed through a 40µm strainer and plated for Air-Liquid Interface cultures.

24-well 6.5mm transwells with 0.4µm pore polyester membrane inserts were coated with 0.2mg/mL collagen type I dissolved in 60% ethanol and allowed to air dry. ABSCs were seeded at 100,000 cells per well directly onto collagen-coated transwells and allowed to grow in the submerged phase of culture for 4-5 days with 500µl media in the basal chamber and 200µl media in the apical chamber. ALI cultures were then established and cultured with only 500µl media in the basal chamber, and cultures were harvested at varying timepoints for IF studies. Media was changed every other day and cultures were maintained at 37°C and 5% CO 2 .

Human ABSCs/NHBEs were grown in TEC Plus media and TEC serum-free media during the submerged and ALI phases of culture, respectively. TEC base media is DMEM/Ham's F12 50/50 (Corning 15090CV). Table S1 indicates the media components and concentrations for TEC Plus and TEC serum-free media.

A sterile chamber of 0.4 cu. ft. volume, with an attached vacuum pump, was used to generate and deliver CS. The ALI plates (without lids) were placed inside the chamber and exposed to the CS of a 1R3F research cigarette (University of Kentucky, Lexington, KY), which was attached to the vacuum chamber tube with a cigarette holder. The cigarette was burned 10% of its length and the CS was introduced into the chamber via suction pump. A previously optimized treatment of 3-minute exposure/day was continued for 4 days.

SARS-CoV-2, Isolate USA-WA1/2020, was obtained from Biodefense and Emerging Infectious (BEI)

were conducted in the UCLA BSL3 high-containment facility with appropriate institutional biosafety approvals. SARS-CoV-2 was passaged once in Vero-E6 cells and viral stocks were aliquoted and stored at -80 o C. Virus titer was measured in Vero-E6 cells by TCID 50 assay.

ALI cultures on the apical chamber of transwell inserts were infected with SARS-CoV-2 viral inoculum (MOI of 0.1; 100 µl/well) prepared in ALI TEC media. The basal chamber of the transwell contained 500 µl of ALI media. For mock infection, ALI media (100 µl/well) alone was added. The inoculated plates were incubated for 2 hr at 37 °C with 5% CO 2 . At the end of incubation, the inoculum was removed from the apical chamber.

At selected timepoints live cell images were obtained by bright field microscopy. Striking cytopathic effect 

Interferonβ-1 Drug Study.

Once the ALI-SARS-CoV-2 infection system was established, we evaluated the effect of interferonβ-1 (200ng/ml) and remdesivir (10 µM), as a control. The ALI cultures in 24-well plates were pretreated with Interferonβ-1 for 1 hour, then SARS-CoV-2 inoculum (MOI 0.1) was added. DMSO vehicle treated cells, with or without viral infections, and with and without smoking exposure were included as controls. At 72 hpi, the cells were fixed and immunostained with Polyclonal Anti-SARS-CoV to assess viral genome replication ( Figure 1C ).

ALI cultures were fixed in 4% paraformaldehyde for 15 minutes followed by permeabilization with 0.5% Triton-X for 10 minutes. Cells were then blocked using serum-free protein block (Dako X090930) for one hour at room temperature and overnight for primary antibody incubation. After several washes of Tris-Buffered Saline and Tween-20 (TBST), secondary antibodies were incubated on samples for 1 hour in darkness, washed, and mounted using Vectashield hardest mounting medium with DAPI (Vector Labs H-1500). IF images were obtained using an LSM700 or LSM880 Zeiss confocal microscope and composite images generated using ImageJ. The list of antibodies used is provided in the Key Resources Table as part of STAR Methods. Cells were counted manually in a blinded fashion. Approximately equal numbers of cells (around 1,000 cells) were counted for each experimental group. All immunofluorescence images used for scoring cells consisted of a z-series of optical sections captured on the Zeiss LSM 700 or 880 confocal microscopes.

RNA was isolated from the ALI cells with the RNeasy Mini Kit (Qiagen 74104) following manufacturer's protocol and quantified using a NanoDrop Spectrophotometer (ThermoFisher). cDNA synthesis was performed using the TaqMan Reverse Transcription Reagents (ThermoFisher) or iScript cDNA Synthesis Kit (BioRad)s as indicated by the respective company. qRT-PCR for viral load was then performed with Cyber green using the primers in Supplemental Table S2 . Samples were run in triplicate and fold changes in expression were determined using the comparative ∆C T method and GAPDH was used as an endogenous control. qRT-PCR for ACE2 was run with Taqman primer probes set Hs01085333_m1

The ALI cultures were infected with SARS-CoV-2 at MOI of 0.1, in transwell plates. Mock infected cells received only the media used for preparing the SARS-CoV-2 inoculum. After 1-hour incubation at 37°C with 5% CO 2 , the inocula were removed. At each timepoint (days 1, 2 and 3), media from the basal chamber from mock and SARS-CoV-2 infected wells were collected and stored at -80°C. Viral production by infected ALI at each timepoint was measured by quantifying TCID50 (Median Tissue Culture Infectious Dose) as described (Gauger et al., 2014) . In brief, Vero-E6 cells were plated in 96-well plates at a density of 5 x10 3 cells/well. The next day, culture media samples collected from ALI at various timepoints were subjected to 10-fold serial dilutions (10 -1 to 10 -6 ) and inoculated onto Vero-E6 cells. The cells were incubated at 37°C with 5% CO 2 .

After 72 hours, each inoculated well was examined for presence or absence of viral CPE and percent infected dilutions immediately above and immediately below 50% were determined. TCID50 was calculated based on the method of Reed and Muench.

Single cells were obtained by incubating ALI cultures in 500ul Accumax for an hour and 15 minutes. Cells were then fixed in cold methanol per the Illumina protocol and frozen at -80ºC. Cells were then rehydrated in ice cold PBS and RNAse inhibitor per the Illumina protocol. Cells were captured using a 10X Chromium device (10X Genomics) and libraries prepared according to the Single Cell 3' v2 or v3 Reagent Kits User Guide (10X Genomics, https://www.10xgenomics.com/products/single-cell/). Cellular suspensions were loaded on a Chromium Controller instrument (10X Genomics) to generate single-cell Gel Bead-In-EMulsions (GEMs). Reverse transcription (RT) was performed in a Veriti 96-well thermal cycler (ThermoFisher). After RT, GEMs were harvested, and the cDNA underwent size selection with SPRIselect Reagent Kit (Beckman J o u r n a l P r e -p r o o f Coulter). Indexed sequencing libraries were constructed using the Chromium Single-Cell 3' Library Kit (10X Genomics) for enzymatic fragmentation, end-repair, A-tailing, adapter ligation, ligation cleanup, sample index PCR, and PCR cleanup. Libraries QC was performed by the Agilent Technologies Bioanalyzer 2100 using the High Sensitivity DNA kit (Agilent Technologies, catalog# 5067-4626) and quantitated using the Universal Library Quantification Kit (Kapa Biosystems, catalog# KK4824. Sequencing libraries were loaded on a NovaSeq 3000 (Illumina).

Raw sequencing data were filtered by read quality, adapter-and polyA-trimmed, and reads were aligned to a hybrid human hg38-SARS Covid2 transcriptome using the Cell Range software (10X Genomics) and the STAR aligner. Expression counts for each gene were collapsed and normalized to unique molecular identifiers to construct a cell by gene matrix for each library, filtered to keep cells with over 2000 transcripts and genes expressed in at least 0.05% of cells.

Data analysis was performed in R. Expression matrices were normalized by the total number of transcripts per cell in log space by dividing raw counts by the total number of transcripts per cell, then multiplying by 10,000. Two-dimensional visualization was obtained with the UMAP package. To identify major cell types in our normal integrated datasets, previously published lung epithelial cell type specific gene lists (Plasschaert et al., 2018) were used to create cell type-specific gene signatures, with cells assigned by maximal identity score. To identify differentially expressed genes between samples, we selected genes from every given pair of condition comparisons which satisfied an average expression difference of 33% either up or down in log normalized counts, filtered by a Bonferroni corrected p<0.01. Gene ontology enrichments were determined using the Metascape tool (Zhou et al., 2019) . For heatmap generation, the mean expression per condition was calculated and then per gene normalized to the maximal sample value.

J o u r n a l P r e -p r o o f

The percentage of each cell population in each field was determined by dividing marker positive cells by the total number of nuclei. 2-way ANOVA was used to determine statistically significant effects of either smoking or virus on the cell populations. Unpaired t-test was used to determine statistical significance of viral infection between no CS and CS groups. Kruskal-Wallis non-parametric ANOVA was used to compare nuclei counts across different drug treatments. Significance was defined as p<0.05. Statistical details of experiments can be found in the Results section and Figure legends. 

Cigarette smoke suppresses type I interferon-mediated antiviral immunity in lung fibroblast and epithelial cells

Imbalanced Host Response to SARS-CoV-2 Drives Development of COVID-19

Tobacco Smoking Increases the Lung Gene Expression of ACE2, the Receptor of SARS-CoV-2

Inhibition of Antiviral Defense in Airway Epithelial Cells by Cigarette Smoke

Serum virus neutralization assay for detection and quantitation of serum-neutralizing antibodies to influenza A virus in swine

Intermittent exposure to whole cigarette smoke alters the differentiation of primary small airway epithelial cells in the air-liquid interface culture

Culture and Transformation of Human Epithelial Cells

Smoking links to the severity of Covid-19: An update of a meta-analysis

Isolation of basal cells and submucosal gland duct cells from mouse trachea

Isolation and In Vitro Characterization of Basal and Submucosal Gland Duct Stem/Progenitor Cells from Human Proximal Airways

Aldehyde dehydrogenase activity enriches for proximal airway basal stem cells and promotes their proliferation

Cigarette smoking products suppress anti-viral effects of Type I interferon via phosphorylation-dependent downregulation of its receptor

Smoking and COVID-19

ACE2 localizes to the respiratory cilia and is not increased by ACE inhibitors or ARBs

Asymptomatic carrier state, acute respiratory disease, and pneumonia due to severe acute respiratory J o u r n a l P r e -p r o o f syndrome coronavirus 2 (SARS-CoV-2): Facts and myths

Type I Interferon susceptibility distinguishes SARS-CoV-2 from SARS-CoV

Antiviral activities of type I interferons to SARS-CoV-2 infection

Dynamic changes in intracellular ROS levels regulate airway basal stem cell homeostasis through Nrf2-dependent notch signaling

A single-cell atlas of the airway epithelium reveals the CFTR-rich pulmonary ionocyte

Comorbidity and its Impact on Patients with COVID-19

COVID-19 and smoking: A systematic review of the evidence

Clinical and Translational Report Retrospective Multicenter Cohort Study Shows Early Interferon Therapy Is Associated with Favorable Clinical Responses in COVID-19 Patients Clinical and Translational Report Retrospective Multicenter Cohort Study Shows Ear

SoupX removes ambient RNA contamination from droplet based single cell RNA sequencing data

Expression of the SARS-CoV-2 ACE2 Receptor in the Human Airway Epithelium

The impact of COPD and smoking history on the severity of Covid-19: A systemic review and meta-analysis

Metascape provides a biologist-oriented resource for the analysis of systems-level datasets

SARS-CoV-2 Receptor ACE2 Is an Interferon-Stimulated Gene in Human Airway Epithelial Cells and Is Detected in Specific Cell Subsets across Tissues

J o u r n a l P r e -p r o o f Highlights:• Direct cigarette smoke exposure increases the number of SARS-CoV-2 infected cells • SARS-CoV-2 infection inhibits the airway basal stem cell repair response • Cigarette smoke reduces innate immune responses with worse SARS-CoV-2 infection eTOC blurb: Purkayastha and colleagues modelled the direct effects of cigarette smoke on SARS-CoV-2 infection of the airway epithelium. Acute cigarette smoke exposure increased the number of infected and apoptotic cells, prevented the normal airway basal stem cell repair response, and blunted innate immune responses. Improving innate immunity impaired SARS-CoV-2 infection.