Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Sep 1.
Published in final edited form as: Mol Cancer Res. 2024 Mar 1;22(3):240–253. doi: 10.1158/1541-7786.MCR-23-0692

Single-cell transcriptomics reveals pre-existing COVID-19 vulnerability factors in lung cancer patients

Wendao Liu 1,2, Wenbo Li 1,3, Zhongming Zhao 1,2,4,*
PMCID: PMC10922768  NIHMSID: NIHMS1952501  PMID: 38063850

Abstract

COVID-19 and cancer are major health threats, and individuals may develop both simultaneously. Recent studies have indicated that cancer patients are particularly vulnerable to COVID-19, but the molecular mechanisms underlying the associations remain poorly understood. To address this knowledge gap, we collected single-cell RNA sequencing data from COVID-19, lung adenocarcinoma, small cell lung carcinoma patients and normal lungs to perform an integrated analysis. We characterized altered cell populations, gene expression, and dysregulated intercellular communication in diseases. Our analysis identified pathological conditions shared by COVID-19 and lung cancer, including upregulated TMPRSS2 expression in epithelial cells, stronger inflammatory responses mediated by macrophages, increased T cell response suppression, and elevated fibrosis risk by pathological fibroblasts. These pre-existing conditions in lung cancer patients may lead to more severe inflammation, fibrosis, and weakened adaptive immune response upon COVID-19 infection. Our findings revealed potential molecular mechanisms driving an increased COVID-19 risk in lung cancer patients and suggested preventive and therapeutic targets for COVID-19 in this population.

Introduction

SARS-CoV-2 has caused a global pandemic since its outbreak in 2019, leading to millions of patients with severe symptoms or death. In 2022, this new Coronavirus disease 2019 (COVID-19), joined heart disease and cancer as the top three diseases leading to death in the United States, comprising 5.7%, 21.4%, 18.6% of all 3,273,705 deaths respectively. (1) Previous studies have demonstrated that cancer patients are more vulnerable to COVID-19 when compared to healthy individuals (2–4), even if they are fully vaccinated against SARS-CoV-2 (5–7). This heightened susceptibility implies influential factors other than adaptive immune responses, and highlights the urgent need to understand the underlying molecular mechanisms driving the association between COVID-19 and cancer.

While previous studies indicated the connection between COVID-19 and cancer, they have primarily relied on clinical symptoms and medical records to draw the conclusions (4,6,8–10). While these studies have provided valuable insights, they have not fully elucidated the intricate molecular processes at play. Recent review articles have highlighted some shared pathways between COVID-19 and cancer like hypoxia and hypercoagulability (11,12). but a more comprehensive investigation is necessary to gain a deeper understanding of their interplay at the molecular and cellular levels.

In recent years, single-cell RNA sequencing (scRNA-seq) has emerged as a powerful tool in studying complex diseases, including COVID-19 and various types of cancers (13–17). It enables researchers to examine gene expression patterns at the level of individual cells, providing unprecedented insights into the heterogeneity and dynamics of cellular populations. The utilization of scRNA-seq in this context holds great promise for uncovering previously unknown cellular interactions and trajectories, gene expression changes, and signaling pathways that contribute to an elevated COVID-19 risk observed in cancer patients.

Cancer is tissue specific (18). Lung is the primary infection tissue for SARS-CoV-2 and lung cancer patients had significantly higher rates of infection, severe conditions and death (2). This motivated us to investigate and compare the molecular characteristics of lung cancer and COVID-19. Here, we collected lung and bronchoalveolar lavage fluid (BALF) scRNA-seq datasets from COVID-19 patients (13,17), lung tissue scRNA-seq datasets from lung adenocarcinoma (LUAD) and small cell lung carcinoma (SCLC) patients (14,15), as well as a lung tissue scRNA-seq dataset from healthy controls (19). Through a comprehensive analysis of the molecular landscape from COVID-19 patients and lung cancer patients, we aim to delve deeper into the potential molecular mechanisms driving the increased vulnerability of lung cancer patients to COVID-19, which help identify key players and potential targets for intervention.

Material and Methods

Data collection

The scRNA-seq datasets used in this study were all collected from public databases. The normal control samples (n=40) from the Human Lung Cell Atlas (HLCA) were obtained from cellxgene (https://cellxgene.cziscience.com/collections/6f6d381a-7701-4781-935c-db10d30de293, RRID:SCR_021059). For comparison analysis with diseased conditions, we used 134,550 cells from lung parenchymal tissue. The scRNA-seq data from COVID-19 BALF (n=21) and lung (n=20) samples were obtained from GEO database (RRID:SCR_005012) with accession numbers GSE157344 and GSE171524, respectively. Seven control samples in GSE171524 dataset were not included in the analysis, because we used HLCA samples as healthy controls. The LUAD (n=17), SCLC (n=9), and tumor adjacent (n=4) samples were obtained from cellxgene (https://cellxgene.cziscience.com/collections/62e8f058-9c37-48bc-9200-e767f318a8ec). Tumor adjacent samples from LUAD patients were annotated as ‘normal’ in the original dataset, but we reannotated them into ‘tumor adjacent’ samples in our study. The clinical characteristics of included human subjects are listed in Supplementary Table S1.

Data integration and annotation

We used scANVI and scArches from scvi-tools (v0.19.0) for integration and annotation of scRNA-seq datasets (19–21). Each COVID-19 and lung cancer dataset was individually mapped onto the HLCA reference to obtain low dimensional embeddings and cell type annotation prediction of cells. The low dimensional embeddings of all datasets were then merged to create a unified visualization with UMAP using Seurat (v4.1.1, RRID:SCR_016341) (22). For cell type annotations with a five-level classification system, we primarily focused on levels 3 and 4 for downstream analysis. To compare the proportion of different cell populations in a diseased condition with that in normal samples, we calculated the proportion of each cell population within all cells from each sample. The statistical significance was determined using Wilcoxon rank-sum test between samples from two conditions. P values were adjusted using Benjamini-Hochberg procedure (23).

Subpopulation analysis

We performed subpopulation analysis in four large cell populations, macrophages, T cells, epithelial cells, and fibroblasts. For macrophages, we extracted single cells with annotations “Alveolar macrophages” and “Interstitial macrophages”. For T cells, we extracted single cells with annotations “T cell lineage”. For fibroblasts, we extracted single cells with annotations “Fibroblasts”. For epithelial cells, we extracted single cells with annotations including: “AT1”, “AT2”, “AT2 proliferating”, “Suprabasal”, “Basal resting”, “Multiciliated”, “Deuterosomal”, “Club”, “Goblet”, “Transitional Club-AT2”, “Ionocyte”, “Tuft”, “Neuroendocrine”.

When analyzing each cell population, we found that the low-dimensional embeddings from scANVI displayed batch effect within each cell population, suggesting that they were not appropriate for clustering analysis. Therefore we used another method, Harmony (v0.1.1, RRID:SCR_022206) (24), for integrating datasets in this step. We used the default settings in Harmony and considered each sample as a batch for data integration. The low-dimensional embeddings from Harmony was used for constructing KNN graph, clustering with Louvain algorithm, and visualization with uniform manifold approximation and projection (UMAP) using Seurat (22). To compare the proportion of different cell subpopulations between other conditions with normal, we calculated the proportion of each cell subpopulation in each cell population from each sample. The statistical significance is calculated using Wilcoxon rank-sum test between two conditions.

Differential gene expression analysis

For each cell type annotation in levels 3 and 4, we performed differential gene expression analysis between each diseased condition and normal samples. Differential gene expression analysis was performed using “FindMarkers” function from Seurat with Wilcoxon rank-sum test. Identifying marker genes of cell subpopulations was performed by differential gene expression analysis between each subpopulation with all other cells using “FindAllMarkers” function in Seurat. P values were adjusted using Benjamini-Hochberg procedure (23).

Pathway enrichment analysis

We used clusterProfiler (v4.2.2, RRID:SCR_016884) and Gene Ontology (GO, RRID:SCR_002811) biological processes terms for over-representation analysis to find enriched pathways (25). To predict potential functions of macrophage and fibroblast subpopulations, we used marker genes with average log2 fold change (log2FC)>1 of each subpopulation as input. Enriched terms were hierarchically clustered according to semantic similarity. To predict upregulated pathways in all fibroblasts, we used differentially expressed genes with average log2FC>1 between other conditions with normal as input.

Intercellular communication analysis

We used Cellchat (v1.6.1, RRID:SCR_021946) and its built-in database of ligand-receptor pairs for intercellular communication analysis (26). Cellchat measures the intercellular communication probability between each two cell populations using ensemble average gene expression. We followed the official analysis pipeline using default parameters for our study. To predict the communication probability of specific ligand-receptor pairs, such as IL1B-IL1R2 and SPP1-CD44, we extracted the results using the “extractEnrichedLR” function and visualized them using “netVisual_individual”.

To assess the overall signaling strength of different signaling pathways in each condition, we extracted and summed the communication probability for each cell population from the data slot “netP”. Subsequently, we calculated the logarithm of the summed communication probability and scaled the values across conditions to obtain the overall signaling strength. For visualization purposes, we focused on the pathways that were upregulated in diseased conditions. Cell populations with scaled communication probability greater than 0.3 were considered major contributors to each signaling pathway. These populations were then used to construct graphs illustrating the signaling pathways associated with specific cell populations.

Transcription factor analysis

We used pySCENIC (v0.12.1, RRID:SCR_017247) to infer potential transcription factors in pathological fibroblasts (27). Raw count matrix of fibroblasts was used as input. pySCENIC inferred transcription factors and their target genes from co-expression network and then refined results based on enriched motifs. The activity of a regulon, which consists of a transcription factor and its target genes, was scored using AUCell. We identified differentially activated regulons using “FindAllMarkers” function from Seurat. The top two transcription factors with average log2FC>0.1 in the pathological fibroblast population were selected for visualization.

Statistical analysis

All statistical analyses were performed using R software (v4.1.2) packages. The P values of proportional difference of cell populations or subpopulations between samples from two conditions were calculated using Wilcoxon rank-sum test. Differential gene expression analysis between two conditions was performed using “FindMarkers” function from Seurat with Wilcoxon rank-sum test. The P values of GO terms in enrichment analysis were calculated using “enrichGO” function from clusterProfiler with hypergeometric test. All multiple testing P values were adjusted using Benjamini-Hochberg procedure. Adjusted P value<0.05 was considered statistically significant.

Data availability

The processed scRNA-seq data saved as a Seurat object and original code are publicly available at https://doi.org/10.5281/zenodo.8222832. The original data can be accessed from public databases described in “Data collection” section.

Results

An integrated lung cellular landscape of COVID-19 and lung cancers

We collected four scRNA-seq datasets comprising a total of 366,096 cells from COVID-19, lung cancers, and healthy control samples (Figure 1A). For the two COVID-19 datasets, one was derived from lung samples of deceased COVID-19 patients (n=79,636 cells) (17), and the other was derived from BALF samples of severe COVID-19 patients (n=62,580 cells) (13), The lung cancer dataset was derived from lung tumor samples of LUAD patients (n=45,760 cells), tumor samples of SCLC patients (n=31,319 cells), and tumor-adjacent normal tissue sample of LUAD patients (n=12,251 cells) (14,15). The healthy control dataset consisted of cells extracted from the Human Lung Cell Atlas (HLCA, n=134,550 cells) (19). We used softwares scANVI (20) and scArches (21) to integrate datasets for consistent cell type annotation. scANVI, known for its superior performance for atlas-level data integration (28), facilitated the projection of the COVID-19 and lung cancer datasets onto the reference HLCA dataset. Subsequently, scArches transferred the cell type annotation from the HLCA dataset to all single cells in COVID-19 and lung cancer datasets. The cell type annotation followed a hierarchical structure with five levels. At the lowest level, the cells are labeled into broad categories including immune, epithelial, stromal, and endothelial cells (ECs). At higher levels, the cells are labeled with finer annotations (Figure S1A). In addition, we generated consistent low-dimensional embeddings of the COVID-19 and lung cancer datasets using scANVI, which were then visualized together using UMAP (Figure 1B,C). We observed cohesive clusters for the curated intermediate level cell type annotation (Figure 1B), as well as for other levels of cell type annotation (Figure S1B).

Figure 1. Analysis of cellular landscape of COVID-19 and lung cancers using lung scRNA-seq datasets.

Figure 1.

Open in a new tab

(A) Overview of the study design. (B) Uniform manifold approximation and projection (UMAP) plot showing the intermediate level annotation of cell populations. (C) UMAP plot showing cells from different lung conditions. (D) The percentage of cell populations in different conditions. The significant differences between other conditions and normal samples are labelled with asterisks in the plot. (E) The expression level of top three marker genes of each cell population. See also Figure S1.

To examine the alteration in cell composition across different conditions, we compared the proportion of various cell types in COVID-19 and lung cancer samples with that in normal controls (Figure 1D). While substantial within-group variations were observed in many cell types, we could identify significant differences between the diseased conditions and the normal controls. Notably, COVID-19 BALF samples exhibited an enrichment of monocytes due to emergency myelopoiesis and defective monocyte activation, while COVID-19 lung samples presented an enrichment of fibroblasts, a characteristic feature of pulmonary fibrosis. SCLC samples displayed a remarkably high proportion of airway epithelial cells, which were also identified as tumorous epithelial cells in the original study (14). To further characterize these cell types, we conducted differential gene expression analysis to identify their top marker genes (Figure 1E). Our results confirmed the presence of many canonical marker genes, such as C1QA for macrophages, CD3D for T cells, and COL1A2 for fibroblasts, thereby validating the accurate cell type annotation achieved by scArches.

Upregulated TMPRSS2 in LUAD epithelial cells implies an increased risk of SARS-CoV-2 entry

The alveolar and airway epithelia are the primary infection sites for SARS-CoV-2. Two vital cell surface receptors, ACE2 and TMPRSS2, are involved in SARS-CoV-2 infection. ACE2 acts as the main entry receptor for SARS-CoV-2, allowing the virus to attach and gain entry into host cells. TMPRSS2 facilitates viral entry by priming the spike protein of the virus, thereby enabling its fusion with the host cell membrane (29). The presence of ACE2 and TMPRSS2 in alveolar and airway epithelia makes these cells particularly susceptible to infection by SARS-CoV-2 and contributes to the respiratory symptoms associated with COVID-19 (30,31).

We focused on epithelial cell populations to analyze the expression of genes ACE2 and TMPRSS2. Alveolar epithelial cell subpopulations include alveolar type I (AT1) cells and alveolar type II (AT2) cells. Airway epithelial cell subpopulations include basal resting, suprabasal, multiciliated, club, goblet cells as well as some rare populations including deuterosomal, ionocyte, tuft, and neuroendocrine cells (Figure 2A,B). By comparing TMPRSS2 expression in epithelial cell subpopulations between diseased conditions and normal, we observed higher expression of TMPRSS2 in AT1 cells from COVID-19 lung samples and slightly elevated expression in multiciliated cells from COVID-19 BALF samples. Note that these findings were not reported in the original studies (13,17), possibly due to our inclusion of more normal samples and cells from the HLCA for comparison in our study. Particularly, we found significant upregulation of TMPRSS2 in large epithelial populations, including AT1, AT2, club, goblet, multiciliated, and transitional club-AT2 cells, in LUAD tumor tissue (Figure 2C), which is consistent with several other studies (32,33). However, such upregulation was not observed in SCLC or tumor adjacent tissue, potentially because SCLC tumor cells derive from pulmonary neuroendocrine cells rather than the cell types highly expressing TMPRSS2, such as ciliated cells, secretory cells and AT1 (34). This suggests that the upregulation of TMPRSS2 is a unique feature of LUAD tumor. In contrast, we found that the expression of ACE2 was generally lower than TMPRSS2 in all epithelial cell populations, with no significant difference observed between diseased conditions and normal controls (Figure S2).

Figure 2. Epithelial cell subpopulations and TMPRSS2 expression in COVID-19 and lung cancers.

Figure 2.

Open in a new tab

(A) UMAP plot showing different alveolar and airway epithelial cell subpopulations. (B) UMAP plot showing epithelial cells from different conditions. (C) TMPRSS2 expression in epithelial cell subpopulations in different conditions. (D) TMPRSS2 expression by sex in epithelial cell subpopulations in different conditions. Cells in grey color indicates insufficient number of cells for differential gene expression analysis. (E) The composition of epithelial cell subpopulations in different conditions. See also Figure S2.

The previous studies have reported that TMPRSS2 expression is influenced by sex hormones (35,36). Here, we analyzed its expression by sex (Figure 2D). COVID-19 lung samples from both female and male patients showed higher TMPRSS2 expression in major epithelial cell, with significantly higher expression in AT1 cells. In LUAD patients, both females and males exhibited significantly higher TMPRSS2 expression in transitional club-AT2 and multiciliated cells, while AT1 and AT2 cells showed significantly higher expression only in females. Notably, tumor-adjacent tissue displayed significantly higher TMPRSS2 expression in AT1, goblet, and multiciliated cells in males, but not in females.

Furthermore, we characterized the composition of different epithelial cell populations in diseased conditions and normal lung tissue (Figure 2E). The cell populations with significant upregulation of TMPRSS2 constituted the largest proportion of epithelial cells in LUAD and tumor-adjacent samples, accounting for approximately 90% of all epithelial cells. Collectively, our findings demonstrate that upregulation of TMPRSS2 is a common feature in most alveolar and airway epithelial cells of LUAD patients, but not in SCLC patients. Both female and male LUAD patients had this TMPRSS2 upregulation to some extent, and even tumor-adjacent tissue exhibited this upregulation. These observations suggested a higher susceptibility to SARS-CoV-2 infection through facilitated viral entry in LUAD patients (37–39).

Pre-existing proinflammatory macrophages in lung cancers implies an increased risk of hyperinflammation

Excessive inflammation is a hallmark of severe COVID-19. Upon infection with SARS-CoV-2, various immune cells are activated, triggering the production of proinflammatory cytokines (16). The heightened inflammation can cause damage to healthy tissues and organs, leading to severe complications in COVID-19 patients, such as acute respiratory distress syndrome (ARDS), organ dysfunction, and in some cases, multiorgan failure.

To investigate the potential role of existing immune cells in inflammation, we focused on macrophages, the largest tissue-resident myeloid cell population known for their involvement in inflammation. By analyzing integrated data from COVID-19, lung cancers, and normal controls, we identified nine macrophage clusters (Figure 3A). Notably, cluster 2 (Mφ2) exhibited high expression level of canonical proinflammatory cytokines, including IL1B, IL6, and TNF, suggesting that these macrophages primarily exhibited a proinflammatory phenotype (Figure 3B). We further compared the proportions of macrophage clusters across different conditions (Figure 3C, S3A,B). Interestingly, the proportion of Mφ2 was significantly higher in COVID-19 BALF and LUAD than that in the normal lung tissue. We observed the same trend (high expression) in SCLC samples as well, even though not significantly. Of note, nearly no macrophages in COVID-19 lung was in Mφ2, potentially because these samples were from deceased patients whose immune responses were suppressed. Additionally, the largest macrophage cluster, cluster 0 (Mφ0), was significantly reduced in both COVID-19 and lung cancer cases. These consistent proportional changes in COVID-19 and lung cancers suggested the presence of shared inflammatory changes in macrophages in these conditions. To gain insight into the biological functions of the different macrophage clusters, we conducted pathway enrichment analysis using cluster marker genes. Mφ0 was primarily involved in antigen presentation through major histocompatibility complex class II (MHC-II) and antibody-mediated immunity. In contrast, Mφ2 primarily participated in response to IL1 and TNF, as well as immune cell chemotaxis (Figure S3C). These findings collectively indicated a consistent increase in proinflammatory macrophages and a decrease in antigen presentation macrophages in both COVID-19 and lung cancer cases.

Figure 3. Macrophage subpopulations and inflammation in COVID-19 and lung cancers.

Figure 3.

Open in a new tab

(A) UMAP plot showing 9 macrophage clusters (Mφ0~8). (B) The expression of proinflammatory cytokines in 9 macrophage clusters. (C) The percentage of Mφ0 and Mφ2 in different conditions. The significant differences between each diseased condition and normal samples are labelled in the plot. (D) Significantly differentially expressed genes shared by macrophages from COVID-19 and lung cancers compared to normal samples. (E) The predicted intercellular communication among macrophage clusters and other immune cells through IL1B and SPP1. See also Figure S3.

In addition to the clustering analysis, we performed the differential gene expression analysis in macrophages to identify consistently upregulated and downregulated genes in COVID-19 and lung cancers compared to normal samples (Figure 3D). Twelve consistently upregulated genes and twenty downregulated genes were identified with the strict criteria (log2 Fold Change>1.5, adjusted p-value<0.05). Most of the results were shared in alveolar and interstitial macrophages. Several proinflammatory cytokines, including IL1B, CCL3, CCL4, and CXCL8, were consistently and strongly upregulated in COVID-19 BALF, LUAD, and SCLC. Other consistently upregulated genes included early response genes IER3, HSPA6, and HSPH1, as well an interferon-stimulated gene IFI27. Another noteworthy upregulated gene was SPP1, which was reported to be upregulated in multiple pathological conditions like idiopathic pulmonary fibrosis (IPF), COVID-19 and multiple cancers (14,17,40,41). It plays roles in chronic inflammation and T cell inhibition (42–44). Among the consistently and strongly downregulated genes, FABP4 stood out, as it could be considered a marker of macrophages in healthy controls. A previous IPF study has shown a dramatic decrease in FABP4hi macrophages alongside a dramatic increase in SPP1hi macrophages (41). Our results indicated that the downregulation of FABP4 and upregulation of SPP1 in IPF macrophages were shared among COVID-19, LUAD, and SCLC.

Considering that macrophages play a key role in mediating inflammation through cytokines, we explored the intercellular communication between macrophage clusters with other immune cells using CellChat (Figure 3E, S3D) (26). We observed that IL1 and CCL3 signaling were mainly present between macrophages and monocytes, and between macrophages and dendritic cells (DCs). The strongest communication via IL1 occurred between proinflammatory Mφ2 macrophages and classical monocytes, potentially promoting the differentiation of monocytes into conventional DCs and proinflammatory macrophages (45). Furthermore, we found SPP1-CD44 signaling between macrophages (Mφ2 or Mφ4) and most immune cell populations. This signaling pathway may inhibit the T cell response (42,44). Collectively, our findings revealed that macrophage-mediated proinflammatory signaling was consistently upregulated in COVID-19 and lung cancers. In cases where lung cancer patients were infected with SARS-CoV-2, the pre-existing proinflammatory signaling could exacerbate pathological inflammation or drive hyperinflammation in the lungs, ultimately leading to severe COVID-19 symptoms (46).

Suppressed T cell response in lung cancers implies the ineffective antiviral response

T cell inhibition in the tumor microenvironment (TME) is a crucial mechanism facilitating immune evasion by tumor cells. Inhibitory immune checkpoints expressed on T cells such as PD1 and CTLA4 can be engaged by ligands on tumor or immune cells, leading to the suppression of T cell function (47). Additionally, regulatory T cells (Tregs) play a significant role in T cell inhibition through various mechanisms (48). In the TME, it is likely that T cells targeting SARS-CoV-2 infected cells could also be suppressed through these mechanisms.

In the context of lung cancers and COVID-19, we analyzed T cell subpopulations in our integrated scRNA-seq data. Using reported marker genes, we annotated T cell subsets, including effector memory T cells (Tem) expressing GZMK or PRF1, exhausted T cells (Tex) expressing PDCD1 and LAG3, and interferon-stimulated gene (ISG)+ subpopulations expressing IFIT1 and IFIT3 (Figure 4A,B). Interestingly, we observed two abnormal T cell subpopulations in the data, characterized by low expression of common T cell markers and effector genes such as CD3E, CD8A, and CD4. These subpopulations were predominantly derived from COVID-19 lung samples, indicating potential immune suppression in deceased patients (Figure S4A,B).

Figure 4. T cell subpopulations and T cell inhibition in COVID-19 and lung cancers.

Figure 4.

Open in a new tab

(A) UMAP plot showing annotated T cell subpopulations. (B) The expression of selected marker genes for T cell subpopulations. (C) The percentage of Tem, Tex, Tact, Th, and Treg subpopulations in different conditions. The significant differences between each diseased condition and normal samples are labelled in the plot. (D) Heatmap showing top differentially expressed genes between CTLA4+ Th and Treg. (E) The predicted intercellular communication between DCs and T cell subpopulations through CD86. (F) The expression of coinhibitory receptor genes and FOXP3 in CD4+ T cell subpopulation. (G) The expression of coinhibitory receptor genes in CD8+ T cell subpopulation. See also Figure S4.

Next, we compared the proportion of T cell subpopulations between diseased conditions and normal samples (Figure 4C, S4A). Among large CD8+ T cell subpopulations, PRF1+ Tem showed a consistent and significant decrease in COVID-19 and lung cancers. Conversely, Tex exhibited a consistent increase in COVID-19 BALF and lung cancers, with the most significant increase observed in COVID-19 BALF. Among large CD4+ T cell subpopulations, the activated T cell (Tact) subpopulation consistently and significantly decreased in COVID-19 and lung cancers. Treg showed a significant increase in lung cancers but not in COVID-19. Interestingly, we also identified another inhibitory helper T cell (Th) subpopulation characterized by high expression of CTLA4 but lack of expression of FOXP3. This CTLA4+ Th subpopulation consistently and significantly increased in COVID-19 BALF and lung cancers. This subpopulation aligns with the concept of exhausted CD4+ T cells with lost helper function that has been recently proposed (49,50). Differential gene expression analysis between CTLA4+ Th with Treg revealed that CTLA4+ Th highly expressed CXCL13 but not Treg markers FOXP3 and IL2RA (Figure 4D). To explore the intercellular communication between T cells and professional antigen-presenting cells (APCs) which is crucial for T cell activation, we predicted the intercellular communication between T cells and DCs. We found a high probability of the inhibitory interaction between CD86 on DCs and CTLA4 on CTLA4+ Th or Treg cells. In contrast, the probability of the activating interaction between CD86 on DCs and CD28 on T cell subpopulations was much lower (Figure 4E). Considering the report that CTLA4 can inhibit CD28 costimulation by depletion of CD86 on APCs (51), the high proportion of T cells in CTLA4+ Th and Treg subpopulations in the TME and COVID-19 BALF could inhibit the activation of other T cells through inactivation of DCs. In addition, we identified the interaction of another immunosuppressive ligand LGALS9 on DCs with receptors on T cell subpopulations, including LGALS9-HAVCR2 which contributes to the persistence of PDCD1+HAVCR2+ T cells (52), and LGALS9-CD44 which promotes Treg differentiation and maintenance (Figure S4C) (53).

To further evaluate T cell inhibition and exhaustion, we examined the expression of coinhibitory receptor genes, LAG3, HAVCR2, PDCD1, TIGIT, and CTLA4, in both CD4+ and CD8+ T cells (Figure 4F,G). Among CD4+ T cell subpopulations, naïve or memory T cell (Tn/Tm) and Tact showed minimal expression of coinhibitory receptor genes, while Treg cells presented high expression across all conditions. In contrast, CTLA4+ Th had high expression of these genes in COVID-19 BALF and lung cancers, but very low expression in normal and tumor adjacent tissues (Figure 4F). Similarly, among CD8+ T cell subpopulations, Tex, ISG+ Tem, and GZMK+ Tem displayed high expression of these genes exclusively in COVID-19 BALF and lung cancers (Figure 4G). Collectively, these results indicated the presence of similar patterns of T cell inhibition and exhaustion in both COVID-19 and lung cancers. In cases where lung cancer patients were infected with SARS-CoV-2, the immunosuppressive TME could hinder the activation and effector response of T cells against SARS-CoV-2 infected cells, resulting in an impaired antiviral response that may not be mounted effectively or in a timely manner (46).

Pre-existing profibrotic fibroblasts in lung cancers implies an increased risk of pulmonary fibrosis

Pulmonary fibrosis has emerged as a significant complication in some COVID-19 patients , with approximately 44.9% of the survivors developing this condition according to a recent meta-analysis (54). Following SARS-CoV-2 infection, the virus can cause direct damage to lung cells and trigger an exaggerated inflammatory response. This persistent inflammation stimulates the activation of fibroblasts, which are responsible for producing collagen and other components of the extracellular matrix (ECM). The abnormal accumulation of collagen and fibrotic tissue in the lungs results in the formation of scar tissue (55).

To assess the risk of pulmonary fibrosis in COVID-19 and lung cancers, we conducted an analysis focusing on fibroblasts, because fibroblasts are the predominant stromal cell population and play a crucial role in pulmonary fibrosis. We excluded the samples of COVID-19 BALF and tumor adjacent tissues because the number of captured fibroblasts was limited (approximately 30 cells). We identified nine fibroblast clusters along with a pericyte cluster (Figure 5A). Two large fibroblast clusters, 0 and 2 (FB0 and FB2), showed significant proportional difference between diseased conditions and normal (Figure 5B). FB2 was predominantly found in normal samples, while FB0 was enriched in both COVID-19 and lung cancers (Figure S5A,B). The top marker genes of FB0 included COL1A1, COL3A1, POSTN, and FN1, all of which are ECM components and associated with pulmonary fibrosis (Figure 5C) (56,57). Another noteworthy marker gene, CTHRC1, was identified as a hallmark gene for pathological fibroblasts in idiopathic pulmonary fibrosis and COVID-19 (17,58). Using pySCENIC (27), we further identified the top two transcription factors in FB0, TWIST1 and CREB3L1 (Figure 5C). TWIST1 enhances fibroblast activation by amplifying TGFβ signaling (59), while CREB3L1, activated by TGFβ, induces the transcription of genes involved in collagen-containing ECM assembly (60). Pathway enrichment analysis revealed that FB0 primarily functioned in ECM organization, while FB2 was associated with energy metabolism (Figure S5C). Together, these results indicated a substantial increase of a profibrotic fibroblast subpopulation in both COVID-19 and lung cancers.

Figure 5. Fibroblast subpopulations and TGFβ signaling in COVID-19 and lung cancers.

Figure 5.

Open in a new tab

(A) UMAP plot showing 9 fibroblast clusters (FB0–8). In particular, pericyte (FB6) is considered a subpopulation of fibroblasts according to HLCA annotation. (B) The percentage of FB0 and FB2 in different conditions. The significant differences between each diseased condition and normal samples are labelled in the plot. (C) Heatmap showing the expression of some pathological fibroblast marker genes and transcription factors, as well as the regulon activity of transcription factors in fibroblast clusters. (D) Significantly differentially expressed genes shared by fibroblasts from COVID-19 and lung cancers compared to normal samples. (E) Enriched biological functions of all fibroblasts in COVID-19 and lung cancers compared to normal samples. (F) The predicted intercellular communication among fibroblast clusters and immune cells through TGFβ. (G) TGFB1 expression in fibroblasts and immune cells from different conditions. See also Figure S5.

Next, we focused on identifying consistently upregulated and downregulated genes in fibroblasts from COVID-19 and lung cancers by comparing with normal samples (Figure 5D). We observed that many of these genes were shared across different fibroblast subtypes, including adventitial, alveolar, peribronchial fibroblasts, and myofibroblasts. The consistently and significantly upregulated genes included several ECM component genes COL1A1, COL3A1, COL5A1, COL5A2, POSTN, and SULF1. Conversely, the expression of many other genes encoding extracellular protein was consistently downregulated, such as SPARKL1, CFD, CLU, GSN, and DCN. In our pathway enrichment analysis of differentially expressed genes between all fibroblasts in each condition and normal samples, we revealed consistent functional associations with fibrosis in COVID-19 and lung cancers (Figure 5E). These functions encompassed ECM organization, cell-matrix adhesion, and Wnt signaling pathway, which is needed for TGFβ-mediated fibrosis (61).

Interestingly, we observed that multiple collagen genes were upregulated not only in fibroblasts but also in various other cell types (Figure S5D). Pericytes, smooth muscle cells, epithelial cells, and ECs all demonstrated upregulation of certain collagen genes in both COVID-19 and lung cancers. Furthermore, more cell populations exhibited significant upregulation of collagen genes in LUAD than in SCLC and COVID-19. Conversely, tumor adjacent tissues showed a similar pattern to normal samples, with nearly no significant upregulation of collagen genes. These findings indicated the involvement of multiple cell types in collagen deposition and the formation of aberrant ECM.

Considering the crucial role of TGFβ signaling in fibrosis development (62,63), we further investigated the intercellular communication between fibroblast clusters and other cells (Figure 5F). Our results pointed out that immune cells, such as T cells, B cells, NK cells, and macrophages, were the primary sources of TGFβ release. Notably, the fibroblast clusters FB0 and FB9, which exhibited high expression levels of profibrotic genes and TGFβ signaling-related transcription factors (Figure 5C), were identified as key recipients of signaling from these immune cells. This observation suggested that the expression of profibrotic genes within these fibroblast clusters was likely driven by TGFβ signaling under certain pathological conditions. Additionally, we observed a significant upregulation of TGFB1, the gene encoding TGFβ, in fibroblasts, B cells, T cells, and NK cells in both COVID-19 and lung cancers (Figure 5G), indicating a potential contribution to both fibrosis and immune suppression (64). Collectively, our results demonstrated an increased proportion of profibrotic fibroblasts and upregulated profibrotic genes in both COVID-19 and lung cancers, which were likely driven by TGFβ signaling. In cases where lung cancer patients were infected with SARS-CoV-2, the pre-existing profibrotic signaling and cell populations might experience further elevation in response to lung tissue damage, ultimately heightening the risk of pulmonary fibrosis development.

Dysregulated signaling patterns in COVID-19 and lung cancers

Given the crucial role of cytokine signaling in the development of pathological inflammation and fibrosis in COVID-19, we conducted a characterization of global signaling patterns in COVID-19 and lung cancers. Intercellular communication was predicted among cell populations, as shown in Figure 1B, with ligand-receptor pairs summarized into signaling pathways. We identified the major upregulated signaling pathways in COVID-19 and lung cancers (Figure 6A) and determined the primary cell populations contributing to each pathway in different conditions (Figure 6B, S6).

Figure 6. Dysregulated signaling in COVID-19 and lung cancers.

Figure 6.

Open in a new tab

(A) The predicted relative strength of signaling pathways in different condition. (B) The major signaling pathways (in black color) and cell populations (in red color) contributing to each pathway in different conditions. See also Figure S6.

The most prominently upregulated signaling pathways in COVID-19 and lung cancers were associated with immune responses. These included MHC-I signaling in T and NK cells, MHC-II signaling in macrophages and DCs, galectin signaling in ECs and some immune cells, visfatin signaling in ECs and some immune cells, ICAM signaling in epithelial cells, ECs, and some immune cells, CCL signaling in some immune cells, CXCL signaling in ECs and some immune cells, and CD22 signaling in B cells (Figure 6A, B). Many are involved in immune cell recruitment and regulation during inflammation and play significant roles in immune responses (65–67). Particularly, COVID-19 exhibited very low level of these signaling pathways, indicating a suppressed immune response consistent with the analysis results of macrophages and T cells. In addition to immune response-related signaling pathways, we observed pathways that promote tumor growth in lung cancers or pulmonary fibrosis in COVID-19. These pathways included various growth factor pathways, primarily upregulated in LUAD and COVID-19 lung tissue, such as VEGF, PDGF, TGFβ, FGF, and IGF (68–71). The primary cell populations associated with these signaling pathways were fibroblasts, ECs, and epithelial cells (Figure 6A, B). These findings highlighted the involvement of these cell types in promoting tumor growth and fibrosis development. Another noteworthy upregulated signaling pathway in both COVID-19 BALF and lung cancers was SPP1 signaling, which showed significantly increased expression in macrophages (Figure 3D). SPP1 acts as a link between cancer, inflammation, and fibrosis, and its upregulation further underscores the interconnectedness of these processes (41,72,73).

In summary, our analyses revealed the shared signaling pathways between COVID-19 and lung cancers that contribute to inflammation and pulmonary fibrosis. These pathways involve the release of paracrine cytokines, which can impact nearby normal lung tissue. Accordingly, we proposed the following scenario. In cases where lung cancer patients were infected with SARS-CoV-2, the presence of pre-existing proinflammatory and profibrotic signaling could enhance the development of pathological inflammation and pulmonary fibrosis.

Discussion

In this study, we integrated multiple scRNA-seq datasets from lung tissue of the COVID-19 patients, lung cancer patients, and healthy controls for a comprehensive study of the cellular landscape of the lung in these diseases. Despite the distinct pathogenesis of COVID-19 and lung cancers, we identified many shared molecular features at the cellular level, as well as unique characteristics of each disease. Our results provided valuable insights into the potential factors contributing to the heightened vulnerability of lung cancer patients to COVID-19 and shed light on the interplay between COVID-19 and lung cancers.

Through the analysis of major immune and non-immune cell populations in COVID-19 and lung cancers, we characterized several molecular aspects that may contribute to the increased vulnerability of lung cancer patients to COVID-19. In the early stage, LUAD patients with high expression of TMPRSS2 in alveolar and epithelial cells are more likely to be infected by SARS-CoV-2 by increased spike protein priming. Next, high proportion of tissue-resident proinflammatory macrophages in lung cancers, characterized by high expression of IL1B, are more likely to promote excessive inflammation. The release of proinflammatory cytokines such as IL1B, CCL3, and CCL4 recruits other proinflammatory immune cells. After SARS-CoV-2 infection, the effector functions of virus-specific T cells are suppressed in the TME due to the high proportion of inhibitory Th and Treg. CD8+ T cells also display a more exhausted phenotype with increased expression of coinhibitory receptors. The impaired ability to eliminate infected cells hinders the control of SARS-CoV-2 proliferation, and inflammation cannot be timely resolved. Patients will likely have severe symptoms at this stage (46). Later, virus-mediated tissue damage and sustained inflammation may lead to pathological wound healing, characterized by excessive deposition of ECM components. High proportion of collagen-expressing fibroblasts, potentially driven by TGFβ signaling from multiple cell populations, likely play a central role in this process. Other cell populations, such as collagen-expressing smooth muscle cells, epithelial cells, and ECs, as well as SPP1+ macrophages may also contribute to fibrosis. Consequently, lung cancer patients with COVID-19 are likely to experience worse outcomes.

It is noteworthy that several of the identified mechanisms are not relevant to adaptive immune responses. For lung cancer patients who are fully vaccinated against SARS-CoV-2 and develop effective responses, T cell suppression in TME could be mitigated. However, pre-existing upregulated TMPRSS2 expression, the high proportion of proinflammatory macrophages and profibrotic fibroblasts, as well as proinflammatory and profibrotic signaling are unlikely to be influenced by vaccination. This could partially explain why some cancer patients remain at high risk to COVID-19 breakthrough and severe symptoms (6,74).

While our study generated reliable results by analyzing published high-quality scRNA-seq datasets with a large number of samples and cells, it has some limitations. First, we could not find scRNA-seq data from lung cancer patients who also developed COVID-19. Currently, there are no published datasets available for such patients, and acquiring lung samples from this specific group for scRNA-seq is not feasible. To meet this challenge, we inferred potential risk factors for COVID-19 in lung cancer patients based on consistent pathological features observed in both conditions. Future studies utilizing samples from lung cancer patients with COVID-19 will further elucidate the molecular mechanisms underlying the vulnerability of lung cancer patients to COVID-19. Second, we observed strong differences between lung tissue samples from deceased COVID-19 patients and BALF samples from severe COVID-19 patients, mainly in immune cell populations. Immune cells, such as macrophages and T cells, generally exhibited unique subpopulation compositions and significantly lower expression of immune response-related genes in lung tissue from COVID-19 patients (Figure 3D, S3A, S4A). We believe that immune responses were generally suppressed in the deceased patients; thus, our focus was primarily on comparing immune cells between COVID-19 BALF samples and lung cancer samples. Third, we included only LUAD as the representative non-small cell lung carcinoma (NSCLC) but did not include other NSCLC subtypes. Our study could be expanded given the release of such NSCLC subtype scRNA-seq data in future. Finally, to validate and prove the mechanistic roles of our findings, additional experiments using the appropriate tissues from the patients or the animal models are needed, but such tissues are currently difficult to obtain.

In summary, our comprehensive analysis of multiple scRNA-seq datasets revealed four molecular mechanisms contributing to the heightened vulnerability of lung cancer patients to COVID-19: upregulated TMPRSS2 expression in epithelial cells, stronger inflammatory responses mediated by macrophages, suppressed T cell response, and an increased risk of fibrosis due to pathological fibroblasts. These findings offer some valuable insights into potential preventive and therapeutic targets for COVID-19 in lung cancer patients.

Supplementary Material

1
2
3
4
5
6
7

Implications:

Our work reveals the potential molecular mechanisms contributing to the vulnerability to COVID-19 in lung cancer patients.

Acknowledgements

W. Liu is a CPRIT Predoctoral Fellow in the Biomedical Informatics, Genomics and Translational Cancer Research Training Program (BIG-TCR) funded by Cancer Prevention & Research Institute of Texas (CPRIT RP210045). Z.Z. was partially supported by National Institutes of Health grants (R01LM012806, U01AG079847, and R01CA276513) and Cancer Prevention and Research Institute of Texas (CPRIT RP180734 and RP210045). The visual overview and Figure 1A were created with BioRender.com.

Footnotes

Declaration of interests

The authors declare no competing interests.

References

  • 1.Ahmad FB, Cisewski JA, Xu J, Anderson RN. Provisional Mortality Data - United States, 2022. MMWR Morb Mortal Wkly Rep 2023;72(18):488–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Dai M, Liu D, Liu M, Zhou F, Li G, Chen Z, et al. Patients with Cancer Appear More Vulnerable to SARS-CoV-2: A Multicenter Study during the COVID-19 Outbreak. Cancer Discov 2020;10(6):783–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Robilotti EV, Babady NE, Mead PA, Rolling T, Perez-Johnston R, Bernardes M, et al. Determinants of COVID-19 disease severity in patients with cancer. Nat Med 2020;26(8):1218–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Wang Q, Berger NA, Xu R. Analyses of Risk, Racial Disparity, and Outcomes Among US Patients With Cancer and COVID-19 Infection. JAMA Oncol 2021;7(2):220–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Ehmsen S, Asmussen A, Jeppesen SS, Nilsson AC, Osterlev S, Vestergaard H, et al. Antibody and T cell immune responses following mRNA COVID-19 vaccination in patients with cancer. Cancer Cell 2021;39(8):1034–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Schmidt AL, Labaki C, Hsu CY, Bakouny Z, Balanchivadze N, Berg SA, et al. COVID-19 vaccination and breakthrough infections in patients with cancer. Ann Oncol 2022;33(3):340–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Tran S, Truong TH, Narendran A. Evaluation of COVID-19 vaccine response in patients with cancer: An interim analysis. Eur J Cancer 2021;159:259–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Kuderer NM, Choueiri TK, Shah DP, Shyr Y, Rubinstein SM, Rivera DR, et al. Clinical impact of COVID-19 on patients with cancer (CCC19): a cohort study. Lancet 2020;395(10241):1907–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Bakouny Z, Labaki C, Grover P, Awosika J, Gulati S, Hsu CY, et al. Interplay of Immunosuppression and Immunotherapy Among Patients With Cancer and COVID-19. JAMA Oncol 2023;9(1):128–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Grivas P, Khaki AR, Wise-Draper TM, French B, Hennessy C, Hsu CY, et al. Association of clinical factors and recent anticancer therapy with COVID-19 severity among patients with cancer: a report from the COVID-19 and Cancer Consortium. Ann Oncol 2021;32(6):787–800. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Leyfman Y, Emmanuel N, Menon GP, Joshi M, Wilkerson WB, Cappelli J, et al. Cancer and COVID-19: unravelling the immunological interplay with a review of promising therapies against severe SARS-CoV-2 for cancer patients. J Hematol Oncol 2023;16(1):39. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Zong Z, Wei Y, Ren J, Zhang L, Zhou F. The intersection of COVID-19 and cancer: signaling pathways and treatment implications. Mol Cancer 2021;20(1):76. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Bost P, De Sanctis F, Cane S, Ugel S, Donadello K, Castellucci M, et al. Deciphering the state of immune silence in fatal COVID-19 patients. Nat Commun 2021;12(1):1428. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Chan JM, Quintanal-Villalonga A, Gao VR, Xie Y, Allaj V, Chaudhary O, et al. Signatures of plasticity, metastasis, and immunosuppression in an atlas of human small cell lung cancer. Cancer Cell 2021;39(11):1479–96 e18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Laughney AM, Hu J, Campbell NR, Bakhoum SF, Setty M, Lavallee VP, et al. Regenerative lineages and immune-mediated pruning in lung cancer metastasis. Nat Med 2020;26(2):259–69. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Liu W, Jia J, Dai Y, Chen W, Pei G, Yan Q, et al. Delineating COVID-19 immunological features using single-cell RNA sequencing. Innovation (Camb) 2022;3(5):100289. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Melms JC, Biermann J, Huang H, Wang Y, Nair A, Tagore S, et al. A molecular single-cell lung atlas of lethal COVID-19. Nature 2021;595(7865):114–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Kim P, Park A, Han G, Sun H, Jia P, Zhao Z. TissGDB: tissue-specific gene database in cancer. Nucleic Acids Res 2018;46(D1):D1031–D8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Sikkema L, Ramirez-Suastegui C, Strobl DC, Gillett TE, Zappia L, Madissoon E, et al. An integrated cell atlas of the lung in health and disease. Nat Med 2023;29(6):1563–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Xu C, Lopez R, Mehlman E, Regier J, Jordan MI, Yosef N. Probabilistic harmonization and annotation of single-cell transcriptomics data with deep generative models. Mol Syst Biol 2021;17(1):e9620. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Lotfollahi M, Naghipourfar M, Luecken MD, Khajavi M, Buttner M, Wagenstetter M, et al. Mapping single-cell data to reference atlases by transfer learning. Nat Biotechnol 2022;40(1):121–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Hao Y, Hao S, Andersen-Nissen E, Mauck WM 3rd, Zheng S, Butler A, et al. Integrated analysis of multimodal single-cell data. Cell 2021;184(13):3573–87 e29. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal statistical society: series B (Methodological) 1995;57(1):289–300. [Google Scholar]
  • 24.Korsunsky I, Millard N, Fan J, Slowikowski K, Zhang F, Wei K, et al. Fast, sensitive and accurate integration of single-cell data with Harmony. Nat Methods 2019;16(12):1289–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Wu T, Hu E, Xu S, Chen M, Guo P, Dai Z, et al. clusterProfiler 4.0: A universal enrichment tool for interpreting omics data. Innovation (Camb) 2021;2(3):100141. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Jin S, Guerrero-Juarez CF, Zhang L, Chang I, Ramos R, Kuan CH, et al. Inference and analysis of cell-cell communication using CellChat. Nat Commun 2021;12(1):1088. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Van de Sande B, Flerin C, Davie K, De Waegeneer M, Hulselmans G, Aibar S, et al. A scalable SCENIC workflow for single-cell gene regulatory network analysis. Nat Protoc 2020;15(7):2247–76. [DOI] [PubMed] [Google Scholar]
  • 28.Luecken MD, Buttner M, Chaichoompu K, Danese A, Interlandi M, Mueller MF, et al. Benchmarking atlas-level data integration in single-cell genomics. Nat Methods 2022;19(1):41–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Hoffmann M, Kleine-Weber H, Schroeder S, Kruger N, Herrler T, Erichsen S, et al. SARS-CoV-2 Cell Entry Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven Protease Inhibitor. Cell 2020;181(2):271–80 e8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Ziegler CGK, Allon SJ, Nyquist SK, Mbano IM, Miao VN, Tzouanas CN, et al. SARS-CoV-2 Receptor ACE2 Is an Interferon-Stimulated Gene in Human Airway Epithelial Cells and Is Detected in Specific Cell Subsets across Tissues. Cell 2020;181(5):1016–35 e19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Bost P, Giladi A, Liu Y, Bendjelal Y, Xu G, David E, et al. Host-Viral Infection Maps Reveal Signatures of Severe COVID-19 Patients. Cell 2020;181(7):1475–88 e12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Singh MK, Mobeen A, Chandra A, Joshi S, Ramachandran S. A meta-analysis of comorbidities in COVID-19: Which diseases increase the susceptibility of SARS-CoV-2 infection? Comput Biol Med 2021;130:104219. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Wang Q, Li L, Qu T, Li J, Wu L, Li K, et al. High Expression of ACE2 and TMPRSS2 at the Resection Margin Makes Lung Cancer Survivors Susceptible to SARS-CoV-2 With Unfavorable Prognosis. Front Oncol 2021;11:644575. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Schuler BA, Habermann AC, Plosa EJ, Taylor CJ, Jetter C, Negretti NM, et al. Age-determined expression of priming protease TMPRSS2 and localization of SARS-CoV-2 in lung epithelium. J Clin Invest 2021;131(1). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Foresta C, Rocca MS, Di Nisio A. Gender susceptibility to COVID-19: a review of the putative role of sex hormones and X chromosome. J Endocrinol Invest 2021;44(5):951–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Gebhard C, Regitz-Zagrosek V, Neuhauser HK, Morgan R, Klein SL. Impact of sex and gender on COVID-19 outcomes in Europe. Biol Sex Differ 2020;11(1):29. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Guo Y, Guo Y, Ying H, Yu W, Chen S, Zhang Y, et al. In-hospital adverse outcomes and risk factors among chronic kidney disease patients infected with the omicron variant of SARS-CoV-2: a single-center retrospective study. BMC Infect Dis 2023;23(1):698. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Yarza R, Bover M, Paredes D, Lopez-Lopez F, Jara-Casas D, Castelo-Loureiro A, et al. SARS-CoV-2 infection in cancer patients undergoing active treatment: analysis of clinical features and predictive factors for severe respiratory failure and death. Eur J Cancer 2020;135:242–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Lemos AEG, Silva GR, Gimba ERP, Matos ADR. Susceptibility of lung cancer patients to COVID-19: A review of the pandemic data from multiple nationalities. Thorac Cancer 2021;12(20):2637–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Cheng S, Li Z, Gao R, Xing B, Gao Y, Yang Y, et al. A pan-cancer single-cell transcriptional atlas of tumor infiltrating myeloid cells. Cell 2021;184(3):792–809 e23. [DOI] [PubMed] [Google Scholar]
  • 41.Morse C, Tabib T, Sembrat J, Buschur KL, Bittar HT, Valenzi E, et al. Proliferating SPP1/MERTK-expressing macrophages in idiopathic pulmonary fibrosis. Eur Respir J 2019;54(2). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Sangaletti S, Tripodo C, Sandri S, Torselli I, Vitali C, Ratti C, et al. Osteopontin shapes immunosuppression in the metastatic niche. Cancer Res 2014;74(17):4706–19. [DOI] [PubMed] [Google Scholar]
  • 43.Lund SA, Giachelli CM, Scatena M. The role of osteopontin in inflammatory processes. J Cell Commun Signal 2009;3(3–4):311–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Klement JD, Paschall AV, Redd PS, Ibrahim ML, Lu C, Yang D, et al. An osteopontin/CD44 immune checkpoint controls CD8+ T cell activation and tumor immune evasion. J Clin Invest 2018;128(12):5549–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Kaneko N, Kurata M, Yamamoto T, Morikawa S, Masumoto J. The role of interleukin-1 in general pathology. Inflamm Regen 2019;39:12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Merad M, Blish CA, Sallusto F, Iwasaki A. The immunology and immunopathology of COVID-19. Science 2022;375(6585):1122–7. [DOI] [PubMed] [Google Scholar]
  • 47.Thommen DS, Schumacher TN. T Cell Dysfunction in Cancer. Cancer Cell 2018;33(4):547–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Takeuchi Y, Nishikawa H. Roles of regulatory T cells in cancer immunity. Int Immunol 2016;28(8):401–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Dong Y, Li X, Zhang L, Zhu Q, Chen C, Bao J, et al. CD4(+) T cell exhaustion revealed by high PD-1 and LAG-3 expression and the loss of helper T cell function in chronic hepatitis B. BMC Immunol 2019;20(1):27. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Miggelbrink AM, Jackson JD, Lorrey SJ, Srinivasan ES, Waibl-Polania J, Wilkinson DS, et al. CD4 T-Cell Exhaustion: Does It Exist and What Are Its Roles in Cancer? Clin Cancer Res 2021;27(21):5742–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Qureshi OS, Zheng Y, Nakamura K, Attridge K, Manzotti C, Schmidt EM, et al. Trans-endocytosis of CD80 and CD86: a molecular basis for the cell-extrinsic function of CTLA-4. Science 2011;332(6029):600–3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Yang R, Sun L, Li CF, Wang YH, Yao J, Li H, et al. Galectin-9 interacts with PD-1 and TIM-3 to regulate T cell death and is a target for cancer immunotherapy. Nat Commun 2021;12(1):832. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Wu C, Thalhamer T, Franca RF, Xiao S, Wang C, Hotta C, et al. Galectin-9-CD44 interaction enhances stability and function of adaptive regulatory T cells. Immunity 2014;41(2):270–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Hama Amin BJ, Kakamad FH, Ahmed GS, Ahmed SF, Abdulla BA, Mohammed SH, et al. Post COVID-19 pulmonary fibrosis; a meta-analysis study. Ann Med Surg (Lond) 2022;77:103590. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Tanni SE, Fabro AT, de Albuquerque A, Ferreira EVM, Verrastro CGY, Sawamura MVY, et al. Pulmonary fibrosis secondary to COVID-19: a narrative review. Expert Rev Respir Med 2021;15(6):791–803. [DOI] [PubMed] [Google Scholar]
  • 56.Naik PK, Bozyk PD, Bentley JK, Popova AP, Birch CM, Wilke CA, et al. Periostin promotes fibrosis and predicts progression in patients with idiopathic pulmonary fibrosis. Am J Physiol Lung Cell Mol Physiol 2012;303(12):L1046–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Upagupta C, Shimbori C, Alsilmi R, Kolb M. Matrix abnormalities in pulmonary fibrosis. Eur Respir Rev 2018;27(148). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Tsukui T, Sun KH, Wetter JB, Wilson-Kanamori JR, Hazelwood LA, Henderson NC, et al. Collagen-producing lung cell atlas identifies multiple subsets with distinct localization and relevance to fibrosis. Nat Commun 2020;11(1):1920. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Palumbo-Zerr K, Soare A, Zerr P, Liebl A, Mancuso R, Tomcik M, et al. Composition of TWIST1 dimers regulates fibroblast activation and tissue fibrosis. Ann Rheum Dis 2017;76(1):244–51. [DOI] [PubMed] [Google Scholar]
  • 60.Chen Q, Lee CE, Denard B, Ye J. Sustained induction of collagen synthesis by TGF-beta requires regulated intramembrane proteolysis of CREB3L1. PLoS One 2014;9(10):e108528. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Akhmetshina A, Palumbo K, Dees C, Bergmann C, Venalis P, Zerr P, et al. Activation of canonical Wnt signalling is required for TGF-beta-mediated fibrosis. Nat Commun 2012;3:735. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 62.Meng XM, Nikolic-Paterson DJ, Lan HY. TGF-beta: the master regulator of fibrosis. Nat Rev Nephrol 2016;12(6):325–38. [DOI] [PubMed] [Google Scholar]
  • 63.Fernandez IE, Eickelberg O. The impact of TGF-beta on lung fibrosis: from targeting to biomarkers. Proc Am Thorac Soc 2012;9(3):111–6. [DOI] [PubMed] [Google Scholar]
  • 64.Batlle E, Massague J. Transforming Growth Factor-beta Signaling in Immunity and Cancer. Immunity 2019;50(4):924–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 65.Stofkova A. Resistin and visfatin: regulators of insulin sensitivity, inflammation and immunity. Endocr Regul 2010;44(1):25–36. [DOI] [PubMed] [Google Scholar]
  • 66.Bui TM, Wiesolek HL, Sumagin R. ICAM-1: A master regulator of cellular responses in inflammation, injury resolution, and tumorigenesis. J Leukoc Biol 2020;108(3):787–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 67.Lazennec G, Richmond A. Chemokines and chemokine receptors: new insights into cancer-related inflammation. Trends Mol Med 2010;16(3):133–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 68.Krein PM, Winston BW. Roles for insulin-like growth factor I and transforming growth factor-beta in fibrotic lung disease. Chest 2002;122(6 Suppl):289S–93S. [DOI] [PubMed] [Google Scholar]
  • 69.Bonner JC. Regulation of PDGF and its receptors in fibrotic diseases. Cytokine Growth Factor Rev 2004;15(4):255–73. [DOI] [PubMed] [Google Scholar]
  • 70.Chaudhary NI, Roth GJ, Hilberg F, Muller-Quernheim J, Prasse A, Zissel G, et al. Inhibition of PDGF, VEGF and FGF signalling attenuates fibrosis. Eur Respir J 2007;29(5):976–85. [DOI] [PubMed] [Google Scholar]
  • 71.Witsch E, Sela M, Yarden Y. Roles for growth factors in cancer progression. Physiology (Bethesda) 2010;25(2):85–101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 72.Lamort AS, Giopanou I, Psallidas I, Stathopoulos GT. Osteopontin as a Link between Inflammation and Cancer: The Thorax in the Spotlight. Cells 2019;8(8). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 73.O’Regan A. The role of osteopontin in lung disease. Cytokine Growth Factor Rev 2003;14(6):479–88. [DOI] [PubMed] [Google Scholar]
  • 74.Choueiri TK, Labaki C, Bakouny Z, Hsu CY, Schmidt AL, de Lima Lopes G Jr., et al. Breakthrough SARS-CoV-2 infections among patients with cancer following two and three doses of COVID-19 mRNA vaccines: a retrospective observational study from the COVID-19 and Cancer Consortium. Lancet Reg Health Am 2023;19:100445. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1952501-supplement-1.pdf (713KB, pdf)
2
NIHMS1952501-supplement-2.pdf (141.3KB, pdf)
3
NIHMS1952501-supplement-3.pdf (721.9KB, pdf)
4
NIHMS1952501-supplement-4.pdf (602.8KB, pdf)
5
NIHMS1952501-supplement-5.pdf (508.6KB, pdf)
6
NIHMS1952501-supplement-6.pdf (198KB, pdf)
7
NIHMS1952501-supplement-7.xlsx (34.9KB, xlsx)

Data Availability Statement

The processed scRNA-seq data saved as a Seurat object and original code are publicly available at https://doi.org/10.5281/zenodo.8222832. The original data can be accessed from public databases described in “Data collection” section.

RESOURCES