Characterization of the pleural microenvironment niche and cancer transition using single-cell RNA sequencing in EGFR-mutated lung cancer

Abstract

Background: Lung cancer is associated with a high mortality rate and often complicated with malignant pleural effusion (MPE), which has a very poor clinical outcome with a short life expectancy. However, our understanding of cell-specific mechanisms underlying the pathobiology of pleural metastasis remains incomplete. Methods: We analyzed single-cell transcriptomes of cells in pleural effusion collected from patients with lung cancer and congestive heart failure (as a control), respectively. Soluble and complement factors were measured using a multiplex cytokine bead assay. The role of ferroptosis was evaluated by GPX4 small interfering RNA (siRNA) transfection and overexpression. Results: We found that the mesothelial-mesenchymal transition (MesoMT) of the pleural mesothelial cells contributed to pleural metastasis, which was validated by lung cancer/mesothelial cell co-culture experiments. The ferroptosis resistance that protected cancer from death which was secondary to extracellular matrix detachment was critical for pleural metastasis. We found a universal presence of immune-suppressive lipid-associated tumor-associated macrophages (LA-TAMs) with complement cascade alteration in the MPE of the lung cancer patients. Specifically, upregulated complement factors were also found in the MPE, and C5 was associated with poor overall survival in the lung cancer patients with epidermal growth factor receptor mutation. Plasmacytoid dendritic cells (pDCs) exhibited a dysfunctional phenotype and pro-tumorigenic feature in the primary cancer. High expression of the gene set extracted from pDCs was associated with a poor prognosis in the lung cancer patients. Receptor-ligand interaction analysis revealed that the pleural metastatic niche was aggravated by cross-talk between mesothelial cells-cancer cells/immune cells via TNC and ICAM1. Conclusions: Taken together, our results highlight cell-specific mechanisms involved in the pathobiological development of pleural metastasis in lung cancer. These results provide a large-scale and high-dimensional characterization of the pleural microenvironment and offer a useful resource for the future development of therapeutic drugs in lung cancer.

Keywords: Pleural metastasis; complement factor; ferroptosis; lung cancer; mesothelial cell.

Figure 1

Cell profiling in pleural fluid assessed by single-cell RNA-seq analysis. (A) Workflow depicting the collection and processing of primary tissue and pleural fluid from patients with congestive heart failure or with lung cancer for single-cell RNA-sequencing and further study. (B) Visualization of 15 cell types on the UMAP plot. (C) The heatmap of cell markers used for cluster annotation. (D) The proportions of all cell types in the normal lung (NL), primary lung cancer (PLC), pleural fluid of heart failure (HP) and lung cancer with pleural metastasis (LCP) groups. AT2, alveolar type II epithelial cells; DCs, dendritic cells; NK cells, natural killer cells.

Figure 2

Alterations in pleural mesothelial cells (PMCs) in pleural metastasis. (A) The volcano plot of differentially-expressed genes (DEGs) of PMCs obtained from the HP and LCP groups. (B) The IPA canonical pathways of DEGs. Dot plot showing the related expressions of mesenchymal (C), extracellular matrix (ECM) remodeling, MesoMT related factors, epithelial (D), and transcription factors regulating the MesoMT (E). (F) Lung cancer cells stimulated MesoMT of MeT-5A mesothelial cells. (G) The levels of tenascin C, VEGFA, fibronectin, ICAM-1, and PAI-1 in the pleural fluid obtained from the HP and LCP groups. (H) The levels of tenascin C and VEGFA in pleural fluid of wild type (WT) or mutated (MT) EGFR lung cancer patients. Data in Figure 2G and 2H are presented as mean ± SD. *p < 0.05, **p < 0.01, ****p < 0.0001.

Figure 3

Ferroptosis resistance contributed to pleural metastasis in lung cancer. (A) InferCNV profiles of pleural and primary cancer cells. (B) KEGG pathway of DEGs. (C) The upregulated expressions of GPX4 and FTL in pleural cancer cells. (D) IHC revealed higher levels of GPX4 protein (red arrow head) in pleural cancer than in primary cancer of a paired lung adenocarcinoma tissue. (E) The expressions of ferroptosis-related genes in A549, CL1-5, H1975 and HCC827 cells. (F) The viability of A549, CL1-5, H1975 and HCC827 cells in pleural fluid of pleural metastasis lung cancer patients. (G) MPE fluid increased expressions of GPX4, FTL, as well as NUPR1, and decreased the expression of ACSL4 in A549 and H1975 cells. (H) Inhibition of GPX4 decreased the viability of H1975 cells in MPE by siRNA. (I) Overexpression of GPX4 prevented cell death induced by MPE in HCC827 cells. Lung cancer cells were cultured in the pleural fluid of MPE for 48 h. The cell viability was determined by WST-1. The expressions of various genes were measured by qRT-PCR after 24 h incubation. Data are presented as mean ± SD. ns, not significant, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. MPE, malignant pleural effusion.

Figure 4

Lipid-associated tumor-associated macrophages (LA-TAMs) exerted an immunosuppressive phenotype in the pleural microenvironment. (A) UMAP plots of myeloid cells in the HP, LCP, NL, and PLC groups. (B) The percentage of different myeloid cell subsets. (C) The functional scores of monocytes/macrophages. (D) The phenotype of two TAM subsets. (E) KEGG pathway analysis of LA-TAMs. (F) Trajectory analysis of LA-TAMs. (G) Expressions of different transcriptional factors across monocytes and TAMs, ordered by Monocle 2 analysis in pseudotime. (H) CD68⁺APOE⁺ LA-TAMs presence and CD68⁺ZNF331⁺ LA-TAMs in pleural lung cancer. Arrowhead indicated LA-TAMs.

Figure 5

The levels of complement factors in fluid of MPE. (A) Scheme of the complement pathways. The concentrations of complement factors of (B) classical, (C) lectin pathway, and (D) alternative pathways. The levels of (E) C3 and (F) C5 -related factors. (G) The levels of complement factors H and I in pleural fluid. (H) The level of C5 in pleural fluid of patient with WT or MT EGFR. (I) C5 were associated with poor overall survival in lung cancer patients with MT EGFR. Various complement factors in pleural fluid of HP and LCP were determined by EMD Millipore's MILLIPLEX^® Complement Panel 1 and 2 Magnetic Bead Panel. HR, hazard ratio. Data are presented as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 6

Dysregulation of pDCs at the primary tumor site. (A) Visualization of six clusters of DCs on the UMAP plot. (B) Cell markers used to annotate DC clusters. (C) Cell populations of various DC subsets. (D) Functional scores of the DC subsets. (E) Expressions of pro-tumorigenic factors in the DC subsets. (F) The levels of HB-EGF, amphiregulin, CCL3, CXCL2, CXCL8, and IL-1β in the pleural fluid. (G) The impact of CXCL8 and IL-1β in the overall survival of lung cancer patients. (H) Trajectory analysis of DCs in pleural fluid inferred by Monocle 2 software. (I) Expression of ZNF331 across DC subsets. (J) The gene set of pDCs in PLC was associated with shorter overall survival in the lung cancer patients. Data are presented as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, MPE, malignant pleural effusion.

Figure 7

Cross-talk of mesothelial cells, cancer cells, and LA-TAMs in the pleural niche. (A) Heatmap showing the putative receptor-ligand interactions between mesothelial cells and immune cells. (B) Circular plot displayed the impact of mesothelial cells and cancer cells or DCs. Lines originated from the ligand and connected to its receptor as indicated by the arrowhead. (C) Interactions of pleural cancer cells with mesothelial cells. (D) Circular plot displayed the impact of pleural cancer cells on mesothelial cells. (E) Interactions of pleural cancer cells with IFN-TAMs. (F) Circular plot displays the impact of pleural cancer cells on immune cells. (G) Cell-cell interactions between LA-TAMs and cancer cells/mesothelial cells. (H) Circular plot of LA-TAMs and cancer/mesothelial cells axis. (I) Levels of galectin-9, resistin, and MIF in the pleural fluid. Data are presented as mean ± SD. **p < 0.01, ****p < 0.0001.