Single-cell RNA sequencing elucidates potential mechanisms of endothelial cells in lung region-specific repair and remodeling in combined pulmonary fibrosis and emphysema

Abstract

Background

Combined pulmonary fibrosis and emphysema (CPFE) is a severe and progressive lung disease with limited therapeutic options and poor prognosis. CPFE manifests as emphysema and fibrosis in different lung regions and is frequently associated with pulmonary hypertension. Pulmonary vascular endothelial cells may regulate these processes, but their role in CPFE remains unclear.

Methods

Single-cell RNA sequencing and multiplex immunohistochemistry were performed on upper/lower lung tissues from CPFE patients and healthy controls. Cellular heterogeneity, gene expression, and intercellular communication networks were analyzed.

Results

The upper lung of CPFE exhibited marked depletion and impaired function of endothelial cells, along with neutrophil accumulation and M1 macrophage-driven inflammation. Conversely, the lower lung of CPFE displayed higher proportions of ACKR1 + venous endothelial cells and enhanced angiogenesis. Endothelial cells demonstrated profibrotic signaling via THBS1-SDC4, PDGFB-PDGFRα, and JAG1-NOTCH3 signaling targeting epithelial cells, fibroblasts, and smooth muscle cells, respectively. These endothelial cells also exhibited strong interactions with macrophages through chemokines and adhesion molecules such as CCL14 and ICAM1, and expressed factors promoting M2 macrophage polarization. Multi-color staining confirmed the proximity of these endothelial cells with specific lung resident cells and immune cells, with clear spatial compartmentalization of endothelial cells with their target cell populations in the upper and lower lungs of CPFE.

Conclusions

This study provides a preliminary cellular characterization of CPFE, implicating endothelial cells as potential regulators of its regional pathology through angiocrine signaling and macrophage crosstalk. The findings provide novel mechanistic insights into the histopathological heterogeneity of CPFE and highlight these pathways as therapeutic targets.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12931-025-03460-x.

Introduction

Combined pulmonary fibrosis and emphysema (CPFE) is a chronic and progressive lung disease characterized by a history of heavy smoking, exercise-induced hypoxemia, upper lung-predominant emphysema, lower lung-predominant fibrosis, reduced vital capacity, and decreased diffusing capacity for carbon monoxide [1]. The prognosis and mortality of CPFE are significantly worse than those of pulmonary fibrosis or emphysema alone. CPFE is frequently complicated by pulmonary hypertension, which occurs in 47% of patients with CPFE, markedly worsens the prognosis and is associated with a 1-year mortality of 60% [2]. Currently, lung transplantation remains the only effective treatment option for CPFE. The pathogenic mechanisms underlying the coexistence of emphysema and pulmonary fibrosis remain poorly understood. Pulmonary fibrosis is characterized by fibroblast proliferation and excessive extracellular matrix (ECM) deposition [3], whereas emphysema involves increased matrix metalloproteinase activity [4], leading to ECM degradation. These seemingly paradoxical processes likely involve region-specific mechanisms in the lung, as well as a complex interplay of multiple cell types.

Recent evidence highlights the critical role of vascular endothelial cells in organ repair and regeneration across various diseases. Pulmonary endothelial cells have been categorized into six subtypes, including lymphatic endothelial cells, arterial endothelial cells, aerocytes, general capillary endothelial cells, pulmonary-venous endothelial cells, and systemic-venous endothelial cells [5]. ACKR1 is a specific marker exhibiting high expression in both venous endothelial cell types [6].​ Tissue-specific endothelial cells establish specialized vascular niches and secrete angiocrine factors, including growth factors, cytokines, chemokines, ECM components, and exosomes [7]. These angiocrine signals play crucial roles in maintaining tissue homeostasis and regulating repair processes, influencing nearby epithelial cells, vascular smooth muscle cells, and fibroblasts in the lung microenvironment [8]. Clinical and experimental data suggest that endothelial dysfunction and loss of endothelial cells contribute to the development of emphysema [9]. Endothelial dysfunction has been implicated in the early stages of chronic obstructive pulmonary disease (COPD) and in smokers, which results in impaired endothelium-dependent vasorelaxation and thickening of small arterioles. Exposure to cigarette smoke induces endothelial injury and dysfunction, leading to increased endothelial monolayer permeability in vitro and exacerbation of acute lung injury in vivo [10]. Interestingly, therapeutic strategies targeting endothelial cells have shown promise in experimental models, where intravenous administration of endothelial cells in mice with elastase-induced emphysema restored lung endothelial cell numbers and ameliorated the disease [11]. In idiopathic pulmonary fibrosis (IPF), non-fibrotic regions show elevated capillary density, whereas areas of active fibrosis exhibit significant vascular compression, particularly surrounding fibroblastic foci, which are devoid of blood vessels [12]. Maladaptive endothelial cells contribute to fibrosis by promoting excessive and dysregulated repair responses. For example, in pulmonary fibrosis, endothelial cells stimulate perivascular fibroblasts and secrete chemokines to recruit macrophages highly expressing tissue inhibitors of metalloproteinase, thereby driving fibrotic progression [13]. Intriguingly, the lung tissue of CPFE patients may exhibits concurrent pathological alterations, including endothelial cell injury and dysfunction, as well as endothelial cell-mediated repair and regenerative processes. This provides a unique model for investigating the context-specific roles of endothelial cells and their interactions with other cell types in tissue remodeling and repair, in a condition with coexisting fibrotic and emphysematous pathologies.

Given that these pathophysiological processes likely involve distinct cellular populations with diverse functional roles, investigations at single-cell resolution are imperative to dissect their functional heterogeneity, molecular signatures, and pathogenic contributions in CPFE. In this study, we used single-cell RNA sequencing to analyze issues from the upper and lower lung lobes of CPFE patients and healthy donors, in order to examine the regional differences between the two lung regions. We investigated the alterations in cellular composition and gene expression patterns, with a focus on communications between key cell populations. We identified important ligand-receptor pairs and validated them experimentally in tissue samples, uncovering the mechanisms through which endothelial cells regulate multiple cell types and determine the development of emphysema versus fibrosis.

Methods

Human lung tissue collection

The study protocol involving the collection of human specimens was approved by the Clinical Research Ethics Committee of China-Japan Friendship Hospital (Approval No. 2024-KY-039). Written informed consent were obtained from all participants prior to tissue collection. Fresh lung specimens were obtained from two CPFE patients undergoing lung transplantation and unused lungs from two healthy donors. Two tissue blocks (~ 2 cm³ each) were collected from each donor: one from the apex of the upper lung lobe and another from the basal region of the lower lobe, distal to the hilum. The tissues were maintained in DMEM medium at 4 °C and subsequently aseptically divided into three portions under sterile conditions. Approximately 1 cm³ was fixed in 4% paraformaldehyde, approximately 0.5 cm³ was allocated for single-cell RNA sequencing, and the remaining tissue was snap-frozen in liquid nitrogen. The study collected surplus tissues for lung transplantation from two male donors, aged 43 and 56 years, respectively. CPFE patient 1 was​​ a 66-year-old male, who had a 30 pack-year smoking history (cessation achieved 10 years prior) and presented with pulmonary hypertension, with pulmonary artery pressures of 56/22 mmHg (systolic/diastolic) and a mean pulmonary arterial pressure (mPAP) of 33.3 mmHg measured by catheterization.​​ CPFE patient 2 was​​ a 68-year-old male, who had a 32 pack-year smoking history (cessation achieved 11 years prior) and presented with pulmonary hypertension, with pulmonary artery pressures of 90/26 mmHg (systolic/diastolic) and a mean pulmonary arterial pressure (mPAP) of 47.3 mmHg measured by catheterization.

Tissue processing and single-cell sequencing

Tissues were processed following the 10X Genomics Cell Preparation Guide. Single-cell suspensions were loaded onto Chromium microfluidic chips with 3’ chemistry, and cells were barcoded using the 10X Chromium Controller (10X Genomics). Subsequently, RNA from the barcoded cells was reverse-transcribed, and sequencing libraries were constructed using the Chromium Single Cell 3’ v3 reagent kit (10X Genomics) according to the manufacturer’s protocol. Sequencing was conducted on an Illumina HiSeq 2000 platform following the standard operating procedure provided by Illumina.

Sequencing data processing

For all samples, raw sequencing data were processed using the 10X Genomics CellRanger pipeline (v 7.1.0) for alignment to the GRCh38 (human) and generation of gene count matrices. Subsequent analysis was carried out using Seurat [14] (v 4.4.0). Cells with nFeature_RNA < 500, nCount_RNA < 500 or mitochondrial gene percentage > 25% were identified as low-quality cells and removed. Doublet cells were filtered using DoubletFinder [15] (v 2.0.4). We used Seurat FindVariableFeatures to identify the top 5000 high variable genes, which were applied to scale and center expression matrix using Seurat ScaleData. Principal component analysis was performed with the first 50 principal components. Next, datasets were integrated with harmony [16] (v 1.2.3) using the RunPCA and RunHarmony functions with default parameters. Uniform manifold approximation and projection analysis was performed to illustrate cell clusters, which were calculated by the Louvain algorithm with a resolution 0.6. The Seurat modules of FindMarkers and FindAllMarkers were performed to find differentially expressed genes (DEGs) with |avg_log2FC|>0.5 and p_val_adj < 0.05. Cell identities were determined by comparing top DEGs of the cell clusters. To explore the functions of DEGs, clusterProfiler [17] (v 4.12.6) was used based on annotations from Gens Ontology (GO) and Kyoto Encyclopedia of Genes (KEGG). Significantly enriched pathways were identified using a p-value < 0.05 and a q-value < 0.2.

Cell-cell interaction analysis

To evaluate cell-cell communication and interactions, we used CellChat [18] (v 1.6.1), which integrates gene expression data with a database of known interactions between signaling ligands, receptors, and their cofactors (CellChatDB). The gene count matrices were used to construct a CellChat object via the createCellChat function, followed by standard preprocessing with default parameters for the analysis of individual datasets. Communications involving fewer than 10 cells were excluded from the analysis for statistical robustness. Network centrality scores for all inferred communication networks were calculated using the netAnalysis_computeCentrality function. Various functions, including rankNet, netVisual_heatmap, netAnalysis_signalingRole_scatter, netVisual_bubble, netVisual_individual, and netVisual_chord_gene, were used to generate visualizations of the cell-cell communication patterns identified.

M1 and M2 macrophage gene set scoring

To assess the polarization of macrophages into M1 and M2 phenotypes, we compiled gene sets for M1 and M2 macrophages based on previously published literature [19–21]. Genes associated with the M1 phenotype were assigned a score of + 1, while genes associated with the M2 phenotype were assigned a score of −1. The gene expression matrix for each cell was then analyzed using Imogimap [22] (v 0.0.0.9000) to compute a gene set score for each cell. A positive score (> 0) indicated an M1 macrophage phenotype, whereas a negative score (< 0) reflected an M2 macrophage phenotype.

Histopathological analysis

Fixed lung tissues were paraffin-embedded and cut into 5 μm-thick sections. Tissue sections were deparaffinized in xylene, rehydrated through a graded ethanol series, and stained with a hematoxylin and eosin kit (Servicebio G1005, Beijing China) or a Masson’s trichrome kit (Servicebio G1006, Beijing China), according to manufacturer’s protocols. Brightfield microscopic images were captured using a Leica DM6 microscope.

Immunohistochemical staining

Immunohistochemical (IHC) staining for α-smooth muscle actin (α-SMA) was performed to assess pulmonary vascular remodeling. Deparaffinized and rehydrated tissue sections underwent antigen retrieval in citrate buffer (pH 6.0). Sections were permeabilized with 0.1% Triton X-100 for 10 min and incubated with 3% hydrogen peroxide. After blocking with 5% goat serum for 1 h at room temperature, sections were incubated overnight at 4 °C with a primary anti-α-SMA antibody (Table S1). Subsequently, horseradish peroxidase-conjugated secondary antibodies (Table S1) were applied for 1 h at room temperature. Chromogenic detection was performed using 3-amino-9-ethylcarbazole (AEC) as a substrate. Brightfield images were acquired using a Leica DM6 microscope.

Immunofluorescence staining

Deparaffinized and rehydrated tissue sections underwent antigen retrieval in citrate buffer (pH 6.0). Sections were permeabilized with 0.1% Triton X-100 for 10 min and blocked with 5% goat serum for one hour at room temperature. Primary antibodies against CD31 and ACKR1 (Table S1) were applied and incubated overnight at 4 °C. After PBS washes, fluorophore-conjugated secondary antibodies (Table S1) were incubated for one hour at room temperature. The sections were mounted with 4’,6-diamidino-2-phenylindole (DAPI)-containing mounting medium. Microscopic images were captured on a Leica DM6 microscope.

Multiplex immunohistochemistry

Deparaffinized and rehydrated tissue sections were processed using a multiplex fluorescence immunohistochemistry kit (Alpha X, BXT35025011, Beijing, China). Sequential staining cycles included antigen retrieval, incubation with primary antibodies (Table S1) for 1 h at 37 °C, incubation with peroxidase-conjugated secondary antibodies (Table S1) 10 min at 37 °C, and incubation with fluorophore-conjugated tyramide (FITC/Cy3/Texas Red/Cy5) in tyramide signal amplification (TSA) solution for 5 min at room temperature. After all staining cycles, the sections were mounted with DAPI-containing mounting medium and imaged using a Leica Stellaris 5 confocal microscope.

Results

CPFE upper and lower lung lobes showed distinct cellular compositions

We profiled a total of 77,916 cells from the upper and lower lung lobes of two CPFE patients (39102 cells) and two healthy donors (38814 cells) (Fig. 1A). Based on canonical lineage markers, we annotated ten major lung cell types: epithelial cells, endothelial cells, fibroblasts, smooth muscle cells, pericyte, monocytes/macrophages, T/NK cells, neutrophils, mast cells, B/plasma cells. Key cell types, including monocytes, macrophages, neutrophils, epithelial cells, endothelial cells, and fibroblasts, were further subdivided into 26 distinct cell subtypes (Fig. 1B). Markers genes for the eight major cell types were identified, with the top three markers for each cell type shown in Fig. 1C.

Fig. 1.

Single-cell description and tissue morphology reveal difference between upper and lower lungs of CPFE A Schematic overview of process of tissue collection and single-cell RNA sequencing. The upper and the lower lung tissues of healthy donors (H-upper and H-lower) and CPFE patients (CPFE-upper and CPFE-lower) were collected from the edge of the lung apex and the distal margin at the base of the lung, respectively. B Uniform manifold approximation and projection (UMAP) representation of 77,916 cells from upper and lower lungs of two CPFE patients (CPFE1 and CPFE2) and two healthy donors (H1 and H2). Each dot represents a single cell and cells are labeled as discrete cell populations based on their gene expression pattern. C Bubble plot showing the expression levels and percentages of cell expressing canonical markers of the major cell types from the integrated data. D-E Microscopic images of D Masson staining and E α-smooth muscle actin (α-SMA) immunohistochemistry (IHC) staining of the upper and lower lungs of healthy donors and CPFE patients, showing airway remodeling and loss of blood vessels in the upper lobe, and interstitial fibrin deposition and medial thickening of arterioles in the lower lobe of CPFE lungs. All stained images are presented at a magnification of ×200. NK, natural killer; Macro, Macrophage; Fibro, Fibroblast; Endo, Endothelial cell; Epi, Epithelial cell; CAP2, alveolar aerocytes; CAP1, general capillary; AT, alveolar type 1 or 2 cells

While all major cell types were observed in both CPFE and healthy donor lungs, the proportions of various cell subtypes differed significantly between the upper and lower lungs, and between CPFE and healthy donors. Immune cell composition varied between CPFE and healthy donor lungs: the CPFE-upper lung contained a large number of activated neutrophils (C1), characterized by high expression of CCL3 and CCL4, and a substantial presence of transitional neutrophils (C2) (Fig S1A, S1E, S1I). Tissue-resident macrophages (TrMs), including alveolar and interstitial macrophages [23], exhibited distinct characteristics in CPFE lung tissues. The abundance of alveolar macrophages, specifically FABP4 + and LAMP + macrophages, was markedly decreased in CPFE and nearly absent in the CPFE-upper lung. Notably, SPP1 + macrophages were predominantly elevated in the CPFE-upper lung, while F13A1 + macrophages were exclusively increased in the CPFE-lower lung. Monocytes and CCL17 + monocyte-derived macrophages were unevenly distributed between the upper and lower lungs, with monocytes predominantly concentrated in the CPFE-upper lung and CCL17 + macrophages showing an increased proportion in the CPFE-lower lung (Fig S1B, S1F, S1J). In CPFE lungs, the number and proportion of alveolar epithelial cells (AT1 and AT2) were markedly reduced compared to healthy donor lungs, while airway epithelial cells, such as CFAP299 + ciliated cells and club cells, were increased (Fig S1C, S1G, S1K). Fibroblasts also exhibited differential distribution, with the CPFE-upper lung showing a reduction in fibroblast numbers, while the CPFE-lower lung demonstrated a marked increase, including subsets such as adventitial fibroblasts and myofibroblasts. Smooth muscle cells were nearly absent in the CPFE-upper lung, while their abundance was higher in CPFE-lower lung compared to healthy lower lung tissues (Fig S1D, S1H, S1L).

Consistent with the variations in cellular composition, histopathological analysis revealed profound differences in tissue morphology. Hematoxylin and eosin staining (Fig S1M) and Masson’s trichrome staining (Fig. 1D) revealed pronounced inflammatory cell infiltration and tissue remodeling in CPFE lungs. In the CPFE-upper lung, inflammatory cells infiltrated small airways and dilated alveoli, while in the CPFE-lower lung, they accumulated around fibrotic regions. The CPFE-upper lung displayed small airway remodeling and muscularization of airway walls, whilst the CPFE-lower lung exhibited interstitial fibrosis and collagen fiber deposition. Immunohistochemical staining for α-smooth muscle actin (α-SMA) further demonstrated medial thickening of arterioles in the CPFE-lower lung (Fig. 1E). These histopathological results confirmed the characteristic features of emphysema in the upper lobe, pulmonary fibrosis in the lower lobe, and vascular remodeling in CPFE lungs.

Endothelial cells and their subtypes were altered in CPFE

Marked differences were observed in the number and transcription of endothelial cells in CPFE. Compared to the H-upper lungs, the CPFE-upper lung exhibited a decrease in endothelial cell proportion, whereas the CPFE-lower lung showed no apparent differences in endothelial cell abundance compared with the H-lower lungs (Fig. 2A). Given the pivotal role of endothelial cells in emphysema, pulmonary fibrosis, and pulmonary hypertension, we examined differential gene expression profiles in endothelial cells between the CPFE-upper and -lower lungs. Our analysis identified 392 genes highly expressed in endothelial cells of the CPFE-upper lung and 181 genes in endothelial cells of the CPFE-lower lung (Fig. 2B). To elucidate the functional roles of these genes, we performed GO (Fig. 2C, Table S2A, S2B) and KEGG (Fig S2A, S2B, Table S2C, S2D) pathway enrichment analyses. In the CPFE-upper lung, genes highly expressed in endothelial cells were enriched in pathways including cytoplasmic translation, ribosome, responses to tumor necrosis factor, and neutrophil chemotaxis. Enriched pathways in endothelial cells of the CPFE-lower lung included focal adhesion, regulation of angiogenesis, epithelial cell development, and endothelium development. These findings indicated a range of transcriptional differences in endothelial cells between the CPFE-upper and -lower lungs.

Fig. 2.

Marked differences in the number and function of endothelial cells between the upper and lower lungs in CPFE A Proportions of major cell types as a percentage of all cells in the upper and lower lobes of CPFE and healthy donor lungs. B Volcano plot showing the differentially expressed genes (DEGs) of endothelial cells between the upper and lower lobes of CPFE lung. C GO enrichment analysis of highly expressed genes in endothelial cells of the CPFE-upper and CPFE-lower lungs. D Bubble plot showing the interaction strength of pro-angiogenesis- and anti-angiogenesis-related ligand-receptor pairs between endothelial cells and other cell types. E UMAP embedding of endothelial cells annotated by cell subtypes from jointly analyzed data of CPFE and healthy lungs (left) and visualized separately (right). F Proportions of cell subtypes as a percentage of all endothelial cells in the upper and the lower lobes of CPFE and healthy lungs. G Heatmap of marker gene expression for the five endothelial cell subtypes. H Merged channels microscopic images of multiplex immunofluorescence staining of CD31 (green, an endothelial marker), ACKR1 (red, a venous endothelial marker) and DAPI (blue, nuclear stain) in the upper and lower lobes of CPFE and healthy lungs, venules in the CPFE-lower lungs showed CD31 and ACKR1 expression in their vessel wall (200 × magnification); The inset shows a magnified view of the blood vessels (400 × magnification). I Quantification of the proportion of ACKR1-positive vessels among CD31-positive vessels

To investigate intercellular interactions, we analyzed ligand-receptor pairings between endothelial cells and other cell types in the CPFE-upper and -lower lungs. In the CPFE-upper lung, The transcriptional levels of CXC chemokines (CXCL10 and CXCL11) were upregulated in endothelial cells. These chemokines are known to facilitate lymphocyte migration while inhibiting angiogenesis [24]. In the CPFE-lower lung, interactions between endothelial cells and other cell types showed a clear pattern of pro-angiogenic signaling, as evidenced by VEGF, ANGPT2, and CXCL2/3/8 signaling (Fig. 2D). Collectively, endothelial cells were reduced and exhibited anti-angiogenic signaling in the CPFE-upper lung, while showing pro-angiogenic signaling in the CPFE-lower lung.

To further characterize endothelial cells, we classified them into five subtypes: general capillary (CAP1), aerocyte capillary (CAP2), venous endothelial cells, arterial endothelial cells, and lymphatic endothelial cells [5, 25]. In the CPFE-upper lung, the numbers of CAP2, CAP1, and arterial endothelial cells were markedly reduced compared to the H-upper lungs. Conversely, in the CPFE-lower lung, both capillary and arterial endothelial cells were relatively diminished, while venous endothelial cells showed a remarkable increase (Fig. 2E and F). Figure 2G shows the unique gene expression profiles for each of these five endothelial subtypes. To validate these findings, we conducted immunofluorescence staining using CD31 as a pan-endothelial marker and ACKR1 as a venous endothelial cell marker [6]. Results revealed CD31 and ACKR1 expression in venous blood vessels of healthy lung tissue. The CPFE-upper lung exhibited decreased vascular density in bullous regions, yet numerous small blood vessels were observed surrounding remodeled and thickened small airways, but these small blood vessels did not express ACKR1. In the CPFE-lower lung, abundant microvessels were found encircling fibrotic regions. These vessels, smaller in diameter than typical veins, exhibited CD31 and ACKR1 expression in their endothelial cells (Fig. 2H and I, Fig S2I).

To investigate the functional alterations of venous endothelial cells, we compared the differentially expressed genes (DEGs) in venous endothelial cells between healthy and CPFE lung tissues (Fig S2C, S2D), and performed GO (Fig S2E, S2F, Table S3A, S3B) and KEGG (Fig S2G, S2H, Table S3C, S3D) pathway enrichment analyses. In the CPFE-upper lung, venous endothelial cells showed significant enrichment in pathways related to metabolism, protein translation, and cell chemotaxis compared to those from healthy donor lungs. In contrast, venous endothelial cells from the CPFE-lower lung were enriched in pathways associated with cell adhesion, ECM organization, leukocyte proliferation and activation, and angiogenesis. Interestingly, the transcriptomic profile of venous endothelial cells in the CPFE-upper and -lower lungs closely resembled the overall endothelial cells, suggesting that venous endothelial cells might contribute substantially to the endothelial cell-related processes in CPFE.

Endothelial cells exerted markedly different effects on epithelial, fibroblast and smooth muscle cells in CPFE-upper and -lower lungs

To elucidate how endothelial cells regulate lung resident cell types involved in pulmonary emphysema, fibrosis, and hypertension, we examined their interactions with epithelial cells, fibroblasts, and smooth muscle cells in the CPFE-upper and -lower lungs. In the CPFE-upper lung, compared with endothelial cells, interactions between airway epithelial cells, such as club cells, basal cells, and ciliated cells and other cells were prominent, whereas the CPFE-lower lung showed strong interactions between venous endothelial cells with fibroblasts, and club cells. Among these, venous endothelial cells in the CPFE-lower lung displayed markedly higher signaling activity compared to all other cell types, which was not observed in the CPFE-upper lung (Fig. 3A-D). Notably, the relative information flow plot indicates that adhesion, angiogenesis, and fibroblast proliferation pathways, illustrated by PECAM1, VCAM, VEGF, ANGPT, and FGF signaling, were pronounced among the enhanced intercellular communication within the CPFE-lower lung (Fig S3A).

Fig. 3.

Intercellular interaction analysis reveals reparative and proliferative effects exerted by endothelial cells on other cell types in the CPFE-lower lung A Heatmap showing the interaction strength between endothelial cells and epithelial cells, smooth muscle cells and fibroblasts in the CPFE-upper lungs compared to H-upper lungs. Red indicates stronger interactions. B Bubble plot illustrating the outgoing and incoming signal strength of each cell subtype in the interactions between endothelial cells and epithelial cells, smooth muscle cells, and fibroblasts in the CPFE-upper lungs. C Heatmap showing the interaction strength between endothelial cells and epithelial cells, smooth muscle cells and fibroblasts in the CPFE-lower lungs compared to H-lower lungs. Red indicates stronger interactions. D Bubble plot illustrating the outgoing and incoming signal strength of each cell subtype in the interactions between endothelial cells and epithelial cells, smooth muscle cells, and fibroblasts in the CPFE-lower lungs. E Bubble plot showing the interaction strength of ligand-receptor pairs between endothelial cells and epithelial cells. F Bubble plot showing the interaction strength of ligand-receptor pairs between endothelial cells and smooth muscle cells. G Bubble plot showing the interaction strength of ligand-receptor pairs between endothelial cells and fibroblasts

In the CPFE-lower lung, endothelial cells, particularly venous endothelial cells, exhibited increased expression of THBS1, which interacted with epithelial cells through SDC4. This interaction was notably stronger than in the CPFE-upper lung (Fig. 3E). The chord diagram illustrates THBS1-SDC4 signaling among endothelial cells, epithelial cells, fibroblasts, and smooth muscle cells (Fig. 4A). In contrast to the CPFE-upper lung, where this pathway was primarily found between lipo-fibroblasts and epithelial cells, the CPFE-lower lung maybe enhanced interaction strength. Venous endothelial cells, smooth muscle cells, and fibroblasts showed a clear increase in outgoing signals, indicating a more active role in mediating intercellular communication via this pathway (Fig. 4A). To validate these findings, multiplex immunohistochemistry (mIHC) was performed to examine the distinct spatial expression patterns of THBS1 between CPFE-upper and -lower lungs. In the CPFE-upper lung, THBS1 expression was predominantly observed in epithelial cells, with minimal expression in endothelial cells. Conversely, in the CPFE-lower lung, THBS1 showed marked endothelial-specific localization, as evidenced by co-staining with CD31, while epithelial expression was minimal or absent (Fig. 4B and C, S3C), indicating clear differences in endothelial-epithelial signaling.

Fig. 4.

Endothelial cells exhibited enhanced THBS1-, PDGFB-, and JAG1-mediated interactions with epithelial cells, fibroblasts, and smooth muscle cells in the CPFE-lower lung A Cell communication analysis of THBS1-SDC4 signaling among endothelial cells and other cell types in the CPFE-upper and CPFE-lower lung. B Merged channel microscopic images of multi-color fluorescence immunohistochemistry (mIHC) staining of THBS1 (green, a ligand expressed in endothelial cells), SDC4 (red, a receptor for THBS1 expressed in epithelial cells), CD31 (cyan, an endothelial marker), SFTPB (magenta, an epithelial marker) and DAPI (blue, nuclear stain) in the upper and lower lobes of CPFE and healthy lungs (200 × magnification); The inset shows a magnified view of the blood vessels (400 × magnification). C Quantification of the proportion of THBS1-positive endothelial cells among CD31-positive endothelial cells. D Cell communication analysis of PDGFB-PDGFRA signaling among endothelial cells and other cell types in the CPFE-upper and CPFE-lower lungs. E Merged channel microscopic images of mIHC staining of PDGFB (green, a ligand expressed in endothelial cells), PDGFRA (red, a receptor for PDGFB and a marker of fibroblasts), CD31 (cyan) and DAPI (blue) in the upper and lower lobes of CPFE and healthy lungs (200 × magnification); The inset shows a magnified view of the region positive for PDGFB and PDGFRa (400 × magnification). F Quantification of PDGFB and PDGFRa-positive fluorescent signals. G Cell communication analysis of JAG1-NOTCH3 signaling among endothelial cells and other cell types in the CPFE-upper and CPFE-lower lungs. H Merged channel microscopic images of mIHC staining of JAG1 (green, a ligand expressed in endothelial cells), NOTCH3 (red, a receptor of JAG1 expressed in smooth muscle cells), CD31 (cyan), α-SMA (magenta, a smooth muscle cell marker) and DAPI (blue) in the upper and lower lobes of CPFE and healthy lungs (200 × magnification); The inset shows a magnified view of the region positive for both JAG1 and NOTCH3 (400 × magnification). I Quantification of JAG1 and NOTCH3 dual positive cells

In the CPFE-lower lung, endothelial cells showed may enhanced interactions with fibroblasts via PDGFB-PDGFRα signaling (Fig. 3F), while this ligand-receptor pair was minimally expressed in the CPFE-upper lung. In the CPFE-lower lung, PDGFB was predominantly derived from CAP1 cells and targeted various fibroblast populations, suggesting a shift in the cellular sources and recipients of this pathway compared to the CPFE-upper lung (Fig. 4D). mIHC staining revealed high expression of PDGFRα in fibroblasts within fibrotic regions of the CPFE-lower lung, while endothelial cells in these regions exhibited high PDGFB expression, as evidenced by co-staining with CD31. In contrast, endothelial cells in the CPFE-upper lung showed less PDGFB expression and were not found near PDGFRα + fibroblasts (Fig. 4E and F, S3D).

Notably, endothelial cell, particularly venous endothelial cells, showed maybe more interactions with smooth muscle cells in the CPFE-lower lung compared to the CPFE-upper lung, with a marked increase in JAG1-NOTCH3 ligand-receptor interactions (Fig. 3G). In the CPFE-upper lung, the interaction strength of JAG1-NOTCH3 was relatively weak and primarily restricted to epithelial-epithelial cell communication. However, in the CPFE-lower lung, JAG1 was predominantly derived from venous endothelial cells, CFAP299 + ciliated epithelial cells, and smooth muscle cells, with smooth muscle cells as the primary target, indicating a shift in signaling dynamics (Fig. 4G). mIHC confirmed the alterations in JAG-NOTCH signaling between endothelial cells and smooth muscle cells. In the CPFE-upper lung, JAG1 was primarily found in cells within alveolar spaces. In contrast, in the fibrotic regions of the CPFE-lower lung, JAG1 expression was prominent on endothelial cells, as indicated by co-staining with CD31. Furthermore, JAG1-expressing endothelial cells and NOTCH3-expressing smooth muscle cells were found in close proximity within fibrotic regions and around pulmonary arterioles of the CPFE-lower lung, suggesting strengthened ligand-receptor interactions in these areas (Fig. 4H and I, S3E).

Endothelial cells secreted factors to recruit macrophages and other immune cells

To explore the potential influence of endothelial cells on the distribution and function of immune cells, we analyzed cell-cell interactions between endothelial cells, myeloid cells and neutrophils. In the CPFE-upper lung, monocytes, SPP1 + macrophages and neutrophils were the most active partners of cellular interactions with endothelial cells. In contrast, in the CPFE-lower lung, interactions between FABP4 + macrophages and venous endothelial cells maybe more prominent. (Fig. 5A-D). The relative information flow diagram illustrates in the CPFE-upper lung, interactions were primarily mediated by molecules such as SPP1, IL6, TNF, and CD80. In contrast, in the CPFE-lower lung, EGF, FGF, PECAM1, WNT, and VCAM signaling predominated, indicating a potential shift in intercellular communication dynamics between the two regions (Fig S4A).

Fig. 5.

Endothelial cells in the CPFE-lower lung exerted chemotactic, adhesive, and M2-polarizing effects on macrophages A Heatmap showing the interaction strength between endothelial cells, myeloid cells and neutrophils in the CPFE-upper lungs compared to H-upper lungs. Red indicates stronger interactions. B Bubble plot illustrating the outgoing and incoming signal strength of each cell subtype in the interactions between endothelial cells, myeloid cells and neutrophils in the CPFE-upper lungs. C Heatmap showing the interaction strength between endothelial cells, myeloid cells and neutrophils in the CPFE-lower lungs compared to H-lower lungs. Red indicates stronger interactions. D Bubble plot illustrating the outgoing and incoming signal strength of each cell subtype in the interactions between endothelial cells, myeloid cells and neutrophils in the CPFE-lower lungs. E Cell communication analysis of CCL14-associated chemotactic signaling among endothelial cells and myeloid cells in the CPFE-upper and CPFE-lower lungs. F Cell communication analysis of ICAM1-associated adhesion signaling among endothelial cells and myeloid cells in the CPFE-upper and CPFE-lower lungs. G Cell communication analysis of MIF-associated M2-polarizing signaling among endothelial cells and myeloid cells in the CPFE-upper and CPFE-lower lungs

Ligand-receptor pair analysis revealed important interactions may mediated by chemokines, adhesion molecules and molecules targeting macrophages. Among chemokine-related ligand-receptor interactions, CXCL12-CXCR4 signaling was predominantly observed between arterial endothelial cells and immune cells in the CPFE-lower lung. Venous endothelial cells in the CPFE-lower lung also exhibited stronger chemotactic interactions with myeloid cells via CCL14-CCR1 and CCL23-CCR1 compared to the CPFE-upper lung (Fig S4B). More specifically, CCL14 signaling from venous endothelial cells was enhanced in its reception by FABP4 + macrophages in the CPFE-lower lung (Fig. 5E). Regarding adhesion molecules, endothelial cells in the CPFE-lower lung showed higher expression levels of SELE, PECAM1, JAM2, ICAM1 and CD99, potentially facilitating immune cell adhesion. Among these, interactions between venous endothelial cells and FABP4 + or SPP1 + macrophages were the most robust (Fig S4C). Compared to the CPFE-upper lung, venous endothelial cell-derived ICAM1 signaling in the CPFE-lower lung suggested interactions with more macrophage subsets through a wider array of adhesion receptors, including SPN, ITGAL, and related molecules (Fig. 5F). Furthermore, venous endothelial cells in the CPFE-lower lung expressed higher levels of bioactive factors targeting macrophages, including GAS6, GRN, MDK, MIF, NAMPT, POSTN, and TGFB1 (Fig S4D). Compared to the CPFE-upper lung, MIF signaling originating from venous endothelial cells in the CPFE-lower lung demonstrated enhanced reception by a broader spectrum of macrophage subsets through interaction with receptors such as ACKR3 (Fig. 5G). These ligands are known to promote M2 polarization of macrophages, which is associated with tissue remodeling and fibrosis [26–32]. Taken together, these results indicated that the distinct transcriptional profiles of genes associated with chemotaxis, adhesion, and polarization in endothelial cells from the upper and lower lobes of CPFE lungs might be associated with differences in the recruitment and activation of immune cells. This could potentially shape the local immune microenvironment and contribute to divergent disease manifestations.

In the CPFE-upper lung, where endothelial cell proportions and specific subtypes were reduced, signaling interactions with immune cells were primarily dominated by epithelial cells (Fig S5A-D). Epithelial cells exhibited chemotactic signaling via CCL3/4-CCR5 (with monocytes/SPP1 + macrophages) and C3-C3AR1 (with neutrophils/myeloid cells) (Fig S5E), alongside adhesion interactions via ICAM1 between epithelial cells and myeloid cells (Fig S5F). These enhanced epithelial-immune cell interactions likely drove the accumulation of monocytes, SPP1 + macrophages, and neutrophils, shaping the distinct immune microenvironment in the CPFE-upper lung.

Distinct macrophage phenotype in CPFE-upper and -lower lungs were related to emphysema and fibrosis

Given that endothelial cells played a pronounced role in regulating macrophage recruitment and activation, macrophage polarization state was assessed using a score based on the expression levels of characteristic genes. Higher scores indicated a phenotype closer to M1 macrophages and lower scores suggested a phenotype resembling M2 macrophages. The scores of healthy donor lungs were near a neutral value of 0, while macrophages in the CPFE-upper lung exhibited significantly higher scores compared to those in the CPFE-lower lung (Fig. 6A). UMAP visualization of macrophage polarization scores revealed that almost all macrophages in the CPFE-upper lung were likely to be the M1 phenotype, whereas both M1 and M2 macrophages were present in the CPFE-lower lung, with a predominance of M2 macrophages (Fig. 6B). To further investigate their functional differences, 1355 DEGs between macrophages from the CPFE-upper and -lower lungs were analyzed. Macrophages in the upper lung exhibited 749 upregulated and 606 downregulated genes compared to the lower lung (Fig. 6C). GO pathway enrichment analysis of these DEGs revealed that, in the upper lung, macrophage-expressed genes were enriched in several inflammatory pathways, including response to lipopolysaccharide, response to bacterial molecules, regulation of innate immune response, response to tumor necrosis factor, canonical NF-κB signaling, response to type II interferon, response to interleukin-1 and granulocyte migration. In contrast, DEGs in the lower lung were primarily associated with macrophage phagocytosis, including pathways such as cytoplasmic vesicle lumen, endocytic vesicle, MHC class II protein complex binding, lysosomal membrane, and phagocytosis (Fig. 6D, Table S4A, S4B). These findings suggested distinct roles of macrophages, with those in the CPFE-upper lung involved in acute inflammatory responses, and those in the lower lung participating in fibrotic processes through phagocytic activity and tissue remodeling.

Fig. 6.

Macrophages exhibited phenotypic and functional differences in the upper and lower lungs of CPFE A Violin plot showing the polarization scores of macrophages in CPFE lungs and healthy lungs. Higher scores indicate a gene expression profile closer to M1 macrophages, whereas lower scores suggest a gene expression profile closer to M2 macrophages. B UMAP plots showing the polarization scores of individual macrophage subtypes in the CPFE-upper and CPFE-lower lungs. Red indicates a gene expression profile closer to M1 macrophages, while blue indicates a profile closer to M2 macrophages. C Volcano plot showing the DEGs of macrophages between the CPFE-upper and CPFE-lower lungs. D GO enrichment analysis of genes highly expressed in macrophages in the CPFE-upper and CPFE-lower lungs. E t-distributed stochastic neighbor embedding (tSNE) plots showing the expression of the M1 macrophage marker gene CD80 in the CPFE-upper and CPFE-lower lungs. F tSNE plots showing the expression of the M2 macrophage marker genes CD163 in the CPFE-upper and CPFE-lower lungs. G Merge channels of mIHC staining for CD80 (green, an M1 macrophage marker), CD163 (red, an M2 macrophage marker) and CD68 (magenta, a macrophage marker) in the upper and lower lobes of CPFE and healthy lungs. The alveolar spaces in the CPFE-upper lung are extensively infiltrated by M1 macrophages, whereas M2 macrophages predominate in the lung interstitium of the CPFE-lower lung. (200 × magnification). The inset shows a magnified view of the macrophages (400 × magnification). H Quantification of the proportion of the CD80-positive macrophages. G Quantification of the proportion of the CD163-positive macrophages

To validate the macrophage polarization state in CPFE, we evaluated the expression levels of CD80, a marker for M1 macrophages, and CD163, a marker for M2 macrophages, across all macrophage subtypes. Macrophages in the CPFE-upper lung exhibited higher expression of CD80 compared to the CPFE-lower lung (Fig. 6E). In contrast, macrophages in the CPFE-lower lung demonstrated higher levels of CD163 expression than those in the upper lung (Fig. 6F). Consistently, mIHC staining revealed that, in the CPFE-upper lung, enlarged alveolar spaces of emphysema were infiltrated by numerous inflammatory cells, which were predominantly macrophages expressing CD80, consistent with an M1 phenotype. In contrast, in the CPFE-lower lung, macrophages were densely distributed within the lung interstitium and predominantly expressed CD163, with minimal to no expression of CD80, suggesting an M2 phenotype (Fig. 6G, H and I, S6A). These findings further reinforced the differential macrophage polarization between the upper and lower lungs in CPFE, with M1 macrophages predominating in the upper lobe and M2 macrophages in the lower lobe.

To further explore the impact of macrophage phenotypes on the development of emphysema and pulmonary fibrosis, we analyzed cellular interactions between macrophages and other cell types (Fig S6B, S6D). In the CPFE-upper lung, monocytes, SPP1 + macrophages, and ciliated epithelial cells were active in both sending and receiving signals. In contrast, fibroblasts emerged as prominent signal senders in the CPFE-lower lung, along with FABP4 + and SPP1 + macrophages (Fig S6C, S6E). In the CPFE-upper lung, interactions between SPP1 + macrophages and monocytes with airway epithelial cells via EREG/HBEGF-ERBB signaling were notably stronger than the CPFE-lower lung (Fig S6F). Compared to the CPFE-upper lung, SPP1 + macrophages in the CPFE-lower lung showed elevated expression of SPP1, which specifically targeted fibroblasts (Fig S6G). Macrophages from the CPFE-lower lung also expressed the profibrotic growth factor GRN, which interacted with the SORT1 receptor on smooth muscle cells, a pattern not observed in the CPFE-upper lung (Fig S6H). Collectively, these findings indicated that macrophage-mediated signaling in the CPFE-upper and -lower lungs differed substantially, contributing to the region-specific processes of inflammation and fibrosis in the lungs.

Discussion

In this study, we present the first single-cell transcriptomic profile of lung tissue of CPFE, offering novel insights into the role of endothelial cells in the pathogenesis of pulmonary emphysema and fibrosis. We uncovered distinct functional and transcriptional heterogeneity in endothelial cells between the CPFE-upper and -lower lungs, highlighting their differential functions and interactions with other cell types. In the CPFE-upper lung, endothelial cells showed marked reductions in number, along with impaired functions and intercellular communications. Conversely, in the CPFE-lower lung, endothelial cells retained their functional integrity, possibly promoting angiogenesis and tissue repair, and regulated a range of resident and immune cells. These findings reveal a previously unappreciated mechanism by which endothelial cells orchestrate multiple cell types and modulate the immune microenvironment, offering potential therapeutic targets for pulmonary diseases characterized by tissue damage, remodeling, or excessive repair (Fig. 7).

Fig. 7.

Cellular alterations in the upper and lower lungs of CPFE and the proposed underlying mechanisms Significant differences in both the number and function of endothelial cells were observed between the upper and lower lobes of CPFE lungs. In the upper lung, endothelial cells were reduced in number and displayed impaired function, possibly due to damage caused by smoking and microbes. This led to weakened communication with other cell types, potentially impairing damage repair and contributing emphysema. In contrast, the lower lung showed an increase in venous endothelial cell numbers and enhanced activity. These endothelial cells recruited and adhered with a large number of macrophages and lymphocytes, altering the immune microenvironment. Additionally, communications between endothelial cells and epithelial cells, fibroblasts, and smooth muscle cells were enhanced, promoting excessive tissue repair and proliferation, potentially leading to fibrosis

In this study, we examined the tissue morphology and cellular compositions of different anatomical regions in CPFE. In humans, the lungs exhibit significant regional heterogeneity in perfusion, ventilation, lymphatic flow, metabolism, and mechanics due to factors such as embryonic development and gravity [33]. This regional heterogeneity ​​provides a pathological basis for the development of different lung pathologies, and their coexistence in the same individual​​. In our study, the CPFE-upper lung exhibited enlargement of distal airspaces, alveolar fusion, and extensive infiltration of inflammatory cells, consistent with the characteristic features of emphysema and COPD [34, 35]. In contrast, the CPFE-lower lung showed structural remodeling with honeycomb fibrosis, dense scar margins featuring fibroblastic foci, interstitial fibrin deposition, along with a higher proportion of fibroblasts and a lower proportion of alveolar epithelial cells, aligning with the histopathological patterns of fibrosis or IPF [36, 37]. To elucidate the cellular landscape and underlying mechanisms, we performed single-cell RNA sequencing on these key anatomical regions.

By comparing the cellular compositions of the CPFE-upper and -lower lungs, we found striking differences in the abundance and proportions of vascular endothelial cells. These findings were consistent with the loss of vasculature in the upper lung and the supportive role of endothelial cells in tissue repair in the lower lung. Among the five endothelial cell subtypes, venous endothelial cells exhibited the most distinctive quantitative and functional alterations between the upper and lower lung regions. These cells, marked by ACKR1, have been reported in fibrotic diseases. For instance, in liver fibrosis, increased numbers of ACKR1 + endothelial cells are found within fibrotic regions. They express profibrotic genes (e.g., PDGFD, PDGFB, LOX, LOXL2) and exhibit an immunoregulatory phenotype, which promotes leukocyte recruitment and facilitates crosstalk with stromal cells, leading to myofibroblast activation [6, 38]. In addition, a subset of tumor endothelial cells expressing ACKR1 has been reported to contribute to angiogenesis, modulate immune cells, and promote ECM production, highlighting their multifaceted roles in fibrosis [39]. Taken together, ACKR1 + venous endothelial cells in the CPFE-lower lung likely played a crucial role in leukocyte recruitment and fibrosis development.

Endothelial cells are increasingly recognized to serve as a dynamic and instructive niche for maintaining tissue homeostasis and directing regeneration, where they coordinate other cells using angiocrine factors [7]. In the CPFE-lower lung, we observed that endothelial cells highly expressed THBS1, PDGFB, and JAG1 transcripts, which could target epithelial cells, fibroblasts, and smooth muscle cells, respectively. ​​THBS1 encodes thrombospondin-1, which regulates angiogenesis, contributes to wound healing and fibrosis by activating latent TGF-β, and promoting epithelial migration and matrix deposition [4, 40]. PDGFB​​ is a well-established mitogen for fibroblasts, ​​playing a central role in myofibroblast expansion and ECM deposition​​​​, which distorts alveolar architecture and impairs gas exchange [41]. Notably, Nintedanib, an antifibrotic drug for IPF, inhibits the PDGF receptor [42]. Regarding endothelial-smooth muscle interaction, JAG1 is a NOTCH ligand ​​critical for vascular development and smooth muscle cell differentiation/maturation [43, 44]. Collectively, these findings reveal divergent expression profiles of endothelial-derived factors between the CPFE-upper and -lower lungs. Given their previously documented in fibrosis and regeneration, our results offer potential therapeutic targets for this complex disease.

In addition to regulating lung resident cells, endothelial cells also exerted pivotal effects on immune cells, particularly in regulating monocyte-macrophage populations through chemokines and adhesion molecules. Furthermore, in the CPFE-lower lung, endothelial cells expressed several factors, including GAS6, GRN, MDK, MIF, NAMPT, POSTN, and TGFB1, which acted on macrophages to promote their activation and polarization. GAS6 promotes M2 macrophage polarization by activating tyrosine kinase receptors [45], GRN regulates M2 macrophage polarization through TNFR2 [27], and MDK promotes M2 polarization by interacting with LRP1 [28]. Recently, NAMPT has been reported to drive M2 polarization in macrophages, which is implicated in bleomycin-induced pulmonary fibrosis [32]. Additionally, MIF [46], TGFB1 [31], and POSTN [47] are well-documented regulators of M2 macrophage polarization. Consistently, macrophages in the CPFE-upper lung exhibited a predominantly M1 phenotype, while those in the CPFE-lower lung displayed a mixed M1/M2 phenotype. Macrophages critically contribute to both emphysema and fibrosis:​​ in COPD, they promote inflammation and tissue damage, while in IPF, M1 macrophages drive early inflammation and M2 macrophages promote fibrosis [29, 48]. Furthermore, in the CPFE-upper lung, where endothelial cells exhibited reduced abundance and function, airway epithelial cells predominated in interactions with immune cells. These combined effects of epithelial and endothelial cell-mediated chemotaxis and adhesion may provide a potential explanation for the immune cells recruited, contributing to the formation of distinct immune microenvironments and the subsequent histopathological changes.

In practice, this spatial heterogeneity in pathological distribution of CPFE creates treatment dilemmas. Monotherapy targeting emphysema or fibrosis may inadvertently exacerbate disease progression due to the conflicting pathological processes. For instance, bronchodilators for emphysema could worsen ventilation-perfusion mismatch in fibrotic regions, whereas anti-fibrotic agents are ineffective against emphysematous destruction [49]. More importantly, bronchodilators and antifibrotic drugs tend to lack efficacy against pathological alterations in the pulmonary vasculature, which not only drive pulmonary hypertension, but also play a fundamental role in emphysema and fibrosis [50], as demonstrated by the endothelial cells in our study. To address this, our findings suggest a potential strategy of targeting dysfunctional or overactive endothelial cell populations to restore their homeostatic function and rebalance ineffective repair versus excessive tissue remodeling. Further studies are needed to validate key intercellular interactions and test new therapies targeting these molecules in experimental settings that resemble the heterogeneous microenvironment in CPFE. It is important to point out that​ the insights gained from CPFE patients may also be relevant to​ the more common conditions of emphysema and pulmonary fibrosis. Our findings may therefore have broader impacts and inform​ therapeutic strategies for these diseases.

We acknowledge the limitations of our study: given the rarity of CPFE and the even more limited subset of patients undergoing lung transplantation, the data presented here were derived from two CPFE patients, we have to admit that the prevalence of CPFE is much lower than pulmonary fibrosis, COPD, or smokers. A study investigated the prevalence and incidence of CPFE among US veterans, annual CPFE prevalence ranged 71–100 per 100,000, and incidence ranged 16–39 per 100,000 [51]. Further confirmatory studies are needed to validate our findings in a larger number of CPFE patients, particularly by collecting samples from distinct regions of the lung. It would also be interesting to compare CPFE patients with and without pulmonary hypertension ​​to verify the contribution of key endothelial cell subtypes. Given the relative rarity of CPFE compared to IPF or COPD, along with the scarcity of patients undergoing lung transplantation, CPFE data are exceptionally rare and valuable. Our study provides the first analysis of CPFE at the cellular level, offering unprecedented insights into its cell type-specific pathobiology.

Conclusions

We present the first cellular characterization of CPFE and identify endothelial cells as pivotal regulators of lung region-specific pathology in CPFE, orchestrating tissue repair and remodeling through angiocrine signaling. This study provides new insights into the cellular and transcriptomic landscape of CPFE, paving the way for future investigations into its underlying molecular mechanisms and potential for therapy. It also identifies endothelial cell-mediated pathways, particularly their interactions with macrophages, as a potential therapeutic targets.

Abbreviations

ACKR1

Atypical Chemokine Receptor 1

α-SMA

alpha-Smooth Muscle Actin

AT1

Alveolar Type 1

AT2

Alveolar Type 2

CCL14

C-C Motif Chemokine Ligand 14

CPFE

Combined Pulmonary Fibrosis and Emphysema

COPD

Chronic Obstructive Pulmonary Disease

DAPI

4',6-Diamidino-2-Phenylindole

DEGs

Differentially Expressed Genes

ECM

Extracellular Matrix

GO

Gene Ontology

H-lower

Healthy Lower Lung

H-upper

Healthy Upper Lung

ICAM1

Intercellular Adhesion Molecule 1

IHC

Immunohistochemistry

IPF

Idiopathic Pulmonary Fibrosis

JAG1

Jagged Canonical Notch Ligand 1

KEGG

Kyoto Encyclopedia of Genes and Genomes

mPAP

Mean Pulmonary Arterial Pressure

NOTCH3

Neurogenic Locus Notch Homolog Protein 3

PDGFB

Platelet-Derived Growth Factor Subunit B

PDGFRa

Platelet-Derived Growth Factor Receptor Alpha

PH

Pulmonary Hypertension

scRNA-seq

single-cell RNA sequencing

SDC4

Syndecan 4

THBS1

Thrombospondin 1

TrMs

Tissue-Resident Macrophages

TSA

Tyramide Signal Amplification

Authors’ contributions

Author contributions: H.H., C.L., and P.Y. Conceptualization. H.H., J.Z. W.S., C.L., and P.Y. Methodology. C.L., P.Y., Funding Acquisition. H.H., J.Z., Y.G., X.J., and H.Y. Investigation. Z.G., and W.S. Data Curation. Z.G., Y.H., H.Z., and W.S. Formal Analysis. H.H., and Z.G. Writing – Original Draft. W.S., and P.Y. revised the manuscript. J.Z., Y.H., H.Z. and C.W. Supervision, Project Administration.

Funding

This study was funded by the Noncommunicable Chronic Diseases-National Science and Technology Major Project (2024ZD0522303, 2024ZD0528700), National Natural Science Foundation of China (Excellent Youth Scholars Program, 82270062), Beijing Municipal Natural Science Foundation (7242096), Chinese Academy of Medical Sciences Innovation Fund for Medical Sciences (2021-I2M-1-001, 2023-I2M-2-001), Non-Profit Central Research Institute Fund of the Chinese Academy of Medical Sciences (2021-RC310-016), and the State Key Laboratory Special Fund (2060204).

Data availability

Data are available from the corresponding author (P.Y.) upon request.

Declarations

Ethics approval and consent to participate

The study protocol involving the collection of human specimens was approved by the Clinical Research Ethics Committee of China-Japan Friendship Hospital (Approval No. 2024-KY-039). All procedures performed in this study were in accordance with the ethical standards of the Declaration of Helsinki.​ Written informed consent was obtained from all participants prior to their enrollment.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.