Distinct endothelial cells in chronic thromboembolic pulmonary hypertension

Fig. 1. Single-cell RNA sequencing identifies 20 distinct cell populations in human pulmonary endarterectomy specimens. a Experimental setup: pulmonary endarterectomy specimen tissue was dissociated physically and digested enzymatically. Dissociated cells underwent 3’end, droplet-based single-cell RNA sequencing, generating single-cell expression profiles. Principal component analysis, clustering, differential gene expression, and universal manifold approximation and projection (UMAP) visualizations were performed. Figure created with BioRender.com. b UMAP visualization of clustering revealed 20 cell populations. Population identities were determined based on marker gene expression (Supplementary Table 2). c Heatmap of top 5 markers for each cell cluster illustrating a subset of marker gene expression patterns between clusters.

Fig. 2. Differences in single-cell expression profiles between control and CTEPH endothelial cells. a UMAP visualization and clustering of endothelial cells from control and CTEPH. Left: UMAP visualization with cells parsed by control vs CTEPH. Right: UMAP visualization with 12 endothelial cell clusters shown. b Proportion of cells for each endothelial cell cluster in control and CTEPH. c Number of endothelial cells included in analysis for control and CTEPH. d Heatmap of top 5 markers for each endothelial cell cluster.

Fig. 3. Endothelial dysregulation in thrombogenesis and thrombus resolution in CTEPH ECs. Dot Plots split by condition (control vs CTEPH) showing expression of markers of (a) endothelial thrombotic regulation, (b) endothelial activation, (c) TFG-β signaling, and (d) angiogenesis. Control ECs are represented in red and CTEPH ECs are represented in blue.

Fig. 4. Individual ECs in CTEPH PEA specimens express both pulmonary artery (PA) and bronchial artery EC markers. a Dot plot showing expression of lung endothelial marker genes in control and CTEPH endothelial clusters. Gen, general endothelial cell marker; Art, PA; Vein, pulmonary vein; Cap-a, capillary aerocyte; Cap, general capillary cell; Bronchial, bronchial vessel; lymph, pulmonary lymphatic endothelial cell. b Venn diagrams showing expression of pulmonary arterial endothelial cell marker, GJA5, and pulmonary bronchial endothelial marker, SPRY1, in individual endothelial cells.

Fig. 5. Analysis of gene expression in endothelial cells expressing PA endothelial markers exclusively. a UMAP visualization and clustering of endothelial cells expressing PA endothelial markers from control and CTEPH endothelial cells. Left: UMAP visualization with cells parsed by control vs CTEPH. Right: UMAP visualization with 5 PA endothelial cell clusters shown. b Proportion of PA endothelial cells for each cluster in control and CTEPH. c Heatmap of top 5 markers for each PA endothelial cell cluster. d–g Expression of markers of (d) endothelial thrombotic regulation, (e) endothelial activation, (f) TFG-β signaling, and (g) angiogenesis in PAECs. Control ECs are represented in red and CTEPH ECs are represented in blue.

Fig. 6. Trajectory analysis of endothelial cells. a Trajectory analysis overlaid on UMAP visualization of endothelial cell clusters. Multiple trajectories are indicated. b Pseudotime overlaid on UMAP visualization of clusters with E.2 (predominantly PA endothelial cells) used as a root for pseudotime analysis. c Clusters ordered by pseudo time with E.2 as a root. Note that E.2, E.3 and E.5 are “PA” endothelial cell clusters, E.6 and E.12 are “bronchial artery” endothelial cell clusters. d Examples of genes that are differentially expressed in pseudotime.

Fig. 7. Comparison of CTEPH pulmonary endarterectomy specimens with atherosclerosis surgical specimens. a UMAP visualization and clustering of integrated single-cell RNA sequencing from atherosclerosis specimens (GSE159677) and pulmonary endarterectomy specimens. 20 cell populations are indicated. b UMAP visualization of subclustered endothelial cells. 12 endothelial subclusters are shown. c UMAP visualization of EC subclusters with cells parsed by atherosclerosis (athero) vs CTEPH.