Single-nucleus profiling of Ossabaw pig atherosclerosis model

Figures

• 

  Figure 1. Characterization of HF/HFru diet-induced changes in aortas of Ossabaw minipigs (A) Schematic design of the experiment. (B and D) Representative histological staining of aortic samples (B) with Movat pentachrome-modified Russell-Movat staining (blue, intima; red, media; yellow, adventitia) and (D) with von Kossa silver staining; black, calcium deposits. Scale bar, 250 μm. (C) Quantification of vessel stenosis, percentage of plaque to total lumen area. Pale dots represent technical replicates, and outlined dots represent biological replicates, i.e., mean of technical replicates per each animal, p values < 0.0001, Mann-Whitney test. (E) Calcification quantification as a percent of total vessel area, p value = 0.04, unpaired Student’s t test. (F) Uniform manifold approximation and projection (UMAP) of the snRNA-seq data from combined control and atherosclerotic pig aortas. (G) Violin plot of cell type-specific marker gene enrichment per cluster. (H) Stack plot showing percentage of cells in each cluster provided from control or atherosclerotic samples. (I and J) Bar plot showing log2 fold change in atherosclerotic samples of (I) inflammation-associated genes and (J) lipid-processing…

• 

  Figure 2. Atherosclerosis-induced changes in gene expression across vascular cell types (A) Number of DEGs per cell type in atherosclerosis. (B) Correlation of significant DEGs between cell types. Numbers indicate Pearson’s correlation coefficient (R value). (C) Top 25 upregulated genes across all cell types, log2 fold change. (D) Dot plot of GO terms enriched in upregulated genes. Dot size indicates the number of genes found in the GO term, and color indicates the Benjamini-Hochberg adjusted p value of enrichment. (E) Top 25 downregulated genes across all cell types, log2 fold change. (F) Dot plot of GO terms enriched in downregulated genes. Dot size indicates the number of genes found in the GO term, and color indicates the Benjamini-Hochberg adjusted p value of enrichment.

• 

  Figure 3. Subclustering and analysis of SMCs and ECs of aortas (A) UMAP of subclustered SMCs. (B) Distribution of cells per subcluster in each condition; χ2 test; ∗ indicates p values < 0.05. (C) Heatmap showing Z score of genes specific for synthetic (blue), contractile (green), and activated (red) SMCs in each subcluster. (D) GSEA plots of genes enriched in SMC_2 and SMC_3 versus SMC_1 and SMC_4 clusters. (E) GSEA plots of genes enriched in SMC_1 and SMC_4 versus SMC_2 and SMC_3 clusters. (F) UMAP of subclustered ECs. (G) Distribution of cells per subcluster in each condition; χ2 test; ∗ indicate p values < 0.05. (H) UMAP of arterial module score enrichment. (I) UMAP of vasa vasorum module score enrichment. (J) UMAP of chemotaxis module score enrichment. (K) UMAP of TGF-β signaling pathway module score enrichment. (L) GSEA plot of genes upregulated in EC_3 in atherosclerosis versus control. (M) Selected top GO terms enriched in genes upregulated in EC_3 in atherosclerosis versus control. p-adj = Benjamini-Hochberg adjusted p value.

• 

  Figure 4. Subclustering and analysis of immune and other cell types of aortas (A) Subclustering of immune cells. (B) Distribution of cells per subcluster in each condition; χ2 test; ∗ indicate p values < 0.05. (C) GO terms enriched in up- and downregulated genes of T cells. (D) UMAP showing SMC-like signature module score. (E) GO terms enriched in SMC-like macrophages. (F) UMAP showing expression of TREM2. (G) GO terms enriched in TREM2+ and foamy MF subclusters. (H) Representative immunohistochemistry images showing CD68+ cells (pan-myeloid marker) in control and atherosclerotic aortas; scale bar, 100 μm. (I) Bar plot of macrophage nuclus percentage in control and atherosclerosis snRNA-seq datasets; χ2 test; ∗∗∗ indicate p values < 0.001. (J) Subclustering of “other” non-EC, non-immune, and non-SMC cell types. (K–N) GO terms enriched in (K) downregulated genes in pericytes, (L) upregulated genes in pericytes, (M) downregulated genes in FBs, and (N) upregulated genes in FBs. p-adj = Benjamini-Hochberg adjusted p value.

• 

  Figure 5. Cell-cell communication analysis of vascular cell types in atherosclerosis and control aortas (A) Interaction plot of interactions between cell types changed in aortas of atherosclerotic vs. control aortas. Thickness of the lines represents the strength of interaction; blue, reduced interaction numbers; red, increased interaction numbers. (B) Heatmap of interaction strength changes in atherosclerotic vs. control aortas. (C and D) Word cloud representing most prevalent ligands in (C) control and (D) atherosclerotic conditions. (E–H) Interaction maps showing (E) SPP1 signaling in atherosclerosis, (F) SEMA3 signaling in control and atherosclerotic samples, (G) TGF-β signaling in atherosclerosis, and (H) thrombospondin signaling in atherosclerosis. Thickness of the lines represents the strength of interaction.

• 

  Figure 6. Comparison of Ossabaw pig SMC, EC, and immune subtypes and atherosclerosis-induced changes to humans and mice (A) Schematic of inter-species comparison. (B–E) Similarity heatmaps between pig abdominal aorta atherosclerosis snRNA-seq and integrated human atherosclerosis scRNA-seq datasets37,55,56,57: (B and C) SMC and FB subpopulations of (B) humans and (C) mice, (D) EC subpopulations, and (E) immune subpopulations. (F and G) Correlation of DEGs between pig abdominal aorta atherosclerosis and (F) human coronary artery atherosclerosis scRNA-seq dataset55 or (G) mouse ascending aorta atherosclerosis scRNA-seq.37 Significant DEGs (p-adj ≤ 0.001, |log2FC| ≥ 1) from the pig dataset were utilized for these comparisons; R, Pearson’s correlation coefficient.