High-Resolution Spatial Profiling Unveils Cellular Heterogeneity in Murine Atherosclerosis

Advances in high-parametric cellular analysis such as single-cell RNA sequencing have uncovered the intricate cellular diversity and plasticity in murine and human atherosclerotic plaques.[1] [2] [3] However, these methods require tissue dissociation, which disrupts spatial context and obscures the relationships between cell states and their precise anatomical locations.

We employed the 10x Genomics Xenium v1 pipeline with 379 pre-selected pan-mouse cell marker genes to uncover spatially resolved cellular heterogeneity in 10-μm histological sections of fresh-frozen and OCT-embedded atherosclerotic aortic arches isolated from 40-week-old male ApolipoproteinE-deficient (Apoe^−/− ) mice fed a chow diet. Cell boundaries were determined by staining extracellular (ATP1A1, CD45, E-cadherin), intracellular (18S, αSMA, vimentin), and nuclear (DAPI) targets ([Fig. 1A]).

Fig. 1 In situ gene expression with subcellular resolution in murine atherosclerosis. (A) Representative immunofluorescent image showing cellular segmentation in the aortic arch of a 40-week-old, male Apoe^−/− mouse. (B) Spatial overlay of major cell clusters in the aortic arch, brachiocephalic trunk (BCT), left common carotid artery (LCCA), and left subclavian artery (LSA) (left) obtained from the uniform manifold approximation and projection (UMAP) visualization of major cell clusters (n = 10,869 cells) (right). (C) Dot plot showing normalized expression levels of the top differentially expressed genes across identified cell clusters. Expanded Footnote: To measure in situ gene expression in the atherosclerotic mouse aorta by Xenium v1, fresh-frozen tissue sections were obtained from an OCT-embedded aortic arch of one 40-week-old, male Apoe^−/− mouse. Tissue preparation followed the 10x Genomics Tissue Preparation Guide (CG000579 Rev E). Excess OCT was carefully trimmed from the tissue using a single-edge carbon blade (Electron Microscopy Science). For downstream analysis, 10-μm tissue sections were placed on a dedicated Xenium slide and stored at −80°C. Following proper tissue placement, fresh-frozen sections were fixed and permeabilized according to the 10x Genomics protocol (CG000581 Rev D). Subsequent steps followed the 10x Genomics User Guide (CG000749 Rev B), spanning a total of 3 days. Briefly, probes from the pre-designed mouse tissue atlassing gene expression panel (containing 379 target genes; full list available on the Xenium website) were hybridized on the tissue overnight. Sections were washed, and probes ligated and amplified. Following the cell segmentation staining, autofluorescence was quenched overnight. Data acquisition with the Xenium instrument required 3 days. Spatial analyses were conducted with the Xenium Explorer 3 software (10x Genomics), which enabled individual and simultaneous visualization of the acquired immunofluorescent image ([Fig. 1A]), automated cell clusters ([Fig. 1B]), and total or preselected gene transcripts (not shown). The software provided gene and cell counts, which were subsequently used for differential gene expression analysis ([Fig. 1C]). Statistical analyses were performed with GraphPad Prism 10.

We obtained sufficient counts from 379 pre-selected marker genes for graph-based clustering, which revealed 13 distinct cell populations within their preserved spatial context ([Fig. 1B]). The top 87 most differentially expressed genes uncovering six smooth muscle cell (SMC)-like clusters, three (myo)-fibroblast-like clusters, one endothelial cell–like cluster, two leukocyte-like clusters (including monocytes and macrophages), and a cluster resembling cells undergoing epithelial–mesenchymal transition (EMT) are depicted in [Fig. 1C]. Other previously described cell types within atherosclerotic plaques, such as dendritic cells (DCs), T cells, and Trem2+ foamy macrophages, were not identified. This absence is likely explained by the reduced granularity resulting from the limited immune cell gene selection in the pre-designed panel. Quantification of identified cell types in non-atherosclerotic and atherosclerotic vessel segments within the same tissue section revealed dynamic changes with increasing cellular complexity with plaque progression ([Fig. 2A, B]).

Fig. 2 Spatial mapping of 13 distinct cell clusters during murine atherogenesis. (A) (left to right): High-resolution images of cell segmentation, gene transcripts, and cell cluster overlays in murine atherosclerosis: adventitia (A), fibrous cap (FC), intima (I), vessel lumen (L), media (M), and necrotic core (NC). (B) Relative cell composition in the non-atherosclerotic wall, and a developing and an advanced atherosclerotic lesions. Expanded Footnote: To investigate the dynamic process of atherogenesis within the same mouse aorta, three regions of interest (ROIs) were selected from the histological overview ([Fig. 1A]), representing the non-atherosclerotic vessel wall, developing plaque, and advanced plaque ([Fig. 2A]). ROIs were manually defined using the freehand selection tool in Xenium Explorer 3, and total area (µm²), total and predesigned gene transcript counts, as well as total and cluster-specific cell counts were extracted. Cellular frequencies within each region were extrapolated and are presented in [Fig. 2B].

The approach utilized here enables the identification of dynamic changes in cellular composition and gene expression across different stages of atherosclerotic development, providing crucial insights into spatiotemporal disease mechanisms.

Publication History

Received: 11 December 2024

Accepted: 18 March 2025

Article published online:
08 April 2025

• References

• 1 Zernecke A, Erhard F, Weinberger T. et al. Integrated single-cell analysis-based classification of vascular mononuclear phagocytes in mouse and human atherosclerosis. Cardiovasc Res 2023; 119 (08) 1676-1689
• 2 Horstmann H, Michel NA, Sheng X. et al. Cross-species single-cell RNA sequencing reveals divergent phenotypes and activation states of adaptive immunity in human carotid and experimental murine atherosclerosis. Cardiovasc Res 2024; 120 (14) 1713-1726
• 3 Gastanadui MG, Margaroli C, Litovsky S. et al. Spatial transcriptomic approach to understanding coronary atherosclerotic plaque stability. Arterioscler Thromb Vasc Biol 2024; 44 (11) e264-e276