Spatial Transcriptomic Approach to Understanding Coronary Atherosclerotic Plaque Stability

Abstract

Background: Coronary atherosclerotic plaques susceptible to acute coronary syndrome have traditionally been characterized by their surrounding cellular architecture. However, with the advent of intravascular imaging, novel mechanisms of coronary thrombosis have emerged, challenging our contemporary understanding of acute coronary syndrome. These intriguing findings underscore the necessity for a precise molecular definition of plaque stability. Considering this, our study aimed to investigate the vascular microenvironment in patients with stable and unstable plaques using spatial transcriptomics.

Methods: Autopsy-derived coronary arteries were preserved and categorized by plaque stability (n=5 patients per group). We utilized the GeoMx spatial profiling platform and Whole Transcriptome Atlas to link crucial histological morphology markers in coronary lesions with differential gene expression in specific regions of interest, thereby mapping the vascular transcriptome. This innovative approach allowed us to conduct cell morphological and spatially resolved transcriptional profiling of atherosclerotic plaques while preserving crucial intercellular signaling.

Results: We observed intriguing spatial and cell-specific transcriptional patterns in stable and unstable atherosclerotic plaques, showcasing regional variations within the intima and media. These regions exhibited differential expression of proinflammatory molecules (eg, IFN-γ [interferon-γ], MHC [major histocompatibility complex] class II, proinflammatory cytokines) and prothrombotic signaling pathways. By using lineage tracing through spatial deconvolution of intimal CD68⁺ (cluster of differentiation 68) cells, we characterized unique, intraplaque subpopulations originating from endothelial, smooth muscle, and myeloid lineages with distinct regional locations associated with plaque instability. In addition, unique transcriptional signatures were observed in vascular smooth muscle and CD68⁺ cells among plaques exhibiting coronary calcification.

Conclusions: Our study illuminates distinct cell-specific and regional transcriptional alterations present in unstable plaques. Furthermore, we characterize spatially resolved, in situ evidence supporting cellular transdifferentiation and intraplaque plasticity as significant contributors to plaque instability in human coronary atherosclerosis. Our results provide a powerful resource for the identification of novel mediators of acute coronary syndrome, opening new avenues for preventative and therapeutic treatments.

Keywords: acute coronary syndrome; atherosclerosis; coronary thrombosis; macrophages; muscle, smooth, vascular.

Figure 1.

Unstable coronary atherosclerotic plaques express unique pro-inflammatory transcriptional patterns regardless of spatial anatomy. (A) Stable and (B) Unstable atherosclerotic plaques were reviewed by an independent clinical pathologist and classified by both H&E and histologic analysis. Macrophage (and macrophage-like) cells were fluorescently labeled with a CD68 marker (yellow, inset image), while coronary SMCs were labeled with a smooth muscle myosin heavy chain (SMMHC) marker (red). Nuclei were labeled by nucleic acid stain (SYTO13, blue). (C) Independent of spatial anatomy, unstable atherosclerotic plaques differentially expressed 95 genes, compared to 12 genes from stable plaques (significance defined as p =0.02 and log2 fold change of 0.5). (D) As predicted, gene set enrichment analysis (GSEA) of unstable plaques revealed pro-inflammatory and immune regulatory transcriptional programming (FDR q < 0.05). Scale bars are illustrated for each image. N = 5 patients/group.

Figure 2.

Spatial transcriptomics reveal unique pathologic transcriptional differences in the tunica media of stable and unstable coronary atherosclerotic plaques. Coronary SMCs within the tunica media (A) of stable and unstable coronary atherosclerotic plaques were identified by immunofluorescence (SMMHC) and selectively sampled by spatial transcriptomics (white shapes highlight representative ROIs). (B) A total of 229 genes were differentially expressed within medial SMCs from stable (155 upregulated) and unstable (74 upregulated) plaques (data analyzed per ROI, N = 14 stable and 16 unstable ROIs from 5 patients in each group). KLF4, a mediator of SMC phenotype switching, was also found to be upregulated in patients exhibiting unstable plaques. Despite these differences, GSEA failed to identify upregulation of known signaling pathways (FDR q < 0.05). Significance defined as p=0.02 and log2 fold change of 0.5. Scale bars are 250μm for each image.

Figure 3.

Spatial transcriptomics reveal unique anatomic and pathologic transcriptional differences in tunica intima of stable and unstable coronary atherosclerotic plaques. Coronary SMCs within the tunica intima (A) of stable and unstable coronary atherosclerotic plaques were identified by immunofluorescence (SMMHC) and selectively sampled by spatial transcriptomics (white shapes highlight representative ROIs). Intimal SMCs were defined by their spatial relationship to the internal elastic lamina (IEL) and vessel lumen (yellow arrows). (B) A total of 446 genes were differentially expressed within intimal SMCs with 18 stable versus 428 unstable genes being upregulated (N = 4 stable ROIs from 3 patients and 5 unstable ROIs from 4 patients). (C) GSEA of both groups revealed region-specific transcriptional programming unique to the tunica intima (FDR q < 0.05). Significance defined as p = 0.02 and log2 fold change of 0.5. Scale bars are 500μm for each image.

Figure 4.

CD68+ cells from unstable coronary plaques display smooth muscle and endothelial lineage. CD68+ cells within the intima of unstable plaques were selectively sampled through segmentation (N = 12 ROIs from 4 patients). Principle component analysis (A, PCA) was used to compare global patterns for CD68+ cells with medial SMCs (n = 17 SMC ROIs from 5 patients). While SMCs expressed homogenous signatures, CD68+ cells displayed marked heterogeneity and variable programming. PCA and spatial deconvolution analysis with RNA sequencing (B) identified both myeloid and hybrid CD68+ subpopulations. Specifically, hybrid cells displayed programming consistent with myeloid, smooth muscle, and endothelial cells. Interestingly, hybrid cells were intraplaque, whereas myeloid cells were located peripherally (C). (D) 209 genes were differentially expressed between CD68+ subpopulations (p=0.02 and log2 fold change of 0.5), and GSEA (E) revealed mesenchymal transition and matrix remodeling as markers of hybrid CD68+ cells (FDR q < 0.05).

Figure 5.

Medial and intimal SMCs from calcified coronary atherosclerotic plaques express unique transcriptional programing. After histologic review, ROIs from coronary SMCs (SMMHC) within the tunica media and intima were segregated by the presence of macrocalcification (A, white shapes highlight representative ROIs). (B) A total of 423 genes were differentially expressed within SMCs with 147 non-calcific versus 276 calcific genes being upregulated (N = 6 ROIs / group from 5 unstable patients). GSEA revealed upregulation of inflammatory and metabolic programming in patients with calcific atherosclerotic plaques. In a similar comparison, 294 genes were differentially expressed among CD68+ cells from plaques with and without macrocalcification (C). In contrast to SMC, CD68+ cells from non-calcific plaques displayed altered programming with upregulation of pathways in stress response and extracellular matrix regulation. Significance defined as p = 0.02 and log2 fold change of 0.5. FDR q < 0.05 for GSEA.