Atherosclerosis is a complex disease that affects the arteries, the blood vessels that carry oxygen-rich blood from the heart to the rest of the body. It is a leading cause of heart attacks and strokes, making it a significant public health concern. At its core, atherosclerosis involves the buildup of fatty deposits, or plaques, within the arterial walls. These plaques can harden and narrow the arteries, reducing blood flow and leading to serious cardiovascular complications. Understanding the intricate details of how atherosclerosis develops is crucial for finding new ways to treat and prevent it.
Recent technological advances have revolutionized the ability to study the tiny components that make up our bodies, right down to individual cells. One of the most exciting advancements in this area is single-cell RNA sequencing (scRNAseq). This technology allows scientists to examine the genetic activity of single cells, giving us an unprecedented view of the cellular diversity within tissues. By applying scRNAseq to the study of atherosclerosis, researchers can gain a deeper understanding of how different types of cells contribute to the formation and progression of plaques in the arteries.
The research utilizes a cutting-edge method called Smart-Seq2 to analyze the genetic material from individual cells in atherosclerotic tissues. This technique provides detailed information on the different types of cells present in atherosclerotic plaques and how they interact with each other. By examining these interactions, we can identify the specific roles that different cells play in the disease process.
One of the most significant findings from the research is the identification of six distinct subtypes of cells involved in atherosclerosis. Three of these subtypes are related to smooth muscle cells (SMCs), which are typically involved in maintaining the structure of the arterial walls. The other three subtypes are related to macrophages, a type of immune cell that can contribute to inflammation and plaque buildup. Understanding these subtypes and their functions can help us pinpoint the exact mechanisms that drive the progression of atherosclerosis.
The benefits of the research are far-reaching. By providing a detailed map of the cellular landscape in atherosclerotic plaques, we can identify new targets for therapeutic intervention. For example, if we know which specific cell types and genetic pathways are responsible for plaque formation and instability, we can develop treatments that specifically target these components. This targeted approach has the potential to be more effective and have fewer side effects than current treatments, which often affect the entire body.
Moreover, the research highlights the importance of genetic regulation in atherosclerosis. By integrating the findings with data from the STARNET study, which includes genetic information from hundreds of patients with and without coronary artery disease, we can understand how genetic variations influence the disease. This integration allows us to identify key drivers of atherosclerosis—genes that play a critical role in the disease process. Targeting these key drivers could lead to the development of new drugs that prevent or slow down the progression of atherosclerosis.
Another important aspect of our research is the focus on the extracellular matrix (ECM), a network of proteins and molecules that provide structural support to cells. Changes in the ECM can influence cell behavior and contribute to disease. Our findings suggest that certain genetic networks involved in ECM organization and intracellular calcium regulation play a crucial role in atherosclerosis. By understanding these networks, we can develop strategies to modify ECM components and potentially reduce plaque formation and instability.
In summary, the research provides a comprehensive view of the cellular and molecular mechanisms underlying atherosclerosis. By leveraging advanced single-cell technologies and integrative genomics, we can identify new therapeutic targets and develop more effective treatments for this debilitating disease. The ultimate goal is to reduce the burden of atherosclerosis and improve the health and quality of life for millions of people worldwide.
The study…. scientific in-depth analysis
Technological advances in characterizing molecular heterogeneity at the single-cell level have resulted in a deeper understanding of the biological diversity of cells that transform healthy tissues into diseased ones, including the formation of atherosclerotic plaques in the arterial wall. Here, by applying Smart-Seq2, we provide the deepest multi-species scRNAseq data set of atherosclerosis to date, which identified and markedly expanded 3 inflammatory- and osteogenic-related SMC subtypes, and 3 M1-type proinflammatory and Trem2-high lipid-associated-like macrophage subtypes critical for advanced and symptomatic atherosclerosis. Next, we provided pathophysiological and clinical context to these 6 subcellular clusters by integrative analysis with a CAD framework of 135 tissue-specific GRNs identifying 3 arterial wall GRNs; the macrophage-specific GRN33 and GRN122, and the SMC-specific GRN39, of which GRN39 was successfully validated in independent data and experimentally.
The clustering of scRNAseq data has proven critically important in delineating the subtypes of vascular cells promoting atherosclerosis. However, scRNAseq data have yet to be proven useful for network inference. Critical components of -omics data sets to allow network inference are (1) gene (RNA) expression data reflecting the full spectra of variation associated with disease including lowly expressed genes and (2) expression data coupled to genetic diversity representing at least hundreds of individuals. Currently, scRNAseq data sets do not fulfill these criteria. Moreover, given that the primary goal of scRNAseq is to define disease-relevant subtypes of cells, subsequent network inference within these subtypes is simply untenable as the subtype data lack sufficient depth. Furthermore, fluorescent-activated cell sorting introduces cell-type biases. Thus, although a coexpression network method using scRNAseq data has been proposed, the lack of genetic diversity prevents the inference of directional coexpression networks (ie, GRNs), which in turn prevents the identification of key disease drivers and causal-inference of networks with clinical phenotypes. On the other hand, the human cardiometabolic-focused multiomics STARNET study data sets, obtained from hundreds of living patients with and without obstructive coronary atherosclerosis, meet all the abovementioned criteria. Hence, by integrating STARNET GRNs, we could provide key drivers, clinical contexts, heritability contributions, and pathophysiological mechanisms to the subcellular clusters inferred from our scRNAseq data of advanced and symptomatic atherosclerosis.
The top-ranked key driver of the macrophage-dominated GRN33 (Figure 6), KLF4 (Kruppel-like factor 4 (KLF2 and KLF6 were also top key drivers) has previously been implicated in atherogenesis as a deciding factor for SMC-switching into an macrophage-like osteogenic phenotype contributing to both plaque calcification and destabilization. In addition, KLF4 has been reported to be an important regulator within macrophages. Besides KLF2/4/6, other top key drivers in GRN33 were also compelling: ATF3 (a negative regulator of inflammation-promoting cholesterol metabolism in macrophages), CEBPD (a promoter of lipid accumulation in M1-type macrophage in the atherosclerotic lesion), EGR1-3 (early growth response transcription factors regulating the expression of proteins such as IL1B and CXCL2), FOS/FOSB/FOSL1&2 and JUN/JUNB/JUND (FOS and JUN are involved in formation of foam cells and the FOS and JUN families form the transcription factor complex AP-1 a key regulator of cell proliferation, differentiation, and transformation), MYC/MXD1 (forming the MAX transcription factor involved in regulating cellular transformation), NFIL3 (activates ATFs), and NR4A1-3 (the 3 family members of nuclear receptor subfamily 4 group, which is part of the steroid-thyroid hormone-retinoid receptor superfamily). Taken together, the overarching picture of these key drivers suggests that in the advanced stages of atherosclerosis, GRN33 promotes the formation of M1-type macrophages, resulting in symptomatic plaques likely through lipid accumulation and increased inflammation.
While GRN33 has in part already been implicated in atherogenesis, the macrophage-dominated GRN122 has not but stands out as a major contributor to CAD heritability (H2=4.1%; Figure 6E). Although GRN122 contains 16 candidate genes assigned by GWAS of CAD, the 4.1% H2 contribution cannot be explained by these risk genes alone, as the contribution per GWAS risk locus on average is <0.1%. Instead, the genetic regulation of GRN122 genes (ie, the expression quantitative trait locus [eQTLs] in GRN122) carries substantial heritability for CAD. GRN122 is also interesting given that out of its 766 genes, only 30 were identified as key drivers. Moreover, out of these, only seven were top-ranked, namely BCL11B (enhances IL2 expression), HLA-DPA1/HLA-DQA1 (class II molecules expressed in antigen-presenting cells like macrophages), IKZF3 (involved in regulating BCL2 expression and controlling apoptosis in an IL2-dependent manner), STAT4 (regulating the differentiation of T helper cells), TRAF3IP3 (promotes inflammation by stimulating translocation of toll-like receptor 4 to lipid rafts, and ZNF831 (unknown function). Taken together and consistent with its major GO representation, the importance of GRN122 as a major contributor to CAD heritability paralleled by its strong association with the degree of coronary atherosclerosis appears to be explained by an immune-regulatory role essential for advancing the atherosclerotic plaque to become symptomatic.
Notably, unlike GRN33, which was only enriched in genes from mMP5, representing lipid-poor proinflammatory macrophages, GRN122 was the only macrophage GRN enriched in genes from mMP6, representing TREM2-high, lipid-rich suppositively atheroprotective macrophages. However, GRN122 was also partly enriched in mMP5 genes. Thus, GRN122 may have a decisive immune-regulatory role in determining whether lesion macrophages diversify into becoming proatherogenic or atheroprotective.
WNT (wingless-related integration site) signaling is believed to act through 3 main pathways: the canonical Wnt pathway affecting transcription, the noncanonical planar cell polarity pathway modulating cytoskeleton and cell appearance, and the noncanonical Wnt/calcium pathway regulating intracellular calcium levels. While the ECM (extracellular matrix) has been reported to modify WNT signaling, WNT signaling itself alters cell adhesion, inflammation, calcification, and even atherosclerosis. GRN39 contained 30 ECM-related genes (including the top key drivers; ABHD2, ADAMTSL3, CHST3, ECM2, FMOD, IGFBP2, GALNT5&10 and QSOX1; other key drivers; AEBP1, ECM2, COL13A1, and SMOC2; and among regular network nodes; ANOS1, ANTXR1, COL1A2, COL8A1, COL8A2, COL10A1, COL14A1, ENG, FKBP10, FN1, KAZALD1, LAMB2, MMP2, P3H4, SLC2A10, TNFRSF11B, and VCAN), 7 genes involved in proliferation/differentiation (top key driver; AEBP1, IGFBP2, and TBX2; other key drivers; CFH, SMOC2, and PDGFD; nodes; FGF1), 34 genes involving intracellular calcium regulation (including the top key driver, AHNAK2, EPDR1, PCDH10, SHC4, and TBX2; other key drivers; CAMK1D, NELL2, PDGFD, and TNS3; and among regular nodes; AQP1, BCL21, CACNB4, CACNA1H, CCND1, CD55, CDH13, CPNE4, CRIP2, ETV5, FGF1, FN1, FZD7, GAS6, HTRA1, MEGF10, MMP2, PDGFA, PDGFC, PGF, SCG2, SERPINE1, ST8SIA1, SYT12, and WWTR1), and 28 genes involved in cell adhesion (including the top key drivers; ALCAM, ECM2, EPDR1, ITGB5, and ITGBL1; other key drivers; COL13A1, SMOC2, and TNS3; and among regular nodes; ATRNL1, CDH11, CDH13, CERCAM, COL8A1, FN1, FZD7, GA6, GAS-AS2, ITGA1, ITGA10, ITGA11, ITGA3, JAM3, LAMB2, PARVA, PEAK1, SERPINE1, SPRY4, and TJP2). Taking these genes together, GRN39’s involvement in ECM organization and intracellular calcium regulation suggests it modifies Wnt signaling through the noncanonical Wnt/calcium pathway, which in turn activates SMC proliferation and cell adhesion, thus promoting atherosclerosis. In support of this notion, GRN39 top key drivers; DKK3, FRZB, and FZD1 are well-established modifiers of Wnt signaling and ablation of DKK3 has been shown to attenuate the development of atherosclerosis in mice.
Our study has limitations: first, the number of extracted T-cells in the atherosclerotic lesions was low although their role in advanced stage and symptomatic atherosclerosis is likely more important than suggested by our study. For instance, it is evident that the macrophage-dominated GRN122 takes part in regulating cellular immunity. Similarly, the single-cell isolation protocol in our study yielded few luminal ECs relative to ECs from the microvasculature. Carotid plaques isolated by endarterectomy contain few luminal ECs. In contrast, carotid neovascularization has previously been implicated in plaque rupture. In our study, however, among 4 mouse (3 from vasa vasorum) and 6 human (3 from microvasculature) subtypes of ECs, none were enriched in cells isolated during the advanced stages of atherosclerosis, or from symptomatic carotid plaques. Albeit using a combination of human and mouse atherosclerosis tissue sources for our study rather than relying on a singular model system as most prior studies, we recognize that, as surrogates for atherosclerosis in the coronary arteries (ie, CAD), our data from human and mouse aorta and carotid plaques may have limitations. Nevertheless, our combined analyses of bulk and single-cell RNA sequence data suggest that the mechanisms of atherosclerosis at these different arterial wall locations are largely overlapping. We also note that since Smart-seq2 is a fluorescent-activated cell sorting–based technique relying on extracting atherosclerosis-relevant cell types using antibodies some cells may not be recognized and therefore lost in the subsequent analysis. Additionally, although the cultured primary vascular SMCs exposed to serum-free medium for 24 hours have previously been used to model contractile SMCs, establishing the exact phenotypic nature of SMCs in vitro is difficult. Last, like published scRNAseq studies, we relied on a relatively small but compared with earlier studies larger number of carotid plaques and mouse aortic arches.
Driven by the successes of genetic studies of rare disorders, the scientific focus of studies aiming to understand the genetics of complex traits, like atherosclerosis, rely largely on identifying isolated target pathways and genes, most notably studies of individual candidate genes in risk loci identified by GWAS. Besides unraveling new biology, the underlying motivation for this one-by-one gene strategy leans heavily on the idea that there are disease-central pathways that if targeted alone will substantially mitigate the development of atherosclerosis. However, recent insights from the field of genetics tell another story suggesting the one-by-one strategy is suboptimal. Namely, the sheer number of risk alleles (regardless of their individual nature in terms of genomic location, target genes, and pathways) determines the development of clinically significant atherosclerosis. Thus, risk genes are interchangeable as their corresponding pathways do not act in isolation but jointly within larger biological systems of hundreds, or even thousands of connected genes (ie, within GRNs) that together constitute the pathophysiological processes that drive complex diseases like atherosclerosis. In our study, by integrating subtypes of vascular cells inferred from the deepest scRNAseq data set to date with a CAD framework of GRNs, we identified 3 largely cell-specific GRNs essential for the development of advanced and symptomatic atherosclerosis. Finding new therapies targeting the top key drivers of these GRNs is a realistic strategy to substantially mitigate the development of symptomatic atherosclerosis and its deadly consequences, like myocardial infarction and stroke.
In-Depth Analysis and Implications
The Smart-Seq2 method has allowed us to dive deep into the cellular and molecular underpinnings of atherosclerosis, revealing complex interactions and identifying key drivers of the disease. By focusing on single-cell RNA sequencing (scRNAseq) data, we have been able to uncover the heterogeneity within cell populations that traditional bulk RNA sequencing methods might overlook. This depth of analysis is crucial for understanding the precise mechanisms at play in atherosclerosis, which involves a multitude of cell types and states.
The identification of six subtypes of cells—three related to smooth muscle cells (SMCs) and three related to macrophages—provides a detailed map of the cellular landscape in atherosclerotic plaques. These subtypes are not just markers of different cell states but active players in the progression of the disease. For instance, the SMC subtypes show distinct inflammatory and osteogenic characteristics, suggesting that SMCs can undergo phenotypic changes that contribute to plaque formation and instability.
The three identified macrophage subtypes—M1-type proinflammatory and Trem2-high lipid-associated macrophages—highlight the role of inflammation and lipid metabolism in atherosclerosis. The M1-type macrophages are known for their proinflammatory actions, which can exacerbate the disease, while the Trem2-high macrophages are associated with lipid accumulation, a hallmark of atherosclerotic plaques.
Integrating these findings with the CAD framework of gene regulatory networks (GRNs) allows us to place these cellular changes within a broader genetic and pathophysiological context. The identification of GRN33, GRN122, and GRN39 as critical networks provides new targets for therapeutic intervention. Each of these networks contains key drivers—genes that play pivotal roles in the regulation of other genes within the network.
The Role of Key Drivers in Atherosclerosis
The key drivers identified within GRN33, such as KLF4, ATF3, CEBPD, and others, underscore the complexity of gene regulation in atherosclerosis. KLF4, in particular, is a transcription factor that influences SMCs to adopt a macrophage-like phenotype, promoting both plaque calcification and destabilization. This switch is significant because it highlights a potential target for therapies aimed at stabilizing plaques and preventing rupture.
ATF3, another key driver, acts as a negative regulator of inflammation-promoting cholesterol metabolism in macrophages. This dual role of regulating inflammation and cholesterol metabolism places ATF3 at a strategic point in the pathophysiology of atherosclerosis, making it a potential target for reducing plaque inflammation.
Similarly, GRN122, with its significant contribution to CAD heritability, points to the importance of genetic regulation in disease progression. The presence of key drivers like BCL11B and HLA-DPA1/HLA-DQA1 within this network indicates an immune-regulatory role. These genes are involved in immune responses and could influence the inflammatory environment within plaques, driving their progression towards symptomatic atherosclerosis.
GRN39, on the other hand, is heavily involved in extracellular matrix (ECM) organization and intracellular calcium regulation. This network’s involvement in noncanonical Wnt signaling pathways suggests that targeting these pathways could modulate SMC proliferation and cell adhesion, potentially mitigating plaque growth and instability.
Methodological Considerations and Future Directions
Our study’s methodological strengths lie in the use of both human and mouse models, providing a more comprehensive view of atherosclerosis. However, there are limitations, such as the low number of T-cells extracted and the challenges in isolating specific cell types. These limitations highlight the need for continued refinement of single-cell techniques to capture the full diversity of cells involved in atherosclerosis.
Moreover, the integration of scRNAseq data with GRNs derived from the STARNET study offers a powerful approach to understanding the genetic underpinnings of atherosclerosis. The STARNET data, obtained from hundreds of patients, provide a robust framework for identifying key drivers and understanding their role in disease progression. This integration bridges the gap between molecular insights and clinical relevance, offering a path forward for developing targeted therapies.
Conclusion
The advancements in single-cell technologies and integrative genomics have revolutionized our understanding of atherosclerosis. By dissecting the cellular and molecular heterogeneity within plaques, we have identified key drivers and regulatory networks that offer new avenues for therapeutic intervention. The future of atherosclerosis research lies in leveraging these insights to develop strategies that target the complex interplay of genes and cells driving the disease, ultimately aiming to reduce the burden of cardiovascular events associated with advanced atherosclerosis.
reference link : https://www.ahajournals.org/doi/10.1161/CIRCRESAHA.123.323184#d1e2424