If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
The heterogeneous cellular landscape of atherosclerosis: Implications for future research and therapies. A collaborative review from the EAS young fellows
Amsterdam UMC Location University of Amsterdam, Department of Experimental Vascular Medicine, Amsterdam Cardiovascular Sciences, Amsterdam, NetherlandsAmsterdam Cardiovascular Sciences, Atherosclerosis & Ischemic Syndromes, Amsterdam, NetherlandsLaboratory of Angiogenesis and Vascular Metabolism, VIB-KU Leuven Center for Cancer Biology, VIB, BelgiumLaboratory of Angiogenesis and Vascular Metabolism, Department of Oncology, KU Leuven and Leuven Cancer Institute (LKI), Belgium
Single cell omics unveiled the heterogeneous cellular landscape of atherosclerosis.
•
Intraplaque heterogeneity adds complexity to current and future (drug) research.
•
Studying cell activation and cell-cell interactions is the next step forward.
•
More disease and site-specific therapies for atherosclerosis are needed.
Abstract
Single cell technologies, lineage tracing mouse models and advanced imaging techniques unequivocally improved the resolution of the cellular landscape of atherosclerosis. Although the discovery of the heterogeneous nature of the cellular plaque architecture has undoubtedly improved our understanding of the specific cellular states in atherosclerosis progression, it also adds more complexity to current and future research and will change how we approach future drug development. In this review, we will discuss how the revolution of new single cell technologies allowed us to map the cellular networks in the plaque, but we will also address current (technological) limitations that confine us to identify the cellular drivers of the disease and to pinpoint a specific cell state, cell subset or cell surface antigen as new candidate drug target for atherosclerosis.
The advent of single cell technologies fast-forwarded cardiovascular research allowing us to characterise the cellular heterogeneity and complexity of atherosclerotic tissue of both human and mouse specimens. Nowadays, it has been well accepted that all major cell types in atherosclerotic lesions, including endothelial cells (ECs), vascular smooth muscle cells (VSMCs) and immune cells, (macrophages, T cells, B cells) display a great range of heterogeneity and plasticity. Vascular ECs show remarkable heterogeneity with regards to their location in the vascular bed, but also in terms of their dynamic phenotypes and adaptations to microenvironmental changes [
]. Similarly, VSMCs are more plastic than originally thought and undergo phenotypic modulation upon exposure to lipids and cytokines in the atherosclerotic plaque [
]. Besides triggering a switch from a contractile to a synthetic phenotype, VSMCs may transdifferentiate to alternative phenotypes (e.g. macrophage-like, foam cells) and differentially impact plaque stability. The wide variety of plaque immune cells is no longer restricted to macrophage polarisation states and subpopulations but include heterogenous subsets of T cells and B cells, all leading to an imbalance between proinflammatory and pro-resolving mediators in the plaque milieu [
] trials, have provided robust evidence of treating inflammation to reduce major cardiovascular events in humans. As such, studying the heterogeneity of immune cell mechanisms in human atherosclerosis at single cell level has become of great interest, yet newer insights are needed to fully understand the functional complexity of immune cells in the plaque. In this review, we will discuss the high level of heterogeneity and cell plasticity of the different cell populations in atherosclerotic lesions, how this could impact future research and drug development, and which new technological evolutions are needed to define druggable targets for atherosclerosis.
2. Technological advances reveal new cellular states in atherosclerosis
Newly developed techniques, such as single cell sequencing and other ‘omics’ approaches (Fig. 1), have advanced our understanding of the pathophysiology of atherosclerosis.
Fig. 1Overview of single-cell technologies and other omics approaches to advance our understanding of atherosclerosis.
Bulk RNA sequencing is the method for transcriptomic analysis of pooled cell populations or tissue sections. It measures the ‘average’ expression level of individual genes across hundreds to millions of input cells. Single-cell RNA sequencing (scRNA-seq) captures the transcriptome of individual cells. Conventional scRNA-seq technologies initially involved manual isolation of cells using (micro)pipettes or fluorescence activated cell sorting (FACS). Recent advances include valve-based or droplet-based microfluidic devices. Cellular indexing of transcriptomes and epitopes by sequencing (CITE-seq) allows to perform RNA sequencing to gain quantitative and qualitative information of surface proteins on a single cell level using DNA-barcoded antibodies. It uses droplet microfluidic devices to encapsulate individual cells in small volume droplets. This approach integrates cellular protein and transcriptome measurements. Single-cell assay for transposase-accessible chromatin with sequencing (ATAC-seq) has been developed to study cell type-specific chromatin accessibility in tissue samples containing a heterogeneous cellular population. The method requires adapters with Tn5 transposases. This technique aims at elucidating the networks among promoters and enhancers, and transcription factors. Gene activity and accessibility to genetic variants can also be analysed. Cytometry by time-of-flight (CyTOF) measures the abundance of metal isotope labels on antibodies and other tags on single cells using mass spectroscopy. It uses an atomic mass cytometer to detect the time-of-flight (TOF) of each metal. Each atom's TOF is determined by its mass, allowing the composition of metal atoms on each cell to be ascertained.
2.1 New technologies unmask the heterogeneity of plaque cells
The broad spectrum of cells present in plaque tissue has been unravelled via conventional histology, highlighting the classical image of the vulnerable atherosclerotic plaque consisting of a thin cap, compromising smooth muscle cells and infiltrating immune cells. However, technological advances at the single cell level have underscored the biological diversity of cells present in atherosclerotic plaques, and subsequently underpinned the need for improved sub-phenotyping. New subsets of cells have been discovered among cell types previously considered homogeneous and it has become clear that even within one cell type, alteration of gene expression, transcriptomes and proteins/proteome varies with cell activation, cell cycle, apoptosis, stress, or circadian phase [
Initial microarray approaches were utilised to determine ‘bulk’ transcriptomes of atherosclerotic plaques or mixtures of cells like peripheral blood mononuclear cells (PBMC) from patients [
]. Subsequently, deeper sequencing of transcriptomes of fluorescence activated cell sorting (FACS)-sorted cell populations were performed by bulk RNA sequencing (RNA-seq) (Fig. 1). For example, RNA sequencing analysis of macrophages isolated from murine atherosclerotic aorta showed that foamy macrophages expressed few inflammatory genes compared to non-foamy macrophages, suggesting that foamy macrophages might be less inflammatory than originally thought [
]. Using a microfluidic platform for RNA sequencing on sorted CD4+ T cells from murine atherosclerotic aortas, Butcher et al. identified a subpopulation of interferon gamma (IFNγ)+ T helper 1 (Th1)/regulatory T (Treg) cells, addressing Treg plasticity induced by atherosclerosis [
]. Overall, these approaches were useful in elucidating the diversity in the transcriptome of the atherosclerotic plaque and in trying to identify relevant pathways able to drive atherosclerosis progression. However, RNA-seq and its analysis have shown some limitations in the understanding of the heterogeneous nature of plaque tissue due to the fact that they provide the sum of the transcriptome for all cells in the tissue excluding the possibility to identify which cells are responsible for the signal, as well as the effect of those rare but crucial [
]. Although deconvolution methods have provided the opportunity to overcome these restrictions, the analysis' workflow still presents major limitations [
With the aim to develop newer methodologies with more optimal resolution at the cellular levels, a rapid transition from RNA-seq to single-cell RNA-sequencing (scRNA-seq) was made (Fig. 1) [
], these methods cover cellular heterogeneity only incompletely, because they are limited to the discovery of new or rare cell populations depending on the prior knowledge about the phenotype and markers of the cells. Indeed, using CyTOF with a panel of pre-defined 35 antibodies, Cole et al. identified the dynamic changes in the myeloid cell compartment in atherosclerotic aortas from Apoe−/− mice and revealed that high fat feeding skews the myeloid cell subsets towards inflammatory monocyte-macrophage populations rather than to resident macrophage phenotypes, indicating a remarkable plasticity of plaque myeloid cells in response to high fat diet [
]. The advent of scRNA-seq technologies has addressed most of these limitations by facilitating the analysis of the transcriptome of every cell in each sample [
]. Since the first RNA-sequencing study published in 2009, using a microfluidics-based platform for 96-well sorting, today, technologies encapsulating up to 10,000 individual cells into oil droplets are commercially available (drop-sequencing). Importantly, the small volumes required for this technique substantially reduce the risk of external contamination. These methods have successfully been applied to analyse the cellular content of atherosclerotic plaques from mice [
Single-cell genomics reveals a novel cell state during smooth muscle cell phenotypic switching and potential therapeutic targets for atherosclerosis in mouse and human.
]. Using scRNA-seq, apparently similar cell populations can be deconstructed into subpopulations with distinct transcriptional profiles that reflect diverse metabolic states and unique biological functions [
]. For example, a single cell study from Cochain et al. described the phenotypic heterogeneity of aortic macrophages in mouse atherosclerosis. Based on the gene expression signature of the cells, they identified a previously undescribed macrophage population called TREM2hi (i.e. macrophages enriched for the triggered receptor expressed on myeloid cells 2 (Trem2)) with a putative specialised function in lipid metabolism [
Single-cell analysis of fate-mapped macrophages reveals heterogeneity, including stem-like properties, during atherosclerosis progression and regression.
], have identified a foam cell-like population of TREM2hi macrophages, underscoring the strength of the single cell technique despite the differences in experimental design, type of tissue or specie. Recently, several studies have combined scRNA-seq technologies with mass cytometry, particularly to map the immune cell landscape in atherosclerosis [
]. Combining these two approaches, Winkels et al. revealed that the arterial leucocyte population in atherosclerosis is highly diverse (as in surface markers, gene expression program, cytokine secretion) and is relatively dominated by T cells and myeloid cells [
]. Importantly, they found remarkable overlap in the immune cell composition between mouse plaques and human carotid plaques. Overall, these methods have provided new insights into the pathogenesis of atherosclerosis, by revealing cellular heterogeneity with complex and/or rare cell populations, identifying novel subsets of cells, tracking the trajectories of distinct cell lineages and uncovering regulatory mechanisms, and establishing a functional relevance for lesional leukocytes in human atherosclerosis. As such, this technology has been used to construct an atherosclerosis cell atlas of mice and humans [
The next step in single cell platform analysis will include broader multi-omics approaches like DNA methylation by bisulfite conversion sequencing or TAB-seq, noncoding RNAs, chromatin accessibility by the assay for transposase-accessible chromatin with sequencing (ATAC-seq) (epigenomics), histone modification by chromatin immunoprecipitation sequencing, and protein expression levels by cellular indexing of transcriptomes and epitopes by sequencing (CITE-seq, also called AB-seq) (Fig. 1) [
]. Integrating these data with bioinformatics approaches will enable more powerful analysis. Again, this has been mostly applied in studies for immunoprofiling of human atherosclerotic plaques. For example, Fernandez et al. combined scRNA-seq with CyTOF and CITE-seq to immunophenotype human carotid plaques. Plaques from symptomatic vs. asymptomatic patients were characterised by a distinct subset of CD4+ T cells showing transcriptional signatures associated with T cell activation, T cell differentiation and T cell exhaustion [
]. Moreover, macrophages from these plaques displayed alternatively activated macrophage phenotypes, including subsets associated with plaque vulnerability [
]. Depuydt et al. performed scRNA-seq combined with ATAC sequencing on human carotid atherosclerotic plaques to link the transcriptional signature of immune cells with epigenetic changes [
]. They identified specific transcription factors associated with the myeloid subpopulation and T cell cytokine profiles underlying joint activation between both cell types, highlighting the importance of intercellular communication in disease.
All these scRNA-seq methodologies however, require the preparation of single cell suspensions, which leads to the loss of spatial information of cell types within a given tissue [
]. Spatial transcriptomics has the potential to reveal the physical locations of different cell populations in the plaque and to study intercellular communication. Multiple ways can be used to visualise the spatial gene expression in plaque tissue. In turn, spatial barcoding, in situ hybridisation, and in situ sequencing can aid in visualising spatial gene expression. New approaches to fully retrieve spatial information have been introduced such as seqFISH+ [
] that allow for in situ spatial identification of transcripts at near-single cell resolution across a wide array of RNA probes.
2.2 Future developments in single cell technologies and bioinformatic tools
Despite the major advances, there are still limitations to some of these high-throughput techniques. Limitations of scRNA-seq include the limited depth of reads that can be detected per cell, which are often lower than the depth obtained in bulk RNA-seq of sorted cells [
]. Most bulk RNA-seq experiments are sequenced at millions of reads per sample. Shallow sequencing (new NGS sequencing) approaches in scRNA-seq are thought to underrepresent the transcriptional information as a greater sequencing depth increases the number of detectable transcripts. In addition, all scRNA-seq approaches require enzymatic and mechanical tissue dissociation, introducing artefacts [
]. Sequencing RNA of nuclei (snRNA-seq) has been proposed to have several advantages over sequencing intact cells, including (i) reduced dissociation bias (allowing detection of more fragile cell populations, such as ECs, and thus a more reliable representation of cellular landscape changes), (ii) elimination of dissociation-induced transcriptional stress responses, (iii) compatibility with frozen samples (allowing pooling of different samples collected at different time points) [
]. However, thus far, snRNA-seq studies on atherosclerotic tissues have not yet been reported.
Limitations of CITE-seq must be recognised. For instance, CITE-seq is not compatible with intracellular staining and requires a higher technical skill compared to sample preparation for flow cytometry. The workflow requires accessibility to multiple instruments including a FACS sorter, single cell platforms and next generation sequencing sequencers [
Today, the generation of single-cell data is widely accessible thanks to the plethora of available technologies and their various commercial implementations. However, bioinformatic analysis often becomes a bottleneck. Numerous scRNA-seq datasets are already available for the broad scientific community for re-analysis or via online tools like PlaqView [
]. Additionally, numerous computational algorithms have been developed to address the massive amount of gene expression data. However, the proper analysis of single-cell data is often not trivial due to the high complexity of the data complicating the interpretation of the transcriptomic signal. Nonetheless, given the large volumes of data generated, efficient computational and statistical workflows for processing the data have been refined [
3. The cellular heterogeneity in atherosclerosis: current understanding and future challenges
The advent of technologies allowing the identification of multiple cellular parameters and spatial organisation advances our understanding of atherosclerosis, while depicting a heterogeneity of cellular subsets involved both locally and systemically (Fig. 2). This heterogeneity refers not only to different cell populations, but also to their different stages of cell activation, thus tangling the scenario of actors involved in the development, progression and exacerbation of atherosclerotic plaques and its clinical manifestations. By combining these single cell techniques with cell-specific lineage tracing (i.e. fate mapping) studies, we have gained further insights into the mechanisms of plaque development.
Fig. 2Multiple players drive intraplaque heterogeneity in cardiovascular disease.
A schematic overview of the heterogeneous cellular and molecular architecture of the atherosclerotic plaque illustrating the complex nature of the disease and pinpointing the challenge to identify new effective therapies. A high level of heterogeneity and plasticity is found in all major plaque cell types (ECs, VSMCs, immune cells) and may lead to new target discovery for atherosclerosis beyond the traditional lipid species. Dissecting the mechanisms that dictate the acquisition of a ‘plaque stabilising’ phenotype by plaque cells (e.g. VSMCs) would be instrumental towards the identification of novel therapeutic strategies. Whilst great interest is currently going to the targeting of circulatory inflammatory mediators (such as inflammatory cytokines), the identification and targeting of atherosclerosis-specific antigens (e.g. oxidation-specific epitopes) may hold promise to treat atherosclerosis in a disease- and site-specific manner. ECs = endothelial cells, VSMCs = vascular smooth muscle cells.
Vessel wall cell-specific lineage tracing mouse models in combination with single cell technologies have shed new light on how these cells enter the atherosclerotic plaque and have underlined the importance of their high degree of plasticity in atherosclerotic plaque stability.
3.1.1 Endothelial cell heterogeneity
The endothelium is a heterogeneous and crucial barrier between blood and tissue, and is critical for maintaining vascular homeostasis. Damage to the endothelium, defined as endothelial cell dysfunction (characterised by increased EC permeability, leucocyte adhesion, reduced nitric oxide production), is a critical step in the initiation of atherosclerosis. While EC heterogeneity can be envisioned at the organ level, it also exists within vascular beds (e.g., artery vs. vein), between sexes and with ageing, and is partly inflicted by local changes in microenvironmental cues [
One of these well-studied local changes affecting endothelial function in atherosclerotic lesions is the transcription factor Krüppel-like factor 2 (KLF2) [
]. Under high laminar shear conditions, KLF2 promotes endothelial quiescence by upregulating anti-inflammatory and anti-thrombotic proteins and by downregulating pro-inflammatory and pro-thrombotic proteins. However, atherosclerotic lesion sites, which are exposed to disturbed shear stress conditions, are characterised by low KLF2 expression, leading to EC activation, thereby propagating the inflammatory response [
]. These local changes in shear stress regions are a major determinant of the endothelial heterogeneity at atherosclerotic lesion sites. In fact, using (immuno) scanning electron microscopy imaging, Kluza et al. has elegantly shown the diverse landscape of atherosclerotic endothelium [
]. Interestingly, early atherosclerotic lesions in Apoe−/− mice display junctional abnormalities and large transcellular pores, whereas when atherosclerotic plaques progress, the endothelium appears to regenerate, as attested by increased junctional integrity and decreased immune cell adhesion. It is tempting to speculate that this regeneration and junctional stabilisation might be a consequence of intimal remodelling as a protective mechanism to stabilise the atherosclerotic lesion.
Many endothelial scRNA-seq studies underpin the highly heterogeneous endothelial landscape in plaque tissues [
], a state that could affect plaque stability. A particular subset that requires attention is the EC cluster that show signs of endothelial-to-mesenchymal transition. While it has been assumed that myofibroblast-like cells (characterised by the expression of the contractile marker alpha smooth muscle cell actin (ACTA2)) were exclusively derived from VSMCs, recent data show that 20–40% of the ACTA2+ cells arise from non-VSMC sources, including ECs [
]. Combining bulk RNA-seq analysis with in vitro models revealed that this transition is regulated in an IL-1β and TGFβ-mediated manner. Further supported by this notion, histological intraplaque analysis showed both ACTA2 and CD34 expression, suggesting that transdifferentiation of ECs occurs in atherosclerotic lesions [
Besides the heterogeneity of ECs in the vasculature, it is important to highlight that ECs never act alone in the pathology of atherosclerosis. Hence, intricate cellular communication between ECs and their neighbouring cells are pivotal in maintaining homeostasis as well as in driving and modulating atherosclerosis progression. Due to the recent advances in sophisticated bioinformatic analysis, intercellular communication and receptor ligand interaction (RLIs) analysis can now be studied in detail in scRNA-seq datasets, for example using CellPhoneDB or NicheNet. These tools aid in providing valuable insights into which endothelial pathways are the major contributors of atherosclerosis progression and hence could be used therapeutically. For example, a recent study by Depuydt et al. has shown multiple chemotactic interactions between endothelial atypical chemokine receptor ACKR1 and myeloid-derived chemokines such as CCL2, CXCL8, CCL8, and CXCL1 [
]. Furthermore, interaction of platelet-derived growth factor beta (PDGFB) on myeloid subsets with the PDGFB receptor on ECs suggest a possible myeloid-driven induction of angiogenesis, which has been associated with plaque destabilisation [
]. This pioneering research underpins the importance of the endothelial cellular interplay in atherosclerosis. While providing unprecedented insight, it must be noted that while RLIs are able to map putative interactions, these in silico analyses - based on the cellular transcriptomic signature - represent predictions that needs functional validation.
3.1.2 Vascular smooth muscle cell heterogeneity
VSMCs are the other major cell type of the vessel wall and are responsible for maintaining vascular tone through their contractile properties. Upon formation of atherosclerotic plaques, VSMC migrate and proliferate from the media into the lesion to form the fibrous cap. During this process, VSMCs lose their contractile markers and switch from a contractile to a synthetic, de-differentiated phenotype [
Extensive proliferation of a subset of differentiated, yet plastic, medial vascular smooth muscle cells contributes to neointimal formation in mouse injury and atherosclerosis models.
]. Indeed, VSMCs exhibit a much larger degree of plasticity than originally thought, and can transdifferentiate into multiple alternative phenotypes including, ‘macrophage-like’, ‘foam cell-like’, ‘myofibroblast-like’, ‘osteoblast-like’ cells or may even transdifferentiate into EC-like cells (see section 3.1.1) [
]. Single cell data revealed a large transcriptional heterogeneity underlying these VSMC phenotypes, but the relative proportion of these VSMC subtypes likely depends on the disease context, vascular bed type and microenvironmental cues in the plaque. It has been unclear whether contractile VSMCs can directly transdifferentiate into alternative phenotypes or first adopt a de-differentiated state from which other VSMC-derived cell types arise. By combining VSMC fate mapping and scRNA-seq with trajectory analysis, there is now increasing evidence for the existence of a de-differentiated multipotent population called ‘SEM cells’ (stem cell, EC and monocyte), that may represent an intermediate VSMC state [
Single-cell genomics reveals a novel cell state during smooth muscle cell phenotypic switching and potential therapeutic targets for atherosclerosis in mouse and human.
], yet whether this also happens to VSMCs in the (fibrous cap of) atherosclerotic lesions is less clear. All these studies reveal many similarities in the VSMC gene signatures between mouse and human atherosclerosis. This is important because this means we can use mouse models to study the regulatory mechanisms of VSMC fate decisions relevant to human atherosclerosis.
The high level of VSMC plasticity adds to the complexity of the cellular plaque architecture and brings about some limitations, both technical as well as conceptual. For instance, lineage tracing studies have shown that >80% of VSMC-derived cells in mouse plaques are negative for ACTA2 [
] meaning that immunostaining for ACTA2 will greatly underestimate the VSMC content in the plaque. Because VSMCs can become foam cell-like cells, VSMCs have been repeatedly falsely labelled as macrophages. Indeed, >40% of CD68+ cells in human and mouse plaques were found to be of VSMC and not myeloid origin [
]. Although VSMCs can adopt alternatives phenotypes, it is still unclear whether they also take over full functionality. For example, macrophage-like VSMCs are transcriptionally and functionally different from myeloid-derived macrophages, at least in vitro [
]. Hence, it is difficult to predict whether macrophage-like VSMCs are as harmful as bone marrow-derived macrophages for plaque stability.
The success of the single cell technologies lies in the detailed dissection of the VSMC-specific transcriptomic signature of the plaque looking at the co-clustering of cells with either VSMC contractile or de-differentiation genes. The next step is to move away from these associative studies, and advance current technologies to determine the exact function of these phenotypically modulated VSMCs and most importantly, their causal role in plaque vulnerability. One way to do this, is to integrate single cell transcriptomics with epigenomic profiling and human genomics, to identify novel candidate targets that regulate VSMC phenotypic modulation through altering chromatin accessibility and/or as VSMC-associated coronary artery disease (CAD) risk genes [
]. Spatial transcriptomics will be useful in mapping these phenotypically modulated VSMCs to specific regions of the plaque in relation to vulnerability. At this point, the field advocates to tailor future therapeutic interventions at converting the unwanted VSMCs into a beneficial atheroprotective, plaque-stabilising phenotype, although it is not entirely clear yet which cells do, or do not, fit this profile.
3.2 Immune cells
Numerous studies have contributed to characterise the various immune cell populations, identifying cells of both the innate and adaptive arm of the immune response and proposing a different contribution in the physiopathology of atherosclerosis. In fact, the chronic inflammatory nature of atherosclerosis, where the acute and reparative phases occur together, is nowadays a pillar in the understanding of its pathological mechanisms and possible therapeutic strategies. However, whether i) chronic inflammation develops from an excessive effector function rather than an impaired immunosuppression, or ii) specific trigger/s of immune cells activation and the route/s of their recruitment within the atherosclerotic plaque are involved, is still a field of animated research.
3.2.1 Macrophages
Macrophages, one of the first identified and abundant cell populations in the atherosclerotic plaque [
]. A study where they used scRNA-seq combined with genetic fate mapping of CX3C motif chemokine receptor 1 (CX3CR1)+ derived cells during atherosclerosis, identified an unexpected cluster of proliferating monocytes with a stem cell-like signature, suggesting that monocytes may continue in a self-renewal state, rather than immediately differentiate into macrophages upon entry in the plaque [
]. However, there is a recent study that argues that the earliest foam cells in the murine plaque are represented by specialised aortic intimal resident macrophages with limited proliferation capacity, which are then taken over by recruited proliferating monocytes from the circulation [
Once in the atherosclerotic plaque, monocytes differentiate into macrophages, whose phenotype diversity is now acknowledged to be a broader spectrum containing a dynamic continuum of phenotypes [
], with a far greater complexity compared to the classical polarisation states. Thanks to scRNA-seq and CyTOF data from (lineage traced) atherogenic mouse models, five distinct macrophage subtypes have been identified so far, including: 1) the classical foam cell macrophages (characterised by high expression of Trem2, Mmp12, Mmp14, CD11c, cathepsins and markers of steatosis) [
] of which a large portion is considered to be of VSMC origin (as discussed above), 2) CCR2+ macrophages that are believed to derive from circulating monocytes with a pro-inflammatory potential [
], 3) residential vascular macrophages (expressing markers such as Lyve1 and Mrc1) displaying a transcriptomic signature similar to macrophages in healthy arteries [
Single-cell analysis of fate-mapped macrophages reveals heterogeneity, including stem-like properties, during atherosclerosis progression and regression.
Single-cell analysis of fate-mapped macrophages reveals heterogeneity, including stem-like properties, during atherosclerosis progression and regression.
Single-cell analysis of fate-mapped macrophages reveals heterogeneity, including stem-like properties, during atherosclerosis progression and regression.
Single-cell analysis of fate-mapped macrophages reveals heterogeneity, including stem-like properties, during atherosclerosis progression and regression.
]. A recent study from Goossens et al. used a cutting-edge platform of a multiplex immunofluorescent and mass spectrometry imaging approach, to define how these microenvironmental niches are linked to the myeloid cell phenotype. The key advancement in this study was to link the phenotypic heterogeneity of the myeloid cell to its cellular and molecular (i.e. metabolic, lipidomic) spatial context, giving unprecedented insights into how the plaque microenvironment impacts myeloid cell diversity in atherosclerosis [
]. Combining these techniques along with characterisation of cellular function (migration, efferocytosis, lipid accumulation and lysosomal hydrolysis) would further depict the contribution of macrophage dysfunction to inflammation and atherosclerosis progression.
3.2.2 Neutrophils
ScRNA-seq studies have established neutrophil plasticity at the chromatin, transcriptome and proteome levels, which can contribute to neutrophil phenotypic and functional heterogeneity [
]. Research on neutrophil heterogeneity in atherosclerosis is limited because of technical limitations, such as, the limited number of neutrophils that can be detected in an atherosclerotic plaque, low RNA content per cell and high ribonuclease content [
]. However, current studies have shown that the use of modified methodologies for scRNA-seq data analysis can significantly improve the identification and description of neutrophil subtypes in atherosclerosis [
]. Furthermore, the use of antibody-based detection methods, including CyTOF, can serve as a reliable method for detecting neutrophils in the vascular wall and investigating their heterogeneity [
]. In summary, there is currently more data available on the heterogeneity of neutrophils in the circulation than on those located within the atherosclerotic lesion [
]. In the plaque, neutrophils can be represented by two subtypes depending on the level of Sialic acid-binding Ig-like lectin F (SiglecF) expression, similar to neutrophils in the myocardial infarction and tumour microenvironment [
]. SiglecFhi-neutrophils are aged, long-lived cells expressing relatively large amounts of inflammatory and profibrotic cytokines (the concept of neutrophil ageing has been reviewed elsewhere [
]. It is hypothesised that SiglecFhi-neutrophils may have similar characteristics to aged CD62Llo and CXCR4hi circulating neutrophils (i.e. increased ability to form neutrophil extracellular traps and to produce reactive oxygen species), but have unique traits related to the local microenvironment affecting neutrophils during their aging in tissue [
]. The promising pathway for research on the involvement of different neutrophil subtypes in atherogenesis is related to the integration of CyTOF and scRNA-seq, providing new data on phenotypic, transcriptional, and functional features of purified neutrophil subpopulations in atherosclerosis.
3.2.3 T and B cells
T and B lymphocytes play a crucial role in atherosclerosis; CD4+ T cells have been extensively studied in the experimental atherosclerosis population [
]. CD4+ T cells include “inflammatory cell subsets”, such as the Th1 and Th17, which produce inflammatory cytokines (i.e. IFNγ and IL-17), and “tolerogenic cell subsets”, such as Treg cells, which produce pro-resolutive cytokines (IL-10 and TGFβ) [
], adding further complexity to the understanding of their contribution to the disease. Recent scRNA-seq analyses in human and murine atherosclerotic plaques have identified multiple lymphocyte subsets expressing different levels of cytokines, chemokine receptors, and cell exhaustion markers [
]. However, it remains elusive whether these subsets are linked to atherosclerosis severity, and whether they represent pharmacological targets.
However, the presence of lymphocytes activated against atherosclerosis-specific antigens (ASA) (defined as antigens that are selectively generated or being involved due to an atherosclerotic milieu) remains elusive. This is particularly relevant for B lymphocytes that are considered cardinal players in the initiation and progression of atherosclerosis [
]. B cells affect the development of atherosclerotic plaques primarily via the antibodies they secrete. Antibodies are present in four distinct classes, that are IgM, IgG (which include different subclasses: IgG1, IgG2, IgG3, and IgG4 in humans and IgG1, IgG2a/c, IgG2b, and IgG3 in mice), IgE and IgA [
]. Notably, nearly all antibody classes are implicated in atherosclerotic disease. There is a consensus that IgM antibodies display atheroprotective properties, which are primarily linked to the ability of these antibodies to recognise lipid peroxidation-derived products called oxidation-specific epitopes (OSE) [
]. OSEs are modifications that alter the immunogenicity of a host molecule, be that a protein or a lipid, and tag it as an immune target. Such OSEs are present on oxidised LDL and cellular debris that are abundant within atherosclerotic plaques [
]. This is not surprising if one considers the different IgG subclasses and - more importantly - the large variety of antigens that IgG antibodies recognise. On the other hand, IgE antibodies appear to exhibit a strong detrimental role in plaque formation [
However, it remains unknown whether antibodies affect atherosclerosis via recognising ASA. We consider this pending question of critical importance as it could define what immunomodulatory therapeutic strategy would be most effective against atherosclerosis; that is either an antigen-specific or via targeting inflammatory mediators.
Antigen-mediated activation of naïve cells has been shown to occur already in the murine aorta [
]. Indeed, the presence of circulating antigen-experienced lymphocytes, called effector memory T cells, has been associated with atherosclerosis severity in mice and humans [
], is still unknown. Recent work from Depuydt et al., where they combined scRNA-seq with T cell receptor (TCR) sequencing of human carotid plaques with matched PBMC samples, revealed autoimmune-like features of plaque effector T cells [
Single-cell T cell receptor sequencing of paired human atherosclerotic plaques and blood reveals autoimmune-like features of expanded effector T cells.
]. Interestingly, the clonally expanded CD4+ T cells displayed a signature suggestive of antigen-specific activation revealing potential interactions with TREM2+ foam cells [
Single-cell T cell receptor sequencing of paired human atherosclerotic plaques and blood reveals autoimmune-like features of expanded effector T cells.
]. Hence, understanding the dynamics of lymphocyte biology (phenotype, localisation, activation and recruitment) would accelerate the translation of basic research toward the identification of novel targets to limit cardiovascular inflammation.
4. The impact of cellular heterogeneity in atherosclerosis on future research and therapies
The identification of these highly plastic and heterogeneous cells in the atherosclerotic plaque adds an extra layer of complexity to future atherosclerosis research. Indeed, as vascular wall cells can express cell markers conventionally used to identify macrophages, and vice versa, the use of a cell subset-specific marker for histological analyses of plaques is no longer valid to assess the cellular plaque composition, nor to make any statements about the degree of plaque stability. Instead, cells can be identified using lineage tracing tools in which cells and their progeny are permanently labelled in combination with their transcriptional or proteomic signature.
Moreover, the complex and heterogeneous nature of the plaque calls for novel strategies in basic research to identify druggable targets for atherosclerosis. Although single cell transcriptomics have been instrumental in mapping the cellular networks in the plaque, additional information on cell activation and intercellular interactions are necessary to identify the most impactful drivers of the disease. A next step would be to implement new single cell technologies and bioinformatic tools (see section 2.2) allowing to fully address the functional complexity of the cells present in the plaque micro-environment. The integration of cell-specific fate mapping and single cell omics with human genetics will further help to establish a causal relationship between cellular heterogeneity and atherosclerosis.
Furthermore, the technological advancements in lipidomics and metabolomics may further help to pinpoint the disease players at a molecular level. For example, matrix-assisted laser desorption/ionisation (MALDI) mass spectrometry imaging (MSI) can unveil the lipid profiles of specific regions in the plaque; the use of this technique in symptomatic plaques has shown that macrophage-rich regions are enriched in sphingomyelins, while intimal vascular smooth muscle cells are enriched in cholesterol and cholesteryl esters [
]. Thus, the implementation of lipidomics and metabolomics would also unveil how metabolic adaptations, that occur in the diseased arteries, contribute to cellular plasticity and heterogeneity of the atherosclerotic plaque [
]. Therefore, combination of the phenotype with the energetic preference of a cell in a specific region of the plaque would depict a specific local immune-metabolic response that may support the design of a delivery system for “metabolic drug” at the atherosclerotic lesion. Furthermore, this technology would be particularly useful in investigating the spatial immune cell reactivity by increasing the resolution to single cell level. For instance, combination of MALDI MSI with high-throughput technologies (such as single cell-based B and T cell receptor sequencing and transcriptomics) would combine the specific phenotypic profile and antigen-specificity of lymphocytes against ASA with the peculiar milieu of different areas - characterised by an unique composition of lipids, peptides or metabolites - in the arterial lesion. This approach would help to profile the mechanism by which the adaptive immune response is elicited in atherosclerosis. By revealing if and to what extent the effect of immunoglobulins and T cell activation is mediated via recognition of ASA could pave the way for the development of antigen-specific therapeutic management of atherosclerosis.
Despite the ongoing global burden of cardiovascular disease, investment in cardiovascular drug development has stagnated in the last 2 decades, with a relative underinvestment compared to other areas of research such as oncology [
]. Within the atherosclerosis field and, beyond the lipid compartment, we dare to adjudge there is a lack of innovative drugs. There is a growing interest however for treating inflammation to reduce major coronary events. Despite clear evidence of the causality of specific immune subsets in atherogenesis, further investigation is required to pave the road to target inflammatory responses driven by the impaired immune cell activation. We envision future therapies aimed at either targeting specific immune cell subsets (although this may hold certain health risks such as susceptibility for infections) or at targeting their effectors (e.g. secreted antibodies, ASA). The knowledge from the single cell studies that is available today is insufficient to implement the second approach. As discussed above, more insights in the immune mechanisms through advanced techniques at the protein level are needed, to develop therapeutic strategies against ASA. Apart from targeting circulating mediators (such as cholesterol and cytokines), it is time the field explores the possibility to develop site-specific therapies for atherosclerosis. With the emergence of spatial omics techniques in combination with protein profiling, we see a possibility for treating atherosclerosis in a more site-specific manner in the near future. The identification of ASA will allow us to engineer drugs in a way they can be targeted to a specific disease site. Although only used for cancer therapy to date, the identification of ASA may facilitate the implementation of chimeric antigen receptor (CAR) T cell therapy in atherosclerosis in which a patient's T cells are isolated and engineered to be specific to an antigen. After infusion, the CAR T cells will attack only the cells expressing the antigen whilst leaving the healthy cells unharmed. Targeting the chemokine receptor – ligand interaction, using miRNA- or siRNA-based therapies or heterodimer inhibition via peptide and non-peptide antagonists, might be another interesting approach to dampen immune cell activation. Given the rhythmic nature of cytokine release and the dynamics of myeloid cell recruitment, chrono-pharmacology based therapy may offer improved drug efficacy and lower the risk of side effects by timing-optimised therapeutic strategies [
]. Hence, given the different location and dynamic distribution of immune cells versus vascular wall cells, a different approach may be required to target ECs or VSMCs. Rather than targeting an entire cell population, there is support for targeting only subsets of cells – those that are considered detrimental for plaque stability – and/or trying to convert cells into a ‘plaque stabilising’ phenotype. With regard to ECs and VSMCs, we currently lack experimental proof of which phenotypes are truly harmful for plaque stability and which ones are not. Identifying the regulatory pathways that control phenotypic modulation may bring us closer to finding therapeutic options for converting ECs or VSMCs into an atheroprotective phenotype. To avoid the high costs of clinical trials, which is often due to the high failure rate of newly tested drugs, and to prevent the potential risks of adverse side effects, Mendelian randomisation offers another approach to search for future therapeutic targets for atherosclerosis. This has previously been applied for lipids [
Interleukin-6 Receptor Mendelian Randomisation Analysis, C The interleukin-6 receptor as a target for prevention of coronary heart disease: a mendelian randomisation analysis.
], and uses variants in a gene encoding for a specific drug target to profile the effects of pharmacological modification of this target and its likely effectiveness to treat or prevent CAD.
5. Conclusion
Single cell technologies have been shown to be a powerful tool to unravel the complex and heterogeneous cellular architecture of the atherosclerotic plaque. These advances have not only deepened our appreciation for the subtlety of atherogenesis, but have also challenged previous views on the plasticity of certain cell types, thus shaping future research directions. Single cell transcriptomics have been instrumental in mapping the cellular and molecular networks in the plaque, thereby providing a single cell atlas of human and experimental atherosclerosis. However, their most important achievement would be to facilitate the development of new therapeutic approaches to combat atherosclerosis. To accomplish this, we suggest moving away from conventional approaches for targeting circulating molecules that are not disease nor site-specific. Instead, we encourage the field to implement new technologies for studying cell activation and cell-cell interactions, allowing to fully capture the functional complexity of plaque cells. Only then, we will get closer to a more tailored therapeutic solution for atherosclerosis.
Financial support
FB was supported by the Cariplo Foundation [2019–1560] and Roche Foundation for Research 2021 (Italy). XYK was supported by the South‐Eastern Norway Regional Health Authority (No. 2018064) and European Commission Research grant (PainFact, H2020-SC1-2019-2-RTD-848099). JK was supported by the Dutch Heart Foundation (Senior Scientist Dekker grant (03-004-2021-T045)) and the European Research Council (ERC Starting grant (101076407)). DT is supported by the European Research Council (ERC Starting grant (101041206)), the Austrian Science Fund (I4963, P35233) and the European Research Area Network on Cardiovascular Diseases (ERA-CVD) (I4647). MG is supported by ERA-CVD (ENRICH) and the Research Foundation Flanders (FWO - G0H7220 N).
CRediT authorship contribution statement
Fabrizia Bonacina: Writing – original draft. Alessia Di Costanzo: Writing – original draft, Visualization, generation of images. Vadim Genkel: Writing – original draft. Xiang Yi Kong: Writing – original draft. Jeffrey Kroon: Writing – original draft, Visualization, generation of images. Ena Stimjanin: Writing – original draft. Dimitrios Tsiantoulas: Conceptualization, Writing – original draft. Mandy OJ. Grootaert: Conceptualization, Writing – original draft, Writing – review & editing, Visualization, generation of images.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
The figures in this manuscript were created with BioRender.com.
References
Becker L.M.
et al.
Deciphering endothelial heterogeneity in health and disease at single cell resolution: progress and perspectives.
Single-cell genomics reveals a novel cell state during smooth muscle cell phenotypic switching and potential therapeutic targets for atherosclerosis in mouse and human.
Single-cell analysis of fate-mapped macrophages reveals heterogeneity, including stem-like properties, during atherosclerosis progression and regression.
Extensive proliferation of a subset of differentiated, yet plastic, medial vascular smooth muscle cells contributes to neointimal formation in mouse injury and atherosclerosis models.
Single-cell T cell receptor sequencing of paired human atherosclerotic plaques and blood reveals autoimmune-like features of expanded effector T cells.
00138-7&id=ga1.jpg){kind=link}
00138-7&id=gr1.jpg){kind=link}
00138-7&id=gr2.jpg){kind=link}