Article

Identification of Macrophage-related Gene Signature Associated with Unstable Atherosclerotic Plaques through Integrative Multi- omics Analysis

https://doi.org/10.21203/rs.3.rs-3268699/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background: Macrophages play a crucial role in the occurrence and development of atherosclerosis. The purpose of this study is to identify critical candidate genes associated with unstable atherosclerotic plaques in macrophages

Methods: The single-cell sequencing dataset GSE159677 and microarray datasets GSE43292 and GSE100927 were downloaded from the Gene Expression Omnibus database. The intersection of differential analysis results from single-cell transcriptome sequencing data, microarray data, immune infiltration analysis, and WGCNA analysis results yielded macrophage-related genes. Least absolute shrinkage, and selection operator (LASSO) was utilized for screening characteristic genes among atherosclerotic plaque progression- and macrophage-related genes. ROC curves were produced to evaluate the diagnostic performance. Finally, pseudo-time analysis was used to obtain the expression patterns of key genes along the cell differentiation trajectory.

Results: Based on multiple omics analyses, we obtained 59 genes related to macrophages. Then, by using the LASSO algorithm to select features, we ultimately obtained five key genes. Using these five genes, we constructed a nomogram diagnostic model that demonstrated high diagnostic performance in both the training (AUC=0.882) and testing (AUC=0.985) datasets. Pseudo-time analysis shows that STAB1 is exclusively expressed in resident macrophages, while IL1RN expression gradually increases over time. DAB2 is highly expressed in both resident and foamy macrophages. NLRP3 expression is found in some cells of resident and inflammatory macrophages, but not at all in foam cells. HMOX1 expression varies little across different cell types.

Conclusion: Our study has identified a gene signature consisting of STAB1, NLRP3, IL1RN, HMOX1, and DAB2 through integrative multi-omics analysis, which may play a crucial role in the development and progression of atherosclerosis.

Health sciences/Cardiology/Cardiovascular biology/Cardiovascular genetics

Biological sciences/Computational biology and bioinformatics/Data mining

atherosclerosis

macrophage

single-cell sequencing

bioinformatics

Atherosclerosis is a chronic inflammatory condition of the arterial wall, characterized by the continuous presence of immune cells that are responsible for its initiation, progression, and maintenance through their influx, proliferation, and activation^[1–3]. Macrophages play a crucial and irreplaceable role in the development, growth, and rupture of atherosclerotic plaques by promoting inflammation and lipid infiltration, rather than other immune cells^[4–7]. During the initial stages of atherosclerosis, macrophages engulf and accumulate lipids, leading to the formation of foam cells. These foam cells contribute to the development of fatty streaks, which are the earliest visible signs of atherosclerosis^{[8, 9]}. As atherosclerosis progresses, macrophages continue to accumulate within the arterial wall and release pro-inflammatory cytokines, which contribute to the chronic inflammation characteristic of advanced atherosclerosis. Macrophages also contribute to the destabilization of atherosclerotic plaques, which can lead to acute cardiovascular events such as heart attack and stroke^[10].

Once monocytes have migrated into the arterial wall, they differentiate into macrophages and develop different functional phenotypes based on signals from the surrounding microenvironment^[11]. Macrophages can exhibit a pro-inflammatory phenotype (known as "M1-like "), which is characterized by high levels of pro-inflammatory cytokine production, reactive oxygen species, and reactive nitrogen species. Alternatively, macrophages can exhibit a reparative phenotype (known as "M2-like")^[12], which is characterized by the ability to suppress inflammation and promote tissue repair. Atherosclerotic lesions in humans harbor a diverse population of macrophages, the functional features of which remain largely elusive^[13]. Unraveling the mechanisms underlying macrophage polarization in atherosclerosis could pave the way for novel therapeutic strategies to mitigate disease progression^[14].

Our study employed a multi-omics integration approach to identify critical candidate genes associated with unstable atherosclerotic plaques in macrophages. We investigated the expression patterns of these genes during various stages of macrophage differentiation, and also explored their potential roles in this process.

2.1 Data acquisition and preprocessing

The single-cell sequencing dataset GSE159677^[15] and microarray datasets GSE43292^[16] and GSE100927^[17] were downloaded from the Gene Expression Omnibus database. Samples for GSE159677 were collected from three patients with calcified atherosclerotic core (AC) plaques and patient-matched proximal adjacent (PA) segments. GSE43292 contained 32 carotid atherosclerotic samples and 32 carotid non-atherosclerotic samples, and GSE100927 included 69 atheromatous carotid plaques and 35 control carotid arteries without atherosclerotic lesions. Please refer to supplementary Table S1 for detailed information on the datasets. Microarray datasets were performed log2 conversion of gene expression and normalized the data using the limma^[18] package.

2.2 Single-Cell RNA Sequencing Data Analysis

Using the R package Seurat (version 4.0.5), each of the six single-cell samples was created as a Seurat object and merged using the merge function. Specifically, cells with < 200 measured genes and > 10% mitochondrial contamination were defined as low-quality cells and cells with > 4000 measured genes were identified as potential doublets. They were filtered out from each dataset. After being filtered, 33453 cells from AC samples and 10495 cells from PA samples were selected for following processes. The merged object was normalized (function NormalizeData, method = “LogNormalize,” scale. factor = 10,000), the 3000 most variable genes were identified, and the expression levels of these genes were scaled before performing PCA in variable gene space. Next, batch effect was corrected and merged object was integrated by running Harmony (version 1.0). The top 25 harmony dimensions were provided as an input for Uniform Manifold Approximation and Projection for Dimension Reduction (UMAP) and visualized the first two UMAP dimensions at a clustering resolution of 0.8. All steps were performed using functions implemented in the Harmony package and Seurat package (NormalizeData, FindVariableFeatures, ScaleData, RunPCA, FindNeighbours, FindClusters, RunUMAP) with default parameters, except where mentioned.

Next, distinct cell types were labeled by canonical marker genes such as AIF1, CD14, CD68 (Macrophages), VWF, PECAM1, ECSCR (Endothelial cells), CALD1, MYL9, TAGLN (VSMCs), NKG7, XCL1, CTSW (NK cells), CD2, TRAC, CD69 (T cells), CD79A, MS4A1, IGKC (B cells)^[15]. We performed differential analysis between each cell types of AC samples and PA samples using the FindMarkers function and |Log2FC|>=0.5 and adjP < 0.05 were used as the threshold to identify differentially expressed genes (DEGs). We named the differentially expressed genes between the AC and PA groups of macrophages as "MacDEGs" and used them for subsequent analysis. The results of difference analysis of each cell type were analyzed by Gene Set Enrichment Analysis (GSEA) using clusterProfile^[19] package.

2.3 Immune Infiltrate Analysis

The Gene Set Variation Analysis package (release 1.32.0)^[20] was utilized to employ a computational method of single-sample gene set enrichment analysis (ssGSEA). The method was used to analyze the enrichment scores of immune cell subsets from 29 species within the GSE43292 dataset.

2.4 WGCNA network construction

Based on the GSE43292 microarray dataset, the gene co-expression network was constructed by using “WGCNA” package^[21]. Before analysis, we used the hclust function in R language for hierarchical cluster analysis to identify anomalous samples. Appropriate soft powers β (ranging from 1 to 20) were selected using the "pickSoftThreshold" function in the WGCNA package according to the standard of a scale-free network. Next, an adjacency matrix was constructed using the genetic correlation matrix between the calculated soft threshold power β and all gene pairs calculated in Pearson analysis determined using the following formula: aij = |Sij|b, where aij is the adjacency matrix between gene i and gene j, Sij is the similarity matrix composed of the Pearson correlation coefficients of all gene pairs, and b is the soft power value. The topological overlap matrix and corresponding phase dissimilarity (1 - topological overlap matrix) were converted from the adjacency matrix. A hierarchical clustering dendrogram was constructed to classify genes showing similar expression levels into different modules. Finally, the expression profile of each module was summarized using the module signature genes, and the correlation between the module signature genes and the result of the immune infiltrate analysis was calculated. Therefore, we focused on modules with high correlation coefficients for the macrophage, and genes in the module blue(MEblue) with the highest correlation were selected for subsequent analysis. The WGCNA parameters were as follows: soft power β = 14, networkType = “signed,” minModuleSize = 30, mergeCutHeight = 0.25.

2.5 Functional Enrichment Analyses and Protein–Protein Interaction

To filter the differentially expressed genes (DEGs) in the GSE43292 dataset, the expression data of the normalized dataset was input into the R package "limma", with cut-off criteria set at (FDR < 0.05 and |log2 Fold Change(FC)| ≥ 0.5). We used the Venn diagram to intersect "MacDEGs", "DEGs", and "MEblue". Finally, we obtained 59 genes that are related to macrophages. Next, We conducted Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Ontology (GO) analyses to understand the functions of these 59 genes. We utilized the clusterProfiler^[19] package for enrichment analysis to determine the biological activities of these genes and their associated pathways. To analyze protein-protein interactions, a protein-protein interaction (PPI) network was established. An online tool, STRING^[22], was used to search for genes that interact with each other, and the interaction relationships among these 59 genes were determined. The PPI network was visualized using the Cytoscape software.

2.6 Construction of the Diagnostic Model

To create disease profiles associated with genes related to macrophage, we employed least absolute shrinkage and selection operator (LASSO) regression (lambda.min = 0.051) from mlr3 package^[23] for feature selection and constructed a diagnostic nomogram model utilizing the rms package. To assess the predictive power of our model, we generated calibration curves. Moreover, we utilized the R package pROC^[24] to generate ROC curves and calculate the AUC to evaluate the diagnostic effectiveness of both the model and the key genes Furthermore, we validated the nomogram by performing external validation using the GSE100927 dataset and decision curve analysis (DCA) curves to demonstrate its validity.

2.7 Macrophages Subgroup Definition and Pseudotime

We extracted macrophages from the AC samples, and performed integration, dimensionality reduction, and clustering analysis. Based on the markers expressed by these cells, we divided them into four subtypes of macrophages. To define the developmental trajectories of the four macrophage subsets, we employed monocle2 for predicting their developmental trajectories and also evaluated the expression patterns of hub genes during macrophage developmental trajectories.

2.8 Statistical Analysis

All statistical tests were performed using R software 4.1.1. Wilcoxon or Student’s t-test was used to analyze the differences between two groups. Correlations between variables were determined using Pearson or Spearman correlation tests. All statistical p-values were two-sided, and P < 0.05 was considered to indicate statistically significant results. Please refer to Supplementary Figure S1 for the flowchart of this study.

3.1 Single-cell RNA-seq profiling and clustering

Initially, we processed the single-cell sequencing data included in GSE159677 by integrating and performing quality control on the data from all six samples. Figure 1A shows the number of detected genes (nFeature_RNA), sequencing depth (nCount_RNA), and mitochondrial gene ratio (percent.mt) of each sample before data quality control. After filtering according to the mentioned quality control standards in the method, we obtained a total of 44,148 high-quality cells from 3 PA samples and 3 AC samples and a strong positive correlation was observed between nFeature_RNA and nCount_RNA with the Pearson's correlation of 0.94(Fig. 1B). We employed the UMAP method to perform cell clustering, which resulted in the identification of 22 distinct clusters at a resolution of 0.8 (Fig. 1C). The expression profile of marker genes for each cluster was examined subsequently (Fig. 1D). By analyzing the expression patterns of marker genes, we identified clusters 0, 2, 3, 14, and 16 as T cells, clusters 1, 13, 20, and 21 as endothelial cells (ECs), clusters 9 and 17 as B cells, clusters 4, 5, 11, 18, and 19 as vascular smooth muscle cells (VSMCs), clusters 6, 7, 10, 12, and 15 as macrophages, and cluster 8 as NK cells (Fig. 2A). We can observe from Fig. 2B that the cell composition of AC and PA samples differs greatly. The proportion of macrophages and T cells in AC samples is significantly higher than that in PA samples, while the proportion of VSMCs and ECs is significantly lower than that in PA samples. There are no significant changes in the proportions of B cells and NK cells. Figure 2C and 2D display the dot plot and heatmap, respectively, of marker genes for each subpopulation of cells.

3.2 Differential expression analysis and Gene Set Enrichment Analysis

Differential gene expression analysis identified a total of 76 differentially expressed genes in B cells, among which 33 genes were upregulated and 43 genes were downregulated. In the endothelial cells (ECs), 290 genes showed significant differential expression, comprising 154 upregulated and 136 downregulated genes. For macrophages, 255 genes exhibited significant differential expression, comprising 99 upregulated and 156 downregulated genes. In NK cells, a total of 46 differentially expressed genes were identified, with 15 genes upregulated and 31 genes downregulated. T cells showed a total of 48 differentially expressed genes, composed of 16 upregulated and 32 downregulated genes. Finally, there were 260 differentially expressed genes in vascular smooth muscle cells (VSMCs), among which 83 genes were upregulated and 177 genes were downregulated (Fig. 3A).

Further pathway analysis using GSEA revealed that signaling by interleukins was the major upregulated pathway in B cells, while the TGF-beta signaling pathway was downregulated (Fig. 3B). In ECs, the extracellular matrix organization pathway was significantly upregulated, while the interferon gamma signaling pathway was significantly downregulated (Fig. 3C). Moreover, macrophages showed a significant upregulation of the degradation of extracellular matrix pathway and downregulation of toll-like receptor (TLR1/TLR2) cascade pathway (Fig. 3D). For NK cells, the vesicle-mediated transport pathway was significantly upregulated, whereas the CD8 TCR downstream pathway was significantly downregulated (Fig. 3E). Similarly, T cells showed a significant upregulation of the vesicle-mediated transport pathway and downregulation of the AP1 pathway (Fig. 3F). Finally, in VSMCs, the smooth muscle contraction pathway was significantly upregulated, while the Wnt signaling pathway was significantly downregulated (Fig. 3G).

3.3 Immune Infiltrate and WGCNA

Based on the GSE43292 dataset, an immune infiltration analysis was conducted, and the results showed that, except for T cells, the infiltration abundance of other immune cells in AS samples was significantly higher than that in the control samples (Fig. 4A). Then, we proceeded to construct a gene co-expression network. To begin with, we constructed a clustering tree (as shown in Fig. 4B) and identified an outlier sample (GSM1060144), which was subsequently removed from the analysis. In order to ensure a scale-free network, we selected the power of β = 14 (scale-free R2 = 0.89) as the soft-thresholding parameter (Fig. 4C and D). Using dynamic tree-cutting analysis, we constructed 13 co-expression modules, each labeled with a different color. We conducted a Pearson correlation analysis between the results of immune infiltration analysis and these modules (Fig. 5A), and found that genes in MEblue had a strong positive correlation with macrophages (Cor = 0.94, P < 1e-200, Fig. 5B). In addition, we performed differential expression analysis on GSE43292 and identified a total of 1329 differentially expressed genes, including 747 upregulated genes and 582 downregulated genes (Fig. 5C). Only genes that overlapped across the results obtained via single-cell sequencing data analysis (MacDEGs), WGCNA analysis (MEblue), and differential expression analysis (DEGs) were considered significant macrophage-associated candidate genes related to plaque instability. Taking the intersection of genes identified from the three methods, we arrived at a final set of 59 macrophage-related genes (MRGs) for conducting further analysis (Fig. 5D).

3.4 Functional Enrichment Analysis and Protein–Protein Interaction Network Construction

To gain a more comprehensive understanding of the 59 MRGs, we employed GO annotation and KEGG pathway enrichment analyses. The GO biological process (BP) analysis revealed a significant enrichment of these genes in IL-1 beta production, indicating their involvement in inflammatory processes. The GO cell components (CC) analysis further showed that these genes are enriched in endocytic vesicles, vacuolar lumens, and other intracellular organelles where cellular waste is processed and removed. Furthermore, the GO molecular function (MF) analysis demonstrated enrichments of these genes in proteoglycan binding, cargo receptor activity, and other functions related to protein transport and extracellular matrix organization (Fig. 6A). Moreover, the KEGG analysis revealed that these genes are primarily associated with pathways related to inflammation, atherosclerosis, and cell apoptosis, suggesting their critical roles in regulating these physiological processes (Fig. 6B). Additionally, we constructed a protein-protein interaction (PPI) network of these genes to investigate their relationships with each other. Notably, CD68, MMP9, and CCL2 were found to occupy central positions within the entire network, indicating their critical roles in mediating interactions among the genes.

3.5 Feature Selection and Construction of the Diagnostic Model

To identify genes with diagnostic efficacy, we employed LASSO regression to screen the 59 candidate genes (Fig. 6D and E). As a result, we identified 5 hub genes (STAB1, NLRP3, IL1RN, HMOX1 and DAB2) with high diagnostic accuracy for AS. We evaluated the expression levels of these five genes in advanced and early groups, and found that the expression levels of these genes were significantly higher in the advanced group (Fig. 7A). A nomogram diagnostic model has been established using these five genes (Fig. 7B). To validate the predictive effectiveness of the model, we generated calibration curves, ROC curves, and DCA curves. The calibration curves demonstrated excellent agreement between the predicted and actual risk of AS, suggesting that our nomogram model for AS is highly accurate (Fig. 7C). The correctness of the model may also be confirmed using the ROC curve analysis (AUC = 0.88, Fig. 7D). The higher DCA curve (Fig. 7E) compared to the reference curve further confirms that our diagnostic model may have clinical value. We also assessed the independent diagnostic value of these 5 genes and observed that their ROC curves (Fig. 7F) had AUC values greater than 0.8, indicative of strong independent diagnostic potential. Furthermore, we conducted a validation analysis utilizing an independent dataset GSE100927 to confirm the validity of our findings (Fig. 7G, H and I). These findings suggest that these 5 genes may be involved in the occurrence and development of AS.

3.6 Identification of Macrophage Subtypes and Pseudo-time Analysis.

We extracted macrophages from AC samples and conducted a process of dimensionality reduction, clustering, and annotation. According to previous studies on macrophage subtypes, we annotated macrophages into four subtypes labeled Resident Macs, Foamy Macs, IL1B + Inflammatory Macs and TNF + Inflammatory Macs (Fig. 8A). Their marker genes^{[25, 26]} are shown in Fig. 8B. To further investigate the potential role of hub genes in macrophage differentiation, we conducted a pseudo-time analysis. One root and two branches were identified. Resident macrophages are concentrated at the root, while inflammatory macrophages and foam cells are concentrated in two branches, respectively (Fig. 8C, D and E). This result is consistent with our biological understanding that resident macrophages can differentiate into inflammatory macrophages and foam cells during the development of atherosclerosis. Figure 8F displays the expression patterns of hub genes in the pseudo-time. STAB1 expression is limited to resident macrophages, whereas IL1RN expression shows a gradual increase over time. DAB2 is highly expressed in both resident and foamy macrophages. NLRP3 is expressed in a subset of cells in resident and inflammatory macrophages but not at all in foam cells. HMOX1 expression does not vary significantly across the different cell types.

Although lipid-lowering therapy has led to some improvement, atherosclerosis still accounts for a significant number of deaths due to cardiovascular disease, underscoring the urgent need for identifying new potential target genes in clinical practice. Macrophages possess remarkable plasticity and a multitude of biological functions, which has prompted scientists to study their relationship with atherosclerosis for over 50 years, and significant discoveries are still being made today^[5]. Recent advances in single-cell RNA sequencing technology have allowed scientists to uncover the remarkable diversity of macrophage populations within atherosclerotic plaques^[27-30]. These subpopulations exhibit distinct gene expression and functional characteristics, and they play critical roles in the initiation, progression, stabilization, regression, and rupture of atherosclerotic plaques^{[31, 32]}. Therefore, it is essential to systematically explore the expression profile of macrophage-related genes and identify key genes to gain a profound understanding of the pathological features of atherosclerosis. In this study, we constructed a model composed of five macrophage-related genes (STAB1, NLRP3, IL1RN, HMOX1 and DAB2) based on multiple omics data, which showed highly accurate diagnostic performance. This may facilitate the progress of molecular diagnosis and immune-regulatory treatment of atherosclerosis.

STAB1 as a conserved scavenger receptor, it binds and uptakes LDL. However, a study has shown that the sole expression of Stab1 in macrophages is not sufficient to affect aortic atherosclerosis, and only Ldlr-KO mice with Stab1 deficiency limited to the hematopoietic system do not display changes in plaque development^[33]. In a recent study, it was found that mice with Stab1 knocked out throughout the body can significantly inhibit atherosclerosis. The mechanism is that after Stab1 knockout, the altered plasma proteome suppresses both patrolling and inflammatory monocytes, and thus systemically protects against atherogenesis^[34].

NLRP3 inflammasome, an inflammatory multiprotein complex, plays a pivotal role in the pathogenesis of atherosclerosis. Within the complex process of atherosclerosis, the activation of endothelial cells, macrophages, T lymphocytes and other cells leads to the secretion of pro-inflammatory cytokines, which in turn triggers inflammation reactions and further intensifies the progress of atherosclerosis^[35]. As an innate immune sensor, NLRP3 inflammasome is capable of detecting various detrimental stimuli, such as oxidative stress and cholesterol crystals, and subsequently inducing the secretion of potent pro-inflammatory cytokines, including IL-1β, IL-6, and IL-18, into the extracellular space, thereby promoting inflammation at the site of atherosclerotic lesions^[36].

DAB2 is a versatile adapter protein that participates in a variety of cellular functions, such as endocytosis, cell signaling, lipid uptake, cholesterol homeostasis, and cell adhesion^[37]. Previous study^[38] have shown that the expression of DAB2 is increased in M2 macrophages of both mice and humans, while it is suppressed in M1 macrophages, which is consistent with our analytical results. DAB2 modulates macrophage polarization and inflammatory signaling by binding to κ receptor-associated factor (TRAF6) and inhibiting the NF-TRAFB pathway. After knocking out Dab2 in mouse bone marrow, the level of LDL in the serum was significantly increased, as well as the expression of systemic inflammation and cytokines, leading to liver damage. However, its role in atherosclerosis is still unclear^{[39, 40]}.

HMOX1 exhibits a protective effect against atherosclerosis. Numerous studies have demonstrated that HMOX1 not only promotes the polarization of macrophages to the M2 subtype^[41], intensifying their phagocytic capability, but also inhibits pro-inflammatory responses and stimulates the secretion of IL-10^[42]. Additionally, HMOX1 enables macrophages to shield the tissues from oxidative damage^{[43, 44]}.

IL1RN suppresses the functions of interleukin 1, alpha (IL1A) and interleukin 1, beta (IL1B), and regulates multiple immune and inflammatory reactions related to interleukin 1, especially during the acute stage of infection and inflammation^[45]. Its anti-inflammatory effects may play a critical role in atherosclerosis^[46].

In addition, we subdivided macrophages into subpopulations and conducted a pseudo-temporal analysis to evaluate the trajectory of macrophages. M2 polarization appears to be the default choice for resident macrophages, but M2 and M1 macrophages can interconvert in response to specific environmental stimuli^[47-49]. According to our analysis results, promoting the expression of STAB1 and DAB1 while inhibiting the expression of IL1RN and NLRP3 may suppress M2-to-M1 polarization and thus play a role in anti-atherosclerosis.

Although our key genes have demonstrated accurate diagnostic ability and were validated using external datasets, some limitations need to be clarified. Firstly, the establishment and validation of the model were performed on small sample public datasets; large-scale validation and determination of the optimal cutoff value are needed. Secondly, the impact of key genes on macrophage polarization requires further in vivo and in vitro experiments for validation.

In conclusion, our study has identified a gene signature consisting of STAB1, NLRP3, IL1RN, HMOX1, and DAB2 through integrative multi-omics analysis, which may play a crucial role in the development and progression of atherosclerosis.

HANSSON G K, HERMANSSON A. The immune system in atherosclerosis [J]. Nature immunology, 2011, 12(3): 204-12.
FALK E. Pathogenesis of atherosclerosis [J]. Journal of the American College of Cardiology, 2006, 47(8 Suppl): C7-12.
LIBBY P, BURING J E, BADIMON L, et al. Atherosclerosis [J]. Nature reviews Disease primers, 2019, 5(1): 56.
DENG H, SUN Y, ZENG W, et al. New Classification of Macrophages in Plaques: a Revolution [J]. Current atherosclerosis reports, 2020, 22(8): 31.
TABAS I, BORNFELDT K E. Macrophage Phenotype and Function in Different Stages of Atherosclerosis [J]. Circulation research, 2016, 118(4): 653-67.
FLORES A M, HOSSEINI-NASSAB N, JARR K U, et al. Pro-efferocytic nanoparticles are specifically taken up by lesional macrophages and prevent atherosclerosis [J]. Nature nanotechnology, 2020, 15(2): 154-61.
FLYNN M C, PERNES G, LEE M K S, et al. Monocytes, Macrophages, and Metabolic Disease in Atherosclerosis [J]. Frontiers in pharmacology, 2019, 10: 666.
TABAS I. Macrophage death and defective inflammation resolution in atherosclerosis [J]. Nature reviews Immunology, 2010, 10(1): 36-46.
MOORE K J, TABAS I. Macrophages in the pathogenesis of atherosclerosis [J]. Cell, 2011, 145(3): 341-55.
WOOLLARD K J, GEISSMANN F. Monocytes in atherosclerosis: subsets and functions [J]. Nature reviews Cardiology, 2010, 7(2): 77-86.
HOEKSEMA M A, STöGER J L, DE WINTHER M P. Molecular pathways regulating macrophage polarization: implications for atherosclerosis [J]. Current atherosclerosis reports, 2012, 14(3): 254-63.
CHEN W, SCHILPEROORT M, CAO Y, et al. Macrophage-targeted nanomedicine for the diagnosis and treatment of atherosclerosis [J]. Nature reviews Cardiology, 2022, 19(4): 228-49.
DEPUYDT M A C, PRANGE K H M, SLENDERS L, et al. Microanatomy of the Human Atherosclerotic Plaque by Single-Cell Transcriptomics [J]. Circulation research, 2020, 127(11): 1437-55.
STöGER J L, GIJBELS M J, VAN DER VELDEN S, et al. Distribution of macrophage polarization markers in human atherosclerosis [J]. Atherosclerosis, 2012, 225(2): 461-8.
ALSAIGH T, EVANS D, FRANKEL D, et al. Decoding the transcriptome of calcified atherosclerotic plaque at single-cell resolution [J]. Communications biology, 2022, 5(1): 1084.
AYARI H, BRICCA G. Identification of two genes potentially associated in iron-heme homeostasis in human carotid plaque using microarray analysis [J]. Journal of biosciences, 2013, 38(2): 311-5.
STEENMAN M, ESPITIA O, MAUREL B, et al. Identification of genomic differences among peripheral arterial beds in atherosclerotic and healthy arteries [J]. Scientific reports, 2018, 8(1): 3940.
RITCHIE M E, PHIPSON B, WU D, et al. limma powers differential expression analyses for RNA-sequencing and microarray studies [J]. Nucleic acids research, 2015, 43(7): e47.
YU G, WANG L G, HAN Y, et al. clusterProfiler: an R package for comparing biological themes among gene clusters [J]. Omics : a journal of integrative biology, 2012, 16(5): 284-7.
HäNZELMANN S, CASTELO R, GUINNEY J. GSVA: gene set variation analysis for microarray and RNA-seq data [J]. BMC bioinformatics, 2013, 14: 7.
LANGFELDER P, HORVATH S. WGCNA: an R package for weighted correlation network analysis [J]. BMC bioinformatics, 2008, 9: 559.
SZKLARCZYK D, GABLE A L, NASTOU K C, et al. The STRING database in 2021: customizable protein-protein networks, and functional characterization of user-uploaded gene/measurement sets [J]. Nucleic acids research, 2021, 49(D1): D605-d12.
SONABEND R, KIRáLY F J, BENDER A, et al. mlr3proba: An R Package for Machine Learning in Survival Analysis [J]. Bioinformatics (Oxford, England), 2021, 37(17): 2789-91.
ROBIN X, TURCK N, HAINARD A, et al. pROC: an open-source package for R and S+ to analyze and compare ROC curves [J]. BMC bioinformatics, 2011, 12: 77.
VALLEJO J, COCHAIN C, ZERNECKE A, et al. Heterogeneity of immune cells in human atherosclerosis revealed by scRNA-Seq [J]. Cardiovascular research, 2021, 117(13): 2537-43.
WILLEMSEN L, DE WINTHER M P. Macrophage subsets in atherosclerosis as defined by single-cell technologies [J]. The Journal of pathology, 2020, 250(5): 705-14.
MCQUEEN L W, LADAK S S, ABBASCIANO R, et al. Next-Generation and Single-Cell Sequencing Approaches to Study Atherosclerosis and Vascular Inflammation Pathophysiology: A Systematic Review [J]. Frontiers in cardiovascular medicine, 2022, 9: 849675.
SLENDERS L, TESSELS D E, VAN DER LAAN S W, et al. The Applications of Single-Cell RNA Sequencing in Atherosclerotic Disease [J]. Frontiers in cardiovascular medicine, 2022, 9: 826103.
WINKELS H, WOLF D. Heterogeneity of T Cells in Atherosclerosis Defined by Single-Cell RNA-Sequencing and Cytometry by Time of Flight [J]. Arteriosclerosis, thrombosis, and vascular biology, 2021, 41(2): 549-63.
FERNANDEZ D M, RAHMAN A H, FERNANDEZ N F, et al. Single-cell immune landscape of human atherosclerotic plaques [J]. Nature medicine, 2019, 25(10): 1576-88.
WANG Y, WANG Q, XU D. New insights into macrophage subsets in atherosclerosis [J]. Journal of molecular medicine (Berlin, Germany), 2022, 100(9): 1239-51.
WILLIAMS J W, GIANNARELLI C, RAHMAN A, et al. Macrophage Biology, Classification, and Phenotype in Cardiovascular Disease: JACC Macrophage in CVD Series (Part 1) [J]. Journal of the American College of Cardiology, 2018, 72(18): 2166-80.
NAHON J E, HOEKSTRA M, VAN HULST S, et al. Hematopoietic Stabilin-1 deficiency does not influence atherosclerosis susceptibility in LDL receptor knockout mice [J]. Atherosclerosis, 2019, 281: 47-55.
MANTA C P, LEIBING T, FRIEDRICH M, et al. Targeting of Scavenger Receptors Stabilin-1 and Stabilin-2 Ameliorates Atherosclerosis by a Plasma Proteome Switch Mediating Monocyte/Macrophage Suppression [J]. Circulation, 2022, 146(23): 1783-99.
GREBE A, HOSS F, LATZ E. NLRP3 Inflammasome and the IL-1 Pathway in Atherosclerosis [J]. Circulation research, 2018, 122(12): 1722-40.
BERGSBAKEN T, FINK S L, COOKSON B T. Pyroptosis: host cell death and inflammation [J]. Nature reviews Microbiology, 2009, 7(2): 99-109.
CUI K, DONG Y, WANG B, et al. Endocytic Adaptors in Cardiovascular Disease [J]. Frontiers in cell and developmental biology, 2020, 8: 624159.
ADAMSON S E, GRIFFITHS R, MORAVEC R, et al. Disabled homolog 2 controls macrophage phenotypic polarization and adipose tissue inflammation [J]. The Journal of clinical investigation, 2016, 126(4): 1311-22.
ADAMSON S E, POLANOWSKA-GRABOWSKA R, MARQUEEN K, et al. Deficiency of Dab2 (Disabled Homolog 2) in Myeloid Cells Exacerbates Inflammation in Liver and Atherosclerotic Plaques in LDLR (Low-Density Lipoprotein Receptor)-Null Mice-Brief Report [J]. Arteriosclerosis, thrombosis, and vascular biology, 2018, 38(5): 1020-9.
TAO W, MOORE R, MENG Y, et al. Endocytic adaptors Arh and Dab2 control homeostasis of circulatory cholesterol [J]. Journal of lipid research, 2016, 57(5): 809-17.
NAITO Y, TAKAGI T, HIGASHIMURA Y. Heme oxygenase-1 and anti-inflammatory M2 macrophages [J]. Archives of biochemistry and biophysics, 2014, 564: 83-8.
CAMPBELL N K, FITZGERALD H K, DUNNE A. Regulation of inflammation by the antioxidant haem oxygenase 1 [J]. Nature reviews Immunology, 2021, 21(7): 411-25.
BOYLE J J, JOHNS M, KAMPFER T, et al. Activating transcription factor 1 directs Mhem atheroprotective macrophages through coordinated iron handling and foam cell protection [J]. Circulation research, 2012, 110(1): 20-33.
OROZCO L D, KAPTURCZAK M H, BARAJAS B, et al. Heme oxygenase-1 expression in macrophages plays a beneficial role in atherosclerosis [J]. Circulation research, 2007, 100(12): 1703-11.
DINARELLO C A. Biologic basis for interleukin-1 in disease [J]. Blood, 1996, 87(6): 2095-147.
RIDKER P M. Anticytokine Agents: Targeting Interleukin Signaling Pathways for the Treatment of Atherothrombosis [J]. Circulation research, 2019, 124(3): 437-50.
COLIN S, CHINETTI-GBAGUIDI G, STAELS B. Macrophage phenotypes in atherosclerosis [J]. Immunological reviews, 2014, 262(1): 153-66.
KIM K, SHIM D, LEE J S, et al. Transcriptome Analysis Reveals Nonfoamy Rather Than Foamy Plaque Macrophages Are Proinflammatory in Atherosclerotic Murine Models [J]. Circulation research, 2018, 123(10): 1127-42.
HU Q, LYON C J, FLETCHER J K, et al. Extracellular vesicle activities regulating macrophage- and tissue-mediated injury and repair responses [J]. Acta pharmaceutica Sinica B, 2021, 11(6): 1493-512.

No competing interests reported.

Version 1

posted

You are reading this latest preprint version