A vascular biology network model focused on inflammatory processes to investigate atherogenesis and plaque instability
- Héctor De León†1Email author,
- Stéphanie Boué†1,
- Walter K Schlage2,
- Natalia Boukharov3,
- Jurjen W Westra3,
- Stephan Gebel2,
- Aaron VanHooser3,
- Marja Talikka1,
- R Brett Fields3,
- Emilija Veljkovic1,
- Michael J Peck1,
- Carole Mathis1,
- Vy Hoang3,
- Carine Poussin1,
- Renee Deehan3,
- Katrin Stolle1,
- Julia Hoeng1 and
- Manuel C Peitsch1
© De León et al.; licensee BioMed Central Ltd. 2014
Received: 28 February 2014
Accepted: 9 June 2014
Published: 26 June 2014
Numerous inflammation-related pathways have been shown to play important roles in atherogenesis. Rapid and efficient assessment of the relative influence of each of those pathways is a challenge in the era of “omics” data generation. The aim of the present work was to develop a network model of inflammation-related molecular pathways underlying vascular disease to assess the degree of translatability of preclinical molecular data to the human clinical setting.
We constructed and evaluated the Vascular Inflammatory Processes Network (V-IPN), a model representing a collection of vascular processes modulated by inflammatory stimuli that lead to the development of atherosclerosis.
Utilizing the V-IPN as a platform for biological discovery, we have identified key vascular processes and mechanisms captured by gene expression profiling data from four independent datasets from human endothelial cells (ECs) and human and murine intact vessels. Primary ECs in culture from multiple donors revealed a richer mapping of mechanisms identified by the V-IPN compared to an immortalized EC line. Furthermore, an evaluation of gene expression datasets from aortas of old ApoE-/- mice (78 weeks) and human coronary arteries with advanced atherosclerotic lesions identified significant commonalities in the two species, as well as several mechanisms specific to human arteries that are consistent with the development of unstable atherosclerotic plaques.
We have generated a new biological network model of atherogenic processes that demonstrates the power of network analysis to advance integrative, systems biology-based knowledge of cross-species translatability, plaque development and potential mechanisms leading to plaque instability.
KeywordsVascular systems biology Plaque destabilization Vascular biology networks Computational modeling Atherosclerosis modeling
Evidence gathered from in vitro and in vivo experimental systems, as well as population-based observational studies, has led to the recognition of vascular inflammatory processes as central to all stages of atherogenesis, from local endothelial dysfunction to plaque development and rupture[1, 2]. Cigarette smoking has been epidemiologically established as a major risk factor for atherosclerosis and shown to promote plaque development in experimental animal models[3–5]. Mechanistically, endothelial dysfunction is thought to be a key initiating cellular event that results from a variety of pro-atherogenic stimuli including cigarette smoke (CS), dyslipidemia and oxidative stress[6–8].
Recent advances in high-throughput technologies have made the analysis of datasets from cardiovascular cells and tissues possible. Current challenges in the analysis of transcriptomics datasets based on functional annotation or pathway maps (e.g. Gene ontology, KEGG)[10, 11] reside on the forward reasoning assumption that differential expression of genes is directly related to differential protein activity. The variable relationship of mRNA to protein activity due to post-transcriptional, translational and protein and mRNA degradation regulation[12–14] may lead to misinterpretation of gene expression data. Reverse Causal Reasoning (RCR), a backward computational reasoning methodology, uses observed differential expression of genes in datasets to reverse-formulate mechanistic explanations (termed hypotheses [HYPs]) of the observed effects. RCR uses a large database structure of experimentally-driven causal observations (Selventa Knowledgebase, [SK]) as a substrate for reasoning and HYP generation. Subsequent mapping of HYPs to network models that recreate the biology of interest (e.g., atherogenesis) offers a mechanistically integrated evaluation and interpretation of gene expression data captured in large datasets. Combining prior knowledge from published literature with large “omics” datasets (e.g., transcriptomics) into in silico network models accelerates the data interpretation process and our understanding of cellular behavior. Unlike direct network mapping of gene expression data, a network-based HYP evaluation approach allows translating experimentally-determined molecular changes as measurable network perturbations that can be compared between different datasets.
We have previously reported the construction of five network models relating cellular stress, proliferation, DNA damage, autophagy, cell death and senescence, lung inflammation, and tissue repair and angiogenesis in lung and vascular tissues[16–20]. The present work describes the construction and application of the Vascular Inflammatory Processes Network (V-IPN), a network model that combines a molecular framework constructed from publicly available literature and enhanced with RCR data-derived mechanisms, to depict a broad range of inflammatory processes known to occur in vascular tissue during atherosclerotic disease progression. The V-IPN also describes the mechanisms leading to plaque instability, an event in plaque development that often leads to fatal myocardial or cerebral infarction as a result of plaque rupture and vessel occlusion. We used the V-IPN network to assess the degree of biological mechanistic coverage from four different sets of transcriptomics profiling data derived from multiple atherosclerotic-relevant contexts including human endothelial cells (ECs) in culture, coronary arteries from coronary artery disease (CAD) patients and aortas from ApoE-/- mice. The systemic inflammatory status of ApoE-/- mice, a well-established model of atherosclerosis, makes this strain an ideal model in which to study comorbidities associated to cigarette smoking. Our results indicate that the V-IPN captures the key biological mechanisms that underlie the progression of vascular disease in various cellular and tissue contexts and allows for a comprehensive interrogation of transcriptomics datasets related to atherogenesis and cross-species translatability.
Selventa Knowledgebase (SK)
Proportion of vascular-specific evidence statements for the V-IPN subnetworks
Edges >1 evidence annotated with vascular context
Vascular edges/Total (%)
Endothelial Cell Activation
Smooth Muscle Cell Activation
Endothelial Cell – Monocyte Interaction
Foam Cell Formation
RCR-based HYP generation process
The RCR methodology utilized for network augmentation has been described previously and a detailed description may be found in the supplementary methods (Additional file2). Briefly, RCR analysis identifies potential HYPs for the statistically significant mRNA State Changes observed in the transcriptomics datasets. These upstream controllers are termed HYPs, as they represent statistically significant hypotheses that are potential explanations for the observed mRNA State Changes (Figure 1C). Detailed descriptions of the probabilistic scoring metrics (richness and concordance) can be found in Catlett et al., whereas the use of causal assertions in the construction of the V-IPN are further described in the Additional file2: Supplementary Methods.
V-IPN construction: model structure and boundaries
The workflow for the creation of the V-IPN is illustrated in Figure 1A. The initial literature-based network scaffold was defined by specific cell, tissue, species and disease contexts (e.g., ECs, aorta, human and atherosclerosis) known to be implicated in vascular pathobiology. The V-IPN nodes and edges comprising the scaffold were assembled in a sequential process by first using causal connections derived from knowledge published in the scientific literature and captured by the SK (Figure 1B).
The literature-derived framework was further augmented with nodes derived from the RCR analysis of vascular inflammation transcriptomics datasets (referred to as “model building” datasets, Table 1). RCR analysis yielded several dozen additional HYPs that were vetted for biological relevance and incorporated into the network as new nodes. Such nodes were connected to the literature scaffold using causal relationships captured by the SK. The resulting integrated network was manually reviewed by scientists with expertise in vascular biology and inflammation. The modular framework consists of six subnetworks that accompany this manuscript in XGMML and .XLS formats (Additional file3). The network architecture may be viewed from the XGMML files using freely available network visualization software such as Cytoscape (http://www.cytoscape.org/).
Gene expression datasets used for V-IPN construction and evaluation
Datasets analyzed by RCR for V-IPN augmentation and evaluation
Model Building Datasets
16912112 (Gargalovic, 2006)
oxPAPC (40 μg/ml)
19139167 (Grabner, 2009)
32 wk of age
Aortic morphometry, IHC, FACS
Wild-type mice (C57BL/6 J)
78 wk of age
19279231 (Mattaliano, 2009)
HAECs (cell line)
20170901 (Romanoski, 2010)
Ox-PAPC (40 μg/ml)
19997623 (Hagg, 2009)
66 ± 8 yr of age
Angiography, blood cytokines
Paired unaffected artery (IMA)
in vivo (ApoE-/-)
13-16 wk of age; 30 d exposure
Fresh air exposure
Four transcriptomics datasets from isolated human ECs (GSE13139 [Hs_EC_GFP_oxLDL_vs_ct]) and atherosclerotic human coronary arteries (GSE40231 [Hs_athCA_vs_ctIMA]), as well as murine aortas (E-MTAB-1696 [Mm_Ao_16w_ApoE_CS_vs_sham]), were analysed by RCR. RCR results (HYPs) were then used to evaluate network performance by determining HYP-level coverage and odds ratios (OR) across the six subnetworks constituting the V-IPN. Names describing the species and experimental settings for each dataset were created and they were used throughout the results and discussion section to facilitate comparative analyses.
Gene expression datasets used as negative controls
Datasets analyzed by RCR used as negative controls
Exposure to CDK4/6 inhibitor
24 exposure + 8 h after removal of inhibitor
FACS, cell cycle analysis
Primary cardiac microvascular ECs
24 h hypoxia (1% O2)
Normoxia (21% O2)
Primary human pulmonary microvascular ECs
24 h hypoxia (1% O2)
Normoxia (21% O2)
Jurkat T cells
Arsenic trioxide 3 μM
6 h exposure
Calculation of coverage and odds ratio
CS generation and ApoE-/- mice exposure
We set up a study in ApoE-deficient mice in which we investigated the effects of CS on cardiovascular endpoints including plasma lipid profiles, and transcriptomics of aortas. The E-MTAB-1696 [Mm_Ao_16w_ApoE_CS_vs_sham] transcriptomics dataset was generated from aortas displaying evidence of atherosclerotic plaques in ApoE-/- mice exposed to cigarette smoke (CS). All animal experimental procedures and CS exposure were approved by an Institutional Animal Care and Use Committee (IUCAC) and are described in detail in the Additional file2: Supplementary methods. Total cholesterol measurements in plasma, atherosclerotic plaque measurements in the aortic arch and immunohistochemical stainings in vascular tissues of ApoE-/- mice were conducted according to methods detailed in the Addditional file2: Supplementary methods
V-IPN construction and biological integration: description of modular framework and boundaries
Assessment of V-IPN subnetwork-level HYP coverage
Summary statistics of dataset overlapping across the six V-IPN subnetworks
N° of state changes (SC)
N°of HYPs in dataset
NoHYPs overlapping withEC activation
N° HYPs overlapping withplatelet activation
N° HYPs overlapping withEC-monocyte interaction
N° HYPs overlapping withfoam cell formation
N° HYPs O/L with SMC Activation
N° HYPs O/L with Plaque Destabilization
N° HYPs O/L with V-IPN
Total N° of possible HYPs in each subnetwork for human/mouse
Hs_athCA_vs_ctIMA exhibited statistically significant ORs for all of the six subnetworks and the highest coverage of all datasets (except Mm_Ao_78w_ApoE_vs_wt, a model construction dataset) in four of the six subnetworks: Plaque Destabilization, Platelet Activation, EC-Monocyte Interaction and Foam Cell formation. This remarkable finding underlines the value of contrasting gene expression datasets from atherosclerotic arteries (e.g. coronary arteries) with normal vessels from the same subjects (e.g., internal mammary arteries) as performed for this dataset. Interestingly, the murine dataset derived from ApoE-/- aortas of old mice (Mm_Ao_78w_ApoE_vs_wt) displayed a very similar OR pattern to the human dataset across all subnetworks, with the exception of the Platelet Activation subnetwork, which indicates that largely similar biological pathways underlie atherosclerosis in both species (Figure 4). This result also suggests that the process of platelet activation may play a larger role in the development of atherosclerotic plaques in humans compared to advanced-age murine models of atherosclerosis.
Negative Control Datasets
Transcriptomics data from four datasets were used as negative controls to evaluate the specificity of the V-IPN. Hs_NHBE_CDKinh_rel_vs_blk_8h, a negative control dataset obtained from NHBE cells, exhibited a low degree of coverage across all subnetworks (Figure 4). Significant HYPs for Hs_NHBE_CDKinh_rel_vs_blk_8h in the EC Activation and SMC Activation subnetworks were mostly related to cell cycle and growth factor signaling molecules, which are ubiquitously represented across cell cycle and growth factor subnetworks of the Cell Proliferation Network (Additional file6: Table S1). Three additional negative control datasets mapped to the V-IPN rendered even lower degrees of coverage compared to NHBE cells (Figure 4). Hs_MVEC-C_hpx_vs_ct_24h and Hs_MVEC-L_hpx_vs_ct_24h, two datasets obtained from cardiac and lung MVECs subjected to hypoxia, exhibited lower coverage than the dataset from NHBE cells across all subnetworks. Lung MVECs (Hs_MVEC-L_hpx_vs_ct_24h) showed a slight degree of coverage in the EC activation, SMC Activation and the Plaque destabilization subnetworks. Hs_JurkT_ars_vs_ct, a control dataset from Jurkat cells, exhibited the lowest degree of coverage of all datasets examined. The low HYP coverage displayed by the negative control datasets from studies using NHBE cells, MVEC and Jurkat cells further demonstrates the specificity of the biology captured by the V-IPN and highlights its value to evaluate processes proximal to vascular immunopathology.
HYP scoring and HYP directionality
The HYP coverage of Hs_athCA_vs_ctIMA was examined not only across all V-IPN subnetworks, but also across all subnetworks and models we have previously published[16–20]. A bar plot visualization of all overlapping networks is depicted in Additional file9: Figure S4 as coverage (bar length) and OR (grey color intensity). Subnetworks displaying the largest coverage and highest ORs are within the IPN (Inflammatory Process Network), TRAG (Tissue Repair and Angiogenesis) and DACS (DNA damage, Autophagy, Cell death, and Senescence) models and include Dendritic Cell Migration, Neutrophil Chemotaxis, Natural Killer (NK) Cell Activation, Epithelial Cell Barrier Defense, Macrophage activation, Immune Regulation of Angiogenesis and MAP kinases (Mapk). Many of these subnetworks constitute biological processes that have also been implicated in the development of atherosclerotic lesions.
V-IPN evaluation of preclinical data translatability
V-IPN coverage of predicted HYPs from human in vitro datasets
To investigate the ability of the V-IPN to distinguish the effects of different experimental perturbations, we compared the predicted HYPs from three sets of transcriptomics data from HAECs stimulated with Ox-LDL (Hs_EC_GFP_oxLDL_vs_ct, Hs_EC_LOX1_oxLDL_vs_ct) or Ox-PAPC (Hs_EC_oxPAP_vs_ct). The largest HYP coverage by a single dataset was observed with Hs_EC_oxPAP_vs_ct (40 HYPs, 7-15% across all subnetworks), followed by Hs_EC_LOX1_oxLDL_vs_ct (14 HYPs, 2-7%) and Hs_EC_GFP_oxLDL_vs_ct (5 HYPs, 0-3%) (Additional file10: Table S2). Significant HYPs observed to be shared between the three datasets included inflammatory molecules INFB1, IFNG and IL17A. Predicted HYPs known to be transcriptional modulators involved in lipid metabolism of biomembranes (CREB1, SREBF1 and SREBF2) were also commonly observed in the three datasets. Additional significant HYPs in Hs_EC_oxPAP_vs_ct reflects a group of growth factors and cell cycle controllers (PDGF, IGF1, CDK4, CCND1, CDKN1A), inflammatory cytokines and chemokines (CCL2, CCL5, CD40LG), oxidative stress-related molecules (NOS3, SOD1), transcriptional regulators of mitogenesis and inflammation (ATF4, NFKB, SP1), and molecules driving cytoplasmic and intra-organelle signaling events leading to migration, proliferation and cell death (AKT, MAPK8, PKA). The results of this coverage analysis suggest that treatment with Ox-PAPC may be a more potent inducer of processes related to atherogenesis when compared to oxLDL treatment, in the specific context of HAECs stimulated in vitro.
V-IPN coverage of predicted HYPs from human and murine in vivo datasets
To test the power of the V-IPN at capturing mechanisms and biological pathways implicated in advanced vascular lesion development in human arteries, we evaluated coverage across the V-IPN subnetworks with a gene expression dataset from human coronary arteries isolated from CAD patients undergoing bypass surgery (Hs_athCA_vs_ctIMA). Lesion stage- and species-specific mechanistic differences were assessed by comparing the HYP coverage of each V-IPN subnetwork by the human dataset with two murine aortic datasets from early (Mm_Ao_16w_ApoE_CS_vs_sham) and advanced (Mm_Ao_78w_ApoE_vs_wt) atherosclerosis. Aortic tissue was collected from sexually mature 13–16 week-old ApoE-/- mice following 30 days of CS exposure (Mm_Ao_16w_ApoE_CS_vs_sham) and from 78 week-old, unexposed ApoE-/- mice (Mm_Ao_78w_ApoE_vs_wt). The Mm_Ao_78w_ApoE_vs_wt dataset was used in this evaluation as a reference for stage- and species-specific comparisons with the Mm_Ao_16w_ApoE_CS_vs_sham and human Hs_athCA_vs_ctIMA datasets, respectively.
Dataset HYP overlapping between murine and human vascular tissues across the V-IPN subnetworks
No of HYPs
Foam cell formation
Early vs advanced murine atherosclerosis datasets
Predicted HYPs that were common to the mouse datasets from young and old mice (Mm_Ao_16w_ApoE_CS_vs_sham and Mm_Ao_78w_ApoE_vs_wt) indicated a low degree of overlapping across the V-IPN subnetworks (0-3%, Table 5). Coverage analysis showed some mechanisms and molecules being activated solely in the Mm_Ao_16w_ApoE_CS_vs_sham dataset including PPARA, CD40LG, RAC1, PGE2 and SREBF2. Two predicted HYPs were found in four out of the six subnetworks (decreased PPARA and increased CD40LG) including the Plaque Destabilization subnetwork (Figure 5). AGTR1A (angiotensin II receptor 1A) was also a predicted HYP shared by both datasets in three subnetworks. These results indicate that in addition to a small set of mechanisms shared by early and advanced murine lesions, distinct biological pathways may also contribute to lesion formation in CS exposure-induced early vascular lesions compared to those implicated in advanced, older lesions.
Advanced murine vs. advanced human atherosclerosis datasets
In contrast to the murine dataset comparison results, the degree of overlapping HYPs between Mm_Ao_78w_ApoE_vs_wt and Hs_athCA_vs_ctIMA ranged from 10-16% (Table 5), except for the SMC Activation subnetwork (5%). Many significant HYPs were shared between these two datasets within the Plaque Destabilization subnetwork, (Figure 5A). Significant HYPs related to vascular pathobiology were mapped to the V-IPN; they included the processes of inflammation, angiogenesis, and monocyte and macrophage differentiation. Common HYPs included the canonical transcriptional regulators AP1, IRF3, REL and SPI1, nuclear receptors and signal transducers (e.g., NCOR1, STAT1), growth factors (e.g., VEGFA), and chemoattractants (e.g., CCL5). Cytokines involved in the differentiation and function of macrophages and lymphocytes CSF1, CSF2, IFNG, IL1B and IL6, were also shared between the two datasets. A gene functional clustering analysis of common HYPs using DAVID (http://david.abcc.ncifcrf.gov/home.jsp) revealed inflammation, cytokine activity, chemotaxis and the toll-like receptor signaling pathways as the top ranked functional categories (Additional file11: Table S3). This HYP coverage analysis indicates that a shared repertoire of biological mechanisms underlies the development of advanced vascular lesions in both humans and mice.
Early murine vs. advanced human atherosclerosis
A comparison between Mm_Ao_16w_ApoE_CS_vs_sham and Hs_athCA_vs_ctIMA revealed a low degree of overlapping HYPs, ranging from 0-7% spanning all subnetworks (Table 5). Furthermore, coverage analysis within the Plaque Destabilization subnetwork indicated only a few HYPs being shared between these two datasets (increased CCL2, increased “response to hypoxia”, and increased TGFB1) (Figure 5A). Only two additional HYPs, HIF1A and PPARD, were shared by both datasets in other subnetworks. This analysis suggests that despite a comparable atherosclerotic plaque morphology, the molecular events leading to its development in a CS exposure murine model are quite distinct from the molecular pathways leading to plaque formation and destabilization in advanced human lesions.
V-IPN-generated functional causal paths in atherosclerotic coronary arteries of CAD patients
Hs_athCA_vs_ctIMA dataset coverage analysis showed Platelet Activation, Plaque Destabilization, Foam Cell Formation and EC-Monocyte Interaction as the V-IPN subnetworks with the highest HYP coverage (Figure 4). In contrast, SMC Activation exhibited the lowest coverage. V-IPN mapping of significant HYPs reflected a rich set of biological functions and molecules potentially involved in the pathophysiology of advanced, unstable, atherosclerotic lesions. Some of the most relevant groups are delineated below.
A series of significant HYPs related to lipid metabolism were captured by the V-IPN from Hs_athCA_vs_ctIMA. They included ABCG1, LDLR, Ox-LDL and Ox-HDL.
Platelet function and angiogenesis
Platelet-related HYPs included the thrombin receptor (F2R) and angiopoietin 1 (ANGPT1). F2R is involved in the regulation of the thrombotic response, whereas platelet release of ANGPT1 following platelet activation may be related to a role of platelets in maintaining vascular stability. ANGPT1 has roles in vascular development and may be involved in neovascularization within the plaque.
A number of signal transduction molecules including MAP kinases (MAPK1, MAPK3, MAPK8, MAP2K4, MAP2K6, MAP3K5), PI3K as well as transcription factors (CEBPA, EGR1, GATA6) were all predicted as significant HYPs, confirming that multiple signaling pathways account for the behavior of the various cell types present in the diseased atherosclerotic milieu.
Pathobiological content of the V-IPN
Early mechanisms in vascular disease development such as arterial cell dysfunction are amenable to controlled experimental perturbations in animal models or in vitro settings. In contrast, advanced atherosclerotic lesions are challenging to recreate experimentally, which has led to a paucity of sound data on the precise cellular and molecular mechanisms leading to plaque instability and eventual rupture. Disease modelling approaches, such as the implementation of the V-IPN reported here, overcome these barriers by integrating current knowledge and large gene expression datasets into networks reflecting the pathobiology of interest. Each new dataset mapping on the network has the potential to expand our mechanistic knowledge and further refine the network’s structure and content.
Impaired endothelial production of prostacyclin (PGI2) and nitric oxide (NO) in early arterial lesions facilitate vasoconstriction, inflammation and oxidative stress at a time when no morphological changes in the vessel wall have occurred. Cell adhesion molecules (CAMs) are well represented in the V-IPN as they have been extensively documented in the migration of inflammatory cells from the vascular lumen to the subendothelial space. In the presence of hypercholesterolemia and a local excess of reactive oxygen species, oxidation of lipids and lipoproteins leads to activation of phagocytic cells that set in motion a cascade of inflammatory events including the release of growth and chemotactic factors as well as the migration and proliferation of SMCs. Excess lipid uptake by macrophages promotes their differentiation into foam cells, which is the hallmark of fatty streaks observed in early atherosclerosis. We captured the signaling mechanisms responsible for these phenomena in the SMC Activation and Foam Cell Formation subnetworks, respectively. In humans, the final stage of atheroma formation is reached after years of continuous exposure to environmental noxious stimuli, such as cigarette smoke and a diet rich in saturated fats and poor in antioxidants[29, 32]. The natural disease progression results in plaque growth and positive vessel remodeling to maintain a functional lumen size. Although atherosclerotic plaques remain clinically silent for decades, they may evolve to become advanced lesions that are prone to calcification, cap thinning, hemorrhage and rupture. Platelets play a major role in the advanced stages of plaque development where neovascularization, thrombosis, plaque erosion and rupture constitute fatal complications. Interestingly, platelets also participate in early atherogenic events by promoting EC activation and forming microthrombi in fatty streaks. These platelet-related pathways have been captured in the Plaque Destabilization and Platelet Activation subnetworks. Thus, the pathobiology incorporated in the V-IPN represents a comprehensive implementation of our current knowledge of atherogenesis, which includes the primary cellular players, as well as the array of biological processes involved, ranging from EC dysfunction to the formation of fatty streaks, atheromas, and subsequent plaque instability.
Comparisons of RCR-based analyses of transcriptomics datasets from cells in culture and intact murine and human tissues rendered a series of powerful insights. The underlying molecular findings as well as the pathobiological relevance are described in detail below and summarized in Additional file12: Table S4.
The V-IPN captured predicted HYPs from primary HAECs and an immortalized HAEC line that may account for the divergent phenotypes exhibited by the two cell types
ECs uptake oxidized lipids and Ox-LDL through the scavenger receptor LOX-1. Oxidized lipids induce inflammatory responses mediated by NFkB, including upregulation of cytokines, chemokines and CAMs that result in substantial leukocyte recruitment[35, 36]. Predicted HYPs from HAECs overexpressing LOX-1 or GFP (Hs_EC_LOX1_oxLDL_vs_ct; Hs_EC_GFP_oxLDL_vs_ct) differed significantly from the HYP profile of HAECs stimulated by proinflammatory oxidized phospholipid, Ox-PAPC (Hs_EC_oxPAP_vs_ct). A distinction between the three datasets is indicated by the relatively low number of predicted HYPs observed in Ox-LDL-treated HAECs; 5 and 14 HYPs for GFP- and LOX-1-transfected cells, respectively, compared to 40 significant HYPs predicted from Hs_EC_oxPAP_vs_ct, a dataset generated from HAECs treated with Ox-PAPC (Additional file10: Table S2). The marked differences in the number of significant predicted HYPs from each dataset could reflect differing magnitudes of signaling events elicited upon stimulating ECs with atherogenic lipids. Indeed, Ox-PAPC, a purified component of Ox-LDL, may have wider and more potent effects on EC biology compared to Ox-LDL. The analysis presented herein could indicate that Ox-PAPC treatment is a more efficient process by which to induce the molecular features that most closely resemble chronic development of atherosclerotic plaques. Alternatively, the differences could be explained by the phenotypic differences between the cell types used and the number of human donors represented in each dataset. Cells from dataset Hs_EC_LOX1_oxLDL_vs_ct were obtained from an immortalized (SV40-induced) human aortic EC line, whereas primary HAEC cultures from 96 donors were used to generate Hs_EC_oxPAP_vs_ct expression data. Cellular transformation of immortalized cell lines is associated with phenotypic changes at multiple levels including gene expression, biochemical, metabolic and proliferative capacity[40, 41]. Therefore, transformed cell lines may have more limited abilities to respond to experimental induction by atherogenic lipids.
Mechanisms identified by the V-IPN discriminate between early and late vascular lesions in ApoE-/- mice and highlight the commonalities between advanced murine and human atherosclerotic lesions
We have conducted RCR analysis on gene expression datasets from a study utilizing the ApoE-/- mouse strain, a well-established model of atherosclerosis. The minimal overlap of common HYPs (6 total, Table 5) between aortas of young (Mm_Ao_16w_ApoE_CS_vs_sham) and old (Mm_Ao_78w_ApoE_vs_wt) adult ApoE-/- mice demonstrates a substantial divergence of atherogenic processes taking place in young ApoE-/- mice (16-week old) exposed to CS for 30 days compared to advanced atherosclerotic disease observed in older ApoE-/- mice (78 weeks). Strikingly, predicted HYPs from human (Hs_athCA_vs_ctIMA) and old murine (Mm_Ao_78w_ApoE_vs_wt) datasets generated from intact arteries harboring advanced atherosclerotic lesions shared a significantly larger set of causal mechanisms (32 total HYPs, Table 5) indicating that similar mechanisms underlie the development of advanced atherosclerotic lesions in both species. The pathobiological picture that emerged when overlaying predicted HYPs from both datasets onto the V-IPN is one where a complex series of proliferative, apoptotic and inflammatory events driven by intracellular transducers and transcription regulators are all taking place simultaneously. A functional clustering analysis of the common HYPs from both species using DAVID revealed functional categories consistent with the mechanisms described above (Additional file11: Table S3). The vast majority of this set of commonly mapped HYPs, which were predicted increased in the human (Hs_athCA_vs_ctIMA) and old murine (Mm_Ao_78w_ApoE_vs_wt) datasets, should serve as an initial reference for future simulations using murine and human vascular datasets. Our results highlight the power of the V-IPN to assess, at the molecular level, the degree of translatability from murine morphologic and tomographic data on plaque instability to the human clinical setting using transcriptomics data.
V-IPN evaluation reveals that distinct molecular pathways contribute to atherogenesis in a CS-exposure murine model of disease
Among the predicted HYPs from the mouse Mm_Ao_16w_ApoE_CS_vs_sham and Mm_Ao_78w_ApoE_vs_wt datasets, coverage analysis across the V-IPN revealed a remarkably low degree of overlapping mechanisms across the V-IPN subnetworks (0-3%, Table 5). This result suggests that distinct biological pathways contribute to early lesion development in ApoE-/- mice exposed to CS. A comparison between Mm_Ao_16w_ApoE_CS_vs_sham and Hs_athCA_vs_ctIMA also revealed a low extent of overlapping HYPs, ranging from 0-7% spanning all subnetworks (Table 5). Furthermore, broad coverage analysis of the predicted HYPs indicated that the Mm_Ao_16w_ApoE_CS_vs_sham dataset contained 52 HYPs found across the V-IPN, whereas the Hs_athCA_vs_ctIMA dataset contained 80 HYPs (Table 4). This result suggests that a more diverse array of biological mechanisms underlie advanced atherosclerotic lesion development in the human disease. Indeed, coverage analysis specifically within the Plaque Destabilization subnetwork indicated only a few unique mechanisms being activated in the Mm_Ao_16w_ApoE_CS_vs_sham samples (e.g. decreased PPARA and increased CD40LG) (Figure 5). A full characterization of discrete cellular events distinguishing acute atherogenesis in a murine model from advanced-stage murine and human disease further demonstrates the utility of the V-IPN in evaluating novel datasets to better understand comparability between species and/or disease model systems.
V-IPN-generated causal paths were consistent with advanced, unstable lesions, in coronary atherosclerotic arteries of CAD patients
Unlike all other datasets used for model construction and evaluation, Hs_athCA_vs_ctIMA represents paired gene expression profiles from 37 human atherosclerotic coronary arteries and control internal mammary arteries (IMA); each pair was obtained from the same subject (STAGE study). A principal component analysis (PCA) of the gene expression profiles identified two groups of control and diseased arteries (Additional file13: Figure S5). All pairs were clearly clustered, thus providing a sense of how distinct the gene profile of advanced coronary atherosclerosis was, compared to the reference IMA. Pathological evidence indicates that unstable plaques are complex lesions with common morphological features including a thin fibrous cap, a large lipid core, a network of vasa vasorum and focal inflammation. HYP coverage analysis of the human dataset (Hs_athCA_vs_ctIMA) was distinctly consistent with mechanisms driving the morphological features of unstable plaques. Remarkably, the Plaque Destabilization and Platelet Activation subnetworks exhibited the largest ORs for HYP overlap across all datasets examined and through the six V-IPN subnetworks (Figure 4). The ORs for the Hs_athCA_vs_ctIMA dataset were even higher than those obtained for Mm_Ao_78w_ApoE_vs_wt; in the Platelet Activation (2.9 vs. 1.1) and Plaque Destabilization subnetworks (4.1 vs. 3.4); Mm_Ao_78w_ApoE_vs_wt is a dataset from advanced murine aortic lesions used for network construction. This analysis highlights model-level coverage of additional mechanisms unique to the human condition and validates the strength of the V-IPN to differentiate expression data derived from advanced human and murine lesions. Consistent with a decreased activity of SMCs leading to fibrous cap thinning in lesions that are prone to rupture, the SMC Activation network exhibited lower coverage and ORs compared to both murine datasets, Mm_Ao_16w_ApoE_CS_vs_sham and Mm_Ao_78w_ApoE_vs_wt (2.1 vs. 2.7 and 2.8) (Figure 4). Significant HYPs categorized within pathobiological functions linked to plaque destabilization are described below.
Lipid metabolism-related HYPs
Lipid metabolism-related predicted HYPs included ABCG1, an ATP-binding cassette transporter that regulates macrophage cholesterol efflux and phospholipid transport to lipoprotein acceptors. Intact endothelium from ABCG1-deficient mice has been shown to exhibit 4-fold increases in monocyte adhesion. Other lipid metabolism HYPs included the low density lipoprotein receptor (LDLR), a protein expressed by human macrophages and known to be recycled between the plasma membrane and lysosomes upon binding of LDL. Ox-LDL and Ox-HDL were both highlighted as predicted HYPs. Relaxing the concordance and richness p values from 0.05 to 0.1 resulted in a few additional predicted HYPs relevant to atherogenesis including SP1, CD36 and NOS3. S1P receptor is expressed by ECs and it binds its ligand S1P, a bioactive lipid with numerous functions in the immune and cardiovascular systems. Using a lipidomics approach, we have previously shown that CS exposure increases, whereas cessation decreases, the levels of S1P in plasma of ApoE-/- mice[5, 22]. Furthermore, CD36 is a glycoprotein expressed in various vascular and circulatory cell types including monocytes, macrophages, platelets, ECs and adipocytes. It binds collagen, thrombospondin, phospholipids, and Ox-LDL. In macrophages of human atherosclerotic lesions, CD36 acts as a receptor for Ox-LDL and a transporter of long-chain fatty acids. CD36-deficient patients were shown to have hypertriglyceridemia, whereas patients with acute coronary syndrome exhibited 6-fold higher levels of CD36 in circulating monocytes compared to healthy controls. Taken together, these data demonstrate that the scavenger receptor CD36 is involved not only in pro-atherogenic mechanisms but also in the development of acute coronary syndrome symptoms, which are primarily caused by plaque rupture at sites of thrombus formation.
HYPs related to platelet activation and clot formation
The human advanced coronary lesion dataset (Hs_athCA_vs_ctIMA) displayed a high degree of coverage and enrichment within the Platelet Activation subnetwork, highlighting the involvement of platelets in thrombus formation and plaque instability. CD36 expression on the surface of platelets serves as an adhesion molecule and a receptor for thrombospondin and Ox-LDL, which was also a predicted HYP mapped to the V-IPN. CD36 has been demonstrated to play a role in platelet activation and thrombus formation in experiments where immobilized thrombospondin and Ox-LDL activate platelets via CD36 through a Syk kinase-dependent signaling mechanism. In agreement with the multiple functions of CD36, the CD36 node is present in three V-IPN subnetworks, Platelet Activation, EC-Monocyte Interaction and Plaque Destabilization; CD36 was a predicted HYP in both the murine and human dataset. A predicted HYP unique to the human dataset was the thrombin receptor. The first step of the coagulation cascade involves cleavage of coagulation factor II to form thrombin. Protease-activated receptors (PARs [FR2]) are activated in response to TF-VIIa-Xa, a ternary complex that is also linked to inflammation within plaques prone to rupture. Plaque vulnerability has been shown to be correlated with PAR1 expression in ApoE-/- mice.
Angiogenesis-related causal paths were captured by the V-IPN
Other significant HYPs predicted by the datasets included NOS3 and ANGPT1, both of which regulate vascular tone and permeability, as well as blood vessel maturation and stability. Intra-plaque neovascularization may lead to hemorrhage, fissure development and plaque rupture. The neovascularization process that occurs in advanced lesions indicates that various pro-angiogenic factors are being secreted within the plaque and that signaling pathways and associated molecules are all operating in advanced atherosclerotic lesions. In addition to ANGPT1, vascular endothelial growth factor A (VEGFA), HIF1A, a master regulator of a cellular homeostatic response to hypoxia that activates transcription of genes involved in angiogenesis, PTAFR and STAB1, were also identified as significant HYPs and potentially modulated mechanisms. STAB1, also known as CLEVER-1, is a glycoprotein involved in scavenging, angiogenesis and cell adhesion, and has also been shown to mediate transmigration of leukocytes.
RCR-based models do not operate with integrated feedbacks and non-linear elements that contribute to regulating a dynamic output. In order to make accurate predictions, mathematical models incorporate feedback elements tuned to match phenotypic constrains (e.g., blood pressure values). In RCR, all biological feedbacks are implicitly integrated in the datasets. RCR-based models do not dynamically model the regulatory processes controlling biological pathways. RCR-based models are tools to extract biological processes embedded in large sets of molecular data driven by specific experimental perturbations, and to contextualize those findings within a body of knowledge.
In summary, we have demonstrated that RCR analysis of large gene expression datasets coupled with HYP mapping to the V-IPN was able to discern the mechanistic variability underpinning the development of atherosclerotic lesions in a variety of experimental and species contexts. The mapping of predicted HYPs to the V-IPN was able to successfully distinguish between early and advanced murine lesions, as well as advanced murine and human atherosclerosis, thus pointing to a distinct subset of mechanisms that are translatable to the human condition. Importantly, our computational model proved to be a powerful tool to further our pathophysiological understanding of vascular inflammation, atherogenesis and plaque destabilization. The dynamic nature of the model’s structure allows for further refinement as additional datasets become available and represents a useful tool for the interrogation of cross-species translatability in the context of cardiovascular disease.
Reverse Causal Reasoning (RCR): A computational methodology for identifying potential upstream controllers leading to differential molecular profiles.
Selventa Knowledgebase (SK): A network representing a working set of knowledge fit for a specified use. The SK is used as a substrate for RCR. It encodes prior scientific knowledge as a network of nodes that are connected by edges.
Biological Expression Language (BEL): The knowledge representation language used to build the SK.
Node: A biological entity or process in the SK.
Edge A causal relationship (e.g., increase, decrease, subset) connecting two nodes in the knowledgebase.
State Change (SC): A differential measurement across a sample group (e.g., treated and control) that is converted to a discrete value of increase, decrease, or no change, based on two statistical metrics: richness and concordance.
Hypothesis (HYP): A small, directed causal network containing an upstream node representing a biological entity or process connected by a causal increase, decrease or ambiguous edges to downstream nodes representing measured entities.
HYP upstream node: A controller of downstream nodes in a HYP and a potential explanation for state changes (SC) mapped to the downstream nodes.
HYP downstream nodes: Nodes in a HYP mapped to quantities measured in the dataset.
HYP causal edges: The causal relationships (i.e., increases or decreases) connecting the HYP upstream node to each downstream node.
Richness: A measure of the relevance of a HYP to the changes observed in an experimental dataset.
Concordance: A measure of the consistency of the direction of the changes observed in an experimental dataset.
Coverage (sensitivity): An estimate of the fraction of possible HYPs in a subnetwork that are significant in a dataset. Coverage is a measure of HYP enrichment.
Odds ratio (OR): The probability of having significant dataset HYPs within a network. The higher the OR, the better the network encompasses the biology embedded in a given dataset.
X: Protein abundance of X
taof(X): Transcriptional activity of X
exp(X): RNA expression of X
gtpof(X): GTP-bound activity of X
kaof(X): Kinase activity of X
paof(X): Phosphatase activity of X
catof(X): Catalytic activity of X
X P@Y: Abundance of X phosphorylated at Y
Biological expression language
Reverse causal reasoning
The authors wish to acknowledge Anita Iskandar, Michael Maria, and Natalie Catlett for their critical review of the manuscript. This work was supported financially by Philip Morris International under a joint research collaboration between Selventa and Philip Morris International.
- Ross R: Atherosclerosis-an inflammatory disease. N Engl J Med. 1999, 340: 115-126.View ArticlePubMedGoogle Scholar
- Libby P, Ridker PM, Maseri A: Inflammation in atherosclerosis. Arterioscler Thromb Vasc Biol. 2012, 32: 2045-2051.PubMed CentralView ArticlePubMedGoogle Scholar
- Howard G, Wagenknecht LE, Burke GL, Diez-Roux A, Evans GW, McGovern P, Nieto FJ, Tell GS: Cigarette smoking and progression of atherosclerosis: The Atherosclerosis Risk in Communities (ARIC) Study. JAMA. 1998, 279: 119-124.View ArticlePubMedGoogle Scholar
- von Holt K, Lebrun S, Stinn W, Conroy L, Wallerath T, Schleef R: Progression of atherosclerosis in the Apo E-/- model: 12-month exposure to cigarette mainstream smoke combined with high-cholesterol/fat diet. Atherosclerosis. 2009, 205: 135-143.View ArticlePubMedGoogle Scholar
- Boue S, Tarasov K, Janis M, Lebrun S, Hurme R, Schlage W, Lietz M, Vuillaume G, Ekroos K, Steffen Y, Peitsch MC, Laaksonen R, Hoeng J: Modulation of atherogenic lipidome by cigarette smoke in apolipoprotein E-deficient mice. Atherosclerosis. 2012, 225: 328-334.View ArticlePubMedGoogle Scholar
- Celermajer DS, Sorensen KE, Georgakopoulos D, Bull C, Thomas O, Robinson J, Deanfield JE: Cigarette smoking is associated with dose-related and potentially reversible impairment of endothelium-dependent dilation in healthy young adults. Circulation. 1993, 88: 2149-2155.View ArticlePubMedGoogle Scholar
- Barua RS, Ambrose JA, Saha DC, Eales-Reynolds LJ, DeVoe MC, Zervas JG: Smoking is associated with altered endothelial-derived fibrinolytic and antithrombotic factors: an in vitro demonstration. Circulation. 2002, 106: 905-908.View ArticlePubMedGoogle Scholar
- Chen XP, Xun KL, Wu Q, Zhang TT, Shi JS, Du GH: Oxidized low density lipoprotein receptor-1 mediates oxidized low density lipoprotein-induced apoptosis in human umbilical vein endothelial cells: role of reactive oxygen species. Vascul Pharmacol. 2007, 47: 1-9.View ArticlePubMedGoogle Scholar
- Barrett T, Edgar R: Gene expression omnibus: microarray data storage, submission, retrieval, and analysis. Methods Enzymol. 2006, 411: 352-369.PubMed CentralView ArticlePubMedGoogle Scholar
- Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, Davis AP, Dolinski K, Dwight SS, Eppig JT, Harris MA, Hill DP, Issel-Tarver L, Kasarskis A, Lewis S, Matese JC, Richardson JE, Ringwald M, Rubin GM, Sherlock G: Gene ontology: tool for the unification of biology: The Gene Ontology Consortium. Nat Genet. 2000, 25: 25-29.PubMed CentralView ArticlePubMedGoogle Scholar
- Kanehisa M, Goto S, Furumichi M, Tanabe M, Hirakawa M: KEGG for representation and analysis of molecular networks involving diseases and drugs. Nucleic Acids Res. 2010, 38: D355-360.PubMed CentralView ArticlePubMedGoogle Scholar
- Chen G, Gharib TG, Huang CC, Taylor JM, Misek DE, Kardia SL, Giordano TJ, Iannettoni MD, Orringer MB, Hanash SM, Beer DG: Discordant protein and mRNA expression in lung adenocarcinomas. Mol Cell Proteomics. 2002, 1: 304-313.View ArticlePubMedGoogle Scholar
- Guo Y, Xiao P, Lei S, Deng F, Xiao GG, Liu Y, Chen X, Li L, Wu S, Chen Y, Jiang H, Tan L, Xie J, Zhu X, Liang S, Deng H: How is mRNA expression predictive for protein expression? A correlation study on human circulating monocytes. Acta Biochim Biophys Sin (Shanghai). 2008, 40: 426-436.View ArticleGoogle Scholar
- Vogel C, Marcotte EM: Insights into the regulation of protein abundance from proteomic and transcriptomic analyses. Nat Rev Genet. 2012, 13: 227-232.PubMed CentralPubMedGoogle Scholar
- Catlett NL, Bargnesi AJ, Ungerer S, Seagaran T, Ladd W, Elliston KO, Pratt D: Reverse causal reasoning: applying qualitative causal knowledge to the interpretation of high-throughput data. BMC Bioinformatics. 2013, 14: 340-PubMed CentralView ArticlePubMedGoogle Scholar
- Schlage WK, Westra JW, Gebel S, Catlett NL, Mathis C, Frushour BP, Hengstermann A, Van Hooser A, Poussin C, Wong B, Lietz M, Park J, Drubin D, Veljkovic E, Peitsch MC, Hoeng J, Deehan R: A computable cellular stress network model for non-diseased pulmonary and cardiovascular tissue. BMC Syst Biol. 2011, 5: 168-PubMed CentralView ArticlePubMedGoogle Scholar
- Gebel S, Lichtner RB, Frushour B, Schlage WK, Hoang V, Talikka M, Hengstermann A, Mathis C, Veljkovic E, Peck M, Peitsch MC, Deehan R, Hoeng J, Westra JW: Construction of a computable network model for DNA damage, autophagy, cell death, and senescence. Bioinform Biol Insights. 2013, 7: 97-117.PubMed CentralPubMedGoogle Scholar
- Westra JW, Schlage WK, Hengstermann A, Gebel S, Mathis C, Thomson T, Wong B, Hoang V, Veljkovic E, Peck M, Lichtner RB, Weisensee D, Talikka M, Deehan R, Hoeng J, Peitsch MC: A modular cell-type focused inflammatory process network model for non-diseased pulmonary tissue. Bioinform Biol Insights. 2013, 7: 167-192.PubMed CentralPubMedGoogle Scholar
- Westra JW, Schlage WK, Frushour BP, Gebel S, Catlett NL, Han W, Eddy SF, Hengstermann A, Matthews AL, Mathis C, Lichtner RB, Poussin C, Talikka M, Veljkovic E, Van Hooser AA, Wong B, Maria MJ, Peitsch MC, Deehan R, Hoeng J: Construction of a computable cell proliferation network focused on non-diseased lung cells. BMC Syst Biol. 2011, 5: 105-PubMed CentralView ArticlePubMedGoogle Scholar
- Park JS, Schlage WK, Frushour BP, Talikka M, Toedter G, Gebel S, Deehan R, Veljkovic E, Westra JW MJP, Boue S, Kogel U, Gonzalez-Suarez I, Hengstermann A, Peitsch MC, Hoeng J: Construction of a computable network model of tissue repair and angiogenesis in the lung. J Clinic Toxicol. 2013, S12: 002-View ArticleGoogle Scholar
- Nakashima Y, Plump AS, Raines EW, Breslow JL, Ross R: ApoE-deficient mice develop lesions of all phases of atherosclerosis throughout the arterial tree. Arterioscler Thromb. 1994, 14: 133-140.View ArticlePubMedGoogle Scholar
- Lietz M, Berges A, Lebrun S, Meurrens K, Steffen Y, Stolle K, Schueller J, Boue S, Vuillaume G, Vanscheeuwijck P, Moehring M, Schlage W, De Leon H, Hoeng J, Peitsch M: Cigarette-smoke-induced atherogenic lipid profiles in plasma and vascular tissue of apolipoprotein E-deficient mice are attenuated by smoking cessation. Atherosclerosis. 2013, 229: 86-93.View ArticlePubMedGoogle Scholar
- O'Donnell SM, Holm GH, Pierce JM, Tian B, Watson MJ, Chari RS, Ballard DW, Brasier AR, Dermody TS: Identification of an NF-kappaB-dependent gene network in cells infected by mammalian reovirus. J Virol. 2006, 80: 1077-1086.PubMed CentralView ArticlePubMedGoogle Scholar
- Mattaliano MD, Huard C, Cao W, Hill AA, Zhong W, Martinez RV, Harnish DC, Paulsen JE, Shih HH: LOX-1-dependent transcriptional regulation in response to oxidized LDL treatment of human aortic endothelial cells. Am J Physiol Cell Physiol. 2009, 296: C1329-1337.View ArticlePubMedGoogle Scholar
- Hagg S, Skogsberg J, Lundstrom J, Noori P, Nilsson R, Zhong H, Maleki S, Shang MM, Brinne B, Bradshaw M, Bajic VB, Samnegard A, Silveira A, Kaplan LM, Gigante B, Leander K, de Faire U, Rosfors S, Lockowandt U, Liska J, Konrad P, Takolander R, Franco-Cereceda A, Schadt EE, Ivert T, Hamsten A, Tegner J, Bjorkegren J: Multi-organ expression profiling uncovers a gene module in coronary artery disease involving transendothelial migration of leukocytes and LIM domain binding 2: the Stockholm Atherosclerosis Gene Expression (STAGE) study. PLoS Genet. 2009, 5: e1000754-PubMed CentralView ArticlePubMedGoogle Scholar
- Costello CM, Howell K, Cahill E, McBryan J, Konigshoff M, Eickelberg O, Gaine S, Martin F, McLoughlin P: Lung-selective gene responses to alveolar hypoxia: potential role for the bone morphogenetic antagonist gremlin in pulmonary hypertension. Am J Physiol Lung Cell Mol Physiol. 2008, 295: L272-284.View ArticlePubMedGoogle Scholar
- Shao J, Katika MR, Schmeits PC, Hendriksen PJ, van Loveren H, Peijnenburg AA, Volger OL: Toxicogenomics-based identification of mechanisms for direct immunotoxicity. Toxicol Sci. 2013, 135: 328-346.View ArticlePubMedGoogle Scholar
- Li JJ, Huang YQ, Basch R, Karpatkin S: Thrombin induces the release of angiopoietin-1 from platelets. Thromb Haemost. 2001, 85: 204-206.PubMedGoogle Scholar
- Pratico D: Antioxidants and endothelium protection. Atherosclerosis. 2005, 181: 215-224.View ArticlePubMedGoogle Scholar
- Amberger A, Maczek C, Jurgens G, Michaelis D, Schett G, Trieb K, Eberl T, Jindal S, Xu Q, Wick G: Co-expression of ICAM-1, VCAM-1, ELAM-1 and Hsp60 in human arterial and venous endothelial cells in response to cytokines and oxidized low-density lipoproteins. Cell Stress Chaperones. 1997, 2: 94-103.PubMed CentralView ArticlePubMedGoogle Scholar
- Chai YC, Binion DG, Macklis R, Chisolm GM: Smooth muscle cell proliferation induced by oxidized LDL-borne lysophosphatidylcholine. Evidence for FGF-2 release from cells not extracellular matrix. Vascul Pharmacol. 2002, 38: 229-237.View ArticlePubMedGoogle Scholar
- Chelland Campbell S, Moffatt RJ, Stamford BA: Smoking and smoking cessation – the relationship between cardiovascular disease and lipoprotein metabolism: a review. Atherosclerosis. 2008, 201: 225-235.View ArticlePubMedGoogle Scholar
- Mayranpaa MI, Heikkila HM, Lindstedt KA, Walls AF, Kovanen PT: Desquamation of human coronary artery endothelium by human mast cell proteases: implications for plaque erosion. Coron Artery Dis. 2006, 17: 611-621.View ArticlePubMedGoogle Scholar
- Masuda J, Ross R: Atherogenesis during low level hypercholesterolemia in the nonhuman primate: I Fatty streak formation. Arteriosclerosis. 1990, 10: 164-177.View ArticlePubMedGoogle Scholar
- Mehta JL, Chen J, Hermonat PL, Romeo F, Novelli G: Lectin-like, oxidized low-density lipoprotein receptor-1 (LOX-1): a critical player in the development of atherosclerosis and related disorders. Cardiovasc Res. 2006, 69: 36-45.View ArticlePubMedGoogle Scholar
- Berliner JA, Watson AD: A role for oxidized phospholipids in atherosclerosis. N Engl J Med. 2005, 353: 9-11.View ArticlePubMedGoogle Scholar
- Itabe H, Obama T, Kato R: The Dynamics of Oxidized LDL during Atherogenesis. J Lipids. 2011, 2011: 418313-PubMed CentralView ArticlePubMedGoogle Scholar
- Chadwick CC, Shaw LJ, Winneker RC: TNF-alpha and 9-cis-retinoic acid synergistically induce ICAM-1 expression: evidence for interaction of retinoid receptors with NF-kappa B. Exp Cell Res. 1998, 239: 423-429.View ArticlePubMedGoogle Scholar
- Romanoski CE, Lee S, Kim MJ, Ingram-Drake L, Plaisier CL, Yordanova R, Tilford C, Guan B, He A, Gargalovic PS, Kirchgessner TG, Berliner JA, Lusis AJ: Systems genetics analysis of gene-by-environment interactions in human cells. Am J Hum Genet. 2010, 86: 399-410.PubMed CentralView ArticlePubMedGoogle Scholar
- Mignotte B, Larcher JC, Zheng DQ, Esnault C, Coulaud D, Feunteun J: SV40 induced cellular immortalization: phenotypic changes associated with the loss of proliferative capacity in a conditionally immortalized cell line. Oncogene. 1990, 5: 1529-1533.PubMedGoogle Scholar
- RayChaudhury A, Frazier WA, D'Amore PA: Comparison of normal and tumorigenic endothelial cells: differences in thrombospondin production and responses to transforming growth factor-beta. J Cell Sci. 1994, 107 (Pt 1): 39-46.PubMedGoogle Scholar
- Coleman R, Hayek T, Keidar S, Aviram M: A mouse model for human atherosclerosis: long-term histopathological study of lesion development in the aortic arch of apolipoprotein E-deficient (E0) mice. Acta Histochem. 2006, 108: 415-424.View ArticlePubMedGoogle Scholar
- Rosenfeld ME, Polinsky P, Virmani R, Kauser K, Rubanyi G, Schwartz SM: Advanced atherosclerotic lesions in the innominate artery of the ApoE knockout mouse. Arterioscler Thromb Vasc Biol. 2000, 20: 2587-2592.View ArticlePubMedGoogle Scholar
- Cilingiroglu M, Oh JH, Sugunan B, Kemp NJ, Kim J, Lee S, Zaatari HN, Escobedo D, Thomsen S, Milner TE, Feldman MD: Detection of vulnerable plaque in a murine model of atherosclerosis with optical coherence tomography. Catheter Cardiovasc Interv. 2006, 67: 915-923.View ArticlePubMedGoogle Scholar
- Bond AR, Jackson CL: The fat-fed apolipoprotein E knockout mouse brachiocephalic artery in the study of atherosclerotic plaque rupture. J Biomed Biotechnol. 2011, 2011: 379069-PubMed CentralView ArticlePubMedGoogle Scholar
- Yla-Herttuala S, Bentzon JF, Daemen M, Falk E, Garcia-Garcia HM, Herrmann J, Hoefer I, Jukema JW, Krams R, Kwak BR, Marx N, Naruszewicz M, Newby A, Pasterkamp G, Serruys PW, Waltenberger J, Weber C, Tokgozoglu L: Stabilisation of atherosclerotic plaques. Position paper of the European Society of Cardiology (ESC) Working Group on atherosclerosis and vascular biology. Thromb Haemost. 2011, 106: 1-19.View ArticlePubMedGoogle Scholar
- Whetzel AM, Sturek JM, Nagelin MH, Bolick DT, Gebre AK, Parks JS, Bruce AC, Skaflen MD, Hedrick CC: ABCG1 deficiency in mice promotes endothelial activation and monocyte-endothelial interactions. Arterioscler Thromb Vasc Biol. 2010, 30: 809-817.PubMed CentralView ArticlePubMedGoogle Scholar
- Yamashita S, Hirano K, Kuwasako T, Janabi M, Toyama Y, Ishigami M, Sakai N: Physiological and pathological roles of a multi-ligand receptor CD36 in atherogenesis; insights from CD36-deficient patients. Mol Cell Biochem. 2007, 299: 19-22.View ArticlePubMedGoogle Scholar
- Piechota M, Banaszewska A, Dudziak J, Slomczynski M, Plewa R: Highly upregulated expression of CD36 and MSR1 in circulating monocytes of patients with acute coronary syndromes. Protein J. 2012, 31: 511-518.PubMed CentralView ArticlePubMedGoogle Scholar
- Nergiz-Unal R, Lamers MM, Van Kruchten R, Luiken JJ, Cosemans JM, Glatz JF, Kuijpers MJ, Heemskerk JW: Signaling role of CD36 in platelet activation and thrombus formation on immobilized thrombospondin or oxidized low-density lipoprotein. J Thromb Haemost. 2011, 9: 1835-1846.View ArticlePubMedGoogle Scholar
- Riewald M, Ruf W: Orchestration of coagulation protease signaling by tissue factor. Trends Cardiovasc Med. 2002, 12: 149-154.View ArticlePubMedGoogle Scholar
- Yang L, Zhou X, Guo R, Shi Y, Liang X, Heng X: Role of Kruppel-like factor 2 and protease-activated receptor-1 in vulnerable plaques of ApoE(-/-) mice and intervention with statin. Can J Cardiol. 2013, 29: 997-1005.View ArticlePubMedGoogle Scholar
- Cheng C, Chrifi I, Pasterkamp G, Duckers HJ: Biological mechanisms of microvessel formation in advanced atherosclerosis: the big five. Trends Cardiovasc Med. 2013, 23: 153-164.View ArticlePubMedGoogle Scholar
- Salmi M, Koskinen K, Henttinen T, Elima K, Jalkanen S: CLEVER-1 mediates lymphocyte transmigration through vascular and lymphatic endothelium. Blood. 2004, 104: 3849-3857.View ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.