Identification of key differential genes in intimal hyperplasia induced by left carotid artery ligation

Introduction

A growing body of evidence has shown that atherosclerosis is a chronic inflammatory process that narrows and hardens the arteries due to an excessive build-up of plaque in the tunica intima of the arterial wall (Daugherty, Tabas & Rader, 2015). Although percutaneous transluminal coronary angioplasty and drug-eluting stents are widely used in clinical treatment for coronary artery disease, they both inevitably trigger a high rate of in-stent restenosis (>10%) (Giacoppo et al., 2015). The major causes of in-stent restenosis are intimal hyperplasia and vascular remodeling (Liang et al., 2019). However, the pathogenesis of intimal hyperplasia is complex and not completed understood. Previous studies have suggested that angiogenesis, intimal formation, vascular remodeling, platelet aggregation, endothelial activation, inflammation, vascular smooth muscle cell (VSMC) proliferation, migration, and phenotypic transformation may be closely related to intimal hyperplasia (Chan et al., 2008; Esper et al., 2008; Mount et al., 2008). Therefore, elucidating the gene changes and mechanisms involved into intimal hyperplasia is vital to understand the process of neointimal hyperplasia and for developing new therapeutic strategies in the prevention of vascular restenosis.

Gene chip microarray technology is a recent development commonly used to detect changes in gene expression with high specificity and sensitivity. For example, Woods et al. (2002) identified that the tyrosine kinase receptor, EphB2, as a differentially expressed gene (DEGs), had an anti-proliferative effect in vascular smooth muscle cells (VSMCs) in response to continuous intravenous heparin administration in the rabbit model of arterial injury. Moreover, Hokamura et al. (2010) analyzed changes in the gene expression profiles of injured blood vessels of mice using DNA microarray assays. They demonstrated that bacteremia induced by P. gingivalis led to intimal hyperplasia associated with over-expressions of S100A9 and Smemb (Hokamura et al., 2010). Similarly, Stone et al. (2001) reported that the early and sustained under-expression of some proteasome genes may alter cell cycle control and matrix protein signaling, contributing to the unregulated proliferation of smooth muscle cells and the extracellular matrix in anastomotic intimal hyperplasia after prosthetic arterial grafting.

Herein, we downloaded and analyzed the original microarray dataset of intimal hyperplasia induced by the ligation of the left carotid artery from the Gene Expression Omnibus (GEO) (i.e. GSE56143) to obtain potential DEGs related to intimal hyperplasia. In addition, their functions and pathways were analyzed using the Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses, respectively. The major key genes related to intimal hyperplasia were identified and the transcriptional regulation based on DEGs transcription factors (TFs) was predicted using TRRUST2.0 software. Lastly, we established a mouse model of vascular intimal hyperplasia by ligating the left carotid artery and verified the changes of the key genes at the mRNA and protein levels.

Methods

Results

Identification and GO enrichment analysis for DEGs

We used the GEO2R tool to analyze the DEGs in order to obtain the changes of the vascular transcriptome in the intimal hyperplasia model induced by carotid artery ligation. As shown in Fig. 1A and Table S2, a total of 373 DEGs were identified in the LCA/RCA group, including 199 upregulated genes and 174 downregulated genes.

Next, to understand the biological function of DEGs, the enrichment results were obtained through GO using DAVID version 6.8. The DEGs mainly enriched several biological processes, including immune system process, DNA methylation on cytosine, cell adhesion, DNA replication-dependent nucleosome assembly, positive regulation of gene expression, epigenetics, protein heterotetramerization, positive regulation of angiogenesis, nucleosome assembly, chromatin silencing at ribosome DNA (rDNA), and inflammatory response (Fig. 1B and Table S3). DEGs were mainly enriched in the extracellular exosome, nucleosome, extracellular matrix, membrane, nuclear chromosome, extracellular space, proteinaceous extracellular matrix, cell surface, extracellular region, and focal adhesion for cellular components (Fig. 1C and Table S4). Molecular function analysis showed that DEGs were mainly enriched in protein binding, histone binding, nucleosomal DNA binding, protein kinase binding, laminin binding, protein heterodimerization activity, actin binding, enzyme binding, carbohydrate binding, and haptoglobin binding (Fig. 1D and Table S5).

KEGG pathway enrichment analysis for DEGs

KEGG pathway enrichment analysis was performed using DAVID (version 6.8). We observed that the DEGs were significantly enriched in systemic lupus erythematosus, alcoholism, staphylococcus aureus infection, tuberculosis, phagosome, malaria, cell adhesion molecules (CAMs), asthma, intestinal immune network for IgA production, and regulation of actin cytoskeleton (Fig. 2 and Table S6).

PPI networks and prediction of TFs for DEGs in mouse

We used STRING to construct an initial PPI network from 373 DEGs to study the interaction of these DEGs in the LCA/RCA gene set (Fig. S1). The initial PPI network was then imported into Cytoscape to construct a sub-network (Fig. 3A). A total of 10 candidate genes were identified as key genes and their edge degrees were more than 48 in the PPI analysis. According to the edge degree rank, the 10 key genes were Ptprc (degree = 73), Fn1 (degree = 69), Tyrobp (degree = 59), Emr1 (degree = 58), Itgb2 (degree = 57), Itgax (degree = 54), CD44 (degree = 54), Ctss (degree = 49), Ly86 (degree = 48) and Aif1 (degree = 48). TRRUST2.0 was used to analyze the promoter binding motifs and to identify the transcription factors associated with each gene. The results showed that Nfkb1, Sp1, and Trp53 transcription factor families played an important role in the regulation of the DEGs expression (Fig. 3B).

Key gene validation

To further verify the results of the above bioinformatics analysis, we performed the carotid artery ligation in mice to establish a model of intimal hyperplasia. There was a certain degree of intimal formation in the carotid artery 7 days after ligation, as characterized by a slight increase in the thickness of the intima and the ratio of I/M compared with the RCA group (Figs. 4A–4D). In addition, a small number of macrophages infiltrated in the carotid arteries 7 days after injury by immunohistochemistry (Fig. S2). RTqPCR analysis showed that the mRNA levels of Ptprc, Fn1, Tyrobp, Emr1, Itgb2, Itgax, CD44, Ctss, Ly86, and Aif1 in the intimal hyperplasia group were significantly increased (P < 0.05) compared to those in the RCA group. The results were consistent with the above-mentioned bioinformatic analysis (Figs. 4E–4N). In addition, the DEGs were further verified by immunohistochemistry (Fig. 5) and western blot analysis (Fig. 6), which revealed that the expression levels of these DEGs were upregulated in LCA mice compared with RCA mice 7 days after injury.

Discussion

In this study, the global expression of DEGs was evaluated in intimal hyperplasia by analyzing the microarray dataset of RCA and LCA group in mice from the GEO database. A total of 373 DEGs were identified, including 199 upregulated genes and 174 downregulated genes. Furthermore, GO enrichment analysis and KEGG revealed the diversity of functions and pathways of DGEs. PPI network analysis showed that 10 DEGs with high node connectivity were selected as key genes. Three major transcription factors (i.e. Nfkb1, Sp1, Trp53) were predicted to potentially regulate the expression of the 373 DEGs. Finally, the expression changes of the key DEGs were verified at the mRNA and protein levels. Our bioinformatic analysis showed that the 10 key DEGs were closely related to immune inflammation and cell adhesion, despite not knowing the types of cells these DEGs originated from. Our previous studies demonstrated that that vascular smooth cells and infiltrated inflammatory cells play an essential role in the development of neointima formation after vascular injury (Du et al., 2013; Yao et al., 2018). Therefore, future studies will focus on determining the effects of key DEGs on certain types of cells and the development of vascular remodeling.

Accumulating evidence has shown that within hours to days following endothelial injury, inflammatory cells begin to infiltrate the injured area and secrete cytokines and growth factors. This has the potential to induce the migration and proliferation of VSMCs (Hansson, 2005). These cells secrete extracellular matrix components, including elastin and collagen, to form intimal hyperplasia and restenosis. Our KEGG pathway enrichment analysis indicated that cell adhesion molecules (CAMs) were enriched. Changes in the actin cytoskeleton are known as a prerequisite for vascular contractility and remodeling (Zhou, Gensch & Liao, 2011) and the regulation of the actin cytoskeleton is essential to maintaining vascular permeability, endothelial cell junction stability (Radeva & Waschke, 2018) and leukocyte migration (Marelli-Berg & Jangani, 2018). KEGG pathway analysis showed that the regulation of the actin cytoskeleton was significantly enriched.

However, after using PPI network analysis, only 10 up-regulated DEGs including Ptprc, Fn1, Tyrobp, Emr1, Itgb2, Itgax, CD44, Ctss, Ly86, and Aif1 were selected as key DEGs. RTqPCR, immunohistochemistry, and western blotting were used to detect the expression of these key DEGs at the mRNA and protein levels. We found that the mRNA and protein levels of key DEGs were also up-regulated in the carotid arteries of mice ligated for 7 days, which was consistent with the results of bioinformatics analysis.

Among the key DEGs, Ptprc (protein tyrosine phosphatase receptor type C), also known as CD45, can encode some members from the protein tyrosine phosphatase (PTP) family and plays a critical role in the process of cell growth, differentiation, mitosis, and carcinogenic transformation (Rheinlander, Schraven & Bommhardt, 2018). Previous studies have confirmed the role of CD45 on T cell receptor (TCR) and B cell receptor (BCR) signaling. The effects of CD45 in mast cells, macrophages, DCs and leucocyte adhesion and migration have been explored. The Janus kinases (JAKs) were identified as potential CD45 substrates (Saunders & Johnson, 2010). Fibronectin 1 (Fn1) encodes fibronectin as a glycoprotein on the cell surface and in the extracellular matrix, which is involved in cell adhesion and migration processes, including host defense, blood coagulation, metastasis and wound healing (Aota, Nomizu & Yamada, 1994; Aziz-Seible & Casey, 2011; Brentnall et al., 2014; Singh & Schwarzbauer, 2012; Veevers-Lowe et al., 2011). Fn1 is essential for smooth muscle cell phenotype regulation, proliferation, and cell adhesion (Tamura et al., 2000), and was implicated in vascular remodeling in radiation-induced brain injury (Andrews et al., 2017). Immune signaling adaptor TYROBP, also known as DAP12, encodes a transmembrane signaling polypeptide. This adaptor, initially characterized in NK cells, is associated with multiple cell-surface activating receptors expressed in both lymphoid and myeloid lineages and has multiple biological functions (Tomasello & Vivier, 2005). A previous study revealed that DAP12 signaling augmented the response to microbial products and amplified inflammation and thus contributed to mortality in sepsis (Turnbull et al., 2005). Consistently, a recent study of vascular transcriptomics demonstrated that the trend of TYROBP mRNA was increased in high fat/cholesterol (HFC) diet-fed Tibetan minipig atherosclerosis models (Pan et al., 2020).

Emr1, known as Adgre1, encodes the F4/80 antigen, which is restricted to leukocytes and plays an important role in the immune response by moderating the cell adhesion and inflammatory response. A previous study showed a trend of up-regulation for Emr1 in carotid artery balloon injury, consistent with our findings (Zhang et al., 2014). Additionally, the cytoplasmic region of Emr1 is required for regulating the number of ERMES foci, which regulates the number of foci of the endoplasmic reticulum-mitochondria encounter structure complex (Rasul et al., 2021). The integrin subunit beta2 (Itgb2), also known as CD18, is responsible for encoding an integrin beta chain and participating in cell adhesion and cell-surface mediated signaling. As a risk factor, Itgb2 had been shown to accelerate the development of myocardial infarction, atherothrombotic cerebral infarction (Lehmkuhl et al., 1996; May et al., 2002; Yamaguchi et al., 2006), and diabetic nephropathy (Geng et al., 2019) through the cell adhesion molecule pathway. Integrin subunit alpha X (Itgax), also known as CD11c, is a typical marker on the membrane of dendritic cells (DCs), which mediates cell-cell interaction during inflammatory responses. There is growing evidence that CD11c+DC may be involved in angiogenesis. Wu et al. (2009) showed that knocking out ITGAX in mice resulted in decreased vascular plaque formation, suggesting that ITGAX plays a pathogenic role of in atherogenesis. CD44, a cell adhesion molecule, is involved in angiogenesis, endothelial cell proliferation, and migration. It is a multi-faceted receptor, which exists in multiple activation states, variant isoforms, as well as intracellular and soluble forms (Pure & Cuff, 2001). Previous studies have shown that it plays an important role in the development of atherosclerotic lesions characterized by VSMC proliferation (Schultz, Rasmussen & Ledet, 2005; Zhao et al., 2008). In response to inflammation, CD44 is upregulated and functionally activated on vascular endothelial, smooth muscle and inflammatory cells (Johnson & Ruffell, 2009). We demonstrated that CD44 was significantly increased in the intimal hyperplasia from injured vessels. Notably, two previous studies provided strong evidence to support the protective role of CD44 in the pathological remodeling process (Vendrov et al., 2006; Zhao et al., 2011). Cysteine protease cathepsin S (Ctss) is mainly over-expressed in human and animal atherosclerosis and abdominal aortic aneurysms (Qin et al., 2013). Some studies have demonstrated that Ctss promotes neovascularization by producing pro-angiogenic factors, stimulates cell proliferation, and enhances the formation of endothelial capillary-like tubules (Premzl, Turk & Kos, 2006). Lymphocyte antigen 86 (Ly86), also known as MD1, has been shown to be involved in various pathophysiological processes including immune regulation, obesity, insulin resistance, and inflammation. Peng et al. (2017) demonstrated that the loss of MD1 aggravated the left ventricular structure and electrical remodeling in response to chronic pressure overload. Allogeneic transplantation inflammatory factor 1 (Aif1) encodes a protein that binds to actin and calcium, and may link inflammation with cell proliferation (Zhao, Yan & Chen, 2013). Several studies have shown that Aif1 and CD68+ macrophages co-localized in human atherosclerotic arteries (Tian, Kelemen & Autieri, 2006), and the severity of atherosclerosis was increased in Aif1 transgenic mice (Sommerville et al., 2012). These studies shown that these DEGs play an important role in immune inflammation, proliferation, migration, and adhesion of intimal hyperplasia.

We used the TRRUST2.0 to analyze the promoter binding motifs and to identify the transcription factors for the key genes in order to directly demonstrate the potential transcriptional regulation mechanism of the above-mentioned key genes in the intimal hyperplasia. We found that the most prominent transcription factor was Nfkb1, which regulates inflammatory response-related genes (Cartwright, Perkins & L. Wilson, 2016) and played a role in the neointima formation after vascular injury (Yoshimura et al., 2001). Additionally, Sp1 is involved in many cellular processes (O’Connor, Gilmour & Bonifer, 2016), including cell differentiation (Thomas et al., 2007), cell growth, apoptosis (Torabi et al., 2018), immune response (Kong et al., 2020), response to DNA damage (Bu et al., 2008), and chromatin remodeling (Cakouros et al., 2001). Sp1 plays an important role in the study of related mechanisms of intimal hyperplasia (Yang et al., 2013; Garrido-Martin et al., 2013). Trp53 responds to a variety of cellular stresses and is involved in the induction of cell cycle arrest, apoptosis, senescence (Levine & Oren, 2009; Freed-Pastor & Prives, 2012), and inhibited the intimal hyperplasia of the carotid artery in rats with balloon injury (Jacob, Hingorani & Ascher, 2012).

Interestingly, this dataset revealed that the deubiquitinating enzyme BRCC36 mRNA expression has an increased trend on the 7th day after surgery in the ligated arteries compared with the control arteries. By contrast, previous study showed that BRCC36 expression increased significantly in the 2th week (Fig. S3). The results suggested that the key DEGs may be involved in the dynamic regulation of intimal formation. Future studies should focus on addressing the detailed alternation, effects, and mechanism of the key DEGs in the development of vascular injury and remodeling. These studies may provide us some new insights in developing effective therapeutic strategies to prevent various vascular diseases such as coronary atherosclerosis, pulmonary hypertension, and restenosis. It is noted that the present study had some limitations. For example, the lack of more public data sets about intimal hyperplasia limited our detailed analysis, discussion, and verification. It is worth mentioning that the right carotid artery taken from an animal subjected to partial ligation of the left carotid artery was used as the control in the original study, and the original method-partial carotid ligation model in mice was adapted (Nam et al., 2009; Dunn et al., 2014). However, in our experiments, we conducted the complete ligation of the common carotid artery to explore a biological question not considered in the original publication, and the results are quite consistent. In addition, the right carotid artery taken from an animal subjected to ligation of the left carotid artery was also used as the control, which is a generally accepted method (Du et al., 2013; Yao et al., 2018; Godin et al., 2000).

Conclusion

We revealed new insights into the underlying mechanism of vascular remodeling by exploring the key DEGs in the development of neointima formation. The results are promising and future study will focus on the key targets from the DEGs. A better understanding of their role will contribute to develop new therapeutic strategies in the prevention of vascular remodeling.

Supplemental Information