Verify Hub Genes of expression profile in aortic dissection

Background To assess the mRNAs expression profile and explore the hub mRNAs and potential molecular mechanisms in the pathogenesis of human thoracic aortic dissection (TAD). Methodology: mRNA microarray expression signatures of TAD tissues (n=6) and no TAD tissues (NT;n=6) were analyzed by Arraystar human mRNAs microarray. Real-time PCR (qRT-PCR) were used to validate the result of mRNAs microarray. Bioinformatic tools including gene ontology, and Kyoto Encyclopedia of Genes and Genomes pathway analysis were utilized. The protein-protein interaction networks were constructed based on data from the STRING database. Molecular Complex Detection (MCODE) and cytohubba analysis were used to infer the most hug gene and pathways. Results: The top 10 hub genes CDK1, CDC20, CCNB2, CCNB1, MAD2L1, AURKA, C3AR1, NCAPG,CXCL12 and ASPM were identified from the PPI network. Module analysis revealed that TAD was associated with cell cycle, oocyte meiosis, p53 signaling pathway, progesterone-mediated oocyte maturation. The qRT-PCR result showed that the expression of all hug genes was significantly increased in TAD samples (p < 0.05). Conclusions: These candidate genes could be used as potential diagnostic biomarkers and therapeutic targets of TAD. Author summary Many basic characteristics underlying the establishment of aortic dissection have not been studied in detail. The presented work sought to understand the pathogenesis of human thoracic aortic dissection by employing bioinformatic tools to explore the hub mRNAs and potential molecular mechanisms of thoracic aortic dissection. Many pathway were thought to have relevant with this disease, but the most important pathway was not define. We used bio-mathematical analysis to explore the potential functions in thoracic aortic dissection and identified the hub genes and explored the intrinsic molecular mechanisms involved in thoracic aortic dissection between two microarray analysis. Finally, we indentified the cell cycle maybe the key pathway in thoracic aortic dissection.


Introduction
Thoracic aortic dissection (TAD) is a common and a life-threatening aortic disease 1 . Despite improvements in medical therapy and surgical or endovascular techniques in recent years, TAD still remains a high morbidity and mortality rate 2 . Owing to the poor results of the existing treatment methods, necessary understanding of the molecular mechanism maybe provided new insights into therapeutic targets for TAD.
Many reports show the degradation of extracellular matrix (ECM) and depletion of vascular smooth muscle cell (VSMC) of the aortic wall are the main histopathological findings [3][4][5] . However, it is remains unclear of the key molecular mechanism of TAD pathogenesis.
In recent years, mRNAs have been reported to participate in the regulation of pathophysiological conditions and have been proved to involve in the progression of cardiovascular disease 6 . Nowadays, researches have focused on the Genomewide Association Studies (GWAS) 7 , which could find relevant genetic variants that may be used as potential biomarkers for diagnosis and targeted therapy. Although high-throughput sequencing technology have provided many diverse expressed genes, different expression profile in TAD provided various results and no reliable results have been identified up to now [8][9][10] . More samples and advanced bio-informatics methods should be used in further study.
In the present study, we used mRNA microarray to acquire differential expression profiles in human TAD tissues and CAD tissues. Subsequently, bio-mathematical analysis were used to explore the potential functions in TAD and identified the hub genes and explore the intrinsic molecular mechanisms involved in TAD. Thus, these data would provide a foundation for new biomarkers and therapeutic targets for human TAD.

Differential expression profiles of mRNAs in TAD and NT group
Volcano plots revealed that mRNAs were differentially expressed in human TAD aortic tissues through microarray technology ( Figure 1). A total of 2834 mRNAs were differentially expressed in TAD group compared with NT group. In total, 1928 mRNAs were up-regulated, and 906 mRNAs were down-regulated (fold change >2.0, P value < 0.05). The top ten up-and down-regulated mRNAs are listed in Table 3.

GO Functional Enrichment Analysis.
To know the functions of all DEGs, we used DAVID online tool, the DEGs functions of GO function enrichment were divided into three groups including BP, CC, and MF ( Figure 2). As shown in the Figure 2 and All the DEGs (fold change >4.0) were analysis using STRING database. Then we put these data into Cytoscape software to construct a PPI network which containing 433 nodes and 348 edges ( Figure 5). In these DEGs, 10 hub genes including CDK1, CDC20, CCNB2, CCNB1, MAD2L1, AURKA, C3AR1, NCAPG,CXCL12, ASPM after calculating. In these 10 hub genes, CDK1 present with the highest degree

Disscussion
TAD is a life-threatening event that carries a high mortality rate 1 . Patients always complain with acute chest and back pain and are often misdiagnosed as acute myocardial infarction 15 . Surgical treatment seems to the best method of this disaster disease 16 . However, the traditional surgical treatment of TAD is very complex and time-consuming with high mortality and poor prognosis. Therefore it is necessary to research of the new biomarkers and therapeutic targets. Up to now, the molecular mechanism of this severe disease remains unclear. TAD such as Marfan syndrome and Ehlers-Danlos syndrome had been confirmed with deficiency of the glycoprotein 17 and abnormal type-III precollagen 18 respectively. However, most patients with TAD do not exhibit such explicit syndromes. These TAD patients always complain with hypertension, atherosclerosis and trauma 15 . In addition, lots of studies had pointed that degradation of ECM and depletion of VSMC play an important role in un-heredity TAD 19 . So studies are primarily focused on the protein-coding genes of ECM 20 and VSMC such as COL3A1, FBN1, LOX, FLNA, ACTA2, MYH11. These protein-coding genes play role in pathological processes of TAD. However, the key pathogenesis of the TAD has not been confirmed up to now.
Modern molecular biology believe that the disease was induced by the changing of various gene expression profiles of tissues or cells. Collection, summary and analysis of the large numbers of these gene can help us to understand the mechanism of the diseases. High-throughput sequencing technology make the identity of the changing gene in tissue available and have been widely used to predict potential targets gene.
Up to now, there are many studies have been performed and thousands of differentially expressed genes had been screened. However, different gene profile are various with each study [8][9][10] . So more further analysis should be carried out and hub genes and pathway should be identify.
This study identify 2834 mRNAs DEGs which included 1928 upregulated DEGs and 906 downregulated DEGs (fold change >2.0, P value < 0.05). Like other studies 8-10 , we classified into three groups including BP,CC, PF by GO terms. GO functional enrichment analysis showed that immune response, response to stress, defense response and immune system process, which was consistent with other studies.
Inflammatory and immunological may both probably lead to the degree of vascular damage of aortic wall [21][22][23] . Reports had presented that inflammatory mechanisms participated in medial degeneration of aortic dissection tissue. The macrophages and activated T lymphocytes were also found in the dissection tissue. T helper 2 response was reported to have relevant with the growth of aneurysms [24][25] . hyper-expression IL-6 and IL-8 in aortic dissection was also present that immunologic pathways were critical in the aortic wall damage 26 . In addition, neutrophils, CD8+, CD28-,IL-6, TNF-α , IL-8, and MCP-1 suggestted that cytotoxic and innate cells were mainly relevant with the pathogenesis of TAD 27 . T helper lymphocytes and activated macrophages, mast cells, T and B lymphocytes that immune system process and inflammatory pathways were involved in the weakening of the aortic wall.
Furthermore, the enriched KEGG pathways of downregulated genes were mainly enriched in dilated cardiomyopathy, arrhythmogenic right ventricular cardiomyopathy, hypertrophic cardiomyopathy, adrenergic signaling in cardiomyocytes and vascular smooth muscle contraction. VSMC was the major cell in aortic media. The function of VSMC such as proliferation and migration acted critical biomechanical properties of the aortic wall. It was affluence by many factor including lymphocytes induce apoptosis of VSMCs and synthesis of MMPs. So many studies aim to research the function of VSMC. Talin-1 had been reported to main exist in aortic media and it was significant downregulation in AD aortic tissue. Further study confirmed that Talin-1 regulating VSMC proliferation and migration and finally caused of pathologic vascular remodeling to change vascular media structure and function and lead to AD 28 . Many gene such as YAP1, Sirtuin-1, PCSK9, polycystin-1, brahma-related gene 1 had also been reported to associated with the pathophysiologic processes of aortic dissection through influence proliferation and migration of VSMC [29][30][31] .
Like previous high-throughput sequencing studies, we also got many relevant pathways through analysising, but we can not sure which one is the most relevant mechanism about TAD and the most hug gene about AD is also not sure. So we carry out some further analysis such as PPI Network and Modular Analysis including many studies aim to research the VSMCs from dissected aorta and indentidy this SMC can proliferated more rapidly than normal VSMCs tissues, and the genes participate in proliferation exhibited an increased expression. In our study, the 10 hug genes are all up expression which show the up regulaition defferent expression will act an important role in TAD which is coninsistance with previous study.
In summary, by means of high-throughput sequencing and data processing as well as qRT-PCR validation, the hub genes including CDK1， CDC20， CCNB2， CCNB1，AURKA，NCAPG and ASPM may have the potential to be used as drug targets and diagnostic markers of TAD. Cell cycle maybe the key pathway in TAD.
However, there were still some limitations: normal aortic tissue used as control group maybe more accuracy. Further experimental studies with larger sample size need to confirm cell cycle pathway in TAD.

Tissue collection
This study was conducted in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of the Second Hospital of Jilin University.
Ascending aortic specimens near the intimal tear were obtained from TAD patients undergoing surgical repair (n=6) if the informed consent was obtained. Aortas specimens derived from patients without aortic diseases undergoing coronary artery bypass ( n=6) graft were obtained as no TAD (NT) group if the informed consent was obtained. General information of all patients was shown in Table 1. There were no significant differences in age, gender, obesity, smoking between the two groups.
Patients diagnosed with TAD were confirmed by the computed tomography angiography (CTA) and excluded from any heredity TAD. All the specimens were immediately frozen in liquid nitrogen, and preserved at -80°C for microarray analysis and qRT-PCR or further usage.

RNA extraction and quality control
Tissue RNA from TAD and NT group was extracted according to the  12 . In this study, the STRING online tool was used to construct a PPI network of upregulation and downregulation diferentially expressed genes (DEGs), with a confidence score >0.7 defined as significant. Then the interaction data were typed into the Cytoscape software 13 to structure a PPI network. Based on the above data, we used Molecular Complex Detection (MCODE) 14 , a built-in APP in Cytoscape software, to analyze the interaction relationship of the DEGs encoding proteins and screening hub gene. The parameters of network scoring and cluster finding were set as follows: degree cutoff = 2, node score cutoff = 0.2, k-core = 2, and max depth = 100.
Quantitative reverse transcription-PCR (qRT-PCR) Validation and Statistical Analysis. qRT-PCR was used to verify the core genes. Total RNA was reverse-transcribed to cDNA using PrimeScript RT reagent Kit with gDNA Eraser (TaKaRa, Japan) according to the manufacturer's instructions. Primer 5.0 software (PREMIER Biosoft, Palo Alto, CA, USA) was used to design primers, and a QuantStudio 7 Flex real-time PCR system (Applied Biosystems, Carlsbad, CA, USA) was used. All primers used in this study were listed in Table 2. All samples were normalized to GAPDH. And the relative expression levels of each gene were calculated using 2−ΔΔCt methods.

Statistical analysis
The data were presented as the mean ± standard deviation. Statistical analyses were performed using SPSS version 20.0 (SPSS Inc., Chicago, IL, USA). The raw data were preprocessed by affy package in R software and limma package in R software .
Comparisons between groups were performed using unpaired Student's t-test. Fisher's exact test was used to evaluate the significance of GO terms and Pathway identifiers enrichment. The false discovery rate (FDR) controlling was used to correct p-value.
FDR and P-value less than 0.05 was considered statistically significant.  [32] Li Z, et al.

Integration of Gene Expression Profile Data to Screen and Verify Hub Genes Involved in Osteoarthritis.
Biomed Res Int. 14, 9482726 (2018).        Red nodes represent hug GENE analysis by cytohubba.

Data availability
Data available on request for the Ethics Committee of the Second Hospital of Jilin University .