Bioinformatics analysis of potential common molecular pathogenesis assumed by intracranial aneurysm, aortic aneurysm and aortic dissection

Chao Zhao (  32279950@qq.com ) Rizhao People's Hospital https://orcid.org/0000-0003-3552-7737 Xinchun Cui Qufu Normal University Rizhao Campus Guodong Liu The Second A liated Hospital of Chongqing Medical University Jianlong Li Rizhao People's Hospital Jinxing Liu Qufu Normal University Rizhao Campus Junliang Shang Qufu Normal University Rizhao Campus Shuai Wang Rizhao People's Hospital Ronghua Shi Rizhao People's Hospital Aihong Wu Qufu Normal University


Introduction
Intracranial aneurysm (IA),abdominal aortic aneurysms (AAA) and aortic dissection (AD) have yet lingered as life-threatening diseases. Currently, endovascular intervention and surgical treatment are the two main preventative methods for them.The molecular mechanisms leading to their initiation, progression and rupture remain incompletely interpreted and were discussed respectively in the past. As a result, no safe and effective noninvasive IA, AAA or AD therapies have been identi ed and implemented in clinical practice by now.AD and thoracic aortic aneurysm(TAA) share common etiologies and feature common pathological characteristics to a great extent 1

.Generally IAs have saccular shapes, whereas
AAAs and TAAs are more often spindle-shaped; Atherosclerosis plays a more signi cant role in AAA, compared with IA and TAA 2,3 .Although these three diseases have different internal molecular mechanisms leading to different clinical manifestations of occurrence and development 4,5 , they are also related in genetic and pathological processes. For example, the prevalence of IA is higher in patients with AD or aortic aneurysm; Most aneurysms are named according to the local parent vessels, but they are often accompanied with systematic vascular lesions 6 .
Expression microarray technology and bioinformatic analysis for genetic dysfunction research have been extensively applied so far, identifying the differentially expressed genes (DEGs) and signal pathways leading to the pathogenesis and progression of many diseases. In order to understand the common molecular mechanisms among these three diseases better, the author research them via bioinformatics methods. In this study, three mRNA microarray datasetsfrom Gene Expression Omnibus (GEO) were included and detected to reveal common DEGs among IA, AAA and AD. Subsequently, Gene Ontology(GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis and protein-protein interaction(PPI) network analyses were conducted to illustrate the molecular mechanisms in their pathogenesis and progression.

Identifying differentially expressed genes
Gene expression data tables of GSE75436 GSE7084 and GSE52093 were obtained via GEO2R webpage application (http://www.ncbi.nlm.nih.gov/geo/geo2r/), which is an R programming languages-based tool for identifying DEGs 7 . In GEO2R, query for each datasets were conducted with de ning and assigning "disease" and "control" groups and selecting "Top 250" DEGs, respectively. Difference of p value 0.05 and|log 2 foldchange | 1 is considered statistically signi cant. Common differently expressed genes cDEGs of these three datasets, the overlapped genes, are illustrated via Venny mapping (https://bioinfogp.cnb.csic.es/tools/venny/).

Functional and pathway enrichment analysis
The Database for Annotation,Visualization and Integrated Discovery Version 6.8 (DAVID) was used to perform GO term and KEGG pathway analysis for the cDEGs (https://david.ncifcrf.gov/).Gene Ontology (GO) enrichment annotation is to locate functional gene products at the subcellular structures(cellular component CC), describe activities that occur at the molecular level(molecular function, MF) and the larger processes accomplished by multiple molecular activities (biological process BP). The Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway database, analyzing the potential relevant biological function, provides useful structured information of a gene network. p value<0.05 and FDR<0.05 was the cut-off criteria in GO and KEGG pathway analysis.
The PPI networks and analysis of most signi cant moduleThe Search Tool for the Retrieval of Interacting Genes (STRING) database Version 11.0 (http://string-db.org/) is a database of known and predicted PPI, consisting of direct (physical) associations and indirect (functional) associations 8 . Input the gene symbols of cDEGs into the "multiple proteins" selection bar of STRING and then the interactions between proteins encoded by cDEGs acquired with cut-off criterion "combined score > 0.4". Then Cytoscape Version 3.7.2 9 were utilized to visualize the PPI network of the cDEGs. Molecular Complex Detection (MCODE) plugin was used to select the most signi cant module from the PPI network with degree cutoff =2, haircut on, node score cutoff = 0.2, k-score=2, maximum depth=100, and nodes more than 5.The GO and KEGG pathway enrichment analyses for genes in this module were performed using DAVID.Hub genes selection and analysis.CytoHubba is a plugin software of Cytoscape. It predicts and explores important nodes and sub-networks in a given network by 12 topological analysis methods including Degree, EPC, MNC, et al. The hub genes, acting more importantly in biological processes than the other DEGs, were selected with degrees≥10 by CytoHubba. Then,the biological process analysis of them was performed and visualized using Biological Networks Gene Ontology tool (BinGO) plugin of Cytoscape with the signi cance level 0.02. BiNGO is a software to select statistically overrepresented Gene Ontology (GO) categories in a set of genes or a subgraph of a biological network, and maps the predominant functional themes of a given gene set on the GO hierarchy. GO and KEGG function analysis of these hub genes were performed in DAVID (p value< 0.05).

PPI network construction and module analysis
The PPI network of cDEGs was mapped intuitively and expansively to illustrate the possible association between down-stream proteins with 137 nodes and 336 edges ( Figure 3A). After MCODE plugin of Cytoscape processing, the most signi cant module was selected with nodes=10 and edges=43 (Fig. 3B), consisting of SMTN Smoothelin MYH11 Myosin Heavy Chain 11 TAGLN Transgelin ACTG2 actin gamma 2, smooth muscle LMOD1 Leiomodin 1 MYLK Myosin light chain kinase VCL Vinculin CNN1 Calponin 1 ACTC1 Actin alpha cardiac muscle 1 MYL9 Myosin light chain 9 . Functional enrichment analysis indicated that these genes in the module were enriched in GO terms and KEGG pathways in Table 1 Functional and pathway enrichment analysis of hub genes A total of 20 hub genes were screened out, and their gene symbols and aliases were shown in Table2 with PPI network in Fig.4A .The biological processes resulted from BingGO analysis with different color depths of each node according to their p value (Fig.4B).The functional analysis of hub genes was conducted through DAVID (Table 3). Muscle contraction, smooth muscle contraction, actomyosin structure organization, in ammatory response co-existed in Fig.4B and Table 3.

Discussion
Thoracic aortic diseases (TADs) include thoracic aortic aneurysms (TAAs) and aortic dissections (ADs), associating with a widely common genetic etiology, and are usually discussed as a whole by many scholars 10,11 . TADs are different from abdominal aortic aneurysms (AAAs) in clinical features, inheritance modes, et al. But TADs and AAAs share several joint pathogenic processes, for example, proteolytic elastic tissue degeneration and smooth muscle dysfunction 5 . In an autopsy study, these thoracoabdominal artery lesions can co-exist 12 . So, there may be same pathogenic molecular mechanisms between TADs and AAAs.The prevalence of IA in patients with aortic dissection or aneurysms is higher than those without 13,14,15 .Degeneration of the elastic tissueand smooth muscle dysfunction are also explicated in IAs formation and rupture 16 .These evidences suggest associations in these diseases pathogenetic process, but still understood unclearly. In this paper, the common differentially expressed genes of the three diseases IA, AD and AAA, and their possible signaling pathways was detected in silico, so as to provide basis for future studies.
In this study, to reveal possible common pathogenesis-related genes of IA, AAA and AD, the author selected three microarray pro le datasets ,GSE75436 of IAs, GSE7084 of AAAs and GSE52093 of ADs.
DEGs of these three datasets screened through GEO2R, respectively. In summary, 178 common DEGs of the three datasets were identified. The enrichment analysis demonstrated these DEGs were primarily enriched in biological processes, including muscle cell contraction, cell chemotaxis, cell adhesion and protein localization to cell surface.To date, smooth musclecells (SMCs), leukocytes, complements, immunoglobulins, cytokines have been a rmed to be contributors to IA pathogenesis.Their roles in the biological processes mentioned above need further investigation. In GO cellular component enrichment analysis, the most enriched component was Z disc, whereas in GO molecular function analysis, it was actin binding. In KEGG pathway analysis, DEGs were mostly enriched in vascular smooth muscle contraction, neurotrophin signaling pathway, cGMP-PKGsignaling pathway, osteoclast differentiation, arginine and proline metabolism, et al. in Fig.2D. Only smooth muscle contraction, Toll-like receptor signaling pathway and cGMP-PKG signaling pathway have been researched extensively by now 17,18,19 .
SMCs contractile dysfuction (dedifferentiation) refers to the degeneration from an initial contractile status to a in ammatory and matrix remodeling status under the stimulation of the cellular and extracellular environment. It is an integral part of the in ammatory response along with subsequent cell death. It is a key predisposing process of aortic dissection, aortic aneurysms and intracranial aneurysms 20,21,22,23  in vitro experiments implicated that ACTG2, a target of miR-193a-3p, participated in the development of AD 28 . Feng J reported CNN1 was downregulated in aortic dissection vascular smooth muscle cells which might be a key pathogenetic gene in vascular diseases 29 . LMOD1 is a highly specific contractile gene for smooth muscle lineages. myosin light chain 9 MYL9 is involved in in ammatory response, which is a main biological process of aneurysmal diseases 30 . Defects of cytoskeleton proteins VCL and actin ACTC1 are the cause of dilated cardiomyopathy. These genes in the most signi cant module except MYH11 and MYLK have not been discussed in aneurysmal diseases and need continous research in the future.
In this study, smooth muscle contraction, actomyosin structure organization, mesenchyme migration and platelet aggregation are common biological processes shared by genes in the most signi cant module and hub genes. Smooth muscle contraction, actomyosin structure organization are important in preventing aneurysm (or dissections) pathogenesis, validated by many researches 5 . Our findings are in accordance with previously reported data on the role of vascular SMCs in aneurysm pathogenesis. Nevertheless, mesenchyme migration and platelet aggregation have not been understood completely. Platelet aggregation and thrombosis are the subsequent events of elastin integrity loss after AAA initiation. Expounded by quite limited articles, platelet aggregation plays a signi cant role in AAA onset and progression, and platelet inhibitors can attenuate aneurysm formation in rat AAA model, though lacking of clinical evidence 31 . In ammatory factors TGFβ released during platelet aggregation, can promote the transformation of SMC phenotype from contractile status to in ammatory status, along with in ammatory responses, thereby contribute to aneurysm formation 32 . These evidences lead us to speculateon a positive feedback between aneurysm pathogenesis and platelet aggregation.
Focal adhesion signaling pathway co-exhibited in Fig 2D, Table 1 Table 3. Focal adhesions function as the transmembrane signal transduction interface between extracellular matrix and intracellular cytoskeleton, and has not been elucidated extensively in aneurysm pathogenesis. Testin (TES), a focal adhesion scaffold protein, was validated as a potential contributor to TAA onset 33 . THSD1, a novel nascent adhesion protein, is provided likely to cause intracranial aneurysm in both familial and sporadic patients 34 . MYL9, VCL and MYLK mined in this article may play signi cant roles in aneurysmal diseases pathology via focal adhesion pathway, which worthy of further investigation.
Genetic variation is animportant factor of vascular diseases, and the understanding of its role is gradually enriched. It can lead to biological degeneration in blood vessels, including cellular and tissue events contributing to intracranial and aortic aneurysms (or dissection), which may provide new therapeutic targets. Genetic testing is regular procedure of clinical diagnosis and treatment. The genes discussed in this paper, may be developed as potential predictors of systemicaneurysms in genetic test or targets of vascular therapy.
There are some limitations to this study. The rst one is merely single-aspect analysis. Other methods, e.g. gene set enrichment analysis (GSEA), can provide understanding of genes function from another aspect. The second one is lacking experimental and clinical feature validation, such as in vitro and in vivo experiments, patients' age, sex, different types of aneurysms, and so on. And in addition, the sample count of each group is limited. False-positive incidences in independent microarray analysis hinder reliable data acquisition.
To date, genetic sharing among IA, AAA, and AD has been considered mainly within families 35,36 , but possible genetic corelations in sporadic aneurysmal cases are also worthy of attention. Therefore, further research focusing on the aforementioned hub genes and biological processes calls for more evidences with a systematic view.

Conclusion
This article aiming to nd possible pathogenesis-related genetic overlap of IA, AAA and AD, identi ed 178 DEGs, which include SMTN, MYH11, TAGLN, ACTG2, CNN1, MYLK, LMOD1, MYL9,VCL and ACTC1 in the the most signi cant module. Except for those con rmed biological processes, mesenchyme migration and platelet aggregation are common biological processes shared by genes in the most signi cant module and the hub genes. Focal adhesionssignaling pathway highlighted in this analysis, has not been elucidated extensively in aneurysm pathogenesis by now. Basic and clinical experiments are needed to Tables   Table 1 GO terms and KEGG pathways analysis of genes in the most significant module   Category  Term  PValue  Genes  FDR   GOTERM_BP_DIRECT GO:0006936~muscle contraction  1.18E-09  LMOD1, MYH11, MYL9     legend.docx