Screening of hub inflammatory bowel disease biomarkers and identification of immune-related functions based on basement membrane genes

Inflammatory bowel disease (IBD) is an inflammatory disease involving the ileum, rectum, and colon, among which ulcerative colitis (UC) and Crohn's disease (CD) are the most common. As a systemic disease, IBD affects the quality of life of patients, and severe cases increase the risk of colon cancer [1, 2]. The main pathological feature of IBD is intestinal inflammation and causes varying degrees of intestinal mucosal damage, including extensive epithelial cell death, ulcers, crypt abscesses, and the formation of fibrosis [3, 4]. At present, the pathogenesis of IBD is still unclear, and the possible causes include genetic susceptibility, commensal flora disturbance, epithelial barrier defect, immune dysregulated response, and environmental factors [5]. Therefore, it is of great significance to further explore new molecular markers and molecular mechanisms of IBD.

The BM is the evolutionary conserved animal extracellular matrix and forms sheet-like structures that epithelialize and surround most tissues [6‐8]. Two independent planar networks of laminin and type IV collagen molecules associate with cell surface interactors and provide a scaffolding structure for building BMs along tissues. These functions will suggest that the BM has an important relationship with the occurrence and development of IBD [9]. Recently, Jayadev et al.[10] systematically defined 222 humans BM genes, which identified the enormous complexity of BM and important impact on human health. We will further use these 222 humans BM genes to explore new molecular markers of BM closely related to IBD. BM has a variety of components that can be used to resist mechanical stress, determine tissue shape, and create diffusion barriers. They also provide cues directing cell polarity, differentiation, migration, and survival. Variations in more than 20 BM genes underline their diverse and fundamental functions that underlie human disease. BM proteins are targets of autoantibodies in immune diseases, and defects in BM protein expression and turnover are hub pathogenic aspects of cancer, inflammation, and fibrosis [11].

Advances in high-throughput sequencing technologies such as RNA-seq and microarrays have provided opportunities to comprehensively characterize the molecular features of tumorigenesis [12]. Furthermore, the application of these high-throughput technologies has facilitated the identification of promising biomarkers for cancer diagnosis and prognostic assessment. Appropriate IBD data were obtained through Gene Expression Omnibus (GEO). Based on BM genes, we grouped and compared them separately, performed a series of bioinformatics analyses, and mined new molecular markers to predict related drugs and miRNA, which aims to lay the foundation for the development of IBD treatment, identification of subtypes, and further in-depth research.

In recent years, with the continuous research on IBD, the expected therapeutic effect has still not been achieved. The course of IBD is usually prolonged and even requires lifelong treatment. Its pathogenic mechanism is complex [29‐31] and is closely related to the inflammatory immune response of the intestine [32]. Therefore, the in-depth study of its immune regulation mechanism has always been a clinical research hotspot. The pathological manifestations of IBD are mainly in the congestion and erosion of the intestinal tissue mucosa, the structure of the matrix is damaged, accompanied by the infiltration of a large number of inflammatory cells [33, 34]. BM is an extracellular matrix structure, which is formed by the combination of collagen and adhesions protein to form a tissue structure, which is closely related to the formation of the intestinal mucosal structure. Recently, Jayadev et al. [10] successfully identified the BM gene set by network analysis and screening of elegans and zebrafish, which finally clearly defined 222 human BM set and affirmed the important impact of BM genes on human health. It is composed of 160 basement members matrix and 62 cell surface interactors, which are the scaffolding structure of the tissue and play an important role as a protective barrier. However, there are fewer studies on BM genes on diseases, which the role of BM genes in IBD diseases has not been carried out, thus BM genes are important for IBD studies. We extracted the BM gene set for the study to discover the key pathogenic genes of IBD. A further prediction of related drugs and miRNAs has important clinical application and guiding significance for basic experiments.

We are interested in a multi-faceted exploration of IBD through BM genes. In both CD and UC, we were surprised to find that these two disease subtypes collectively focus on the ADAMTS family [35, 36]. At present, the ADAMTS family has been studied to be associated with the occurrence of various diseases and has a hub role in tissue development and homeostasis. Its functions are widely determined by their interactions with the extracellular matrix and proteins in the extracellular matrix [37]. It is also suggested that its functional form of action may be closely related to the histopathological destruction and immune cell infiltration of IBD. In CD, we further screened out the central hub gene ADAMTS17 by machine learning. It is related to fibrillin biology [38‐41], and its downregulation is associated with dysplasia, and ECM substrates are functionally linked, consistent with our analysis that low expression promotes the development of CD [42]. It has also been found that there is a significant negative correlation with immune functions such as T lymphoid helper cells, paracrine, and type I interferon responses, which may be involved in the immune response of CD. In UC, central hub genes ADAMTS17 and ADAMTS9 were identified. Among them, ADAMTS9 belongs to the hub enzyme of proteoglycan degradation [43‐45], and its harmful role in osteoarthritis, rheumatoid arthritis, and intervertebral disc degeneration have been widely described [46]. Possibly related to disrupting the function of the extracellular structural matrix, consistent with our analysis. In the immune response, it is significantly related to the promotion of B cells, CD8 + T cells, and DCs, the promotion of inflammatory responses, and the inhibition of NK cells and other immune cells. Chen et al. suggested that there is a strong interaction force between BM and immune cells, while regulating immune cell levels [11, 47, 48]. BM genes are also involved in the regulation of immune cells including mast cells and may be associated with the activation of loaded inflammatory networks including mast cells when inflammatory factors bind to the BM [49‐51].

In addition, we are also interesting to find that there are also significant differences between CD and UC [52]. The clinical manifestations of CD and UC are different, but some patients are difficult to identify, especially in the early stage of the disease. Therefore, it is of great clinical significance to find new biomarkers for identification. We found that the BM-related differential expression genes of the two subtypes were still enriched in the cell–matrix organization and structural organization, which indicated that the degree of structural changes in the stromal organization was different between them [53]. The analysis showed that the hub genes of the difference center were SPARC, POSTN, and ADAMTS2. The gene expression levels of these three genes in UC were higher than those in CD, and the AUC of the SPARC gene was 0.71, indicating a more accurate prediction effect. It is mainly involved in the regulation of cell adhesion, proliferation, migration, and tissue remodeling, and is related to the expression of fibroblasts [54, 55]. This may suggest that the differences in the pathological and immune-inflammatory responses developed between the two are closely related.

Our model has a good prediction of the pathogenic factors of CD and UC, and also explains the differences between the two to a certain extent, which can guide clinical treatment and basic research plans. As our findings were derived from bioinformatic analysis, there are deficiencies in the authenticity and credibility of the findings without experimental evidence of protein, and further experimental studies are required to confirm these results.

In conclusion, we constructed and validated a nomogram model of CD and UC composed of BM genes through a series of multiple group bioinformatics comparative analyses. In CD, the central key gene is ADAMTS17; in UC, the central key gene is ADAMTS17 and ADAMTS9, which are closely related to the progression of disease, hoping to provide a new direction for the diagnosis and treatment of IBD. Additionally interesting, we found that the important differential significance of SPARC, POSTN and ADAMTS2 between CD and UC, which was able to clarify the similarities and differences between CD and UC, provides new insights into the identification of IBD subtypes, and has important research significance.