Introduction

Ischemic stroke is the most common type of cerebrovascular disease, with four high characteristics such as high morbidity, high disability, high recurrence and high mortality, which brings a heavy medical and economic burden to patients' families and society. By 2030, almost 3.9% of the adult population in the United States would have had a stroke, considering that there has been a 20.5% increase in its prevalence since 20121. Of all strokes, 87% are ischemic strokes (ISs)2. In the United States, strokes have caused serious and long-term neurological disability in3 approximately 3% of the males and 2% of the females that have suffered them. It is predicted that the total direct medical costs of stroke-related diseases will increase between 2015 and 2035 from $36.7 billion to $94.3 billion, with much of the anticipated increase arising from affliction of people more than 80 years of age4. The Global Burden of Disease Study 2019 (GBD 2019) showed that the global prevalence of stroke was 104.2 million and that of IS was 82.4 million5.

Considering the high rates of morbidity and mortality, ferreting out the causes and the potential molecular mechanisms of stroke and identifying molecular biomarkers for early diagnosis, preventive treatment, and precise therapy are critically important. Familial epidemiological studies suggest that stroke has a genetic component. Genes modestly associated with stroke have been identified in several studies. Several heritable disorders associated with stroke have also been reported6.

Inflammatory bowel disease (IBD), including ulcerative colitis (UC) and Crohn's disease (CD). It was first identified in the western developed countries, and the cause is still unknown, but the number of patients in Europe and the United States has accounted for 0.5% of the world population. It is an idiopathic inflammatory disease of the intestinal tract involving the ileum, rectum and colon. Clinical manifestations include diarrhea, abdominal pain, and even bloody stools. In the past 20 years, the incidence and prevalence of inflammatory bowel disease (IBD) has risen steeply in developing countries in Asia, South America, the Middle East, and Africa, with the rate of increase particularly significant in South America and East Asia, and has become a global disease A 2017 retrospective study of the incidence and prevalence of IBD worldwide published in The Lancet2 showed that the incidence in Europe and the United States has reached a plateau while non-Western countries are in the first phase of new case growth. More than 2 million Europeans and 1.5 million North Americans currently suffer from inflammatory bowel disease. And according to 2014 data from the Chinese Center for Disease Control and Prevention, the total number of IBD cases in China between 2005 and 2014 was about 350,000. One study estimated that by 2025, the number of people with IBD in China will reach 1.5 million7.

Compared with the general population, there was a significant difference in the risk of stroke in patients with IBD8. Patients with IBD have an abnormal immune response to microbial dysbiosis, leading to a sustained stimulated pro-inflammatory state, and pro-inflammatory Th1, Th17, and γδ T cells are often associated with increased inflammatory injury, while subsets of T cells can help or exacerbate ischemic brain injury9,10. In addition, inflammatory bowel disease increases patients' risk of atrial fibrillation11, predisposes patients to recurrent ischemic cerebrovascular events, and stroke survivors with inflammatory bowel disease are more likely to have a worse prognosis than those without inflammatory bowel disease12. In addition, one study used two-sample bidirectional Mendelian randomization (MR) to analyze the relationship between stroke and osteoarthritis (OA), a common general inflammatory disease. Among the three categories of OA at any site, knee OA and hip OA, hip OA was found to be a potential risk factor for overall stroke, IS and small vessel IS. The mechanisms linking this site-specific OA to stroke remain to be further studied13. Potential mechanisms between inflammatory bowel disease and ischemic stroke may include activation of the innate immune system, systemic inflammation leading to vascular endothelial dysfunction, increased cholesterol biosynthesis, and thrombotic states. All of these may exacerbate atherosclerotic lesions and increase the risk of atherosclerosis, thereby promoting cerebrovascular events, including large- and small-vessel cerebral infarcts. To this end, protein interaction data obtained from a wide range of cellular and biochemical model systems can be used to build protein–protein interaction (PPI) networks from genes associated with inflammatory bowel disease and stroke, which would enable a better understanding of genetics for ischemic stroke pathogenesis in terms of biological pathway and functional analysis, integrating relevant genetic risk factors. As biotechnology continues to advance, PPI network analysis may be a powerful time- and cost-efficient approach to identify potential biological pathways in ischemic stroke, key players or candidate genes involved in the inflammatory bowel disease-stroke linkage.

Therefore, the aim of this study was to identify key proteins and bioregulatory pathways involved in the physiopathology of inflammatory bowel disease and stroke through bioinformatic analysis and application of surveillance-related proteins.

Materials and methods

Data collection

We downloaded data on inflammatory bowel disease and ischemic stroke from the NHGRI-GWAS database by entering the keywords ischemic stroke and inflammatory bowel disease, respectively, in the search field of the NHGRI-GWAS database (https://www.ebi.ac.uk/gwas/, accessed on 27 January 2022). The GWAS-related studies on ischemic stroke included more than 2,951,164 individuals covering Europeans, East Asians, South Asians, African Americans or African Caribbean, Hispanics or Latinos. The GWAS-related studies on inflammatory bowel disease included approximately 640,515 individuals from sub-Saharan African, European, East Asian, South Asian, African American or African Caribbean, Hispanic, or Latino ethnic populations. The GWAS data were all from different populations because none of the included studies combined data from both ischemic stroke and inflammatory bowel disease conditions.

Data organization and analysis

We used Microsoft Office Excel to perform basic processing of the data. After removing duplicates by excel, the total number of genes involved in ischemic stroke was 418, while the number involved in inflammatory bowel disease was 400.We then accessed the string online website (version 11.5) and selected the multi-protein model to construct the protein–protein interaction network. We set the maximum number of interactors displayed to 10 and the minimum required interaction score of 0.4 for the first analysis. Based on the analysis, we then determined if a higher interaction score of 0.7 or 0.9 would be chosen as the cutoff score for the analysis.

Construction and analysis of protein–protein interaction network

The STRINGdb v11.5 package (https://string-db.org/cgi/input.pl?) was used toconstruct a PPI network of the proteins encoded by the genes. Furthermore, Cytoscape v3.7.1 was utilized to construct a PPI network, considering confidence a score ≥ 0.9 and maximum number of interactors = 10 as cut-off criteria, and the interaction of the candidate gene-encoded proteins was analyzed. Moreover, the Network Analyzer plug-in was used to calculate node degrees, i.e., the number of interconnections useful to filter hub genes from the PPI network. The Molecular Complex Detection (MCODE) plug-in of the Cytoscape tool was employed to visualize the significant gene clusters in IS with a degree cutoff = 2, node score cutoff = 0.2, k-core = 2, and max. depth = 100. The criteria for selecting the two most significant clusters were set as MCODE scores ≥ 4 and number of nodes ≥ 4. The corresponding proteins in the central nodes might be key proteins and hub candidate genes with important physiological regulatory functions.

GO and biological pathway enrichment analysis

GO and pathway enrichment analyses of genes were performed using multiple online databases. We submitted the genes to the DAVID online program (https://david.ncifcrf.gov/; version: 6.8) with p < 0.05 as the cut-off criterion. Besides, GO and pathway analyses were carried out using KEGG (http://www.genome.jp/kegg), Reactome (http://www.reactome.org), BioCyc (http://biocyc.org), and PANTHER (http://www.pantherdb.org)14. The enriched GO terms were ranked according to their p-values and visualized as bar charts; p < 0.05 was considered as statistically significant15,16.

Corroboration of key pathways

To further validate the key signaling pathways obtained from the above bioinformatic analysis, we used data-independent acquisition mass spectrometry (DIA-MS), one of the highly regarded mass spectrometry acquisition techniques in recent years, to identify and quantify proteins in samples from 8 blood samples (4 ischemic stroke samples and 4 normal population samples). The main steps were as follows: first, whole blood was collected using a vacuum blood collection tube and gently mixed by inverting up and down 5–6 times; the blood was left to stand for 30 min at 4 °C to allow coagulation; the supernatant was immediately taken by centrifugation at 1700g for 10 min at 4 °C; the upper layer of yellowish serum was transferred to a centrifuge tube using a pipette; liquid nitrogen was snap frozen for 5 min and stored at − 80 °C for backup. Then, protein extraction was performed, and then a portion of the prepared total protein solution was trypsinized, and the enzymatically cleaved peptides were desalted using a SOLA™ SPE 96-well plate for LC–MS/MS high-resolution mass spectrometry detection (Full MS: Resolution = 120,000, AGC target = 3e6, Maximum injection time = 50 ms, Scan range = 350–1250 m/z; MS2: Resolution = 15,000, AGC target = 1e5, Maximum injection time = 35 ms, Scan range = 200–2000 m/z, Isolation window = 1.4 m/z, Normalized Collision Energy = 28), and the high-abundance proteins were removed using a MilliPore de-abundance kit before detection. The LC–MS/MS raw files were imported into Spectronaut Pulsar software for matching, and the machine signals were transformed into peptide and protein sequence information, and then combined with sequence information, peptide retention time, and fragment ion information for the spectral library building to facilitate the subsequent DIA analysis. Based on obtaining plausible proteins, we adopted the following criteria to screen the differential proteins: FoldChange ≥ 1.2, P ≤ 0.05 as up-regulated proteins; FoldChange ≤ 0.833, P ≤ 0.05 as down-regulated proteins. To further understand the biological functional information of differential proteins, we performed GO and pathway enrichment analysis on candidate DEPs using multiple online databases.

Results

Identification of genes in ischemic strokes and inflammatory bowel disease

NHGRI-GWASCatalog supplies a publicly hand-curated collection of published GWAS assaying more than 100,000 single-nucleotide polymorphisms (SNPs) and all SNP-trait associations with p < 1 × 10–5.The Catalog includes 5595 curated publications of 195,610 SNPs and 326,947 associations. In addition to SNP-trait association data, Catalo is a global public source that archives and makes freely available high-throughput functional genomics data submitted by researchers. The GWAS datasets of ischemic stroke (EFO ID: HP_0002140, include 17 reported traits with 314 Associations) and inflammatory bowel disease (EFO ID: EFO_0003767, include 11 reported traits with 526 Associations, excluding child trait data) were taken from this repository, we extracted 418, and 400 genes. After performing an integrated bioinformatical analysis using jvenn (https://www.vandepeerlab.org/?q=tools/venn-diagrams)17, the intersection of these datasets yielded 12 genes, and the total number of unique elements is 806 (Table 1).

Table 1 Identification of genes in ischemic strokes and inflammatory bowel disease (calculate and draw custom Venn diagrams).
Full size table

Construction and analysis of protein–protein interaction network

Using the online web tool STRINGdb and the Cytoscape software platform, a total of 589genes out of the 818 identified, were used to construct a PPI network complex that contained 587 nodes and 366 edges (Fig. 1). According to their degree of importance, we chose five significant clusters from the PPI network complex for further analysis using Cytoscape MCODE. Pathway enrichment analysis showed that cluster 1 (Fig. 2A) consisted of 14 genes primarily associated with Cytokine Signaling in Immune system and Interleukin-23 signaling, cluster 2 (Fig. 2B) consisted of 6 genes primarily associated with U2-type spliceosomal complex and Spliceosome, cluster 3 (Fig. 2C) consisted of 6 genes primarily associated with Degradation of GLI1 by the proteasome and Proteasome, cluster 4 (Fig. 2D) consisted of 5 genes primarily associated with Blood clotting cascade and Fibrinolysis, whereas cluster 5 (Fig. 2E) consisted of 9 genes primarily associated with Interleukin-4 and Interleukin-13 signaling and Signaling by Interleukins. Furthermore,The interactions of the 40 hub node genes are visualized using a heatmap plot (Fig. 3).

Figure 1
figure 1

Protein–protein interaction (PPI) network between ischemic stroke and inflammatory bowel disease with minimum required interaction score 0.900, the highest confidence as the cut-off criterion.

Full size image
Figure 2
figure 2

Five significant clusters from the PPI network complex for further analysis using Cytoscape MCODE. (A) Cluster 1 consisted of 14 genes primarily associated with Cytokine Signaling in Immune system and Interleukin-23 signaling; (B) cluster 2 consisted of 6 genes primarily associated with U2-type spliceosomal complex and Spliceosome; (C) cluster 3 consisted of 6 genes primarily associated with Degradation of GLI1 by the proteasome and Proteasome; (D) cluster 4 consisted of 5 genes primarily associated with Blood clotting cascade and Fibrinolysis; (E) cluster 5 consisted of 9 genes primarily associated with Interleukin-4 and Interleukin-13 signaling and Signaling by Interleukins.

Full size image
Figure 3
figure 3

The interactions of the 40 hub node genes.

Full size image

GO and pathway enrichment analysis

GO and pathway enrichment analyses of these significant genes were performed using multiple online databases. The functional GO terms were classified into three categories: molecular function (MF), biological process (BP), and cellular component (CC). As shown in Fig. 4 and Table 2, in the BP category, the genes were primarily enriched in Cellular response to cytokine stimulus, Cell surface receptor signaling pathway, Response to cytokine, et al. In the MF category, the genes were primarily enriched in Protein binding, Protein kinase binding, Kinase binding, et al. In the CC category, the genes were primarily enriched in Plasma membrane, Cell periphery, Integral component of plasma membrane, et al.. These results show that most of the identified genes were significantly enriched in protein binding, innate immune response, defense response to virus, cytosol, and negative regulation of apoptotic process terms.

Figure 4
figure 4

Gene ontology analysis and significant enriched GO terms of genes associated with Ischemic stroke and inflammatory bowel disease. GO analysis classified the DEGs into three groups (i.e., biological process, molecular function and cellular component).

Full size image
Table 2 GO and pathway enrichment analyses of significant genes in ischemic strokes and inflammatory bowel disease.
Full size table

Pathway enrichment analysis

Functional and signaling pathway enrichment analyses of these significant genes were conducted using multiple online databases. By KEGG pathway analysis, these genes were mainly enriched in Pathways in cancer, Inflammatory bowel disease, Th17 cell differentiation, Human T-cell leukemia virus 1 infection, Cytokine-cytokine receptor interaction. Furthermore, We used 'Reactome pathways' to analyze these important genes and calculated the false discovery rate using Fisher's Exact, and the results showed that Cytokine Signaling in Immune system, Signal Transduction, Signaling by Interleukins, Immune System, Diseases of signal transduction by growth factor receptors and second messengers, and Transcriptional regulation by RUNX3 were the significant signaling pathways. In addition, we also used "WikiPathways" analysis as a complement to show that these significant genes aremainly enriched in Neuroinflammation and glutamatergic signaling, T-cell activation SARS-CoV-2, Non-genomic actions of 1,25 dihydroxyvitamin D3, Allograft Rejection, and Regulatory circuits of the STAT3 signaling pathway. Together, these results showed that these significant genes had several pathways in common, including those involved in immune system, cytokine signaling in immune system, adaptive immune system (Fig. 5).

Figure 5
figure 5

Significantly enriched pathway terms of genes associated with ischemic stroke and inflammatory bowel disease.

Full size image

Corroboration of predicted signaling pathways

To further corroborate the key signaling pathways obtained from the above bioinformatic analysis, we used DIA-MS, one of the highly regarded mass spectrometry acquisition techniques in recent years, to identify and quantify proteins in samples from 8 blood samples (4 ischemic stroke samples and 4 normal population samples) (Table 3), 60 down-regulated and 18 up-regulated protein (Tables 4, 5) followed by pathway enrichment analysis with Reactome pathways, and revealed some aspects of over-representation in Complement cascade, Regulation of Complement cascade, Scavenging of heme from plasma, innate immune system and binding and uptake of ligands by scavenger receptors (Table 6).

Table 3 Clinical characteristics of 4 ischemic stroke (IS) patients and healthy (HE) (n = 4).
Full size table
Table 4 78 differentially expressed proteins (DEPs) were identified from four profile datasets, including 18 up-regulated and 60 down-regulated proteins in ischemic stroke.
Full size table
Table 5 The gene ontology (GO) enrichment analysis of differentially expressed proteins/genes in ischemic stroke.
Full size table
Table 6 The significant enriched analysis of differentially expressed proteins (DEPs) in ischemic stroke.
Full size table

Discussion

For this exploratory bioinformatics analysis, potential PPI networks between ischemic stroke and inflammatory bowel disease were predicted using the open catalog of human GWAS studies. Although this predicted PPI network requires careful interpretation and preclinical and clinical validation, it shows new evidence that will contribute to new studies on the relationship between ischemic stroke and inflammatory bowel disease. In the following we propose different hypothetical biological scenarios depending on our PPI network, by which inflammatory bowel disease can increase the risk of ischemic stroke.

It has long been known that cardioembolism, large artery atherosclerosis, and small vessel occlusion are three main pathophysiological mechanisms in patients with IS18. Moreover, cerebral ischemia is responsible for producing various immune cell mediators, which can exacerbate ischemic brain injury. CNS inflammation induced by ischemic conditions plays an essential role in stroke pathophysiology and exacerbates infarct formation at the injury site19,20.

As previously mentioned, in our study, the five most significant clusters from the PPI network were identified by integrated bioinformatics analysis. These clusters contained 40 hub genes. The first one consisted of 14 genes (KAT2B, IRF4, IL23R, SMARCA4, IRF1, IRF5, IRF6, RELA, IL12B, STAT4, STAT3, JAK2, TYK2, ARID1A) (Fig. 4B) whose biological functions could be associated with Cytokine Signaling in Immune system, Interleukin-23 signaling, Signaling by Interleukins, Th17 cell differentiation, and JAK-STAT signaling pathway. By analyzing the first cluster, we found that various immune cell mediators influence the various processes of atherosclerosis mainly by affecting inflammation, with cytokines and interleukin-related signaling having a particularly significant impact. IRF5 expression was significantly increased in peripheral blood mononuclear cells (PBMCs) and colonic inflammatory tissue in patients with IBD and was significantly associated with IBD activity21. The activated IRF5 increased macrophage infiltration and lipid accumulation, with concomitant reductions in smooth muscle cells and collagen content, weakening plaque stability and promoting the formation of atherosclerotic plaques 22. KAT2B upregulation promotes acetylation of HMGB1, promotes pro-inflammatory cell polarization and macrophage recruitment, and promotes atherosclerotic progression 23. Dysregulation of KAT2B may lead to inhibition of IL-10, a key anti-inflammatory cytokine, disrupting innate and adaptive inflammatory responses and promoting the development of inflammatory bowel disease24. In peripheral inflammation, interferon regulatory factor-5 (IRF5) and IRF4 regulate M1 and M2 activation of macrophages, respectively, leading to a greater risk of ischemia in the elderly brain25. Traditional T cells coordinate the inflammatory response in atherosclerosis and their function is altered by the lipoprotein milieu and complement activity. Hypercholesterolemic states are associated with upregulation of DAF expression in circulating T cells and increased levels of nuclear factor kappa B (NF-kB) and interferon regulatory factor 4 (IRF4) 26. The inhibition of IRF5 expression and promotion of IRF4 expression in lamina propria monocytes induces macrophage polarization to the M2 phenotype, which improves the inflammatory response in inflammatory bowel disease27, and increased IRF4 expression results in enhanced M2 activation, reduced pro-inflammatory response, and improved stroke prognosis, whereas downregulation of IRF4 results in increased IRF5 expression, enhanced M1 activation, increased pro-inflammatory response, and functional poorer recovery28. The inflammatory response, immune response, and cytokine-cytokine receptor interactions may play an important role in the progression of atherosclerosis. Co-operative inflammatory signaling by TLRs, STAT/IRF, and IFNs leads to M1 polarization of macrophages, which contributes to the development of atherosclerosis29. In response to TNF-α stimulation, Ataxin-10 promoted the cytoplasmic localization of IRF-1, thereby inhibiting VCAM-1 transcription 30. IL23R-dependent gamma-delta T cells can release IL-17 and GM-CSF, which together induce inflammation and necrosis in macrophages. Il-23R + gamma-delta T cells locally promote early lesion formation in the aortic root and contribute to the expansion of the necrotic core, a hallmark of vulnerable atherosclerotic lesions 31. Interferon regulatory factor 1 (IRF-1) contributes to the pathological phenotype of VSMCs during atherosclerosis by increasing CCL19 transcription, whereas silencing IRF-1 inhibits angiotensin proliferation and migration and downregulates CCL19 expression, thereby suppressing atherosclerosis 32. RelA-driven escalation of pro-inflammatory gene responses in intestinal epithelial cells exacerbates inflammatory cell infiltration and colonic lesions, driving inflammation associated with inflammatory bowel disease (IBD) 33. Upregulation of RELA /p65 inhibits Hif-1α, thereby suppressing neuroinflammatory activity after ischemic stroke, and attenuating brain injury caused by cerebral ischemia34. SMARCA4/BAF190A may promote neural stem cell self-renewal/proliferation by enhancing Notch-dependent proliferative signaling while insensitizing neural stem cells to SHH-dependent differentiation cues. Inflammation of endothelial cells induces inflammatory cytokine production and monocyte adhesion, which are key events in initiating atherosclerosis. aberrant activation of STAT3 has been shown to contribute to the development and progression of atherosclerosis, and signal transducer and activator of transcription 3 (STAT3) have key roles in endothelial cell dysfunction, macrophage polarization, inflammation, and immunity during atherosclerosis35. Reduced phosphorylation of signal transducer and activator of transcription 3 (STAT3) and nuclear levels of STAT3 and interferon-regulated transcription factor-1 (IRF1), which in turn inhibits the STAT3/IRF1 pathway in vascular endothelial cells, produce anti-atherosclerotic effects on TNF-α-induced inflammation36. As the primary pathological basis of ischemic stroke disease, its pathogenesis has been shown to involve an imbalance in anti-inflammatory/pro-inflammatory processes. IL-23R can enhance its antigen-presenting capacity through an autocrine pathway, allowing it to infiltrate the lesion site, promoting its secretion of large amounts of inflammatory factors, and upregulating pro-inflammatory DCs and macrophages. The IL-23 binds to IL-23R on the surface of target cells through Janus kinase 2/signal transducer and transcriptional activator channels to deliver signals involved in chronic inflammatory and autoimmune diseases37. Among the nine IRFs, at least three (IRF-1, IRF-5, and IRF-8) are involved in pro-inflammatory M1 commitment, while IRF-3 and IRF-4 control M2 polarization38. IRF6 is a potential causative gene for ischemic stroke39. STAT4 deletion reduces perivascular and visceral AT inflammation, which may reduce atheromatous plaque formation through direct or indirect effects40. The mRNA levels of IL-23 and IL-23R were significantly higher in carotid plaques compared to non-atherosclerotic vessels. The coalition showed co-localization with plaque macrophages. IL-23 accentuated tumor necrosis factor release in monocytes from patients with carotid atherosclerosis, and high plasma levels of IL-23 may increase mortality. IL-23 is associated with disease progression in patients with carotid atherosclerosis and may be involved in IL-17-related mechanisms41. Inhibition of TNF-α-induced phosphorylation of IkappaB-α and nuclear translocation of NF-kappaB P65 (RELA) attenuated TNFα-induced expression of ICAM-1, VCAM-1 and E-selectin, which had a significant benefit in delaying the progression of inflammatory diseases, including atherosclerosis42. ARID1A, a subunit of the SWItch/Sucrose Non-Fermentable (SWI/SNF) chromatin remodeling complex, is localized to promoters and enhancers to influence transcription to affect the progression of atherosclerosis43.

Spliceosome and mRNA Splicing-Major Pathway are the most relevant signaling pathways in the second cluster, and this cluster consists of BUD13, RBM22, CDC5L, PRF8, PLRG1, and AQR. The main molecular events after ischemic stroke are translation-related, which may be caused by subsequent signaling pathway-related effects caused by upregulated cell membrane ribosomal and spliceosomal complex proteins44. BUD13, a subunit of the retention and splicing complex, may significantly increase the risk of ischemic stroke by affecting APOA1 leading to elevated TG and VLDL45,46. PLRG1 is a core component of the complex encoding the cell division cycle 5-like (CDC5L) complex. The CDC5L complex is part of the spliceosome, and both are highly conserved spliceosomal proteins across species. They are both shown to be part of the secondary spliceosomal protein complex and their encoded proteins play a critical role in alternative splice site selection. The interaction between CDC5L and PLRG1 is important for pre-mRNA splicing and essential for human pre-mRNA splicing47. Transcript variants of this gene with alternative splicing have been observed, encoding multiple isoforms48. CDC5L is significantly associated with the diagnosis of atrial fibrillation and ischemic stroke49. Knockdown of AQR in HepG2 facilitates glucose uptake, decreases PCK2 expression levels, increases GSK-3β phosphorylation, and restores insulin sensitivity50. The Innate immune genes are differentially expressed during either adaptation to hypoxia or recovery from H/R stress. Prolyl hydroxylase EGL-9, a known regulator of adaptation to hypoxia and innate immune response, inhibits the rapid recovery from H/R stress by activating AQR activity in O2-sensing neurons51. PPRF8 (pre-mRNA processing factor 8), a core component of the spliceosome, is an important mediator of hypoxia-induced mitosis52, and a high-fat diet upregulates the expression of genes related to the spliceosome53, increasing the risk of ischemic stroke. Adjusting the diet may enhance functional recovery after stroke by modulating insulin resistance and normalizing glucose metabolism54.

The third cluster contains six genes, SPRED2, PSMA4, PSMC4, PSMB3, SUFU, PSMD8, which are mainly involved in Degradation of GLI1, GLI2and GLI3 by the proteasome and Proteasome signaling pathways. Atherosclerosis is the primary pathological basis of ischemic stroke. In addition to its intensive inflammatory character, atherosclerosis also has features of enhanced proliferation, apoptosis, and oxidative stress. Besides playing an essential role in the degradation of dysfunctional and oxidatively damaged proteins, the ubiquitin–proteasome system (UPS), the major intracellular degradation system in eukaryotic cells, is involved in many processes that influence the progression of atherosclerotic disease55, which in turn affects the outcome of ischemic brain injury. SPRED2 is a negative regulator of the extracellular signal-regulated kinase (ERK) pathway and is important for cell proliferation, neuronal differentiation, plasticity, and survival. The decrease in SPRED-2 after injury may be related to the activation of the ERK pathway, and SPRED-2 signaling may play a role in the cell proliferation phase of neural repair in the zebrafish brain after injury56. SPRED2 affects neurological recovery after stroke by reducing cell migration through the ERK/c-Fos/MMPs pathway57. By binding to Gli, Sufu regulates the expression of target genes downstream of the Hedgehog (Hh) signaling pathway, and Sufu alternates between "open" and "closed" conformations, with the "closed" form of Sufu stabilized by Gli binding and inhibited by Hh processing. In contrast, the "open" state of Sufu is facilitated by the separation of Gli and Hh signaling58, which in turn affects neural regeneration in the brain. Down-regulation of three subunits of the proteasome (PSMA2, PSMA3, and PSMA4) genes to regulate cellular functions against oxidative stress, which has the effect of reducing atherosclerosis59. Specific proteasome inhibitor PR957 inhibited LMP7 and significantly attenuated histological damage to the cerebral white matter and the cognitive function in ischemic stroke by suppressing the inflammatory response60.

The fourth cluster is mainly related to thrombosis and coagulation signaling pathways, containing five genes PLAU, FGG, F2, FGB, and FGA. The three genes FGG, FGB, and FGA encode fibrinogen, which is a separate risk factor for ischemic stroke, like hypertension and diabetes mellitus61. In addition, single nucleotide polymorphisms (SNPs) in FGG and FGA mediate an increase in D-dimers62, which are significantly associated with stroke progression63. Fibrinogen (Fg) levels decrease in the colon of inflammatory bowel disease (IBD), which plays a crucial role in the pathogenesis of IBD by regulating vascular permeability (VP) through activation of AKT and subsequent microfilament depolymerization64. The level of intraplatelet PLAU is closely related to the production of intraplatelet plasma proteins and the secondary degradation of α-granulin, which affects fibrinolysis65. Moreover, PLAU may play a role in the treatment of ulcerative colitis66. Binding of recombinant PLAU (uPA) or endogenous uPA to uPAR induces membrane recruitment and activation of β1 integrins via low-density lipoprotein receptor-associated protein-1 (LRP1), causing activation of the Rho family small GTPase Rac1 and Rac1-induced axonal regeneration, which promotes neuronal growth during development and functional improvement after ischemic injury67,68. The F2 gene encodes thrombospondin (also known as coagulation factor II), which is proteolytically cleaved in multiple steps to form activated serine protease thrombin. The activated thrombin plays an essential role in thrombosis and hemostasis by converting fibrinogen to fibrin during clot formation, stimulating platelet aggregation, and animating other coagulation factors.

Interestingly, the fifth cluster also mainly involves the interleukin signaling pathway, and the interaction of FOS and CCL2 genes in this cluster may play a bridging role. FOS is a potential target gene for Crohn's disease (CD)69, where downregulation of FOS significantly enhances oxidative stress levels, accelerates neuronal apoptosis, and inhibits mitochondrial function, increasing the risk of ischemic stroke70. CCL2 is intimately associated with unstable atherosclerosis as upregulated based on inflammatory stimuli. Inhibition of CCL2 expression attenuates the polarization of pro-inflammatory microglia and inhibits the release of inflammatory cytokines such as TNF-α, IFN-γ, IL-1β, IL-6, IL-12, IL-17, and IL-23 while acting as a neuroprotective agent71,72. Overexpression of CCL2 is closely associated with colonic inflammation in inflammatory bowel disease73. IL-23 and IL-17 induce platelet-activating factor receptor (PAF-R) expression on activated T cells, which is invoked by PAF binding to PAF-R. PAF-R is co-expressed with IL-17 and similarly regulated with the Th17 markers IL-17A, IL-17F, IL-22, and RORC74. Through STAT3, IL-6 and IL-21 promote the expression of RORA and RORC genes encoding RORα and RORγt, respectively, RORγ(t) is an essential regulator of Th17 cell differentiation. These two RORs then drive the expression of IL-17A/F and IL-22. Inverse RORγ(t) agonists cause a rebalancing of the Th17/Treg ratio in favor of anti-inflammatory Tregs75. RORA is not only overexpressed in both Crohn's disease (CD) and ulcerative colitis (UC)76, but is also involved in cholesterol metabolism77. RORgammat( +) Treg cells in patients with inflammatory bowel disease are progressively enlarged and have pro-inflammatory properties78. Endothelial-to-mesenchymal transition (EndMT) plays a significant role in atherosclerosis, and Twist-179, a prime regulator of T helper cells adapted to chronic inflammatory metabolism, and its activation induces this process80. Many cytokines and chemokines are involved in the process of atherosclerotic plaque formation, either accelerating the progression of atherosclerotic plaque or slowing down this process. Plasma levels of IL-8/CXCL8 in patients with acute ischemic stroke tend to increase over time81. Serum CXCL8 levels are associated with infarct size and functional outcome in patients with ischemic stroke82. And IL10 causes IL-1beta and TNF-alpha to be down-regulated, which may effectively reduce inflammation in atherosclerotic plaques83. Chemokine CXCL12, on the other hand, exerts a protective role against atherosclerosis84. Furthermore, atherosclerosis is an inflammatory disease of the vessel wall characterized by the activation of the innate immune system with macrophages as the main actors and the adaptive immune system with Th1 as the primary factor. In advanced atherosclerosis, a defect in outflow cells drives disease progression, and the expansion of regulatory T cells (TREG) ameliorates the role of outflow cells. Treg cells secrete interleukin-13 (IL-13), which stimulates macrophage production of IL-10. Endocrine paracrine signaling of IL-10 induces Vav1 in macrophages and promotes phagocytosis by apoptotic cells85. In addition, Treg cells promote macrophage efflux and improve inflammatory conditions by enhancing the transcytotic signaling pathway of apoptotic cell internalization86.

Finally, the results of bioinformatics analysis of proteins significantly differentially expressed in four ischemic stroke patients and four healthy subjects suggested that the Complement cascade related signaling pathways, Innate Immune System, and hemostasis related signaling pathways were the most significantly associated signaling pathways in ischemic stroke. Although this is not entirely consistent with the signaling pathways associated with ischemic stroke and inflammatory bowel disease obtained from previous analyses, it also suggests that the immune system, signal transduction, and hemostasis-related pathways are key signaling pathways in ischemic stroke. There are two main limitations of this study. First, due to the lag in the inclusion of proteins and protein interactions in bioinformatics such as string, many newly discovered proteins are not available with corresponding matching data, and the results may be further updated with the addition of data from subsequent related studies. However, because of the universality of this study, which included more than one million people and involved multiple populations, the results are of good value. Second, we did not perform further protein validation, but since this study used the DIA-MS technique, which has high accuracy, so further validation was carried out in the next phase. Our study may help us understand the relationship between ischemic stroke and inflammatory bowel disease and, with the help of advances in proteomics such as DIA-MS, help further understand the molecules and cytokines involved in inflammatory signaling in the gut-brain axis, which is essential to our understanding of neuroimmune communication and help develop immunomodulatory therapies for ischemic stroke.

Conclusion

In conclusion, by performing bioinformatic analysis of two datasets downloaded from NHGRI-GWAS, including ischemic stroke and inflammatory bowel disease, we constructed a protein interaction network containing 589 genes, identified a core protein-related interaction network composed of 40 hub genes, and determined that cytokine and interleukin-related signaling pathways, spliceosome, ubiquitin–proteasome system (UPS), thrombotic and anticoagulant signaling pathways are closely associated with ischemic stroke and inflammatory bowel disease. And these findings could significantly improve our understanding of the potential molecularly linked events in IS and IBD. Furthermore, these candidate genes and pathways could serve as predictive and therapeutic targets for both states (Supplementary Information S1).