Systematic Analysis of CXC Chemokine–Vascular Endothelial Growth Factor A Network in Colonic Adenocarcinoma from the Perspective of Angiogenesis

Background Tumor angiogenesis plays a vital role in tumorigenesis, proliferation, and metastasis. Recently, vascular endothelial growth factor A (VEGFA) and CXC chemokines have been shown to play vital roles in angiogenesis. Exploring the expression level, gene regulatory network, prognostic value, and target prediction of the CXC chemokine-VEGFA network in colon adenocarcinoma (COAD) is crucial from the perspective of tumor angiogenesis. Methods In this study, we analyzed gene expression and regulation, prognostic value, target prediction, and immune infiltrates related to the CXC chemokine-VEGFA network in patients with COAD using multiple databases (cBioPortal, UALCAN, Human Protein Atlas, GeneMANIA, GEPIA, TIMER (version 2.0), TRRUST (version 2), LinkedOmics, and Metascape). Results Our results showed that CXCL1/2/3/5/6/8/11/16/17 and VEGFA were markedly overexpressed, while CXCL12/13/14 were underexpressed in patients with COAD. Moreover, genetic alterations in the CXC chemokine-VEGFA network found at varying rates in patients with COAD were as follows: CXCL1/2/17 (2.1%), CXCL3/16 (2.6%), CXCL5/14 (2.4%), CXCL6 (3%), CXCL8 (0.8%), CXCL11/13 (1.9%), CXCL12 (0.6%), and VEGFA (1.3%). Promoter methylation of CXCL1/2/3/11/13/17 was considerably lower in patients with COAD, whereas methylation of CXCL5/6/12/14 and VEGFA was considerably higher. Furthermore, CXCL9/10/11 and VEGFA expression was notably correlated with the pathological stages of COAD. In addition, patients with COAD with high CXCL8/11/14 or low VEGFA expression levels survived longer than patients with dissimilar expression levels. CXC chemokines and VEGFA form a complex regulatory network through coexpression, colocalization, and genetic interactions. Moreover, many transcription factor targets of the CXC chemokine-VEGFA network in patients with COAD were identified: RELA, NFKB1, ZFP36, XBP1, HDAC2, SP1, ATF4, EP300, BRCA1, ESR1, HIF1A, EGR1, STAT3, and JUN. We further identified the top three miRNAs involved in regulating each CXC chemokine within the network: miR-518C, miR-369-3P, and miR-448 regulated CXCL1; miR-518C, miR-218, and miR-493 regulated CXCL2; miR-448, miR-369-3P, and miR-221 regulated CXCL3; miR-423 regulated CXCL13; miR-378, miR-381, and miR-210 regulated CXCL14; miR-369-3P, miR-382, and miR-208 regulated CXCL17; miR-486 and miR-199A regulated VEGFA. Furthermore, the CXC chemokine-VEGFA network in patients with COAD was notably associated with immune infiltration. Conclusions This study revealed that the CXC chemokine-VEGFA network might act as a prognostic biomarker for patients with COAD. Moreover, our study provides new therapeutic targets for COAD, serving as a reference for further research in the future.


Background
Colon cancer is a common malignant tumor of the digestive tract. The incidence and mortality of colon adenocarcinoma (COAD) are the third highest of all cancer types [1]. Since the early diagnosis of COAD remains difficult, its mortality is increasing yearly [2]. Approximately 50% of COAD patients relapse or die within five years [3]. Although using bioactive materials in anticancer drugs improves their overall therapeutic effects [4,5], finding new biomarkers and therapeutic targets for early diagnosis remains the most critical initial step in the prevention and treatment of COAD.
Chemokines are a family of small heparin-binding proteins 8-10 kDa in size. Four subgroups exist within the chemokine family (CXC, CC, CX3C, and C). The CXC subgroup has been shown to play a crucial role in angiogenesis in physiological and pathological settings [6]. Recently, the role of CXCL in regulating tumor angiogenesis has attracted increasing interest [7]. Different members of the CXC chemokines subgroup can promote or inhibit angiogenesis, thus promoting or inhibiting tumor growth [8].
Multiple factors have been identified as regulators of angiogenesis. However, CXC chemokines are a unique family of cytokines that regulate angiogenesis in several ways [9]. Vascular endothelial growth factor A (VEGFA) is a vital factor that plays an essential role in tumor angiogenesis and development [10]. Sunitinib, a VEGFA inhibitor, has been used to treat advanced renal cell carcinoma. However, the side effects of sunitinib can be quite severe and include kidney and cardiovascular damage [11]. CXC chemokines and VEGFA are heavily regulated during tumor angiogenesis. CXCL12 can promote a malignant phenotype by promoting the clonal growth of colorectal cancer cells and regulating the expression of VEGF and ICAM-1 [12].
Multiple online databases were used to explore the expression level, gene regulation network, prognostic value, and regulation targets of the CXC chemokine-VEGFA network in patients with COAD from an angiogenic perspective in this study. In addition, we aimed to identify the relationship between CXC chemokine and VEGFA expression and the development and prognosis of COAD, as well as to provide new insights into targeted therapies for patients with COAD.
2.3. GEPIA. GEPIA (http://gepia.cancer-pku.cn/index.html) is an analysis tool that delivers RNA sequencing expression data from 9,736 cancerous and 8,587 noncancerous samples [18]. Gene (CXC chemokines and VEGFA), dataset (COAD), and threshold conditions (P value cutoff = 0.05) were set as screening criteria. The expression of CXC chemokines and VEGFA, as well as the pathological stage of COAD, was analyzed using a Student's t-test. The prognosis of patients with COAD was analyzed using the Kaplan-Meier curve [14][15][16]. Data were obtained on February 15, 2022.
2.4. cBioPortal Analysis. cBioPortal (http://cbioportal.org) is a free online database for visualizing, studying, and analyzing cancer genomic data [19]. The analysis of genetic alterations in the CXC chemokine-VEGFA network was conducted using cBioPortal in this study. Overall, 636 samples of COAD were analyzed. A z-score threshold of ±2.0 was used to calculate mRNA expression z-scores for all samples (log RNA Seq V2 RSEM). CXC chemokines and VEGFA were the chosen genes [14][15][16]. Data were obtained on February 15, 2022.
2.5. STRING Analysis. STRING (https://string-db.org/cgi/ input.pl) is a free online database that helps researchers analyze all publicly available sources of protein-protein interaction (PPI) data [20]. We created the PPI network interaction using STRING in this study. The screening criteria were set as follows: (1) confidence: 0.400 and (2) species: Homo sapiens [14,15]. Data were obtained on February 16, 2022.
2.6. GeneMANIA Analysis. GeneMANIA (http://www .genemania.org) is a free online database that creates PPI networks and analyzes gene function [21]. The interaction networks were built using this database to explore the roles of CXC chemokines and VEGFA [14][15][16]. Data were obtained on February 16, 2022.
2.7. Metascape Analysis. Metascape (https://metascape.org) is a free online gene function analysis tool that assists users in using current common bioinformatics analysis approaches to batch gene and protein analysis to predict function [22]. We conducted Gene Ontology (GO) function and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses of the CXC chemokine-VEGFA network in COAD using Metascape [14][15][16]. Data were obtained on February 17, 2022.
2.9. LinkedOmics Analysis. LinkedOmics (http://www .linkedomics.org/) is a free database that provides methods for analyzing and comparing cancer multiomics data [24]. 2 BioMed Research International The "LinkInterpreter" module of LinkedOmics was used to derive biological insights into miRNA target enrichment and transcription factor target enrichment of the CXC chemokine-VEGFA network. A minimum number of three genes (size), cancer type (COAD), a simulation of 500, gene (CXC chemokines and VEGFA), and target dataset (RNAseq) were chosen in this study [14][15][16]. Data were obtained on February 18, 2022.
2.10. TIMER Analysis. TIMER (https://cistrome.shinyapps .io/timer/) is a free online platform for systematically analyzing tumor-infiltrating immune cells [25]. The "Gene module" of TIMER was used to assess the correlation between the expression level of the CXC chemokine-VEGFA network and tumor-infiltrating immune cells [14][15][16]. Data were obtained on February 18, 2022. ). In addition, immunohistochemical results validated the differential expression of the CXC chemokine-VEGFA network between patients with COAD and those without COAD ( Figure 2). The pathological stage of COAD and the differential expression of the CXC chemokine-VEGFA network were assessed in this study. The pathological stage in patients with COAD and the expression of CXCL9/10/11 and VEGFA were found to have a significant correlation (P < 0:05) ( Figure 3). Subsequently, the prognostic ability of the CXC chemokine-VEGFA network expression in COAD patients was evaluated. The overall survival was longer in COAD patients when levels of CXCL8/11/14 expression were higher (P ≤ 0:05) (Figures 4(a)-4(c)) or when levels of VEGFA expression were lower (P < 0:05) (Figure 4(d)).
3.3. CXC Chemokines and VEGFA Interaction Network. The potential interactions between CXC chemokines and VEGFA in patients with COAD were explored. Overall, 13 nodes and 68 edges were obtained in the PPI network using STRING software (Figure 7(a)). The average node degree and local clustering coefficient of the PPI network were 10.5 and 0.908, respectively. Furthermore, the CXC chemokine-VEGFA network (33 genes and 2,152 edges) was linked to a complex interaction network through shared protein domains, coexpression, predicted, colocalization, and genetic interactions using GeneMANIA (Figure 7(b)). Moreover, cell chemotaxis, chemokine and cytokine receptor binding, chemokine and cytokine activity, leukocyte chemotaxis, and migration were the major functions of the CXC chemokine-VEGFA network in COAD patients ( Figure 7(b)). In brief, CXC chemokines were connected to and interacted with VEGFA in a complex network.

GO and KEGG Pathway Enrichment Analyses.
Metascape was utilized to analyze the functions of the CXC chemokine-VEGFA network in patients with COAD. We found that the biological processes connected with CXC chemokines and VEGFA were mainly related to leukocyte chemotaxis, myeloid leukocyte migration, positive regulation of leukocyte chemotaxis, lymphocyte migration, and regulation of multiorganism processes (Figure 8(a)). Moreover, chemokine and cytokine activity, heparin binding, and growth factor activity were the main molecular functions of CXC chemokine-VEGFA network expression ( Figure 8(b)). The KEGG pathway of the CXC chemokine-VEGFA network in COAD was mainly involved in cytokine-cytokine receptor interaction, rheumatoid arthritis, interleukin-(IL-) 17 signaling pathway, and nuclear factor kappa B (NF-κB) signaling pathway (Figure 8(c)).

Discussion
Tumor angiogenesis plays a vital role in tumorigenesis, proliferation, and metastasis. In recent years, studies have identified VEGFA and CXC chemokines as important participants in angiogenesis, particularly tumor angiogenesis [14,15,[26][27][28]. The expression levels of CXC chemokines and VEGFA have been studied in a range of tumor types; however, findings are contradictory with regard to colonic adenocarcinomas [29,30]. This study investigated expression level, gene regulatory network, prognostic value, and target prediction of the CXC chemokine-VEGFA network for COAD from a tumor angiogenesis perspective.
In this study, we also examined the potential correlation between pathological stage and differential expression of COAD. The expression of CXCL1/2/3/5/6/8/11/16/17 and VEGFA was upregulated in patients with COAD compared with that in individuals without COAD. Patients with COAD also showed downregulated CXCL12/13/14 expression. The results were similar to those reported in a previous study in patients with COAD [30] and contradicted those    BioMed Research International reported previously in patients with colorectal cancer [29]. This may be due to the small sample size and the variable types of colorectal cancer. We further attempted to explain the different expression levels by investigating promoter methylation and gene alteration in patients with COAD, as these factors affect tumor cell proliferation, angiogenesis, and metastasis. We observed that patients with COAD had different rates of genetic alteration in their genes. Moreover, the promoter methylation levels of CXCL5/6/12/14 and VEGFA were higher in patients with COAD than those in healthy individuals. Conversely, the promoter methylation levels of CXCL1/2/3/11/13/17 were lower in patients with COAD. Thus, we hypothesized that genetic methylation and alteration within the CXC chemokine-VEGFA network might be the leading cause of abnormal gene expression levels in patients with COAD.
We also observed a notable correlation between the CXCL9/10/11 and VEGFA expression and the pathological stage of COAD. Furthermore, the survival of patients with COAD was higher with low VEGFA or high CXCL8/11/14 expression levels. Therefore, the expression levels of CXCL8/11/14 and VEGFA may be potential prognostic indicators for COAD. CXCL8/11/14 and VEGFA promote tumor angiogenesis in different ways [31][32][33]. Thus, they may affect the prognosis of patients with COAD through multiple biological functions.
The potential functions and interactions of the CXC chemokine-VEGFA network were further explored in this study. They were found to be complex and tightly connected. CXC chemokines and VEGFA may promote cancer progression, and this could be through a potential interaction network. Genes in the network were mainly involved in cytokine receptor binding, chemokine and cytokine activity, leukocyte chemotaxis, and migration, all of which are closely related to angiogenesis. For instance, IL-8 (CXCL8) promotes tumor angiogenesis by binding to CXCR1 and CXCR2 receptors [34]. In addition, increasing the antitumor activity of cytokine-induced killer cells could reduce tumor proliferation and angiogenesis [35]. Collectively, these results suggest that the CXC chemokine-VEGFA network may influence the development of COAD by increasing tumor angiogenesis.
Furthermore, the functions of the CXC chemokine-VEGFA network in patients with COAD were mainly related to chemokine activity, cytokine activity, and growth factor activity, as demonstrated by GO enrichment analysis, all of which are closely related to tumor angiogenesis. More studies are needed to confirm the mechanism by which this happens. In this study, we further found through KEGG pathway analysis that the cytokine-cytokine receptor interaction signaling pathway, IL-17 signaling pathway, and NF-κB signaling pathway were highly involved in the CXC chemokine-VEGFA network in COAD patients, all of which are highly related to tumor angiogenesis [36,37]. Therefore, the respective regulation of these pathways may serve as a potential treatment strategy for patients with COAD.
The correlation between CXC chemokine-VEGFA network expression and differentially expressed genes in COAD        (l) Figure 11: Continued. 16 BioMed Research International patients was explored in this study. We found that in patients with COAD, approximately 20,000 genes were negatively or positively correlated with CXC chemokine-VEGFA network expression. From these, we screened for genes with the highest correlation with CXC chemokines and VEGFA. Some of the genes with the highest correlation (ZC3H12A, IL24, MMP3, IL1B, OSM, IDO1, NPR1, and TIGIT) were positively associated with tumor angiogenesis [53,54]. Regulation of these cancer-related genes may offer an alternative therapeutic strategy for the treatment of patients with COAD. Immune infiltration is highly related to the clinical prognosis of tumors. Immune cells reach the tumor site through vascular transport, and vascularization of tumors is a process mediated by angiogenesis. We observed that CXC chemokine-VEGFA network expression, which regulates angiogenesis, is correlated with the infiltration of immune cells. This infiltration involved CD4+ T cells, CD8+ T cells, neutrophils, macrophages, and dendritic cells. Improving immune cell infiltration in COAD by developing drugs that act on the CXC chemokine-VEGFA network or CXC chemokines and VEGFA-related regulatory targets may serve as a viable therapeutic oncology approach.

Conclusions
In this study, we determined the expression levels and gene regulatory network of the CXC chemokine-VEGFA network, which plays a vital role in angiogenesis in COAD. We also identified new prognostic biomarkers and therapeutic targets. These findings provide insight into the study and treatment of COAD.