Effect of interleukin-8 receptor B (IL8RB) rs1126579 C>T variation on the risk to cancer

ABSTRACT Chemokines are a type of cytokine that participate in the migration of macrophages and monocytes to inflammatory cells. In particular, CXC chemokines are involved in the development of many cancers. Evidence for the association between interleukin-8 receptor B (IL8RB) rs1126579 C > T variation and cancer risk remains contradictory. Here, we utilized a comprehensive analysis containing odds ratios (ORs), regression, and in silico tools to evaluate the effect of IL8RB polymorphism on cancer risk. We further employed Gene set enrichment analysis combined with ELISA to evaluate the IL8RB expression in patients with prostate cancer (PRAD). A total of 5,187 cancer cases and 6,691 controls were included in the present analysis. Individuals with the TT genotype were associated with an increased risk of cancer compared to those with the TC+CC genotype. In a subgroup analysis by type of cancer, individuals with the TT genotype had a 39% increased risk of urinary cancer compared to those with the CC genotype. A subgroup analysis by ethnicity showed that Asians carrying the TC genotype had a 26% lower risk of cancer than those carrying the CC genotype. We found that the expression of IL8RB was down-regulated in PRAD. Compared to that in PRAD subjects carrying the CC genotype, the expression of IL8RB was decreased in patients with the TT+TC genotype. In conclusion, the IL8RB rs1126579 C > T variation may be associated with cancer risk, especially in Asian populations and patients with PRAD.


Introduction
Cancer remains a major threat to human health worldwide [1]. In Western countries, the incidence and mortality of malignant tumors are growing, and the situation is similar in developing countries [2]. In 2020, the United States was expected to have 1,806,590 new cancer patients and 606,520 cancer-related deaths [3]. In Australia, the annual incidence of cancer was 31.7 per 100,000 people and the mortality rate was 4.1 in 100,000 from 2000 to 2009 [4]. The main reasons for the low survival rate of most malignant tumors are late diagnosis of the tumor at an advanced stage, metastasis, and resistance to treatment [5]. At present, specific markers for accurate diagnosis of many malignant tumors have not been identified. Therefore, it is necessary to develop some specific molecular markers to predict the prognosis of cancer patients and to provide effective targets for the treatment of these patients [6].
Chemokines belong to a large class of small protein molecules that play a vital role in many cellular activities, including cell recruitment and migration [7,8]. Previous researchers shown evidence that CXC chemokines and receptors can participate in a variety of functional activities including embryogenesis, angiogenesis, migration of leukocytes, and metastasis of malignancies [9][10][11]. Interleukin-8 (IL-8, CXCL8) is a small molecular basic protein that belongs to the ELR+-CXC chemokine subgroup. CXC chemokine receptor-2 (CXCR2) is a crucial IL-8 receptor and mediates angiogenesis induced by ELR+-CXC chemokines [12]. CXCR2, encoded by the IL8RB gene, is a seven-transmembrane G protein-coupled receptor (GPCR) that exists on the cell membrane of endothelial and cancer cells. Previous research revealed that the biological effect of IL-8 is mediated through the binding of CXCR1 or CXCR2 to IL-8 [13]. Subsequent studies demonstrated that CXCR2, but not CXCR1, is the crucial functional chemokine receptor that participates in angiogenesis induced by the chemokine and chemotaxis activity of endothelial cells [14][15][16]. The binding of CXCR2 and IL-8 can promote a series of tumor cell activities including proliferation, angiogenesis, and invasion [17,18]. The high affinity of CXCR2 to chemokines has been shown to be associated with the prognosis of patients with many cancers, including glioblastoma, colon cancer, lung cancer, hepatocellular carcinoma, and pancreatic cancer [19][20][21][22][23].
Genetic polymorphisms of IL8RB may directly influence the development of malignant tumors by inducing tumor angiogenesis and the immune response pathway [24,25]. Some studies revealed that expression of IL8RB is a potential adverse prognostic marker for individuals with nonmetastatic renal clear cell carcinoma after nephrectomy [26]. However, several studies in other tumors failed to demonstrate a positive correlation between IL8RB and the recurrence and survival of patients [27,28]. The IL8RB rs1126579 C > T variation has been evaluated in rectal, prostate, stomach, bladder, esophageal, breast, colon, and lung cancer. Nevertheless, the relationship between this polymorphism and cancer risk remains incomprehensive. The aim of the current study was to comprehensively assess the association between IL8RB rs1126579 C > T variation and cancer risk based on all eligible case-control studies [29][30][31][32][33][34][35][36][37][38]. In addition, we used in silico analysis to evaluate the expression of IL8RB in prostate, bladder, breast, and lung cancer. We also investigated the correlation between the expression of IL8RB and the N stages of these patients. Moreover, we employed Gene Set Enrichment Analysis (GSEA) to investigate the expression of IL8RB in prostate cancer (PRAD), and ELISA to verify the findings in patients recruited from our centers.

Search strategy
A database search was conducted according to the Embase, National Center for Biotechnology Information (NCBI) database, Chinese Wanfang, and Google Scholar. The following keywords were used: ('rs1126579 OR 'interleukin-8 receptor B' OR 'IL8RB') AND ('cancer' OR 'tumor' OR 'carcinoma') AND ('mutation' OR 'variation' OR 'Single Nucleotide Polymorphism' OR 'SNP' OR 'mutant'). The last search update was 1 March 2021. We also searched the references and supplementary information of published articles to elevate the number of included studies.

Inclusion and exclusion criteria
An appropriate study could be included in the current analysis if it met the following criteria: (a) case-control studies on the association between IL8RB rs1126579 C > T variation and susceptibility to cancer; (b) containing enough genotype data to measure odds ratios (ORs); and (c) manuscripts with available full text. The exclusion criteria were: (a) no available data in control; (b) insufficient information to assess the ORs; and/or (c) not related to IL8RB rs1126579 C > T variation and cancer risk.

Data extraction
Data was classified according to the following characteristics: name of the author, year of publication, origin of patients, type of cancer, ethnicity of population, source of control, genetic data regarding IL8RB rs1126579 C > T variation, P value of Hardy-Weinberg equilibrium (HWE) in control, and method of genotyping. Cancers of different systems were divided into separated subgroups. The urinary cancer subgroup included prostate and bladder cancer. The digestive system cancer subgroup involved gastric, esophageal, colon, and rectal cancer. One study addressed to Kaposi's sarcoma, and was defined as 'other cancer'.

Statistical analyses
Odds ratios (ORs) and 95% confidence intervals (CIs) were used to evaluate the strength of the correlation between IL8RB rs1126579 C > T variation and susceptibility to cancer. We employed five genetic models to evaluate the overall ORs of the rs1126579 C > T variant: allelic comparison (T allele vs. C allele), heterozygous (TC vs. CC), homozygous (TT vs. CC), dominant genetic model (TT + TC vs. CC), and recessive comparison (TT vs. TC + CC). The heterogeneity of the included studies was measured using a Q statistic test. If the P-value of heterogeneity (P heterogeneity ) was < 0.05, a random effects calculation was adopted (DerSimonian and Laird) [39]. Conversely, a fixed effects method was selected (Mantel-Haenszel) if the P heterogeneity was > 0.05 [40]. The P value of HWE (P HWE ) was evaluated using Fisher's exact test. Studies with a P HWE > 0.05 were defined as high-quality groups. Otherwise, studies were defined as low-quality groups. If a study had a sample size greater than 1000, it was classified as a large sample group. Subgroup analysis contained the type of cancer, source of control, ethnicity, sample size, and quality of studies. Sensitivity analysis of IL8RB rs1126579 C > T variation was conducted by excluding every single study in turn. Publication bias was measured using Begg's and Egger's tests. P > 0.05 indicates that there is no evidence of publication bias among studies. Additionally, we used regression analysis to estimate the functional relationship of the log OR with the study characteristics. All the statistical analyses were performed using STATA software (v11.0, Stata Company, College Station, TX, USA).

In silico and ELISA analysis
The minor allele frequencies (MAFs) in various populations were assessed using the Single Nucleotide Polymorphism database (dbSNP) of the NCBI repository (https://www.ncbi.nlm.nih. gov/snp). The gene expression profiles of IL8RB were investigated using the Gene Expression Profiling Interactive Analysis (GEPIA, http:// gepia.cancer-pku.cn/index.html) and TNMplot databases (https://www.tnmplot.com/). Gene-gene interaction and expression of IL8RB in different populations were assessed using the Ualcan database (http://ualcan.path.uab.edu/analysis.html). The protein-protein correlation of CXCR2 was measured using the STRING online server (https://string-db.org/cgi/input.pl). The GSEA of the transcriptomes in the PRAD samples was determined using GSEA software (version 4.1.0, http://software.broadinstitute.org/gsea/index.jsp), a joint project produced by UC San Diego and the Broad Institute [41]. The immune pattern and landscape distribution between the high and low expression subtypes were investigated using the CIBERSORT computational method. We adopted the CIBERSORT algorithm to measure the proportions of tumor-infiltrating immune cells (TICs) in the PRAD samples [42]. Samples with a P-value less than 0.05 were chosen for follow-up analyses. ELISA analysis, according to the manufacturer's instructions (CUSABIO Co., Ltd), was conducted based on samples from PRAD volunteers recruited in our centers [43,44]. A total of 220 needle biopsy-confirmed PRAD patients (by) were included from the Affiliated Changzhou No.2 People's Hospital of Nanjing Medical University and the Affiliated Hospital of Jiangnan University. Every enrolled patient provided 2 mL of peripheral blood after signing the informed consent form. We then used ELISA to detect the serum expression of IL8RB. The above study was approved by the Ethics Committee of our hospitals.

Results
In the current study, we used ORs, 95 CIs, and regression analysis to comprehensively assess the association between IL8RB rs1126579 C > T variation and cancer risk based on all eligible casecontrol studies. Moreover, we used in silico analysis, GSEA, and ELISA to explore the expression of IL8RB.

Characteristics of studies
In total, 13 case-control studies with 5,187 cases and 6,691 controls were involved in the current analysis (Table 1). In the stratified analysis by cancer type, four studies were based on digestive cancer, and three studies each were based on urinary system tumors and lung cancer, respectively. Two studies focused on breast cancer, and one study was on Kaposi's sarcoma and was classified as 'other cancer'. There were six separate studies on Caucasians and Asians in the stratified analysis by ethnicity. An additional study was on an African population. In the subgroup analysis by control source, nine population-based (PB) and four hospital-based (HB) studies were included. The stratification analysis by genotyping method revealed seven studies utilizing the polymerase chain reaction (PCR) method, whereas the rest of the studies used either the Golden Gate method, Taqman assay, or iPLEX Gold method. In the  Figure 1).

Main results
Odds ratios (ORs) and 95% Cls were employed to assess the relationship between IL8RB rs1126579 C > T polymorphism and cancer risk. Compared to those with the TC+CC genotype, individuals carrying the TT genotype were associated with an elevated risk of cancer (OR = 1.15, 95% CI = 1.05-1.25, P = 0.003, Table 2). In the subgroup analysis by cancer type, individuals with the TT genotype had a 39% increased risk of urinary cancer compared to those with the CC genotype (95% CI = 1.06-1.83, P = 0.018, Figure 2(a)). In the homozygous comparison, the IL8RB rs1126579 C > T polymorphism was also associated with increased breast and digestive  Figure 3(a)). In the stratified analysis by control source, the TT genotype was associated with enhanced risk of cancer in PB studies using the recessive model (95% CI = 1.03-1.24, P = 0.010, Figure 3(b)). The stratification analysis by quality of study revealed that IL8RB rs1126579  For the large-sample size studies, patients carrying the T-allele had a 16% augmented risk compared to those with the C-allele (95% CI = 1.01-1.33, P = 0.033).

In silico and ELISA analysis
The TNMplot database was used to evaluate the expression of IL8RB in PRAD and bladder cancer (BLCA) patients. The expression of IL8RB was down-regulated in both the PRAD (Figure 4(a), p < 0.05) and BLCA (Figure 4(b), p < 0.05) subjects. Effect of IL8RB expression on PRAD patients' disease-free survival (DFS) and overall survival (OS) time was shown in Figure 4(c,e). In the first 50 months, PRAD patients with low IL8RB expression had a shorter DFS time than that in the high expression groups. The expression of IL8RB on BLCA participants' DFS and OS time was described in Figure 4(d,e). No significant  difference on the DFS and OS time was indicated between the low and high expression groups among BLCA subjects. We further used the Ualcan database to verify the expression of IL8RB in various types of cancer. Expression of IL8RB was down-regulated in both PRAD and BLCA ( Figure 5(a,b), p < 0.05). Expression of IL8RB was also down-regulated in lung and breast cancer patients ( Figure 5(c,d), p < 0.05). Furthermore, ELISA was employed to assess the serum expression of IL8RB in PRAD patients recruited from our hospitals. Compared with that in PRAD subjects carrying the CC genotype, the expression was decreased in patients with the TT + TC genotype ( Figure 6, P < 0.05). Moreover, we investigated the correlation between IL8RB expression and the N stages of cancer patients. For PRAD, expression was decreased in both N0 and N1 patients compared to that in their normal counterparts (P < 0.05, Figure 7(a)). For BLCA, the expression of IL8RB was only down-regulated in N3 patients (P < 0.05, Figure 7(b)). For lung cancer, the expression was attenuated in patients with N0, N1, and N2 stage cancer (P < 0.05, Figure 7 (c)). For breast cancer, the expression of IL8RB was also diminished in patients with N0, N1, and N2 stage cancer (P < 0.05, Figure 7(d)). Moreover, we used R language to investigate the gene-gene correlations of IL8RB. More than 24 genes were involved in interactions with the IL8RB gene in PRAD (Figure 8(a)). The most correlated genes included VSTM2L (V-set and transmembrane domain containing 2 like gene, P < 0.05, R = −0.19, Figure 8(b)), CRISP3 (cysteine rich secretory protein 3 gene, P < 0.05, R = −0.08, Figure 8(c)), and DLX1 (distal-less homeobox 1 gene, P < 0.05, R = −0.29, Figure 8(d)). We further employed the STRING database to investigate the protein-protein correlation of CXCR2. More than 20 proteins could interact with the CXCR2 protein (Figure 9(a)). The correlations of the top ten proteins were described in Figure 9(b). In addition, we performed GSEA analysis to investigate the potential associated signaling pathways correlated with expression of IL8RB. As described in Figure 10(a), signaling pathways including aldosterone regulated sodium reabsorption, extracellular matrix (ECM) receptor, focal adhesion, and regulation of actin cytoskeleton were associated with high expression of IL8RB. Meanwhile, signaling pathways such as glyoxylate and dicarboxylate metabolism, oxidative phosphorylation, and pyrimidine metabolism were correlated with low expression of IL8RB (Figure 10(b)). Furthermore, we used the CIBERSORT computational method to evaluate the abundance profile of TICs in PRAD samples (Figure 11(a,b)). Compared with that in the low IL8RB expression group, the proportion of T regulatory cells was significantly attenuated in the high expression group (Figure 11 (c)). Moreover, the proportion of monocytes was relatively augmented in the high expression group (Figure 11(d)).

Sensitivity, regression analysis, and publication bias
Sensitivity analysis was employed to reveal the effect of a single study on the overall ORs. Publication bias was assessed using Begg's and Egger's tests. As shown in Figure 12(a), no single study was found to have a significant impact on the ORs when assessing the IL8RB rs1126579 C > T variation. Begg's ( Figure 12(b), P > 0.05) and Egger's tests (Figure 12 (c), P > 0.05) also identified no evidence of publication bias in the studies of IL8RB polymorphism. Moreover, we adopted regression analysis to determine whether there was a significant correlation between the studied ethnicity, genotyping method, sample size, and combined OR. No evidence of heterogeneity among the studies was revealed (Figure 13(a-d)).

Discussion
Cancer is a huge health problem worldwide. Although most cancer patients receive standard treatment including operation, radiotherapy, chemotherapy, or immunotherapy, not everyone can benefit from these strategies. Previous studies have shown evidence that expression of IL8RB is related to necrosis and development of several cancers [23,45,46]. Furthermore, expression of IL8RB can act as an autocrine or paracrine growth factor in the invasion and migration of cancer [47]. Genetic variants of IL8RB may affect the function of the protein by influencing gene expression. The correlation between IL8RB variations and cancer risk has been evaluated in previous studies [32][33][34][35][36][37][38]. However, the conclusions have been contradictory. Singh et al. assessed the IL8RB rs1126579 C > T variation in Indians and revealed that the T allele was associated with an increased risk of BLCA (P = 0.003, OR = 1.29) [36]. Another researcher assessed the IL8RB variant in two races (Caucasian and Asian) and observed that rs1126579 C > T variation was associated with a decreased risk of lung cancer [37]. A meta-analysis published in 2017 showed that IL8RB expression in cancer was related to poor prognosis of patients [47]. One year later, another meta-analysis demonstrated that IL8RB expression is a poor predictor for digestive cancer patients [48]. However, the above studies did not assess the correlation between IL8RB rs1126579 C > T variation and susceptibility to cancer. Therefore, we performed a comprehensive analysis based on 5,187 cancer cases and 6,691 controls from 13 case-control studies. We observed a positive association between IL8RB rs1126579 C > T mutation and cancer risk.
In the subgroup analysis by cancer type, we revealed that IL8RB rs1126579 C > T variation was correlated with an elevated risk of urinary and breast cancer, as well as cancer in the digestive system. For lung cancer, individuals with the TT genotype had a 30% decreased risk compared to those with the CC genotype. Our results are in line with those reported in previous studies [37]. In the stratification analysis by ethnicity, Asian individuals carrying the TC genotype had a 26% lower risk of cancer than those carrying the CC genotype. However, we observed no positive results in the Caucasian or African participants. The possible reason may be that the sample size of studies during the subgroup analysis was relatively small. However, there were some studies indicating that IL8RB rs1126579 C > T variation was associated with an elevated risk of breast cancer in African populations [34]. As described in the stratified analysis by sample size, the results of studies with a large sample size may be different from those with a small sample size. Therefore, further research with large sample sizes on IL8RB rs1126579 C > T polymorphism in African descendants is required in the future. Furthermore, in silico analysis was utilized to explore the expression of IL8RB in urinary cancer based on the race Figure 7. The correlation between the IL8RB expression and the N stages of cancer patients. For PRAD, the expression was decreased in both N0 and N1 patients than that in normal counterparts (P < 0.05, Figure A). For BLCA, the expression of IL8RB was only downregulated in N3 patients (P < 0.05, Figure B). For lung cancer, the expression was attenuated in patients with N0, N1, and N2 stage (P < 0.05, Figure C). For breast cancer, the expression of IL8RB was also diminished in patients with N0, N1, and N2 stage (P < 0.05, Figure D).  of patients. Expression of IL8RB was diminished in BLCA patients of Caucasian, African-American, and Asian descent. The expression was also mitigated in Caucasian and African-American PRAD patients. To verify the reliability of the results obtained from the online database, we used ELISA to detect the serum IL8RB expression in the pathologically confirmed PRAD patients recruited from our centers. It showed that IL8RB expression was attenuated in PRAD patients with the TT+TC genotype, which was consistent with the results of the present analysis. In addition, expression of IL8RB was down-regulated in several cancers including PRAD, BLCA, lung cancer, and breast cancer. For PRAD, the expression was decreased in both N0 and N1 patients. For BLCA, the expression of IL8RB was only downregulated in N3 patients. For lung cancer, the expression was attenuated in patients with N0, N1, and N2 stage cancer. For breast cancer, the expression of IL8RB was also diminished in patients with N0, N1, and N2 stage cancer. Previous study has adopted TCGA database to explore prognostic factors for testicular germ cell tumors [49]. In addition, researchers also used this database to identify a genomic lncRNA signature to provide guidance for the treatment of patients with BLCA [50]. In the present study, we employed the TCGA database to investigate signaling pathways associated with expression of IL8RB. Several signaling pathways, such as ECM receptor, focal adhesion, regulation of actin cytoskeleton, and aldosterone regulated sodium reabsorption, were associated with high IL8RB expression. Moreover, we used the CIBERSORT method to investigate the TIC abundance in PRAD samples between the high and low IL8RB expression groups. Compared with that in the low IL8RB expression group, the proportion of T regulatory cells was significantly attenuated in the high expression group. Meanwhile, the proportion of monocytes was relatively augmented in the high expression group. Besides, there are several limitations in the above analysis. First, the number of studies on African populations is fairly small. More studies of these populations with large sample sizes are required. Second, the sample size of case-control studies on IL8RB rs1126579 C > T variation remains insufficient. The number of studies for the subgroup analysis of cancer types was also insufficient, especially for PRAD, BLCA, and renal cell carcinoma. Third, we revealed that IL8RB rs1126579 C > T polymorphism may be related to an elevated risk of PRAD. Further research is still needed to ascertain whether this variant can affect the expression of IL8RB in Figure 11. The relationship between the expression of IL8RB and the proportion of tumor-infiltrating immune cells (TICs). Violin plot displayed the difference of 22 kinds of immune cells with low or high expression of IL8RB to the median level of IL8RB expression in PRAD ( Figure A and B). Compared with low IL8RB expression group, the proportion of T regulatory cells was significantly attenuated in high expression group (Fig. C). Meanwhile, the proportion of monocytes was relatively augmented in high expression group (Fig. D).
PRAD. Since a single mutation cannot have a great impact on the occurrence and development of cancer, future studies on gene-gene or geneenvironment interactions are still warranted.

Conclusion
Taken together, the current study summarizes all eligible genetic data for association between IL8RB  . Regression analysis of log odds ratio versus subgroup analysis of study ethnicity, genotyping method, sample size. No evidence of heterogeneity among the studies was revealed (Fig. A, B, C, and D). rs1126579 C > T variation and cancer risk. Our study revealed that the IL8RB rs1126579 C > T polymorphism is associated with an increased risk of urinary, breast, and digestive cancer, especially in individuals of Asian descent. IL8RB rs1126579 C > T variation may also be correlated with the risk of PRAD.

Highlights
(1) We used odds ratios, regression, and in silico analysis to assess the effect of IL8RB variation on cancer risk. (2) The current analysis showed that IL8RB rs1126579 C>T variation may be associated with risk of cancer. (3) Expression of IL8RB is decreased in PRAD patients. (4) GSEA showed that signaling pathways including ECM receptor and focal adhesion were associated with high expression of IL8RB.