Impacts of Chemokine (C-X-C Motif) Receptor 2 C1208T Polymorphism on Cancer Susceptibility

Background The CXC chemokines belong to a unique family of cytokines that participates in the progression and development of many malignant tumors. Evidence for the relationship between chemokine (C-X-C motif) receptor 2 (CXCR2) C1208T polymorphism and susceptibility to cancer remains inconsistent. Methods Odds ratios (ORs), 95% confidence intervals (CIs), and combined analysis were used to investigate the effect of CXCR2 variation on cancer risk. Gene Set Enrichment Analysis (GSEA) and enzyme-linked immunosorbent assay (ELISA) were also used to evaluate the expression of CXCR2 in prostate cancer (PCA). Results Across 11 case-control studies, 4,909 cases and 5,884 controls were involved in the current analysis. Individuals with a TT genotype were associated with increased risk of digestive cancer, compared to those with a TC+CC genotype (OR = 1.16, 95%CI = 1.02-1.31, P = 0.025). Individuals carrying the TT genotype had a 39% higher risk of urinary cancer than those carrying CC genotype (OR = 1.39, 95%CI = 1.04-1.87, P = 0.025). Individuals with a TT genotype showed a 56% augmented breast cancer risk, compared to those with a CC genotype (OR = 1.56, 95%CI = 1.03-2.35, P = 0.034). It was found that CXCR2 expression was downregulated in PCA. Compared with PCA subjects carrying the CC genotype, the expression of CXCR2 was decreased in patients with the TT genotype. Conclusions The CXCR2 C1208T variation was associated with elevated risk of urinary, breast, and digestive cancer. However, the C1208T polymorphism was correlated with attenuated risk of lung cancer.


Introduction
Cancer continues to be a major public health problem all over the world [1]. The incidence and mortality of cancer are increasing in both developing and developed countries [2]. Previous studies have shown that there were approximately 14.1 million new cancer subjects and 8.2 million deaths worldwide in 2012 [3]. In China, the number of new cancer patients was more than 4.3 million in 2018 [4]. Late diagnosis, metastasis, and drug resistance lead to a poor survival rate for most malignant tumors [5]. Thus far, specific tumor biomarkers have not been developed for many solid tumors. Therefore, it is necessary to investigate universal molecular markers to predict prognosis and provide targets for cancer patient treatment [6].
Chemokines are a large class of structurally related small protein molecules that play an important role in cell recruitment and migration [7,8]. Previous studies have shown that CXC chemokines and their receptors are involved in leukocyte migration, angiogenesis, embryogenesis, tumor growth, and metastasis [9][10][11]. CXC chemokine receptor-2 (CXCR2) is one of the main receptors in the CXC superfamily [12]. It is a seven-transmembrane G protein-coupled receptor (GPCR) often expressed on the cell membrane of endothelial cells, leukocytes, and tumor cells. Previous research indicates that binding of interleukin-8 (IL-8) to CXCR1 or CXCR2 on  Journal of Immunology Research the cell surface mediates the biological effect of IL-8 [13]. Subsequent studies confirmed that CXCR2, rather than CXCR1, is the main functional chemokine receptor mediating chemokine-induced angiogenesis and endothelial cell chemotaxis [14][15][16]. The binding of CXCR2 and chemokine induces proliferation, angiogenesis, and invasion of tumor cells [17,18]. CXCR2 has a high affinity for chemokines and is thought to be associated with the prognosis of several cancers, including colon cancer, hepatocellular carcinoma, and pancreatic cancer [19][20][21][22][23].
Previous studies have shown that genetic variants of CXCR2 may affect the development of cancer by regulating tumor angiogenesis and the antitumor immune response pathway [24,25]. An et al. revealed that CXCR2 was associated with poor prognosis in patients with nonmetastatic renal clear cell carcinoma and that CXCR2 can be used as a new prognostic factor [26]. However, several other studies demonstrated no positive correlation of CXCR2 with the outcomes of pancreatic carcinoma or esophageal cancer [27,28]. The CXCR2 C1208T polymorphism has been investigated in several cancers, including those of the prostate, bladder, breast, colon, and lung. However, the relationship between this variant and susceptibility to cancer remains controversial. The purpose of the present study was to comprehensively investigate the association between CXCR2 C1208T polymorphism and cancer risk combined with all eligible case-control studies [29][30][31][32][33][34][35][36]. Furthermore, we used combined analysis to explore the effect of CXCR2 variation. Additionally, Gene Set Enrichment Analysis (GSEA) and enzyme-linked immunosorbent assay (ELISA) were performed to evaluate the expression of CXCR2 in prostate cancer (PCA).

Search Strategy.
We conducted an online database search based on Embase, The National Library of Medicine (NLM), Chinese Wanfang, and Google Scholar. The keywords were ('+1235C/T' OR 'C1208T' OR 'CXCR2' OR 'Chemokine C-X-C motif receptor 2') AND ('cancer' OR 'carcinoma') AND ('polymorphism' OR 'variant' OR 'mutant'). The last search update was on May 31, 2021. To increase the number of selected publications, we also retrieved studies by examining the references in published articles.

Inclusion and Exclusion Criteria.
A suitable study was included in our analysis when it met all of the following criteria: (a) case-control studies focused on the relationship between the CXCR2 C1208T polymorphism and susceptibility to cancer, (b) studies containing essential genotype information for calculating ORs, (c) control group must be consistent with Hardy-Weinberg equilibrium (HWE) balance, and (d) manuscripts written in English or Chinese. The exclusion criteria were as follows: (a) data for the control group was not available, (b) not enough data to measure the odds ratios (ORs), and/or (c) no relevance to the CXCR2 C1208T polymorphism and risk of cancer.

Data
Extraction. Study characteristics were as follows: name of the author, publication year, origin, cancer type, race of population, control source, genetic distribution of CXCR2 variant, sample size of case and control, HWE, and age range of the case and control groups. In the subgroup analysis, PCA and bladder cancer (BLCA) were classified as the urinary cancer group. The digestive system cancer included gastric, esophageal, colon, and rectal cancers. One study focused on Kaposi's sarcoma, and it was classified as "other cancer." 2.4. Statistical Analyses. We applied ORs and 95% confidence intervals (CIs) to investigate the strength of association between the CXCR2 C1208T variant and susceptibility to cancer. Five genetic models were adopted to evaluate the overall ORs. For the CXCR2 C1208T polymorphism, the five models were allelic contrast (T allele vs. C allele), heterozygous comparison (TC vs. CC), homozygous model (TT vs. CC), dominant (TT+TC vs. CC), and recessive (TT vs. TC+CC) model [37,38]. A Q statistic test was used to calculate the heterogeneity among the included studies. A P value of heterogeneity (P heterogeneity ) of less than 0.05 indicated heterogeneity among studies. In this case, the random effects method was selected (DerSimonian and Laird). If P heterogeneity > 0:05, the fixed effects method was chosen (Mantel-Haenszel) [39,40]. We performed Fisher's exact test to measure the P value of HWE (P HWE ). Studies that were not in accordance with the HWE balance were removed. Studies with a sample size greater than 1000 were classified as large-sample groups. Stratification analysis included cancer type, control source, race, sample size, and quality of studies. Sensitivity analysis for the CXCR2 C1208T mutation was carried out by removing every single study in turn. Publication bias was evaluated by Begg's funnel plot and Egger's test. P > 0:05 indicated no evidence of publication bias. All statistical analyses were conducted using STATA software (v11.0, Stata Company, College Station, TX, USA).        [41]. A total of 220 PCA patients (confirmed by needle biopsy) were enrolled from Changzhou No.2 People's Hospital and the Affiliated Hospital of Jiangnan University. After signing the written informed consent, each patient donated 2 mL of peripheral blood for the detection of serum CXCR2 by ELISA. The tissue expression of CXCR2 was evaluated by immunohistochemical staining (IHS) in PCA patients from our hospitals. For ELISA analysis, we collected the patient's blood in a standard cube that did not contain anticoagulants. A serum separation tube was used to solidify and centrifuge the sample at 1000 × g (15 min). The serum was immediately taken out for determination and stored     Figure 1).  Figure 2(a)). However, in the subgroup analysis by cancer type, individuals carrying the TT genotype had a 39% higher risk of urinary cancer than those carrying the CC genotype (homozygous comparison, 95%CI = 1:04 -1.87, P = 0:025, Figure 2(b)). In homozygous comparison, the CXCR2 C1208T variation was also associated with ele-     Journal of Immunology Research findings were observed in the allelic contrast, heterozygous comparison, and dominant models. In the subgroup analysis by sample size, the results were different between the large-and small-sample studies. In small-sample size studies, individuals with the TT+TC genotype had a 15% decreased risk compared to those with the CC genotype (95%CI = 0:74-0.98, P = 0:026). In large-sample size studies, individuals carrying the TC genotype had a 26% increased risk, compared to those with the CC (95%CI = 1:10-1.45, P = 0:001, Figure 2(c)). In the subgroup analysis by source of control, the TT genotype was associated with increased risk of cancer in hospital-based studies using the recessive model (OR = 1:47, 95%CI = 1:04-2.09, P = 0:030, Figure 2(d)). However, no positive association was revealed in population-based studies (OR = 1:12, 95%CI = 0:96-1.30, P = 0:162).

Combined Analysis of CXCR2.
We used the THPA database to evaluate the expression of CXCR2 in human tissues. CXCR2 was mainly expressed in the human appendix, spleen, bladder, and esophagus (Figure 3(a)). We further used GEPIA database to explore the CXCR2 expression in various cancers. CXCR2 expression was downregulated in BLCA, lymphoid neoplasm diffuse large B-cell lymphoma (DLBC), esophageal carcinoma (ESCA), head and neck squamous cell carcinoma (HNSC), lung adenocarcinoma (LUAD), skin cutaneous melanoma (SKCM), and thymoma (Thymoma) (Figure 3(b)) subjects. Nevertheless, CXCR2 expression was augmented in patients with acute myeloid leukemia (LAML). Moreover, ELISA was employed to assess the serum expression of CXCR2 in PCA patients recruited from our hospitals. Compared with PCA subjects carrying the CC genotype, the expression of CXCR2 was decreased in patients with the TT genotype (Figure 4(b), P < 0:05). Furthermore, we used IHS analysis to verify the results of bioinformatical findings. As shown in Figure 4(c), the expression of CXCR2 was attenuated in more advanced PCA, compared with less advanced patients (T3+T4 versus T1+T2, P < 0:05). The expression of CXCR2 was decreased in PCA participants, relative to that in normal counterparts (P < 0:05, Figure 4(a)). We used R language to investigate the gene-gene correlation of CXCR2 in PCA. The heatmap plot was shown in Figure 5(a). The most correlated genes include regulator of G protein signaling 11 (RGS11, Figure 5(b)), progestin and adipoQ receptor family member 6 (PAQR6, Figure 5(c)), and copine 7 (CPNE7, Figure 5(d)). We further employed the STRING database to investigate the protein-protein correlations of CXCR2. More than 10 proteins were involved in interacting with the CXCR2 protein ( Figure 6(a)). They were interleukin-8 (CXCL8), C-X-C motif chemokine 5 (CXCL5), growth-regulated alpha protein (CXCL1), C-X-C motif chemokine 2 (CXCL2), platelet basic protein (PPBP), C-X-C motif chemokine 6 (CXCL6), C-X-C motif chemokine 3 (CXCL3), stromal cell-derived factor 1 (CXCL12), C-X-C chemokine receptor type 1 (CXCR1), and C-C motif chemokine 5 (CCL5). We further performed GSEA to investigate potential associated signaling pathways correlated with CXCR2 expression. Plot of normalized enrichment score (NES) versus false positive rate (FDR) was indicated in Figure 6(b). The GSEA showed evidence that the expression of CXCR2 is correlated with pathways in cancer ( Figure 6(c), P < 0:05). Signaling pathways, including mitogen-activated protein kinase (MAPK) signaling (Figure 6(d)) and calcium signaling system (Figure 6(e)), were associated with a high expression of CXCR2. Furthermore, we used the TIMER database to evaluate the prognostic relevance of immune cell infiltration in PCA, BLCA, and KIRC samples. The survival time of BLCA patients with high CD8+ T cells was significantly shorter than that of BLCA patients with low infiltration (P = 0:006). The survival time of KIRC patients with a high expression of CXCR2 was significantly longer than that in the low CXCR2 group (P = 0:039, Figure 7).

Sensitivity Analysis and Publication Bias.
We performed a sensitivity analysis to evaluate the effect of each single  11 Journal of Immunology Research study on overall ORs. Publication bias was investigated using Begg's funnel plots and Egger's tests. As shown in Supplemental Figure 1A, no single study was found to have a significant impact on the ORs when assessing the CXCR2 C1208T polymorphism. Begg's funnel plots (Supplemental Figure 1C, P > 0:05) and Egger's tests (Supplemental Figure 1B, P > 0:05) also revealed no evidence of publication bias among the studies on the CXCR2 variants.

Discussion
Carcinoma remains a significant health issue worldwide, even though a large number of previous studies have explored the susceptibility and high-risk factors of malignant tumors. However, specific markers for several cancers have not been identified up to now [44,45]. It is necessary to investigate universal molecular markers to form a prognosis for the treatments of cancer patients. Previous studies showed that the expression of CXC chemokine receptors, including CXCR2, is associated with necrosis and development of several cancers [23,46,47]. Moreover, CXCR2 acts as an autocrine or paracrine growth factor to induce tumor invasion and migration [48]. Genetic polymorphisms of CXCR2 may influence the function of the protein by regulating gene expression. Several previous studies have investigated the correlation of CXCR2 mutations with risk of cancer [32][33][34][35][36][37][38]. However, the conclusions of these studies were not consistent. Ryan et al. [36] evaluated the CXCR2 polymorphism in Indians and revealed that C1208T was associated with an increased risk of bladder cancer (P = 0:003, OR = 1:29). Peng et al. [37] assessed the CXCR2 variant in two independent populations (European and Asian) and found that C1208T was associated with a reduced risk of lung cancer. In 2017, Yang et al. [48] performed a meta-analysis and indicated that CXCR2 expression in tumor tissues was correlated with poor prognosis in solid tumor patients. In 2018, Qiao et al. [49] conducted another meta-analysis and revealed that CXCR2 was an unfavorable predictor of overall survival and recurrencefree survival in patients with cancers, except for digestive cancer. However, these researchers did not evaluate the association between the CXCR2 C1208T polymorphism and susceptibility to cancer, which remains controversial. Recently, Zhu et al. conducted a pooled analysis and found that interleukin-8 receptor B rs1126579 C>T variation may be correlated with susceptibility to cancer [50]. However, they did not include all the available data on this gene polymorphism. In the current study, we pooled 14 eligible casecontrol studies on the CXCR2 C1208T variant, comprising 4,909 cancer cases and 5,884 controls. Our study revealed a significant association between the CXCR2 C1208T polymorphism and risk of cancer. The CXCR2 gene variation played a different role in various tumors.
In the stratified analysis by cancer type, we observed that the C1208T variant was correlated with increased risk of urinary, breast, and digestive cancers. For lung cancer, individuals with the TT genotype had a 30% decreased risk compared to those with the CC genotype, consistent with the reports in previous studies by Peng et al. [37]. In the subgroup analysis by race, no positive results were demonstrated in participants of European or African descent. One possible reason may be that the number of studies on Africans included in our analysis was relatively small. However, results from Bondurant et al. [34] indicate that CXCR2 C1208T polymorphism is associated with increased breast cancer risk in African patients. As described in the subgroup analysis by sample size, the results derived from the largeand small-sample studies may be different. Further research with large-sample sizes on the CXCR2 C1208T variation in African population is warranted in the future. In addition, we used combined analysis to explore the expression of CXCR2 in urinary cancer according to the race of Log-rank P = 0.14 Log-rank P = 0.563 Log-rank P = 0.85 Log-rank P = 0.363 Log-rank P = 0.07 Log-rank P = 0.608 Log-rank P = 0.457 Log-rank P = 0.006 Log-rank P = 0.889 Log-rank P = 0.107 Log-rank P = 0.901 Log-rank P = 0.523 Log-rank P = 0.541 Log-rank P = 0.305 Log-rank P = 0.126 Log-rank P = 0.363 Log-rank P = 0.099 Log-rank P = 0.669 Log-rank P = 0.865 Log-rank P = 0.039 Log-rank og-rank P P = 0.808 0 808 Log-rank Log-ran P P = 0.14 = 0 14 Log-ra Log o nk Log rank P P = 0.563 0 5 Log-rank Log-ran P P = 0.85 = 0 Log-rank og rank P P = 0.363 0 363 Log-rank og-ran P P = 0.07 = 0 Log-rank og-ra k P P = 0.608 0 608 Log-rank Log rank P P = 0.457 0 4 Log-rank Log ra k P P = 0.006 = 0 00 Log-rank Log rank P P = 0.889 = 0 88 Log-rank Log rank P P = 0.107 = 0 10 Log-rank Log rank P P = 0.901 0 9 Log-rank Log rank P P = 0.523 = 0 52 Log-rank Log ra k P P = 0.541 0 54 Log-rank Log rank P P = 0.305 0 3 Log-rank Log-ra k P P = 0.126 = 0 12 Log-rank Log rank P P = 0.363 = 0 36 Log-rank Log rank P P = 0.099 = 0 09 Log-rank Log rank P P = 0.669 0 6 Log-rank Log rank P P = 0.865 = 0 86 Log-rank Log-ra k P P = 0.039 0 0 participants. CXCR2 expression was downregulated in BLCA patients of European and African American descent. The expression of CXCR2 was also downregulated in PCA patients of European descent. Expression profiles of CXCR2 in PCA patients of Asian descent were not available from the online database. We used ELISA to investigate the serum expression of CXCR2 in needle biopsy confirmed PCA patients enrolled in our centers. It was observed that the expression of CXCR2 was decreased in PCA participants with the TT genotype, consistent with results derived from the pooled analysis. In addition, CXCR2 expression was decreased in PCA participants, compared with controls.
Our results show that patients with advanced PCA had lower levels of CXCR2, compared with early-stage patients. It is suggested that the detection of CXCR2 may provide guidance for the prognosis of patients with PCA. We further used GSEA to explore the signaling pathways potentially correlated with CXCR2 expression. Signaling pathways, such as MAPK signaling and calcium signaling system, were associated with high CXCR2 expression. Moreover, we used the TIMER database to evaluate the prognostic relevance of immune cell infiltration in PCA, BLCA, and KIRC samples. The survival time of BLCA patients with high CD8+ T cells was significantly shorter than that of BLCA patients with low infiltration. The survival time of KIRC patients with high expression of CXCR2 was significantly longer than that of the low CXCR2 group. There are, however, several limitations to the current analysis. First, there is only one case-control study on participants of African descent based on the inclusion criteria. Future studies with large-sample sizes focused on African populations are warranted. Second, the sample size of the included studies on the CXCR2 C1208T polymorphism remains insufficient. The number of studies on tumors including KIRC, PCA, and BLCA is quite limited. Third, we revealed that the CXCR2 C1208T polymorphism may be associated with increased risk of PCA. Further research is warranted to confirm whether the CXCR2 C1208T mutation has an impact on CXCR2 expression in PCA. As a single variation cannot have a huge impact on the development of cancer, it is necessary to investigate the gene-gene or gene-environment interactions in future research.
In conclusion, the present study summarizes all eligible genetic data for correlation between the CXCR2 C1208T variant and risk of cancer. Our study revealed that the CXCR2 C1208T polymorphism is associated with increased risk of urinary, breast, and digestive cancers. CXCR2 C1208T may also be associated with risk of PCA. Finally, CXCR2 expression was negatively correlated with the degree of PCA malignancy.

Data Availability
All the data in this study can be provided by corresponding authors if the request is reasonable.