RESEARCH ARTICLE

Impact of chemokines CCR5Δ32, CXCL12G801A,and CXCR2C1208Ton bladder cancer susceptibilityin north Indian population

Vibha Singh & Praveen Kumar Jaiswal & Rohit Kapoor &

Rakesh Kapoor & Rama Devi Mittal

Received: 13 September 2013 /Accepted: 5 January 2014# International Society of Oncology and BioMarkers (ISOBM) 2014

Abstract Chemokines are small inducible pro-inflammatorycytokines. In the present study, we tested association of che-mokine single nucleotide polymorphisms (SNPs) viz.,CCR5Δ32, CXCL12G801A and CXCR2C1208T genes inbladder cancer (BC) patients and normal healthy controls ofnorth Indians. Genotyping of the above SNPs were done in200 BC cases and 200 healthy controls, using RFLP andamplification refractory mutation system–polymerase chainreaction methodology. A significant association was foundin CXCL12G801Awith BC risk. In case of CXCL12G801Apolymorphism, the heterozygous (GA) genotype showed sig-nificantly high risk (p<0.001, odds ratio (OR)=2.72), whereasA allele carrier (GA + AA) also showed risk with BC(p<0.001, OR=2.44). In CXCR2C1208T polymorphism, thevariant genotype (TT) showed significant risk for BC (p=0.028, OR=1.58). The variant allele (T) of CXCR2C1208Tpolymorphism was found to be associated with BC risk (p=0.003, OR=1.29). Interestingly, smoking was also found tomodulate 1.16-fold risks for BC in case of CXCR2C1208T,variant genotype (TT). Upon analyzing the gene–gene inter-action between CXCR2C1208Tand CXCL12G801A, the com-bination CT-GA showed 4-fold risk for BC (p=0.009). Ourresults indicated that polymorphism in CXCR2C1208T andCXCL12G801A showed high risk for BC in north Indianpopulation. However, CCR5Δ32 exhibited no association.Studywith large sample size and diverse ethnicity are requiredto validate these observations.

Keywords Chemokinegenepolymorphism .Bladdercancer .

Bacillus Calmette–Guerin . RFLP

Introduction

Bladder cancer (BC) is the fourth most prevalent cancer andthe second most frequently diagnosed urologic cancer in men.BC is a multifactorial and polygenic malignancy, and inaddition to genetic status of individuals, gene environmentalinteractions are thought to play apparent role. Polymorphismsin many genes have been studied in connection with bladdercancer risk of which inflammation is playing important role indisease progression. Inflammatory response is mainlygoverned by chemokines. The role of chemokines in thepathogenesis of bladder cancer is currently of considerableinterest. Chemokines and chemokine receptors are key regu-lators in inflammation and cancer. They have been reported toplay important roles in tumor progression, invasion, and me-tastasis. Chemokines are produced by tumor cells as well asby cells of the tumor microenvironment including cancer-associated fibroblasts, mesenchymal stem cells, endothelialcells, and more recently tumor-associated neutrophils [1].The human CC chemokine receptor 5 (CCR5), an importantco-receptor for macrophage-tropic strains (R5) of HIV-1, en-codes a cell surface receptor that is the main co-receptor forHIV-1 entry into CD4+ T lymphocytes [2]. Chemokine C-Cmotif receptor 5 (CCR5) is a receptor for the pro-inflammatory chemokine macrophage inflammatory proteins1a (MIP-1a), MIP-1b, and regulated on activation, normal Tcell expressed and secreted (RANTES) [3, 4]. The expressionof RANTES (ligand for CCR5) has been reported in differentcancers. CCR5 plays a crucial role in antitumor immunitythrough immune cell recruitment. Human CCR5 gene ismapped at chromosomal position 3p21.3. CCR5 encodes a

V. Singh : P. K. Jaiswal : R. Kapoor : R. Kapoor : R. D. Mittal (*)Department of Urology and Renal Transplantation, Sanjay GandhiPost Graduate Institute of Medical Science, Raebareli Road,Lucknow 226014, Uttar Pradesh, Indiae-mail: ramamittal@gmail.com

R. D. Mittale-mail: Rdm_pgi@rediffmail.com

Tumor Biol.DOI 10.1007/s13277-014-1624-7

352-amino acid protein with molecular mass of 40,600Da [5].The most reported polymorphism is a 32-base pair deletion inthe open-reading frame (ORF) of CCR5 (CCR5Δ32) [6].

The CXCL12, also known as a stromal cell-derived factor(SDF-1), belongs to C-X-C (Cys-Xxx-Cys) subfamily of che-mokine [7]. This gene located on chromosome 10q 11.1,27revealed that a single nucleotide polymorphism (SNP), aguanine to adenine (G > A), at position 801 of the 3′ untrans-lated gene region may affect the expression of SDF-1 chemo-kine [8]. CXCL12 has been reported to be involved in cellproliferation cell migration, and cell invasion, while the che-mokine CXCL12 and its receptor CXCR4 have been found tobe associated with cancer metastasis [9]. The variant A/Ahomozygote genotype of CXCL12 has been suggested to alterthe production of CXCL12 and is associated with the risk ofcarcinogenesis of various origins, including oral, lung, andbreast cancers and leukemia [10]. CXCL12 plays importantroles in both normal embryonic development and migrationand metastasis of cancer cells [11]. CXCR2 is predominantlyexpressed on neutrophils, and lower level of expression isfound on monocytes and lymphocytes [12]. CXCR2C1208Tare encoded by two separate genes, IL8RA and IL8RB, re-spectively, the gene located on 2q34-q3517. The CXCR2genehas 11 exons, and their differential splicing gives rise todistinct mRNA variants [13].

We aimed to evaluate the influence of genetic polymor-phism of CCR5Δ32I/D, CXCL12G801, and CXCR2C1208Twith the susceptibility of BC. We also looked for the possiblerole of confounding factors such as sex, age of onset of thedisease, and smoking and whether the risk of BC was modu-lated by the genes selected.

Materials and methods

Study subjects

The bladder cancer patients in this analysis were enrolled froman ongoing case–control study of bladder cancer, whichstarted patient recruitment in 2010. All enrolled patients wereincident cases of histological confirmed invasive or superficialbladder cancer and were recruited from the UrologyDepartment at Sanjay Gandhi Postgraduate Institute ofMedical Sciences, a tertiary care center. A total of 200 patientswith histological confirmed transitional urothelial BC (meanage, 58.5 years; 175 men and 25 women) were recruited forthe study. Those with previous history of other cancer, cancermetastasized to the bladder from another origin, and previousradiotherapy were excluded. Healthy and genetically unrelat-ed individuals visiting the hospital for a routine checkup orhealth awareness camps and hospital employees were recruit-ed as the controls (n=200). All the controls were age and sexmatched with similar ethnicity and had no evidence of

malignancy or chronic disease. The mean age of the controlswas 56.8 years, andM/F ratio as 179:21. The disproportionateratio between male and female bladder cancer in our popula-tion could be largely due to the increased prevalence in males(3:1). The participation rate was 100 %, and blood sampleswere available for all subjects. An epidemiologic question-naire was designed for study participants to collect data ondemographic characteristics, smoking history, occupation his-tory, and other lifestyle factors that were employed. At the endof the interview, a 5-ml blood sample was drawn into codedtubes. Informed consent was taken from all subjects wheninterviewing for the demographic details and blood samplecollection. The Ethical Review Board of the Institute ap-proved the study.

Epidemiology data collection

The demographic details were obtained by interviewing eachindividual in cases and controls. The response rate for theinterview was 75 % for the subjects. Individuals who smokedonce a day for more than 5 years were defined as smokers. Theindividuals who had never smoked in their lifetime wereregarded as nonsmokers. At the conclusion of the interview,a 5 ml of blood sample was drawn into coded vials.

Clinical data collection

The demographic and clinical characteristics of the patientsare presented in Table 1. The clinical information about tumorsize, number, stage and tumor grade, intravesical therapy anddates of recurrence, chemotherapy, radical cystectomy, andpathological findings at cystectomy were provided by theUro-Oncologist in our department. The tumor stages wereclassified as per American Joint Committee on Cancer’sTNM staging system [14] Of the 200 total patients enrolledin the study, 149 patients had non-muscle invasive bladdercancer (NMIBC), while the rest, 51, had muscle invasivebladder cancer (MIBC). Patients with NMIBC at high risk(high-grade, multiple, and large tumor) were treated withintravesical Bacillus Calmette–Guerin (BCG) (n=78). Thepatients with NMI cancer of low risk (low grade and singlesmall tumor) were kept on cystoscopic surveillance and con-sidered as non-BCG patients. Subsequently, all the patientswere examined by cystoscopy after every 3 months in first andsecond years and later at six monthly intervals as long as therewas no tumor recurrence. BCG treatment consisted of 6weekly instillation induction BCG (n=78). Since the numberof patients receiving maintenance BCG was too low, we didnot categorize the patients according to BCG regime forstatistical analysis. The end point of the study included tumorrecurrence, defined as a newly found bladder tumor followinga previous negative follow-up cystoscopy or end of study time(60 months). Patients with invasive BC (n=51) were treated

Tumor Biol.

with radical cystectomy with or without adjuvant chemother-apy, which included cisplatin and gemcitabine, followed byperiodical cystoscopy. Blood sample was collected in EDTAfrom all subjects for genotyping at the time of enrollment andstored at −70ºC.

Genotyping

Genomic DNAwas extracted from peripheral blood lympho-cytes by salting out method [15]. CCR5Δ32 andCXCL12G801A gene polymorphisms were analyzed by usingpolymerase chain reaction–restriction fragment length poly-morphism, and CXCR2C1208T gene polymorphism was ana-lyzed by using the methodology ARMS–PCR (amplification

refractory mutation system–polymerase chain reaction).Details of primers and PCR conditions for CCR5Δ32 [16],CXCL12G801A [9], andCXCR2C1208T [13] were taken fromsomewhere else. Genotyping was done on 15 % PAGE and2 % agarose gel, using molecular weight markers and visual-ized after staining with ethidium bromide. Positive and nega-tive controls were used in each genotyping assay, and 10 % ofthe samples were randomly selected and run in duplicates with100 % concordance. The results were reproducible with nodiscrepancy in genotyping.

Statistical analysis

The power of the study was calculated using Quanto software,version 1.0 (available from: http://hydra.usc.edu/gxe) withinput of the following variables: case–control study design,significance level (alpha) >0.05 (two-sided), model of inher-itance was log additive, allele frequency was 0.28, and thegenetic effect for odds ratio (OR) was 1.65 or greater. Thepresent study achieved 80 % of the statistical power. Thegoodness-of-fit chi-square test was used to analyze any devi-ation from the Hardy–Weinberg equilibrium in controls. Abinary logistic regression model was used to estimate the risk,as the OR was at 95 % confidence interval (CI). Bonferroni’scorrection was applied in case of multiple comparisons usingthe the formula pc=p×n (pc represents corrected value where nis the number of comparisons performed). The statisticalanalysis was done using the Statistical Package for SocialSciences software, version 15.0 (SPSS; Chicago, IL), andp<0.05 was considered statistically significant.

Results

Characteristics of subjects

There was no significant age difference between the cases(58.5±12.4 years) and the controls (56.8±10.8 years) (p=0.117). The cases had significantly higher percentage ofsmokers (69.9 %) than the controls (22.5 %) (p<0.001). Thedemographic details of the study subjects and clinical charac-teristics of the patients are presented in Table 1.

Chemokine gene polymorphisms in bladder cancer

The genotype and allele frequencies of chemokine gene poly-morphism in healthy individuals (controls) and BC patientsare presented in Table 2. The genotype frequencies of controlswere in Hardy–Weinberg equilibrium (HWE). We found sig-nificant association in CXCL12G801A and CXCR2C1208Tpolymorphism with BC risk, whereas CCR5Δ32 polymor-phism showed no association with BC. In CXCR2C1208Tpolymorphism, the variant genotype (TT) showed significant

Table 1 Demographical details of bladder cancer patients and healthycontrols

Variables Cases n=200 Controls n=200 Chi-squaren (%) n (%) p value

Sex

Female 25 (12.5) 21 (10.5) 0.531Male 175 (87.5) 179 (89.5)

Age (years)

Mean age ± SD 58.5±12.4 56. 8±10.8 0.117a

Smokingb

Nonsmokers 47 (30.1) 155 (77.5) <0.001Smokers 109 (69.9) 45 (22.5)

Tumor number

Single 122 (60.8) – –Multiple 78 (39.2) –

Tumor size (cm)

<1 49 (24.3) – –1–3 101 (50.7) –

>3 50 (25.0) –

Stage

Ta 64 (32.0) – –T1 85 (42.5) –

T2 51 (25.5) –

Grade

G1 67 (33.5) – –G2 43 (21.5) –

G3 90 (45.0) –

Intravesical therapy

Non-treated 71 (47.7) – –BCG induction(BCG i + m)

78 (52.3) –

Event

Recurrence 65 (43.9) – –Nonrecurrence 84 (56.1) –

BCG i + m Bacillus Calmette–Guerin induction + maintenancea Student t test was used to determine the p valueb The sum could not add up to the total due to some missing values

Tumor Biol.

risk for BC (p=0.028, OR=1.58). The variant allele (T) ofCXCR2C1208T polymorphism was found to be associatedwith BC (p=0.003, OR=1.29). In CXCL12G801A polymor-phism, the heterozygous genotype showed significant risk forbladder cancer (p<0.001; OR=2.72; 95 % CI, 1.78–4.15). Oncombining heterozygous and variant genotype, it showed highrisk for bladder cancer (p<0.001; OR=2.44; 95 % CI, 1.61–3.68). While looking at allelic level, there was no significantrisk for BC in case of any chemokine.

Association of chemokine genotypes with smoking

We evaluated the gene–smoking interaction to study themodulation of BC risk with respect to CCR5Δ32,CXCR2C1208T, and CXCL12G801A. The patients weregrouped as nonsmokers and smokers. In the case ofCXCL12G801A gene, the variant genotype was associ-ated with the reduced risk of BC (p=0.025; OR, 0.16),whereas in the case of CXCR2C1208T, the variant genotype(TT) was found to modulate a 1.16-fold risk for BC withsmoking (Table 3).

Association of chemokine genotypes with tumor stage/grade

The patients with similar stage but with different gradesrespond to treatment differently. Hence, we stratified the pa-tients into three subgroups according to stage/grade [TaG1

(low-risk NMIBC), TaG2, 3 + T1G1−3 (high-risk NMIBC),and T2+ (muscle invasive)]. TaG1 was taken as a reference.No significant association was observed in case CCR5Δ32,CXCR2C1208T, and CXCL12G801A in BC patients (Table 4).

Modulation of genotype variants and outcome after BCGimmunotherapy

To analyze the association of CCR5Δ32, CXCR2C1208T, andCXCL12G801Agene polymorphisms and risk of recurrence inNMIBC patients, further analysis was restricted to NMIBCpatients only (n=148). The median follow-up of NMIBCpatients was 14 months (3–60 months). We analyzed theassociation of genotypes and risk of recurrence after BCGimmunotherapy. We grouped patients into BCG treated (n=78) and non-treated (n=70), as these were patients of low-grade tumors and did not require BCG immunotherapy. Noneof the three polymorphisms were associated with the risk ofrecurrence-free survival (Fig. 1).

Gene–gene interaction of CXCR2C1208T–CXCL12G801Agene polymorphisms

To analyze the combined effect of these polymorphisms, weconducted gene–gene interaction. Combination of CXCR2–CXCL12 showed statistically significant results. WhenCXCR2 and CXCL12were analyzed to see the potential effectin combination, combination of CXCR2 and CXCL12 re-vealed, (CC-GA), (CT-GA), and (TT-GA), different geno-types of CXCR2 combined with different genotype CXCL12.They showed to be associated with the increased BC risk.After applying the Bonferroni correction, only the combina-tion CT-GA showed significant risk for BC (pc=0.009)(Table 5).

Discussion

Chemokine/chemokine receptor interactions can protectagainst tumor development/growth or can stimulate melano-ma tumor progression, tumor growth, and metastasis.Metastatic melanoma cells express chemokine receptors thatplay a major role in the specifying the organ site for metasta-sis, based upon receptor detection of the chemokine gradientelaborated by a specific organ/tissue. The expression ofchemokines and chemokine receptors are often strongly up-regulated during tumorigenesis in case of different cancerslike breast, lung, prostate, colon, ovary, and bladder [1].

Table 2 Genotype and allele frequency distribution of CCR5Δ32,CXCR2C1208T, and CXCL12G801A gene polymorphisms in bladdercancer patients and healthy controls

Genotype Controlsn=200

Casesn=200

p value ORa (95 % CI)

n (%) n (%)

CCR5Δ32I/I 184 (92.0) 192 (96.0) Ref Ref

I/D 15 (7.5) 8 (4.0) 0.113 0.42 (0.15–1.23)

D/D 1 (0.5) 0 (0.0) NC NC

I/D + D/D 16 (8.0) 8 (4.0) 0.221 0.79 (0.53–1.15)

I allele 383 (95.7) 392 (98.0) Ref Ref

D allele 17 (4.3) 8 (2.0) 0.098 0.48 (0.20–1.15)

CXCR2C1208T

CC 112 (56.0) 96 (48.0) Ref Ref

CT 73 (36.5) 90 (45.0) 0.155 1.45 (0.87–2.44)

TT 15 (7.5) 14 (7.0) 0.028 1.58 (0.59–4.27)

CT + TT 88 (44.0) 110 (55.0) 0.364 1.56 (1.05–2.31)

C allele 297 (74.3) 276 (69.0) Ref Ref

T allele 103 (25.7) 124 (31.0) 0.003 1.29 (0.95–1.76)

CXCL12G801A

GG 101 (50.5) 59 (29.5) Ref Ref

GA 83 (41.5) 132 (66.0) <0.001 2.72 (1.78–4.15)

AA 16 (8.0) 9 (4.5) 0.930 0.96 (0.40–2.32)

GA + AA 99 (49.5) 141 (70.5) <0.001 2.44 (1.61–3.68)

G allele 285 (71.3) 250 (62.5) Ref Ref

A allele 115 (28.7) 150 (37.5) NC NC

aOR (95 % CI) age, gender, and smoking-adjusted odds ratio and 95 %confidence interval

Tumor Biol.

CCR5Δ32 is expressed preferentially by CD4+ Th1 cells andplay crucial role in bladder cancer. In our study, we did notfind any significant association between CCR5Δ32 polymor-phismwith BC risk. In case of ESRD patients, same result wasfound in north Indian population [17]. The reason for noassociation with BC risk may be attributed to the ethnicvariation and/or due to the absence of variant genotype/allelein the case of bladder cancer patients. Our result however wascompatible to the diabetic nephropathy in north Indian popu-lation [18]. One of the studies from our lab in renal transplantfound no association with CCR5Δ32 [16]. A significant asso-ciation was observed in CCR5Δ32 allele and multiple sclero-sis in Sicilian population [19]. Srivastava et al. reported thatpatients with CCR5+/Delta32 genotype (p=0.028, OR=2.850, 95 % CI=1.1–7.2) and CCR5Δ32 allele (p=0.012,

OR=3.145, 95 % CI=1.2–7.7) showed a high risk of gallbladder cancer [20]. Another study of the risk of BC in aTurkish population is contradictory to our observations inbladder cancer patients [21]. CCR5Δ32molecules are capableof forming heterocomplexes with normal CCR5 in heterozy-gous individuals, retaining normal CCR5 molecules in theendoplasmic reticulum and reducing cell surface expressionof CCR5 [22]. In vitro expression studies have also demon-strated that mutant CCR5Δ32 protein is retained in the endo-plasmic reticulum and exerts a transdominant negative effecton wt CCR5, impairing its transport to the cell surface. Thetruncated CCR5Δ32 protein can further act as a negativeregulator of wild-type (wt) CCR5 [23].

The chemokine CXCL12 and its receptor CXCR4 havebeen found to be associated with cancer metastasis in prostate

Table 3 Association ofCCR5Δ32, CXCR2C1208T, andCXCL12G801A gene polymor-phisms with smoking habit in BCsusceptibility

Genotype Patients (nonsmokers) Patients (smoker) p value OR (95 % CI)n=47 n=109n (%) n (%)

CCR5Δ32II 44 (93.6) 106 (97.2) Ref Ref

ID 23 (6.4) 3 (2.8) 0.293 0.42 (0.08–2.14)

DD 0 (0.0) 0 (0.0) NC NC

CXCR2C1208T

CC 23 (48.9) 54 (49.5) Ref Ref

CT 21 (44.7) 47 (43.2) 0.707 1.64 (1.08–2.47)

TT 3 (6.4) 8 (7.3) 0.019 1.16 (0.53–2.53)

CXCL12G801A

GG 8 (17.0) 31 (28.4) Ref Ref

GA 34 (72.4) 75 (68.8) 0.208 0.57 (0.40–1.90)

AA 5 (10.6) 3 (2.8) 0.025 0.16 (0.03–0.79)

Table 4 Association of CCR5Δ32, CXCL12G801A, and CXCR2C1208T gene polymorphisms with tumor grade/stage categories of BC patients

(a) (b) (c) p value (a–b) OR, 95 % CI* p value (a–c) OR, 95 % CI*n=37 n=112 n=51n (%) n (%) n (%)

CCR5Δ32 TaG1 TaG2−3,T1G1−3 T2+

I/D 35 (94.6) 111 (99.1) 46 (90.2) Reference

I/D 2 (5.4) 1 (0.9) 5 (9.8) 0.136 0.16 (0.02–1.79) 0.458 1.90 (0.35–10.38)

D/D 0 (0.0) 0 (0.0) 0 (0.0) NC NC NC NC

CXCL12G801A TaG1 TaG2−3,T1G1−3 T2+ ReferenceGG 14 (37.8) 31 (27.6) 14 (27.5)

GA 21 (56.8) 75 (67.0) 36 (71.5) 0.24 1.61 (0.73–3.57) 0.25 1.71 (0.68–4.28)

AA 2 (5.4) 6 (5.4) 1 (1.0) 0.73 1.36 (0.24–7.57) 0.59 0.50 (0.04–6.17)

CXCR2C1208T TaG1 TaG2−3,T1G1−3 T2+

CC 13 (35.1) 51 (45.5) 26 (51.0) Reference

CT 23 (62.2) 50 (44.6) 23 (45.1) 0.140 0.55 (0.25–1.21) 0.123 0.50 (0.21–1.21)

TT 1 (2.7) 11 (9.9) 2 (3.9) 0.344 2.80 (0.37–13.73) 1.000 1.00 (0.08–12.07)

*OR odds ratio is adjusted with confounding factors viz. an age, gender and smoking at 95 % CI confidence interval for present study

Tumor Biol.

cancer [7]. CXCL12 plays an important role in the metastasisof cancer cells [8]. In our hospital-based case–control study ofnorth Indian population, we found that in the CXCL12G801Apolymorphism, the heterozygous genotype showed significantrisk for bladder cancer (p<0.001; OR=2.72; 95 % CI, 1.78–4.15). Combining heterozygous and variant genotype, highrisk for bladder cancer (p<0.001; OR=2.44; 95 % CI, 1.61–3.68) was revealed. The probable reason could be the presenceof one risk allele in heterozygous condition (GA) and two riskalleles in case of variant allele carrier genotype (GA + AA).Another reason may be due to the higher percentage of het-erozygous condition (GA) in BC patients. SDF1-3′A variantmay be associatedwith an increased risk of laryngeal cancer inPolish population [24]. In the case of breast cancer, theirinvestigation did not support the CXCL12-3′A gene variantas a risk factor in Polish population. In a further study by Tenget al., they found that A/G heterozygous genotype of SDF-1had a higher risk of 1.86-fold to develop oral cancer inTaiwanese populations [25]. No significant association was

found at genotypic level as well as allelic level in colorectalcancer in Swedish population and immune thrombocytopenia(ITP) [26, 27]. G801A polymorphism of the SDF-1 gene isassociated with advanced stages of oral cancer, especially inalcohol abusers in Athens ethnicity [28]. No significant asso-ciation was found in SDF-1 gene polymorphisms and occur-rence of SLE inMalaysian population [29]. Kucukgergin et al.found that AA genotype was associated with BC risk. Thepatients with AA genotype of SDF1-3′A gene presented ahigher risk for developing prostate cancer to patients withGG homozygotic condition [30]. Kim et al. observed thatCXCL12 expressed in liver, lung, and bone marrow promotemetastasis of melanoma tumor cells to these organs [31],whereas Crazzolara et al. found that SDF-1 mRNAwas pre-sented in 50 % of grades I and II, 67 % of grade III, and 37 %of grade IVof glioblastoma multiforme tissue [32].

CXCR2C1208T located in the noncoding region of CXCR2gene is assumed to provide valuable information for thepathogenesis and the susceptibility to chronic inflammatorydisease. In our study of CXCR2C1208T gene polymorphism,the variant genotype (TT) showed significant risk for BC (p=0.028, OR=1.58). The variant allele (T) of CXCR2C1208TC/T polymorphism was also found to be associated with BCrisk (p=0.003, OR=1.29). It may be assumed that perhaps thepresence of two risk alleles in variant genotype/allele (TT)could be the reason for the risk. Ethnicity again may accountto some extent the variations in subjects worldwide compar-atively. In a study by Javor et al., they found reduced risk inacute pyeloephritis [33], whereas in Turkish population, noassociation was observed with allelic or genotypic level ofCXCR2with Behçet’s disease [34]. A highly significant asso-ciation was found between the homozygous CXCR2C1208TTT genotype (adjusted OR=2.89, p=0.008) and breast carci-noma [35]. The CXCR2 variant genotype was significantlymore frequent (p<0.001, OR=2.76) with systemic sclerosis[36]. Ben-Baruch et al. have demonstrated that, beside the roleof IL-8 and CXCR2 inmediating the recruitment of the tumor-infiltrating leucocytes to tumor site, their expression may alsoaffect neoplastic proliferation and metastasis [37].

In our study, we evaluated the gene–smoking interaction tostudy the modulation of bladder cancer risk with respect toCCR5Δ32,CXCL12G801A, andCXCR2C1208T. However, nosignificant association was observed in the case of CCR5Δ32with BC risk. But significant association was found in the caseof heterozygous condition of CCR5Δ32 having p=0.035 andOR=3.27 with smoking gall bladder cancer patients [35]. Inthe case of CXCL12G801A gene, the variant genotype wasassociated with reduced risk of BC (p=0.025; OR, 0.16),whereas in case of CXCR2C1208T, the variant genotype (TT)was found to modulate a 1.16-fold risk for BC with smoking.

To analyze whether chemokine genes polymorphism pre-dispose enhanced risk to invasive BC, we compared genotypedistribution between superficial and invasive tumor patients

Fig. 1 Influence ofCCR5Δ32,CXCR2C1208T, andCXCL12G801Agenepolymorphisms on the risk of recurrence in BCG-treated NMIBC patients

Table 5 Gene–gene interaction of CCXCR2C1208T–CXCL12G801Apolymorphism in bladder cancer patients and healthy controls

Controls Patients OR (95 % CI) p valuen (%) n (%)

CCXCR2C1208T–CXCL12G801A

CC-GG 52 (26.0) 29 (14.5) Ref Ref

CC-GA 49 (24.9) 56 (28.0) 2.05 (1.13–3.71) 0.018*

CC-AA 11 (5.5) 5 (2.5) 0.82 (0.26–2.58) 0.728

CT-GG 40 (20.0) 27 (13.5) 1.21 (0.62–2.35) 0.575

CT-GA 30 (15.0) 64 (33.5) 4.01 (2.14–7.49) <0.001**

CT-AA 3 (1.5) 2 (1.0) 1.19 (0.19–7.57) 0.850

TT-GG 9 (4.5) 3 (1.5) 0.59(0.15–2.38) 0.466

TT-GA 4 (2.0) 9 (4.5) 4.03(1.14–14.25) 0.030***

TT-AA 2 (1.0) 2 (1.0) 1.79(0.24–10.51) 0.569

*pc=0.16; **pc=0.009;*** pc=0.27 (after Bonferroni correction)

Tumor Biol.

group. The patients with similar stages but with differentgrades respond to treatment differently. Hence, we stratifiedthe patients into three subgroups according to stage/grade[TaG1 (low-risk NMIBC), TaG2, 3 + T1G1−3 (high-riskNMIBC), and T2+ (muscle invasive)]. TaG1 was taken as areference, and no significant association was observed statis-tically in the case of CCR5Δ32, CXCL12G801A, andCXCR2C1208T with any of the stage/grade categories ofbladder cancer. To analyze the combined effect of these poly-morphisms, we conducted gene–gene interaction.Combination of CXCR2C1208T–CXCL12G801A showed sta-tistically significant results. When CXCR2C1208T andCXCL12G801A were analyzed to see the potential effect incombination, CC-GA, CT-GA, and TT-GA combination isassociated with the increased BC risk. After applyingthe Bonferroni correction, only CT-GA showed signifi-cant risk. None of the CCR5Δ32, CXCL12G801A, andCXCR2C1208T polymorphisms were associated with therisk of recurrence-free survival in NMIBC patients withBCG immunotherapy.

Our study has several strengths, like both the chemokinegene polymorphisms in our control subjects were followingHWE, all BC cases were histopathologically confirmed, andthe study subjects enrolled were of similar ethnicity since bothBC patients as well as controls were of north Indian ethnicity.

Conclusion

In conclusion, our findings also reinforce the role attributed toinflammation, angiogenesis, and their mediators as majorcontributing factors in the process of bladder tumor develop-ment , progression, and by using the CCR5Δ32,CXCL12G801A, and CXCR2C1208T polymorphisms aloneor in combination with other genetic polymorphisms in an-giogenic and inflammatory genes to predict bladder canceroutcome, and prognosis may therefore have an importantclinical significance. Our study though had some limitationsdue to small sample size that may lead to a relatively lowerstatistical power, particularly in subgroups. In addition, de-tailed stage-specific molecular and cellular expression studiesin tissue biopsy specimens of bladder cancer might help indetermining the precise functional effects of CCR5Δ32,CXCL12G801A, and CXCR2C1208T polymorphisms and ar-riving at definitive conclusions regarding the critical linkbetween these chemokines and bladder cancer. To the best ofour knowledge, this is the only study examining the associa-tion between chemokine gene polymorphisms and risk of BCfrom India.

Acknowledgments We are thankful to the urologists in our departmentfor providing bladder cancer samples and relevant clinical information ofpatients. Vibha Singh is thankful to ICMR New Delhi for providing

Senior Research Fellowship (SRF) and Praveen Kumar Jaiswal is thank-ful to UGC New Delhi for providing Junior Research Fellowship (JRF).

Conflict of interest None

References

1. Raman D, Baugher PJ, Thu YM, et al. Role of chemokines in tumorgrowth. Cancer Lett. 2007;256:137–65.

2. Morgan D, Mahe C, Mayanja B, et al. Progression to symptomaticdisease in people infected with HIV-1 in rural Uganda: prospectivecohort study. BMJ. 2002;324:193–6.

3. Horuk R. Molecular properties of the chemokines receptor family.Trends Pharmacol Sci. 1994;15:159–65.

4. Ruster M, Sperschneider H, Funfstuck R, Stein G, Grone HJ.Differential expression of beta-chemokines MCP-1 and RANTESand their receptors CCR1, CCR2, CCR5 in acute rejection andchronic allograft nephropathy of human renal allografts. ClinNephrol. 2004;61:30–9.

5. Liu R, Paxton WA, Choe S, et al. Homozygous defect in HIV-1coreceptor accounts for resistance of some multiply-exposed individ-uals to HIV-1 infection. Cell. 1996;86:367–77.

6. Huang Y, Paxton WA, Wolinsky SM, et al. The role of a mutantCCR5 allele in HIV-1 transmission and disease progression. NatMed. 1996;2:1240–3.

7. Borish LC, Steinke JW. Cytokines and chemokines. J Allergy ClinImmunol. 2003;111:S460–75.

8. Apostolakis S, Baritaki S, Kochiadakis GE, et al. Effects of poly-morphisms in chemokine ligands and receptors on susceptibility tocoronary artery disease. Thromb Res. 2007;119:63–71.

9. Hirata H, Hinoda Y, Kikuno N, et al. CXCL12 G801A polymor-phism is a risk factor for sporadic prostate cancer susceptibility. ClinCancer Res. 2007;13:5056–62.

10. de Oliveira CE, Cavassin GG, Perim Ade L, et al. Stromal cell-derived factor-1 chemokine gene variant in blood donors and chronicmyelogenous leukemia patients. J Clin Lab Anal. 2007;21:49–54.

11. Balkwill F. Cancer and the chemokine network. Nat Rev Cancer.2004;4:540–50.

12. Brant E, Ludwig A, Petersen F, Flad HD. Platelet-derived CXCchemokines: old players in new games. Immunol Rev. 2000;177:204–16.

13. Renzoni E, Lympany P, Sestini P. Distribution of novel polymor-phisms of IL-8 and CXC receptor 1 and 2 genes in systemic sclerosisand cryptogenic fibrosing alveolitis. Arthritis Rheum. 2000;43:1633–40.

14. Colombel M, Soloway M, Akaza H. Epidemiology, staging, gradingand risk stratification of bladder cancer. Eur Urol. 2008;7:618–26.

15. Miller SA, Dykes DD, Polesky H. A simple salting out procedure forextracting DNA from human nucleated cells. Nucleic Acids Res.1988;16:1215.

16. Singh R, Kapoor R, Srivastava A, et al. Impact of chemokine receptorCCR2 and CCR5 gene polymorphism on allograft outcome in northIndian renal transplant recipients. Scand J Immunol. 2009;69:51–6.

17. Manchanda PK, Singh R, Mittal RD. Cytokine (IL-10–1082 and -819) and chemokine receptor (CCR2 and CCR5) gene polymorphismin north Indian patients with end-stage renal disease. DNA Cell Biol.2009;28:177–83.

18. Ahluwalia TS, Khullar M, Ahuja M, et al. Common variants ofinflammatory cytokine genes are associated with risk of nephropathyintype 2 diabetes among Asian Indians. PLoS One. 2009;4:51–68.

19. D’Angelo R, Crisafulli C, Rinaldi C, et al. CCR5Δ32 polymorphismassociated with a slower rate disease progression in a cohort of RR-MS Sicilian patients. Mult Scler Int. 2011;2011:153–82.

Tumor Biol.

20. Srivastava A, Pandey SN, Choudhuri G, et al. CCR5 Delta32 poly-morphism associated with gallbladder cancer susceptibility. Scand JImmunol. 2008;67:516–22.

21. Kucukgergin C, Isman FK, Dasdemir S, et al. The role of chemokineand chemokine receptor gene variants on the susceptibility andclinicopathological characteristics of bladder cancer. Gene.2012;10(511):7–11.

22. Zafiropoulos A, Crikas N, Passam AM, Spandidos DA. Significantinvolvement of CCR2–64I and CXCL12–3a in the development ofsporadic breast cancer. J Med Genet. 2004;41:e59.

23. Paterlini MG. Structure modeling of the chemokine receptor CCR5:implications for ligand binding and selectivity. Biophys J. 2002;83:3012–31.

24. Kruszyna L, Lianeri M, Rubis B, et al. CCL2–2518 A/G singlenucleotide polymorphism as a risk factor for breast cancer. MolBiol Rep. 2010;38:1263–7.

25. Teng YH, Liu TH, Tseng HC, et al. Contribution of genetic poly-morphisms of stromal cell-derived factor-1and its receptor, CXCR4,to the susceptibility and clinicopathologic development of oral can-cer. Head Neck. 2009;31:1282–8.

26. Dimberg J, Hugander A, Löfgren S, et al. Polymorphism andcirculatinglevels of the chemokine CXCL12 in colorectal cancerpatients. Int J Mol Med. 2007;1:11–5.

27. Ku FC, Tsai CR, Der Wang J, et al. Stromal-derived factor-1 genevariations in pediatric patients with primary immune thrombocyto-penia. Eur J Haematol. 2013;90:25–30.

28. Vairaktaris E, Vylliotis A, Spyridonodou S, et al. DNA polymor-phism of stromal-derived factor-1 is associated with advanced stagesof oral cancer. Anticancer Res. 2008;28:271–5.

29. Lian LH, Kee BP, Ng HL, et al. Lack of association betweenRANTES-28, SDF-1 gene polymorphisms and systemic lupus

erythematosus in the Malaysian population. Genet Mol Res.2011;10:2841–50.

30. Işman FK, Kucukgergin C, Daşdemir S, et al. Association betweenSDF1-3′A or CXCR4 gene polymorphismswith predisposition toandclinicopathological characteristics of prostate cancer with or withoutmetastases. Mol Biol Rep. 2012;39:1073–9.

31. Kim J, Mori T, Chen SL, Amersi FF, Martinez SR, Kuo C, et al.Chemokine receptor CXCR4 expression in patients with melanomaand colorectal cancer liver metastases and the association with dis-ease outcome. Ann Surg. 2006;244:113–20.

32. Crazzolara R, Kreczy A, Mann G, et al. High expression of thechemokine receptor CXCR4 predicts extramedullary organ infiltra-tion in childhood acute lymphoblastic leukaemia. Br J Haematol.2001;115:545–53.

33. Javor J, Bucova M, Cervenova O, et al. Genetic variations ofinterleukin-8, CXCR1 and CXCR2 genes and risk ofacutepyelonephritis in children. Int J Immunogenet. 2012;39:338–45.

34. Duymaz-Tozkir J, Yilmaz V, Uyar FA, et al. Polymorphisms of theIL-8 and CXCR2 genes are not associated with Behçet’s disease. JRheumatol. 2005;32:93–7.

35. Snoussi K, Mahfoudh W, Bouaouina N, et al. Combinedeffects of IL-8 and CXCR2 gene polymorphisms on breast cancersusceptibility and aggressiveness. BMC Cancer. 2010;12:10–283.

36. Salim PH, Jobim M, Bredemeier M, et al. Combined effects ofCXCL8 and CXCR2 gene polymorphisms on susceptibility to sys-temic sclerosis. Cytokine. 2012;60:473–7.

37. Ben-Baruch A. Inflammatory cells, cytokines and chemokines inbreast cancer progression: reciprocal tumor-microenvironment inter-actions. Breast Cancer Res. 2003;5:31–6.

Tumor Biol.