Prostate cancer is one of the most common types of cancer and is the sixth leading cause of cancer-related deaths among Japanese men [1]. It is now a serious public health issue in Japan, with the death rate increasing over the past two decades [1]. In most cases, death from prostate cancer results from metastatic disease. Understanding the mechanisms underlying its progression will facilitate the development of biomarkers and novel therapeutic strategies to control this devastating malignancy.

A strong association exists between states of chronic inflammation and cancer, and it is believed that mediators of inflammation may be responsible for this phenomenon [2]. Chronic inflammation can lead to carcinogenesis by damaging DNA through radical oxygen and nitrogen species, enhancing cell proliferation, and stimulating angiogenesis [3]. Moreover, some single nucleotide polymorphisms (SNPs) in cytokine genes have been proven to influence protein expression and/or activity, which may predispose affected individuals to develop particular cancers [4–6].

Cyclooxygenase (COX) is involved in several inflammatory pathways, and the two COX isoforms, COX1 and COX2, differ both in their regulation and tissue distribution. COX2, or prostaglandin-endoperoxide synthase 2 (PTGS2), is an enzyme inducible by cytokines, among other factors, that catalyzes the conversion of arachidonic acid to prostaglandins in response to inflammatory stimuli [7]. It is also involved in ovulation. Prostaglandins may be involved in the pathogenesis of cancers and are induced in many malignant tissues through overexpression of COX2 [8], which also cause hyperproliferation, transformation, tumor growth, invasion, and metastasis, thus potentially increasing the risk of prostate cancer. Several studies showed that COX2 is overexpressed in prostate cancer tissue compared with benign tissue from the same patient, while in vitro COX2 overexpression inhibits apoptosis and induces tumor angiogenesis [9].

The association between COX2 SNPs and prostate cancer susceptibility has previously been examined in Western [10–12] and African populations [13], but never in Japanese individuals. The COX2 gene on chromosome 1q25.2–q25.3 is a candidate gene for prostate cancer susceptibility because of its proximity to the first mapped locus for this on chromosome 1q24–q25 (hereditary prostate cancer 1, HPC1), recognized in a 1996 genetic linkage study [14].

Emerging evidence for a role for COX2 in carcinogenesis prompted us to investigate the relationship between different alleles of this gene and prostate cancer. We therefore aimed to determine the genotypic frequency of the COX2 G1195A polymorphism and its association with prostate cancer susceptibility. We also analyzed the relationship between COX2 polymorphisms and clinicopathological characteristics, such as the Gleason grade and prostate-specific antigen (PSA) grade. To the best of our knowledge, this is the first study to evaluate the contribution of COX2 polymorphisms to prostate oncology in a Japanese population.

Materials and methods

Study participants

The study consisted of a total of 220 Japanese men, including 134 histologically confirmed cases of prostate cancer and 86 healthy age-, ethnicity-, and smoking status-matched controls. The patients with prostate cancer were treated at the Department of Urology, Miyazaki Medical University Hospital (Miyazaki, Japan) and its related hospitals between August 2011 and October 2012. Tumor grade was evaluated using the Gleason scoring system. Controls were selected randomly from healthy individuals with no history of prostate cancer or any other cancers. All controls received a PSA test to detect occult prostate cancer. All participants were informed of the details, procedures, and objectives of this study. During the study period, critical information such as age and smoking status was collected from the participants using a standardized questionnaire. This study was approved by the Ethics Committee of Miyazaki Medical University and related hospitals.

Assessment of smoking status

Information on demographics, smoking history, family history of cancer, and medical history was collected during the interview. Participants were asked about their smoking status and were classified as “smokers” or “nonsmokers.” Those who had never smoked or had smoked only a few packs of cigarettes during their lifetime were defined as nonsmokers. Those who had smoked cigarettes regularly for more than 20 years were defined as smokers. Both cases and controls were subjected to similar protocols/questionnaires by the same interviewer.

COX2 genotyping

Genomic DNA was extracted from peripheral blood leukocytes using the DNA Extractor WB Kit (Wako Pure Chemical Industries Ltd., Osaka, Japan) according to the manufacturer’s instructions and eluted with 100 μl TE buffer (Nacalai Tesque, Tokyo, Japan). COX2 G1195A genotypes were determined using the polymerase chain reaction (PCR)-based restriction fragment length polymorphism assay, as described previously [15, 16]. The following primers were used: COX2 G1195A, forward, 5'-CCCTGAGCACTACCCATGAT-3′, and reverse, 5′-GCCCTTCATAGGAGATACTGG-3′. Reactions were performed using KAPA Taq PCR Kits (Nippon Genetics, Tokyo, Japan) in a thermal cycler (TaKaRa PCR Thermal Cycler Dice; Takara, Tokyo, Japan).

PCR conditions consisted of 1 cycle at 95 °C for 5 min and then 40 cycles of 95 °C for 30 s, 61 °C for 40 s, and 72 °C for 45 s, with a final extension at 72 °C for 10 min. The PCR products were digested overnight with PvuII restriction enzyme (Takara, Tokyo, Japan) at 37 °C, and the fragments were separated on a 3 % agarose gel. A 273-bp band corresponded to COX2 AA; 273-, 220-, and 53-bp bands represented heterozygous individuals, and bands of 220- and 53-bp corresponded to the homozygous GG genotype (Fig. 1).

Fig. 1
figure 1

Polymerase chain reaction-based restriction analysis of the G1195A (rs689465) polymorphism of the COX2 gene by 3 % agarose gel electrophoresis. Marker is 20 bp DNA size

Full size image

Statistical analysis

Statistical analysis was performed using the R i386 2.15.1 software package (Vienna University of Economics and Business Administration, Vienna, Austria). The significance of differences in COX2 G1195A genotypes between cases and controls was determined by Pearson’s χ 2 tests. Probability values of <0.05 were regarded as statistically significant. Odds ratios (ORs) and 95 % confidence intervals (CIs) for prostate cancer were calculated by multivariate logistic regression analysis after adjusting for several confounding variables such as age and smoking status.

Results

Clinicopathological features are shown in Table 1. The mean age of cases and controls was 68.3 ± 7.4 (range, 61–75) years and 66.9 ± 8.3 (range, 59–75) years, respectively. There were no significant differences between prostate cancer cases and controls in terms of mean age distribution (p = 0.21, not significant (NS)) or relative frequencies of smokers and nonsmokers (p = 0.24, NS).

Table 1 Clinicopathological features
Full size table

COX2 G1195A genotypic and allelic frequencies are shown in Table 2. The genotype and allele frequencies were in Hardy–Weinberg equilibrium. Genotypic distributions (p = 0.028) and allelic frequencies (p = 0.014) differed significantly between prostate cancer and control groups in terms of the COX2 G1195A polymorphisms. The frequency of A allele carriers was higher in cases than in controls. Logistic regression analysis of outcomes (adjusted for age at diagnosis and smoking status) showed that the AA genotype was associated with increased susceptibility to prostate cancer (OR = 3.15, 95 % CI = 1.27–8.08, p = 0.014). Subset analysis to investigate possible associations between COX2 polymorphisms and clinicopathological characteristics such as Gleason grade and PSA grade revealed no significant associations.

Table 2 Distribution of COX2 G1195A (rs689465) polymorphism among patients with prostate cancer and controls
Full size table

COX2 genotype and risk associated with Gleason grade

COX2 genotypes were further analyzed for a risk associated with less aggressive or highly aggressive disease, based on Gleason grade. Patients were then categorized into three groups based on this combined score (Gleason score of ≤6, low grade; Gleason score 7, intermediate grade; Gleason score of ≥8, high grade). The results demonstrated no significant associations between genotype and Gleason grade (Table 3).

Table 3 Association between COX2 G1195A (rs689465) polymorphism and clinicopathological characteristics, including Gleason grade and PSA grade
Full size table

COX2 genotype and risk associated with PSA grade

Patients were categorized into three groups based on PSA value (PSA of <10.0 ng/ml, low grade; PSA of ≥10 and <20.0 ng/ml, intermediate grade; PSA of ≥20.0 ng/ml, high grade). No significant associations were found between COX2 polymorphisms and PSA grade in patients with prostate cancer (Table 3).

Discussion

In recent years, interest in cancer genetic susceptibility has increased, with attention being focused on the study of SNPs in genes associated in carcinogenesis as they may contribute to individual susceptibility to cancer [17]. We consider COX2 located in the HPC1 region 1q25 [14] to be of particular interest in the search for prostate cancer susceptibility genes. COX2 is inducible by a variety of factors such as cytokines, growth factors, and tumor promoters and is highly expressed in several human cancers and cancer cell lines, including prostate cancer [18].

Associations of COX2 polymorphisms with prostate cancer indicate that inflammation is an important factor in the development of this cancer. However, to date, these SNPs have not been extensively explored. We selected the COX2 G1195A polymorphism from the National Center for Biotechnology Information website and found a significant association of the A variant with higher susceptibility for prostate cancer in our Japanese population.

Similar results were previously reported by Zhang et al. and Gao et al. [19, 20], who showed that the G1195A A variant was significantly associated with esophageal squamous cell carcinoma and breast cancer, respectively, in Chinese populations. Similarly, in a case–control study of 377 oral squamous cell carcinoma (OSCC) patients and 442 controls from Taiwan, AA homozygotes were shown to have a potential genetic risk of developing OSCC. The strongest risk was observed in AA homozygote betel chewers, while AG and AA genotype non-betel chewers who smoked and consumed alcohol had increased risks of 15.1- and 32.1-fold, respectively [21]. However, the GA and AA genotypes showed no significant association with head and neck cancer risk in a Dutch population [22]. This could be explained by the fact that COX2 polymorphisms vary in different tumors and between different ethnic groups. Moreover, the presence of COX2 GA or AA polymorphisms was reported to create a c-myb binding site and to enhance transcriptional activity of the COX2 gene [19].

Recently, several studies demonstrated that COX2 variants were associated with the risk of prostate cancer [11, 12, 23, 24]. In a Swedish population, two COX2 variants, C3100T (rs689470) and C8365T (rs2043), were associated with a reduced risk of prostate cancer [12]. Another study focusing on advanced prostate cancer patients in African-Americans and European Americans showed that three of the nine SNPs examined demonstrated significant associations with prostate cancer risk, with the most compelling polymorphism, rs2745557, associated with a lower risk of disease [11]. Finally, Wu et al. reported that the G allele of COX2 promoter G765C appears to be associated with prostate cancer development and may be a useful biomarker for its early detection [23].

In summary, our results are the first to indicate that prostate cancer susceptibility and risk are influenced by a common COX2 genetic polymorphism, with the COX2 G1195A A allele shown to increase the risk of developing prostate cancer. They are also the first to suggest that the COX2 G1195A polymorphism plays an important role in prostate cancer susceptibility in the Japanese population. However, we found no significant associations between this polymorphism and clinicopathological characteristics associated with prostate cancer progression. This apparent discrepancy could be attributed to the small number of cases (n = 134) in this pilot study, so further studies of larger sample sizes are needed to clarify the relationship between the G1195A polymorphism and clinicopathological characteristics in prostate cancer. The polymorphism also merits further study as a potential genomic marker for the early detection of prostate cancer. Also, 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors, commonly known as statins, have been hypothesized to be anti-inflammatory in some studies and function synergistically with drugs containing COX2 activity such as aspirin and nonsteroidal anti-inflammatory drugs (NSAIDs) [25, 26]. Therefore, statins may prevent the prostate cancer susceptibility in the group of A variant in particular. More sophisticated gene–gene and gene–environment interactions [27], such as the effects of environmental exposure to specific carcinogens, together with genotype–phenotype correlations should also be investigated in future studies.