Endothelial nitric oxide synthase gene polymorphisms and prostate cancer risk in Serbian population

Abstract

Genome-wide association studies (GWAS) have identified over 46 SNPs associated with human prostate cancer (PCa). Some studies have shown correlation of the nitric oxide synthase (NOS) NOS3 gene polymorphisms with the risk and/or progression of PCa. This study aimed to evaluate the association of NOS3 gene polymorphisms (−786T>C, −764A>G, −714G>T, −690C>T, −649G>A and 894G>T) with PCa risk and progression. 150 patients with PCa, 150 patients with BPH and 100 age-matched healthy controls were recruited in this study. Genotyping of promoter polymorphisms was performed by bi-directional DNA sequencing, and for 894G>T by RFLP analysis. There was no significant association between the alleles and genotypes of these genetic variants and PCa risk. For −786T>C polymorphism, we found that C allele is associated with absence of metastases, assuming dominant genetic model (P = 0.049; OR, 0.50; 95% CI, 0.25–1.00). It was found that, compared with NOS3 −690C>T variant CC genotype, CT and TT genotypes confer decreased risk of developing metastases (dominant model, P = 0.015, OR, 0.24; 95% CI, 0.07–0.88) and show association with low clinical tumour stage, compared with stages T₃ and T₄ (dominant model, P = 0.046, OR, 0.20; 95% CI, 0.04–1.02). Genetic variants −764A>G, −714G>T, −649G>A were not detected in our study group. There is evidence of an inverse correlation of the NOS3 894G>T minor allele with high serum PSA (>20 ng/ml) (dominant model, P = 0.013, OR, 0.37; 95% CI, 0.17–0.82). Our results suggest that NOS3 gene polymorphisms are genetic susceptibility factors for the progression of PCa and patient outcome.

Introduction

Prostate cancer (PCa) is a disease with considerable heterogeneity in biological aggressiveness and prognosis. Onset and progression of PCa involve accumulation of both genetic and epigenetic alterations (Konishi et al. 2005; Yu & Luo 2007). The incidence of PCa has a strong age, ethnic origin and geographical dependence (Hällström & Laiho 2008). Almost 899,000 PCa cases and 258,000 PCa deaths are estimated to have occurred in 2008 worldwide, with 72% of cases and 53% of deaths in developed countries (Center et al. 2012). Prostate cancer in Serbian population shows increasing trend of newly discovered cases, from 662 in 1999 to 1.673 in 2009 (Vrdoljak et al. 2011; Cancer Registry of Central Serbia, Institute of Public Health of the Republic of Serbia 1999–2009). In the past decade, numerous studies analysed association of SNPs with the risk and/or progression of malignant diseases (Goh et al. 2012).

Nitric oxide (NO) is a pleiotropic molecule critical to a number of physiological and pathological processes (Mocellin et al. 2007). Nitric oxide synthases (NOSs) are a family of enzymes that catalyse the production of NO from L-arginine and L-citrulline amino acids (Forstermann et al. 1998). Different members of the NOS family are encoded by separate genes. There are three known isoforms, two are constitutive (NOS1, NOS3) and the third is inducible (NOS2). NOS3 gene is located at 7q35–q36, and it is expressed in endothelial cells. The endothelial isoform is the primary signal generator in the control of vascular tone, insulin secretion and airway tone. Furthermore, it is involved in the regulation of cardiac function and angiogenesis (Grande et al. 2000). For the last 40 years, there is some experimental evidence showing that tumour growth and metastasis are dependent upon tumour angiogenesis (Gimbrone et al. 1972). Furthermore, there is evidence that tumour angiogenesis correlates with metastasis in invasive PCa (Weidner et al. 1993). Also, the production of endogenous NO is associated with apoptosis of tumorigenic cells (Salvucci et al. 2001; Torok et al. 2002; Wartenberg et al. 2003; Kim & Tannenbaum 2004).

Several single nucleotide polymorphisms (SNPs) have been identified in the NOS3 gene. Two common polymorphisms are −786T>C (rs2070744) and 894G>T (rs1799983), but there are also reported −1468T>A, −922G>A, −764A>G, −714G>T, −690C>T, −649G>A and 4a4b polymorphism (27-bp repeat, intron 4) (Wang & Wang 2000; Marangoni et al. 2008; Lee et al. 2009). These polymorphisms have a possible role in the patogenesis of stroke (Howard et al. 2005), high blood pressure (Cruz-González et al. 2009; Niu & Qi 2011), decreased insulin sensitivity (Yoshimura et al. 2003), diabetic nephropathy (He et al. 2011), non-arteritic anterior ischaemic optic neuropathy (Sakai et al. 2007), coronary spasm (Nakayama et al. 1999), erectile disfunction (Rosas-Vargas et al. 2004; Sinici et al. 2010), male infertility (Safarinejad et al. 2010) and carcinogenesis (Lala & Orucevic 1998).

The results of association studies have shown that several polymorphisms in NOS3 gene are associated with PCa (Lee et al. 2009). A common single nucleotide genetic variant located in exon 7 (894G>T) of NOS3 gene, results in an amino acidic substitution at position 298 (Glu298Asp), which is implicated in low NOS3 level due to reduced protein stability (Tesauro et al. 2000). The wild type genotype at position 894 was identified as a NO-related genetic factor predictive of advanced disease and bone metastasis (Medeiros et al. 2002).

Earlier reports showed that the incorporation of the C allele at position −786 creates a binding site for replication protein A1 (RPA1) (Miyamoto et al. 2000) and reduces promoter activity (Nakayama et al. 1999). At the same time, the incorporation of the C allele is associated with increased levels of NOS3 transcripts in the peripheral blood of presurgical samples from patients with PCa. These results identify −786T>C polymorphism as the most important promoter alteration of the NOS3 gene that may affect the PCa progression, but not its occurrence (Marangoni et al. 2008).

Therefore, in the present study, we examined the potential association of 894G>T and promoter polymorphisms of the NOS3 gene with prostate cancer in Serbian population.

Methods

Patients

The NOS3 polymorphisms were genotyped in 303 consecutively enrolled patients, of which 150 patients with sporadic PCa (mean age: 69.8 years; range: 45–96 years) and 150 patients with benign prostate hyperplasia (BPH) (mean age: 68.7 years; range: 33–85 years) who were treated at Clinical Centre ‘dr Dragiša Mišović’, Belgrade, Serbia, in the period from 2009 to 2011. Three cases were excluded because of the lack of biopsy results. All patients were diagnosed with PCa confirmed by histopathological examination of specimens obtained by transrectal ultrasound (TRUS) biopsies, transurethral resection of the prostate and radical prostatectomy (RP). Grading was established according to the Gleason score (GS) differentiation system (Gleason & Mellinger 1974). Blood prostate-specific antigen (PSA) levels were determined using Hybritech monoclonal immunoassay (Beckman Hybritech assay; Beckman Coulter, Inc., Fullerton, CA, USA), with cut-off value of 4.0 ng/μl (Catalona et al. 1994). One hundred healthy volunteers (mean age: 67.8 years; range: 58–83 years) with normal PSA, normal DRE and no previous personal and family history of PCa nor BPH comprised the control group. Controls were recruited after passing standard annual physical examination. This study was approved by the Ethics Committee of Clinical Centre ‘dr Dragiša Mišović’, Belgrade, Serbia.

The risk of progression was determined using two classification systems: one proposed by D'Amico et al. 1998 and the other by Medeiros et al. 2002. According to D'Amico et al. recommendations, PCa patients were divided into three groups with low risk (PCa patients with PSA ≤10 ng/ml, clinical stage ≤ T2a, and Gleason score ≤6), medium risk (PCa patients with PSA from 10–20 ng/ml, or clinical stage T2b-c, or Gleason score 7) and high risk (PCa patients with PSA >20 ng/ml, or clinical stage ≥ T3, or Gleason score ≥8) (D'Amico et al. 1998). Following the instructions of Medeiros et al., patients were stratified into two groups: with high risk (Gleason ≥7, or advanced clinical stage (T3 and T4, or presence of bone metastasis)) and low risk of cancer progression (low grade, early stage and absence of bone metastasis).

Peripheral blood samples were collected in vacutainer tubes containing Na-citrate and maintained at 4 °C. All samples were obtained with the informed consent of participants before their inclusion in the study.

Methods

Genomic DNA was extracted from 200 μl of peripheral blood using the QIAamp DNA blood mini kit (Qiagen, Hilden, Germany) following the supplier's instructions.

Amplification and genotyping of the promoter polymorphisms

The presence of the −786T>C, −764A>G, −714G>T, −690C>T and −649G>A polymorphisms in the 5′-flanking region of the NOS3 gene was determined by PCR amplification with the primers 5′-ATG CTG CCA CCA GGG CAT CA-3′ and 3′-GTC CTT GAC TCT GAC ATT AGG G-5′ (Nakayama et al. 1999). A volume of 50 μl was used for each PCR reaction, which contained five ρmoles of both primers, 200 μM of each dNTP (deoxyribonucleotide triphosphate), 1.5 mmol/l MgCl₂ and 1.5 U Taq DNA polymerase (AmpliTaq Gold; Roche, Basel, Switzerland) with 5 μl of genomic DNA (about 10–20 ng DNA). Temperature profile was as follows: 3 min at 97 °C, 35 cycles of 95 °C for 60 s, 51 °C for 60 s, and 72 °C for 60 s and final extension 10 min at 72 °C. The amplified fragments were separated by electrophoresis in 1.5% agarose gel with ethidium bromide and purified using QIAquick PCR Purification Kit (QIAquick PCR Purification Kit; Qiagen, Hilden, Germany). Purified PCR samples were used in sequencing reaction using BigDyeTerm v1.1 CycleSeq Kit (Applied Biosystems, Foster City, CA, USA) with reverse primer. Afterwards, the products of sequencing reaction were purified using EDTA/ethanol purification (Wallis & Morrell 2011) and analysed by capillary gel electrophoresis in genetic analyzer (ABI PRISM 3100 Genetic Analyzer, Applied Biosystems).

RFLP analysis for NOS3 894G>T polymorphism

Genotypization of NOS3 894G>T polymorphism was performed by RFLP analysis with forward and reverse primers (5^′-CAT GAG GCT CAG CCC CAG AAC-3^′ and 5^′-AGT CAA TCC CTT TGG TGC TCA C-3^′) (Wilcox et al. 1997). A volume of 25 μl for PCR reactions contained 0.2 mM of both primers, 200 μM of each dNTP and 1 U Taq DNA polymerase (Taq polymerase, Kappa) with 5 μl of genomic DNA. After initial denaturation at 97 °C for 3 min PCR reactions were run for 35 cycles: 95 °C for 60 s, 54 °C for 60 s, and 72 °C for 60 s; and final extension 10 min at 72 °C.

The products of PCR amplification were incubated at 37 °C overnight with 1 U of the restriction enzyme MboI (Fermentas). Digested fragments' lengths were 119 bp and 87 bp for TT genotype. The restricted fragments were separated on 3% agarose gels with ethidium bromide.

Ten randomly selected samples were analysed by capillary gel electrophoresis as a control of RFLP analysis.

Statistical analyses

Analyses of data were performed using statistical software spss for Windows (Version 17.0), SNPStats software (Catalan Institute of Oncology, Barcelona, Spain) and plink statistical software (Purcell et al. 2007). Logistic regression analysis was used to compare categorical variables. A 5% level of significance was used in the analysis. Exact test implemented in plink was used to determine whether experimental observations departed from Hardy–Weinberg equilibrium (Purcell et al. 2007). For each SNP, potential dominant effects were evaluated by combining homozygous and heterozygous minor allele carriers for comparison with the reference group. The odds ratio (OR) and its 95% confidence interval (CI) were calculated as the measure of the association of NOS3 polymorphisms genotypes with PCa risk. Results were adjusted for the age confounder.

Results

Patients characteristics

This study was performed on the groups of 150 PCa patients, 150 BPH patients and 100 healthy control subjects. The clinical and histopathological characteristics of study groups are shown in Table 1.

Table 1.

The clinical and histopahological characteristics

Characteristics		PCa patients	BPH patients	Controls
Number		150	150	100
Age		69.8 (45–96)	68.7 (33–85)	67.8 (58–83)
Serum PSA (ng/ml)	0–10	49	124	100
	10–20	39	20	–
	>20	62	6	–
Tumour stage	T1	28	–	–
	T2	72	–	–
	T3, T4	50	–	–
Gleason score	<7	71	–	–
	=7	56	–	–
	>7	22	–	–
Metastasis	Absence	95	–	–
	Presence	55	–	–
Risk of progression (D'Amico et al. 1998)	Low	81	–	–
	Medium	55	–	–
	High	14	–	–
Risk of progression (Medeiros et al.2002)	Low	55	–	–
	High	95	–	–

Genotyping

The distributions of NOS3 894G>T, −786T>C and −690C>T genotypes among cases of patients with PCa, BPH and controls are shown in Table 2. These distributions were consistent with Hardy–Weinberg equilibrium. There was no significant association between the alleles and genotypes of these genetic variants and PCa risk.

Table 2.

Distribution of the genotype of NOS3 894G>T, −786T>C and −690C>T polymorphisms in patients with PCa, BPH and control subjects

	No. of PCa (%)	No. of BPH (%)	P-value*	OR† (95% CI)	No. of controls (%)	P-value*	OR† (95% CI)
894G>T
Genotype
GG	76 (50.67)	78 (52)	0.62	1 (referent)	54 (54)	0.60	1 (referent)
GT	65 (43.33)	59 (39.33)		1.13 (0.70–1.81)	40 (40)		1.15 (0.68–1.95)
TT	9 (6)	13 (8.67)	0.46	1.71 (0.29–1.76)	6 (6)	0.89	1.08 (0.36–3.23)
Alleles
G	217 (72.33)	215 (71.67)	0.85	0.97 (0.67–1.39)	148 (74.00)	0.65	1.10 (0.72–1.68)
T	83 (27.67)	85 (28.33)			52 (26.00)
−786T>C
Genotype
TT	54 (36)	57 (38)	0.95	1 (referent)	34 (34)	0.69	1 (referent)
TC	68 (45.33)	73 (48.66)		0.98 (0.60–1.62)	51 (51)		0.89 (0.50–1.57)
CC	28 (18.67)	20 (13.33)	0.23	1.53 (0.77–3.05)	15 (15)	0.60	1.22 (0.57–2.64)
Alleles
T	176 (58.67)	187 (62.33)	0.36	1.17 (0.84–1.61)	119 (59.50)	0.74	1.06 (0.74–1.53)
C	124 (41.33)	113 (37.67)			81 (40.50)
−690C>T
Genotype
CC	130 (86.67)	130 (86.67)	0.89	1 (referent)	85 (85)	0.87	1 (referent)
CT	19 (12.67)	20 (13.33)		0.96 (0.49–1.88)	14 (14)		0.94 (0.44–1.99)
TT	1 (0.66)	0 (0)	0.24	NA (0.00–NA)	1 (1)	0.73	0.61 (0.04–9.91)
Alleles
C	279 (93.00)	280 (93.33)	0.86	1.06 (0.56–2.02)	184 (92.00)	0.77	1.15 (0.46–1.77)
T	21 (7.00)	20 (6.33)			16 (8.00)

^*Logistic regression for the difference in genotype distributions and allelic frequencies between the PCa and BPH patients/controls, adjusted for age.

^†Odds Ratios adjusted for age. P-values less than 0.05 were considered significant.

Next, we compared the minor allele frequencies in the probands with the values of standard prognostic parameters for PCa progression. Table 3 displays the NOS3 894G>T, −786T>C and −690C>T genotype distribution towards standard prognostic parameters and the risk of progression in PCa patients. There is evidence of an inverse correlation of the NOS3 894G>T minor allele with high serum PSA (>20 ng/ml) (dominant model, P = 0.013, OR, 0.37; 95% CI, 0.17–0.82).

Table 3.

Association of NOS3 894G>T, −786T>C and −690C>T polymorphisms with values of standard prognostic parameters and the risk of PCa progression

	894G>T genotype				−786T>C genotype				−690C>T genotype
	GG	GT+TT	P-value*	OR (95% CI)*	TT	TC+CC	P-value*	OR (95% CI)*	CC	CT+TT	P-value*	OR (95% CI)*
Age, mean ± SD	68.9 ± 8	70.6 ± 8.1			70.5 ± 7.9	69 ± 8.3			70 ± 8.3	67.6 ± 9.1
Serum PSA (ng/μl)
<10	19	30	0.40	1 (referent)	16	33	0.96	1 (referent)	40	9	0.63	1 (referent)
10–20	19	20	0.69 (0.29–1.64)	12	27	0.98 (0.39–2.47)	33	6	0.76 (0.24–2.40)
>20	38	24	0.013	0.37 (0.17–0.82)†	26	36	0.34	0.68 (0.31–1.50)	57	5	0.11	0.40 (0.12–1.28)
Tumour stage
T1	13	15	0.94	1 (referent)	10	18	0.69	1 (referent)	23	5	0.78	1 (referent)
T2	33	39	1.03 (0.43–2.49)	22	50	1.21 (0.48–3.06)	60	12	0.85 (0.26–2.73)
T3, T4	30	20	0.32	0.62 (0.24–1.60)	22	28	0.37	0.64 (0.24–1.71)	47	3	0.046	0.20 (0.04–1.02)
Gleason score
<7	33	38	0.50	1 (referent)	26	45	0.72	1 (referent)	61	10	0.82	1 (referent)
=7	30	27	0.79 (0.39–1.59)	19	38	1.14 (0.55–2.38)	48	9	1.12 (0.42–3.00)
>7	13	9	0.17	0.50 (0.18–1.38)	9	13	0.84	0.90 (0.33–2.46)	21	1	0.22	0.30 (0.04–2.63)
Metastasis
Absence	45	50		0.71 (0.36–1.40)	29	66		0.50 (0.25–1.00)	78	17		0.24 (0.07–0.88)
Presence	31	24	0.32		25	30	0.049		52	3	0.015
Risk of progression (D'Amico et al.)
Low	5	9	0.51	1 (referent)	5	9	0.5	1 (referent)	12	2	0.46	1 (referent)
Medium	25	30	0.67 (0.20–2.26)	14	41	1.56 (0.44–5.52)	42	13	1.79 (0.35–9.13)
High	46	35	0.14	0.42 (0.13–1.37)	35	46	0.6	0.73 (0.22–2.37)	76	5	0.33	0.40 (0.07–2.28)
Risk of progression (Medeiros et al.)
Low	24	31		0.64 (0.33–1.26)	18	37		0.79 (0.39–1.59)	45	10		0.52 (0.20–1.34)
High	52	43	0.2		36	59	0.50		85	10	0.18

^*Adjusted for age.

^†Statistically significant results are shown in bold.

For −786T>C polymorphism, we found that carriers of minor allele have 50% reduced risk of developing metastases (dominant model, P = 0.049; OR, 0.50; 95% CI, 0.25–1.00).

Furthermore, a statistically significant difference was noted in the −690 C>T genotype distribution between patients with and without metastases (dominant model, P = 0.015, OR, 0.24; 95% CI, 0.07–0.88). We also observed that combined genotypes CT and TT confer the reduced risk of high tumour stage (T₃, T₄) (dominant model, P = 0.046, OR, 0.20; 95% CI, 0.04–1.02).

Other promoter polymorphisms (−764A>G, −714G>T, −649G>A) are found to be monomorphic in Serbian population.

Capillary gel electrophoresis of ten per cent of randomly selected samples confirmed the results of RFLP analysis.

Discussion

Molecularly, PCa cells carry multiple genetic and epigenetic alterations that generate malignant phenotype capable of uncontrolled growth, avoiding apoptosis and invasion – metastasis to other organs (Dasgupta et al. 2012). Recently, genome-wide association studies (GWAS) have identified over 46 SNPs associated with human PCa (Taylor et al. 2010; Goh et al. 2012).

Earlier report suggests that NO may play opposing roles in tumour growth and metastasis (Lala & Orucevic 1998). Alonso et al. reported that NOS3 mRNA levels from patients' blood show significant differences between PCa and BPH groups, with an occurrence chance for PCa 5.8-fold higher than BPH disease (Alonso et al. 2009). Furthermore more, Marangoni et al. have shown that the incorporation of the −786T>C minor allele is associated with the increased levels of NOS3 transcripts (Marangoni et al. 2008).

The first study involving −786 T>C polymorphism showed that this promoter alteration may affect the PCa progression, but not its occurrence (Marangoni et al. 2008). More recent study reported a significant difference in both genotype distribution and allele frequency between PCa patients and healthy controls, showing that patients with NOS3 TC and CC genotype have a 2.43-fold and 3.62-fold increased risk of PCa, respectively (Safarinejad et al. 2012). Similarly to the results of Marangoni et al., our study shows no association between the incorporation of −786T>C minor allele and the increased risk of PCa, but unlike in their study, there is an association of this allele with the decreased risk of metastasis.

We have also shown that −690C>T minor allelic variant arises only in presence of −786T>C minor allele. This is in concordance with the earlier studies (Marangoni et al. 2008). Also, we found a significant difference in −690C>T genotype distributions between patients with high tumour stage (P = 0.046), as well as between patients with and without metastasis, assuming dominant genetic model (P = 0.015).

Along with promoter polymorphisms, our study involved the polymorphism 894G>T in exon 7. One of the previous studies demonstrated strong association between 894G>T GG genotype and advanced disease with bone metastases (Medeiros et al. 2002). Our results showed that 894G>T minor allele is in negative correlation with the increased serum PSA (>20 ng/ml) (P = 0.013). 894G>T substitution results in an amino acid alteration, glutamic to aspartic acid, which leads to lower protein level (Senthil et al. 2005). A study involving NOS3 polymorphisms in coronary artery disease reported that plasma NO level significantly depends on the genotypes of the 894G>T polymorphism; plasma NO was increased in those individuals with 894T allelic variant, but only in the control group (Yoon et al. 2000). On the other hand, study on human umbilical vein endothelial cell culture (HUVAC) reported that rare allele TT genotype is associated with low protein level (Senthil et al. 2005). They argued that 894T allelic variant in exon 7 could also affect bioavailable NOS3 by reducing protein stability. This argument was based on earlier study that had observed, beside regular 135 kDa band, nonspecific 100 kDa band in cell lysates from three primary human endothelial cell lines, one with the 894TT genotype and two with the 894GT genotype, but not in the one with the 894GG genotype (Tesauro et al. 2000). Unlike the earlier results (Medeiros et al. 2002, 2002; Marangoni et al. 2006), and ours as well, a recent study showed no association with PCa incidence nor PCa clinical and pathological features (Safarinejad et al. 2012). One explanation for the differences that have been observed is that they arise from the different ethnic backgrounds of the study populations. Thus earlier studies included patients from Portugal (Medeiros et al. 2002, 2002) and Brasil (Marangoni et al. 2006), whereas our study was performed on patients from Serbia. A further study, which also showed with opposing results (Safarinejad et al. 2012) was conducted on an Iranian population.

In summary, the analysed polymorphisms are not associated with PCa risk. 894T allele is inversely correlated with increased serum PSA. −786T>C and −690C>T minor allele carriers were found to have the reduced risk of developing metastases. Also, the obtained results show negative correlation of −690C>T combined CT and TT genotypes with high tumour stage.

We are fully aware that the present study is not without limitations. Because of the small subgroups of PCa patients, conclusions drawn from these subanalyses must be interpreted with caution. Since the controls were volunteers, they may not be representative of the general population. Finally, as this study is restricted to one particular population group, the Serbian population, it is not clear that these results can be generalized to other populations, since there may be important ethnic and cultural variations that need to be taken into account.

Conclusion

Our results indicate that NOS3 894G>T, −786T>C and −690C>T genetic polymorphisms are not associated with prostate cancer risk in Serbian population but may be relevant as prognostic factors for the progression of prostate cancer and patients' outcome.