BER gene polymorphisms associated with key molecular events in bladder cancer

Smal M.P.*1, Kuzhir T.D.1, Savina N.V.1, Nikitchenko N.V.1, Rolevich A.I.2, Krasny S.A.2, Goncharova R.I.1

Summary. Aim: Base excision repair (BER) gene polymorphisms are known to play an independent role in predisposition to developing different cancers as well as to be associated with clinicopathological traits of the disease modifying its clinical outcomes. One of the underlying mechanisms is presumed to include interplay between BER gene polymorphisms and key mutational, epigenetic and chromosomal events in tumor tissues. The present study was aimed at elucidating potential gene-gene interaction and assessing their mutual effects in bladder cancer (BC). Materials and Methods: The earlier obtained data on genotyping patients with verified diagnosis of BC for OGG1 rs1052133 (Ser326Cys) and XRCC1 rs25487 (Arg399Gln) polymorphisms were used for this study. The tumor tissue samples from the same patients were analyzed for mutations, epigenetic variations and losses of heterozygosity in some key genes involved in divergent pathogenic pathways of BC. Results: It was shown that the OGG1 (326 codon) heterozygous genotype as well as the minor 326Cys allele can intensify a mutational response of the RAS locus in urothelial carcinomas in the total cohort of patients simultaneously decreasing the mutation rates in the PIK3CA locus in smokers. The XRCC1 (399 codon) heterozygous genotype as well as the minor 399Gln allele reduced the frequency of LOH in the PTEN and TNKS genes, but did not affect the mutational variability in any locus tested. Both polymorphisms influenced the methylation status, carriers of OGG1 326Ser/Cys or Ser/Cys+Cys/Cys genotypes demonstrating increased frequency of methylated RUNX3 and ISL1 genes whereas the similar effect of XRCC1 polymorphism concerning methylation of p16 and TIMP3 genes. When dividing the total cohort into groups based on the extent of tumor spread, the observed associations were characteristic of non-muscle invasive BC. Conclusion: The BER gene polymorphisms contributed to modification of key molecular events in urothelial carcinomas. Their mutual effects mainly manifested in non-muscle invasive BC. The underlying mechanisms as well as possible clinical outcomes need to be further explored to propose novel prognostic biomarkers for BC.

Submitted: August 06, 2018.
*Correspondence: E-mail: marharyta.smal@gmail.com
Abbreviations used: BC — bladder cancer; BER — base excision repair; LOH — loss of heterozygosity; MIBC — muscle invasive bladder cancer; NMIBC — non-muscle invasive bladder cancer; SNP — single nucleotide polymorphism; SSCP — single strand conformation polymorphism.

Bladder cancer (BC) is one of the most common malignancies of the urinary tract. Its incidence rate is 12.8 cases per 100,000 population in Belarus per year [1]. According to epidemiologic data, this cancer prevails among men and is strongly associated with age [2, 3]. The risk factors for BC include occupational chemical exposures and tobacco smoking, the latter causes almost half the cases in men [2]. Mutagen-induced DNA lesions initiate genome instability that, as a possible outcome, may result in cell malignancy. DNA repair pathways controlled by over than 130 genes maintain genome integrity and are active in the cellular responses to DNA damage [4]. In the context of cancer prevention, excision repair systems are of principal importance, as they act in a “cut-and-patch” manner, excising the damaged segment of DNA and filling the single-stranded gap by using the intact complementary strand as a template [5].

The base excision repair (BER) pathway is involved in removal of damaged DNA bases and single-strand breaks, and variant BER proteins contribute to genomic instability [6]. The 8-oxo-guanine DNA glycosy­lase (OGG1) acts as a glycosylase to excise the highly mutagenic DNA lesion, 8-oxo-7,8-dihydroguanine (8-oxoGua), from DNA and then as an AP lyase nicks the nascent apurinic/apyrimidinic site by β-elimination [7, 8]. XRCC1 is the scaffold protein interacting with DNA glycosylases, AP endonuclease-1 (APE-1), DNA polymerase β (POLβ), DNA ligase III (Lig III), poly(ADP-ribose) polymerase 1 (PARP-1), polynucleotide kinase (PNK) and thereby coordinating the subsequent enzymatic steps of BER [9].

Many environmental agents, including cigarette smoke constituents, induce DNA damage through a reactive oxygen species (ROS)-mediated mechanism [10]. Current findings indicate both an initiating role of the oxidatively generated DNA damage in cell malignancy and their accumulation in tumors. Therefore, the impairment of the catalytic function of the OGG1 is assumed to affect susceptibility to cancer and other oxidative pathologies [11]. Functional studies have revealed inhibition of OGG1 enzymatic activity due to OGG1 Ser326Cys polymorphism as well as modulation of the cellular genotoxic response and BER efficiency affected by certain polymorphic variants of the XRCC1 gene [11, 12]. Both OGG1 Ser326Cys (rs1052133) and XRCC1 Arg399Gln (rs25487) polymorphisms have been reported to decrease the capacity of human blood lymphocytes to repair oxidatively and radiation-induced DNA damage [13, 14] suggesting their potential impact on cancer risk. The decline in BER activity has been discussed by Markkanen et al. [15] to promote the development of sporadic cancers. Furthermore, inter-individual variability in DNA repair capacity of tumor cells may influence the removal of chemically induced DNA adducts or radiation damage to alter the clinical responses to treatment [16, 17].

It is generally accepted that the “driver” mutations in a few key genes trigger certain (sometimes alternative) pathways of cancer pathogenesis. In BC, the mutations in FGFR3 gene are strongly associated with non-muscle invasive tumors, whereas mutations in TP53 gene are characteristic of muscle invasive cancer [18]. However, the molecular analysis of tumor tissue samples from Belarusian patients has shown that more than a third of them have the wild type genotype with respect to both genes suggesting another or even multiple genetic origins of urothelial carcinomas [19]. In spite of the fact that genetic variations in BER genes are not attributed to driver mutations in BC [20], they might modulate susceptibility to cancer initiation and cancer progression due to their cooperation with such important events as mutations in various key genes, epigenetic variations, loss of heterozygosity (LOH) and so on. Findings in this field are fragmentary and concern mostly TP53 mutations associated with DNA repair and antioxidant defense gene polymorphisms [21–23]. Therefore, it would be of great interest to reveal the relationship between OGG1 and XRCC1 polymorphic variants, on the one hand, and mutations, aberrant methylation and LOH in some oncogenes and tumor suppressor genes, on the other hand, to evaluate their mutual contribution to deve­loping BC that might provide an accurate prognosis of the clinical course and outcomes. In the framework of the present study, the associations indicated were analyzed depending on patient age and lifestyle habits as well as on the extent of tumor spread.

Materials and methods

Study subjects. The group analyzed comprised 340 BC patients who were examined and treated at the Department of Urology of the N.N. Alexandrov National Cancer Centre of Belarus over 2010–2014. All urothelial carcinoma diagnoses were verified histologically. The T stages were determined using the International TNM classification, as well as the grade of tumor tissue differentiation was established according to WHO classifications of 1973 and 2004 [24, 25].

The study was approved by the Ethics Committee at the N.N. Alexandrov National Cancer Center (Republic of Belarus). Informed consent was obtained from each participant included in the study before the collection of blood and tumor samples. All participants were interviewed to complete a questionnaire covering medical, residential and occupational history as well as age, gender and the tobacco smoking status. The smoking status was summarized as “smokers” (combining current smokers and ex-smokers) or “non-smokers” (including never smoking persons). The demographic data of patients and the clinicopathological description of tumors are presented in Table 1.

Table 1. The demographic features of BC patients and clinicopathological parameters of tumors
Features Patients
n Frequency, %
Gender Males 278 81.8
Females 62 18.2
Age, years Min 31
Max 88
Mean ± SD 67.0 ± 10.9
Median 68
Smoking Smokers 228 67.1
Non-smokers 97 28.5
Not specified 15 4.4
Tumor stage Ta 70 20.6
T1 165 48.5
T2 58 17.1
T3 22 6.5
T4 25 7.4
Tumor grade 1973 G1 120 35.3
G2 153 45.0
G3 63 18.5
Not specified 4 1.2
2004 PUNLMP/Low 213 62.6
High 125 36.8
Not specified 2 0.6
Note: PUNLMP — papillary urothelial neoplasm of low malignant potential.

DNA for genotyping procedures, mutation, methy­lation and LOH analyses was extracted from blood and tumor samples using the traditional phenol-chloroform technique. Only individuals whose blood samples were genotyped for BER polymorphisms simultaneously with tumor samples tested for mutations or other key events were included in the present study. Since the inaccessibility and quality of biological material and quantity of extracted DNA did not always allow synchronous analyses, the sizes of samples varied.

SNP genotyping. Single nucleotide polymorphisms (SNPs) in DNA repair genes (OGG1 rs1052133 and XRCC1 rs25487) were determined by the polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) method under conditions used in the previous works [26, 27]. The PCR products were digested with restriction enzymes, electrophoresed through 2.5% agarose gels containing ethidium bromide, and visualized under UV light.

It should be noted that OGG1 rs1052133 is traditionally designated as Ser326Cys, its (G) minor (rarer) allele encodes the cysteine (Cys). XRCC1 rs25487 is also known as Arg399Gln; its (A) minor allele encodes the glutamine (Gln).

Mutation, methylation and LOH analyses. FGFR3, RAS, PIK3CA mutational status was determined using a previously described SNaPshot assay [28] that allowed screening for the most common point mutations of these genes. Regions comprising known FGFR3, HRAS, KRAS, NRAS, and PIK3CA mutations were amplified in multiplex PCR, followed by extension of internal primers for each mutation with a labeled dideoxynucleotide. Extended primers were separated by capillary electrophoresis in an automatic sequencer ABI Prism 3500 (Applied Biosystems, USA), and the presence or absence of a mutation was defined by the incorporated dideoxynucleotide.

Detection of TP53 gene molecular alterations was conducted by means of single strand conformation polymorphism (SSCP), as described elsewhere [19]. Briefly, after amplification of the regions comprising 5–8 exons, PCR-products were separated into single strands by denaturation and electrophoresed on 10% polyacrylamide gel containing 5% glycerol. Samples that showed aberrant bands by the SSCP were subjected to direct sequencing, which was performed using the Big Dye Terminator v 3.1 cycle sequencing kit (Applied Biosystems, USA) in accordance with the manufacturer’s instructions.

For DNA methylation analysis, sodium bisulfite modification of genomic DNA was carried out using EZ DNA Methylation-Gold Kit (Zymo Research, USA) according to the manufacturer’s recommendations. The methylation status of RUNX3, p16, and TIMP3 genes was assessed by methylation-specific PCR (MSP) with two sets of primers for each gene specific for either the methylated or unmethylated DNA. Primer sequences, annealing temperatures and MSP conditions were described previously [29, 30]. PCR products were loaded onto 8% nondenaturing polyacrylamide gels, stained with ethidium bromide and visualized under UV light.

Quantitative analysis of ISL1 gene methylation was performed by means of methylation-sensitive single nucleotide primer extension (Ms-SNuPE). The converted DNA samples were amplified in a bisulfite specific PCR with the following primers: forward primer 5ʹ-GGGTTTGAAAGATAAAAAGTTAGTT-3ʹ, reverse primer 5ʹ-АААССТААССТАССТТААСТАСТСС-3ʹ. After completion, PCR products were purified with a mixture of Exonuclease I and FastAP Thermosensitive Alkaline Phosphatase (Thermo Scientific, Lithuania) and subjected to Ms-SNuPE reaction which was carried out using ABI Prism SNaPshot Multiplex Kit (Applied Biosystems, USA) according to the manufacturer’s instructions. Two different internal primers were designed for quantification of methy­lation at two distinct CpGs in ISL1 gene promoter: 5ʹ-GACTGAAGGGTTTGAAAGATAAAAAGTTAGTT-3ʹ, 5ʹ-GTTAAAACAATCAAATTAAAAC-3ʹ. Ms-SNuPE primers were extended with a labeled dideoxynucleotide, dephosphorylated with FastAP and analyzed on an ABI Prism 3500 sequencer. The percentage of ISL1 methy­lation was calculated as a mean value of methylation levels of every analyzed CpG. Methylation levels were determined using the peak heights of the methylated and unmethylated allele in the GeneMapper software (Applied Biosystems, USA) [methylation% = methy­lated allele/(methylated allele + unmethylated allele)]. To analyze the association of ISL1 methylation with the SNPs, the methylation values were dichotomized with a median cut-off point amounted to 30%. The level of methylation less or more than 30% was considered as “non-methylated” and “methylated” ISL1, respectively.

LOH in RB1, PTEN, and TNKS loci was determined by microsatellite DNA analysis using the following polymorphic markers: D13S153, D10S1765, and te­tranucleotide repetition motif 5ʹ-TTGC-3ʹ (located in intron 3 of TNKS gene). The primers for D13S153 and D10S1765 were obtained from the Genome Data Bank (http://genome-euro.ucsc.edu/). For detection of TNKS allelic imbalance, 5ʹ-AGGTAGTCTTTGTGGGACTGA-3ʹ and 5ʹ-GGCAACAAAATAGGCCAAACA-3ʹ were used as forward and reverse primers, respectively. All forward primers were labeled at the 5ʹ-end with the fluorescent dyes FAM. PCR was performed according to standard procedures. The PCR products were denatured, separated on an ABI Prism 3500 sequencer and analyzed with GeneMapper software. The case was called informative when two alleles of the marker were present in the DNA extracted from blood. LOH was defined when the intensity ratio of the germline DNA peak to tumor DNA peak was 1.35:1 or greater.

Statistical analysis. The Pearson’s χ2 test (or the Fisher’s exact test when necessary) was used to verify the significance of differences between the groups of BC patients stratified by individual age, life style (smokers or non-smokers) and the extent of tumor spread. The statistical significance for deviation from Hardy — Weinberg equilibrium was determined using the χ2 test. The strength of association between SNPs and molecular alterations in tumor tissues was measured by odds ratios (OR) with 95% confidence intervals (CI). The p ≤ 0.05 values were considered significant.

Results

The impact of BER SNPs rs1052133 and rs25487 on some key cancer-related events in urothelial carcinomas in the total study cohort. The results of gene-gene interactions (with statistically significant effects) are presented in Table 2. When studying the influence of BER SNPs on mutational, epigenetic and chromosomal alterations in the total patient cohort, the following results were of inte­rest. The heterozygous OGG1 326Ser/Cys genotype as well as genotypes containing at least one minor allele was associated with a high frequency of RAS (HRAS + KRAS + NRAS) mutations (OR = 2.93, 95% CI 1.28–6.70 and OR = 2.61, 95% CI 1.14–5.99, respectively). The same OGG1 genotypes increased the levels of RUNX3 and ISL1 methylation (OR = 1.79, 95% CI 1.07–2.99 and OR = 2.70, 95% CI 1.00–2.65 for RUNX3; OR = 1.76, 95% CI 1.04–2.98 and OR = 1.82, 95% CI 1.11–3.00 for ISL1). In the latter case, statistically significant difference between frequencies of methylated and non-methylated ISL1 gene was also found in carriers of OGG1 326Cys minor allele (OR = 1.64, 95% CI 1.07–2.51).

Table 2. Key cancer-related events affected by polymorphic variants of two BER genes (rs1052133 and rs25487) in the total BC patients cohort
Gene polymorphism Genotype, allele Event, n (%) No event, n (%) p OR 95% CI
Impact on RAS mutability
OGG1
rs1052133 (C>G)
Ser326Cys
CC 11 (42.3) 136 (65.7) 0.03 0.38 0.17–0.88
CG 14 (53.8) 59 (28.5) 2.93 1.28–6.70
GG 1 (3.8) 12 (5.8) 0.65 0.08–5.21
CG + GG 15 (57.7) 71 (34.3) 0.02 2.61 1.14–5.99
C 36 (69.2) 331 (80.0) 0.07 0.56 0.30–1.07
G 16 (30.8) 83 (20.0) 1.77 0.94–3.35
Impact on RUNX3 methylation
OGG1
rs1052133(C>G)
Ser326Cys
CC 119 (62) 93 (72.7) 0.09 0.61 0.38–1.00
CG 64 (33.3) 28 (21.9) 1.79 1.07–2.99
GG 9 (4.7) 7 (5.5) 0.85 0.31–2.34
CG + GG 73 (38.0) 35 (27.3) 0.048 2.70 1.00–2.65
C 302 (78.6) 214 (83.6) 0.12 0.72 0.48–1.09
G 82 (21.4) 42 (16.4) 1.38 0.92–2.09
Impact on ISL1 methylation
OGG1
rs1052133(C>G)
Ser326Cys
CC 124 (61.7) 91 (74.6) 0.06 0.55 0.33–0.90
CG 65 (32.3) 26 (21.3) 1.76 1.04–2.98
GG 12 (6.0) 5 (4.1) 1.49 0.51–4.32
CG + GG 77 (38.3) 31 (25.4) 0.02 1.82 1.11–3.00
C 313 (77.9) 208 (85.2) 0.02 0.61 0.40–0.93
G 89 (22.1) 36 (14.8) 1.64 1.07–2.51
Impact on PTEN LOH
XRCC1
rs25487
(G>A)
Arg399Gln
GG 18 (60.0) 38 (37.6) 0.05 2.49 1.08–5.73
GA 11 (36.7) 47 (46.5) 0.67 0.29–1.54
AA 1 (3.3) 16 (15.8) 0.18 0.02–1.44
GA + AA 12 (40.0) 63 (62.4) 0.03 0.40 0.17–0.93
G 47 (78.3) 123 (60.9) 0.01 2.32 1.18–4.57
A 13 (21.7) 79 (39.1) 0.43 0.22–0.85
Impact on TNKS LOH
XRCC1
rs25487
(G>A)
Arg399Gln
GG 17 (65.4) 14 (28.0) 0.006 4.86 1.76–13.43
GA 8 (30.8) 27 (54.0) 0.38 0.14–1.03
AA 1 (3.8) 9 (18.0) 0.18 0.02–1.53
GA + AA 9 (34.6) 36 (72.0) 0.003 0.21 0.07–0.57
G 42 (80.8) 55 (55.0) 0.002 3.44 1.55–7.60
A 10 (19.2) 45 (45.0) 0.29 0.13–0.64
Impact on p16 methylation
XRCC1
rs25487
(G>A)
Arg399Gln
GG 4 (22.2) 66 (48.9) 0.06 0.30 0.09–0.95
GA 12 (66.7) 53 (39.3) 3.09 1.09–8.75
AA 2 (11.1) 16 (11.9) 0.93 0.20–4.42
GA + AA 14 (77.8) 69 (51.1) 0.043 3.35 1.05–10.69
G 20 (55.6) 185 (68.5) 0.12 0.57 0.28–1.16
A 16 (44.4) 85 (31.5) 1.74 0.86–3.53
Impact on TIMP3 methylation
XRCC1
rs25487
(G>A)
Arg399Gln
GG 4 (23.5) 66 (47.8) 0.06 0.34 0.10–1.08
GA 12 (70.6) 55 (39.9) 3.62 1.21–10.85
AA 1 (5.9) 17 (12.3) 0.44 0.06–3.57
GA + AA 13 (76.5) 72 (52.2) 0.07 2.98 0.93–9.59
G 20 (58.8) 187 (67.8) 0.3 0.68 0.33–1.41
A 14 (41.2) 89 (32.2) 1.47 0.71–3.05
Impact on p16 + TICMP3 methylation
XRCC1
rs25487
(G>A)
Arg399Gln
GG 6 (22.2) 64 (50.8) 0.005 0.28 0.10–0.73
GA 19 (70.4) 46 (36.5) 4.13 1.68–10.18
AA 2 (7.4) 16 (12.7) 0.55 0.12–2.55
GA + AA 21 (77.8) 62 (49.2) 0.01 3.61 1.37–9.55
G 31 (57.4) 174 (69.0) 0.1 0.60 0.33–1.10
A 23 (42.6) 78 (31.0) 1.66 0.91–3.02

The heterozygous XRCC1 399 Arg/Gln genotype increased the frequency of p16 and TIMP3 methy­lation (OR = 3.09, 95% CI 1.09–8.75 and OR = 3.62, 95% CI 1.21–10.85, respectively). In addition, statistically significant differences were found between frequencies of methylated and non-methylated p16 in carriers of the genotypes containing at least one XRCC1 399Gln minor allele (OR = 3.35, 95% CI 1.05–10.69). The reverse effects were revealed with respect to allelic imbalance in PTEN and TNKS loci. Homozygous wild type genotype was associated with LOH of both genes (OR = 2.49, 95% CI 1.08–5.73 and OR = 4.86, 95% CI 1.76–13.43, respectively), whereas the genotypes containing at least one minor allele or the XRCC1 399Gln allele itself decreased the frequency of these events (OR = 0.40, 95% CI 0.17–0.93 and OR = 0.43, 95% CI 0.22–0.85 for PTEN; OR = 0.21, 95% CI 0.07–0.57 and OR = 0.29, 95% CI 0.13–0.64 for TNKS).

The impact of BER SNPs rs1052133 and rs25487 on some key cancer-related events in urothelial carcinomas depending on etiological factors. The effects of gene-gene interaction were analyzed depending on such important etiological factors as patient gender, age and smoking status (Tables 3, 4). The results presented in Table 3 indicated the similar trends concerning the impact of the OGG1 polymorphism on key cancer-related events in urothelial carcinomas in men and smokers as well as in the group of BC patients under the age of 70 years. The heterozygous 326Ser/Cys genotype increased the frequencies of methylated RUNX3 and ISL1 genes in all analyzed groups, and the minor OGG1 326Cys allele possessed the same effects. As to genome mutability, stratification of the study population into separate groups and reduction of samples resulted in disappearance of the association between the OGG1 polymorphism and mutations in the RAS genes, but their association with PIK3CA mutability became apparent in smokers, the heterozygous 326Ser/Cys genotype as well as the minor OGG1 326Cys allele decreasing the mutation rate.

Table 3. Key cancer-related events affected by the OGG1 gene polymorphism rs1052133 in BC patients depending on their gender, age and smoking status
Gene polymorphism Genotype/allele Event, n (%) No event, n (%) p OR 95% CI
Impact on RUNX3 methylation in men
OGG1
rs1052133 (C>G)
Ser326Cys
CC 97 (60.6) 80 (77.7) 0.008 0.44 0.25–0.78
CG 56 (35.0) 18 (17.5) 2.54 1.39–4.65
GG 7 (4.4) 5 (4.9) 0.90 0.28–2.90
CG + GG 63 (39.4) 23 (22.3) 0.02 2.26 1.29–3.96
C 250 (78.1) 178 (86.4) 0.02 0.56 0.35–0.91
G 70 (21.9) 28 (13.6) 1.78 1.10–2.87
Impact on RUNX3 methylation in patients under the age of 70
OGG1
rs1052133 (C>G)
Ser326Cys
CC 45 (50.6) 57 (71.3) 0.02 0.41 0.22–0.78
CG 38 (42.7) 21 (26.3) 2.09 1.09–4.02
GG 6 (6.7) 2 (2.5) 2.82 0.55–14.39
CG + GG 44 (49.4) 23 (28.8) 0.006 2.42 1.28–5.59
C 128 (71.9) 135 (84.4) 0.006 0.47 0.28–0.81
G 50 (28.1) 25 (15.6) 2.11 1.23–3.61
Impact on RUNX3 methylation in current and former smokers
OGG1
rs1052133 (C>G)
Ser326Cys
CC 80 (60.2) 67 (77.0) 0.02 0.45 0.25–0.83
CG 46 (34.6) 15 (17.2) 2.54 1.31–4.92
GG 7 (5.3) 5 (5.7) 0.91 0.28–2.97
CG + GG 53 (39.8) 20 (23.0) 0.009 2.22 1.21–4.08
C 206 (77.4) 149 (85.6) 0.03 0.58 0.35–0.96
G 60 (22.6) 25 (14.4) 1.74 1.04–2.90
Impact on ISL1 methylation in men
OGG1
rs1052133 (C>G)
Ser326Cys
CC 99 (60.7) 79 (78.2) 0.01 0.43 0.24–0.76
CG 55 (33.7) 18 (17.8) 2.35 1.28–4.30
GG 9 (5.5) 4 (4.0) 1.42 0.42–4.73
CG + GG 64 (39.3) 22 (21.8) 0.003 2.32 1.32–4.09
C 253 (77.6) 176 (87.1) 0.006 0.51 0.31–0.83
G 73 (22.4) 26 (12.9) 1.95 1.20–3.18
Impact on ISL1 methylation in patients under the age of 70
OGG1
rs1052133 (C>G)
Ser326Cys
CC 56 (54.4) 47 (71.2) 0.09 0.48 0.25–0.93
CG 40 (38.8) 17 (25.8) 1.83 0.93–3.61
GG 7 (6.8) 2 (3.0) 2.33 0.47–11.59
CG + GG 47 (45.6) 19 (28.8) 0.03 2.08 1.07–4.01
C 152 (73.8) 111 (84.1) 0.03 0.53 0.30–0.93
G 54 (26.2) 21 (15.9) 1.88 1.07–3.29
Impact on ISL1 methylation in current and former smokers
OGG1
rs1052133 (C>G)
Ser326Cys
CC 85 (61.1) 64 (75.3) 0.11 0.53 0.29–0.96
CG 44 (31.9) 17 (20.0) 1.87 0.99–3.55
GG 9 (6.5) 4 (4.7) 1.41 0.42–4.74
CG + GG 53 (38.4) 21 (24.7) 0.03 1.90 1.04–3.46
C 214 (77.5) 145 (85.3) 0.04 0.60 0.36–0.99
G 62 (22.5) 25 (14.7) 1.68 1.01–2.80
Impact on PIK3CA mutability in current and former smokers
OGG1
rs1052133 (C>G)
Ser326Cys
CC 38 (80.9) 112 (62.9) 0.06 2.49 1.13–5.47
CG 7 (14.9) 54 (30.3) 0.40 0.17–0.95
GG 2 (4.3) 12 (6.7) 0.61 0.13–2.85
CG + GG 9 (19.1) 66 (37.1) 0.02 0.40 0.18–0.88
C 83 (88.3) 278 (78.1) 0.03 2.21 1.08–4.17
G 11 (11.7) 78 (21.9) 0.47 0.24–0.93

The common trends with respect to the XRCC1 polymorphism affecting some cancer-related events in the total patient cohort were confirmed in the groups of men, patients aged over 70 years, and smokers (Table 4). So, the wild type XRCC1 399Arg allele increased the frequency of LOH in PTEN and TNKS loci, whereas the minor XRCC1 399Gln allele possessed a protective effect against these events. The XRCC1 Arg399Gln heterozygous genotype as well as those containing at least one minor allele influenced the methylation status of the p16+TIMP3 loci in smokers in the same manner as in the total population.

Table 4. Key cancer-related events affected by the XRCC1 gene polymorphism (rs25487) in BC patients depending on their gender, age and smoking status
Gene polymorphism Genotype/allele Event, n (%) No event, n (%) p OR 95% CI
Impact on PTEN LOH in men
XRCC1
rs25487
(G>A)
Arg399Gln
GG 15 (62.5) 29 (36.7) 0.03 2.87 1.12–7.39
GA 9 (37.5) 38 (48.1) 0.65 0.25–1.65
AA 0 (0) 12 (15.2) 0.11 0.01–1.93
GA + AA 9 (37.5) 50 (63.3) 0.03 0.35 0.14–0.89
G 39 (81.3) 96 (60.8) 0.009 2.80 1.27–6.18
A 9 (18.8) 62 (39.2) 0.36 0.16–0.79
Impact on PTEN LOH in patients over the age of 70
XRCC1
rs25487
(G>A)
Arg399Gln
GG 11 (61.1) 14 (31.8) 0.09 3.37 1.08–10.53
GA 6 (33.3) 22 (50.0) 0.50 0.16–1.57
AA 1 (5.6) 8 (18.2) 0.26 0.03–2.29
GA + AA 7 (38.9) 30 (68.2) 0.047 0.30 0.09–0.93
G 28 (77.8) 50 (56.8.8) 0.04 2.66 1.09–6.49
A 8 (22.2) 38 (43.2) 0.38 0.15–0.92
Impact on PTEN LOH in current and former smokers
XRCC1
rs25487
(G>A)
Arg399Gln
GG 13 (59.1) 26 (39.4) 0.07 2.22 0.83–5.94
GA 9 (40.9) 29 (43.9) 0.88 0.33–2.35
AA 0 (0) 11 (16.7) 0.11 0.01–1.90
GA + AA 9 (40.9) 40 (60.6) 0.14 0.45 0.17–1.20
G 35 (79.5) 81 (61.4) 0.04 2.45 1.09–5.52
A 9 (20.5) 51 (38.6) 0.41 0.18–0.92
Impact on TNKS LOH in men
XRCC1
rs25487
(G>A)
Arg399Gln
GG 13 (61.9) 9 (26.5) 0.03 4.51 1.41–14.46
GA 7 (33.3) 20 (58.8) 0.35 0.11–1.09
AA 1 (4.8) 5 (14.7) 0.29 0.03–2.67
GA + AA 8 (38.1) 25 (73.5) 0.01 0.22 0.07–0.71
G 33 (78.6) 38 (55.9) 0.02 2.89 1.20–6.97
A 9 (21.4) 30 (44.1) 0.35 0.14–0.83
Impact on TNKS LOH in patients over the age of 70
XRCC1
rs25487
(G>A)
Arg399Gln
GG 6 (75.0) 6 (24.0) 0.06 9.50 1.50–60.11
GA 2 (25.0) 15 (60.0) 0.22 0.04–1.33
AA 0 (0) 4 (16.0) 0.28 0.01–5.80
GA + AA 2 (25.0) 19 (76.0) 0.015 0.11 0.02–0.67
G 14 (87.5) 27 (54.0) 0.02 5.96 1.23–29.02
A 2 (12.5) 23 (46.0) 0.17 0.03–0.82
Impact on TNKS LOH in current and former smokers
XRCC1
rs25487
(G>A)
Arg399Gln
GG 13 (72.2) 8 (25.8) 0.006 7.48 2.02–27.65
GA 4 (22.2) 18 (58.1) 0.21 0.06–0.77
AA 1 (5.6) 5 (16.1) 0.31 0.03–2.85
GA + AA 5 (27.8) 23 (74.2) 0.003 0.13 0.04–0.49
G 30 (83.3) 34 (54.8) 0.008 4.12 1.50–11.30
A 6 (16.7) 28 (45.2) 0.24 0.09–0.67
Impact on p16 + TIMP3 methylation in current and former smokers
XRCC1
rs25487
(G>A)
Arg399Gln
GG 6 (26.1) 41 (51.9) 0.04 0.33 0.12–0.92
GA 15 (65.2) 28 (35.4) 3.42 1.25–9.05
AA 2 (8.7) 10 (12.7) 0.66 0.13–3.24
GA + AA 17 (73.9) 38 (48.1) 0.03 3.06 1.09–8.56
G 27 (58.7) 110 (69.6) 0.17 0.62 0.31–1.22
A 19 (41.3) 48 (30.4) 1.61 0.82–3.18

The impact of BER SNPs rs1052133 and rs25487 on some key cancer-related events depending on the BC type. To find out what stage of tumorigenesis is more susceptible to the influence of BER polymorphisms cooperated with key molecular events, two groups of patients with BC (non-muscle invasive BC — NMIBC and muscle invasive BC — MIBC) were compared to each other (Table 5). In this analysis, the OGG1 Ser326Cys polymorphism (the heterozygous genotype and the minor allele) increased the probability of RUNX3 and ISL1 methylation in NMIBC, but decreased PIK3CA mutability in smokers of the same group. Likewise, the XRCC1 Arg399Gln polymorphism (genotypes containing at least one minor allele or this allele itself) inhibited LOH in the PTEN and TNKS loci, but increased the probability of TIMP3 and p16 methylation in patients with NMIBC. Consequently, the trends observed in the total patient cohort were due to associations pronounced in NMIBC suggesting that influence of BER polymorphisms on key events was limited to earlier stages of tumorigenesis.

Table 5. Key cancer-related events affected by polymorphic variants of two BER genes (rs1052133 and rs25487) in NMIBC and MIBC patients
Gene polymorphism Genotype/allele NMIBC MIBC
OR 95% CI p OR 95% CI p
Impact on RUNX3 methylation
OGG1
rs1052133 (C>G)
Ser326Cys
CC 0.54 0.31–0.96 0.05 0.78 0.30–2.02 0.92
CG 2.10 1.14–3.85 1.26 0.46–3.40
GG 0.80 0.25–2.57 1.25 0.12–12.58
CG + GG 1.85 1.05–3.28 0.03 1.28 0.49–3.34 0.61
C 0.67 0.42–1.08 0.1 0.81 0.35–1.84 0.61
G 1.49 0.92–2.40 1.24 0.54–2.84
Impact on ISL1 methylation
OGG1
rs1052133 (C>G)
Ser326Cys
CC 0.53 0.29–0.95 0.1 0.59 0.23–1.50 0.60
CG 1.87 1.00–3.48 1.56 0.58–4.17
GG 1.39 0.40–4.75 1.91 0.20–17.79
CG + GG 1.89 1.05–3.41 0.03 1.70 0.67–4.36 0.26
C 0.60 0.36–0.99 0.04 0.62 0.27–1.39 0.24
G 1.67 1.01–2.75 1.63 0.72–3.67
Impact on PIK3CA mutability in current and former smokers
OGG1
rs1052133 (C>G)
Ser326Cys
CC 3.00 1.15–7.81 0.035 1.52 0.37–6.30 1.0
CG 0.28 0.09–0.84 0.93 0.22–3.88
GG 0.88 0.17–4.45 0.49 0.03–9.41
CG + GG 0.33 0.13–0.87 0.02 0.66 0.16–2.71 0.74
C 2.24 1.01–4.99 0.04 1.78 0.49–6.45 0.42
G 0.45 0.20–0.99 0.56 0.16–2.03
Impact on PTEN LOH
XRCC1
rs25487
(G>A)
Arg399Gln
GG 2.86 0.82–10.05 0.14 1.77 0.55–5.65 0.74
GA 0.78 0.23–2.71 0.63 0.19–2.08
AA 0.16 0.01–2.90 0.61 0.06–6.31
GA+AA 0.35 0.10–1.23 0.11 0.57 0.18–1.81 0.39
G 2.90 1.02–8.23 0.04 1.56 0.61–4.00 0.37
A 0.34 0.12–0.98 0.64 0.25–1.64
Impact on TNKS LOH
XRCC1
rs25487
(G>A)
Arg399Gln
GG 5.74 1.14–28.79 0.1 2.00 0.41–9.71 0.39
GA 0.25 0.04–1.37 0.75 0.15–3.72
AA 0.57 0.06–5.34 0.17 0.01–4.62
GA+AA 0.17 0.03–0.87 0.037 0.50 0.10–2.43 0.44
G 2.85 0.85–9.60 0.1 2.14 0.59–7.84 0.31
A 0.35 0.10–1.18 0.47 0.13–1.71
Impact on TIMP3 methylation
XRCC1
rs25487
(G>A)
Arg399Gln
GG 0.13 0.02–1.09 0.03 0.69 0.14–3.42 0.82
GA 6.00 1.20–29.89 2.04 0.41–10.13
AA 0.66 0.08–5.64 0.66 0.03–13.54
GA+AA 7.53 0.92–61.94 0.04 1.45 0.29–7.18 0.70
G 0.53 0.21–1.33 0.17 0.98 0.28–3.39 1.0
A 1.90 0.75–4.82 1.02 0.30–3.54
Impact on p16 + TIMP3 methylation
XRCC1
rs25487
(G>A)
Arg399Gln
GG 0.23 0.06–0.86 0.03 0.34 0.08–1.50 0.17
GA 4.08 1.31–12.71 4.23 0.96–18.65
AA 0.77 0.16–3.79 0.44 0.02–8.81
GA+AA 4.33 1.16–16.24 0.03 2.92 0.67–12.75 0.18
G 0.56 0.27–1.19 0.13 0.68 0.24–1.89 0.45
A 1.78 0.84–3.76 1.48 0.53–4.15

The results as a whole demonstrated the comparable effects of polymorphisms in both BER genes on the epigenetic status of some tumor suppressor genes in urothelial carcinomas: the OGG1 SNP was associated with increased frequencies of methylated RUNX3 and ISL1 loci, and XRCC1 SNP similarly influenced methylation of the p16 and TIMP3 loci. Allelic imbalance was affected by the XRCC1 SNP with stable protective effects of the XRCC1 399Gln allele against losses of heterozygosity in TNKS and PTEN genes. The association of studied SNPs with mutations in some key genes was less evident and differently manifested itself depending on the target gene/event and important etiological factors: OGG1 SNP increased the mutational rate in RAS family genes in the total study cohort, but decreased the mutational rate in PIK3CA in smokers, whereas the XRCC1 SNP was inactive with respect to mutability of all genes tested.

When studying relationship between BER gene polymorphisms and both mutation frequency in the FGFR3 and TP53 loci and LOH in the RB1 locus, differences were found neither in the total group of patients, nor in the subgroups under study (the data are not shown). And finally, observed associations were characteristic of NMIBC rather than of MIBC.

Discussion

Both genetic and epigenetic changes have been repeatedly discussed to contribute to development of human cancer [31–33]. Genomic changes ranging from point mutations to gross chromosomal aberrations result in modified gene expression profiles and disturbance of signaling networks that control cell division and other cellular functions. In addition to DNA and chromosome damage, DNA methylation is one of the key epigenetic events involved in regulation of gene expression and genomic stability. Aberrant methylation is considered as a potential predictor of malignant growth [34]. Therefore, investigation of the interplay between genetic and epigenetic alterations is of scientific interest and great clinical importance for better understanding the molecular mechanisms underlying tumorigenesis and clinical outcomes in different cancers including BC.

In our previous studies, the OGG1 (codon 326) heterozygous genotype decreased BC risk, especially in smokers with OR = 0.55, 95% CI 0.34–0.89 (p = 0.014) in this group, as well as prevented from high grade tumors as compared to neoplasms of low malignant potential. Polymorphism in XRCC1 locus (codon 399) did not affect susceptibility of the Belarusian population to developing BC and was associated neither with tumor stage, nor with tumor grade [26, 27].

The results of studying mutational and epigenetic variability as well as allelic imbalance in a set of cancer-related genes [19, 35–37] were in accordance with the previously proposed model of developing NMIBC and MIBC [18, 33]. Original findings concerning FGFR3, TP53, RAS, PIK3CA mutability, RUNX3 and other gene methylation and RB1, PTEN, TNKS LOH, which were published earlier and partly presented herein (Table 6), showed that certain molecular alterations result in diverse pathogenic pathways and clinical outcomes. An increased mutational rate in FGFR3 and HRAS loci was preferably observed in non-muscle invasive tumors with a favorable prognosis whereas increased mutability of TP53 and KRAS loci, hypermethylation of the RUNX3 gene and LOH in the RB1 locus were characteristic of muscle invasive carcinomas having a poor prognosis.

Table 6. Characteristics of genes under study and related molecular alterations in tumor tissues
Gene Chromosome Function of encoded protein Alterations Frequencies
in NMIBC in MIBC
Oncogenes
FGFR3 4p16 Protein tyrosine kinase growth factor receptor Point mutations 60–70%162.5%2 5–20%120.6%2
HRAS 11p15 Cytoplasmic GTPase Point mutations 5–10%110.1%2 5–6%11.2%2
KRAS 12p12 Cytoplasmic GTPase Point mutations 5%12.4%2 5%14.9%2
PIK3CA 3q26 Catalitic subunit α (p110 α) Point mutations 25%121.5%2 9–20%113.6%2
Tumor suppressor genes
TP53 17p13 Transcriptional factor; cell cycle and stress response; apoptosis Inactivation mutations 0–14%17.7%2 24–56%140.0%2
RUNX3 1p36 Runt-domain transcription factor Promoter hypermethy­lation 65%154.4%2 85%171.1%2
PTEN 10q23 Protein and lipid phosphatase; negative regulator of AKT signalling Hemizygous deletion;Homozygous deletionLOH 6–8%10%117.1%2 25–58%14–6%134.6%2
RB1 13q14 Negative regulator of the cell cycle; heterochromatin stability Inactivation mutationsLOH No info115.7%2 11–13%138.8%2
TNKS 8p23.1 Tankyrase (Poly-ADP-ribosyl-transferase) LOH No info116.3%2 No info164.3%2
p16 or CDKN2A 9p21 Cyclin-dependent kinase inhibitor 2A Promoter hypermethy­lation Conflicting data110.7%2 Conflicting data112.7%2
TIMP3 22q12.3 Tissue inhibitor of metalloproteinases 3 Promoter hypermethy­lation No info19.4%2 No info112.7%2
ISL1(threshold ≥ 30%) 5q11.1 A member of the LIM/homeodomain family of transcription factors. The encoded protein binds to the enhancer region of the insulin gene regulating its expression Promoter hypermethy­lation No info160.3%2 No info167.3%2
Note: 1 Knowles, Hurst [33]; our data.

When comparing the literature data and results of our study, one can see an analogy between them. Among oncogenes, KRAS seemed to be only exception, as its mutation frequency in our study was significantly higher in MIBC compared to NMIBC [36]. Mutations in the PIK3CA gene often coexisted with those in the FGFR3 locus and were associated with favorable clinical parameters. Our results indicating ISL1, p16, TIMP3 hypermethylation, and RB1 LOH agreed reasonably well with the trends observed by other authors [20, 38–40].

In the framework of the present study we attempted to estimate an impact of SNPs (rs1052133 and rs25487) in two functionally important BER genes on molecular events contributing to BC. The defined relationships between some genetic, epigenetic and chromosomal events in BC are presented in the Figure.

2486 620 BER gene polymorphisms associated with key molecular events in bladder cancer
Figure. BER polymorphisms affected mutation frequencies, methylation and LOH in tumor suppressor genes and oncogenes

Among mutated genes under study only RAS and PIK3CA were affected by OGG1 polymorphism, the frequency of PIK3CA mutations being reduced in carriers of the OGG1 326Cys alleles. PIK3CA mutations are known to be strongly associated with FGFR3 mutations and to be a common event in superficial papillary bladder tumors [28, 41]. Moreover, PIK3CA mutations were associated with FGFR3 mutations in low-grade tumors and reduced recurrence of NMIBC [42]. Significant decrease in oncogenic PIK3CA mutations in carriers of the OGG1 326Ser/Cys heterozygous genotype might underlie the protective effect of this SNP against developing primary NMIBC arising from accumulation of such mutations in smokers.

Our results with respect to association of the OGG1 326Cys variant with RAS mutations in tumor tissues, to some extent, corresponded to findings [43] indicating a 2-fold increased risk of BRAF tumor mutations in homozygous carriers of the OGG1 326Cys minor allele (OR = 2.2, 95% CI 1.02–4.9). It should be noted that encoded protein belongs to the RAF family of serine/threonine protein kinases, which are the first effectors of RAS in the MAP kinase/ERK signaling cascade regulating cell growth, proliferation, and differentiation in response to growth factors, cytokines, and hormones [44]. Mutations in the BRAF proto-oncogene are not very frequent events in bladder carcinomas, but BRAF G469A mutations have been identified in a rare form of BC, inverted urothelial papilloma with a generally accepted benign clinical course [45]. Prevalence of mutations in HRAS and FGFR3 loci in inverted urothelial papilloma and NMIBC reported by other authors [36, 43, 46] may suggest that HRAS and FGFR3 share a similar function and the same pathogenic pathway that has been emphasized by Knowles, Hurst [33]. Based on the data above, one can assume that  OGG1 polymorphism is involved in the pathogenesis of NMIBC due to increase in RAS mutational rate. This mechanism seems to be also appropriate to interpret preferential development of neoplasms of low malignant potential as compared to high grade tumors in carriers of the germline OGG1 326Cys variant [27].

We found the impact of two BER gene polymorphisms (rs1052133 and rs25487) on methy­lation of some tumor suppressor genes (Figure). The association was earlier established between methylation of HOXA9, ISL1, and ALDH1A3, which decreased gene expression levels, and aggressive clinicopathological traits of NMIBC [47]. In the other work, methylated genes including CDH1, FHIT, LAMC2, RASSF1A, TIMP3, SFRP1, SOX9, PMF1, and RUNX3 were associated with poor survival of patients with MIBC suggesting their diagnostic and prognostic values [48]. According to Kitchen et al. [38], CpG islands in the ISL1 promoter were more frequently methylated in recurrent and progressive high-grade BC than in non-recurrent tumors (60.0% vs 18.2%, p = 0.008). Thus, a set of effective biomarkers predicting clinical outcome and therefore allowing the adequate treatment approach have been revealed among methylated genes.

The interplay between DNA methylation status and BER gene polymorphisms is poorly understood; however, the current publications could somewhat elucidate an underlying mechanism. A novel model of DNA demethylation involving OGG1 has been recently proposed by Zhou et al. [49]. According to their results, oxidative stress induces the formation of 8-oxo­guanine (8-oxoG) which stimulates demethylation of adjacent 5-methyl cytosines (5mC). OGG1 binds to 8-oxoG lesions, interacts with and recruits TET proteins to 8-oxoG lesions, converting step by step 5mC to 5-hydroxymethyl cytosine, 5-formylcytosine and 5-carboxylcytosine, which are recognized and excised by thymine DNA glycosylase (TDG) generating an apyrimidinic (AP) site. The AP site is then corrected by specific base-excision repair mechanism with the replacement of normal cytosine [50]. Thus, OGG1 was found to cause specific DNA demethylation in response to oxidative stress, for example, at the promo­ter CpG islands of the BACE1 gene [49]. XRCC1 was also shown to perform a central role in coordinating DNA repair pathways, including DNA demethylation, in both animals and plants [51]. Collectively, these data suggest that BER has an evolutionary conserved role in active DNA demethylation. In mouse germ line, the perturbation of BER enzymes by genetic and pharmacological inhibition resulted in the partial block of global DNA demethylation [52], and deficient DNA demethylation could lead to DNA hypermethylation. Presumably, BER polymorphisms reducing activity of related enzymes modify the demethylation process that is confirmed by observed relationship between the OGG1 326 and XRCC1 399 SNPs and increased frequencies of certain methylated genes in our study.

Recent reports on molecular subtypes of BC focused on allelic imbalance, e.g. LOH within 17p13, 13q14, 10q23, and 9p21 loci including TP53, RB1, PTEN, and CDKN2A/ARF genes, as biomarkers of clinically advanced tumors [33, 53, 54]. However, the influence of DNA repair impairment on allelic imba­lance has not been practically studied; there is a limited body of research indirectly concerning this problem. For example, the relationship between allelic loss of the DNA repair gene OGG1 and 8-oxoG accumulation was revealed in esophageal cancer [55], and similar data were reported for epithelial ovarian cancer [56]. In the latter study, four of the six hete­rozygous OGG1 326 Ser/Cys samples exhibiting allelic imbalance displayed loss of the Ser326 variant. According to the authors, in spite of the fact that the OGG1 326Cys allele appears to play a minor role in conferring increased risk of ovarian cancer, the allelic imbalance of the OGG1 locus observed in association with TP53 somatic mutations may have a diagnostic and prognostic value.

Herein, the XRCC1 polymorphism (rs25487) was associated with LOH in the PTEN and TNKS loci, demonstrating the protective effect against these events. Interestingly, when patients from the total cohort were stratified into groups of NMIBC and MIBC, XRCC1 399Gln allele was associated with the lower frequency of LOH in the PTEN locus in patients with NMIBC (OR = 0.34, 95% CI 0.12–0.98, p = 0.04). In carriers of XRCC1 genotypes containing at least one minor allele, the frequency of LOH in TNKS locus was also decreased in NMIBC (OR = 0.17, 95% CI 0.03–0.87, p = 0.037), and allelic imbalance in both loci were not significantly affected in MIBC. When comparing the strength of other observed associations in these two subtypes of BC (Table 5), the statistical significance was confirmed for NMIBC indicating the involvement of BER polymorphisms into molecular pathways responsible for developing less advanced tumors.

Summing up, BER polymorphisms may affect earlier stages of tumorigenesis through the preferable activation of RAS-associated pathway of BC pathogenesis and/or abnormal methylation of tumor suppressor genes, such as RUNX3, ISL1, p16, and TIMP3. Presumably, these molecular changes confer a selective advantage for tumor cell survival and growth preventing from gross chromosomal alterations that might explain the protective effect of the XRCC1 SNP against PTEN and TNKS LOH in NMIBC. Subsequent accumulation of key molecular events, which trigger progression of NMIBC to MIBC, appears to weaken effects of BER polymorphisms.

Conclusion

The present study demonstrated dual or even multiple effects of BER polymorphic variants (rs1052133 and rs25487) on key molecular events triggering diverse pathogenic pathways of BC. The OGG1 326Ser/Cys heterozygous genotype was associated with increased frequency of mutations in the RAS loci in the total cohort of BC patients whereas the same genotype decreased the frequency of PIK3CA mutations in current and former smokers. The XRCC1 genotypes containing at least one 399Gln allele as well as the minor allele itself were shown to inhibit the allelic imbalance in the PTEN and TNKS loci. Both BER gene polymorphisms affected the promoter methylation in RUNX3, ISL1, p16, and TIMP3. The associations revealed were characteristic of NMIBC rather than of MIBC indicating involvement of BER gene SNPs in regulation of key molecular events at the earlier steps of tumorigenesis leading to development of a less aggressive form of cancer. Elucidation of a pathogenic role and underlying mechanisms of such interplay as well as its influence on a clinical course of the disease can be the goal of a new special study.

REFERENCES

No Comments » Add comments
Leave a comment

ERROR: si-captcha.php plugin says GD image support not detected in PHP!

Contact your web host and ask them why GD image support is not enabled for PHP.

ERROR: si-captcha.php plugin says imagepng function not detected in PHP!

Contact your web host and ask them why imagepng function is not enabled for PHP.