The impact of receptor of advanced glycation end‐products polymorphisms on prostate cancer progression and clinicopathological characteristics

Abstract The receptor for advanced glycation end products (RAGE) overexpression was suggested to be associated with prostate cancer development and poor prognosis. In this study, we focused on the correlations between the clinicopathological characteristics and susceptibility of prostate cancer and RAGE single‐nucleotide polymorphisms (SNPs). In 579 prostate cancer patients, the RAGE SNPs rs1800625, rs1800624, rs2070600 and rs184003 in patients with or without grade group upgrade were analysed with real‐time polymerase chain reaction. The results demonstrated that the prostate cancer patients who carried the RAGE SNPs rs2070600 ‘GA’ genotypic variants were significantly associated with lower risk to develop grade group upgrade. Moreover, patients with the RAGE rs1800625 ‘TC + CC’ genotypic variants were associated with higher risk of perineural invasion. In 343 prostate cancer patients who carried the RAGE rs1800625 ‘TC + CC’ genotype without grade group upgrade were correlated with higher risk of biochemical recurrence and perineural invasion. In the analysis of TCGA database, significant differences of the RAGE mRNA level were found between the normal controls and prostate cancer patients (p < 0.0001), and the pathologic stage N1 and N0 patients (p = 0.0027). The prostate cancer patients with high RAGE expression were associated with lower overall survival rate (p = 0.025). In conclusion, our results have revealed that the RAGE SNPs rs2070600 and rs1800625 were associated with the grade group upgrade of prostate cancer and clinical status. The RAGE polymorphisms may provide as a pivotal predictor to evaluate prostate cancer disease progression and prognosis.


| INTRODUC TI ON
Prostate cancer (PCa) is a global health problem with considerable diversity in epidemiology and genomics. 1 In Taiwan, prostate cancer is the fifth most prevalent cancer and ranks the seventh highest cancer-related mortality rate. [2][3][4] Epidemiological risk factors such as ageing and high fat consuming diet were suggested to raise the incidence of PCa in Taiwan. 2,5 RAGE, or the AGER, is the receptor for advanced glycation end products (AGEs). 6,7 The AGEs are nonenzymatic protein modifications, which were produced during ageing. 7,8 In prostate cancer, overexpression of RAGE and its ligand amphoterin were found to be correlated with tumour development and poor prognosis. [9][10][11] The RAGE expression was observed to be correlated with apoptosis induction and inhibition of prostate tumour growth, 12 and the RAGE quantification of human prostate cancer samples has been confirmed that increased uptake of RAGE was corresponding to increasing of Gleason scoring. 13 The polymorphisms of RAGE were suggested to be associated with various cancers, [14][15][16] including oral cancer, 17 breast cancer, [18][19][20][21][22][23][24] lung cancer, [25][26][27][28] gastric cancer, 29-31 hepatocellular carcinoma (HCC), 32 pancreas cancer, 33 cervical cancer, 34,35 urothelial cell carcinoma 36 and colorectal cancer. 37 Previous studies revealed that the RAGE rs1800625 polymorphism was correlated with the increasing of cancer risk in various cancers including oral cancer and gastric cancer. 15,17,29 Moreover, the 'TT' polymorphisms of rs184003 were suggested to be correlated with poorer disease-specific survival on urothelial cell carcinoma, 36 and individuals who carried the rs184003 T allele were found to exhibit increased risk of breast cancer. 20 However, the RAGE polymorphisms to prostate cancer progression and clinicopathologic characteristics remained not well-investigated. In this study, we focused on four SNPs of RAGE rs1800625, rs1800624, rs2070600 and rs184003, and try to elucidate their correlations to clinicopathologic characteristics and susceptibility of prostate cancer.

| Study subjects
In the current study, 579 prostate cancer patients with adenocarcinoma were enrolled as the study group. During 2012-2017, the patients who involved in our study have received robotic assisted radical prostatectomy at Taichung Veteran General Hospital. The informed consent was confirmed and acquired from each individual who enrolled in our study (IRB No. CE19062A). The medical information including the age at diagnosis (years), initial PSA level at diagnosis (ng/ml), clinical and pathological TNM staging, pathologic Gleason grade group, perineural invasion, seminal vesicle invasion, lymphovascular invasion, biochemical recurrence and D'Amico classification was acquired from the personal medical records for each patient. 38 Before this study started to initiate, the certification and approval was confirmed by the Institutional Review Broad (IRB) of the Taichung Veteran General Hospital.

| Sample preparation and DNA extraction
For genomic DNA extraction, the peripheral blood specimens from normal controls and prostate cancer patients who enrolled in our study were collected. The samples of peripheral whole blood were preserved in EDTA containing tubes and centrifuged with the settings of 3000 g for 10 min. The buffy coats extracted from centrifuged whole blood specimens were further applied for the DNA extraction. 39 The Tris-EDTA (TE) buffer was used to dissolve DNA

| Selection of RAGE SNPs and RAGE SNPs genotyping
In our current study, a total of four SNPs of RAGE rs1800625, rs1800624, rs2070600 and rs184003 were selected from the International HapMap Project database. 41 The RAGE rs1800624 polymorphism was suggested to contribute to increase breast cancer and lung cancer risk. 23,42 The RAGE rs2070600 polymorphism was associated with significant breast cancer and gastric cancer risk. 20,30 The assessment of allelic discrimination for the RAGE rs184003, rs2070600, rs1800624 and rs1800625 SNP was performed with ABI StepOne Software v2.3 Real-Time PCR System. The genotyping was analysed with the TaqMan assay. The SDS 7000 series software (Applied Biosystems) was applied for the analysis and calculation of the final data of genotyping.  Figure 1A), and pathologic stage N1 and N0 patients (p = 0.0027, Figure 1C). However, no significant differences of the RAGE mRNA expression between pathologic N0 stage and N1 stage were observed ( Figure 1B). The prostate cancer patients who posses higher RAGE expression were correlated with lower overall survival rate (Log Rank p = 0.025, Figure 1D).  (Tables 2 and 3).

| Statistical analysis
We further examined the correlations between the RAGE SNPs and clinical status of prostate cancer. Intriguingly, we found that although the RAGE rs1800625 polymorphisms were not associated with the grade group upgrade of prostate cancer (Tables 2 and 3), however, the RAGE rs1800625 genotypic variants 'TC + CC' were found to be significantly associated with perineural invasion of prostate cancer (p = 0.005, Table 4). Moreover, in 343 prostate cancer patients with no grade group upgrade, the RAGE rs1800625 polymorphic variants 'TC + CC' were also found to be associated with perineural invasion (p = 0.014) and biochemical recurrence (p = 0.039) ( Table 5). The RAGE rs1800625 polymorphisms were suggested to be associated with increased cancer risk in various cancers. 15,17,36,56,57 Previous study has suggested that the C allele of rs1800625 may induce the expression of RAGE, and leads to chronic inflammatory conditions in diabetic retinopathy. 58 Besides, the variant of the RAGE rs1800625 SNP was suggested to be associated with the hypomethylation of the promoter region of RAGE and contribute to the ulcerative colitis risk. 59 Furthermore, after we analysed the TCGA database, we found that the RAGE mRNA level was significantly associated with prostate cancer tumorigenesis ( Figure 1A) and pathologic N1 stage development ( Figure 1C).
The higher RAGE expression was also observed to be associated with lower overall survival rate in prostate cancer patients (Log Rank p = 0.025, Figure 1D) and rs1800625 polymorphisms may provide as pivotal markers to predict tumour aggressiveness, recurrence and prognosis in prostate cancer.

CO N FLI C T S O F I NTE R E S T
The authors declare that there is no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data used to support the findings of the present study are available from the corresponding author upon request.