A DNA adductome analysis revealed a reduction in the global level of C5-hydroxymethyl-2'-deoxycytidine in the non-tumoral upper urinary tract mucosa of urothelial carcinoma patients

Abstract

Background: DNA adducts, covalent modifications to DNA due to exposure to specific carcinogens, cause the mispairing of DNA bases, which ultimately results in DNA mutations. DNA methylation in the promoter region, another type of DNA base modification, alters the DNA transcription process, and has been implicated in carcinogenesis in humans due to the down-regulation of tumor suppressor genes. Difficulties are associated with demonstrating the existence of DNA adducts or chemically modified bases in the human urological system. Apart from aristolochic acid-DNA adducts, which cause urothelial carcinoma and endemic nephropathy in a particular geographical area (Balkan), limited information is currently available on DNA adduct profiles in renal cell carcinoma and upper urinary tract urothelial carcinoma, including renal pelvic cancer and ureteral cancer.

Method: To elucidate the significance of DNA adducts in carcinogenesis in the urothelial system, we investigated 53 DNA adducts in the non-tumoral renal parenchyma and non-tumoral renal pelvis of patients with renal cell carcinoma, upper urinary tract urothelial carcinoma, and other diseases using liquid chromatography coupled with tandem mass spectrometry. A comparative analysis of tissue types, the status of malignancy, and clinical characteristics, including lifestyle factors, was performed.

Results: C5-Methyl-2'-deoxycytidine, C5-hydroxymethyl-2'-deoxycytidine (5hmdC), C5-formyl-2'-deoxycytidine, 2'-deoxyinosine, C8-oxo-2'-deoxyadenosine, and C8-oxo-2'-deoxyguanosine (8-OHdG) were detected in the renal parenchyma and renal pelvis. 8-OHdG was more frequently detected in the renal pelvis than in the renal cortex and medulla (p = 0.048 and p = 0.038, respectively). 5hmdC levels were significantly lower in the renal pelvis of urothelial carcinoma patients (n = 10) than in the urothelium of patients without urothelial carcinoma (n = 15) (p = 0.010). Regarding 5hmdC levels in the renal cortex and medulla, Spearman's rank correlation test revealed a negative correlation between age and 5hmdC levels (r = - 0.46, p = 0.018 and r = - 0.45, p = 0.042, respectively).

Conclusions: The present results revealed a reduction of 5hmdC levels in the non-tumoral urinary tract mucosa of patients with upper urinary tract urothelial carcinoma. Therefore, the urothelial cell epithelia of patients with upper urinary tract cancer, even in non-cancerous areas, may be predisposed to urothelial cancer.

Keywords: C5-hydroxymethyl-2′-deoxycytidine; DNA adduct; DNA adductome; DNA adductomics; Oxidative DNA damage; Renal cell carcinoma; Upper urinary tract urothelial carcinoma.

Fig. 1

Examples of sampling sites from resected kidneys. a A macroscopic view of multisite sampling in a kidney from a patient with renal pelvic cancer who underwent nephroureterectomy. The normal renal pelvic mucosa (N1 and N2) was taken from regions other than the tumor (T). Samples of the renal cortex (N3) and renal medulla (N4) were also collected. b The sites of tissue sampling in a kidney with renal cell carcinoma (T). Renal pelvis tissues were collected and named N1 to N4. Samples of the renal cortex (N5) and renal medulla (N6) were taken from the non-tumoral renal parenchyma

Fig. 2

Representative chromatograms of 5hmdC. The x-axis represents the retention time. The y-axis represents the intensity of the target molecule and the arrowhead shows the “best retention time” corresponding to the highest peak in the peak boundary. A blank sample (a) and standard sample (b) that contained a chemical compound of 5hmdC were used as controls. Representative chromatograms are shown from every group (c: Cortex-RCC group, d: Cortex-non-RCC group, e: Medulla-RCC group, f: Medulla-non-RCC group, g: Pelvis-UTUC group, h: Pelvis-non-UTUC group). The best retention times of all groups were consistent

Fig. 3

Relationships between herb usage and DNA adducts. The relationships between herb usage and 5mdC (a) and 5hmdC (b) levels in the renal cortex are shown as dot plots. The y-axis represents the molar ratio (number/base) of each DNA adduct

Fig. 4

The comparison of 5hmdC in the pelvis-UTUC group. The molar ratios (number/bases) of 5hmdC were shown according to tumor multicentricity (a) and the presence of bladder cancer history (b). We described the sample size under the dot plots

Fig. 5

The molar ratio of DNA adducts in each group. The molar ratios (number/bases) of DNA adducts (a: 5mdC, b: 5hmdC, c: 5fdC, d: dI, e: 8-OHdA, f: 8-OHdG) on the y-axis were compared: Cortex-RCC group vs Cortex-non-RCC group; Medulla-RCC group vs Medulla-non-RCC group; Pelvis-UTUC group vs Pelvis-non-UTUC group. We described the sample size under the group name. As described in the main text and Table 3, 5hmdC levels were higher in the Pelvis-non-UTUC group than in the Pelvis-UTUC group (p = 0.010)

Fig. 6

Relationships between 5hmdC levels, age, and eGFR. The molar ratios of 5hmdC in the renal cortex (a and b) and renal medulla (c) were plotted with age (a and c) and eGFR (b). The regression coefficient (R²) and p value were calculated by a linear regression analysis and indicated on the scatterplot with a regression line. The correlation coefficient and p value calculated by Spearman’s rank correlation analysis are described under the graphs

Fig. 7

Relationship between 5mdC and dI. The relationships among all DNA adducts are shown in Supplementary Table S8. The molar ratios of 5mdC and dI in the renal cortex (a) and renal medulla (b) were plotted. The regression coefficient (R²) and p value were calculated by a linear regression analysis and indicated on the scatterplot with a regression line. The correlation coefficient and p value calculated by Spearman’s rank correlation analysis are described under the graphs