Samples were obtained from 106 consecutive patients. All analyses were performed within 6 months of collection. The study groups were classified as either non-cancer controls (n=55), consisting of individuals with no evidence of malignancy, or patients with operable bladder cancer (n=51). All samples were obtained from the archives of the Department of Urology, Chungnam National University from 2010-2011. This study was approved by the Chungnam National University Hospital Institutional Review Board, and all participants signed informed consent forms. In the control group, we included patients with benign prostate hyperplasia (n=22), a urinary tract stone (n=18), trauma (n=11), and urethral stricture (n=4). We excluded patients with chronic or recurrent urinary tract infections, unevaluated gross hematuria, and malignancy. In the bladder cancer group, postoperative histological confirmation of urothelial cell carcinoma, including grade and stage, was recorded. No chemotherapy or radiotherapy was administered prior to surgery. Thirty-eight patients underwent transurethral resection of bladder cancer and 13 patients underwent radical or partial cystectomy. Blood (8-10 mL) was drawn from an antecubital vein in patients with bladder cancer or healthy control subjects using a 22-G needle and placed in a vacuum tube. Serum was separated by centrifugation at 3,000 rpm for 10 minutes at room temperature, and then re-centrifuged at 5,000 rpm for 5 minutes to obtain cell-free serum, which was stored in liquid nitrogen until use.

The clinical and pathological characteristics of the bladder cancer subjects, which were divided according to nonmuscle invasive bladder cancer (NMIBC) and muscle invasive bladder cancer (MIBC), are shown in Table 1. Clinical staging of bladder cancer was performed using cystoscopy, computed tomography, bone scanning, and simple chest radiography. All tumor tissues were analyzed by a single genitourinary pathologist according to a standardized protocol. Pathological staging was recorded in accordance with the 2003 TNM classification and the tumors were assigned a grade according to the World Health Organization classification [16,17]. Tumors were thus classified as Ta (noninvasive papillary carcinoma), T1 (invasive up to the subepithelial connective tissue), T2 (invading the muscle), T3 (invading the perivesical fat), and T4 (invading an adjacent organ). Tumors were also graded based on the likelihood of recurrence and progression into grade I (tumors may recur but have only a low risk of progression), grade II (tumors are more likely to recur and progress compared with grade I), and grade III (tumors are very likely to recur and progress). Bladder cancer patients were followed-up, and cystoscopy and urine cytology were performed every 3 months for the first year, every 6 months for the second year, and annually thereafter. Recurrence was defined as transitional cell carcinomas with lower or equivalent pathologic stage after treatment. For MIBC patients who had undergone partial cystectomy or radiation treatment, reappearance of bladder cancer including Ta and carcinoma in situ was considered recurrence.

A sandwich enzyme-linked immunosorbent assay (ELISA) was used to quantify serum APE1/Ref-1 levels. Briefly, 96-microwell plates (Nunc, Penfield, NY) were precoated overnight with 100 μL of a 1:1,000 dilution of a rabbit anti-APE1/Ref-1 antibody (Abcam, Cambridge, UK) in coating buffer (0.5 M carbonate buffer, pH 9.6) in each well. After blocking with blocking buffer (5% bovine serum albumin in phosphate buffered saline [PBS] containing 0.05% Tween 20 [PBS-T]) at room temperature for 60 minutes, 100 μL of sample was added to the wells and the plates were incubated at 4°C for 90 minutes, and then washed five times with PBS-T. This was followed by addition of 100 μL of a 1:1,000 dilution of mouse anti-APE1/Ref-1 antibody (Abcam), and further incubation at room temperature for 2 hours. The plate was then washed seven times with PBS-T, and 100 μL of horseradish peroxidase–conjugated secondary antibody (1:5,000) was added, followed by incubation at room temperature for 30 minutes. After further washing, 100 μL of freshly prepared tetramethyl benzidine substrate was added to the wells. The color development reaction was stopped by addition of 100 μL of 2.5 M H₂SO₄, and the absorbance was measured at 450 nm using an automatic microtiter plate reader (Sunrise Xfluor 4, Tecan Systems Inc., San Jose, CA). Each sample was assayed in duplicate, and mean values were determined. To establish a standard curve, purified recombinant human APE1/Ref-1 (1 mg/mL) was serially diluted (5-fold) and used in a concentration series from 0.16-20 ng/100 μL.

Human full length APE1/Ref-1 DNA was inserted into the pET28b expression vector (Novagen, Gibbstown, NJ), containing a 6-histidine tag for easy purification [10]. pET28b-APE1/Ref-1 plasmids were then transformed into the BL21(DE3) strain of Escherichia coli. Following induction with isopropyl β-D-1-thiogalactopyranoside (IPTG), the cells were sonicated in lysis buffer (100 mM NaCl, 20 mM HEPES), and the recombinant protein was purified on a nickel-nitrilotriacetic acid agarose column (Qiagen, Valencia, CA). After washing, the isolated APE1/Ref-1 protein was eluted with 250 mM imidazole buffer followed by desalting on a PD-10 column (Amersham Pharmacia Biotech, Liverpool, UK) in PBS, and frozen in 10% glycerol at −80°C. This protein was subsequently used for the standard curve in the APE1/Ref-1 ELISA.

Imunohistochemical staining of APE1/Ref-1 was determined in paraffin-embedded sections of bladder tissues, which were stored in the tissue bank of Chungnam National Hospital. Tissue specimens were cut into 4-μm sections from randomly selected tumor and non-tumor blocks. Distant, apparently normal tissue samples from the same patient were used as the non-tumor blocks. A monoclonal antibody against human APE1/Ref-1 was used at a 1:600 dilution (Novus Biological, Littleton, CO). Staining was developed using 3, 3′diaminobenzidine (DAKO, Carpinteria, CA). The sections were lightly counterstained with hematoxylin and mounted using Immu-Mount (Thermo Shandon, Midland, Canada). Negative control slides were incubated with the mouse IgG1 negative control reagent provided with the kit for evaluation of nonspecific staining.

Tumor tissues from bladder cancer patients were obtained from the tissue bank of Chungnam National Hospital. Distant, apparently normal tissue samples from the same patient were used as the non-tumor tissues. Tissues were suspended in PBS containing protease and phosphatase inhibitors, cut into small pieces, and homogenized using a polytron. The homogenates were centrifuged at 14,000 rpm for 40 minutes and the resultant supernatant fractions were used for immunoblotting. The proteins were resolved by sodium dodecyl sulfate polyacrylamide gel electrophoresis and transferred onto a polyvinylidene fluoride membrane. After blocking with 5% skimmed milk powder, the blots were incubated with anti APE1/Ref-1 antibody (1:1,000, Abcam). Immunoreactive bands were visualized by enhanced chemiluminescence. Each membrane was stripped and re-probed with an anti–glyceraldehyde-3-phosphate dehydrogenase (GAPDH) antibody to ensure equivalent protein loading. Immunoblotting for each protein was performed at least twice using independently prepared lysates. Changes in protein levels were quantified by densitometric scanning of the immunoreactive band and normalized against GAPDH loading control.

Values are expressed as the mean±SE. A statistical evaluation was performed using a one-way analysis of variance and post-hoc analysis. SPSS ver. 18.0 (SPSS Inc., Chicago, IL) was used for evaluation of the data. Differences were considered statistically significant if the null hypothesis could be rejected with > 95% confidence interval (p < 0.05). Nonparametric receiver operating characteristic (ROC) curves were generated by plotting the sensitivity value against the false-positive rate (1–specificity).We assessed the potential predictive value of APE1/Ref-1 for bladder cancer as well specific cancer stages and grades, and recurrence, by calculating the area under the curve (AUC). We estimated the sensitivity and specificity of APE1/Ref-1 at the optimal cutoff value (Youden index) to maximize the sum of sensitivity and specificity [18].

Serum levels of APE1/Ref-1 in bladder cancer patients were significantly elevated compared to those of the control group (3.548±0.333 ng/100 μL [n=51] for bladder cancer vs. 1.547±0.319 ng/100 μL [n=55] for the control group) (Fig. 1A). The US National Cancer Institute recommends use of ROC curve for evaluation of the performance of potential cancer detection markers [19]. Serum samples from 55 non-cancer controls and 51 patients with bladder cancer were used for the ROC curve analysis to determine the sensitivity and specificity of APE1/Ref-1 in detection of bladder cancer. The ROC curve for all tumors resulted in an AUC of 0.824 (95% confidence interval, 0.74 to 0.90) (Fig. 1B).

The ROC is drawn through points representing different decision cut-off values. The range of sensitivity and specificity at these different values was 0.51-0.95 and 0.51-0.86, respectively (Table 2). The sensitivity and specificity of APE1/Ref-1 were estimated at the optimal cutoff value in order to maximize the sum of sensitivity and specificity [18]. On this basis, the optimal combination of sensitivity and specificity were determined to be 0.93 and 0.59, using a cut-off value of 2.83 ng/100 μL (Table 2).

The grade of bladder cancer provides important prognostic information and can help guide treatment, and is based on the degree of cellular abnormality. We found that serum APE1/Ref-1 levels were associated with the grade of bladder cancer (Fig. 2A). In particular, the mean values of serum APE1/Ref-1 levels of patients with grade II and III bladder cancer (2.719±0.465 ng/100 μL for grade II, n=12; 4.891±0.389 ng/100 μL for grade III, n=23) were significantly greater than those of the control group (1.547±0.319 ng/100 μL, n=55). The ROC-AUC for grade I, grade II, and grade III cancers were 0.70, 0.71, and 0.96, respectively (Fig. 2B), suggesting that serum APE1/Ref-1 may serve as a predictor of tumor grade.

In addition to grade, bladder cancer is also staged by depth of invasion into the bladder wall. Thus, we attempted to determine whether serum APE1/Ref-1 is also associated with the tumor stage (Fig. 2C). Serum APE1/Ref-1 levels were significantly elevated among patients with Ta (2.450±0.365 ng/100 μL), T1 (3.251±0.305 ng/100 μL), and T2-T3 (5.711±0.437 ng/100 μL) bladder tumors compared to the control groups. The ROC AUC for Ta, T1, and T2-T4 were 0.67, 0.87, and 0.99, respectively (Fig. 2D). In addition, we evaluated the change of serum APE1/Ref-1 in NMIBC and MIBC. Serum APE1/Ref-1 levels in MIBC (5.711±0.437 ng/100 μL) were significantly elevated, compared with NMIBC (2.801±0.249 ng/100 μL) (Fig. 2E). The ROC-AUC for NMIBC and MIBC were 0.76 and 0.99, respectively (Fig. 2F).

Next, we attempted to determine whether serum APE1/Ref-1 levels were also elevated in bladder cancer recurrence. Among the 51 patients with bladder cancer, 20 patients (39%) had recurrent bladder cancer. The mean serum APE1/Ref-1 level titer in these patients was (4.480±0.430 ng/100 μL), significantly greater than in patients with non-recurrent tumors (2.940±0.324 ng/100 μL) (Fig. 2F). In results of ROC analysis, the AUC for non-recurrent tumors was 0.75, and the AUC for recurrent tumors was 0.93 (Fig. 2G), suggesting a high predictive value of serum APE1/Ref-1 for recurrent bladder tumors. Serum APE1/Ref-1 levels in patients with multifocal tumors did not differ significantly from those who had single tumors (data not shown).

To determine whether APE1/Ref-1 protein was differentially expressed in bladder cancer, immunoblotting for APE1/Ref-1 was performed in non-tumor tissues and tumor tissues of bladder cancer (Fig. 3A). Differences in protein level were quantified by densitometric scanning of the immunoreactive bands and normalized against the GAPDH loading control. This revealed that APE1/Ref-1 protein was expressed at a far higher level in tumor tissue (T) compared to non-tumor tissue as normal (N) (Fig. 3A). APE1/Ref-1 protein expression in tumor and non-tumor tissues of bladder cancer was visualized using immunohistochemistry (Fig. 3B). APE1/Ref-1 in non-tumor tissues and tumor tissues were mainly localized in the nucleus. However, nuclear expression of APE1/Ref-1 was significantly increased in tumor tissues. In addition, mononuclear cells including monocytes and lymphocytes were infiltrated close to tumor regions.