Urinary APE1/Ref-1: A Potential Bladder Cancer Biomarker

Bladder cancer (BCa) is one of the most common urothelial cancers with still noticeable incidence rate. Early detection of BCa is highly correlated with successful therapeutic outcomes. We previously showed that apurinic/apyrimidinic endonuclease 1/redox factor-1 (APE1/Ref-1) was expressed at an increased level in the serum of BCa patients when compared to the level in healthy controls. In this study, we investigated whether urinary APE1/Ref-1 was also elevated in patients with BCa. In this case-control study, voided urine was collected from 277 subjects including 169 BCa patients and 108 non-BCa controls. Urinary APE1/Ref-1 level was assessed by enzyme-linked immunosorbent assay (ELISA). APE1/Ref-1 levels were significantly elevated in BCa patients relative to levels in non-BCa controls and were correlated with tumor grade and stage. Urinary APE1/Ref-1 levels were also higher in patients with recurrence history of BCa. The receiver operating characteristics (ROC) curve of APE1/Ref-1 showed an area under the curve of 0.83, indicating the reliability and validity of this biomarker. The optimal combination of sensitivity and specificity was determined to be 82% and 80% at a cut-off value of 0.376 ng/100 μL for detection of APE1/Ref-1 in urine. In conclusion, urinary APE1/Ref-1 levels measured from noninvasively obtained body fluids would be clinically applicable for diagnosis of BCa.


Introduction
Bladder cancer (BCa) is the second most common of all genitourinary malignancies in the United States [1] and Korea [2]. Most individuals with BCa who are diagnosed early, show no muscle invasion, and have superficial urothelial carcinoma can expect a 5-year survival rate of more than 90% [3]. If the BCa is invasive, however, with tumor cells spreading beyond the bladder to the surrounding tissue or to nearby lymph nodes, or organs, signs of late stage BCa, the 5-year survival rate drops sharply. Therefore, early intervention can dramatically increase the probability of a BCa patient's survival. Patients with noninvasive BCa frequently show a high rate of recurrence and progression of the disease within 2 years of transurethral resection [4], and continuous follow-up testing is required. Several studies have focused on the development of tools for the diagnosis and prognosis of BCa using urinary biomarkers [5][6][7][8][9].
In our previous report, we proposed a new BCa diagnostic protein, apurinic/apyrimidinic endonuclease 1/redox factor-1 (APE1/Ref-1) in serum [10]. APE1/Ref-1 protein was originally identified as a multifunctional protein involved in DNA repair and redox signaling. APE1/Ref-1 levels were found to be elevated with dysregulated cellular proliferation, as is typically seen in cancers [10][11][12]. APE1/Ref-1 is mainly localized in the nucleus and shows dynamic shuttling between the nucleus and cytoplasm in response to various stress stimuli. Furthermore, extracellular secretion of APE1/Ref-1 into the circulation suggests this protein could be used as a serologic biomarker [13,14]. APE1/Ref-1 secretion from cells is also supported by the presence of autoantibody against APE1/Ref-1 in the blood of patients with lung cancer [15] and systemic lupus erythematosus [16].
The clinical need for specific and sensitive urothelial tumor diagnostics remains as urgent issue. The ideal diagnostics would measure the level of a BCa protein biomarker 2 Disease Markers in a single step with a noninvasive sampling method. The protein components of urine are qualitatively similar to those of blood although they are more diluted [17]. A diagnostic based on patient urine has some advantages as urine is available in large quantities and can be obtained by noninvasive means for repeated measurements, continuous surveillance, or establishing of assay reproducibility.
In this study, we compared the urinary APE1/Ref-1 levels of patients with BCa and healthy subjects. Using an existing, quantitative APE1/Ref-1 serum assay that was modified for use with urine sample, we aimed to quantify the level of APE1/Ref-1 in urine samples from patients with results from cystoscopic examination of the bladder to establish a new, noninvasive urinary BCa biomarker.

Patients and Urine
Samples. Urine samples were obtained from 277 patients. All analyses were performed within 6 months of collection. The study subjects were classified as either noncancer controls ( = 108), individuals with no evidence of malignancy, or patients with bladder tumors ( = 169). Control patients were diagnosed as having nonmalignant urological diseases including benign prostate hyperplasia ( = 37), neurogenic bladder ( = 23), urolithiasis ( = 21), stress urinary incontinence ( = 19), and urethral stricture ( = 8). Tumors were staged according to the 2011 tumor-node-metastasis (TNM) classification [18] and graded according to the World Health Organization (WHO) system [19] by two pathologists that were blind to this study. In the BCa group, postoperative histological confirmation of urothelial cell carcinoma, including grade and stage, was recorded. All patients' clinical pathological data, including prior medical history and treatment, recurrence, and clinical stage, were retrieved from electronic medical records (Table 1). Voided urine samples were obtained from patients with BCa before cystoscopy and then frozen in liquid nitrogen and stored at −70 ∘ C until use. All samples were obtained from the archives of the Department of Urology, Chungnam National University, from 2011 to 2012. The collection and analysis of all samples were approved by the Chungnam National University Hospital institutional review board, and informed consent was obtained from all study subjects.

Urine Sample Preparation.
Each urine sample was thawed and sedimented by centrifugation (10,000 ×g for 10 min) to remove impurities. The supernatant was carefully collected and prepared without an additional protein enrichment step.

Measurement of Urinary APE1/Ref-1.
The level of APE1/Ref-1 in each urine sample was quantitatively analyzed using a sandwich enzyme-linked immunosorbent assay (ELISA) as described previously [10], with some modification. Briefly, 96-microwell plates (Thermo Fisher Scientific, Waltham, MA, USA) were precoated overnight with 100 L of a 1 : 5000 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. Plates were washed five times with phosphate buffered saline (PBS) containing 0.05% Tween 20 (PBS-T) between all incubation steps. After blocking with blocking buffer (5% bovine serum albumin in PBS-T) at room temperature for 2 h, plates were washed five times with PBS-T. Urine sample, standards, and blank (100 L/well) were added to the wells. Plates were incubated at 37 ∘ C for 90 min and then washed five times with PBS-T to ensure the removal of remnants that could affect antibody recognition. This was followed by the addition of 100 L of a 1 : 5000 dilution of a mouse anti-APE1/Ref-1 antibody (Abcam, Cambridge, UK) and further incubation at room temperature for 2 h. The plates were then washed seven times with PBS-T, and 100 L of horseradish peroxidase-conjugated anti-mouse antibody (1 : 5000) was added; the plate was incubated at room temperature for 30 min. After washing, 100 L of freshly prepared tetramethyl benzidine substrate was added to the wells. The color development reaction was stopped by adding 100 L of 2.5 M H 2 SO 4 , and the absorbance was measured at 450 nm with an automatic microtiter plate reader (Sunrise Xfluor4; Tecan Systems, Inc., San Jose, CA, USA). Each sample was assayed in duplicate, and mean values were determined. A 6-point standard curve was established using purified recombinant human APE1/Ref-1 (rh APE1/Ref-1) [10]. The protein (1 g/mL) was serially diluted (5-fold) in diluent and used at 0.0064-4 ng/100 L.

Statistical
Analysis. Data were expressed as the mean ± standard error of the mean (SEM). A repeated-measures ANOVA was used to compare more than three variables, while Student's t-test was used to evaluate differences between two variables. Receiver operating characteristic (ROC) curves were generated by plotting the sensitivity value against the false-positive rate (1 − specificity). We assessed the predictive value of APE1/Ref-1 for BCa by calculating the area under the curve (AUC), and we estimated the optimal cut-off value (Youden index), based on maximum sensitivity and specificity [20]. SPSS version 18.0 for Windows was used to analyze the data. Differences were considered statistically significant if the null hypothesis could be rejected with >95% confidence ( < 0.05).  (Figure 1(a))). ROC analysis of control versus BCa groups (Figure 1(b)) yielded an AUC of 0.83; the ROC curve and corresponding AUC show that urine APE1/Ref-1 as a biomarker has the predictive ability to discriminate between patients with BCa and normal subjects. The optimal cut-off value was set to maximize the sum of sensitivity and specificity. Based on these results, the optimal combination of sensitivity (81.7%) and specificity (79.6%) was obtained at a cut-off value of 0.3765 ng/100 L ( Table 2).

Urine Levels of APE1/Ref-1 in Patients with Hematuria.
False-positive results in the nuclear matrix protein 22 (NMP22) test have been observed in cases with hematuria [21]. Therefore, we investigated the effect of hematuria on urine APE1/Ref-1 levels in human subjects. As shown in Figure 3

Discussion
In the present study, to measure urinary APE1/Ref-1 levels, we modified an APE1/Ref-1 ELISA that was initially developed for the detection of serum APE1/Ref-1 [10]. We obtained an ROC curve by applying results of human urine samples testing and determined a Youden index-based, optimal cut-point for the urinary diagnostic test [20]. A cut-off APE1/Ref-1 value of 0.3765 ng/100 L resulted in a sensitivity of 81.7% and a specificity of 79.6% of the assay. We also observed that APE1/Ref-1 protein levels were elevated in BCa patients, even those with low-grade tumors, compared with the levels in normal controls. Furthermore, the increased urinary APE1/Ref-1 levels were associated with BCa severity, suggesting that this protein could be used as a reliable marker for BCa diagnosis.
We previously reported serum APE1/Ref-1 as a BCa biomarker [10]. However, to date, there are still no bloodbased biomarker tests for use in diagnosis or surveillance of BCa in clinical practice although markers in blood are useful in determining prognosis and informing therapeutic decisions. However, particularly for monitoring bladder diseases, biomarkers in urine have distinct advantages, including larger sample quantities, noninvasive sampling for repeated measurements, and less concern about pathogens transmission than with a blood sample. In general, the proteins in urine originate from glomerular filtration of blood, excretion from epithelial cells in the urinary tract, sloughing of epithelial cells and casts, and formation of urinary exosomes [22]. Changes in urine protein components and concentrations may directly indicate dysfunction of urothelial cells that line the uropoetic organs within the bladder. Therefore, measurement of urinary APE1/Ref-1 levels for BCa monitoring is a rational strategy. However, further studies are needed to determine the origin of urinary APE1/Ref-1 in the bloodstream or bladder itself, the mechanism of excretion into urine, and the meaning of the blood to urine APE1/Ref-1 ratio; urinary APE1/Ref-1 can be established as a new BCa biomarker. However, the results from our urinary APE1/Ref-1 analysis demonstrate an acceptable sensitivity, specificity, and reliability as a BCa diagnostic. APE1/Ref-1 has been implicated in the development and progression of various cancers [10][11][12]23]. Genetic variants of APE1/Ref-1 have been also studied to determine the relationship between specific polymorphisms and cancer susceptibility. Because of its frequency, the most well studied genetic variant is Asp148Glu; D148E [24]. Interestingly, APE1/Ref-1 has been shown to exhibit an atypical subcellular distribution pattern and cytoplasmic localization in many cancer types [25]. Based on our finding and those of previous reports, we propose that the high frequency and cytoplasmic localization of the genetic variant D148E APE1/Ref-1 in BCa cells are associated with the existence of this protein in urine.
In this study, we used frozen urine samples. To avoid alteration of the protein composition during frozen storage, we minimized the number of freeze-thaw cycles (1-2) by preparing small aliquot of each urine sample. Further, precipitates containing cellular debris, which may bias test results, were removed from the urine before storage. However, it is possible that freshly voided urine samples will yield different results. The cellular contribution of proteins derived from