Urine gamma-synuclein as a biomarker for the diagnosis of bladder cancer

Gamma-synuclein (SNCG) is secreted from tumor cells and elevated in the urine of bladder cancer (BCa) patients, however, the diagnostic and prognostic values of urine SNCG for BCa remain unknown. Here, we used enzyme immunoassay and western blotting to measure urine SNCG levels. Patients with BCa or other urological diseases and healthy controls were enrolled at four Chinese hospitals from April 2010 to November 2014. Diagnostic performance was evaluated by analyzing the area under receiver operating characteristic curves (AUROCs). The AUROC was 0.903 ± 0.019 (95% confidence interval [CI], 0.867 - 0.940) for the test and 0.929 ± 0.015 (95% CI, 0.901 - 0.958) for the validation cohort. The optimal cutoff value yielded sensitivities of 68.4%, 62.4% and specificities of 97.4%, 97.8% for the test and validation cohort, respectively. Urine SNCG levels were decreased after tumor resection, but were higher in BCa patients with recurrence than those without (P = 0.001). The urine SNCG levels in patients with urological benign diseases were significantly lower than BCa patients (all P < 0.05) but higher than healthy controls (all P < 0.05). Hematuria did not interfere with the SNCG detection by spiking urine specimens with whole blood. Compared with a nuclear-matrix-protein-22 assay in an additional cohort excluding hematuria, SNCG showed a similar sensitivity and higher specificity. In summary, our results demonstrated that urine SNCG can discriminate BCa from urinary diseases, and is a useful prognosticator of postsurgical recurrence.


SUPPLEMENTARY DATA Western blotting
For the analysis of SNCG protein expression in bladder cancer tumors, cell lysates were prepared by homogenizing tumor tissues in a buffer containing 10 mM pH 8.0 PBS, 1% Triton x-100 (PBST), and 0.25% sodium deoxycholate and protease inhibitor cocktail (Roche Applied Science, Indianapolis, IN). Cell lysates or urine samples were analyzed by sodium dodecyl sulfatepolyacrylamide gel electrophoresis in a 15% acrylamide gel. The resolved protein bands electrophoretically transferred to nitrocellulose membranes (GE Healthcare, Waukesha, WI), and blocked for 2 h in 4% nonfat dry milk in PBST. The blots were probed using the antibodies indicated. The protein bands were developed using the Super Signal Kit (GE Healthcare), according to the manufacturer's instructions.
Paired urine and tumor samples were analyzed to compare the levels of SNCG expression. The tissue samples were obtained immediately after surgical resection and the serum samples, urine samples were also stored at -80°C. Serum SNCG was detected by the ELISA method, SNCG in tumor tissue was detected by western blotting, and SNCG in urine was valued using both the ELISA and western blotting to validate the ELISA-based quantification of SNCG. Serum samples were diluted 5-fold in assay diluent, and urine was not diluted for the ELISA.

Evaluation of the SNCG ELISA
Prior to the sample analyses, the interassay variability was assessed based on the analysis of 10 plates using three different SNCG standards and urine samples with low, medium, and high SNCG levels. The assay precision was satisfactory, with a mean coefficient of variation (CV) of 2.9% ± 1.9% for intra-assay variation and 8.0% ± 2.1% for interassay variation (Supplementary  Table 1), thereby demonstrating high reproducibility. The spike recovery in human urine ranged from 84.6% ± 6.5% to 101.5% ± 6.0% (Supplementary Table 2). The dilution recovery of SNCG from urine ranged from 97.4% ± 9.7% to 110.5% ± 10.8% (Supplementary Table 3).
We assessed the performance of the sandwich ELISA by evaluating the calibration curve, detection limit, recovery, specificity, precision, and dilution linearity. A series of standards (10 to 0.16 ng/mL) were prepared by diluting purified SNCG in assay diluent. Quality control samples containing three different concentrations of SNCG were analyzed in each microtiter plate. We performed an analytical evaluation of the ELISA-based detection SNCG in urine.
The lowest limit of detection was determined from 20 buffer samples, and calculated as the mean plus three standard deviations. Three different concentrations of SNCG standard and three urine samples (Urine 1 to 3) were analyzed 10 times to determine the intra-assay and interassay variation. Replicate samples were assayed, and the assay variability was expressed as % CV = (SD/ mean) × 100. Urine samples (Urine 4 to 8) were spiked with different amounts of purified SNCG for recovery analysis based on the differences between the expected and observed values. For the linearity analysis, 5 urine samples (Urine 9 to 13), which had high concentrations of SNCG, were diluted 1:2, 1:4, and 1:8, and the percentage of difference between the expected and observed values was calculated.