Combined proteome and transcriptome analyses for the discovery of urinary biomarkers for urothelial carcinoma.

BACKGROUND: Proteomic discovery of cancer biomarkers in body fluids is challenging because of their low abundance in a complex background. Altered gene expression in tumours may not reflect protein levels in body fluids. We have tested combining gene expression profiling of tumours with proteomic analysis of cancer cell line secretomes as a strategy to discover urinary biomarkers for bladder cancer. METHODS: We used shotgun proteomics to identify proteins secreted by three bladder cancer cell lines. Secreted proteins with high mRNA levels in bladder tumours relative to normal urothelium were assayed by ELISA in urine samples from 642 patients. RESULTS: Midkine and HAI-1 were significantly increased in bladder cancer patients, with the highest levels in invasive disease (area under the receiver operating characteristic curve 0.89 vs non-cancer). The urinary concentration of both proteins was too high to be explained by bladder cancer associated haematuria and most likely arises by direct tumour secretion. CONCLUSIONS: This 'dual-omic' strategy identified tumour secreted proteins whose urine concentrations are increased significantly by bladder cancer. Combined secretome-transcriptome analysis may be more useful than direct proteomic analysis of body fluids for biomarker discovery in both bladder cancer and other tumour types.

Urinary protein profiling by liquid chromatography/tandem mass spectrometry: ADAM28 is overexpressed in bladder transitional cell carcinoma.

Bladder cancer is the most common urological cancer with higher incidence rate in the endemic areas of Blackfoot disease (BFD) in southern Taiwan. The aim of this study was to utilize the proteomic approach to establish urinary protein patterns of bladder cancer. The experimental results showed that most patients with bladder cancer had proteinuria or albuminuria. The urine arsenic concentrations of bladder cancer patients in BFD areas were significantly higher than those patients from non-BFD areas. In the proteomic analysis, the urinary proteome was identified by nano-high-performance liquid chromatography/electrospray ionization tandem mass spectrometry (nano-HPLC/ESI-MS/MS) followed by peptide fragmentation pattern analysis. We categorized 2782 unique proteins of which 89 proteins were identified with at least three unique matching peptide sequences. Among these 89 proteins, thirteen of them were not found in the control group and may represent proteins specific for bladder cancer. In this study, three proteins, SPINK5, ADAM28 and PTP1, were also confirmed by Western blotting and showed significant differential expression compared with the control group. ADAM28 may be used as a possible biomarker of bladder cancer.

Proteomic techniques identify urine proteins that differentiate patients with interstitial cystitis from asymptomatic control subjects.

OBJECTIVE: The purpose of this study was to identify differences in urine proteins between patients with interstitial cystitis (IC) and asymptomatic control (AC) subjects with the use of proteomic techniques. STUDY DESIGN: Nine patients with IC and their age-, race-, and sex-matched AC subjects volunteered a urine specimen. Urine proteins were separated with the use of 2-dimensional polyacrylamide gels. Differing proteins underwent digestion and matrix-assisted laser desorption ionization-time of flight mass spectrometry. Computer-assisted data analysis was used to identify the corresponding protein. Differences in urine protein responses between patients with IC and AC subjects were evaluated by the Mann-Whitney U test to account for the nonnormal frequency distribution of the parameter estimate or chi-square when data were bimodal. RESULTS: Four proteins differed significantly between patients with IC and AC subjects. The AC subjects had a greater concentration of a uromodulin (P = .019) and two kininogens (P = .023, .046). The patients with IC had a greater concentration of inter-alpha-trypsin inhibitor heavy chain H4 (P = .019). CONCLUSION: These urine protein isoforms may be biomarkers for IC.

Identification of potential urine proteins and microRNA biomarkers for the diagnosis of pulmonary tuberculosis patients.

This study identified urinary biomarkers for tuberculosis (TB) diagnosis. The urine proteomic profiles of 45 pulmonary tuberculosis patients prior to anti-TB treatment and 45 healthy controls were analyzed and compared using two-dimensional electrophoresis with matrix-assisted laser desorption/ionization time of flight mass spectrometry. Nineteen differentially expressed proteins were identified preliminarily, and western blotting and qRT-PCR were performed to confirm these changes at the translational and transcriptional levels, respectively, using samples from 122 additional pulmonary tuberculosis patients and 73 additional healthy controls. Two proteins, mannose-binding lectin 2 and a 35-kDa fragment of inter-α-trypsin inhibitor H4, exhibited the highest differential expression. We constructed a protein-microRNA interaction network that primarily involved complement and inflammatory responses. Eleven microRNAs from microRNA-target protein interactions were screened and validated using qRT-PCR with some of the above samples, including 97 pulmonary tuberculosis patients and 48 healthy controls. Only miR-625-3p exhibited significant differential expression (p < 0.05). miR-625-3p was increased to a greater extent in samples of smear-positive than smear-negative patients. miR-625-3p was predicted to target mannose-binding lectin 2 protein. A binary logistic regression model based on miR-625-3p, mannose-binding lectin 2, and inter-α-trypsin inhibitor H4 was further established. This three-biomarker combination exhibited better performance for tuberculosis diagnosis than individual biomarkers or any two-biomarker combination and generated a diagnostic sensitivity of 85.87% and a specificity of 87.50%. These novel urine biomarkers may significantly improve tuberculosis diagnosis.