Multiplexed quantification of 63 proteins in human urine by multiple reaction monitoring-based mass spectrometry for discovery of potential bladder cancer biomarkers

Abstract

Three common urological diseases are bladder cancer, urinary tract infection, and hematuria. Seventeen bladder cancer biomarkers were previously discovered using iTRAQ — these findings were verified by MRM-MS in this current study. Urine samples from 156 patients with hernia (n = 57, control), bladder cancer (n = 76), or urinary tract infection/hematuria (n = 23) were collected and subjected to multiplexed LC–MRM/MS to determine the concentrations of 63 proteins that are normally considered to be plasma proteins, but which include proteins found in our earlier iTRAQ study. Sixty-five stable isotope-labeled standard proteotypic peptides were used as internal standards for 63 targeted proteins. Twelve proteins showed higher concentrations in the bladder cancer group than in the hernia and the urinary tract infection/hematuria groups, and thus represent potential urinary biomarkers for detection of bladder cancer. Prothrombin had the highest AUC (0.796), with 71.1% sensitivity and 75.0% specificity for differentiating bladder cancer (n = 76) from non-cancerous (n = 80) patients. The multiplexed MRM-MS data was used to generate a six-peptide marker panel. This six-peptide panel (afamin, adiponectin, complement C4 gamma chain, apolipoprotein A-II precursor, ceruloplasmin, and prothrombin) can discriminate bladder cancer subjects from non-cancerous subjects with an AUC of 0.814, with a 76.3% positive predictive value, and a 77.5% negative predictive value. This article is part of a Special Section entitled: Understanding genome regulation and genetic diversity by mass spectrometry.

Introduction

Three of the common urological diseases are bladder cancer (BC), urinary tract infection (UTI), and hematuria (HU). Bladder cancer is one of the most common urinary tract carcinomas. According to the most recent estimates of the American Cancer Society, in 2009 there were 70,980 new cases of bladder cancer in the United States and 14,330 deaths from bladder cancer annually [1]. In Taiwan, there were 2050 new cases of bladder cancer (accounting for 2.71% of all cancers) and 804 deaths (1.99% of all cancers) in 2007 [2], [3]. The earlier this cancer is found and treated, the better the outcome [1]. Thus, there is a compelling need to develop more reliable bladder cancer markers for early detection.

Urinary tract infection (UTI) is a bacterial infection that affects any part of the urinary tract and is a common urological disease. The main etiologic agent is Escherichia coli which can get into the bladder or kidney and multiply in the urine, and cause UTI. The most common type of UTI is acute cystitis often commonly referred to as a “bladder infection”. Bacteria may also ascend the ureter to the kidney and establish a secondary infection called acute pyelonephritis [4].

Hematuria (HU), the presence of red blood cells in the urine, can be a sign of a kidney stone or a tumor in the urinary tract (kidneys, ureters, urinary bladder, prostate, and urethra) which can range from minor to lethal [5]. If white blood cells are found in addition to red blood cells, then it is a signal of UTI.

Inguinal hernia is a protrusion of the peritoneum which occurs through the muscles of the anterior abdominal wall at the level of the inguinal canal in the groin. The hernia may extend into the scrotum and can cause discomfort or ache. The main risk factors for inguinal hernia are male sex and increasing age [6]. Since this condition should not affect the urinary protein compositions and levels, hernia patients are used as the control group in this study.

Previously, we used iTRAQ to discover potential bladder cancer biomarkers in urine samples, and found that the levels of numerous classical “plasma” proteins in urine were statistically higher in bladder cancer than in hernia and UTI/HU patients [7]. In the present study, we optimized the protocol for tryptic digestion of urine proteins and applied an LC–MRM/MS approach using 65 SIS peptides to quantify 63 plasma proteins to urine specimens from a total of 156 patients with one of the three common urological diseases (bladder cancer, UTI and/or HU).

Although this study was initially designed to examine the levels of our previously-discovered biomarker proteins in a larger number of samples in order to verify their utility as biomarkers of bladder cancer and other urological diseases, multiplexed MRM-MS allows the simultaneous quantification of a large number of urinary proteins. Therefore these multiplexed MRM-MS experiments are also a biomarker discovery tool that can provide a valuable starting point for future development of urinary protein biomarkers.

Mass spectrometry-based approaches in quantitative proteomics have become a powerful tool for the analysis of complex proteomes [8], [9], [10], [11], [12]. With the incorporation of multi-dimensional fractionation, immuno-enrichment or depletion, the cataloging of proteins has been improved [12], [13], [14], [15], [16]. However, because of the limited sampling rate of mass spectrometers using untargeted analyses, and because of technical variations in multidimensional proteomics studies, conventional bottom-up proteomics has only limited reproducibility for protein quantitation purposes. However, tens to hundreds of candidate biomarker proteins are often listed, and are currently awaiting validation, which requires analysis of clinical specimens from a large cohort of patients. The validation step represents one of the major bottlenecks for translation of protein biomarker data to the clinic.

Enzyme-linked immunosorbent assay (ELISA) is currently the most common method for measuring the concentration of a target protein in biological samples. The high sensitivity and reasonable specificity of ELISA allow the detection of proteins with concentration ranges of low ng/ml to pg/ml in plasma [17]. Some multiplexed immunoassays have been developed for the measurement of several proteins in a single assay [18], [19], [20]. However, commercially-available immunoassays are usually very costly, and multiplexing of ELISAs may be limited due to cross-reactivity of the antibodies. The availability of specific antibodies against novel candidate proteins and time required for development of new ELISA assays create other bottlenecks in the biomarker pipeline [21]. Thus, sufficient sensitivity and multiplexing capability are urgently needed for quantitation of proteins.

Multiple-reaction-monitoring mass spectrometry (MRM-MS) is an MS scanning mode involving two stages of mass analysis, that is commonly performed using triple quadrupole MS instruments. In this technique, specific transitions from precursor peptides to fragment product ions, produced under collision-induced dissociation, are selected. MRM-MS has been widely used as a “gold standard” approach for pharmaceuticals as well as for other low-molecular-weight compounds in a variety of fields [22], [23], [24], [25], [26]. Given the worldwide focus on biomarker discovery, MRM-MS is now being used as an effective tool for accurate quantitation of candidate proteins. Dozens of targeted candidates can be quantified in a single LC–MRM/MS run by detecting Q1/Q3 ion pairs of signature peptides, “proteotypic peptides”, which are representatives of precursor proteins [27], [28], [29], [30], [31]. High to mid ng/ml protein concentrations in unfractionated plasma have been reported with high reproducibility within and across laboratories and instrument platforms [32]. MRM coupled with stable isotope-coded peptides as internal standards (SIS peptides) results in coefficients of variation less than 15% [32].

By incorporating HPLC fractionation, depletion of abundant proteins, and/or peptide immunoaffinity enrichment using antibodies prior to MRM-MS analyses, the limit of quantitation (LOQ) can be close to 1–10 ng/ml [17], [30], [33], [34]. Software algorithms have been developed for optimization of instrumental parameters and to search for putative post-translational-modified sites in a target protein, which further extends the application of this technique [35], [36], [37], [38]. Thus, multiplexed MRM-MS assays offer an attractive alternative to ELISA through the development of a comprehensive panel of biomarkers.

MRM-MS has been used to validate relative abundances of urinary biomarkers discovered in label-free quantitative LC–MS experiments [39], [40]. One of the major challenges in urine biomarker discovery, however, is the high biological variation between individuals. Therefore, after the discovery phase, potential urinary biomarkers must be verified in a large number of samples. Quantitation of urinary proteins in each sample must be part of this verification step, but quantitation of a large number of urine proteins has not yet been systematically performed, and the knowledge of urinary protein concentrations in healthy and diseased patients is still very limited.

Biofluids, including blood plasma, serum, and urine, are routinely used for clinical assays. The profiles of the human urinary proteome, particularly those proteins that are associated with kidney function and the urinary system, vary with disease progression or drug treatment. Thus, interest in exploring the urinary proteome has broadened the search for new biomarkers, as well as disease etiology studies [7], [13], [41], [42], [43], [44]. Protein biomarker discovery in urine has gained prominence in recent years, and levels of specific proteins in urine have been found to increase or decrease during disease in numerous studies, including kidney disease, bladder disease, and cancers [7], [45], [46], [47]. Selevsek and Domon have recently demonstrated the quantification of sixteen urinary proteins using MRM-MS [48].

Our study has yielded information about the concentrations of the major urinary proteins under disease conditions, which allowed us to verify and evaluate the diagnostic efficacy of these proteins as urological disease biomarkers. Using APOA1 as a model protein, we also compared the quantitative results from MRM-MS (peptide level quantitation) and two other immunoassay-based techniques (ELISA and Bio-Plex systems, protein-level quantitation).