www.impactjournals.com/oncotarget/ Oncotarget, 2017, Vol. 8, (No. 3), pp: 4960-4976 Research Paper Targeted proteomics in urinary extracellular vesicles identifies biomarkers for diagnosis and prognosis of prostate cancer Tamara Sequeiros1, Marina Rigau1, Cristina Chiva2,3, Melania Montes1, Iolanda Garcia-Grau1, Marta Garcia1, Sherley Diaz4, Ana Celma5, Irene Bijnsdorp6, Alex Campos8, Primiano Di Mauro9, Salvador Borrós9, Jaume Reventós10,11, Andreas Doll1, Rosanna Paciucci1, Michiel Pegtel7, Inés de Torres1,4, Eduard Sabidó2,3, Juan Morote1,5, Mireia Olivan1 1 Group of Biomedical Research in Urology, Vall d’Hebron Research Institute (VHIR) and Universitat Autònoma de Barcelona (UAB), Barcelona, Spain 2 Proteomics Unit, Centre de Regulació Genòmica (CRG), Barcelona, Spain 3 Proteomics Unit, Universitat Pompeu Fabra, Barcelona, Spain 4 Department of Pathology, Vall d'Hebron University Hospital, Barcelona, Spain 5 Department of Urology, Vall d'Hebron University Hospital, Barcelona, Spain 6 Department of Urology, VU University Medical Center, Amsterdam, The Netherlands 7 Department of Pathology, VU University Medical Center, Amsterdam, The Netherlands 8 Sanford-Burnham Medical Research Institute, La Jolla, California, USA 9 Sagetis-Biotech; Grup d'Enginyeria de Materials (GEMAT) Institut Químic de Sarrià, Barcelona, Spain 10 Departement of Basic Science, International University of Catalonia, Barcelona, Spain 11 IDIBELL-Bellvitge Biomedical Research Institute, Barcelona, Spain Correspondence to: Mireia Olivan, email: mireia.olivan@vhir.org Keywords: prostate cancer, biomarkers, urine, extracellular vesicles, diagnosis Received: September 22, 2016     Accepted: November 07, 2016     Published: November 26, 2016 ABSTRACT Rapid and reliable diagnosis of prostate cancer (PCa) is highly desirable as current used methods lack specificity. In addition, identification of PCa biomarkers that can classify patients into high- and low-risk groups for disease progression at early stage will improve treatment decision-making. Here, we describe a set of protein-combination panels in urinary extracellular vesicles (EVs), defined by targeted proteomics and immunoblotting techniques that improve early non-invasive detection and stratification of PCa patients.We report a two-protein combination in urinary EVs that classifies benign and PCa patients (ADSV-TGM4), and a combination of five proteins able to significantly distinguish between high- and low-grade PCa patients (CD63-GLPK5-SPHM- PSA-PAPP). Proteins composing the panels were validated by immunohistochemistry assays in tissue microarrays (TMAs) confirming a strong link between the urinary EVs proteome and alterations in PCa tissues. Moreover, ADSV and TGM4 abundance yielded a high diagnostic potential in tissue and promising TGM4 prognostic power. These results suggest that the proteins identified in urinary EVs distinguishing high- and low grade PCa are a reflection of histological changes that may be a consequence of their functional involvement in PCa development. In conclusion, our study resulted in the identification of protein-combination panels present in urinary EVs that exhibit high sensitivity and specificity for PCa detection and patient stratification. Moreover, our study highlights the potential of targeted proteomic approaches–such as selected reaction monitoring (SRM)–as diagnostic assay for liquid biopsies via urinary EVs to improve diagnosis and prognosis of suspected PCa patients. www.impactjournals.com/oncotarget 4960 Oncotarget INTRODUCTION rather complex, due to the presence of protein aggregates and other types of vesicles (such as microvesicles and Prostate cancer (PCa) is the most frequently apoptotic bodies), which are frequently co-isolated with diagnosed cancer and the second leading cause of cancer the population of interest [15]. Accordingly, the best related death among men in developed countries [1]. procedure for EVs purification remains a discussion Nevertheless, the currently used diagnostic methods for topic in the field, with ultracentrifugation being the most PCa detection are far from ideal. Both, prostate-specific frequently used method [16, 17]. antigen (PSA) serum measurement, and digital rectal Nevertheless, previous reports have indicated examination (DRE) present a low specificity. As a result, urinary EVs to be an excellent source of PCa protein a significant rate of unnecessary prostate biopsies (PB) is biomarkers. In 2009, Mitchell et al. analyzed the practiced [2]. There is therefore still a clear need for new proteomic content of urinary EVs from 10 healthy donors biomarkers for a fast and reliable diagnosis of PCa. and 10 PCa patients who were undergoing hormonal Because of the prostate location in the body, in therapy prior to radical radiotherapy. In this cohort, PSA direct contact with the urethra, prostate (cancer) secrete and Prostate-Specific Membrane Antigen (PSMA) were products can be detected in urine. Consequently, urine found to be present in almost all of the PCa specimens, has been intensively studied as a liquid biopsy source but not in the healthy donor specimens [18]. A later study of biomarkers for PCa [3–6]. However, the low protein characterized EVs preparations in pooled post-DRE urine concentration, the presence of salts and the vast dynamic samples using a shotgun proteomics procedure identifying range of protein expression in urine turn it into a around 900 proteins [19, 20]. Since then, several studies particularly complicated fluid for the discovery of protein- have emerged to determine the role of extracellular based biomarkers [7]. vesicles in urological malignancies [21–23]. A rich source for prostate-derived products in In our study, we used a targeted proteomics urine are extracellular vesicles (EVs) such as exosomes, approach to specifically quantify a set of 64 proteins in 50-150 nm sized membrane vesicles that are shed urinary EVs in the context of PCa biomarker validation in by many mammalian cell types, including malignant a cohort of 107 individual urine-derived human samples. cells and formed within the endosomal network and Our study resulted in the identification of a protein released upon fusion of multi-vesicular bodies with the biomarker combination present in urinary EVs that plasma membrane [8]. In the past decade, EVs have exhibits better sensitivity and specificity than the currently been recognized as potent vehicles of intercellular available biomarkers used for PCa detection and patient communication, both in prokaryotes and eukaryotes. This stratification. Therefore, the protein combination reported is due to their capacity to transfer proteins, lipids and here improves the detection and management of PCa, nucleic acids, thereby influencing various physiological avoiding the over-diagnosis and over-treatment associated and pathological functions of both, recipient and parent with the currently used screening methods. cells [8]. Consequently they have been pointed out as a promising easily accessible biomarker reservoir, as their RESULTS content (such as proteins, lipids, DNA, and RNA) is thought to reflect the molecular composition of their tissue In this work we used a targeted proteomics approach of origin [9–11]. The potential of using EVs as source of to specifically quantify a set of 64 proteins in urinary EVs PCa biomarkers as a strategy to overcome the dynamic in the context of PCa biomarker validation in a cohort of range challenge in urine has generated considerable 107 individual urine-derived human samples. Initially a interest in the last years. When EVs are purified from cohort of 107 urine samples was collected after DRE and the whole urine sample, their protein content is highly EVs were isolate. We then measured protein abundance enriched while at the same time high-abundance soluble changes in urinary EVs by SRM, and we built protein- proteins are removed, enhancing the detectability of low- based panels for PCa diagnosis and prognosis in urinary abundance proteins [12, 13]. The analysis of the content EVs, which we also checked in tissue microarrays of EVs harvested from urine appears to have important analysis. advantages such as i) urine collection is a non-invasive procedure, ii) the proteomic and genomic material within EVs isolation from post-DRE urine samples EVs is protected from enzyme degradation by the vesicles lipid bilayer [14], and iii) EVs are stable after long-term Several approaches were followed in order to storage at -80°C, which makes prospective studies feasible assess the purity of the isolated urinary EVs. Urine EVs [13]. were observed using a transmission electron microscope One of the most significant challenges involving (TEM). The TEM images showed vesicles of the expected the use of EVs for the discovery of new biomarkers is size, presenting the cup-shaped morphology typical the lack of standardized and reliable isolation methods. of EVs (Figure 1A) [24]. In addition, NTA and protein The isolation of exosomes from biological fluids is quantification data for all samples (n=107) were used www.impactjournals.com/oncotarget 4961 Oncotarget to assess the protocol performance. A linear regression (Table 2), with all proteins exhibiting higher levels in low- analysis showed a good correlation (r=0.85 with p<0.001) grade patients (44 proteins). between the number of vesicles and total amount of protein To evaluate whether the proteins that were extracted, thus indicating that at least an important part of significantly changing in abundance as found by SRM- the protein content in the EVs pellet did indeed belong to proteomics could be validated by an independent the vesicles quantified by NTA (Figure 1B). Finally, the method (Figure 2A-2B), we performed antibody-based presence of known EVs markers, such as TSG101 (tumor immunoblotting in a subset of urinary EVs samples from susceptibility gene 101), CD81 or Rab5, was confirmed by the same cohort of patients. Among the known EVs Western blot (Figure 1C). markers such as CD9, TSG101 or CD63, protein CD81 was used as an EVs-related marker in our study. Although Validation of protein abundance changes in these are generally accepted by the scientific community urinary EVs by SRM as proteins commonly found in EVs [25], they have been found to present different abundance levels in the presence The 107 urine samples collected after DRE were of disease [11]. The results obtained were in agreement divided into two groups: PCa patient samples (n=53), with the protein quantitation obtained by SRM (Figure which include 22 low-grade PCa (Gleason score ≤ 7 2C-2D). (3+4)) and 31 high-grade PCa (Gleason score ≥ 7 (4+3)); and control samples (n=54). Protein-based panels for PCa diagnosis and A total of 64 protein candidate biomarkers were prognosis in urinary EVs selected for quantitation by SRM (Supplementary Table ST1). One proteotypic peptide was selected per protein Individual ROC curves and the AUC values were and a scheduled-SRM method was used to allow the calculated for each individual quantified protein. Previous monitoring of all 64 proteins in one single run. To ensure studies described that combining multiple proteins might that the selected peptides correctly represented the true improve the diagnostic performance over the use of single fold-changes of the targeted proteins, we used a previously biomarkers, as single markers may not necessarily reflect in-house shotgun dataset to identify quantotypic peptides the multifactorial nature of PCa neither have the necessary for the selected proteins (see Methods section). As a result prediction power for patient classification [26]. Therefore, of these SRM experiments, we observed that 14 out of in order to compute a protein combination with better the 64 initial proteins exhibited a significant different diagnostic power we fitted abundances of differentially abundance between benign and PCa samples (Table 1), expressed proteins (Figure 3A–3B) into a logistic of which 11 had higher protein levels in urine EVs of PCa regression model and checked the AUC for each protein patients while 3 showed higher levels in urine EVs of combination. We identified a protein combination formed control patients. Interestingly, when low- and high-grade by TGM4 and ADSV (Figure 3C), which had better PCa samples were compared, 45 out of the 64 measured sensitivity and specificity (AUC = 0.65, CI95: 0.55-0.76) proteins showed significantly different abundance levels while the accuracy obtained by each individual protein Figure 1: Quality assessment of the EVs isolation process. A. Transmission electron microscopy images of EVs isolated from post-DRE urine; B. Linealregression analysis showing a good correlation between the number of EVs counted by NTA and the total amount of protein recovered from the same sample; C. Western blot analysis of TSG101, CD81 and Rab5, described in the literature as EVs markers, were performed in EVs isolated from post-DRE urine. Benign and PCa samples were pooled to obtain representative and sufficient material. www.impactjournals.com/oncotarget 4962 Oncotarget Table 1: Targeted proteins exhibiting abundance changes in urinary exosomes between PCa patients and benign controls Gene symbol Protein Uniprot accession FDR FC (PCa vs. Benign) ADSV Adseverin Q9Y6U3 0.002 1.34 N-acetylglucosamine-6- GNS P15586 0.002 1.40 sulfatase TGM4 Transglutaminase-4 P49221 0.002 0.60 CA4 Carbonic anhydrase 4 P22748 0.002 1.40 Tumor susceptibility gene TSG101 Q99816 0.011 1.22 101 protein Vacuolar protein sorting- VPS28 associated protein 28 Q9UK41 0.011 1.20 homolog Choline transporter-like SLC44A4 Q53GD3 0.013 1.24 protein 4 SLC26A4 Pendrin O43511 0.020 1.21 ITGB3 Integrin β3 P05106 0.020 1.22 PRSS8 Prostasin Q16651 0.020 0.83 SDCBP2 Syntenin-2 Q9H190 0.020 1.20 TSPAN9 Tetraspanin-9 O75954 0.022 1.17 ITGAV Integrin αV P06756 0.047 1.20 PPAP Prostatic acid phosphatase P15309 0.047 0.75 Proteins showing significant protein abundance changes in the SRM analysis (adjusted p-value < 0.05) between benign and PCa patients are shown. Proteins with higher abundance in PCa are represented by fold change values > 1, while values < 1 indicate decreased protein levels in PCa. False discovery rate (FDR); fold change (FC). Table 2: Targeted proteins exhibiting abundance changes in urinary exosomes between high and low-grade patients Gene symbol Protein Uniprot accession FDR FC (High vs. Low) Prostatic acid PPAP P15309 < 0.001 0.37 phosphatase Prostate-specific PSA P07288 < 0.001 0.44 antigen CD63 CD63 antigen P08962 < 0.001 0.46 N-sulphoglucosamine SPHM P51688 < 0.001 0.45 sulphohydrolase Putative glycerol GLPK5 Q6ZS86 < 0.001 0.47 kinase 5 FAM177A1 Protein FAM177A1 Q8N128 < 0.001 0.46 Golgi-associated plant GLIPR2 pathogenesis-related Q9H4G4 < 0.001 0.50 protein 1 DNASE1 Deoxyribonuclease-1 P24855 < 0.001 0.54 (Continued ) www.impactjournals.com/oncotarget 4963 Oncotarget Gene symbol Protein Uniprot accession FDR FC (High vs. Low) Metalloreductase STEAP2 Q8NFT2 < 0.001 0.58 STEAP2 TOLLIP Toll-interacting protein Q9H0E2 < 0.001 0.55 Phospholipid- ATP8B1 O43520 < 0.001 0.70 transporting ATPase IC Ras-related protein RRAS P10301 < 0.001 0.67 R-Ras Calcium and integrin- CIB1 Q99828 < 0.001 0.69 binding protein 1 SLC26A2 Sulfate transporter P50443 < 0.001 0.65 CD82 CD82 antigen P27701 < 0.001 0.73 GALK1 Galactokinase P51570 < 0.001 0.69 TOM1L2 TOM1-like protein 2 Q6ZVM7 < 0.001 0.64 V-type proton ATPase ATP6V0D1 P61421 < 0.001 0.68 subunit d 1 Transmembrane TMPRSS2 O15393 < 0.001 0.65 protease serine 2 Lipid phosphate PPAP2A O14494 < 0.001 0.57 phosphohydrolase 1 GNG4CRYZDBNL Drebrin-like protein Q9UJU6 0.001 0.82 40S ribosomal protein RPSA P08865 0.001 0.74 SA GSS Glutathione synthetase P48637 0.001 0.71 Metalloreductase STEAP4 Q687X5 0.002 0.68 STEAP4 Glycogen PYGL phosphorylase, liver P06737 0.002 0.74 form Bis(5'-nucleosyl)- NUDT2 tetraphosphatase P50583 0.002 0.65 [asymmetrical] TSPAN9 Tetraspanin-9 O75954 0.003 0.80 UBC Polyubiquitin-C P0CG48 0.004 0.73 TMBIM1 Protein lifeguard 3 Q969X1 0.004 0.70 FAM49B Protein FAM49B Q9NUQ9 0.005 0.79 Tyrosine-protein PTPN13 phosphatase non- Q12923 0.006 0.78 receptor type 13 Lysophosphatidic acid LPAR3 Q9UBY5 0.007 0.72 receptor 3 DPP3 Dipeptidyl peptidase 3 Q9NY33 0.010 0.73 N-acetylglucosamine- GNS P15586 0.013 0.74 6-sulfatase (Continued ) www.impactjournals.com/oncotarget 4964 Oncotarget Gene symbol Protein Uniprot accession FDR FC (High vs. Low) Mannose-6-phosphate MPI P34949 0.015 0.76 isomerase TSG101DPYS Dihydropyrimidinase Q14117 0.021 0.77 Ethanolamine- PCYT2 phosphate Q99447 0.021 0.80 cytidylyltransferase VTN Vitronectin P04004 0.024 1.31 Alpha/beta hydrolase PDCD6IPVPS26AABHD17C domain-containing Q6PCB6 0.031 0.87 protein 17C ADSV Adseverin Q9Y6U3 0.034 0.81 Proteins showing significant protein abundance changes in the SRM analysis (adjusted p-value < 0.05) between high- and low-grade PCa patients are shown. Proteins with higher abundance in high-grade patients are represented by fold change values > 1, while values < 1 indicate decreased protein levels in high grade patients (i.e. patients with bad prognosis). False discovery rate (FDR); fold change (FC). was low (TGM4 AUC = 0.58; ADSV AUC = 0.58). This proteins were generated, obtaining an AUC of 0.81 (CI95: result evidenced that the combination of two different 0.74-0.88) for TGM4, and an AUC of 0.73 (CI95: 0.65- proteins improved the ability to distinguish benign from 0.81) for ADSV (Figure 4C). Additionally, taking into PCa samples. account the clinical progression of each patient, we tested Next, we also evaluated protein combinations ADSV and TGM4 as potential prognostic biomarkers and that could correctly classify low and high-grade tumors. assessed their performance in distinguishing between i) Individual markers such as CD63 (AUC = 0.65), GLPK5 low- versus high-grade patients (as defined for the urinary (AUC = 0.64), PSA (AUC = 0.66), PPAP (AUC = 0.64) EVs experiments); ii) groups of patients who presented and SPHM (AUC = 0.61) were among the single-protein biochemical recurrence (BCR) versus patients who did panels with best sensitivity and specificity. However, it not. The analyses revealed that TGM4 could indeed was the combination of these five protein abundances that differentiate between the low and high-grade groups with rendered the best protein panel to distinguish patients with high sensitivity and specificity (p < 0.001; AUC = 0.82, low-grade and high-grade tumors (PPAP + PSA + CD63 + CI95: 0.71-0.92) (Figure 4D–4E) as well as between SPHM + GLPK5; AUC = 0.70, CI95: 0.56-0.84) (Figure the BCR versus no BCR groups (p < 0.001; AUC = 3D). 0.80, CI95: 0.69-0.91) (Figure 4F–4G). These results suggest that the proteins identified in urinary EVs reflect Tissue microarrays analysis for tissue validation histological changes and may have a functional role in of PCa biomarkers PCa development since we also find them altered in PCa tissue. The most promising EVs biomarkers found in the The profile of urinary EVs proteins obtained by protein combination models, such as ADSV and TGM4 SRM analysis after comparing low versus high-grade (for the benign vs. PCa comparison) and CD63, GLPK5 PCa patients (CD63, GLPK5 and SPHM), was also and SPHM (for the low-grade vs. high-grade comparison) validated in tissue samples by using TMAs. CD63 was were further evaluated in patient tissues from radical detected at significantly different levels in tissue between prostatectomies using tissue microarrays (TMAs). low- versus high-grade PCa patients (p = 0.021), but In agreement with the expression in urine EVs, resulted a modest patient classifier (AUC = 0.62, CI95: IHC on prostate TMA-tissues revealed that both PCa 0.51-0.74) (Figure 5A–5C). Similarly, when classifying diagnostic biomarkers, ADSV and TGM4, were also patients with BCR versus patients without BCR, GLPK5 altered in terms of protein detection (ADSV p < 0.001; was able to differentiate between the two groups (p = TGM4 p < 0.001) when comparing 98 benign prostatic 0.013; AUC = 0.64, CI95: 0.54-0.73) (Figure 5D–5F) tissues versus 136 PCa tissues (Figure 4A–4B). We then and, furthermore, taking into account only the patients analyzed the ability of these two proteins to correctly of Gleason 7, the patients that had been developed BCR classify patients when directly detected in prostate tissue. showed significantly lower levels of GLPK5 (p = 0.0015) ROC curves for both, ADSV and TGM4, individual (Supplementary Figure SF1). Finally, SPHM did not www.impactjournals.com/oncotarget 4965 Oncotarget show, neither significant value when comparing low- biomarkers is under intense investigation, especially versus high-grade PCa patients, nor when comparing PCa because they can be obtained by non-invasive methods patients with and without BCR. using urine samples. Recently, transcriptomic analysis of urinary EVs DISCUSSION from PCa patients has been carried out with promising results for Cadherin 3, which shows a decreased Nowadays, there is still a clear clinical need to abundance in PCa samples as a source of biomarkers [27] identify new biomarkers that improve the early non- and the use of microRNAs as markers for this disease invasive detection and stratification of PCa patients. has also been extensively reported [28]. In this work, we Within this context, the use of EVs as source of new have performed targeted proteomics and immunoblotting Figure 2: Abundance changes and diagnostic/prognostic evaluation of targeted proteins in urinary EVs. Volcano plots represent differentially expressed target proteins in urinary EVs from A. PCa vs. benign patients and B. from high- vs. low-grade patients; C. Western blot of TGM4 and ADSV in a selected set of benign and PCa samples; D. Western blot of CD63, SPHM, TSG101 and PSA in a selected set of low and high-grade PCa samples. CD81 was used as EVs related marker. www.impactjournals.com/oncotarget 4966 Oncotarget techniques to address this question. Our results show that have been analyzed across multiple studies; the AUC alterations detected in urinary EVs reflect the protein value of total serum PSA (biomarker used in the clinic) changes in the prostate tissue and highlight the potential has been estimated to be around 0.6 [30]. of SRM targeted proteomics as diagnostic assay for liquid In our study, a total of 64 previously identified biopsies via urinary EVs to improve diagnosis and detect candidate biomarkers for PCa in urinary EVs were PCa patients with a poor prognosis. validated in a large cohort of samples (n=107) by SRM. Whereas immunological methods like TMA or ELISA PCa diagnosis and prognosis protein-based represent the traditional way of validation, targeted MS- panels in urinary EVs based approaches like SRM are emerging as additional alternatives [31]. In this study, SRM methodology Current diagnostic techniques for PCa are based on was applied in order to facilitate the simultaneous a measurement of serum PSA and DRE, but decisive PCa quantification of a large number of candidate proteins in a diagnosis is based on PB, which is indicated when patients large cohort of samples. The presence of several of these present serum PSA above 4 ng/mL. Furthermore, repeat PB proteins had already been described in prostate-related is indicated for patients who have a prior negative biopsy urinary EVs by others [19], which proved to be good but continue to have an elevated serum PSA or abnormal candidates as PCa biomarkers. DRE, or as follow-up of previous pathologic diagnoses of In terms of diagnosis, our urinary EVs diagnostic pre-malignant HGPIN [29]. The lack of specificity of the panel (AUC=0.65) slightly improves the diagnostic above-mentioned tools urges for the identification of new performance of serum PSA. However, considering the biomarkers using non-invasive methods that improve the promising potential that EVs have shown as a source of early diagnosis of PCa. Although different PSA isoforms biomarkers, further studies will be conducted to verify Figure 3: Protein-based panels for PCa diagnosis and prognosis. A-B. Scatter plots representing the protein abundance levels of selected proteins that are part of diagnostic and prognostic panels, respectively; C-D. ROC curves of protein-based panels for PCa diagnosis (ADSV + TGM4) and prognosis (CD63 + GLPK5 + SPHM + PSA + PPAP). www.impactjournals.com/oncotarget 4967 Oncotarget Figure 4: Assessment of PCa diagnostic biomarkers in tissue microarrays. A. Tissue microarrays images with immunohistochemistry results for ADSV and TGM4; B. Scatter plots representing immunohistochemistry scores (IHC) of ADSV and TGM4 in benign and PCa FFPE tissue from PCa patients; C. Diagnostic performance represented by a ROC curve of ADSV and TGM4 individually. Prognostic value of TGM4 represented as IHC score (D) and ROC curve (E) between low vs. high-grade patients. Relation between TGM4 and PCa progression is showed as IHC score (F) and ROC curve (G) between patients with and without biochemical recurrence (BCR). Values that are significantly different by the Mann-Whitney test from the control group are indicated by p-value < 0.001 (*** vs. benign; +++ vs. No BCR). www.impactjournals.com/oncotarget 4968 Oncotarget if this diagnostic panel, in combination with current Changes in EVs protein abundances reflect screening tools, could increase the efficiency in detecting alterations in prostate tissue PCa and avoid unnecessary biopsies. From a prognosis point of view, controversy TMAs of PCa patients were used to directly detect exists regarding the potential of PSA as a prognostic each selected protein profile and determine if the observed biomarker. Although often serum PSA has higher protein content in urinary EVs reflected changes in the expression in more aggressive PCa patients [32], it prostate tissue. All patients included in the study presented is often found decreased together with several EVs > 7 years of follow up after surgery, therefore, we were markers, including PSMA, CD9, TSG101, miR-21 able to correlate the results not only with the presence of and miR-375 among others. In agreement with these the disease and Gleason score but also with their clinical observations, we detected decreased PSA levels in progression (BCR after RP). urinary EVs [18, 33]. The abundance of PSA within The protein abundance changes observed in urinary EVs in combination with the protein abundance urinary EVs of ADSV and TGM4 were validated in of four more proteins (CD63, GLPK5, SPHM and prostate tissue by IHC analysis and resulted to have even PAPP) results in an AUC value of 0.70, for detecting a better AUC value. This result suggest that developing PCa patients with a poor prognosis. These findings are new techniques to pre-selecting certain subtypes of in agreement with a previous report describing that a vesicles may lead to better diagnosis performance. combination of PSA to CD63 or CD9 improves the ADSV, also known as scinderin (SCIN), has been detection of PCa. Whether the decrease in prostate- implicated in the translocation of secretory vesicles related markers is due to a decrease in the quantity of directed to be exocytosed [35, 36]. ADSV has been prostate secreted EVs number within urine, or a change found highly expressed in human PCa tissue, and has in protein expression profile within prostate (cancer) been described to be critical for the proliferation of PCa secreted EVs remains to be clarified [34]. cells [37]. Figure 5: Assessment of aggressive PCa biomarkers in tissue microarrays. A-B. Scatter plots representing immunohistochemistry scores (IHC) of CD63, GLPK5 and SPHM in PCa FFPE tissue from low and high-grade PCa patients; C. ROC curve of CD63 biomarker performance; D-E. Scatter plots representing IHC scores of CD63, GLPK5 and SPHM in PCa FFPE tissue from patients with and without biochemical recurrence (BCR); F. ROC curve of GLPK5 biomarker performance. Significantly different values were assessed by the Mann- Whitney test and are indicated by *p-value < 0.05. www.impactjournals.com/oncotarget 4969 Oncotarget Table 3: Clinico-pathological conditions of patients included in the study SRM experiments (urine) TMAs Validation (FFPE) Benign PCa Benign PCa No. of samples 54 53 98 136 Age (yr)* 65.6 (53 - 78) 67.7 (51 - 87) 64 (53-72) 64 (53-73) Serum PSA (ng/mL)* 8.1 (2.5 - 41.1) 17.3 (1.0 - 245.4) 9,8 (2,5-48) 11 (1,1-66) 5,1*1012 (5,3*1011- 4,1*1012 (4,6*1011- No. of vesicles/mL* - - 2,2*1013) 1,4*1013) Total protein (μg/mL)* 1.1 (0.2 - 8.0) 0.8 (0.2 - 4.9) - - Low Grade (GS ≤ - 22 51 50 7(3+4)) High Grade (GS ≥ - 31 48 86 7(4+3) No BCR - - 49 72 BCR - - 49 64 * Values are represented as mean (range); GS (Gleason's Score) In addition, the IHC analysis also identified prostate epithelium and detected at lower levels in a relationship between TGM4 protein levels and PCa [41]. These observations are consistent with the Gleason score as well as BCR development decreasing levels we have detected for these two proteins in the significantly in those patients with a poor prognosis. urinary EVs. Among the prognostic proteins, it is TGM4 is a protein almost uniquely expressed in the worth mentioning that CD63 is mainly associated prostate gland [38] and it has been described to be with membranes of intracellular vesicles, although down-regulated in PCa tissue [39]. This protein was cell surface expression may also be induced. It has previously measured in urinary secretions from patients been shown that prostate basal epithelial cells do not with extra-capsular or organ-confined PCa. Although its express the characteristic CD antigens of secretory diagnostic power was not evaluated, TGM4 was shown to cells, however the expression levels of CD63 found in be down-regulated in poor prognostic PCa [40]. Despite cancer cells are similar to that of secretory cells [44]. the controversy about the expression of TGM4 in PCa, Surprisingly, our study revealed that decreased protein these previous studies combined with our findings may levels of CD63 in EVs fractions as well as in prostate indicate a possible role of TGM4 as a tumor suppressor tissue could indicate poor prognosis of PCa. Similar gene and, after a further validation in a larger cohort of results were obtained with GLPK5 (glycerol Kinase 5) patients, its potential use not only as a PCa diagnostic taking into account the BCR development, however, it biomarker in urinary EVs, but also to identify the poor is worth noting that this protein was able to distinguish prognosis patients using biopsies or tissue samples from patients with Gleason score 7 who developed BCR from radical prostatectomy. Thus, ADSV and TGM4 appear to those who did not, therefore, this candidate could be be promising candidates to be confirmed in future studies useful in predicting tumor behavior in patients with aimed to establish their possible roles in PCa clinical uncertain risk of progressing. From a functional point outcome. of view, further studies will be required to understand The defined prognostic urinary EVs protein the role of this protein in PCa development and profile (CD63, GLPK5 and SPHM) was also validated progression. by TMA. PSA and PPAP were discarded from the TMA study due to extensive previous research already Urinary EVs as a source of disease biomarkers available in the literature. PSA is strongly expressed in liquid biopsies in the prostate, both in benign and neoplastic tissue. However, IHC staining detected a lower abundance in PCa is a clinically heterogeneous and often cancer compared to the adjacent benign epithelium, multifocal disease with a clinical outcome difficult which decreased according to less differentiate PCa to predict. Proteomics data give a different level of [41–43]. Similarly, PPAP immunoreactivity has also understanding and together with the innovative high- been demonstrated to be more intense in the benign throughput technologies is a promising way to identify www.impactjournals.com/oncotarget 4970 Oncotarget new biomarkers for PCa detection, prognosis and therapy Urine samples collection and extracellular [45]. In this work we combined this strategy with the vesicles isolation isolation of urinary EVs to identify new protein biomarker panels for PCa diagnosis (TGM4 and ADSV) and for Urine was collected after DRE, a procedure included in patient stratification (PSA, PPAP, CD63, GLPK5 and the standard process for detection of PCa. Urine (30-50 mL) SPHM). was collected in urine collection cups, kept on ice, transported The strategy presented in this manuscript to the lab and processed within 2 h of its collection. The highlights the potential of liquid biopsies via urinary samples were centrifuged at 2,500 g for 10 min at 4 °C and EVs isolation for the diagnosis and PCa stratification, the supernatant, containing the EVs, was supplemented with and its associated clinical research. Biomarker a cocktail of protease inhibitors (Sigma-Aldrich, St Louis, detection from urinary EVs is capable of distinguishing MO, USA) and stored at –80 °C until its use. aggressive from clinically insignificant PCa and other For EVs isolation, cell-free urine samples were first benign conditions beyond serum PSA and, thus, it centrifuged at 16,500 g, 20 min, to remove larger vesicles might avoid PCa related over-diagnosis and over- and any possible remaining cell debris. The pellet from this treatment. Taking all together, we conclude that the centrifugation was treated with dithiothreitol (DTT) (37 future of the diagnosis and prognosis in PCa will °C, 10 min) in order to break THP (Tamm-Horsfall protein) benefit from research of biomarkers in EVs, leading to fibers and release EVs, and centrifuged again at 16,500 g, an improvement in therapeutic decision making with 20 min. Supernatants from the two centrifugations were liquid biopsies. mixed together and filtered through a 0.2 μM pore size filter. Samples were then ultracentrifuged at 100,000 g for 120 min at 4 °C. The resulting pellet was washed with METHODS PBS (phosphate buffered saline) and centrifuged again at 100,000 g, 60 min at 4 °C. The final pellet was resuspended Experimental design and statistical rationale in 50 μL PBS, of which 5 μL were set aside and stored at -80 °C for nanoparticle tracking analysis (NTA) and the This is a retrospective case-control study to rest was mixed with 50 μL lysis buffer (Tris 20 mM pH identify biomarkers associated to PCa. The main 8.8, NaCl 150 mM, EDTA 5 mM, Triton X-100 1%, and objectives are to assess the potential of urinary EVs as protease inhibitors) and stored at -20 °C until its use. a source of PCa biomarkers and to evaluate a different biomarker candidates previously identified in a larger Transmission electron microscopy cohort of patients. The validation cohort was integrated by a total of 107 urine samples obtained after DRE Transmission electron microscopy (TEM) imaging (Table 3). For statistical analysis, PCa cases were divided of EVs was performed using negative-staining technique. into low grade PCa (Gleason score ≤ 7(3+4); n=22) and Briefly, pellets recovered from the ultracentrifugation, high grade PCa (Gleason score ≥ 7(4+3); n=31). The containing the EVs, were fixed with 4% paraformaldehyde control group included patients without PCa (n=54), and then deposited on Formvar/Carbon-coated grids, such as patients with benign prostatic hyperplasia (BPH), which were negatively stained with uranyl acetate. EVs inflammation or high-grade prostatic intraepithelial preparations were examined using a transmission electron neoplasia (HGPIN). microscope JEOL JEM 1010 (Japan Electron Optics Laboratory Co., Tokyo, Japan). Patients selection and inclusion criteria Nanoparticle tracking analysis This study obtained approval from the Vall Hebron University Hospital (Barcelona, Spain) institutional Vesicles present in purified samples were analyzed by review board (PR(IR)56/2014). Written informed NTA using the NanoSight LM14 system (NanoSight Ltd., consent was obtained from all the study participants Amesbury, UK), configured with a high sensitivity digital and samples were coded to ensure sample tracking and camera system (Hamamatsu C11440 ORCA-Flash2.8, confidentiality on patient/donor identity. All patients Hamamatsu City, Japan). Three videos of 60s duration were were men with suspicion of PCa according to abnormal recorded for each sample and average measurements and DRE and/or serum PSA levels higher than 4 ng/mL, standard deviations were calculated. Videos were analyzed referred for a first PB at the Urology Service of the Vall using the NTA-software (version 2.3), with the minimal d’Hebron University Hospital (Barcelona, Spain) from expected particle size, minimum track length, and blur set 2008 to 2012. Definitive diagnosis was achieved after to automatic. Camera shutter speed and camera gain were PB. Patients with a diagnosis of PCa and age matched set to maximum. Camera sensitivity and detection threshold controls were selected. Patients with an unrelated chronic were set close to maximum (15 or 16) and minimum (2 to or acute severe illness and/or previous PCa therapies 5), respectively, to reveal small particles. Samples were were excluded from the study. diluted in ultrapure water. www.impactjournals.com/oncotarget 4971 Oncotarget Sample preparation for mass spectrometry All samples were individually digested (see above) and one isotopically labeled reference peptide at Urinary EVs were disrupted by sonication C-terminal lysine (13C6,15N2-Lys) or arginine (13C6,15N4- (LABSONIC M, Sartorius Stedim Biotech, Goettingen, Arg) per protein was spiked into the digested samples Germany) at 100 % amplitude for 4 pulses of 5 s each prior mass spectrometry (MS) acquisition (Supplementary separated by 5 s pauses on ice. The extracted proteins Table ST1). were stored at –20 °C. An aliquot of each preparation was Selected reaction monitoring (SRM) measurements used for protein quantity estimation using the DC Protein were performed with unfractionated samples on a Assay (Bio-Rad, Bio-Rad Laboratories, Hercules, CA, hybrid triple quadrupole / ion trap mass spectrometer USA), following the manufacturer’s instructions. Samples, (5500 Q-Trap; AB Sciex Instruments, Foster, CA, USA) usually diluted 1:10, were compared in triplicates against equipped a reversed-phase chromatography 25-cm column serially diluted Bovine Serum Albumin (BSA) as standard. with an inner diameter of 75 μm, packed with 1.9-μm C18 Filter-aided sample preparation (FASP) was particles (Nikkyo Technos, Tokyo, Japan) and a 2-cm performed using a 10 kDa molecular weight cut-off filter pre-column (Acclaim PepMap 100, C18, 15 μm, 100- (Merck Millipore) as previously described [46]. Briefly, A; Dionex, Sunnyvale, CA, USA). Loading buffer: H2O 20 μg of sample in RIPA buffer were loaded in the filter + 0.1% formic acid; Elution buffer: ACN + 0.1% formic unit and washed twice with 8 M urea by centrifugation acid. Flow rate: 250 nL/min. Gradient: From 7 to 40% at 14,000 g for 15 min. Proteins were reduced with eluting buffer in 60 min. Blank runs were performed 10 mM DTT for 1 h at RT, and alkylated with 30 mM between the SRM measurements of biological samples iodoacetamide (IAA) 30 min in the dark. The reaction was to avoid sample carryover. Measurements were done in stopped with 37.5 mM n-acetylcysteine (NAC) during 15 scheduled SRM mode, using a SRM detection window of min, and the solutions were removed by centrifugation 300 seconds and a total cycle time of 2.5 seconds. For at 14,000 g for 15 min. The samples were then washed each peptide of interest five transitions were monitored for once with 1 M urea. The resulting pellet was diluted with both the endogenous and the isotopically labeled peptides . 40 μL of 1 M urea, containing 20 μg of trypsin, and it Transition groups corresponding to the targeted was incubated overnight 37 °C for protein digestion. peptides were evaluated with Skyline v2.5 based on Finally, tryptic peptides were collected in a clean tube the correlation of transition intensities between the by centrifugation at 14,000 g for 10 min, and this filtrate endogenous and reference SRM traces as well as with the was acidified with 0.3-0.5 μL of concentrated formic acid. reference spectral library, on the relative retention times Samples were stored at –20 °C until further analysis. across runs, and on the co-elution of endogenous peptide and spiked-in references [47]. Transition areas were used Selected reaction monitoring to calculate protein ratios between the two groups using the linear mixed-effects model as implemented in the R A total of 64 proteins were selected for the targeted package MSstats v2.0 [48]. proteomic analysis based on previous in-house discovery The raw proteomics data have been deposited to the proteomics experiments and additional information ProteomeXchange Consortium [49] via the PASSEL [50] retrieved from the literature. Literature search was focused partner repository with the dataset identifier PASS00843. on previous studies in which differentially expressed proteins in PCa had been identified using tissue, urine and Western blotting different PCa cell lines as a source of biomarkers [11, 19, 21, 23]. Proteins were separated by 10% SDS-PAGE under One quantotypic peptide per protein was selected reducing or non-reducing conditions and transferred based on a previously in-house shotgun dataset. Briefly, to PVDF membranes. For blocking, membranes were to ensure that the selected peptides correctly represented soaked in 5% non-fat dried milk in TBS-Tween20 the true fold-changes of the targeted proteins, we extracted (Tris-buffered saline, 0.01% Tween20). Proteins were all ion intensity chromatograms from MS1 scans for the immunodetected using antibodies against TSG101 identified peptide precursors of the selected proteins (1:500; Abcam, Cambridge, UK), CD81 (1:100; Santa across 24 samples. All the ions were visually inspected Cruz Biotechnology, Santa Cruz, CA, USA), CD63 (non- and area integration of all the precursor ions was manually reduced conditions, 1:1000, Merck Millipore, Germany), reviewed. The integrated areas were used to statistically PSA/KLK3 (1:100, Dako, Denmark), TGM4, ADSV, assess protein fold-changes. Peptides were selected for the SPHM (1:1000, 1:100, 1:100, Sigma-Aldrich, MO, USA), SRM experiment according the following criteria: i) the in overnight incubations at 4°C. Afterwards, membranes difference between the protein and the peptide fold-change were washed and incubated with a secondary HRP- (the lower the better); and ii) the intensity of the peptide coupled antibody (horseradish peroxidase-conjugated (the more intense the better). In the case of proteins that antibody) for 1h at room temperature. Finally, HRP signal exhibited not significant abundance changes, peptides was revealed using the Immobilon Western HRP Substrate selection was based on peptide intensity. (Merck Millipore, Darmstadt, Germany). www.impactjournals.com/oncotarget 4972 Oncotarget Tissue microarrays positive) + (3x % cells staining strongly positive) to a maximum of 300. For the tissue microarrays (TMAs) a total of 165 samples of radical prostatectomy (RP) were selected Statistical analysis of mass spectrometric data from patients with > 7 years of clinical follow up. Low and high-grade patients and patients with and without SRM peak areas were normalized based on the areas biochemical recurrence (BCR) were included. BCR is corresponding to the isotopically labeled internal peptide defined as the first post-operative PSA value > 0.4 ng/ references, which were used to first equalize the median mL, confirmed by at least 1 subsequent increasing value abundance (log2-areas) for all reference peptides across (persistent PSA increase) after undetectable PSA post- all runs, and then all endogenous peptide areas in a run operatively. From each PCa patient, both cancerous and were corrected by a same bias. Comparisons of relative benign peripheral tissues were inspected. Formalin- protein abundance between groups were performed using fixed paraffin-embedded tissue (FFPE) samples of PCa the MSstats R package [48]. available from the archives of the Pathology Department For predictive analysis, the whole patient cohort of the Vall Hebron University Hospital (Barcelona, was used. Initially, the quantity of each protein per sample Spain) were used. Clinico-pathologic features of these was estimated based on a relative log2-transformed using patients are detailed in Table 1 . For TMA construction, the MSstats R package, and these protein quantities were a hematoxylin and eosin-stained section was made from then used as input variables to a logistic regression model each block to define representative tumor regions. Tissue between the tested patient groups. Each protein was fitted cylinders with a diameter of 1 mm were then punched in the model between two groups and its classification from selected tumor and non-malignant areas of each ability was evaluated by the area under the curve donor tissue block in triplicates and brought into a (AUC) obtained in the receiver operating characteristic recipient paraffin block using a custom-made precision (ROC) curve. AUC was used as it summarizes well the instrument (Advanced Tissue Arrayer Chemicon classification performance of each protein in terms of International USA). Sections (3 μm) of the resulting TMA specificity and sensitivity, with candidates with better block were obtained with a Leica RM 2255 microtome specificity and sensibility having an AUC value close (Finesse ME+ A77500016 microtome, Thermo Scientific) to 1. Additional proteins were iteratively added and and transferred to glass slides and stained with different the variation of AUC was checked and only those that antibodies as described below. increased the AUC value were kept for further iterations [51]. The pROC package in R was used to draw ROCs, Immunohistochemistry calculate the AUC values and estimate the confidence intervals of the ROC curves. Mann-Whitney test and the Slides were incubated at 55°C overnight, then pROC package mentioned above were used for the IHC treated in xylene to de-paraffinize them, and subsequently data analysis. re-hydrated through graded alcohol rinses. TGM4, ADSV, GLPK5, CD63 and SPHM immunohistochemical (IHC) staining were performed. Heat-induced antigen retrieval CONFLICTS OF INTEREST STATEMENT was performed by immersing the slides in 10 mM citrate buffer (C8H8O7.H2Odd 0.01M + C6H5O7.H2Odd; pH 6.0). Michiel Pegtel has an ownership interest in Nonspecific staining was avoided using 10 % normal- exBiome, the other authors declare that they have no matched serum and 0.3 % H2O2 in PBS buffer for 30 conflict of interest. min. Antigen was detected by 1 h incubation at room temperature or overnight at 4°C with the relevant primary GRANT SUPPORT antibody (ADSV 1:1000, TGM4 1:1000, GLPK5 1:50, SPHM 1:600, Sigma-Aldrich, MO, USA; CD63 1:500, The study was supported by grants PI11/02486 Merck Millipore, Germany), followed by an appropriate and CD12/00475 of Instituto de Salud Carlos III secondary antibody conjugated to a peroxidase complex (ISCIII), AECC-JB-2013-02 of Asociación Española (Envision+ poly-HRP system; DAKO Cytomation, Contra el Cáncer, 7th Framework Programe, IRSES Glostrup, Denmark). Color development was done using (PROTBIOFLUID –269285) – Belgium, pre-doctoral DAB+ Chromogen (DAKO Cytomation) followed by fellowship of Vall d’Hebron Research Institute (VHIR) and counterstaining with hematoxylin. Two-experienced by Instituto de Salud Carlos III, grant RD12/0036/0035 pathologist evaluated the staining of the TMA and cofinanced by the European Regional Development calculated a histoscore based on the percentage of stained Fund(ERDF). The CRG/UPF Proteomics Unit is part of the cells and the intensity of the staining (ranging from 1 “Plataforma de Recursos Biomoleculares y Bioinformáticos -weakly positive- to 3 -strongly positive). Histoscores (ProteoRed)” supported by grant PT13/0001 of ISCIII were calculated from the sum of (1x % cells staining and Spanish Ministry of Economy and Competitiveness. weakly positive) + (2x % cells staining moderately We acknowledge support of the Spanish Ministry of www.impactjournals.com/oncotarget 4973 Oncotarget Economy and Competitiveness, “Centro de Excelencia 12. Moon P-G, You S, Lee J-E, Hwang D, Baek M-C. Urinary Severo Ochoa 2013-2017”, SEV-2012-0208, and from exosomes and proteomics. Mass Spectrom Rev. 2011; 30: “Secretaria d’Universitats i Recerca del Departament 1185–202. doi: 10.1002/mas.20319. d’Economia i Coneixement de la Generalitat de Catalunya” 13. Zhou H, Yuen PST, Pisitkun T, Gonzales PA, Yasuda H, (2014SGR678). Dear JW, Gross P, Knepper MA, Star RA. Collection, storage, preservation, and normalization of human urinary REFERENCES exosomes for biomarker discovery. Kidney Int. 2006; 69: 1471–6. doi: 10.1038/sj.ki.5000273. 1. Siegel RL, Miller KD, Jemal A. Cancer statistics, 14. Keller S, Sanderson MP, Stoeck A, Altevogt P. Exosomes: 2016. CA Cancer J Clin. 2016; 66: 7–30. doi: 10.3322/ from biogenesis and secretion to biological function. caac.21332. Immunol Lett. 2006; 107: 102–8. doi: 10.1016/j. 2. Loeb S. Guideline of guidelines: prostate cancer screening. imlet.2006.09.005. BJU Int. 2014; 114: 323–5. doi: 10.1111/bju.12854. 15. Wachalska M, Koppers-Lalic D, van Eijndhoven M, Pegtel 3. Mlcochova H, Hezova R, Stanik M, Slaby O. Urine M, Geldof AA, Lipinska AD, van Moorselaar RJ, Bijnsdorp microRNAs as potential noninvasive biomarkers in urologic IV. Protein Complexes in Urine Interfere with Extracellular cancers. Urol Oncol. 2014; 32: 41.e1–9. doi: 10.1016/j. Vesicle Biomarker Studies. J Circ Biomark. 2016; : 1. doi: urolonc.2013.04.011. 10.5772/62579. 4. Rigau M, Olivan M, Garcia M, Sequeiros T, Montes M, 16. Royo F, Zuñiga-Garcia P, Sanchez-Mosquera P, Egia A, Colás E, Llauradó M, Planas J, Torres I de, Morote J, Perez A, Loizaga A, Arceo R, Lacasa I, Rabade A, Arrieta Cooper C, Reventós J, Clark J, et al. The present and future E, Bilbao R, Unda M, Carracedo A, et al. Different EV of prostate cancer urine biomarkers. Int J Mol Sci. 2013; 14: enrichment methods suitable for clinical settings yield 12620–49. doi: 10.3390/ijms140612620. different subpopulations of urinary extracellular vesicles 5. Royo F, Zuñiga-Garcia P, Torrano V, Loizaga A, Sanchez- from human samples. J Extracell Vesicles. 2016; 5: 29497. Mosquera P, Ugalde-Olano A, González E, Cortazar 17. Torrano V, Royo F, Peinado H, Loizaga-Iriarte A, Unda AR, Palomo L, Fernández-Ruiz S, Lacasa-Viscasillas I, M, Falcón-Perez JM, Carracedo A. Vesicle-MaNiA: Berdasco M, Sutherland JD, et al. Transcriptomic profiling extracellular vesicles in liquid biopsy and cancer. Curr of urine extracellular vesicles reveals alterations of CDH3 Opin Pharmacol. 2016; 29: 47–53. doi: 10.1016/j. in prostate cancer. Oncotarget. 2016; 7: 6835–46. doi: coph.2016.06.003. 10.18632/oncotarget.6899. 18. Mitchell PJ, Welton J, Staffurth J, Court J, Mason MD, 6. Sequeiros T, Bastarós JM, Sánchez M, Rigau M, Montes Tabi Z, Clayton A. Can urinary exosomes act as treatment M, Placer J, Planas J, de Torres I, Reventós J, Pegtel DM, response markers in prostate cancer? J Transl Med. 2009; 7: Doll A, Morote J, Olivan M. Urinary biomarkers for the 4. doi: 10.1186/1479-5876-7-4. detection of prostate cancer in patients with high-grade 19. Principe S, Jones EE, Kim Y, Sinha A, Nyalwidhe prostatic intraepithelial neoplasia. The Prostate. 2015; 75: JO, Brooks J, Semmes OJ, Troyer DA, Lance RS, 1102–13. doi: 10.1002/pros.22995. Kislinger T, Drake RR. In-depth proteomic analyses of 7. Wood SL, Knowles MA, Thompson D, Selby PJ, Banks exosomes isolated from expressed prostatic secretions RE. Proteomic studies of urinary biomarkers for prostate, in urine. Proteomics. 2013; 13: 1667–71. doi: 10.1002/ bladder and kidney cancers. Nat Rev Urol. 2013; 10: 206– pmic.201200561. 18. doi: 10.1038/nrurol.2013.24. 20. Nyalwidhe JO, Betesh LR, Powers TW, Jones EE, White 8. Yáñez-Mó M, Siljander PR-M, Andreu Z, Zavec AB, Borràs KY, Burch TC, Brooks J, Watson MT, Lance RS, Troyer FE, Buzas EI, Buzas K, Casal E, Cappello F, Carvalho DA, Semmes OJ, Mehta A, Drake RR. Increased bisecting J, Colás E, Cordeiro-da Silva A, Fais S, et al. Biological N-acetylglucosamine and decreased branched chain glycans properties of extracellular vesicles and their physiological of N-linked glycoproteins in expressed prostatic secretions functions. J Extracell Vesicles. 2015; 4: 27066. associated with prostate cancer progression. Proteomics 9. Giusti I, Dolo V. Extracellular vesicles in prostate cancer: Clin Appl. 2013; 7: 677–89. doi: 10.1002/prca.201200134. new future clinical strategies? BioMed Res Int. 2014; 2014: 21. Junker K, Heinzelmann J, Beckham C, Ochiya T, 561571. doi: 10.1155/2014/561571. Jenster G. Extracellular Vesicles and Their Role in 10. Hessels D, Schalken JA. Urinary biomarkers for prostate Urologic Malignancies. Eur Urol. 2016; . doi: 10.1016/j. cancer: a review. Asian J Androl. 2013; 15: 333–9. doi: eururo.2016.02.046. 10.1038/aja.2013.6. 22. Nawaz M, Camussi G, Valadi H, Nazarenko I, Ekström K, 11. Øverbye A, Skotland T, Koehler CJ, Thiede B, Seierstad T, Wang X, Principe S, Shah N, Ashraf NM, Fatima F, Neder Berge V, Sandvig K, Llorente A. Identification of prostate L, Kislinger T. The emerging role of extracellular vesicles cancer biomarkers in urinary exosomes. Oncotarget. 2015; as biomarkers for urogenital cancers. Nat Rev Urol. 2014; 6: 30357–76. doi: 10.18632/oncotarget.4851. 11: 688–701. doi: 10.1038/nrurol.2014.301. www.impactjournals.com/oncotarget 4974 Oncotarget 23. Duijvesz D, Burnum-Johnson KE, Gritsenko MA, vesicles. Oncotarget. 2016; 7: 22566–78. doi: 10.18632/ Hoogland AM, Vredenbregt-van den Berg MS, Willemsen oncotarget.8124. R, Luider T, Paša-Tolić L, Jenster G. Proteomic profiling of 34. Hoshino A, Costa-Silva B, Shen T-L, Rodrigues G, exosomes leads to the identification of novel biomarkers for Hashimoto A, Tesic Mark M, Molina H, Kohsaka S, Di prostate cancer. PloS One. 2013; 8: e82589. doi: 10.1371/ Giannatale A, Ceder S, Singh S, Williams C, Soplop N, journal.pone.0082589. et al. Tumour exosome integrins determine organotropic 24. Raposo G, Stoorvogel W. Extracellular vesicles: exosomes, metastasis. Nature. 2015; 527: 329–35. doi: 10.1038/ microvesicles, and friends. J Cell Biol. 2013; 200: 373–83. nature15756. doi: 10.1083/jcb.201211138. 35. Dumitrescu Pene T, Rosé SD, Lejen T, Marcu MG, 25. Jakobsen KR, Paulsen BS, Bæk R, Varming K, Sorensen Trifaró J-M. Expression of various scinderin domains BS, JØrgensen MM. Exosomal proteins as potential in chromaffin cells indicates that this protein acts as a diagnostic markers in advanced non-small cell lung molecular switch in the control of actin filament dynamics carcinoma. J Extracell Vesicles. 2015; 4: 26659. and exocytosis. J Neurochem. 2005; 92: 780–9. doi: 26. Etzioni R, Kooperberg C, Pepe M, Smith R, Gann PH. 10.1111/j.1471-4159.2004.02907.x. Combining biomarkers to detect disease with application 36. Trifaró JM. Scinderin and cortical F-actin are components to prostate cancer. Biostat Oxf Engl. 2003; 4: 523–38. doi: of the secretory machinery. Can J Physiol Pharmacol. 1999; 10.1093/biostatistics/4.4.523. 77: 660–71. 27. Royo F, Zuñiga-Garcia P, Torrano V, Loizaga A, Sanchez- 37. Wang D, Sun S-Q, Yu Y-H, Wu W-Z, Yang S-L, Tan J-M. Mosquera P, Ugalde-Olano A, González E, Cortazar Suppression of SCIN inhibits human prostate cancer cell AR, Palomo L, Fernández-Ruiz S, Lacasa-Viscasillas I, proliferation and induces G0/G1 phase arrest. Int J Oncol. Berdasco M, Sutherland JD, et al. Transcriptomic profiling 2014; 44: 161–6. doi: 10.3892/ijo.2013.2170. of urine extracellular vesicles reveals alterations of CDH3 38. Dubbink HJ, de Waal L, van Haperen R, Verkaik NS, in prostate cancer. Oncotarget. 2016; 7:6835–46. doi: Trapman J, Romijn JC. The human prostate-specific 10.18632/oncotarget.6899. transglutaminase gene (TGM4): genomic organization, 28. Hessvik NP, Sandvig K, Llorente A. Exosomal miRNAs as tissue-specific expression, and promoter characterization. Biomarkers for Prostate Cancer. Front Genet. 2013; 4: 36. Genomics. 1998; 51: 434–44. doi: 10.1006/geno.1998.5393. doi: 10.3389/fgene.2013.00036. 39. Shaikhibrahim Z, Lindstrot A, Buettner R, Wernert N. 29. Wang R, Chinnaiyan AM, Dunn RL, Wojno KJ, Wei JT. Analysis of laser-microdissected prostate cancer tissues Rational approach to implementation of prostate cancer reveals potential tumor markers. Int J Mol Med. 2011; 28: antigen 3 into clinical care. Cancer. 2009; 115: 3879–86. 605–11. doi: 10.3892/ijmm.2011.746. doi: 10.1002/cncr.24447. 40. Kim Y, Ignatchenko V, Yao CQ, Kalatskaya I, Nyalwidhe 30. Lazzeri M, Haese A, Abrate A, de la Taille A, Redorta JP, JO, Lance RS, Gramolini AO, Troyer DA, Stein LD, McNicholas T, Lughezzani G, Lista G, Larcher A, Bini V, Boutros PC, Medin JA, Semmes OJ, Drake RR, et al. Cestari A, Buffi N, Graefen M, et al. Clinical performance Identification of differentially expressed proteins in of serum prostate-specific antigen isoform [-2]proPSA direct expressed prostatic secretions of men with organ- (p2PSA) and its derivatives, %p2PSA and the prostate confined versus extracapsular prostate cancer. Mol Cell health index (PHI), in men with a family history of prostate Proteomics MCP. 2012; 11: 1870–84. doi: 10.1074/mcp. cancer: results from a multicentre European study, the M112.017889. PROMEtheuS project. BJU Int. 2013; 112: 313–21. doi: 41. Darson MF, Pacelli A, Roche P, Rittenhouse HG, Wolfert 10.1111/bju.12217. RL, Young CY, Klee GG, Tindall DJ, Bostwick DG. 31. Megger DA, Bracht T, Meyer HE, Sitek B. Label- Human glandular kallikrein 2 (hK2) expression in prostatic free quantification in clinical proteomics. Biochim intraepithelial neoplasia and adenocarcinoma: a novel Biophys Acta. 2013; 1834: 1581–90. doi: 10.1016/j. prostate cancer marker. Urology. 1997; 49: 857–62. bbapap.2013.04.001. 42. Sequeiros T, García M, Montes M, Oliván M, Rigau 32. March-Villalba JA, Martínez-Jabaloyas JM, Herrero MJ, M, Colás E, de Torres I, Morote J, Reventós J, Doll A. Santamaria J, Aliño SF, Dasí F. Cell-free circulating plasma Molecular markers for prostate cancer in formalin-fixed hTERT mRNA is a useful marker for prostate cancer paraffin-embedded tissues. BioMed Res Int. 2013; 2013: diagnosis and is associated with poor prognosis tumor 283635. doi: 10.1155/2013/283635. characteristics. PloS One. 2012; 7: e43470. doi: 10.1371/ 43. Stenman UH, Leinonen J, Zhang WM, Finne P. Prostate- journal.pone.0043470. specific antigen. Semin Cancer Biol. 1999; 9: 83–93. doi: 33. Koppers-Lalic D, Hackenberg M, de Menezes R, 10.1006/scbi.1998.0086. Misovic B, Wachalska M, Geldof A, Zini N, de Reijke 44. Liu AY, Roudier MP, True LD. Heterogeneity in primary T, Wurdinger T, Vis A, van Moorselaar J, Pegtel M, and metastatic prostate cancer as defined by cell surface Bijnsdorp I. Non-invasive prostate cancer detection by CD profile. Am J Pathol. 2004; 165: 1543–56. doi: 10.1016/ measuring miRNA variants (isomiRs) in urine extracellular S0002-9440(10)63412-8. www.impactjournals.com/oncotarget 4975 Oncotarget 45. Davalieva K, Polenakovic M. Proteomics in diagnosis 49. Vizcaíno JA, Deutsch EW, Wang R, Csordas A, Reisinger of prostate cancer. Pril Makedon Akad Na Nauk Umet F, Ríos D, Dianes JA, Sun Z, Farrah T, Bandeira N, Binz Oddelenie Za Biološki Med Nauki Contrib Maced Acad Sci P-A, Xenarios I, Eisenacher M, et al. ProteomeXchange Arts Sect Biol Med Sci. 2015; 36: 5–36. provides globally coordinated proteomics data submission 46. Manza LL, Stamer SL, Ham A-JL, Codreanu SG, Liebler and dissemination. Nat Biotechnol. 2014; 32: 223–6. doi: DC. Sample preparation and digestion for proteomic 10.1038/nbt.2839. analyses using spin filters. Proteomics. 2005; 5: 1742–5. 50. Farrah T, Deutsch EW, Kreisberg R, Sun Z, Campbell DS, doi: 10.1002/pmic.200401063. Mendoza L, Kusebauch U, Brusniak M-Y, Hüttenhain R, 47. MacLean B, Tomazela DM, Shulman N, Chambers M, Finney Schiess R, Selevsek N, Aebersold R, Moritz RL. PASSEL: GL, Frewen B, Kern R, Tabb DL, Liebler DC, MacCoss MJ. the PeptideAtlas SRMexperiment library. Proteomics. 2012; Skyline: an open source document editor for creating and 12: 1170–5. doi: 10.1002/pmic.201100515. analyzing targeted proteomics experiments. Bioinforma Oxf 51. Borràs E, Cantó E, Choi M, Maria Villar L,álvarez-Cermeño Engl. 2010; 26: 966–8. doi: 10.1093/bioinformatics/btq054. JC, Chiva C, Montalban X, Vitek O, Comabella M, Sabidó 48. Choi M, Chang C-Y, Clough T, Broudy D, Killeen T, E. Protein-Based Classifier to Predict Conversion from MacLean B, Vitek O. MSstats: an R package for statistical Clinically Isolated Syndrome to Multiple Sclerosis. Mol analysis of quantitative mass spectrometry-based proteomic Cell Proteomics MCP. 2016; 15: 318–28. doi: 10.1074/mcp. experiments. Bioinforma Oxf Engl. 2014; 30: 2524–6. doi: M115.053256. 10.1093/bioinformatics/btu305. www.impactjournals.com/oncotarget 4976 Oncotarget