Peroxiredoxins 3 and 4 Are Overexpressed in Prostate Cancer Tissue

Mar 19, 2012 - and Affect the Proliferation of Prostate Cancer Cells in Vitro ... Department of Oncology, Haematology and Bone marrow transplantation ...
0 downloads 0 Views 5MB Size
Download PDF
Article pubs.acs.org/jpr

Peroxiredoxins 3 and 4 Are Overexpressed in Prostate Cancer Tissue and Affect the Proliferation of Prostate Cancer Cells in Vitro Ramesh Ummanni,†,‡,§ Frederico Barreto,†,‡ Simone Venz,∥,⊥ Christian Scharf,# Christine Barett,† Heiko A Mannsperger,¶ Jan Christoph Brase,¶ Ruprecht Kuner,¶ Thorsten Schlomm,□ Guido Sauter,● Holger Sültmann,¶ Ulrike Korf,¶ Carsten Bokemeyer,† Reinhard Walther,∥ Tim H. Brümmendorf,†,△ and Stefan Balabanov*,† †

Department of Oncology, Haematology and Bone marrow transplantation with section Pneumology, Hubertus Wald-Tumour Zentrum (UCCH), University Hospital Eppendorf (UKE), Hamburg, Germany § Center for Chemical Biology, Indian Institute of Chemical Technology (IICT), Hyderabad, India. ∥ Department of Medical Biochemistry and Molecular Biology, University of Greifswald, Greifswald, Germany ⊥ Interfacultary Institute of Genetics and Functional Genomics, University of Greifswald, Greifswald, Germany # Department of Otorhinolaryngology, Head and Neck Surgery, University of Greifswald, Greifswald, Germany ¶ Cancer Genome Research, Deutsches Krebsforschungszentrum (DKFZ), Heidelberg, Germany □ Martini-Clinic, Prostate Cancer Center, University Hospital Eppendorf (UKE), Hamburg, Germany ● Department of Pathology, University Hospital Eppendorf (UKE), Hamburg, Germany △ Klinik für Onkologie, Hämatologie und Stammzelltransplantation, RWTH Aachen University, Aachen, Germany S Supporting Information *

ABSTRACT: The present study aimed to investigate the proteome profiling of surgically treated prostate cancers. Hereto, 2D-DIGE and mass spectrometry were performed for protein identification, and data validation for peroxiredoxin 3 and 4 (PRDX3 and PRDX4) was accomplished by reverse phase protein arrays (RPPA). The Formal Concept Analysis (FCA) method was applied to assess whether the TMPRSS2-ERG gene fusion could influence the degree of overexpression of PRDX3 and PRDX4 in prostate cancer. Lastly, we performed an in vitro functional characterization of both PRDX3 and PRDX4 using the classical human prostate cancer cell lines DU145 and LNCaP. Reverse phase protein arrays verified that the overexpression of both PRDX3 and PRDX4 in tumor samples is negatively correlated with the presence of the TMPRSS2-ERG gene fusion. Functional characterization of PRDX3 and PRDX4 activity in PCa cell lines suggests a role of these members of the peroxiredoxin family in the pathophysiology of this tumor entity. KEYWORDS: prostate cancer, proteomics, peroxiredoxin, cell proliferation, TMPRSS2-ERG

■

INTRODUCTION Prostate cancer (PCa) is the most common type of cancer affecting men and among the leading causes of cancer death after lung cancer in the western world.1 In spite of this, the pathophysiology of prostate cancer development still remains poorly understood.2 Several valuable methods can be currently accomplished to diagnose this life-threatening disease, including the digital rectal examination (DRE), as well as the prostatespecific antigen (PSA) test and biopsy of the prostate. Due to the introduction of PSA screening assays and subsequent biopsy of prostate into the routine clinical practice of PCa diagnosis, early detection rates and thus incidence of PCa have increased dramatically in the last years. Furthermore, this screening strategy resulted in a decreased PCa mortality.3 However, whether the screening by the determination of serum PSA levels leads to a reduction of overall survival of PCa patients is still an open debate4−7 because PSA is copiously produced by prostatic epithelium, periurethral glands, endome© 2012 American Chemical Society

trium, breast tissue, adrenal and renal tumors and is neither tissue nor gender specific. Moreover, PSA can be secreted from benign as well as malignant cells of the prostate. Therefore, serum PSA correlates with benign prostate hyperplasia and cancer. Consequently, PSA test with limited specificity and sensitivity has limitations for its clinical application in screening and diagnosis of PCa, which demands better biomarkers. Novel biomarkers with a higher specificity and sensitivity would potentially improve the clinical management and the outcome of patients with PCa. In this context, recent application of novel high throughput technologies to study PCa have resulted in the identification of PCa associated genome, transcriptome and proteome alterations.8 These findings could potentially lead to a deeper understanding of the molecular biology of PCa initiation and progression.4,9,10 These studies have revealed that Received: November 28, 2011 Published: March 19, 2012 2452

dx.doi.org/10.1021/pr201172n | J. Proteome Res. 2012, 11, 2452−2466

Journal of Proteome Research

Article

cancerous and normal areas as described before.21 The punches were immersed in RNAlater (Qiagen, Germany) for 24 h at room temperature and subsequently stored at −80 °C. To confirm the presence of tumor, all punches were sectioned, and tumor cell content was determined in every 10th section. Only sections containing at least 70% tumor cells were included in the study. Normal prostate tissue samples were obtained from 53 patients who underwent radical prostatectomy for prostate cancer. Only sections containing exclusively normal tissue material with epithelial cell content between 20 and 40% were included in the study. Clinical data obtained for each sample included: age of the patient, PSA values, Gleason score, TNM classifications, TMPRESS-ERG-fusion status (not for all patients). Ten tissue sections (4 μm) were taken from each tissue block for protein isolation. TMPRSS2-ERG fusion events were determined using RT-PCR. cRNA from the Affymetrix Whole Transcript Sense Target Labeling Assay was reversely transcribed and 10 ng of cDNA were used for RT-PCR based validation. Amplification as well as nested PCR was performed using primers described by Jhavar et al.22

many PCa are associated with the chromosomal translocation involving members of the ETS (E-twenty six) family of transcription factors.11,12 Among these molecular events, the gene fusion of TMPRSS2 (Transmembrane protease, serine 2) and ERG (ETS related gene) represents the most frequent recurrent translocation.9,13 Comparative proteome analyses of clinical patient materials have been demonstrated to be informative for the identification of specific molecular biomarkers for diagnosis and therapeutics of cancer.14 To date, several groups have already reported differential proteomic data on clinical prostate cancer specimens.14−19 Recently, our group described the comparative proteome set of pathologically well-characterized cancerous prostate samples and corresponding histologically normal tissue from patients diagnosed for PCa with different degree of histological grading and a wide range of abnormal PSA values.20 Through systematic analysis of proteomic data in this previous study, we could identify candidate proteins, potentially involved in initiation and/or progression of PCa. The present study aimed to investigate the proteome profiling of cryosections of punch biopsies from radical prostatectomies with adverse histological tumor features (pT3 and/or high Gleason grading). Hereto, 2-dimensional differential in-gel electrophoresis system (2D-DIGE) combining 2Delectrophoresis and MALDI-TOF-MS/MS was performed for protein identification. Out of the list of 82 differentially expressed proteins, two peroxidoredoxins (PRDX3 and PRDX4) were validated by reverse phase protein arrays (RPPA) from the same samples used for proteomic analysis, along with further samples processed in an analogous way. To correlate the protein expression with known molecular markers, we have assessed whether the TMPRSS2-ERG gene fusion could exert any influence on the altered protein expression in prostate cancer. Furthermore, to understand the role of differentially expressed proteins that are significantly correlated with TMPRSS2-ERG gene fusion, we performed an in vitro functional characterization of two selected candidate proteins, namely PRDX3 and PRDX4, on prostate cancer progression. From the current study, we were able to report the feasibility of carrying out a proteomic analysis and its subsequent validation on very small tissue amounts, as well as to enumerate a series of proteins associated with intermediate- and high-grade PCa. Validation of 2D-DIGE proteomic data confirmed that overexpression of both PRDX3 and PRDX4 in tumor samples negatively correlated with presence of TMPRSS2-ERG gene fusion. Further functional characterization of PRDX3 and PRDX4 by modulating their expression in PCa cell lines revealed that they play a role in cellular proliferation.

■

Protein Isolation

Briefly, tissue sections were transferred into T-PER lysis buffer containing one complete mini protease inhibitor, PhosStop (Roche) per 10 mL, 2 mM Staurosporine (Roche) and 1 mM EDTA (Applied Biosystems/Ambion) and homogenized using Tissulyser (Qiagen) for 4 min at 30000 rpm. Homogenates were then frozen at −80 °C for 2 h and thawed with constant shaking in a thermomixer (300 rpm) for 15 min at 4 °C. To collect a clear supernatant, samples were centrifuged at 13000 rpm for 12 min at 4 °C and the supernatant was transferred into a Qiashredder tube (Qiagen) and further centrifuged at 13000 rpm, 4 °C for 5 min. The protein concentration was estimated by standard BCA Protein Assay (Pierce) and lysates were stored at −80 °C until further use. 2D-DIGE, Image Analysis and Mass Spectrometry

2D-DIGE was performed as described previously.20 Briefly, fluorescent labeled samples were diluted with 2× rehydration buffer (8 M urea, 4% CHAPS, 13 mM DTT supplemented with 1% (v/v) IPG buffer pH 4−7 and traces of bromophenol blue) to a final volume of 450 μL (IPG strips 24 cm, pH 4−7, GE Healthcare) and passively rehydrated overnight at 20 °C in IPGPhor cassettes. For preparative gels, 750 μg of unlabeled protein pooled from equal amounts of samples was used. Isoelectric focusing followed by SDS-PAGE and visualization of separated protein spots in gels were performed as described previously. Preparative gels were stained with Roti-Blue, a colloidal coomassie brilliant blue G250 stain. Briefly, gels were fixed in a solution of 40% methanol and 15% acetic acid for at least 4 h and then immersed in colloidal staining solution overnight. To remove background staining, gels were washed in 20% methanol. For gel image analysis, Delta 2D differential analysis software version 4.0 (Decodon GmbH, Germany) was used in this study. For individual gel analysis, protein spots were detected, quantified and normalized using the internal standard module. According to this, spots respective to Cy3 and Cy5 images were expressed as volume ratios relative to the corresponding spot volume in the Cy2 internal standard image. All normalized data sets obtained from the gels were collected and analyzed as two independent groups, namely “tumor” and “benign”. This approach allowed the matching of multiple gel images from different samples to provide statistical data on average abundance for each protein spot among the DIGE gels

MATERIALS AND METHODS

Preparation of Tissue Sections

Prostate tissue samples were obtained from the University Medical Centre Eppendorf, Hamburg, Germany. Approval for the study was obtained from the local ethics committee and all patients agreed to provide additional tissue sampling for scientific purposes. Tissue samples from 51 prostate cancer and 53 normal prostate tissues were included (Table S1, Supporting Information). None of the patients had been treated with neoadjuvant radio-, cytotoxic- or endocrine therapy. During radical prostatectomy, tissue samples from the peripheral zone of the prostate were taken with a 6 mm punch biopsy instrument immediately after surgical removal of the prostate from 2453

dx.doi.org/10.1021/pr201172n | J. Proteome Res. 2012, 11, 2452−2466

Journal of Proteome Research

Article

included in the analysis. Student’s t-test was performed to assess statistical significance of differentially expressed proteins. On the basis of the average volume ratio of spots, those whose relative expression was changed at least 1.5 fold (increase or decrease) between benign and tumor groups at 95% confidence level (t-test; p < 0.05) were considered to be significant. For subsequent mass spectrometry analysis, coordinates of significantly altered protein spots were transferred to coomassie-stained preparative gels for spot picking. Protein identification was performed as described previously.20 Briefly, gel pieces were washed in 50 mM ammoniumbicarbonate/50% (v/v) methanol and with 75% (v/v) ACN. After drying, trypsin solution containing 20 ng/μL trypsin in 20 mM ammoniumbicarbonate was added and incubated at 37 °C for 120 min. For peptide extraction, gel pieces were covered with 50% (v/v) ACN/0.1% (w/v) TFA and incubated for 30 min at 37 °C. The peptide containing supernatant was transferred into a new micro plate and the extraction was repeated. The supernatants were pooled and dried completely at 40 °C for 220 min. Peptides were dissolved in 0.5% (w/v) TFA/50% (v/v) ACN and spotted on the MALDI-target. Then, matrix solution (50% (v/v) ACN/0.5% (w/v) TFA) saturated with CHCA was added and mixed with the sample solution by aspirating the mixture five times. Prior to the measurement in the MALDITOF instrument, the samples were allowed to dry on the target 10−15 min. For MALDI-TOF measurement a 4800 MALDI TOF/TOF Analyzer was used. The spectra were recorded in reflector mode in a mass range from 800 to 4000 Da with an internal one-point-calibration on the autolytic fragment of trypsin (monoisotopic (M + H)+ m/z at 2211.104, signal/noise ≥10). Additionally MALDI-TOF-MS/MS analysis was performed for the 5 strongest peaks of the TOF-spectrum after subtraction of peaks corresponding to background or trypsin fragments. The internal calibration was automatically performed as one-point-calibration if the monoisotopic arginine (M + H)+ m/z at 175.119 or lysine (M + H)+ m/z at 147.107 reached a signal-to-noise ratio (S/N) of at least 5. After calibration a combined database search of MS and MS/MS measurements was performed using the GPS Explorer software (Ver. 3.6, Applied Biosystems, Foster City, CA) with the following settings: (i) MS peak filtering: mass range from 800 to 4000 Da; minimum S/N filter of 10; peak density of 50 peaks per range of 200 Da and maximal 200 peaks per protein spot; mass exclusion list contained background peaks and trypsin fragments with an exclusion tolerance of 50 ppm (ii) MS/MS peak filtering: mass range from 60 Da to a mass that was 20 Da lower than the precursor mass; peak density of 50 peaks per 200 Da and maximal 65 peaks per MS/MS; minimum S/N filter of 10 (iii) database search: precursor tolerance 50 ppm and MS/MS fragment tolerance 0.45 Da. Peak lists were compared with the Swiss-Prot database v56.1 human taxonomy. Peptide mixtures that yielded at least a mowse score of 56 for database results were regarded as positive identifications.

constructed, showing the most relevant direct and indirect connections of proteins found to be deregulated between tumor and benign tissue, as well as additional proteins predicted to interact. Gene ontology classification with respect to biological processes and molecular functions were established with David GO annotation tool.23,24 Reverse Phase Protein Arrays

The method used in this study for validation of 2D-DIGE data has been described in detail previously.25 Total cell lysates were prepared as described above and 4 replicate spots were printed per sample. To probe lysate arrays with specific antibodies for PRDX3 and PRDX4, slides were rehydrated in PBS-Tween (0.05%) and blocked in Odyssey Blocking buffer diluted 1:1 with PBS-Tween (0.05%) containing NaF and NaVO3 for 1 h at RT. The slides were then incubated with prevalidated antibodies at 1:300 dilutions in blocking buffer overnight at 4 °C. Primary antibody detection was carried out with nearinfrared (NIR)-dye-labeled goat antirabbit or antimouse secondary antibodies and visualized using Odyssey scanner (LI-COR, Lincoln, NE). Images were analyzed and quantified using GenePix Pro 5.0 software (Axon Instruments, Inc., Foster City, CA). Control slides were detected with FAST Green FCF to determine the total protein content per spot.25 All quantitative data obtained from GenePix Pro were finally exported for further analysis. Data Analysis of Reverse Phase Protein Arrays

Data analysis was performed using the software tool RPPanalyzer implemented in the statistical environment R.26 For each spot, mean value of signal intensities was generated through the GenePixPro image analysis software (Axon instruments) with subtraction of the local background intensities. FAST Green FCF slide signals were used to calculate a spot specific normalization factor.25 Replicate spots were aggregated using the median value. Arrays with signal intensities not significantly higher than the blank level, as well as arrays without linear correlation between signal intensities and protein concentration, were excluded from further analysis. Overall significance was calculated using the Wilcoxon test and defined as p < 0.05. Bioinformatics Analysis of the Proteomic Data

Protein expression data from 2D-DIGE analysis were analyzed by unsupervised hierarchical clustering with using Euclidian distances and complete linkage to find unique protein clusters among the set of all significant proteins which could classify all samples as tumor and benign with high certainty. The unsupervised clustering was performed using Euclidean distance measure and the average agglomeration method of the log-transformed values of all significant differentially expressed proteins across samples included in the analysis set. Formal concept analysis (FCA) has been applied to RPPA data to integrate protein expression data of PRDX3 and PRDX4 and clinical information of cancer patients on a visual level.27 It is a method of discrete mathematics suitable to determine conceptual relation of objects (e.g., clinical samples) according to their attributes (e.g., expression data, clinical data) by drawing them into a clearly arranged lattice. The graphically represented structures can visualize conceptual hierarchies and allows therefore finding out dependencies within different sample attributes. FCA is increasingly applied in conceptual clustering and analysis of gene expression data to identify gene combinations as biomarkers and to understand co regulation of

Protein Network Analysis and GO Classification of Differentially Expressed Proteins

A network analysis was performed using the Ingenuity Pathways Analysis (IPA) (Ingenuity Systems, www.ingenuity. com) algorithm as described.20 The list of differentially expressed proteins was entered into the IPA software to explore relevant biological networks and to assess interactions with other proteins. A global protein interaction network was 2454

dx.doi.org/10.1021/pr201172n | J. Proteome Res. 2012, 11, 2452−2466

Journal of Proteome Research

Article

genes.28,29 The FCA method was applied here to analyze if any relation between the degree of overexpression of PRDX3 and 4 and the TMPRSS2-ERG gene fusion (fusion gene status was determined in 35 tumors among all tumor patients included in the current study). By considering mean centered values, the expression profiling samples were classified into high or low PRDX3 or 4 expressing groups. From RPPA protein expression data, an input matrix under the context of either presence or absence of TMPRSS2-ERG has been created as described in data Table S2 (Supporting Information). By importing this matrix into the concept explorer, conceptual lattices have been built as described before to understand the dependency of PRDX3 and 4 protein expression with the TMPRSS2-ERG gene fusion in prostate tissues.28,30

Virus Production and Infection of Target Cells

The virus production and infection was performed as described recently.32 HEK293T or phoenix amphotropic cells at 80% confluence were transfected with either empty or recombinant vectors using the calcium chloride transfection method. For this purpose, 20 μg plasmid DNA were mixed with 125 mM CaCl2 in 1 mL HBS and the mixture was incubated for 10 min at room temperature. The DNA precipitate was added dropwise to the cells in culture medium containing 25 μM chloroquine. After 12 h of transfection, medium was removed and cells were refeed with fresh medium and further incubated for 24 h. Subsequently, virus containing supernatants were isolated and passed through a 0.45 μm cellulose acetate filter directly on the target cells growing at around 50% confluence. The cells were fed with fresh medium to continue another round of virus collection. Two more infection cycles were repeated 12 h later. After three cycles of infection, target cells were grown in normal cultivation medium for 24 h and selected for integration of the target gene with puromycin at a concentration of 2 μg/ mL until all cells in control dishes had died. Colonies with resistance to puromycin were further propagated and verified for overexpression or knockdown of peroxiredoxins by PCR and Western blot.

Cell Culture

Human prostate carcinoma cell lines DU145 and LNCaP were purchased from DSMZ (Braunschweig, Germany) and maintained in RPMI 1640 (Invitrogen) supplemented with 10% fetal bovine serum (FBS). Cells were grown in an incubator at 37 °C with a constant supply of 5% CO2 and split after reaching 85−90% confluence. Testing for mycoplasma contamination was performed regularly by using the MycoAlert Mycoplasma Detection Kit (Cambrex Bio Science Rockland, Inc., Rockland, ME). Packaging cell lines (phoenix amphotropic and HEK293T) were grown in DMEM with 10% FBS.

Cell Proliferation Assay

For measurement of cell proliferation, an initial density of 1.0 × 106 cells per well was plated in 6-well format in complete growth medium and allowed to grow under optimal culture conditions. At indicated time points, cells were harvested by trypsinization and live cells counted by analysis by trypan blue exclusion with a Vi-CELL Cell Viability Analyzer (Beckman Coulter).

PRDX Cloning Strategy

For overexpression of peroxiredoxins, a recombinant vector was generated by cloning the coding region of human PRDX3 cDNA (accession number NM_007452) into the pMSCVpuro vector (Clontech, Palo Alto, CA, USA), as well as the coding region of human PRDX4 cDNA (accession number NM_006406) into the pLeGO-iG2puro vector.31 The cDNA was prepared by reverse transcription of 1 μ of total RNA derived from tissues using oligo dT primer (15mer) and MMLV reverse transcriptase (Promega Corp.). A specific primer pair was designed using Lablife online tools (Addgene, www. lablife.org) and synthesized by MWG Operon (Eurofins MWG Operon, Germany). The coding sequence (CDS) of the respective peroxiredoxin isoforms was amplified from total cDNA by PCR using Phusion DNA Polymerase (Finnzymes Oy, Finland). Primer sequences are provided as Supporting Information (Table S3). After digestion of the PCR product and respective vectors with XhoI/EcoRI or BamHI/NotI enzymes (Fermentas GmbH, Germany), ligation of the PCR product with the linear vector resulted in a recombinant pMSCV-PRDX3 or pLEGO-PRDX4 construct. To knockdown the expression of peroxiredoxins 3 and 4 in prostate carcinoma cell lines, lentiviral expressing small hairpin vectors against PRDX3 or PRDX4 or scrambled shRNA were purchased (Mission shRNA; Sigma-Aldrich, Germany). The target sequence of the two hairpins used was CCGGCCTAAGCCTTGATGACTTTAACTCGAGTTAAAGTCATCAAGGCTTAGGTTTTTG for PRDX3 (NM_006793; Mission shRNA TRCN0000064840) and CCGGCCACACTCTTAGAGGTCTCTTCTCGAGAAGAGACCTCTAAGAGTGTGGTTTTTG for PRDX4 (NM_006406; Mission shRNA TRCN0000064818). Mission shRNA-pLKO.1-puro control vector was used as negative controls.

Quantitative Real Time PCR (qPCR)

Total RNA were extracted using the All Prep DNA/RNA Mini kit (Qiagen) according to the manufacturer’s instructions. Complementary DNA was synthesized by reverse transcription of 1 μg total RNA using Oligo (dT)18 Primer (Fermentas) and Superscript II (Invitrogen, Germany). cDNA was amplified by 25 cycles of PCR using DreamTaq Green (Fermentas). Quanti Tect primers for PRDX3, PRDX4 and GAPDH (housekeeping gene) were purchased directly from Qiagen. qPCR was performed as described previously.20,33 To test statistical significance, data were analyzed by unpaired student t-test performed and p value 56 with p-value