SH3BGRL3 Protein As a Potential Prognostic Biomarker for Urothelial

Published OnlineFirst August 18, 2015; DOI: 10.1158/1078-0432.CCR-14-3308

Biology of Human Tumors Clinical Cancer Research SH3BGRL3 Protein as a Potential Prognostic Biomarker for Urothelial Carcinoma: A Novel Binding Partner of Epidermal Growth Factor Receptor Cheng-Yao Chiang1, Chin-Chen Pan2, Hong-Yi Chang1, Ming-Derg Lai1,3, Tzong-Shin Tzai4, Yuh-Shyan Tsai4, Pin Ling1,5, Hsiao-Sheng Liu1,5, Bi-Fang Lee6, Hong-Ling Cheng4, Chung-Liang Ho1,7, Shu-Hui Chen8, and Nan-Haw Chow1,7

Abstract

Purpose: Mass spectrometry–based biomarker discovery has SH3BGRL3 is expressed in 13.9% (159/1145) of bladder cancer clinical benefit. To identify novel biomarkers for urothelial car- cohort and is positively associated with muscle invasion (P ¼ cinoma, we performed quantitative proteomics on pooled urine 0.0028). SH3BGRL3 expression is associated with increased risk pairs from patients with and without urothelial carcinoma. of progression in patients with nonmuscle-invasive bladder can- Experimental Design: Shot-gun proteomics using liquid chro- cer (P ¼ 0.032). SH3BGRL3 expression is significantly associated matography-tandem mass spectrometry and stable isotope with a high level of epidermal growth factor receptor (EGFR) in dimethyl labeling identified 219 candidate proteins. The poten- bladder cancer (P < 0.0001). SH3BGRL3 promotes the epithelial– tial implication of SH3 domain binding glutamic acid-rich pro- mesenchymal transition, cell migration, and proliferation of tein like 3 (SH3BGRL3) was examined by immunoblotting of the urothelial carcinoma in vitro. SH3BGRL3 interacts with phos- urine (n ¼ 13) and urothelial tumors (n ¼ 32). Additional phor-EGFR at Y1068, Y1086, and Y1173 through Grb2 by its immunohistochemistry was performed on bladder cancer array proline-rich motif, and activates the Akt-associated signaling (n ¼ 1145) and correlated with tumor aggressiveness. Then, pathway. biologic functions and signaling pathways of SH3BGRL3 were Conclusions: Evaluation of SH3BGRL3 expression status or explored using stable cell lines. urine content may identify a subset of patients with bladder cancer Results: The detectable urine SH3BGRL3 in patients with who may require more intensive treatment. SH3BGRL3 deserves urothelial carcinoma was positively associated with higher his- further investigation as a cotargeting candidate for designing tologic grading and muscle invasiveness of urothelial carcinoma. EGFR-based cancer therapies. Clin Cancer Res; 1–11. 2015 AACR.

Introduction and 14,000 cancer deaths estimated in 2015 (1). Urothelial carcinoma of the upper urinary tract (UC-UUT) is rare, repre- Urothelial carcinoma of the bladder is the sixth most com- senting only 5% of all urothelial carcinoma (2). It is well mon human cancer in the United States, with 74,000 new cases known that urothelial carcinoma is fundamentally multifocal in nature: in patients with bladder cancer, 4% will eventually develop UC-UUT, whereas up to 50% of those with UC-UUT 1Institute of Basic Medical Sciences, College of Medicine, National will develop a lesion in the bladder (3). The established 2 Cheng Kung University, Tainan, Taiwan. Department of Pathology, prognostic factors for patients with urothelial carcinoma are Taipei Veterans General Hospital, Taipei, Taiwan. 3Department of Bio- chemistry and Molecular Biology, College of Medicine, National Cheng age, tumor staging, and histologic grade (3, 4). Therefore, it is Kung University, Tainan, Taiwan. 4Department of Urology, National necessary to identify biomarkers that are useful for diagnosis Cheng Kung University Hospital, Tainan, Taiwan. 5Department of and predicting the prognosis when planning treatment for Microbiology and Immunology, College of Medicine, National Cheng Kung University, Tainan, Taiwan. 6Department of Nuclear Medicine, urothelial carcinoma. National Cheng Kung University Hospital,Tainan,Taiwan. 7Department Liquid chromatography-tandem mass spectrometry (LC-MS/ of Pathology, College of Medicine, National Cheng Kung University, MS)–based proteomics has become a popular tool for studying 8 Tainan, Taiwan. Department of Chemistry, College of Sciences, biomarker. Stable-isotope labeling coupled with shot-gun prote- National Cheng Kung University, Tainan, Taiwan. omics can explore the protein proﬁle of a biologic sample in a Note: Supplementary data for this article are available at Clinical Cancer single run of MS and distinguish protein expression levels Research Online (http://clincancerres.aacrjournals.org/). between two biologic states (5). We have developed a stable S.-H. Chen and N.-H. Chow contributed equally to this article. isotope-based dimethyl labeling method that produces different Corresponding Author: Nan-Haw Chow, Department of Pathology, National isotopic pairs ranging from 2 to 8 Da (mass difference) or up to six Cheng Kung University Hospital, 138 Sheng-Li Road, Tainan 70428, Taiwan. multiplex analyses (6). By using stable isotope dimethyl labeling Phone: 886-6-235-3535, ext. 5288; Fax: 886-6-276-6195; E-mail: coupled with two-dimensional liquid chromatography peptide [email protected] separation and MS/MS analysis, we discovered that nucleophos- doi: 10.1158/1078-0432.CCR-14-3308 min is positively involved in the arsenic-related bladder carcino- 2015 American Association for Cancer Research. genesis (7). The urine proﬁling experiment found that CD14 is a

www.aacrjournals.org OF1

Chiang et al.

before the study began. The study protocol was approved by the Translational Relevance Institutional Review Board of National Cheng Kung University Our prior studies indicate that EGFR, whether by itself or Hospital (Nos. HR-97-117 and B-ER-103-214) and Taipei Veter- coexpressed with ErbB2, ErbB3, or RON, is an important ans General Hospital (#98-10-05A; ref. 14). predictor of tumor recurrence and survival of patients with Preoperative urine samples (at least 50 mL) were collected urothelial cancer. The data support the importance of EGFR in from 13 of the patients, and primary tumor samples from 1,177 the tumorigenesis of human bladder. We report that detectable of the patients who had their tumors surgically resected at our urinary SH3BGRL3 is positively associated with histologic hospital or Taipei Veterans General Hospital. Preoperative grading and muscle invasiveness of urothelial carcinoma. urine samples from three patients with histologic diagnosis of Higher SH3BGRL3 expression in urothelial carcinoma is relat- chronic pyelonephritis were collected as a control. All samples ed to increased risk of recurrence and lower survival rate. were immediately stored at 80C until analysis. Clinicopath- SH3BGRL3 promotes the cell growth in vitro and in vivo. ologic information for all the patients was obtained from their Evidence provided further supports activation of Akt-associ- medical records, including demographics, clinical and surgical ated signaling cascade by SH3BGRL3 and its interaction with tumor staging, tumor size, histologic grade, the extent of EGFR by proline-rich residues through Grb2. Hence, we have disease involvement, and the status of pelvic and para-aortic unveiled that SH3BGRL3 is a novel urinary biomarker for lymph nodes. urothelial carcinoma. SH3BGRL3 deserves investigation as a cotargeting candidate in the design of EGFR-based cancer LC/MS-MS analysis and database searching therapy. Protein (100 mg) was extracted from the urine samples of two pairs of patients with urothelial carcinoma and two pairs of patients without urothelial carcinoma, as previously described (6, 7). Next, the protein mixtures were digested by trypsin and labeled using formaldehyde-C12H and formaldehyde-C13D . potential biomarker for benign prostatic hyperplasia (8). The 2 2 All samples were fractionated using liquid chromatography results support the accuracy and clinical implications of this (NanoAcquity UltraPerformance LC; Waters) and analyzed technical platform. using a mass spectrometer (LTQ-Orbitrap XL; Thermo Fisher SH3 domain binding glutamic acid-rich protein like 3 Scientific). The spectra generated in both MS and CID-MS2 (SH3BGRL3), one of candidate proteins in the gene list, is steps from the cleaved digest were searched using Mascot 2.3 located on chromosome 1p34.3–35. It contains 279 nucleo- (Matrix Science) against the SwissProt 20110921 (532,146 tides that are translated to 93 amino acids. The crystal and sequences; 188,719,038 residues) protein databank for Homo NMR structures of SH3BGRL3 have been described in an sapiens using a mass tolerance of 5 ppm for precursor ions, earlier work (9). The amino acid sequence of SH3BGRL3 and 0.8 Da for product ions. The peptide quantification ratio belongs to the thioredoxin superfamily and is highly similar (heavy/ light, urothelial carcinoma/non-urothelial carcinoma) to Escherichia coli GRX1 (10), an enzyme with oxidoreductase was calculated using Mascot Distiller 2.3 (Matrix Science), activity. Although SH3BGRL3 protein is homologous to N- which uses the average peak area of the first three or four terminal region of the SH3BGR protein (10, 11), it lacks the isotopic peaks across the elution profile for quantification. conserved SH3-binding motif, implying that it has a different Manual inspections were done to exclude incorrect calculations function to those of other subfamily members. SH3BGRL3 because of low-confidence proteins (poor spectrum quality; could inhibit TNFa-induced apoptosis and promote cell sur- Supplementary Materials and Methods). vival, although the molecular mechanism underlying these functions remains unclear (12). In a pig–human xenograft model experiment, SH3BGRL3 was upregulated in xeno-con- Cell culture and stable clone selection junctival epithelial cells to protect them from programmed cell Human embryonic kidney HEK293 and T24 grade III urothelial death (13). SH3BGRL3 is also involved in all-trans retinoic carcinoma cell lines were obtained from the ATCC and authen- fi acid–induced cell differentiation (9). ticated by short tandem repeat (STR) PCR pro ling in 2014 and Here, we describe the detection of SH3BGRL3 in the urine of 2015 (Genomic Center, National Cheng Kung University, Tai- patients with urothelial carcinoma and upregulated SH3BGRL3 in wan). The grade II human urothelial carcinoma cell lines primary bladder cancer. Further investigations revealed that TSGH8301 and 5637 were obtained from the Bioresource Col- SH3BGRL3 participates in cancer cell proliferation in vitro and in lection and Research Center at Food Industry Research and vivo, epithelial–mesenchymal transition (EMT), and cell migra- Development Institute, Taiwan, and authenticated in 2014 by tion. Of particular importance, we provide evidence that STR PCR. Both cell lines were maintained in DMEM or RPMI SH3BGRL3 interacts with epidermal growth factor receptor (Invitrogen) supplemented with 10% FBS (Hyclone) and antibi- (EGFR) at Y1068, Y1086, and Y1173 through Grb2 by its pro- otic/antimycotic solution (Caisson Laboratories). Cells were cul- fi line-rich motif, and activates the Akt-associated signaling path- tured at 37 C in a humidi ed atmosphere of 5% CO2. The day way. Our results suggest that SH3BGRL3 is a novel cotargeting before the experiments, cells were seeded in 6- or 10-cm dishes candidate for EGFR-based cancer therapy. and starved for 24 hours. The cells were then treated with EGF (10 ng/mL, Invitrogen) at different time courses depending on the experiment. Materials and Methods The TSGH8301, 5637, and HEK293 cell lines were trans- Patient enrollment and sample and data collection fected with pEGFP-N1, pEGFP-N1-SH3BGRL3, p3xflag-CMV- A total of 1,190 patients were enrolled and all of the partici- 14, and p3xflag-CMV-14-SH3BGRL3 using Lipofectamine pants with radical cystectomy had signed informed consent forms LTX agent (Invitrogen). The T24 cell line was transfected with

OF2 Clin Cancer Res; 2015 Clinical Cancer Research

SH3BGRL3 Is a Novel Biomarker of Urothelial Carcinoma

pcDNA6.2-GW/EmGFP-miR-neg or pcDNA6.2-SH3BGRL3 developed using an enhanced chemiluminescence solution (Invitrogen). The transfectants were selected using G418 (Millipore) for 60 to 90 seconds. [2 mg/mL, Geneticin Selective Antibiotic (G418 Sulfate); Thermo] or blastidicin (1 mg/mL; Thermo) for 6 to 8 weeks MTT assay and verified using Western blotting. The assay is based on the cleavage of the yellow tetrazolium salt MTT to purple formazan crystals by metabolically active cells. All cell lines were seeded in a 96-well plate at concentrations of 103 Total RNA isolation, reverse transcription-polymerase chain cells/well. The cells were grown in a final volume of 100 mL culture reaction (RT-PCR), and plasmid construction medium per well. The MTT labeling reagent (50 mL, final con- Cells were lysed with Trizole reagent (TRI) and chloroform centration 0.5 mg/mL) was added to each well and these were then wasappliedtoseparatetheRNAat the aqueous phase. The RNA incubated for 4 hours, after which 50 mL of dimethylsulfoxide was then precipitated by isopropanol and centrifuged at 14,000 (DMSO) was added to each well. The absorbance of the formazan rpm for 15 minutes. Finally, the RNA was dissolved with DEPC- product was measured at 570 nm. MQ H2O. Total RNA was annealed by Oligo (dT), then the following Immunohistochemistry, scoring, immunofluorescent staining, components were added in this order: M-MLV RT reaction buffer, and confocal microscopy dNTP, M-MLV reverse transcriptase, and nuclease-free water, to a Immunohistochemical (IHC) staining, immunofluorescent final volume of 20 mL. The test tubes were mixed gently and staining, and confocal microscopy were done as previously incubated at 42 C for 1 hour and at 95 C for 5 minutes. For PCR, described (15). Antibody for SH3BGRL3 (HPA030848; Sigma- the templates were mixed with 10 reaction buffer, dNTPs, Aldrich) at 1:500 dilution or EGFR (sc-31157; Santa Cruz Bio- oligonucleotide primers, Taq polymerase, and sterile MQ water. technology) at 1:200 dilution was incubated overnight at 4C. The reaction program depended on the T value of each pair of m Because the intensity of immune-staining did not vary obviously, primers. the IHC was evaluated according to the percentage of tumor cells The plasmid construction was executed in three steps: First, stained (by N-H Chow) and blinded to clinical information as the A fragment was amplified by a specificprimercontaining described previously with slight modification (4). Tissue sections the BamHI and BglII restriction sites. Second, PCR products and showing immune-staining in less than 5% of tumor cells or lack of p3xflag-CMV-14 were cut by BamHI and BglII restriction any immune-reactivity were classified as negative expression. enzymes for 18 hours. Third, the two DNA fragments were Those with a staining reaction between 5% and 25% of tumor ligated by DNA ligase at 4 C for 18 hours after clean-up, and cells were defined as low level of expression, between 25% and then expressed by competent cells (Supplementary Materials 50% were intermediate level of expression, and those with and Methods). immunostaining in greater than 50% were defined as high level of protein expression. Protein extraction, Western blotting, and Cellsgrownonglasscoverslipswerefixed with methanol/ co-immunoprecipitation (co-IP) acetone (1:1) at 4C for 10 minutes, and permeabilized with Total protein was precipitated from the patients' urine using 0.5% Triton X-100 for 10 minutes at room temperature. To 20% trichloroacetic acid (Sigma-Aldrich) for 24 hours at 4 C. reduce the non-specific background, the cells were blocked with The protein pellet was centrifuged for 30 minutes at 14,000 5% bovine serum albumin (BSA) and washed with phosphate- rpm at 4 C, then redissolved in a solution using RIPA lysis buffered saline with Tween (PBST). Cells on coverslips were buffer containing protease inhibitors and 1% SDS. Total pro- incubated with E-cadherin (ab40772; Abcam), vimentin tein from uroepithelial cell lines was extracted using RIPA with (ab92547; Abcam) SH3BGRL3 (HPA001200; Sigma-Aldrich) protease and phosphatase inhibitors, and then centrifuged for and EGFR (sc-31157; Santa Cruz Biotechnology) primary anti- fi 15 minutes at 14,000 rpm at 4 C. The protein was quanti ed body overnight at 4 C. After washing three times with PBST, the using a Lowry assay and detected using an ELISA reader at an cells were incubated with or Alexa Fluor 594 (rabbit)-conju- optical density (OD) of 630 nm. For co-IP, total protein (1 mg) gated secondary antibody (Invitrogen) at room temperature for m was incubated with primary antibody (2 g) for 18 hours. The 1 hour. DAPI was used for nuclear staining. For image analysis, complex was then captured by protein G beads (Invitrogen) for cells were mounted and analyzed using a confocal microscope 4 hours and washed with cold PBS buffer containing protease (FV1000; Olympus) and photographed at 1800 or 2400 and phosphatase inhibitors. The electrophoresis was done magnification. using a polyacrylamide gel (Ready Gel System; Bio-Rad). Fifty micrograms of protein was separated on a 14% SDS-PAGE at Migration assays 100 V for 2 hours, and it was then transferred to a 0.22-mm The cell migration assay was done in 24-well Transwell plates polyvinylidene fluoride (PVDF) membrane (Millipore) at 100 (8.0 mm, pore size; Millipore). The cells were seeded in the V for 90 minutes. The membrane was blocked with 5% fat-free upper chamber of the Transwell system at a concentration of 2 milk for 1 hour at room temperature, and then hybridized with to 6 104 cells/well in 200 mLofmedium,andthelower SH3BGRL3 (Sigma-Aldrich), E-cadherin (ab40772, Abcam), chamber was filled with 600 mL normal culture medium. After vimentin (ab92547, Abcam), EGFR (sc-31157, Santa Cruz), 36 hours of incubation at 37 C, 5% CO2, the upper sides of the and b-catenin (GTX101435, GeneTex), raf-1 (GTX107763, filters were carefully washed with PBS, and the remaining cells GeneTex), GSK-3b (GTX111192, GSK-3b), at 4Covernight. were removed with a cotton swab. The cells that had migrated The secondary antibody (anti-rabbit, anti-mouse, and anti-goat to the bottom side of the filter were fixed with methanol/ IgG-horseradish peroxidase; Santa Cruz Biotechnology) was acetone (1:1) and stained using 0.2% crystal violet. The number used to amplify the signal as appropriate. The blots were finally of migrated cells was manually counted in three random fields

www.aacrjournals.org Clin Cancer Res; 2015 OF3

Chiang et al.

per filter at 6,200 magnification using a phase contrast Table S2), signifying the potential of SH3BGRL3 as a urine microscope. marker for urothelial carcinoma. Although SH3BGRL3 may be derived from blood incursion as a serum protein, our data are Statistical analysis consistent with the GENT database (gene expression across All values are reported as means SEM. Statistical comparisons normal and tumor tissue; ref. 17), showing a higher level of among the groups were carried out using analysis of variance SH3BGRL3 mRNA in urothelial carcinoma than in normal (ANOVA), followed by a t test and an F test of the means of tissue (Supplementary Fig. S1B). This study was thus designed fi normal, benign tumor, and urothelial carcinoma tissues. Datasets to investigate the biologic signi cance of SH3BGRL3 in urothe- that were not normally distributed were analyzed using the lial carcinoma. Mann–Whitney test for unpaired samples. Time to event distri- bution was estimated by means of cumulative incidence functions Expression profiling of SH3BGRL3 in the urine and primary to take into account the patients who died (competing risk) before tumors of patients with urothelial carcinoma progression (Fine and Gray Model). Cumulative incidence curves To confirm the prediction of proteomic analysis, 13 additional were plotted with the STATA software program version 11 (Stata urine samples from patients with urothelial carcinoma of various Corp.). The mean scores of protein expression are graphically histologic gradings and stages were measured for SH3BGRL3 presented using error bars with 95% confidence intervals (CI). levels. We found a positive relationship between detectable urine Significance was set at P < 0.05. SH3BGRL3 and histologic grading (P ¼ 0.0182; Fig. 2A and B) and muscle invasiveness (P ¼ 0.0006; Fig. 2C). A recent study reported that SH3BGRL3 is upregulated in the secretome of human adipose Results tissue-derived mesenchymal stem cells upon TNFa treatment Identification of SH3BGRL3 in the urine of patients with benign (18). SH3BGRL3 was classified as a nonclassical secretory protein prostatic hyperplasia and urothelial carcinoma (Supplementary Fig. S1C), based on analysis using the online Urinary proteins from patients with and without urothelial SecretomeP 2.0 Server (19). Taking these results together, carcinoma were analyzed using LC/MS-MS. A total of 219 SH3BGRL3 not only may appear in the cytoplasm, but also as candidateproteinswereidentified (Supplementary Table S1). a secretory protein in patients' urine. Themajoritywereidentified by at least two unique peptides. To clarify the clinical implications, expression pattern of The false discovery rate was determined to be less than 1% SH3BGRL3 was analyzed on paired urothelial carcinoma and based on a search of the reverse protein database. Among the nontumor tissue samples (n ¼ 32) by Western blotting (Fig. 2D). candidate proteins, both IgGs and serum albumin had the four Upregulated SH3BGRL3 (T > N) was detected in 71.8% of primary highest sequence coverages (>85%). As shown in Supplemen- tumors. tary Table S1, hemoglobin, IgGs, and serum albumin were A total of 55 cases of nonneoplastic uroepithelium had been more abundant in the pooled urine from patients with urothe- constructed as an array for IHC analysis. No evidence of lial carcinoma than from those without urothelial carcinoma SH3BGRL3 or EGFR expression was found in normal uroe- (ratio > 1). The results are consistent with prevailing notion that pithelium (data not shown). IHC analysis of clinical cohort hematuria is a common occurrence inpatientswithurothelial (14) demonstrated that SH3BGRL3 is expressed in 13.89% carcinoma, and that hemoglobin may therefore be detectable (159/,1145) of bladder cancer (Table 1 and Fig. 2F). during profiling (16). The accuracy of the mass spectra results SH3BGRL3 expression is significantly associated with overex- was confirmed by cytokertin and vimentin using Western pression of EGFR in primary bladder cancer (Table 1, P < blotting (Supplementary Fig. S1A). 0.0001). Higher level of SH3BGRL3 expression is significantly Paired analysis of the urine profiling for benign prostatic relatedtomuscleinvasion(Fig.2G,left;P ¼ 0.0028). In hyperplasia (Supplementary Table S2) and patients with patients with nonmuscle-invasive bladder cancer (NMIBC) urothelial carcinoma (Supplementary Table S1) found that treated by transurethral resection without intravesical instilla- SH3BGRL3 appears in both experiments. SH3BGRL3 was iden- tion (n ¼ 179), high SH3BGRL3 expression significantly cor- tified in this study by four unique peptides (Fig. 1A, under- related with increased risk of progression (Supplementary lined) with high-quality fragmentation spectra (ion score > 40), Table S3; Fig. 2H, P ¼ 0.032). VYSTSVTGSR (Fig. 1B), SQQSEVTR (Fig. 1C), ATPPQIVNGD- QYCGDYELFVEAVEQNTLQEFLK (Fig. 1D), and IQVQLV- Significance of SH3BGRL3 in cell transformation, EMT, and cell DISQDNALR (Fig. 1E), with 71% sequence coverage (Supple- migration mentary Table S1). These proteins also have the highest Expression pattern of SH3BGRL3 was assessed in a panel of sequence coverage in normal urine (8). Labeling (1H12C) uroepithelial cell lines. SH3BGRL3 expression was positively formaldehyde and isotopic (2D13C) formaldehyde resulted in associated with progression of urothelial carcinogenesis in vitro a mass difference of 6 Da for each labeling site or a difference of based on both RNA and protein levels (Supplementary Fig. S2A 3Dam/z for a doubly charged (z ¼ 2) peptide, with one and S2B). Consistent with positive association of detectable labeling site like VYSTSVTGSR (m/z 542.7882, 2þ;Fig.1E) urine SH3BGRL3 with histologic grading and muscle invasive- derived from SH3BGRL3. As indicated, the peptide ratio was ness, SH3BGRL3 was assumed to have a positive effect on cell determined from each isotopic pair (Fig. 1F), and the protein transformation and migration. Survey of the Gene Expression ratio represents the average of all ratios of peptides. Omnibus (GEO) database indicates that SH3BGRL3 may par- SH3BGRL3 was moderately upregulated in the pooled urine ticipate in TGFb-induced EMT (20). To further examine these from patients with urothelial carcinoma compared with those issues, transient transfection was carried out to explore the without urothelial carcinoma (average ratio: 1.4 0.1 [n ¼ 6]). significanceofSH3BGRL3withregardtoEMTphenotypes, It was ranked seventh in the protein list (Supplementary such as E-cadherin, b-catenin, and vimentin expression (21).

OF4 Clin Cancer Res; 2015 Clinical Cancer Research

SH3BGRL3 Is a Novel Biomarker of Urothelial Carcinoma

Figure 1. Identifying SH3BGRL3 in the urine of patients with urothelial carcinoma. A, underlining indicates the amino acid sequence of SH3BGRL3 identified by LC/MS-MS. B, collision-induced dissociation (CID) MS2 spectrum of VYSTSVTGSR tryptic peptide derived from SH3BGRL3. C, CID MS2 spectrum of SQQSEVTR (m/z 481.7492, 2þ) tryptic peptide derived from SH3BGRL3. D, CID MS2 spectrum of ATPPQIVNGDQYCGDYELFVEAVEQNTLQEFLK (m/z 1291.2999, 3þ) tryptic peptide derived from SH3BGRL3. E, CID MS2 spectrum of IQVQLVDISQDNALR tryptic peptide derived from SH3BGRL3. F, the extracted isotopic pair of VYSTSVTGSR precursor ions. The peptide ratio was calculated as urothelial carcinoma (heavy form)/non-urothelial carcinoma (light form). b-Catenin was downregulated by SH3BGRL3 overexpression, Growth promotion and tumorigenicity of SH3BGRL3- although vimentin expression remained unchanged (Supple- overexpressing stable cell lines mentary Fig. S2C). Cell growth kinetics examined by MTT assay showed a positive To elucidate the biologic significance of SH3BGRL3, three relationship between SH3BGRL3 and cell proliferation in stable cell lines were established from TSGH8301 cells TSGH8301, 5637, and T24 stable cell lines (Fig. 3E; , P < 0.05; with different levels of SH3BGRL3 expression (vector , P < 0.01; , P < 0.001). Cyclin D1, a G1–S phase transition control, SH3BGRL3-low and SH3BGRL3-high; Fig. 3A). Both marker, was also upregulated by SH3BGRL3 (Fig. 3E left-top small SH3BGRL3-overexpressing stable cells had an increased nucle- panel). ar–cytoplasmic ratio compared with vector control cells. In terms of tumorigenicity in vivo, xenografts of the SH3BGRL3- Expression of E-cadherin and b-catenin was downregulated high group were significantly larger than those of the vector or (Fig. 3B and D), whereas vimentin expression was obviously SH3BGRL3-low groups (Fig. 3E right, top; Supplementary Fig. higher in SH3BGRL3-overexpressing stable cell lines (Fig. 3C S2D, all Ps < 0.02). The protein expression of SH3BGRL3 in and D). The Transwell assay demonstrated a positive associ- xenografts was confirmed by Western blot analysis (Fig. 3E, ation between SH3BGRL3 expression and cell migration capac- right-top small panel). The hematoxylin and eosin (H&E) staining ity (Fig. 3F; , P < 0.05; , P < 0.01). of xenografts showed a poorly differentiated phenotype and more

www.aacrjournals.org Clin Cancer Res; 2015 OF5

Chiang et al.

Figure 2. Clinical implications of SH3BGRL3 expression in the urine and primary tumors. A, equal amounts of total protein were extracted from the urine of 13 patients with urothelial carcinoma with different stages. The relative levels of SH3BGRL3 protein were analyzed using immunoblotting of preoperatively voided urine. Detectable urinary SH3BGRL3 was associated with histologic grading (P ¼ 0.0182; B) and muscle invasiveness (C; P ¼ 0.0006). Protein extracted from urothelial carcinoma and nonurothelial carcinoma tissues was assayed using immunoblotting. Expression levels of SH3BGRL3 and EGFR were normalized by b-actin. Representative results show an upregulated SH3BGRL3 (D) and EGFR (E) in most of the primary UCs. F, left, showed representative example of high level of SH3BGRL3 expression in the membrane and focal cytoplasm of high-grade urothelial carcinoma. Right, showed aberrant strong cytoplasmic expression of SH3BGRL3 is demonstrated in the high-grade sarcomatous area compared with weak staining in the low-grade urothelial carcinoma element on the upper corner. G, the quantiﬁcation of IHC scoring of SH3BGRL3 in MIBC and NMIBC. Right, showed the quantiﬁcation of IHC scoring of EGFR in MIBC and NMIBC. H, progression cumulative incidence of SH3BGRL3 in patients with urothelial carcinoma. , P < 0.05; , P < 0.01; , P < 0.001 by t test, Mann–Whitney test for unpaired samples.

pleomorphic and bizarre nuclei in the SH3BGRL3-high group E7 immortalized urothelial cells and RT4 well-differentiated compared with the vector or SH3BGRL3-low groups (Supple- cancer cells. However, ﬂuorescent signals were abundant in the mentary Fig. S2E). nuclei of TSGH8301, J82, and T24 cancer cells, and the cytoplasmic signals of SH3BGRL3 were presented only in The interaction of SH3BGRL3 with EGFR by proline-rich TSGH8301, J82 and T24 metastatic cancer cells (Supplementary residues through Grb2 Fig. S3A). Confocal microscopy showed that SH3BGRL3 was localized Schulze and colleagues (22) reported that SH3BGRL family near the plasma membrane, with a few signals in the nuclei of proteins may interact with receptor tyrosine kinases (RTK), such as EGFR and HER2, two important players in urothelial carci- Table 1. The association of EGFR and SH3BGRL3 nogenesis (4). This prediction was supported by colocalization ﬁ SH3BGRL3 status of SH3BGRL3 with EGFR in all ve uroepithelial cell lines, Negative Positive Total irrespective of the nuclear or membranous expression patterns. EGFR status Both co-IP (Fig. 4A and Supplementary Fig. S3E) and confocal Negative and low 925 (89.55%) 108 (10.45%) 1033 (100%) microscopy (Supplementary Fig. S3B) demonstrated the inter- High 61 (54.46%) 51 (45.54%) 112 (100%) action of SH3BGRL3 with phosphor-EGFR from 10 to 30 Total minutes after EGF (10 ng/mL) treatment. The interaction of 986 (86.11%) 159 (13.89%) 1,145 (100%) SH3BGRL3 with EGFR was also demonstrated in cancer cell Pearson x2 ¼ 103.9889 P < 0.0001 lines of other cell types (Supplementary Fig. S3D). With regard

OF6 Clin Cancer Res; 2015 Clinical Cancer Research

SH3BGRL3 Is a Novel Biomarker of Urothelial Carcinoma

Figure 3. Induction of cell transformation, EMT, cell migration, and proliferation by SH3BGRL3. A, H&E staining of SH3BGRL3 stable cell lines showed spindle-shaped phenotypes and discohesiveness compared with the vector control group. Confocal microscopy of EMT marker expression was performed on vector and SH3BGRL3 stable cell lines. B, red fluorescence represents E-cadherin. The nuclei (blue fluorescence) were stained by DAPI. C, red fluorescence represents vimentin. D, immunoblotting of EMT markers, including E-cadherin, b-catenin, and vimentin, was performed on vector and SH3BGRL3 stable cell lines in TSGH8301, 5637, and T24. E, SH3BGRL3-induced cell proliferation in TSGH8301 (left, top), 5637 (left, bottom), and T24 (right, top) was assessed by MTT assay for 3 to 5 days. The results (means SD) were derived from three independent experiments. Expression of cyclin D1 in TSGH8301 stable cell lines is shown in the small panel. SCID mice (6–8 weeks old) were subcutaneously injected with vector or SH3BGRL3-overexpressing stable cell lines (right, bottom). The size of tumors was measured after 6 weeks. Expression level of SH3BGRL3 in tumors is shown in the small panel. F, after starving for 24 hours, TSGH8301 (top), 5637 (middle), and T24 (bottom) stable cell lines (1–2 104) were seeded in the upper chamber of trans-wells and incubated with complete culture medium for 18 to 24 hours. Number of migrated cells was counted by microscopy after staining. Ten fields of migrated cells from each trans-well were counted by microscopy. The results (means SD) were derived from three independent experiments. , P < 0.05; , P < 0.01 by paired t test.

to binding site(s) on the EGFR, SH3BGRL3 binding was abol- phosphor-Akt associated with SH3BGRL3 overexpression, together ished by Y1173F mutant, whereas partial blockage was with downregulated phosphor-Erk1/2 (Fig. 4D). However, this observed in Y1068F and Y1086F mutants (Fig. 4B and Supple- phenomenon was reversed when SH3BGRL3/PP mutant was mentary S3C). It is well known that Y1173 is related to EGFR transfected. A cell-proliferation assay further showed the growth- de-phosphorylation and directly bound by SHP-1, a protein stimulatory effect of SH3BGRL3 in TSGH8301 cells compared tyrosine phosphatase, and indirectly by Grb2 (23). Both Y1068 with SH3BGRL3/PP mutant (Fig. 4E, all, P < 0.0004). Endogenous and Y1086 of EGFR were also bound by Grb2 (24, 25), and SH3BGRL3 interacted with EGFR on the cell membrane, cytosol, associated with the MAPK pathway. and nucleus of TSGH8301 cells after EGF treatment (Fig. 4F). Protein structure analysis suggested that SH3BGRL3 lacks the SH2 domain (10, 26–29) in directly binding to phosphorylated Activation of Akt-associated signaling cascade by SH3BGRL3 residues of EGFR. This result implies that SH3BGRL3 might indi- In terms of signaling events associated with SH3BGRL3, upre- rectly interact with phosphorylated EGFR through other adaptor gulated phosphor-Akt in the SH3BGRL3-overexpressing stable proteins. We thus examined the interactions of SH3BGRL3 with cell line was demonstrated after EGF treatment (10 ng/mL), the SHP-1, Grb2, and PI3K p85 subunits. Co-IP demonstrated together with inhibition of GSK3-b (Ser 9; Fig. 5A). In contrast, the interaction of SH3BGRL3 with Grb2 after EGF treatment both phosphor-Raf-1 and phosphor-Erk1/2 were suppressed (10 ng/mL) by its proline-rich motif, but not with the SHP-1 or compared with vector control. Moreover, b-catenin, a down- PI3K p85 subunits (Fig. 4C). The SOS1 was also captured in stream target of GSK3-b in the gene trans-activation (30), was SH3BGRL3 co-IP protein complex (Supplementary Fig. S3F). trans-located to a greater extent into the nuclei of SH3BGRL3-high Transient transfection of SH3BGRL3 and SH3BGRL3/PP mutant stable cells after EGF treatment, as compared with vector control (proline!alanine) in TSGH8301 cells revealed upregulated (Fig. 5B, P ¼ 0.023). Nuclear translocation of b-catenin was

www.aacrjournals.org Clin Cancer Res; 2015 OF7

Chiang et al.

Figure 4. Interaction of SH3BGRL3 with EGFR through Grb2 by its proline-rich motif. A, HEK293-SH3BGRL3–overexpressing stable cell line was transiently transfected with EGFR for 48 hours, starved for 24 hours, and treated with EGF (10 ng/ mL) at different time points. Western blotting was performed on Co-IP using anti-EFGR antibody. Negative control (NC) was the lysate of the HEK293-SH3BGRL3–overexpressing stable cell line, which had been starved for 24 hours. Positive control (PC) was the lysate of the T24 cell line, which had been starved for 24 hours. B, HEK293-SH3BGRL3–overexpressing stable cell line was transiently transfected with wild-type EGFR or different EGFR mutants for 48 hours, starved for 24 hours, and treated with EGF (10 ng/mL) for 15 minutes. PC was the lysate of the T24 cell line, which had been starved for 24 hours. C, TSGH8301 was transiently transfected with p3xFlag-SH3BGRL3, p3xFlag-SH3BGRL3/PP mutant constructs, or control vector for 24 hours, starved for 24 hours, and stimulated with EGF (10 ng/mL) for 15 minutes. Total lysate was immunoprecipitated by anti-Flag antibody and probed by Grb2, SHP-1, and PI3Kp85-b subunit antibodies, respectively. The input samples were 5% of IP lysate. D, TSGH8301 with wild-type EGFR expression was transiently transfected with wild-type SH3BGRL3 or proline-rich motif mutant, starved for 24 hours and treated with EGF (10 ng/ mL) for 5 minutes. The total lysate was detected by anti-Akt, anti-phosphor-Akt (Ser473), anti-Erk1/2, and anti-phosphor-Erk1/2 (T202 and T204), respectively. Cells treated with Akt inhibitor VIII and PD98059 were used as negative controls. The number of ratio was normalized by total protein. E, TSGH8301 was transiently transfected with p3xFlag-SH3BGRL3, p3xFlag-SH3BGRL3/PP mutant constructs, or control vector for 24 hours. The growth kinetics was examined using MTT assay for 4 days. The results (means SD) were derived from three independent experiments. F, confocal microscopy showed that endogenous SH3BGRL3 (green ﬂuorescence) is colocalized with EGFR (red ﬂuorescence) in TSGH8301 cells. The results (means SD) were derived from three independent experiments. , P < 0.01; , P < 0.001 by two-way ANOVA.

signiﬁcantly inhibited by pretreatment with LY294002 or Akt (Fig. 5D, P ¼ 0.0059). However, PD98059 had no effects on inhibitor VIII (Fig. 5B and Supplementary Fig. S4A and S4B, cell migration and proliferation in the SH3BGRL3-high stable P ¼ 1.095 10 8 and 1.093 10 8, respectively). The stimu- cell line. Knockdown of Raf-1 (Supplementary Fig. S4C) latory effect of SH3BGRL3 on cell migration could be sup- inhibited cell proliferation in the vector group (Supplementary pressed by preincubation with LY294002 and Akt inhibitor Fig. S4D, top), but not in the SH3BGRL3-high group (Sup- VIII (Fig. 5C, P ¼ 0.0019 and 0.0009, respectively). Cell plementary Fig. S4D, bottom). Furthermore, phosphorylation proliferation was also inhibited by Akt inhibitor VIII of Raf-1 was downregulated by SH3BGRL3 and was reactivated

OF8 Clin Cancer Res; 2015 Clinical Cancer Research

SH3BGRL3 Is a Novel Biomarker of Urothelial Carcinoma

Figure 5. EGFR signaling pathway mediated by SH3BGRL3. A, the EGFR signaling pathway was examined in vector and SH3BGRL3 stable cell lines, respectively. All cells were starved for 24 hours, and treated with EGF (10 ng/mL) at different time points. The phosphorylation status of signaling molecules, including EGFR, Akt, Erk1/2, Raf-1, and GSK3-b, were examined by immunoblotting. The ratio of phosphor–Akt and phosphor–Erk1/2 were normalized by total protein. The b-actin and Grb2 were used as internal control. B, nuclear translocation of b-catenin was examined by confocal microscopy, and showed a higher proportion in SH3BGRL3 stable cell lines compared with vector control. Nuclear translocation of b-catenin was inhibited by Akt inhibitor VIII and LY294002. C, the effect of Akt inhibitor VIII, LY294002, and PD98059 on cell migration was assessed using the Transwell assay. The SH3BGRL3 stable cell line was preincubated with inhibitors for 24 hours. Ten ﬁelds of migrated cells from each well were counted by microscopy. D, inhibition of cell proliferation by Akt inhibitor VIII or PD98059 was measured using MTT assay for 4 days. The SH3BGRL3 stable cell line was preincubated with inhibitors for 24 hours. The results (means SD) were derived from three independent experiments. , P < 0.01; , P < 0.001 by two-way ANOVA. E, both Akt- and Erk-related cascades are major signaling pathways of EGFR induced by EGF. SH3BGRL3 interacts with EGFR at Y1068, Y1086, and Y1173 through Grb2 by its proline-rich motif and blocks the Erk-related cascade. Then, Akt-related cascade in cancer cells is activated for cell survival and growth. The downstream molecule-GSK3-b is phosphorylated at serine 9 with inhibition of its kinase activity, resulting in translocation of b-catenin into the nucleus. in the presence of SH3BGRL3/PP mutant (Supplementary Fig. Discussion S4E). These results suggest that SH3BGRL3 suppresses the Raf- We found that detectable urinary SH3BGRL3 is positively 1–Erk1/2 axis signaling, accompanied by activation of the associated with histologic grading and muscle invasiveness of PI3K–Akt-related signaling cascade. UCs. This result seems to concur with bioinformatics analysis of

www.aacrjournals.org Clin Cancer Res; 2015 OF9

Chiang et al.

protein structure (19), which predicts that SH3BGRL3 belongs to SH3BGRL3 may be a molecular switch guiding cells to the the nonclassical secretory protein family (Supplementary Fig. PI3K–Akt axis after EGF treatment. S1C). A comparable expression pattern has been reported for Both Ras–ERK (mitogenic) and PI3K–AKT–mTOR (survival) urinary MMP-9 content (31), carcinoembryonic antigen-related signaling pathways are key mechanisms for cell survival in cell adhesion molecule 1 (32), and fibronectin (33) in patients response to extracellular cues (41). Currently, we have no ratio- with urothelial carcinoma. Therefore, measurement of urine nale to explain why SH3BGRL3 activates AKT, but would inhibit SH3BGRL3 may become a prognostic biomarker of urothelial ERKs in cancer cells. It is well known that cancer cells must develop carcinoma. stress tolerance machinery to survive under harsh conditions, such Tissue array analysis of clinical cohort demonstrated that as nutrient starvation, hypoxia, and pH change in the microenvir- SH3BGRL3 is positively associated with muscle invasion of onments (42, 43). This adaptation capability contributes to bladder cancer and an increased risk of progression for patients malignant progression and correlates with a poor clinical out- with NMIBC. These results concur with the association of come in several types of cancers (44). The results of this study SH3BGRL3 expression with progression of bladder cancer pre- suggest that SH3BGRL3 initiates the cross-inhibition of PI3K– dicted by the GEO database in National Center for Biotechnology AKT–mTOR signaling on the Ras–Erk pathway (41), thereby Information (GDS1479; Supplementary Fig. S1F; ref. 34). In driving cancer cell toward oncogene addiction. The enhanced addition, analysis of Kaplan–Meier Plotter online software sensitivity of SH3BGRL3-overexpressing tumor cells to gefitinib revealed a lower survival rate in patients with a high level of treatment in vitro support our hypothesis (Supplementary Fig. SH3BGRL3 expression in lung adenocarcinoma and breast cancer S4F). Nevertheless, more work is required to elucidate the with nodal metastasis (Fig. S1D and S1E; ref. 35). Accordingly, mechanisms underlying SH3BGRL3 in the pathway integration evaluation of SH3BGRL3 expression status may identify a subset of cancer cells. of patients with bladder cancer who may require more intensive Taking these findings together, SH3BGRL3 is a novel urinary treatment. biomarker for urothelial carcinoma. Evaluation of SH3BGRL3 In SH3BGRL3 stable cells, both spindle-shaped phenotypes and expression status may identify a subset of patients with urothelial discohesiveness of cancer cells are associated with upregulated EMT carcinoma for cotargeting candidates in the design of EGFR-based markers (suppressed E-cadherin and b-catenin with increased cancer therapies. vimentin expression) in vitro.Thefinding that aberrant cytoplasmic expression of SH3BGRL3 appears in the sarcomatous area of Disclosure of Potential Conflicts of Interest bladder cancer (Fig. 2F) supports for our observation in vitro. No potential conflicts of interest were disclosed. Further analysis of the GEO database revealed that SH3BGRL3 expression is positively associated with EMT markers, for example, Authors' Contributions TGFb (GDS3710; Supplementary Fig. S1G; ref. 20), Smad3-related b Conception and design: C.-Y. Chiang, N.-H. Chow pathway (GDS3985; Supplementary Fig. S1H; ref. 36), TGF , Development of methodology: C.-Y. Chiang, H.-S. Liu, S.-H. Chen, N.-H. Chow AP1510 (GDS4361; Supplementary Fig. S1I; ref. 37), and EGF Acquisition of data (provided animals, acquired and managed patients, (GDS2146; Supplementary Fig. S1J; ref. 38). The cell-cycle marker provided facilities, etc.): C.-Y. Chiang, C.-C. Pan, T.-S. Tzai, Y.-S. Tsai, cyclin D1 was also upregulated by SH3BGRL3, suggesting that B.-F. Lee, H.-L. Cheng, S.-H. Chen, N.-H. Chow SH3BGRL3 promotes cell growth in vitro and in vivo by accelerating Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): C.-Y. Chiang, C.-C. Pan, S.-H. Chen, N.-H. Chow the G –S phase transition. The results are consistent with our data 1 Writing, review, and/or revision of the manuscript: C.-Y. Chiang, H.-Y. Chang, that overexpression of SH3BGRL3 promotes the growth and S.-H. Chen, N.-H. Chow mobility of urothelial carcinoma in vitro. Administrative, technical, or material support (i.e., reporting or organizing Using co-IP and confocal microscopy, SH3BGRL3 was sug- data, constructing databases): C.-Y. Chiang, H.-Y. Chang, M.-D. Lai, Y.-S. Tsai, gested to interact with EGFR in vitro. The observation gains P. Ling, H.-S. Liu, C.-L. Ho, S.-H. Chen, N.-H. Chow strong support from the significant association of SH3BGRL3 Study supervision: C.-Y. Chiang, Y.-S. Tsai, H.-S. Liu, N.-H. Chow expression with overexpression of EGFR in primary bladder cancer. EGFR has been proved to be an important prognostic Acknowledgments factor in urothelial carcinoma (39). It is interesting to note The authors are grateful to Dr. Lai Ming-Derg for providing all EGFR-related that both EGFR interactome (22) and protein structure of plasmids, the Q-TOF-MS/MS analysis that was carried out by the Instrument fi SH3BGRL3 (9) have been characterized in earlier works. How- Center of National Cheng Kung University (NCKU), and for the tissue paraf n sections and H&E staining by the Department of Pathology, NCKU Hospital, ever, SH3BGRL3 does not contain an SH2 domain, which is Tainan, Taiwan. responsible for docking to phosphorylated tyrosine residues. We provided evidence that SH3BGRL3 interacts with EGFR Grant Support through Grb2, an adaptor protein involved in the Raf-1– This work was supported by research grants NSC-98-2320-B-006-022-MY3 Erk1/2 axis. In addition, an inhibitory effect on the Raf-1– – and NSC-102-2320-B-006-010-MY3 from the Ministry of Science and Technol- Erk1/2 axis, together with activation of PI3K Akt signaling, was ogy, Taiwan and NHRI-EX100-9930BI from the National Health Research demonstrated in a transient transfection experiment. To clarify Institutes, Taiwan. the importance of the proline-rich motif (10, 29) in the inter- The costs of publication of this article were defrayed in part by the actions with other proteins (40), mutation of this motif in payment of page charges. This article must therefore be hereby marked SH3BGRL3 blocked the interaction of SH3BGRL3 with Grb2, advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. and inhibited cell growth. As shown in the SH3BGRL3-overexpressing stable cell line, nuclear translocation of b-catenin, cell migration, and growth rate were all suppressed by PI3K or Received December 26, 2014; revised July 7, 2015; accepted July 27, 2015; Akt inhibitor as opposed to PD98059. These results suggest that published OnlineFirst August 18, 2015.

OF10 Clin Cancer Res; 2015 Clinical Cancer Research

SH3BGRL3 Is a Novel Biomarker of Urothelial Carcinoma

References 1. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2015. CA Cancer J Clin tyrosine phosphatase SHP-1 to the epidermal growth factor receptor and 2015;65:5–29. attenuation of receptor signaling. J Biol Chem 1998;273:24839–46. 2. Kirkali Z, Tuzel E. Transitional cell carcinoma of the ureter and renal pelvis. 24. Batzer AG, Rotin D, Urena JM, Skolnik EY, Schlessinger J. Hierarchy of Crit Rev Oncol Hematol 2003;47:155–69. binding sites for Grb2 and Shc on the epidermal growth factor receptor. 3. Hall MC, Womack S, Sagalowsky AI, Carmody T, Erickstad MD, Roehrborn Mol Cell Biol 1994;14:5192–201. CG. Prognostic factors, recurrence, and survival in transitional cell carci- 25. Zhang S, Weinheimer C, Courtois M, Kovacs A, Zhang CE, Cheng AM, et al. noma of the upper urinary tract: a 30-year experience in 252 patients. The role of the Grb2-p38 MAPK signaling pathway in cardiac hypertrophy Urology 1998;52:594–601. and fibrosis. J Clin Invest 2003;111:833–41. 4. Chow NH, Chan SH, Tzai TS, Ho CL, Liu HS. Expression profiles of ErbB 26. Nardini M, Mazzocco M, Massaro A, Maffei M, Vergano A, Donadini A, family receptors and prognosis in primary transitional cell carcinoma of the et al. Crystal structure of the glutaredoxin-like protein SH3BGRL3 at urinary bladder. Clin Cancer Res 2001;7:1957–62. 1.6 Angstrom resolution. Biochem Biophys Res Commun 2004;318: 5. Gygi SP, Rist B, Gerber SA, Turecek F, Gelb MH, Aebersold R. Quantitative 470–6. analysis of complex protein mixtures using isotope-coded affinity tags. Nat 27. Mazzocco M, Arrigo P, Egeo A, Maffei M, Vergano A, Di Lisi R, et al. A novel Biotechnol 1999;17:994–9. human homologue of the SH3BGR gene encodes a small protein similar to 6. Hsu JL, Huang SY, Chen SH. Dimethyl multiplexed labeling combined Glutaredoxin 1 of Escherichia coli. Biochem Biophys Res Commun with microcolumn separation and MS analysis for time course study in 2001;285:540–5. proteomics. Electrophoresis 2006;27:3652–60. 28. Berleth ES, Nadadur S, Henn AD, Eppolito C, Shiojiri S, Gurtoo HL, et al. 7. Chen SH, Wang YW, Hsu JL, Chang HY, Wang CY, Shen PT, et al. Nucleo- Identification, characterization, and cloning of TIP-B1, a novel protein inhib- phosmin in the pathogenesis of arsenic-related bladder carcinogenesis re- itor of tumor necrosis factor-induced lysis. Cancer Res 1999;59:5497–506. vealed by quantitative proteomics. Toxicol Appl Pharmacol 2010;242:126–35. 29. Egeo A, Mazzocco M, Arrigo P, Vidal-Taboada JM, Oliva R, Pirola B, et al. 8. Cheng HL, Huang HJ, Ou BY, Chow NH, Chen YW, Tzai TS, et al. Urinary Identification and characterization of a new human gene encoding a small CD14 as a potential biomarker for benign prostatic hyperplasia—discovery protein with high homology to the proline-rich region of the SH3BGR by combining MALDI-TOF-based biostatistics and ESI-MS/MS-based sta- gene. Biochem Biophys Res Commun 1998;247:302–6. ble-isotope labeling. Proteomics Clin Appl 2011;5:121–32. 30. Chandra V, Fatima I, Manohar M, Popli P, Sirohi VK, Hussain MK, et al. 9. Xu C, Zheng P, Shen S, Xu Y, Wei L, Gao H, et al. NMR structure and Inhibitory effect of 2-(piperidinoethoxyphenyl)-3-(4-hydroxyphenyl)- regulated expression in APL cell of human SH3BGRL3. FEBS Lett 2005; 2H-benzo(b)pyran (K-1) on human primary endometrial hyperplasial 579:2788–94. cells mediated via combined suppression of Wnt/b-catenin signaling and 10. Mazzocco M, Maffei M, Egeo A, Vergano A, Arrigo P, Di Lisi R, et al. The PI3K/Akt survival pathway. Cell Death Dis 2014;5:e1380. identification of a novel human homologue of the SH3 binding glutamic 31. Offersen BV, Knap MM, Horsman MR, Verheijen J, Hanemaaijer R, Over- acid-rich (SH3BGR) gene establishes a new family of highly conserved gaard J. Matrix metalloproteinase-9 measured in urine from bladder cancer small proteins related to Thioredoxin Superfamily. Gene 2002;291:233–9. patients is an independent prognostic marker of poor survival. Acta Oncol 11. Egeo A, Mazzocco M, Sotgia F, Arrigo P, Oliva R, Bergonon S, et al. Iden- 2010;49:1283–7. tification and characterization of a new human cDNA from chromosome 32. Tilki D, Singer BB, Shariat SF, Behrend A, Fernando M, Irmak S, et al. 21q22.3 encoding a basic nuclear protein. Hum Genet 1998;102:289–93. CEACAM1: a novel urinary marker for bladder cancer detection. Eur Urol 12. Berleth ES, Henn AD, Gurtoo HL, Wollman R, Alderfer JL, Mihich E, et al. A 2010;57:648–54. novel tumor necrosis factor-alpha inhibitory protein, TIP-B1. Int J Immu- 33. Menendez V, Fernandez-Suarez A, Galan JA, Perez M, Garcia-Lopez F. nopharmacol 2000;22:1137–42. Diagnosis of bladder cancer by analysis of urinary fibronectin. Urology 13. Akinci MA, Turner H, Taveras M, Wolosin JM. Differential gene expression 2005;65:284–9. in the pig limbal side population: implications for stem cell cycling, 34. Zieger K, Dyrskjot L, Wiuf C, Jensen JL, Andersen CL, Jensen KM, et al. Role replication, and survival. Invest Ophthalmol Vis Sci 2009;50:5630–8. of activating fibroblast growth factor receptor 3 mutations in the devel- 14. Sun CH, Chang YH, Pan CC. Activation of the PI3K/Akt/mTOR pathway opment of bladder tumors. Clin Cancer Res 2005;11:7709–19. correlates with tumour progression and reduced survival in patients with 35. Gyorffy B, Lanczky A, Eklund AC, Denkert C, Budczies J, Li Q, et al. An urothelial carcinoma of the urinary bladder. Histopathology 2011;58: online survival analysis tool to rapidly assess the effect of 22,277 genes on 1054–63. breast cancer prognosis using microarray data of 1,809 patients. Breast 15. Liu HS, Hsu PY, Lai MD, Chang HY, Ho CL, Cheng HL, et al. An unusual Cancer Res Treat 2010;123:725–31. function of RON receptor tyrosine kinase as a transcriptional regulator in 36. Yadav H, Quijano C, Kamaraju AK, Gavrilova O, Malek R, Chen W, et al. cooperation with EGFR in human cancer cells. Carcinogenesis 2010;31: Protection from obesity and diabetes by blockade of TGF-b/Smad3 sig- 1456–64. naling. Cell Metab 2011;14:67–79. 16. Yang N, Feng S, Shedden K, Xie X, Liu Y, Rosser CJ, et al. Urinary 37. Ghosh R, Narasanna A, Wang SE, Liu S, Chakrabarty A, Balko JM, et al. glycoprotein biomarker discovery for bladder cancer detection using LC/ Trastuzumab has preferential activity against breast cancers driven by HER2 MS-MS and label-free quantification. Clin Cancer Res 2011;17:3349–59. homodimers. Cancer Res 2011;71:1871–82. 17. Shin G, Kang TW, Yang S, Baek SJ, Jeong YS, Kim SY. GENT: gene expression 38. Liu B, Chen H, Johns TG, Neufeld AH. Epidermal growth factor receptor database of normal and tumor tissues. Cancer Inform 2011;10:149–57. activation: an upstream signal for transition of quiescent astrocytes into 18. Lee MJ, Kim J, Kim MY, Bae YS, Ryu SH, Lee TG, et al. Proteomic analysis of reactive astrocytes after neural injury. J Neurosci 2006;26:7532–40. tumor necrosis factor-alpha-induced secretome of human adipose tissue- 39. Carlsson J, Wester K, De La Torre M, Malmstrom PU, Gardmark T. EGFR- derived mesenchymal stem cells. J Proteome Res 2010;9:1754–62. expression in primary urinary bladder cancer and corresponding metas- 19. Bendtsen JD, Jensen LJ, Blom N, Von Heijne G, Brunak S. Feature-based tases and the relation to HER2-expression. On the possibility to target these prediction of non-classical and leaderless protein secretion. Protein Eng receptors with radionuclides. Radiol Oncol 2015;49:50–8. Des Sel 2004;17:349–56. 40. Yamada E, Bastie CC. Disruption of Fyn SH3 domain interaction with a 20. Sartor MA, Mahavisno V, Keshamouni VG, Cavalcoli J, Wright Z, Karnovsky proline-rich motif in liver kinase B1 results in activation of AMP-activated A, et al. ConceptGen: a gene set enrichment and gene set relation mapping protein kinase. PloS One 2014;9:e89604. tool. Bioinformatics 2010;26:456–63. 41. Mendoza MC, Er EE, Blenis J. The Ras-ERK and PI3K-mTOR pathways: 21. Pirozzi G, Tirino V, Camerlingo R, Franco R, La Rocca A, Liguori E, et al. cross-talk and compensation. Trends Biochem Sci 2011;36:320–8. Epithelial to mesenchymal transition by TGFb-1 induction increases stem- 42. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell 2000;100:57–70. ness characteristics in primary non small cell lung cancer cell line. PloS One 43. Chang HY, Liu HS, Lai MD, Tsai YS, Tzai TS, Cheng HL, et al. Hypoxia 2011;6:e21548. promotes nuclear translocation and transcriptional function in the onco- 22. Schulze WX, Deng L, Mann M. Phosphotyrosine interactome of the ErbB- genic tyrosine kinase RON. Cancer Res 2014;74:4549–62. receptor kinase family. Mol Syst Biol 2005;1:2005 0008. 44. Brizel DM, Scully SP, Harrelson JM, Layfield LJ, Bean JM, Prosnitz LR, et al. 23. Keilhack H, Tenev T, Nyakatura E, Godovac-Zimmermann J, Nielsen L, Tumor oxygenation predicts for the likelihood of distant metastases in Seedorf K, et al. Phosphotyrosine 1173 mediates binding of the protein- human soft tissue sarcoma. Cancer Res 1996;56:941–3.

www.aacrjournals.org Clin Cancer Res; 2015 OF11

SH3BGRL3 Protein as a Potential Prognostic Biomarker for Urothelial Carcinoma: A Novel Binding Partner of Epidermal Growth Factor Receptor

Cheng-Yao Chiang, Chin-Chen Pan, Hong-Yi Chang, et al.

Clin Cancer Res Published OnlineFirst August 18, 2015.

Updated version Access the most recent version of this article at: doi:10.1158/1078-0432.CCR-14-3308

Supplementary Access the most recent supplemental material at: Material http://clincancerres.aacrjournals.org/content/suppl/2015/08/19/1078-0432.CCR-14-3308.DC1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/early/2015/11/17/1078-0432.CCR-14-3308. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.