Protein Expression Profile Related to Cisplatin Resistance in Bladder Cancer Cell Lines Detected by Two-Dimensional Gel Electrophoresis

Biomedical Research (Tokyo) 36 (4) 253-261, 2015

Protein expression profile related to cisplatin resistance in bladder cancer cell lines detected by two-dimensional gel electrophoresis

1 1 2 1 1 Yoshinori TAOKA , Kazumasa MATSUMOTO , Kazuya OHASHI , Satoru MINAMIDA , Masahiro HAGIWARA , Shoji 1 2 2 1 NAGI , Tatsuya SAITO , Yoshio KODERA , and Masatsugu IWAMURA 1 Department of Urology, Kitasato University School of Medicine, Sagamihara, Kanagawa 252-0374, Japan and 2 Department of Phys- ics, Kitasato University School of Science, Sagamihara, Kanagawa 252-0373, Japan (Received 9 June 2015; and accepted 17 June 2015)

ABSTRACT We used a proteomic approach to compare the differentially regulated protein expression profiles of cisplatin-naïve and cisplatin-resistant bladder cancer cell lines to screen candidate molecules related to cisplatin resistance. The cisplatin-resistant cell line T24 was established by the stepwise exposure of T24 cells to up to 40 μM of cisplatin. We performed a comprehensive study of protein expression in bladder cancer cell lines that included cisplatin-naïve (T24) and cisplatin-resistant cells (T24CDDPR) by means of agarose two-dimensional gel electrophoresis followed by analysis of liquid chromatography tandem mass spectroscopy. We identified 25 obviously different spots for T24 and T24 CDDPR. Seven spots had increased expression and 18 spots had decreased expression in T24CDDPR compared to those in T24. Cytoskeletal proteins and enzyme modulators were prominent among differential proteins. Of the 25 proteins, we selected HNRNPA3, PCK2, PPL, PGK1, TKT, SERPINB2, GOT2, and EIF3A for further validation by Western blot. HNRNPA3, PGK1, TKT, and SERPINB2 had more than 1.5-times incremental expression in T24CDDPR compared to that in T24. PCK2 and PPL expressions were decreased less than 20% in T24CDDPR compared to that in T24. The results of 25 new proteins in this study could be valuable and could lead to the development of a new molecular marker.

Bladder cancer is the second most common malig- Cisplatin-based chemotherapy, including methotrex- nant tumor in the urologic field. Patients with locally ate, vinblastine, doxorubicin, and cisplatin (MVAC) advanced bladder cancer at diagnosis will subse- or gemcitabine and cisplatin (GC), is an effective quently have development of local progression or and frequently used modality for advanced and met- distant metastases. In patients with advanced cancer, astatic bladder cancer (21). chemotherapy is the treatment of choice. Since cis- However, resistance to chemotherapeutic agents is diamminedichloroplatinum, known as cisplatin, for a major clinical problem in the effective manage- the treatment of testicular and bladder cancers was ment of cancer. Long-term follow-up has been re- approved by the Food and Drug Administration in ported in patients with advanced bladder cancer, but 1978, platinum-based drugs have been the main treat- progression-free survival and overall survival have ment administered against a variety of cancers (3). not been satisfactory (16). To overcome drug resistance, alteration of chemotherapy has been performed, including increased dose, shortened intervals, Address correspondence to: Kazumasa Matsumoto, M.D. Department of Urology, Kitasato University School of or a variety of combination strategies. Unfortunately, Medicine, 1-15-1 Kitasato, Minami-ku, Sagamihara, in terms of bladder cancer, there are no established Kanagawa 252-0374, Japan therapeutic modalities for patients with progression Tel: +81-42-778-9091, Fax: +81-42-778-9374 or recurrence after cisplatin-based treatment (5, 8). E-mail: [email protected] Proteomics using cancer specimens is promising 254 Y. Taoka et al. for detecting clinically available biomarkers, which to 0.50-mm glass beads (As One Corporation, Osa- eventually provide the information for an appropri- ka, Japan) and centrifuged at 112,000×g for 20 min. ate clinical choice and development of innovative The supernatants were subjected to agarose 2-DE. therapy. Such a proteomic approach has been used not only for a host of tumor tissue biopsy specimens Agarose 2-DE and MS/MS analysis. Protein extracts but also for cultured tumor cells. To identify appro- were separated by agarose 2-DE. The isoelectric fo- priate proteins for biomarkers, we utilized a unique cusing of the first dimension used an agarose gel, two-dimensional gel electrophoresis (2-DE) method and that of the second dimension used an SDS-PAGE that uses an agarose gel for isoelectric focusing on gel. We used 12% SDS-PAGE gel for the proteins the first dimension. Agarose 2-DE can analyze much weighing between 30 kDa and 100 kDa, and we larger quantities of proteins (up to 1.5 mg) and a used a linear gradient of 6% to 10% SDS-PAGE gel wider dynamic range than conventional 2-DE (14). for the proteins weighing more than 50 kDa. Aga- In our previous studies, proteomic analyses were rose 2-DE was performed on three sets of T24 and continuously investigated in urothelial carcinoma to T24CDDPR bladder cancer cell lines. The slab gel identify the candidates for biomarkers, including pro- was stained with CBB R-350 (PhastGel Blue R; GE teins, sera, and autoantibodies (11, 14, 19, 20). Healthcare, Little Chalfont, UK). We calculated the In this present study, we used a proteomic ap- molecular weight of the protein spot using CS ana- proach to compare the differentially regulated protein lyzer 2.0 (Atto, Tokyo, Japan) computer software. expression profiles of cisplatin-naïve and cisplatin- We subjected each spot to protein digestion and tan- resistant bladder cancer cell lines to screen candi- dem mass spectroscopy (MS/MS) analysis using a date molecules related to cisplatin resistance. liquid chromatography tandem mass spectroscopy (LC-MS/MS) system, which consisted of a Nano- space SI-2 HPLC system (Shiseido Fine Chemicals, MATERIALS AND METHODS Tokyo, Japan) and an ion-trap mass spectrometer Cell line and cell culture. Human bladder cancer (LCQ Deca; Thermo Fisher Scientific, San Jose, cell lines T24, RT4, 5637, and TCCSUP were pur- CA, USA). chased from American Type Culture Collection. EJ We used the SEQUEST search program to identi- was kindly gifted from the Kitasato Institution Hos- fy each protein. SEQUEST referenced the nr.Z and pital (Dr. Akira Irie). The cisplatin-resistant cell line Swiss-Prot.Z protein sequence databases download- T24 (T24CDDPR) was established by the stepwise ed from ftp://ftp.ncbi.nih.gov/blast/db/. The two cri- exposure of T24 cells to 13.3 μM, 26.6 μM, and teria used to judge whether our identification was 40 μM of cisplatin (12). RT4ρ0, which was deleted successful were as follows: 1) the SEQUEST score in mitochondrial DNA, was established from RT4 was more than 50; and 2) at least two peptide frag- using continuous exposure of ethidium bromide in ments were reliably detected. If the SEQUEST score conditional medium additionally supplemented with was less than 50, then we confirmed that there were 50 μg/mL uridine and 100 μg/mL sodium pyruvate more than two high-quality and reliable MS/MS (1). These cells were maintained in RPMI 1640 me- spectra available. If there were several names for a dium supplemented with 10% fetal bovine serum single protein, then the most popular one was ap- (GIBCO BRL), penicillin (100 U/mL), and strepto- plied for analysis. Theoretical molecular weight was mycin (100 μg/mL) at 37°C in a 5% CO2 humidified calculated from the amino acid sequence. Accession incubator. numbers were obtained from the protein database on the UniProt Web page (http://www.uniprot.org/). Se- Sample preparation. Bladder cancer cell lines ap- quence coverage shows the accordance rate of pep- proximately 20 mg in weight were homogenized us- tides confirmed by MS/MS analyses in amino-acid ing a Teflon glass homogenizer in an extracting sequences of candidate proteins. Peptide-spectrum solution with 30-fold volume. The extraction solu- matches were obtained from the total number of tion contained 7 M urea, 2 M thiourea, 0.1 M DTT, identified peptide sequences for the protein, includ- 2.5% w/v Pharmalyte (pH 3–10), 2% w/v CHAPS, ing those redundantly identified. and Complete Mini EDTA-free protease inhibitor (Roche Diagnostics, Mannheim, Germany), one tab- Select proteins. To select proteins, we first excluded let of which was dissolved in 10 mL of the solution. myosin or cytokeratin, which were broadly investi- After homogenization, each homogenate was sub- gated in the field of protein analysis. Then, proteins jected to rapid agitation in the presence of 0.35-mm with expression that obviously increased or de- Cisplatin-related proteins 255

Fig. 1 Two-dimensional electrophoresis gels (2-DE) shows 30 proteins with differential expression of cisplatin-naïve T24 (a, c) and cisplatin-resistant T24 (b, d). The 2-DE gels show 20 spots in 12% gels (a, b) and 10 spots in a linear gradient of 6% to 10% gels (c, d). Red arrow shows increased cisplatin-resistant T24 and blue arrow shows decreased cisplatin-resistant T24. creased in 2-DE were selected. Finally, we searched ribonucleoprotein A3 (HNRNPA3) (1 : 200; Abcam, PubMed for relevant articles published up until Feb- Cambridge, UK), glutamate oxaloacetate transami- ruary 2015. The following search terms and their nase 2 (GOT2) (1 : 1000; Abnova, Taipei City, Tai- combinations were used: each protein name and wan), phospoglycerate kinase 1 (PGK1) (1 : 200; “bladder cancer”; “urothelial carcinoma”; or “cispla- Abgent, San Diego, CA, USA), transketolase (TKT) tin.” Inclusion criteria for further immunoblotting (1 : 100; Abcam), plasminogen activator inhibitor 2 were as follow: less than 10 articles matched protein (SERPINB2) (1 : 1000; Novusn Biologicals, Littleton, names with “bladder cancer” or “urothelial carcino- CO, USA), eukaryotic translation initiation factor 3 ma” and less than five articles matched protein subunit A (EIF3A) (1 : 5000; Santa Cruz Biotech, names with “cisplatin.” From these searches, we se- Dallas, TX, USA), or periplakin (PPL) (1 : 1000; lected eight proteins for immunoblotting. Novusn Biologicals). Horseradish peroxidase–conju- gated secondary antibodies (Dako, Glostrup, Den- Immunoblotting. Cell line samples were homogenized mark) were diluted 1 : 10,000 (2% normal swine with a Teflon glass homogenizer in sample buffer serum/TBS). For the calibration of each lane, we (50 mM Tris-HCl, pH 6.8, 0.1 M DTT, 10% glycer- obtained rabbit anti-β actin monoclonal antibody ol, 2% SDS, and 0.1% bromophenol blue). Protein (Cell Signaling Technology, Danvers, MA, USA). extracts were separated under denaturing conditions Antigens on the membrane were detected with Im- in 10% to 20% polyacrylamide slab gel (DRC, To- mobilon Western detection reagents (Millipore, Bill- kyo, Japan). The SDS-PAGE gel was run with a erica, MA, USA). constant voltage of 300 V. After completion of electrophoresis, proteins were transferred onto Hybond- RESULTS P PVDF membranes (GE Healthcare) and detected with mouse monoclonal antibodies against phospho- Identification of cisplatin-related protein in cisplatin- enolpyruvate carboxykinase (PCK2) (1 : 100; Ab- naïve and cisplatin-resistant bladder cancer cell lines gent, San Diego, CA, USA), heterogeneous nuclear Protein expressions were quite different for T24 and 256 Y. Taoka et al.

T24CDDPR on 2-DE gels (Fig. 1). We compared DISCUSSION 2-DE gels of both cell lines and selected 20 spots in 12% gels and 10 spots in a linear gradient of 6% to Proteomics research is a new biological modality 10% gels. There were 30 obviously different spots that contributes greatly to our understanding of gene for T24 and T24CDDPR. Eight spots had increased functions in the post-genomic era and that focuses expression and 22 spots had decreased expression in on protein expression and post-translational modifi- T24CDDPR compared to that in T24. Characteris- cation. There is an increasing interest in proteomic tics of these spots are shown in Table 1. Because techniques because the cells could express the func- five of these spots were duplicated, 25 spots were tion of proteins in various ways, but DNA sequence finally selected. information provides a static frame. The detection The 25 proteins, which were differentiated protein of proteins was based on the consideration that cells expressions of T24 and T24CDDPR, were catego- could produce specific proteins to exert specific rized as 10 functions. The categories were as fol- functions. When cells encounter unusual situations, lows: seven cytoskeletal proteins (Spot No. 8(24), they try to adjust by expressing proteins that may 9(25), 10(28), 12, 13, 26, 27); five enzyme modula- help to deal with the new situation. Such proteins, tors (Spot No. 16, 17, 18, 22, 23); four transferases specifically synthesized on demand, may indicate (Spot No. 3, 4, 11(29), 15); two nucleic acid–binding characteristic disease states and thus may serve as proteins (Spot No. 2, 14); two transporter proteins diagnostic markers. (Spot No. 7, 20); one oxidoreductase (Spot No. 1); Identifying anti-cancer drug resistance proteins one protein synthesis (Spot No. 5); one lyase (Spot using proteomic techniques in bladder cancer has No. 6); one coagulation factor (Spot No. 19(30)); been performed in recent reports. Miura et al. (13) and one translation factor (Spot No. 21). Of the 25 identified 36 differentially expressed proteins; 21 proteins, we selected HNRNPA3, PCK2, PPL, PGK1, proteins were upregulated and 15 were downregulat- TKT, SERPINB2, GOT2, and EIF3A for further val- ed in HT1376 cisplatin-resistant bladder cancer cell idation by Western blot. lines exposed to 5 μM cisplatin. They reported that adseverin (SCIN), which is a calcium-dependent Validation of differentiated protein expression actin-binding protein, was upregulated four-fold in To further verify the identified proteins, Western cisplatin-resistant cells compared to cisplatin-naïve blotting analysis was performed to determine the ex- cells. The mitochondrial fraction was more increased istence of these eight proteins in T24 or T24CDDPR. than the cytoskeletal fraction among the differentiated Representative results are shown in Fig. 2. The proteins. Meng et al. (10) identified 30 proteins (19 results showed that HNRNPA3, PGK1, TK, and upregulated and 11 downregulated proteins) in SERPINB2 had more than 1.5-times the incremental pumc91 bladder cancer cell lines exposed to 1.0 mg/L expression in T24CDDPR compared to that in T24. adriamycin. They demonstrated that annexin A2 PCK2 and PPL had less than 20% decreased expres- (ANXA2), a calcium-dependent phospholipid-bind- sions in T24CDDPR compared to that in T24. GOT2 ing protein, and nucleophosmin (NPM1), a non- and EIF3A had comparable expressions in the two ribosomal nucleolar phosphoprotein, were upregulated cell lines. in adriamycin-resistant cells compared to adriamy- Western blot analysis was also performed to de- cin-naïve cells. Transferase and oxidoreductase were termine six protein expression patterns in various prominent fractions among differential proteins. bladder cancer cell lines (Fig. 3). HNRNPA3, TKT, These proteins, described in previous reports (10, SERPINB2, and PPL were expressed in all bladder 13), play important roles in apoptosis and cell pro- cancer cell lines. PCK2 was expressed in two other liferation. Once cancer cells acquire a resistance to cell lines. In terms of PGK1, there was slight ex- anti-cancer drugs because of a changing protein pro- pression or loss of expression in bladder cancer cell file, they show unregulated and aggressive behavior, lines (except for T24). resulting in patients suffering from bulky tumors af- Expression levels of four proteins in T24CDDPR ter local progression and distant metastasis. in three different concentrations of cisplatin were in- In this study, we performed a proteomics analysis vestigated (Fig. 4). HNRNPA3 and PGK1 showed to identify differential protein expressions of cisplat- increased expression irrespective of the exposure of in-naïve and cisplatin-resistant bladder cancer cell cisplatin concentration. PCK2 and PPL demonstrat- lines. We successfully identified 25 proteins with ed gradually decreased expression, which correlated significantly altered expression levels. Based on with the exposure of cisplatin concentration. their functions, these proteins are mainly involved Cisplatin-related proteins 257 Status Increase Increase Increase Increase Increase Increase Increase Increase Decrease Decrease Decrease Decrease Decrease Decrease Decrease Decrease Decrease Decrease Decrease Decrease Decrease Decrease Decrease Decrease Decrease Decrease Decrease Decrease Decrease Decrease 8 6 6 11 11 11 11 11 21 27

23 27 28 24 21 18

19 15 19

21 18 16 16 21 17 27 21 20 16 20 PSMs 5.81 5.81 9.49 (%) 29.74 25.32 58.57 37.87 52.39 29.74 23.23 16.3 44.03 16.3 25.68 31.78 26.82 38.25 12.45 53.65 25.25 42.57 30.65 52.83 41.15 27.42 16.93 32.09 37.38 27.58 35.71 Seq. Cov. No. P05120 P21980 P05783 P08729 P35579 P05787 P05120 P14618 P06733 P17516 P29401 P51991 P00505 P00558 P08727 P21980 P05783 P08729 P05787 P31689 P13639 P34897 O60437 Q13409 Q14152 Q16822 Q9Y427 Q9Y617 Q8NC51 Q9Y6N5 Accession 72.20 60.2 58.3 Score 110.2 110.23 108.2 208.26 266.27 268.26 222.22 272.28 108.2 220.27 202.22 178.23 186.21 142.3 190.25 204.3 168.3 160.2 154.2 206.2 168.3 104.2 268.28 200.3 190.2 182.24 146.2 PPL TKT G.N. PKM EEF2 KRT7 KRT8 KRT7 KRT8 PCK2 PGK1 ENO1 GOT2 TPM1 EIF3A TGM2 TGM2 MYH9 PSAT1 KRT18 KRT19 KRT18 SQRDL SHMT2 SERBP1 AKR1C4 DNAJA1 DYNC1I2 HNRNPA3 SERPINB2 SERPINB2 Protein Name Protein expression profile in cisplatin-resistant T24 compared to that in T24

1 Table Plasminogen activator inhibitor 2 Protein-glutamine gamma-glutamyltransferase 2 Keratin, type I cytoskeletal 18 Myosin-9 Periplakin Keratin, type II cytoskeletal 7 Cytoplasmic dynein 1 intermediate chain 2 Pyruvate kinase PKM Alpha-enolase Keratin, type II cytoskeletal 8 Eukaryotic translation initiation factor 3 subunit A Plasminogen activator inhibitor 2 Aldo-keto reductase family 1 member C4 Aspartate aminotransferase, mitochondrial Phosphoglycerate kinase 1 Transketolase Heterogeneous nuclear ribonucleoprotein A3 Tropomyosin Tropomyosin alpha-1 chain Keratin, type I cytoskeletal 19 Protein-glutamine gamma-glutamyltransferase 2 Keratin, type I cytoskeletal 18 Keratin, type II cytoskeletal 7 Keratin, type II cytoskeletal 8 Phosphoenolpyruvate carboxykinase DnaJ homolog subfamily A member 1 Phosphoserine aminotransferase Elongation factor 2 Plasminogen activator inhibitor 1 RNA-binding protein Sulfide:quinone oxidoreductase Serine hydroxymethyltransferase M.W. M.W. 46,567 77,209 48,029 51,304 71,413 57,901 47,140 30,760 46,567 37,074 47,446 44,587 67,835 39,595 32,658 44,066 77,209 48,029 51,304 30,760 70,592 44,840 40,397 95,278 44,939 49,929 55,955 (Theo.) 226,392 204,526 166,469 M.W. M.W. 42,000 80,000 45,000 49,000 70,000 55,000 45,000 40,000 42,000 40,000 45,000 45,000 70,000 43,000 41,000 42,000 80,000 45,000 49,000 40,000 70,000 45,000 41,000 95,000 45,000 43,000 47,000 (Obs.) 220,000 210,000 150,000 11 30 29 28 26 27 25 20 21 22 23 24 19 18 15 16 17 14 13 12 10 09 08 06 07 04 05 02 01 03 Spot No. Spot No., spot number; M.W. (Obs.), observed molecular weight; M.W. (Theo.), theoretical number; Seq. molecular Cov. (%), sequence weight; coverage; PSMs, G.N., peptide-spectrum matches; Status, gene change in name; the expression Score, levels of SEQUEST the score; protein in Accession cisplatin-resistant T24 No., compared to accession that in T24. 258 Y. Taoka et al.

Fig. 2 Differentially expressed representative proteins, PCK2 (A) and HNRNPA3 (B). (a) Two-dimensional electrophoresis gel. (b) Gel images, enlarged in triplicate, of the area containing protein spots. Red arrow shows increased expression and blue arrow shows decreased expression. (c) Western blot. (d) Protein expression levels calculated from Western blot. Cisplatin-related proteins 259

Fig. 3 (a) Protein expression in various bladder cancer cell lines using Western blot. Protein was applied to each column (20 μg). Lane 1: cisplatin-naïve T24, Lane 2: cisplatin-resistant T24, Lane 3: RT4, Lane 4: RT4ρ0, Lane 5: 5637, Lane 6: EJ, Lane 7: TCCSUP. (b) Table shows protein expression levels. Each protein level is calculated from Western blot and compared to the level of T24. in cytoskeletal proteins and enzyme modulators. action. Individual HNRNPs also function in several Some of these proteins have been reported to be as- other cellular processes, like transcription, DNA re- sociated with chemotherapy resistance in previous pair, telomere biogenesis, and cell signaling (2, 4, 6). proteomic studies (10, 13). Our study also found As a consequence of their multiple roles in the reg- several novel candidate resistance-associated pro- ulation of gene expression, expression levels of teins. Using PubMed, HNRNPA3 and PCK2 showed HNRNP are tightly linked to cell proliferation (4). no match with “bladder cancer” or “urothelial carci- Our findings showed HNRNPA3 overexpression, noma” and only showed one match with “cisplatin”; which may acquire biological aggressiveness after there was one study of HNRNPA3 with cisplatin, exposure to cisplatin in bladder cancer cells. reporting that the expression of HNRNPA3 was up- PCK activity, which catalyzes the conversion of regurated in cisplatin-resistant breast cancer cells oxaloacetate to phosphoenolpyruvate, is distributed (18). These proteins need to validate whether pro- in cytosol and mitochondria as a result of two enzy- teins would contribute to responding to anti-cancer matically indistinct isozymes, PCK1 and PCK2 (9). therapy. The analysis of protein changes after treat- PCK2 is important for maintaining cell progression ment may aid in clinical decisions and may possibly and survival, especially under stress conditions. Our facilitate optimized personalized treatments. results showed downregulation of PCK2 using cispl- More than 20 distinct HNRNPs have been identified atin-resistant bladder cancer cell lines. Park et al. in human cells, designated HNRNPA1 to HNRNPU (15) investigated the association of PCK2 expres- in increasing molecular size from 32 to 110 kDa. sion with chemoradiation response using 5-fluourac- HNRNPs are RNA-binding proteins (RBP) that are yl (5-FU). They reported that downregulation of essential players in mRNA metabolism, acting as PCK2 may be linked to induced 5-FU resistance, coordinators of post-transcriptional events by partic- but not to intrinsic 5-FU susceptibility, in colon can- ipating in an extensive network of RNA–RBP inter- cer cells. Because low energy metabolism and slow 260 Y. Taoka et al.

Fig. 4 (a) Protein expression in various resistance levels of bladder cancer cell lines using Western blot. Protein was applied (20 μg) in each column. Lane 1: cisplatin-naïve T24, Lane 2: cisplatin-resistant T24 using 13.3 μM of cisplatin, Lane 3: cisplatin-resistant T24 using 26.6 μM of cisplatin, Lane 4: cisplatin-resistant T24 using 40.0 μM of cisplatin. (b) Table shows protein expression levels. Each protein level is calculated from Western blot and compared to the level of T24. proliferation can decrease the susceptibility to che- all the detected proteins would be related to cisplat- motherapy (17, 22), PCK2 downregulation could in resistance. We observed that cancer cells chemi- lead to slow energy metabolism and may subse- cally changed the protein profile, and we think of quently reduce the susceptibility to anti-cancer drugs. this as “cisplatin resistance.” In addition, our results Cisplatin resistance is most often multifactor, mul- demonstrated that a variety of protein characteristics tistep, and multipathway (3). In our study, we found contributed to resistance, including high molecular proteins possibly correlated with cisplatin resistance weight and complex protein profile (10, 13). Finally, and showed that each protein demonstrated each be- the observed proteins were not noted to be fully havior in various concentrations of cisplatin. For ex- shared by other cancer models because of the het- ample, periplakin, which is involved in cellular erogeneity of cancerous cells. Further studies are movement and attachment (7), showed gradually de- warranted to determine the potential application of creased expression correlated with the concentration altered proteins that are possibly connected to cispl- of cisplatin and, on the contrary, HNRNPA3 showed atin resistance. incremental expressions. This finding suggests that each protein would be preferentially expressed at Acknowledgment the time of exposure level of cisplatin, and that it might be difficult to completely inhibit a single This study was supported in part by a Grant-in-Aid mechanism regulating cisplatin resistance to restore for Scientific Research C (15K10607) from The Ja- the susceptibility. pan Society for the Promotion of Science (to K. The results could lead to the development of a Matsumoto). new molecular marker. This would be an important step in establishing a treatment strategy for patients with advanced cancer. However, it is unlikely that Cisplatin-related proteins 261

REFERENCES acterization of acquired resistance to platinum anticancer drugs in human ovarian carcinoma cells. Jpn J Cancer Res 1. Cook CC, Kim A, Terao S, Gotoh A and Higuchi M (2012) 86, 88–94. Consumption of oxygen: a mitochondrial-generated progres- 13. Miura N, Takemori N, Kikugawa T, Tanji N, Higashiyama S sion signal of advanced cancer. Cell Death Dis 3, e258. and Yokoyama M (2012) Adseverin: a novel cisplatin-resis- 2. Dreyfuss G, Kim VN and Kataoka N (2002) Messenger- tant marker in the human bladder cancer cell line HT1376 RNA-binding proteins and the messages they carry. Nat Rev identified by quantitative proteomic analysis. Mol Oncol 6, Mol Cell Biol 3, 195–205. 311–322. 3. Galluzzi L, Senovilla L, Vitale I, Michels J, Martins I, Kepp O, 14. Okusa H, Kodera Y, Oh-ishi M, Minamida S, Tsuchida M, Castedo M and Kroemer G (2012) Molecular mechanisms of Kavvoussi PK, Matsumoto K, Satoh T, Iwamura M, Maeda cisplatin resistance. Oncogene 31, 1869–1883. T and Baba S (2008) Searching for new biomarkers of blad- 4. He Y, Brown MA, Rothnagel JA, Saunders NA and Smith R der cancer based on proteomics analysis. J Electrophoresis (2005) Roles of heterogeneous nuclear ribonucleoproteins A 52, 19–24. and B in cell proliferation. J Cell Sci 118, 3173–3183. 15. Park JW, Kim SC, Kim WK, Hong JP, Kim KH, Yeo HY, 5. Ikeda M, Matsumoto K, Tabata K, Minamida S, Fujita T, Lee JY, Kim MS, Kim JH, Yang SY, Kim DY, Oh JH, Cho Satoh T, Iwamura M and Baba S (2011) Combination of JY and Yoo BC (2014) Expression of phosphoenolpyruvate gemcitabine and paclitaxel is a favorable option for patients carboxykinase linked to chemoradiation susceptibility of hu- with advanced or metastatic urothelial carcinoma previously man colon cancer cells. BMC Cancer 14, 160. treated with cisplatin-based chemotherapy. Jpn J Clin Oncol 16. Saxman SB, Propert KJ, Einhorn LH, Crawford ED, Tannock 41, 1214–1220. I, Raghavan D, Loehrer PJ, Sr. and Trump D (1997) Long- 6. Krecic AM and Swanson MS (1999) hnRNP complexes: term follow-up of a phase III intergroup study of cisplatin composition, structure, and function. Curr Opin Cell Biol 11, alone or in combination with methotrexate, vinblastine, and 363–371. doxorubicin in patients with metastatic urothelial carcinoma: 7. Matsumoto K, Ikeda M, Sato Y, Kuruma H, Kamata Y, a cooperative group study. J Clin Oncol 15, 2564–2569. Nishimori T, Tomonaga T, Nomura F, Egawa S and Iwamura 17. Shin YK, Yoo BC, Chang HJ, Jeon E, Hong SH, Jung MS, M (2014) Loss of periplakin expression is associated with Lim SJ and Park JG (2005) Down-regulation of mitochondri- pathological stage and cancer-specific survival in patients al F1F0-ATP synthase in human colon cancer cells with in- with urothelial carcinoma of the urinary bladder. Biomed Res duced 5-fluorouracil resistance. Cancer Res 65, 3162–3170. (Tokyo) 35, 201–206. 18. Smith L, Welham KJ, Watson MB, Drew PJ, Lind MJ and 8. Matsumoto K, Mochizuki K, Hirayama T, Ikeda M, Nishi M, Cawkwell L (2007) The proteomic analysis of cisplatin resis- Tabata K, Okazaki M, Fujita T, Taoka Y and Iwamura M tance in breast cancer cells. Oncol Res 16, 497–506. (2015) Gemcitabine plus nedaplatin as salvage therapy is a 19. Tabata K, Matsumoto K, Minami S, Ishii D, Nishi M, Fujita T, favorable option for patients with progressive metastatic uro- Saegusa M, Sato Y and Iwamura M (2014) Nestin is an in- thelial carcinoma after two lines of chemotherapy. Asian Pac dependent predictor of cancer-specific survival after radical J Cancer Prev 16, 2483–2487. cystectomy in patients with urothelial carcinoma of the blad- 9. Mendez-Lucas A, Hyrossova P, Novellasdemunt L, Vinals F der. PLoS One 9, e91548. and Perales JC (2014) Mitochondrial phosphoenolpyruvate 20. Tsumura H, Matsumoto K, Ikeda M, Yanagita K, Hirano S, carboxykinase (PEPCK-M) is a pro-survival, endoplasmic Hagiwara M, Nagashio R, Fujita T, Sato Y and Iwamura M reticulum (ER) stress response gene involved in tumor cell (2015) High expression level of preoperative serum uropla- adaptation to nutrient availability. J Biol Chem 289, 22090– kin III is associated with biologically aggressive bladder can- 22102. cer. Asian Pac J Cancer Prev 16, 1539–1543. 10. Meng Q, Lei T, Zhang M, Zhao J, Zhao XH and Zhang M 21. von der Maase H, Hansen SW, Roberts JT, Dogliotti L, (2013) Identification of proteins differentially expressed in Oliver T, Moore MJ, Bodrogi I, Albers P, Knuth A, Lippert adriamycin-resistant (pumc-91/ADM) and parental (pumc-91) CM, Kerbrat P, Sanchez Rovira P, Wersall P, Cleall SP, human bladder cancer cell lines by proteome analysis. J Can- Roychowdhury DF, Tomlin I, Visseren-Grul CM and Conte PF cer Res Clin Oncol 139, 509–519. (2000) Gemcitabine and cisplatin versus methotrexate, vin- 11. Minami S, Nagashio R, Ueda J, Matsumoto K, Goshima N, blastine, doxorubicin, and cisplatin in advanced or metastatic Hattori M, Hachimura K, Iwamura M and Sato Y (2014) De- bladder cancer: results of a large, randomized, multinational, tection of tumor-associated antigens in culture supernatants multicenter, phase III study. J Clin Oncol 18, 3068–3077. using autoantibodies in sera from patients with bladder can- 22. Yoo BC, Ku JL, Hong SH, Shin YK, Park SY, Kim HK and cer. Biomed Res (Tokyo) 35, 25–35. Park JG (2004) Decreased pyruvate kinase M2 activity linked 12. Misawa T, Kikkawa F, Maeda O, Obata NH, Higashide K, to cisplatin resistance in human gastric carcinoma cell lines. Suganuma N and Tomoda Y (1995) Establishment and char- Int J Cancer 108, 532–539.