Overexpression of Peptidyl-Prolyl Isomerase-Like 1Is Associated With

Human Cancer Biology

Overexpression of Peptidyl-Prolyl Isomerase-Like 1 Is Associated with the Growth of Colon Cancer Cells Kazutaka Obama,1Tatsushi Kato,3 Suguru Hasegawa,3 Seiji Satoh,3 Yusuke Nakamura,1and Yoichi Furukawa2

Abstract Purpose and Experimental Design: To discover novel therapeutic targets for colon cancers, we previously surveyed expression patterns among 23,000 genes in colon cancer tissues using a cDNA microarray. Among the genes that were up-regulated in the tumors, we selected for this study peptidyl-prolyl isomerase-like1 (PPIL1) encoding PPIL1, a cyclophilin-related protein. Results: Western blot analysis and immunohistochemical staining using PPIL1-specific antibody showed that PPIL1protein was frequently overexpressed in colon cancer cells compared with noncancerous epithelial cells of the colon mucosa. Colony formation assay showed a growth- promoting effect of wild-type PPIL1 on NIH3T3 and HEK293 cells. Consistently, transfection of short-interfering RNA specific to PPIL1 into SNUC4 and SNUC5 cells effectively reduced expression of the gene and retarded growth of the colon cancer cells.We further identified two PPIL1- interacting proteins, SNW1/SKIP (SKI-binding protein) and stathmin. SNW1/SKIP is involved in the regulation of transcription and mRNA splicing, whereas stathmin is involved in stabilization of microtubules. Therefore, elevated expression of PPIL1may play an important role in proliferation of cancer cells through the control of SNW1/SKIPand/or stathmin. Conclusion: The findings reported here may offer new insight into colonic carcinogenesis and contribute to the development of new molecular strategies for treatment of human colorectal tumors.

Colon cancer is one of the most common causes of cancer Peptidyl-prolyl isomerase-like 1 (PPIL1) was first identified as a death throughout the world. Molecular investigations have cyclophilin-related gene encoding a protein that shares 46.0% provided evidence that multiple genetic alterations are involved identity in amino acid sequence with human cyclophilin A (8). in colorectal tumorigenesis. Early in those studies, loss of It is expressed in heart, skeletal muscle, and liver (8). However, heterozygosity was observed frequently at loci on chromo- its biological function remains unclear. Cyclophilins and somes 1p, 2p, 3p, 5q, 8p, 11q, 17p, 18q, and 22q (1, 2). Later, FK506-binding proteins are two families of peptidyl-prolyl alterations in genes encoding K-ras, adenomatous polyposis isomerases that are extensively characterized. Cyclophilins form coli protein (APC), h-catenin (CTNNB1), p53 (TP53), and/or complexes with cyclosporin A, whereas FK506-binding proteins conductin (AXIN2) were shown to be involved in colonic bind the immunosuppressant drug FK506. A third family of carcinogenesis (3–7). In spite of intensive study at the peptidyl-prolyl isomerases includes parvulin and Pin1. These molecular level, however, the roles of a large number of genes proteins do not bind immunosuppressants but are involved that show altered expression in colorectal cancers remain in cell cycle progression and cell survival (9, 10). Pin1 is a unclear. phosphorylation-dependent peptidyl-prolyl isomerase that catalyzes the cis-trans isomerization of phosphoserine-proline and phosphothreonine-proline bonds in multiple proteins such as CDC25, RNA polymerase II, cyclin D1, p53, and Tau Authors’Affiliations: 1Laboratory of Molecular Medicine; 2Promotion of Genome- (11, 12). It interacts with phosphorylated-p53 when the latter Based Medicine Project, Human Genome Center, Institute of Medical Science,The is activated in response to DNA damage, altering the stability of University ofTokyo,Tokyo, Japan; and 3Department of Gastroenterological Surgery, Graduate School of Medicine, Kyoto University, Kyoto, Japan p53 through isomerization of a motif involving a serine/ Received 3/16/05; revised 10/7/05; accepted 10/10/05. threonine residue followed by a proline (Ser/Thr-Pro; Grant support: Research for the Future Program grant 00L01402 from the Japan refs. 10, 13). All of these peptidyl-prolyl isomerase families Society for the Promotion of Science. carry out cis-trans isomerization of the peptide bond on the The costs of publication of this article were defrayed in part by the payment of page NH -terminal side of Pro residues, forming a tight bend in the charges. This article must therefore be hereby marked advertisement in accordance 2 with 18 U.S.C. Section 1734 solely to indicate this fact. polypeptide backbone that substantially changes the confor- Note: K. Obama and T. Kato contributed equally to this work. mation of the protein. Through this isomerization and Requests for reprints: Yoichi Furukawa, Promotion of Genome-Based Medicine consequent conformational change, peptidyl-prolyl isomerases Project, Human Genome Center, Institute of Medical Science, The University of alter the functions of their target proteins (14). Tokyo, 4-6-1 Shirokanedai, Minato-ku, 108-8639 Tokyo, Japan. Phone: 81-3- 5449-5373; Fax: 81-3-5449-5124; E-mail: [email protected]. To discover target molecules for development of novel F 2006 American Association for Cancer Research. diagnostic strategies and/or therapeutic drugs, we previously doi: 10.1158/1078-0432.CCR-05-0588 carried out a genome-wide analysis of gene expression in

Clin Cancer Res 2006;12(1) January 1, 2006 70 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on September 30, 2021. © 2006 American Association for Cancer Research. PPIL1 as a Novel MolecularTarget for Colon Cancer human colon cancer tissues by means of cDNA microarray siB; and 5V-TCCCCCAGCTTCTAGATGACATATTCAAGAGATATGT- consisting of 23,040 genes (15). In this report, we show that CATCTAGAAGCTGG-3V and 5V-AAAACCAGCTTCTAGATGACA- PPIL1, a gene that is frequently up-regulated in colon cancers, TATCTCTTGAATATGTCATCTAGAAGCTGG-3V for siC. psiH1BX-PPIL1 promotes growth of cancer cells. In addition, we document or psiH1BX-EGFP (17) plasmids were transfected into SNUC4 and SNUC5 cells, which constitutionally express abundant PPIL1. Silencing interactions between PPIL1 and two known proteins, SNW1/ of PPIL1 by the plasmids was examined by semiquantitative reverse SKIP and stathmin, in vitro and in vivo. These data may not only transcription-PCR 48 hours after transfection. Cells transfected with lead to a more profound understanding of colorectal carcino- the plasmids were cultured in the presence of appropriate concentra- genesis but also provide clues for the development of novel tion of geneticin (G418) for 10 days, when viability of transfected cells therapeutic strategies. was examined by cell counting kit according to the protocol of the supplier (Dojindo Laboratories, Kumamoto, Japan). We additionally

analyzed sub-G1 and G0-G1,S,andG2-M populations in cells Materials and Methods transfected with psiH1BX-PPIL1 or psiH1BX-EGFP using flow cytom- eter according to the recommendations of the manufacturer (Becton Cell lines and tissue specimens. Human colon cancer cell lines Dickinson, San Jose, CA). SNUC4 and SNUC5 were obtained from the Korean cell line bank. Bacterial two-hybrid experiment. A bacterial two-hybrid system HEK293 (human embryonic kidney), COS-7 (monkey kidney), and (Stratagene) was done with the BacterioMatch Two-Hybrid System NIH3T3 (mouse fibroblasts) cells were purchased from the American according to the protocols of the manufacturer. We cloned the entire Type Culture Collection (Rockville, MD). All tumor tissues and coding sequence of PPIL1 into the BamHl-Xhol site of pBT vector as a corresponding nontumorous mucosae were excised with informed bait and screened a human fetal-brain cDNA library (Stratagene). consent from 79 patients who underwent colectomy or endoscopic Immunoprecipitation assay. The entire coding region of the polypectomy at the Department of Surgery, Tokyo Hitachi Hospital. human stathmin gene (STMN1) and SNW1/SKIP was amplified by Real-time PCR (TaqMan) assay. Extraction of RNA, preparation of reverse transcription-PCR using primers as follows: 5V-ATTAGAT- cDNA, and real-time PCR were done as described previously (16). The CTTCACCATGGCTTCTGATATCC-3V (forward) and 5V-AATGGT- h-actin gene (ACTB) served as a quantitative control. The sequences ACCTTAGTCAGCTTCAGTCTCGTCAGC-3V(reverse) for stathmin, and of primers and probes are as follows: PPIL1 forward primer, 5V-TGGGAATTCCGGAAGAAGATGGCGCTCACCAGC-3V (forward) and 5V-TGACCCAACAGGGACAGGTC-3V and reverse, 5V-GTTCATCTT- 5V-GTGCCTCGAGCTTCCTCCTCTTCTTGCCTTCATGC-3V (reverse) for CAAACTGTTTGCCAT-3V; probe, 5V-FAM-AGGTGGTGCATCTAT-MGB- SNW1/SKIP. The PCR products were cloned into pCMV-HA or 3V; ACTB forward primer, 5V-GGCACCCAGCACAATGAAG-3V and pcDNA3.1/Myc-His (Invitrogen), respectively. COS-7 cells transfected reverse, 5V-ACACGGAGTACTTGCGCTCA-3V; probe, 5V-FAM-TCAAGAT- with pFLAG-PPIL1 (expressing FLAG-tagged PPIL1), pCMVHA-STMN1 CATTGCTCCTC-MGB-3V. (expressing HA-tagged stathmin), pcDNA-SNW1/SKIP-myc (expressing Immunoblot analysis and immunohistochemical staining using poly- myc-tagged SNW1/SKIP), or a combination of the plasmids were clonal antibody against PPIL1. The polyclonal antibody to PPIL1 was washed with PBS and lysed in TNE buffer. Immunoprecipitation and purified from sera of immunized rabbit with recombinant glutathione immunoblotting were carried out using anti-HA (3F10; Roche, S-transferase–PPIL1 protein prepared in Escherichia coli. Immunohis- Mannheim, Germany), anti-myc (9E10: Santa Cruz Biotechnologies, tochemical staining was carried out using affinity-purified antibody Santa Cruz, CA), anti-FLAG (Sigma), horseradish peroxidase– against human PPIL1. Sections of paraffin-embedded tissues were conjugated goat anti-rat IgG (Santa Cruz Biotechnologies), or horse- subjected to the SAB-PO peroxidase immunostaining system (Nichirei, radish peroxidase–conjugated sheep anti-mouse IgG (Amersham Tokyo, Japan) as described earlier (17). We analyzed a total of 47 Bioscience, Buckinghamshire, United Kingdom) antibodies as de- clinical samples, including 29 cancers and 18 adenomas. The reactivity scribed earlier (18). To prepare plasmids expressing S16A, S25A, was assessed semiquantitatively using the following grading system: À, S38A, or S63A forms of mutant stathmin, we used a QuikChange Site- no staining; +, slight staining; 3+, strong staining; and 2+, staining Directed Mutagenesis kit (Stratagene) and amplified each form using between + and 3+. pCMVHA-STMN1 as a template and different sets of primers according Effect of PPIL1 on cell growth in vitro. The entire coding region of to the recommendations of the manufacturer. human PPIL1 was amplified by reverse transcription-PCR using Statistical analysis. Statistical significance was determined by primers 5V-AGACAAGCTTTCCGCCGCCGGC-3V (forward) and Sheffe’s F test, using a commercially available software, Statview 5.0 5V-GTCTCTCGAGAAGGGTATGCCTTAATGATCTTC-3V (reverse), and (SAS Institute, Cary, NC). A difference of P < 0.05 was considered cloned into expression vectors pcDNA3.1/Myc-His (Invitrogen, statistically significant. Carlsbad, CA) or pFLAG-CMV-5C (Sigma, St. Louis, MO). Using a QuikChange Site-Directed Mutagenesis kit (Stratagene, La Jolla, CA), we additionally prepared plasmids expressing mutant PPIL1 (pcDNA- Results mtPPIL1-myc) that lacked four amino acid residues between codons 63 and 66, a conserved region in homologues of PPIL1. pcDNA- Elevated PPIL1 expression in colorectal cancer cells. Among PPIL1-myc, pcDNA-mtPPIL1-myc, or the control vector (pcDNA- the genes that were up-regulated in the colon cancer tissues LacZ-myc) were transfected into NIH3T3 and HEK293 cells. After investigated on our 23,040 gene microarray, expression of 2 weeks of incubation with an appropriate concentration of geneticin PPIL1 gene was elevated in all informative cases (data not (Invitrogen), cells were fixed with 100% methanol and stained shown). Subsequent real-time PCR (TaqMan) assay showed with Giemsa solution to count the number of colonies. that its expression was enhanced >2-fold in 10 of 14 Gene-silencing effect of PPIL1 small interfering RNAs. Plasmids additional colon cancer tissues compared with the corres- expressing PPIL1 small interfering RNAs (siRNA) together with the ponding noncancerous mucosa (Fig. 1A). To examine neomycin-resistant gene were prepared by cloning the following double- expression of PPIL1 protein in colon cancer tissues, we did stranded oligonucleotides into psiH1BX3.0 as described previously (17): 5V-TCCCGACCTGTAAGAACTTTGCTTTCAAGAGAAGCAAAGTTC- immunoblot analysis with anti-PPIL1 polyclonal antibody TTACAGGTC-3V and 5V-AAAAGACCTGTAAGAACTTTGCTTCTCTTG using an additional five pairs of colon cancer tissues and the AAAGCAAAGTTCTTACAGGTC-3V for siA; 5V-TCCCGATGAACTTCATC- corresponding noncancerous mucosa. In agreement with the CAGACTTTCAAGAGAAGTCTGGATGAAGTTCATC-3Vand 5V-AAAAGAT- elevated PPIL1 transcript, PPIL1 protein was augmented in GAACTTCATCCAGACTTCTCTTGAAAGTCTGGATGAAGTTCATC-3V for all five cancer tissues compared with the noncancerous

www.aacrjournals.org 71 Clin Cancer Res 2006;12(1) January 1, 2006 Downloaded from clincancerres.aacrjournals.org on September 30, 2021. © 2006 American Association for Cancer Research. Human Cancer Biology

of PPIL1 expression in noncancerous tissues of the colon. Notably, cancerous cells at an early stage of tumor localized within mucosal and submucosal layer were also positively stained. On the other hand, no epithelial cells in normal crypts were stained with the antibody (Fig. 2A and B, arrowheads). We further examined its expression in 18 adenomas of colon, which were all negative for PPIL1 (Fig. 2D). These data corroborated elevated PPIL1 expression in colon cancer cells and suggest that PPIL1 may be involved in malignant transformation of colon tumor. Effect of exogenous PPIL1 on growth of mammalian cells. To disclose a potential biological function of PPIL1 in colorectal tumorigenesis, we constructed plasmids expressing wild-type PPIL1 (pcDNA-PPIL1-myc) and mutant PPIL1 (pcDNA-mtPPIL1-myc) and carried out colony formation assays with NIH3T3 and HEK293 cells. Transfection of pcDNA-PPIL1-myc produced a significantly larger number of colonies than did the negative control (pcDNA-LacZ-myc) or mutant plasmids (pcDNA-mtPPIL1-myc; data not shown for HEK293; Fig. 3A and B). This result was confirmed by Fig. 1. Elevated PPIL1expression in colorectal cancer. A, relative expression of three independent experiments, suggesting that elevated ex- PPIL1 in 14 colon cancer tissues analyzed by real-time PCR. Logarithmic ratios of cancer to their corresponding noncancerous mucosae are shown. B, Western blot pression of PPIL1 might confer growth-promoting effects on analysis of PPIL1using an additional five pairs of cancer tissues (T)andthe cancer cells. corresponding normal mucosae (N). Expression of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) served for control. Effect of PPIL1 siRNAs on growth of colon cancer cells. To investigate the role of PPIL1 in colorectal tumorigenesis, we constructed plasmids expressing siRNAs specific to PPIL1 mucosa(Fig.1B).Immunohistochemicalstainingwiththe (psiH1BX-PPIL1), and examined their effects on expression of anti-PPIL1 antibody revealed that nuclei and/or cytoplasms the gene in colon cancer cells and on subsequent growth of of tumor cells were more strongly stained in 26 of 29 colon those cells. Among the three plasmids expressing PPIL1-siRNAs, cancer tissues than their corresponding noncancerous epi- both psiH1BX-siB and psiH1BX-siC markedly suppressed thelial cells (Fig. 2A-C). In addition, lamina propria cells constitutional expression of PPIL1 protein in SNUC5 as well showed weak staining, which might correspond to low levels as in SNUC4 colon cancer cells at 48 hours after transfection,

Fig. 2. Immunohistochemical staining of PPIL1in colon tumors. Representative pictures of colon cancer tissue (A-C)and adenoma of the colon (D). Noncancerous mucosa was not stained with anti-PPIL1 antibody in the same section (A and B, arrowheads). Original magnification, Â100 (A, B,andD)andÂ400 (C). Bars, 100 Am(A, B,andD)and20Am(C).

Clin Cancer Res 2006;12(1) January 1, 2006 72 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on September 30, 2021. © 2006 American Association for Cancer Research. PPIL1 as a Novel MolecularTarget for Colon Cancer whereas mild effect was observed with constructs psiH1BX-siA pcDNA-SNW1/SKIP-myc plasmids that express myc-tagged (Fig. 3C). In accord with this suppression of PPIL1, transfection SNW1/SKIP protein. Immunoprecipitation with anti-FLAG of SNUC5 or SNUC4 cells with psiH1BX-siB or psiH1BX-siC antibody using extract from cotransfected cells revealed coimmu- suppressed their growth at day 10 of transfection; although the noprecipitated myc-tagged SNW1/SKIP protein (Fig. 4A, top, number of colonies was unchanged, the number of viable cells lane 8). On the other hand, immunoprecipitation with anti- was significantly fewer compared with the control, psiH1BX- myc antibody using the extract detected coimmunoprecipitated EGFP (Fig. 3D). In addition, mild growth suppressive effect was FLAG-tagged PPIL1 protein (Fig. 4A, bottom, lane 8). These data detected with psiH1BX-siA, indicating a positive correlation suggested that PPIL1 associates with human SNW1/SKIP. between gene knockdown effect and growth suppressive effect Interaction of PPIL1 with stathmin in vitro and in vivo. To of the siRNA (Fig. 3D). We further analyzed sub-G1,G0-G1,S, explore further the function of PPIL1, we searched for and G2-M population in cell cycle by flow cytometry at day 10. additional PPIL1-interacting proteins using a bacterial two- 6 Although sub-G1 population was unchanged, the population of hybrid screening system. We screened 1.2 Â 10 clones from S phase was significantly decreased in cells transfected with a fetal brain library transfected with pBT-PPIL1 as bait, and psiH1BX-siB (7.3%) compared with those with psiH1BX-EGFP identified 98 positive colonies. Among the 98 positive clones, (13.9%; Fig. 3E). one containing stathmin produced more colonies by simulta- Interaction of PPIL1 with SNW1/SKIP. Because Dictyostelium neous transformation with pBT-PPIL1 than did cells without discoideum cyclophilin Cyp2 or Schizosaccharomyces pombe transfected pBT-PPIL1 (data not shown). We transfected cyclophilin CypE binds to SnwA or Snw1p, respectively, we COS-7 cells with plasmids expressing HA-tagged stathmin examined whether PPIL1 may associate in vivo with SNW1/SKIP, (pCMVHA-STMN1) and carried out Western blotting with cell the human counterpart of SnwA and Snw1p. We transfected extracts and anti-HA antibody. This experiment produced COS7 cells with pFLAG-PPIL1, in combination with or without two bands corresponding to an 18 kDa protein of the expected

Fig. 3. Effect of PPIL1 on cell growth. A, growth-promoting effect of PPIL1was evaluated by colony formation assays using NIH3T3 cells. B, number of colonies of NIH3T3 cells in triplicate. Bars, SD. *, P < 0.001. C, effect of PPIL1-siRNAs on expression of PPIL1protein in human SNUC5 and SNUC4 colon cancer cells, as evaluated byWestern blot analysis of PPIL1in cells transfected with plasmids expressing PPIL1-siRNAs or a control siRNA (EGFP). D, growth-suppressive effect of PPIL1-siRNA on SNUC5 and SNUC4 colon cancer cells analyzed by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay in triplicate. Bars, SD. E, cell cycle analysis of SNUC4 cells treated with PPIL1-siRNA. Hatched boxes, G0-G1phase; closed boxes, S phase; shadowed boxes, G2-M phase. Bars, SD. *, P < 0.001.

www.aacrjournals.org 73 Clin Cancer Res 2006;12(1) January 1, 2006 Downloaded from clincancerres.aacrjournals.org on September 30, 2021. © 2006 American Association for Cancer Research. Human Cancer Biology

Fig. 4. Interaction between PPIL1 and SNW1/SKIP or stathmin in vivo. A, interaction between PPIL1 and SNW1/SKIP in vivo. COS7 cells were transfected with pFLAG-PPIL1, pcDNA-SNW1/SKIP-myc, or their combination. Immunoprecipitation was done with anti-FLAG antibody (top) or anti-myc (bottom). Subsequently, Western blot analyses with anti-myc (top) or anti-FLAG antibody (bottom) were done using whole cell lysates (lanes 1-4) and immunoprecipitants (lanes 5-8). B, modification of wild-type and mutant forms of stathmin. Extracts of COS-7cells that were transfected with pCMVHA-STMN1 or pCMVHA-mutant-STMN1 (S16A, S25A, S38A, and S63A) were examined by Western blot analysis using anti-HA antibody. C, COS-7 cells were transfected with pFLAG-PPIL1,pCMVHA-STMN1,or pCMVHA-S38A-STMN1, or their combination. Immunoprecipitation was done with anti-FLAG or anti-HA antibodies. Immunoprecipitation (IP) and immunoblot (IB) analyses were done with antibodies indicated in the left.

size as well as a 20 kDa molecule, suggesting the existence of showed that PPIL1 was also capable of interacting with the modified form(s) of stathmin (Fig. 4B, lane 1). In line with unphosphorylated form of stathmin (Fig. 4C, lane 5). PPIL1 this notion, we observed 18 kDa forms of stathmin alone, also associated with S16A, S25A, and S63A mutants (data not when the extract was treated with E phosphatase (data not shown). shown). As stathmin seems to have four putative Ser/Thr 16 25 38 63 phosphorylation sites (Ser , Ser , Ser , and Ser ; ref. 19), Discussion we prepared four plasmids, each expressing a S16A, S25A, S38A, or S63A form of stathmin. When each construct In this study, we revealed that PPIL1 interacts with stathmin was transfected into COS-7 cells, the S16A, S25A, or S63A and SNW1/SKIP. Because stathmin is a cytoplasmic protein and mutant expressed both (18 and 20 kDa) forms of stathmin SNW1/SKIP localizes in the nucleus, it is unlikely that these (Fig. 4B, lanes 2, 3, and 5). However, the S38A mutant three proteins form a complex. PPIL1 may have several target expressed the 18 kDa form alone (Fig. 4B, lane 4), suggesting proteins in the cytoplasm as well as in the nucleus. In line with that the heavier protein (20 kDa) was generated as a result of this notion, its subcellular localization was observed in both phosphorylation at the Ser38 site. intracellular compartments. Among other prolyl isomerase To confirm an association between PPIL1 and stathmin in proteins, Pin1, CypA, Cyp40, FKBP25, and FKBP52 have been normal mammalian cells, we transfected COS-7 cells with the shown to localize in the cytoplasm and nucleus (9). These plasmid expressing FLAG-tagged PPIL1 (pFLAG-PPIL1), with or proteins have several target molecules; Pin1 interacts with Tau in without inclusion of pCMVHA-STMN1, and carried out the cytoplasm and RNA polymerase II in the nucleus, whereas immunoprecipitation assays. Western blotting of immunopre- CypA associates with calcineurin in the cytoplasm and apoptosis cipitants with anti-HA antibody after hybridization with anti- inducing factor in the nucleus (11, 20, 21). Because PPIL1, Pin1, FLAG antibody corroborated in vivo binding between PPIL1 and CypA do not contain a defined nuclear localization signal, and stathmin (Fig. 4C, lane 3). On Western blots, the band the target proteins may regulate their subcellular localization. corresponding to the 20 kDa form of stathmin was more These data also indicate that prolyl isomerases should regulate intense than the band corresponding to the 18 kDa form. multiple proteins, which is consistent with the fact that Pin1 Because the amount of both forms were nearly equal in cells regulates a number of proteins, including CDC25, RNA transfected with pCMVHA-STMN1 (Fig. 4C, lane 2), PPIL1 may polymerase II, p53, Tau, and so on (11). have a higher affinity to the 20 kDa protein, a phosphorylated Stathmin, also known as oncoprotein 18 or leukemia- form of stathmin. Additional immunoprecipitation assay with associated phosphoprotein 18, was initially identified as a anti-HA antibody using extract from cells cotransfected with molecule that becomes phosphorylated in response to various pFLAG-PPIL1 and S38A (pCMVHA-S38A-STMN1) stathmin extracellular signals. Stathmin controls the dynamic changes

Clin Cancer Res 2006;12(1) January 1, 2006 74 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on September 30, 2021. © 2006 American Association for Cancer Research. PPIL1 as a Novel MolecularTarget for Colon Cancer of microtubules by its phosphorylation. It destabilizes micro- kinases, such as CDK1, CDK2, mitogen-activated protein tubules by stimulating microtubule catastrophes and seques- kinase, and p38y (19), in combination with elevation of PPIL1 tering tubulin dimers (19). Therefore, elevated expression of expression, stimulates progression of the cell cycle and remo- PPIL1 may play a role in remodeling of microtubules in deling of the cytoskeleton in cancer cells. proliferating and invasive cancer cells. Notably, stathmin We also proved that PPIL1 interacts with human SNW1/ expression is often elevated in leukemias and in carcinomas SKIP, a transcriptional coactivator. In agreement with their of breast, ovary, and prostate (19). Stathmin contains four Ser interaction, both PPIL1 and SNW1/SKIP were included in the phosphorylation sites (Ser16, Ser25, Ser38, and Ser63; ref. 19). spliceosome complex (25). Because SNW1/SKIP protein plays We detected two forms of stathmin with different migration an important role in transcription regulation, pre-mRNA on SDS-PAGE; one was a 20 kDa phosphorylated form and splicing and cell cycle regulation (26), PPIL1 may modulate the other an 18 kDa unphosphorylated form of Ser38. genes associated with proliferation or cell cycle progression Interestingly, immunoprecipitation assay revealed that most through the regulation of SNW1/SKIP. Of note, SNW1/SKIP of PPIL1 associated with the phosphorylated form of synergized with Ski and released cells from G1 arrest induced stathmin, which is reminiscent of a phosphorylation-depen- by Rb (27). Therefore, the PPIL1-SNW1/SKIP interaction may dent association between Pin1 and p53 (10, 13). Pin1 enhance the oncogenic activity of SKI and promote cell cycle associates with multiple Ser/Thr-Pro of target proteins, leading progression. Although SNW1/SKIP is reported to be a to their conformational change, where a WW domain in Pin1 phosphoprotein (28), its phosphorylation site(s) and mecha- is responsible for the recognition. Among the four Ser residues nism(s) of modification remain to be clarified. Interestingly, in stathmin, Ser25 and Ser38 are followed by Pro. Therefore, human SNW1/SKIP contains two prolines preceded by Ser we investigated association of PPIL1 with mutant forms of (Ser-Pro) at codon 225 and 233 within the SKIP domain. In stathmin containing S25A, S38A, or their combination, which addition, these residues are conserved from lower eukaryotes showed an interaction of PPIL1 with all mutant proteins implying the importance of these residues in their function. (Fig. 4C; data not shown). Therefore, phosphorylation in Because the residues are within the responsible region for the other residues may play a role in the association. Alternatively, binding with Ski (29), Smad2/3 (30), Rb (27), or HPV-E7 (31), PPIL1 associates with both unphosphorylated and phosphor- phosphorylation and subsequent conformational change may ylated forms of stathmin, although PPIL1 may have higher affect its interaction with these proteins. Further investigations affinity to the phosphorylated form. Future identification of on phosphorylation of SNW1/SKIP as well as the functional the responsible region(s) in stathmin for the binding and role of the interaction between PPIL1 and SNW1/SKIP may investigation using phosphorylation-specific antibodies to uncover not only the role of elevated PPIL1 in cancer cells but stathmin may uncover the mechanisms of regulation by the also mechanisms underlying transcription, splicing, and poly- interaction with PPIL1. adenylation. In our microarray, expression levels of stathmin were In this report, we documented that PPIL1 showed growth- markedly elevated in the colon cancer tissues we examined. promoting activity in normal mammalian cells, and that During mitosis, stathmin is inactivated by phosphorylation; reduction of its expression by siRNA suppressed the growth this process promotes polymerization and aids assembly of of colon cancer cells. The development of drugs to antagonize mitotic spindles (22). It also has been reported that the the function of PPIL1 may thus be a rational strategy for incorporation of tubulin into microtubules is regulated by the therapy of colon cancer. Because PPIL1 expression is elevated in interaction with stathmin-like domain. In addition, phosphor- cervical, gastric, and pancreatic cancers, and chronic myeloid ylation of the four serines (Ser16, Ser25, Ser38, and Ser63) occurs leukemia in our microarray data (data not shown), inhibitors sequentially at the G2-M transition (19). Recent studies of PPIL1 may be effective anticancer drugs for a wide range of reported that a stathmin phosphorylation gradient is necessary human tumors. Although further investigation of the function for correct spindle formation in mitotic cells (23). Therefore, of PPIL1 will be necessary, the data provided here should phosphorylation of the Ser38 and isomerization of the contribute to a more profound understanding of colonic following Pro by PPIL1 may be essential for the accelerated carcinogenesis and to development of novel therapeutic growth of neoplastic cells. Given that phosphorylated stathmin strategies for colon cancer. induces destabilization of microtubules (19), PPIL1 may induce the dissociation of stathmin from microtubules through the conformational change of stathmin. In addition, suppres- Acknowledgments sion of stathmin expression by antisense oligonucleotides We thank Yumi Nakajima, Noriko Ikawa, Kiyomi Jindo, and Tae Makino for tech- inhibits growth and reverses many of the phenotypes associated nical assistance; Drs. Toyomasa Katagiri, Yataro Daigo, and Toshihiro Tanaka for with malignant transformation (24). These data tempt us to contributions in fabrication of the cDNA microarray; and Dr. Hiroshi Okabe for hypothesize that phosphorylation of stathmin by specific helpful discussions.

References 1. Vogelstein B, Fearon ER, Hamilton SR, et al. Genetic by two different point mutations. Nature 1983;304: 6. Hollstein M, Sydransky D, Vogelstein B, Harris CC. alterations during colorectal-tumor development. 507^13. p53 mutations in human cancers. Science 1991;253: N Engl J Med 1988;319:525 ^ 32. 4. Nishisho I, NakamuraY, Miyoshi Y, et al. Mutations of 49^53. 2. Goel A, Arnold CN, Niedzwiecki D, et al. Character- chromosome 5q21genes in FAP and colorectal cancer 7. Liu W, Dong X, Mai M, et al. Mutations in AXIN2 ization of sporadic colon cancer by patterns of patients. Science1991;253:665 ^ 9. cause colorectal cancer with defective mismatch genomic instability. Cancer Res 2003;63:1608 ^ 14. 5. Morin PJ, Sparks AB, KorinekV, et al. Activation of h- repair by activating h-catenin/TCF signalling. Nat 3. Capon DJ, Seeburg PH, McGrathJP, et al. Activation catenin-Tcf signaling in colon cancer by mutations in Genet 2000;26:146 ^ 7. of Ki-ras2 gene in human colon and lung carcinomas h-catenin or APC. Science 1997;275:1787^ 90. 8. Ozaki K, FujiwaraT,Kawai A, et al. Cloning, expression

www.aacrjournals.org 75 Clin Cancer Res 2006;12(1) January 1, 2006 Downloaded from clincancerres.aacrjournals.org on September 30, 2021. © 2006 American Association for Cancer Research. Human Cancer Biology

and chromosomal mapping of a novel cyclophilin- patic cholangiocarcinoma. Hepatology 2005;41: Antisense RNA inhibition of phosphoprotein p18 related gene (PPIL1)fromhuman fetalbrain. Cytogenet 13 3 9 ^ 4 8 . expression abrogates the transformed phenotype of Cell Genet1996;72:242^ 5. 17. Shimokawa T, Furukawa Y, Sakai M, et al. Involve- leukemic cells. Cancer Res 1996;56:1445 ^ 50. 9. HunterT. Prolyl isomerases and nuclear function. Cell ment of the FGF18 gene in colorectal carcinogenesis, 25. MakarovEM, MakarovaOV, Urlaub H, et al. Small 19 9 8;92 :141 ^ 3. as a novel downstream target of the h-catenin/T-cell nuclear ribonucleoprotein remodeling during catalytic 10. Zheng H,You H, Zhou XZ, et al.The prolyl isomerase factor complex. Cancer Res 2003;63:6116^ 20. activation of the spliceosome. Science 2002;298: Pin1 is a regulator of p53 in genotoxic response. 18. TakahashiM,LinYM,NakamuraY,FurukawaY.Isola- 2205 ^ 8. Nature 2002;419:849 ^ 53. tion and characterization of a novel gene CLUAP1 26. Falk P, Puta F, Skruzny M. Transcriptional coregula- 11. Lu KP, Liou YC, Zhou XZ. Pinning down proline- whose expression is frequently upregulated in colon tor SNW/SKIP: the concealed tie of dissimilar path- directed phosphorylation signaling. Trends Cell Biol cancer. Oncogene 2004;23:9289 ^94. ways.CellMolLifeSci2004;61:629^40. 2002;12:164 ^ 72. 19. Cassimeris L. The oncoprotein 18/stathmin family 27. PrathapamT, Kuhne C, Banks L. Skip interacts with 12. Zhou XZ, Lu PJ, Wulf G, Lu KP. Phosphorylation- of microtubule destabilizers. Curr Opin Cell Biol 2002; the retinoblastoma tumor suppressor and inhibits its dependent prolyl isomerization: a novel signaling 14 :18 ^ 24. transcriptional repression activity. Nucleic Acids Res regulatory mechanism. Cell Mol Life Sci 1999;56: 20. Liu J, Farmer JD, Jr., Lane WS, Friedman J, 2002;30:5261 ^ 8. 788^806. Weissman I, Schreiber SL. Calcineurin is a common 28. Neubauer G, King A, RappsilberJ, et al. Mass spec- 13. Zacchi P, Gostissa M, Uchida T, et al. The prolyl target of cyclophilin-cyclosporin A and FKBP-FK506 trometry and EST-database searching allows charac- isomerase Pin1 reveals a mechanism to control p53 complexes. Cell 1991;66:807 ^ 15. terization of the multi-protein spliceosome complex. functions after genotoxic insults. Nature 2002;419: 21. Cande C,Vahsen N, Kouranti I, et al. AIF and cyclo- Nat Genet1998;20:46^50. 853^7. philin A cooperate in apoptosis-associated chromati- 29. Dahl R, Wani B, Hayman MJ. The Ski oncoprotein 14. Shaw PE. Peptidyl-prolyl isomerases: a new twist to nolysis. Oncogene 2004;23:1514^ 21. interacts with Skip, the human homolog of Drosophila transcription. EMBO Rep 2002;3:521 ^ 6. 22. Larsson N, Marklund U, Gradin HM, et al. Control of Bx42. Oncogene 1998;16:1579 ^ 86. 15. Lin YM, Furukawa Y,Tsunoda T, Yue CT, Yang KC, microtubule dynamics by oncoprotein 18: dissection 30. Leong GM, Subramaniam N, Figueroa J, et al. Ski- NakamuraY. Molecular diagnosis of colorectal tumors of the regulatory role of multisite phosphorylation dur- interacting protein interacts with Smad proteins to by expression profiles of 50 genes expressed differen- ing mitosis. Mol Cell Biol 1997;17:5530 ^ 9. augment transforming growth factor-h-dependent tially in adenomas and carcinomas. Oncogene 2002; 23. NiethammerP,BastiaensP,KarsentiE,etal. transcription. J Biol Chem 2001;276:18243 ^ 8. 21:4120 ^ 8. Stathmin-tubulin interaction gradients in motile and 31. Prathapam T, Kuhne C, Banks L. The HPV-16 E7 16. Obama K, Ura K, Li M, et al. Genome-wide mitotic cells. Science 2004;303:1862^6. oncoprotein binds Skip and suppresses its transcrip- analysis of gene expression in human intrahe- 24. Jeha S, Luo XN, Beran M, Kantarjian H, Atweh GF. tional activity. Oncogene 2001;20:7677 ^ 85.

Article on Overexpression of Peptidyl-Prolyl Isomerase Like-1 Association with Colon Cancer Cell Growth

In the article on Overexpression of Peptidyl-Prolyl Isomerase Like-1 in the January 1, 2006 issue of Clinical Cancer Research, a reference was missing in the Results section. The first sentence in the subheading, Interaction of PPIL1 with SNW1/SKIP, should read, as follows: Because Dictyostelium discoideum cyclophilin Cyp2 or Schizosaccharomyces pombe cyclophilin CypE binds to SnwA or Snw1p, respectively (Skruzny M, Ambrozkova M, Fukova I, et al: Cyclophilins of a novel subfamily interact with SNW/SKIP coregulator in Dictyostelium discoideum and Schizosaccharomyces pombe. Biochim Biophys Acta 2001:1521:146–51), we examined whether PPIL1 may associate in vivo with SNW1/SKIP, the human counterpart of SnwA and Snw1p. Also, the first author in reference 26 is P. Folk.

Obama K, Kato T, Hasegawa S, Satoh S, Nakamura Y, and Furukawa Y: Overexpression of Peptidyl-Prolyl Isomerase-Like 1 Is Associated with the Growth of Colon Cancer Cells. Clin Cancer Res 2006:12:70–6.

F 2006 American Association for Cancer Research. doi:10.1158/1078-0432.CCR-CNR12-6COR

www.aacrjournals.org1947 Clin Cancer Res 2006;12(6) March 15, 2006 Overexpression of Peptidyl-Prolyl Isomerase-Like 1 Is Associated with the Growth of Colon Cancer Cells

Kazutaka Obama, Tatsushi Kato, Suguru Hasegawa, et al.

Clin Cancer Res 2006;12:70-76.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/12/1/70

Cited articles This article cites 31 articles, 11 of which you can access for free at: http://clincancerres.aacrjournals.org/content/12/1/70.full#ref-list-1

Citing articles This article has been cited by 3 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/12/1/70.full#related-urls

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/12/1/70. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.