ANTICANCER RESEARCH 31: 11-22 (2011)

Frequent Absence of Tumor Suppressor FUS1 Expression in Human Bone and Soft Tissue

GUIDONG LI1,2, HIROYUKI KAWASHIMA1, LIN JI3, AKIRA OGOSE1, TAKASHI ARIIZUMI1, HAJIME UMEZU4, YONGJUN XU1, TETSUO HOTTA1 and NAOTO ENDO1

1Division of Orthopedic Surgery, Graduate School of Medical and Dental Sciences, Niigata University, Niigata, Niigata 951-8510, Japan; 2Department of Orthopedic Surgery, Second Affiliated Hospital of Harbin Medical University, Harbin, Heilongjiang 150086, P.R. China; 3Section of Thoracic Molecular , Department of Thoracic and Cardiovascular Surgery, University of Texas M.D. Anderson Cancer Center, Houston, TX 77030, U.S.A; 4Division of Clinical Pathology, Niigata University Hospital, Niigata, Niigata 951-8510, Japan

Abstract. Background: FUS1 is a Molecular genetic studies, such as identification of novel located on human chromosome 3p21.3. Frequent loss of and tumor suppressor genes, will greatly FUS1 protein expression is associated with lung cancer contribute to understanding the pathogenesis and progression development. This study examined FUS1 expression and its of sarcomas, and, thus, will promote the development of possible tumor-suppressive role in bone and soft tissue molecular- for various types of sarcomas. sarcomas. Materials and Methods: The expressions of FUS1 FUS1 is a novel tumor suppressor gene that was identified mRNA and FUS1 protein were assessed in cell in a 120-kb crucial region on human chromosome 3p21.3, lines, sarcoma tissues, benign bone and soft-tissue tumor where homozygous deletions frequently occur in lung and (BST) tissues, and healthy tissues. Exogenous FUS1 gene (5-8). The FUS1 gene product is a multi- transfection was performed on sarcoma cell lines. Results: functional protein and plays an important role in various FUS1 mRNA expression was detected in all sarcoma cell cellular processes, including transcription, signal lines, all benign BSTs and healthy tissues, and almost all transduction, cell-cycle progression and (8). High- sarcoma tissues. In contrast, FUS1 protein expression was level FUS1 mRNA expression has been detected in various frequently lost in sarcoma cells and sarcoma tissues. The healthy human tissues such as heart, brain, placenta and exogenous FUS1 gene delivery induced strong FUS1 protein lung, using multiple-tissue Northern blots (5). Unlike other expression, inhibition of cell viability and apoptosis in tumor suppressor genes, mutation of the FUS1 gene is sarcoma cells. Conclusion: FUS1 may act as a tumor infrequent in the majority of lung carcinomas (5). In contrast, suppressor in bone and soft-tissue sarcomas. expression of the FUS1 protein is frequently lost or reduced in primary lung cancer and is associated with lung cancer Bone sarcoma and soft-tissue sarcoma account for fewer than development and a worse survival rate (9). Overexpression 0.2% and 1% of all human carcinomas, respectively, but they of wild-type FUS1 in FUS1 protein-deficient tumor cells can be life threatening (1). Despite significant advances in dramatically suppressed tumor and induced multidisciplinary treatment, the long-term survival rates for apoptosis in vitro and in vivo (6-8). Systemic treatment with advanced bone and soft tissue sarcomas are poor (2-4). a combination of FUS1 nanoparticles and p53 or cisplatin in a human lung-cancer xenograft mouse model synergistically suppressed tumor growth and induced apoptosis (10, 11). To the Authors’ knowledge, there have been no reports Correspondence to: Hiroyuki Kawashima, MD, Ph.D., Division of regarding FUS1 expression in human bone and soft tissue Orthopedic Surgery, Graduate School of Medical and Dental sarcomas. To clarify whether the tumor suppressor FUS1 Sciences, Niigata University, 1-757 Asahimachi-dori, Niigata, also acts as a tumor suppressor gene in bone and soft-tissue Niigata 951-8510, Japan. Tel: +81 252272272, Fax: +81 2522720782, e-mail: [email protected] sarcomas, this study analyzed FUS1 gene and protein expression in various types of healthy and sarcoma cell lines Key Words: Tumor suppressor gene, FUS1, sarcoma, RT-PCR, and tissue samples, including a series of bone and soft tissue immunohistochemistry, growth inhibition. sarcomas, benign bone and soft tissue tumors (BSTs) and

0250-7005/2011 $2.00+.40 11 ANTICANCER RESEARCH 31: 11-22 (2011)

Table I. Summary of human cell lines used in the study.

Name Cell origin Source

WI-38 Normal human lung fibroblast RCB BMF Normal human bone marrow fibroblast Niigata University NHDF-AD Normal human dermal fibroblast (adult skin) Clonetics HOS Osteosarcoma IFO HuO9 Osteosarcoma JCRRB MG-63 Osteosarcoma HSRRB NOS1 Osteosarcoma Niigata University NOS10 Osteosarcoma Niigata University OST Osteosarcoma RCB SaOS2 Osteosarcoma ATCC U2-OS Osteosarcoma ATCC NMFH1 Malignant fibrous histocytoma Niigata University NMFH2 Malignant fibrous histocytoma Niigata University ASPS-KY Alveolar soft part sarcoma Dr. S. Yanoma HS-SY-II Synovial sarcoma Dr. H. Sonobe NMS-2 Malignant peripheral nerve sheath tumor Niigata University OUMS-27 Chondrosarcoma M. Namba SKNMC Ewing’s sarcoma ATTC SFT8606 Epithelioid sarcoma Dr. H. Iwasaki 402-92 Liposarcoma Dr. P. Åman

RCB, Riken Cell Bank, Tsukuba, Japan; Clonetics, Clonetics/Lonza Walkersville Inc., Walkersville, MD, USA; IFO, Institute for Fermentation, Osaka, Japan; JCRRB, Japanese Cancer Research Resources Bank, Tokyo, Japan; HSRRB, Health Science Research Resources Bank, Osaka, Japan; ATCC, American Type Culture Collection, Bethesda, MD, USA.

healthy bone and soft tissues. Moreover, the study 38, HOS, MG63 and SKNMC cells were maintained in α-MEM determined whether forced exogenous FUS1 gene expression (Invitrogen). All media were supplemented with 10% FBS (PAA, may inhibit the tumor growth of sarcoma cells in vitro. Pasching, Austria) containing 1% antibiotics and antimycotics (Invitrogen), except the WI-38 cells which were antibiotic-free. In Materials and Methods addition, the NHDF-AD cells were cultured in fibroblast basal medium supplemented with FGM-2 SingleQuots (Clonetics/Lonza, Walkersville, MD, USA). All cell cultures were incubated at 37˚C in Cell lines and cell culture. Sarcoma cell lines, including the an atmosphere of 5% CO2 with 100% humidity. osteosarcoma cell lines NOS1 (12) and NOS10 (13), the malignant fibrous histiocytoma (MFH) cell lines NMFH1(14) and NMFH2 Tissue samples. A total of 147 malignant and benign BST tissue (15), the malignant peripheral nerve sheath tumor cell line NMS-2 samples (Table II) were obtained from 135 patients who had (16), and a normal human bone marrow fibroblast (BMF) cell line, received surgical resection or open biopsy (79 males and 56 were established under the approval of the Institutional Review females) from 2001 to 2008 in Niigata University Hospital, Japan. Board (Niigata University Hospital). The alveolar soft part sarcoma These BST samples included 95 non-metastatic sarcomas, 12 cell line ASPS-KY was a gift from Dr. S. Yanoma (Kanagawa metastatic sarcomas and 12 distant metastatic foci, as well as 28 Cancer Center, Yokohama, Japan); the synovial sarcoma HS-SY-II benign BSTs (Table II). After resection, each tumor sample was cell line (17) was a gift from Dr. H. Sonobe (Department of divided into two pieces: one was stored at –80˚C, and the other was Pathology, Kochi Medical School. Kochi, Japan); the formalin-fixed, paraffin-embedded and, then, diagnosed by chondrosarcoma cell line OUMS-27 was a gift from Dr. M. Namba experienced pathologists according to the WHO classification (Department of Cell Biology, Okayama University Medical School, system (20). Healthy tissues including tissues from the skin, bone, Okayama, Japan); the epithelioid sarcoma cell line SFT8606 (18) muscle and fat (three each), peripheral nerve and blood vessels (two was a gift from Dr H. Iwasaki (Fukuoka University School of each) and tendon and fascia (one each) were obtained from patients Medicine, Fukuoka, Japan); and the liposarcoma cell line 402-92 suffering from curative amputation or other non-tumor diseases and (19) was a gift from Dr. P. Åman (Department of Clinical Genetics, served as healthy controls. This study was approved by the University Hospital, Lund, Sweden). The remaining cell lines used Institutional Review Board of Niigata University Hospital and in this study were obtained from commercial sources (Table I). informed consent was obtained from all patients. NOS1, NOS10, SaOS2, HuO9, OST, NMFH1, NMFH2, NMS-2, ASPS-KY, 402-92 and SFT8606 cells were maintained in RPMI- Reverse transcriptase-polymerase chain reaction (RT-PCR) analysis. 1640 (Invitrogen, Carlsbad, CA, USA); U2-OS, HS-SY-II, and Total RNA was extracted from cell lines (Table I), frozen tumor OUMS-27 cells were maintained in DMEM (Invitrogen); BMF, WI- tissue samples (Table II) and healthy bone and soft tissues using

12 Li et al: FUS1 Expression in Sarcomas

Table II. Expression of FUS1 mRNA in BSTs. Table III. Immunohistochemical staining of FUS1 protein in BSTs.

Histological type No. FUS1 mRNA Histological type No. FUS1 protein

+ – + –

A Malignant primary tumors A Malignant primary tumors Non-metastatic Osteosarcoma (localized) 11 0 11 Osteosarcoma 9 9 0 Osteosarcoma (metastatic) 4 0 4 MFH 27 25 2 MFH 27 1 26 ASPS 3 3 0 ASPS 3 0 3 Chondrosarcoma 9 9 0 Chondrosarcoma 3 0 3 Epithelioid sarcoma 3 3 0 Epithelioid sarcoma 3 0 3 Ewing’s sarcoma 6 6 0 Ewing’s sarcoma 3 0 3 Leiomyosarcoma 5 5 0 Leiomyosarcoma 3 1 2 Liposarcoma 14 14 0 Liposarcoma 4 1 3 MPNST 3 3 0 MPNST 3 0 3 Rhabdomyosarcoma 8 8 0 Rhabdomyosarcoma 3 0 3 Synovial sarcoma 7 7 0 Synovial sarcoma 3 0 3 Metastatic B Benign tumors Osteosarcoma 9 9 0 Desmoid 1 0 1 MFH 2 2 0 GCT of bone 2 0 2 Synovial sarcoma 1 1 0 Hemangioma 3 0 3 B Metastatic tissues Lipoma 3 0 3 Lung (osteosarcoma) 8 8 0 Neurofibroma 3 0 3 Retroperitoneum (osteosarcoma) 1 1 0 Schwannoma 3 0 3 Lung (MFH) 2 2 0 Total 85 3 82 Lung (synovial sarcoma) 1 1 0 C Benign tumors MFH, Malignant fibrous histocytoma; ASPS, alveolar soft part sarcoma; Desmoid 5 5 0 MPNST, malignant peripheral nerve sheath tumor; GCT, giant cell GCT of bone 5 5 0 tumor. Hemangioma 4 4 0 Lipoma 5 5 0 Neurofibroma 4 4 0 Schwannoma 5 5 2 appropriate area for further analysis. Three cylindrical cores (2-mm in Total 147 145 2 diameter) were punched out from the areas of interest in each donor paraffin block and were re-embedded in recipient tissue microarray MFH, Malignant fibrous histocytoma; ASPS, alveolar soft part sarcoma; blocks, which were manually prepared. Immunohistochemical staining MPNST, malignant peripheral nerve sheath tumor; GCT, giant cell of FUS1 protein was performed using a rabbit anti-FUS1 polyclonal tumor. that was raised against a synthetic oligopeptide (GASGSKARGLWPFASAA), derived from the NH2-terminal amino sequence of the FUS1 protein (Bethyl Laboratories, Montgomery, TX, Isogen Reagent (Nippongene, Toyama, Japan) according to the USA) (8) and a previously reported protocol (9). Briefly, tissue manufacturer’s instructions. Total RNA (1.0 μg) was converted to microarray slides (4 μm) were deparaffinized and hydrated and cDNA using M-MLV RT (Invitrogen) and oligo dT primers was retrieved by heating in 10 mM sodium citrate buffer (pH 6.0) using (Promega, Madison, WI, USA). An equal amount of cDNA (1 μl) a steamer for 10 min at 98˚C. The slides were then blocked with 3% was amplified by PCR using FUS1-specific oligonucleotide primer hydrogen peroxide in methanol at room temperature for 20 min, pairs: forward, 5’-ATGATGAGGATGGGGATCTG-3’; reverse, 5’- followed by 10% bovine serum albumin in Tris-buffered saline Tween- GAGGATCACAGGGAAATCCA-3’. The PCR reaction conditions 20 for 30 min. After incubation with the anti-FUS1 (1:400) were as follows: 94˚C for 2 min, followed by 30 cycles of 94˚C for for 60 min at room temperature, the slides were washed with PBS and 30 s, 56˚C for 30 s, and 72˚C for 90 s, with a final extension at 72˚C stained using a Histofine Simple Stain PO (MULTI) (Nichirei, for 5 min. A plasmid vector encoding FUS1 cDNA was used as a Tokyo, Japan) for 30 min according to the manufacturer’s instructions. positive control. RT-PCR of the housekeeping gene GAPDH was The signals were then developed using DAB substrate (Nichirei) and, used as an internal control and its amplification was performed as finally, the slides were counterstained with hematoxylin. Human previously described (21). healthy lung bronchial epithelia were used as a positive control.

Immunohistochemistry using a tissue microarray. Eighty-five formalin- Transient transfection. A recombinant plasmid vector encoding the fixed and paraffin-embedded tumor samples (Table III), as well as wild-type FUS1 gene (wt-FUS1) was used for transient exogenous various healthy bone and soft tissue samples, were retrieved from the gene transfection. A plasmid vector encoding the β-galactosidase institutional archive (Niigata University Hospital). Prior to gene (LacZ) was used as a nonspecific negative control. immunohistochemistry, the hematoxylin and eosin staining of all tumor Construction of these plasmid vectors has been described previously samples was carefully examined under a microscope to select the (6, 8). Cultured cells were transfected with plasmid DNA using

13 ANTICANCER RESEARCH 31: 11-22 (2011)

FuGENE 6 Transfection Reagent (Roche Applied Science, analysis. Six sarcoma cell lines (NOS10, SaOS2, ASPS-KY, HS-SY- Indianapolis, IN, USA) according to the manufacturer’s instructions. II, NMS-2 and OUMS-27) were transfected with wt-FUS1 or LacZ A FuGENE 6 reagent:DNA ratio of 3:1 (μl:μg respectively) was plasmid DNA. The expression levels of cleaved caspase-3 and applied to each 60 mm culture dish or each well of a 96-well plate. cleaved PARP were analyzed at 48 and 72 h after transfection using rabbit anti-caspase-3 (1:1000; Cell Signaling Western blot analysis. Cultured cells (Table I) were harvested and Technology, Beverly, MA, USA) and anti-PARP polyclonal subjected to Western blot analysis to determine endogenous antibodies (1:1000; Cell Signaling Technology), respectively. The expression of the FUS1 protein. WI-38, a normal human lung Western blotting procedure was performed as described earlier fibroblast that is known to express endogenous FUS1 protein was except that the proteins were loaded on a 7.5% SDS-PAGE gel for used as a positive control (8). Eight sarcoma cell lines as well as analysis of the cleaved PARP protein. the normal cell line WI-38 were chosen for exogenous FUS1 gene transfection. The cells were plated on 60 mm culture dishes at a Statistical analysis. All data were evaluated using SPSS 14.0 density of 2×105/dish 24 h prior to treatment and, subsequently, the software (SPSS Inc., Chicago, IL, USA). Correlations between cells in each dish were transfected with 2 μg of wt-FUS1 or LacZ tumor histological types and FUS1 expression (mRNA and protein) plasmid DNA using the FuGENE 6 reagent. At 48, 72 and 96 h post were assessed using Fisher’s exact test. The level of statistical transfection, the cells were subjected to Western blotting to assess significance was set to p<0.05. The statistical significance of the expression level of the FUS1 protein. Briefly, the cells were differences between wt-FUS1 and LacZ gene transfection groups washed with ice-cold PBS and were lysed by adding sample buffer were evaluated using a two-tailed paired sample t-test. Data are (62.5 mM Tris, pH 6.8; 2% SDS; 5% glycerol and 6 M urea) presented as mean±standard deviation. A p-value <0.01 denoted containing a complete protease inhibitor cocktail (Roche statistical significance. Diagnostics, Indianapolis, IN, USA) to each dish. The cell lysates were centrifuged at 14,000 ×g for 5 min at 4˚C, the supernatants Results were collected and the protein concentration of the lysates was determined using a BCA protein assay kit (Pierce, Rockford, IL, Loss of FUS1 gene expression is rare in both sarcomas and USA). Dithiothreitol (1 M; Sigma-Aldrich, St. Louis, MO, USA) benign BSTs. FUS1 gene expression was analyzed in three and bromophenol blue were then added to the lysates to a final concentration of 5% (v/v) each and the lysates were boiled for 5 human normal fibroblasts, 17 sarcoma cell lines (Table I), 119 min. Equal amounts of proteins (50 μg/lane) were separated on 15% sarcoma samples, 28 benign BSTs (Table II) and in various SDS-PAGE gels (Bio-Rad, Hercules, CA, USA) for 1 h and were healthy bone and soft tissue samples using RT-PCR analysis. electrotransferred onto Hybond ECL nitrocellulose membranes (GE A FUS1 mRNA product, 164 bp in size, was detected in all Healthcare, Little Chalfont, Buckinghamshire, UK) for 2 h. The normal and sarcoma cell lines (Figure 1A), in all benign membranes were blocked in blocking buffer (5% milk in 1 mM BSTs (100%) and in 117 out of 119 (98%) sarcoma samples Tris-buffered saline containing 0.2% Tween-20) for 1 h, and were (Figure 1B and Table II), as well as in all healthy bone and then probed overnight at 4˚C with a rabbit anti-FUS1 polyclonal antibody (1:1000). The immunoblots were subsequently incubated soft tissue samples (Figure 1C). Complete loss of FUS1 with horseradish peroxidase-labeled anti-rabbit IgG (GE Healthcare) mRNA expression was observed in only two cases, both of for 1 h at room temperature and were developed using ECL which were MFH tumor samples (Figure 1B). There were no detection reagents (GE Healthcare). The blots were reprobed using significant differences in FUS1 mRNA expression between a mouse anti-β-actin monoclonal antibody (1:3000, Sigma-Aldrich) the sarcomas and the benign BSTs or between the non- to ensure uniform sample loading. metastatic and the metastasized sarcoma samples.

Cell viability and morphology. Eight sarcoma cell lines were plated on 96-well plates (2×103 cells/well) 24 h before gene transfection. FUS1 protein expression is frequently absent from sarcomas, The cells in each well were transfected with 0.05 μg wt-FUS1 benign BSTs and healthy bone and soft tissues. FUS1 protein plasmid DNA using 0.15 μl FuGENE 6. As a control, cells were expression was analyzed in normal fibroblasts and in also transfected with LacZ plasmid vectors. Cell viability was sarcoma cell lines (Table I) by Western blot analysis. assessed using the XTT (sodium 3’-[1-(phenylaminocarbonyl)-3,4- tetrazolium]-bis (4-methoxy-6-nitro) benzene sulfonic acid hydrate) Compared with its expression in normal lung fibroblast WI- assay 24, 48, 72 and 96 h after gene transfection using the Cell 38 cells (positive control), FUS1 protein expression was Proliferation Kit-II (Roche Diagnostics) according to the relatively low in two normal fibroblast cell lines of manufacturer’s protocol. In addition, 2×105 cells were also plated mesenchymal origin (BMF and NHDF), was extremely low on 60 mm culture dishes and were, subsequently, transfected with in eight of the 17 sarcoma cell lines (HOS, MG-63, NOS1, 2 μg of wt-FUS1 plasmid DNA. Cell morphology was observed OST, SaOS2, NMFH1, NMFH2 and SFT-8606) and was daily using an Olympus phase-contrast microscope ULWCD 0.30 completely absent from 9 of the 17 sarcoma cell lines (IMT2; Olympus, Tokyo, Japan) and photomicrographs were taken, at a magnification of ×40, 72 h after transfection. (HuO9, NOS10, U2-OS, ASPS-KY, HS-SY-II, NMS-2, OUMS-27, SKNMC and 402-92) (Figure 1D). FUS1 protein Apoptosis analysis. To assess the induction of apoptosis after wt- expression was also analyzed by immunohistochemical FUS1 plasmid DNA transfection, the cleavage of caspase-3 and poly staining of a large number of malignant and benign BSTs (ADP-ribose)-polymerase (PARP) were analyzed by Western blot and of various healthy bone and soft tissues. None of the 70

14 Li et al: FUS1 Expression in Sarcomas

Figure 1. FUS1 expression in human sarcoma cell lines, BST samples and in normal bone and soft tissues. The expression of FUS1 mRNA was determined in 3 human normal fibroblasts and 17 sarcoma cell lines (A); representative cases of sarcomas and benign BSTs (OS1-5, osteosarcoma; MFH1-11, malignant fibrous histiocytoma; Ch1-5, chondrosarcoma; GCT1-5, giant cell tumor of bone; Schw1-5, Schwannoma) (B); and normal bone and soft tissues (C), using RT-PCR analysis and the FUS1 gene-specific primer sets described in ‘Materials and Methods’. A FUS1-expressing plasmid cDNA was used as a positive control and GAPDH mRNA was used as an internal control. Western blot analysis of FUS1 protein expression in human normal fibroblasts and in sarcoma cell lines (D). The FUS1 protein band was detected using a rabbit anti-FUS1 polyclonal antibody as described in the Materials and Methods. Normal human lung fibroblast WI-38 was assayed as a positive control. The blots were reprobed with anti- β-actin that was used as a loading control. Data are representative of three independent experiments.

15 ANTICANCER RESEARCH 31: 11-22 (2011)

Figure 2. FUS1 protein expression in BSTs and normal tissues by tissue microarray immunohistochemstry. A: Positive control, normal human bronchial epithelia. B, C and D: Three sarcomas, including a case of MFH (B), a case of leiomyosarcoma (C) and a case of liposarcoma (D), showing heterogeneous, moderate or mild expression, respectively, of the FUS1 protein in the cytoplasm (arrows). E-L: Representative cases of sarcomas and benign BSTs exhibiting negative expression of the FUS1 protein: osteosarcoma (E); malignant fibrous histiosarcoma (F); alveolar soft part sarcoma (G); chondrosarcoma (H); Ewing’s sarcoma (I); synovial sarcoma (J); hemangioma (K); Schwannoma (L). M-P: FUS1 protein expression in representative cases of normal bone and soft tissues. The expression of FUS1 protein was detected in the axons of peripheral nerve tissue (arrows), but not in myelin sheaths or Schwann cells (M). FUS1 protein expression is completely absent from bone (N), muscle (O) and fat (P) tissue samples. Original magnification: ×200 (main images), ×30 (insets). Scale bar, 50 μm.

sarcoma samples examined displayed detectable expression the myelin sheaths, which are the primary sites of origin of of the FUS1 protein (Table III and Figure 2B-J) with the peripheral nerve sheath tumors (Figure 2M). Thus, it appears exception of three sarcoma cases (4%) that showed patchy that FUS1 protein expression is generally absent from both positive immunoreactivity with the anti-FUS1 antibody. Of sarcomas and benign BST samples. these three cases, one MFH case and one leiomyosarcoma case showed a moderate reaction, while one liposarcoma Forced FUS1 gene expression inhibits the growth of sarcoma case displayed a mild reaction (Figure 2B, C and D). None cells. To test the hypothesis that FUS1 protein can function as of these three cases showed loss of FUS1 mRNA expression. a tumor suppressor in bone and soft tissue sarcomas as it does FUS1 protein expression was also undetectable in benign in other tumor types, the wt-FUS1 gene or the negative control BSTs (Table III and Figure 2K and L) and in healthy bone LacZ gene were transiently transfected into eight sarcoma cell and soft tissues except for peripheral nerve tissues (Figure lines which exhibited extremely low endogenous FUS1 protein 2M-P). The FUS1 protein was expressed in the axons of expression or a complete loss of it. FUS1 protein expression peripheral nerve tissues but it was not expressed in any of was analyzed by Western blotting of cell extracts 48, 72 and

16 Li et al: FUS1 Expression in Sarcomas

Figure 3. Exogenous wt-FUS1 gene transfection induced strong FUS1 protein expression in bone and soft tissue sarcoma cells. Eight sarcoma cell lines were plated at a density of 2×105 on 60-mm dishes 24 h before transfection with a plasmid DNA expressing the wt-FUS1 gene or a control LacZ gene using the FuGENE 6 transfection reagent. The cells were harvested at 48, 72 and 96 h after transfection, and cell lysates were subjected to Western blot analysis using an anti-FUS1 antibody. The normal lung fibroblast cell line, WI-38, was assayed as a normal control. The blots were reprobed with an anti-β-actin antibody to confirm equal loading. Data are representative of three independent experiments.

96 h after exogenous wt-FUS1 gene delivery. This analysis phase-contrast analysis of cell morphology 72 h post revealed that six out of the eight sarcoma cell lines (NOS10, transfection (Figure 4B). However, cell viability analysis did SaOS2, ASPS-KY, HS-SY-II, NMS-2 and OUMS-27) not reveal any statistically significant cell death in the other exhibited strong, time-dependent FUS1 protein expression two sarcoma cell lines (NMFH and SKNMC) after exogenous when transfected with wt-FUS1, but not when transfected with wt-FUS1 gene transfection compared to LacZ transfection. LacZ. The other two FUS1-transfected sarcoma cell lines displayed very low (NMFH2) or undetectable (SKNMC) Activation of apoptosis by exogenous FUS1 expression. FUS1 protein expression (Figure 3). In contrast, there was no Activation (cleavage) of caspase-3 plays a critical role in the significant difference in FUS1 protein expression levels execution of apoptosis (22), and the proteolytic cleavage of between the wt-FUS1-transfected and the control vector LacZ- PARP by activated caspase-3 serves as a marker of apoptosis transfected normal lung fibroblast WI-38 cells. (23). The present study, therefore, determined whether the To examine whether the exogenously expressed FUS1 significant suppression of tumor cell growth induced by wt- protein can suppress sarcoma tumor growth, the viability of FUS1 gene transfection is mediated by the induction of these eight sarcoma cell lines was analyzed at 24, 48, 72 and apoptosis. Using Western blot analysis, the protein expression 96 h after wt-FUS1 or LacZ-gene transfection, using the XTT levels of cleaved caspase-3 and cleaved PARP were examined assay and assigning a viability value of 1 to each LacZ- in the six sarcoma cells (NOS10, SaOS2, ASPS-KY, HS-SY- transfected control. FUS1 transfection significantly, and in a II, NMS-2 and OUMS-27), which exhibited strong FUS1 time-dependent manner, decreased the viability of six of the protein expression after transient gene transfection. The cleaved sarcoma cell lines (NOS10, SaOS2, ASPS-KY, HS-SY-II, caspase-3 protein product (19/17 kDa) was clearly detected 48 NMS-2 and OUMS-27), resulting in viability values relative and 72 h after wt-FUS1 gene transfection in five out of these to LacZ-transfected cells of 0.87-0.77, 0.65-0.59, 0.73-0.56, six sarcoma cells (NOS10, SaOS2, ASPS-KY, HS-SY-II and 0.87-0.71, 0.81-0.68 and 0.86-0.74, respectively, 48-96 h after NMS-2) (Figure 5A). However, activation of caspase-3 in the transfection (p<0.01, paired t-test) (Figure 4A). For three of chondrosarcoma OUMS-27 cells was ambiguous at 48 h and these six sarcoma cell lines (SaOS2, ASPS-KY and NMS-2), was undetectable at 72 h post transfection (Figure 5A). In significant inhibition of tumor cell growth was observed even addition, a strong increase in the expression levels of the as early as 24 h after wt-FUS1 gene transfection, with cleaved PARP protein (89 kDa) was detected in all six FUS1- viabilities of 0.85, 0.83 and 0.89, respectively, compared to transfected sarcoma cells compared with the control LacZ- the control LacZ group (p<0.01, paired t-test). The obvious transfected cells (Figure 5B). Cleavage of the PARP protein cell death mediated by exogenous wt-FUS1 gene transfection observed in the OUMS-27 cells may be mediated by a caspase into these six sarcoma cell lines was also demonstrated by other than caspase-3, such as by caspase-7 (24).

17 ANTICANCER RESEARCH 31: 11-22 (2011)

Figure 4. Exogenous wt-FUS1 gene transfection decreases cell viability and alters cell morphology of bone and soft tissue sarcoma cells. A: Eight sarcoma cell lines were plated on 96-well plates (2×103/well) and were transfected with a plasmid encoding the wt-FUS1 gene or a control LacZ gene using the FuGENE 6 transfection reagent after 24 h incubation. Cell viability was assessed at 24, 48, 72 and 96 h after transfection using an XTT assay. The viability of the LacZ-transfected cells was regarded as 1.0, and the viability of FUS1-transfected cells was compared relative to this level. Asterisks indicate significant differences at p<0.01. Date are presented as the mean ± standard deviation (n=4). B: Photomicrographs of human sarcoma cells after exogenous wt-FUS1 gene transfection. Cell morphology was observed using a phase-contrast microscope and cells were photographed at a magnification of ×40, 72 h after gene transfection.

18 Li et al: FUS1 Expression in Sarcomas

Figure 5. Restoration of FUS1 protein leads to apoptosis in bone and soft tissue sarcoma cell lines in vitro. Six sarcoma cells (NOS10, SaOS2, ASPS- KY, HS-SY-II, NMS-2 and OUMS-27) which exhibited strong exogenous FUS1 protein expression were plated on 60-mm dishes at a density of 2×105 cells the day before transfection. Cells were harvested at 48, 72 and 96 h after transfection with a wt-FUS1 or a LacZ plasmid vector using the FuGENE 6 reagent and induction of apoptosis was analyzed by detection of the cleaved products of caspase-3 (A) and poly(ADP-ribose)polymerase (PARP) (B) using Western blotting. β-Actin was assayed as a loading control. The results are representative of three independent experiments.

Discussion may also act as tumor suppressor in human bone and soft tissue sarcomas. FUS1, which is a novel candidate tumor suppressor gene, Somatic mutation of the FUS1 gene is rare in both lung plays an important role in the earliest steps of lung cancer cancer cell lines and in primary lung cancers (<4%) and the pathogenesis and progression (9). FUS1-deficient mice gene transcription levels in these lung cancer cell lines and showed an increased predisposition to a certain range of tumor samples are similar to those in normal lung and tumors (25). It is unknown whether FUS1 also functions as fibroblast cells (5). The results of the present study are a ‘gate-keeper’ in the pathogenesis of human bone and soft consistent with these previous data. FUS1 mRNA was not tissue sarcomas. This study provided evidence that FUS1 only expressed in all three human normal fibroblast cell lines

19 ANTICANCER RESEARCH 31: 11-22 (2011) examined, but also in all healthy tissue, all benign BSTs, all similar mechanisms of FUS1 down-regulation may also sarcoma cell lines and almost all sarcoma tumors analyzed operate in sarcoma cells. (Figure 1A-C and Table II). These data indicated that loss of The present study involving transfection of exogenous FUS1 gene expression is also a rare event in bone and soft FUS1 into sarcoma cell lines and showing weak, or no, tissue sarcomas. endogenous FUS1 protein expression, suggested that FUS1 In contrast to FUS1 mRNA expression, it has been shown does indeed play a tumor-suppressor function in sarcoma cells. that endogenous FUS1 protein expression cannot be detected in It was shown that six of the eight FUS1-transfected sarcoma lung cancer cell lines by Western blotting using FUS1-specific cells showed high protein expression of exogenous FUS1 that polyclonal antibodies, in spite of its expression in normal was accompanied by decreased cell viability compared to human bronchial cells and lung fibroblast cells (8). Loss or LacZ-transfected cells and by induction of apoptosis (Figures reduction of FUS1 protein expression was detected in 82% of 3-5). It is possible that regulation of the protein expression and non-small cell lung cancer specimens and in 100% of small cell the activity of FUS1 in sarcomas may differ from those in lung lung cancer specimens using immunohistochemical tissue carcinomas. FUS1 protein expression in healthy tissue or in microarray analysis. However, FUS1 protein expression was not benign BSTs may be silenced at a post-translational stage, and lost in any of the non-malignant bronchial epithelial specimens may be activated or up-regulated only in response to (9). In the present study, compared with FUS1 protein oncogenic stresses or apoptotic stimuli. However, the number expression in the normal lung fibroblast WI-38 cell line of healthy tissue samples used in this study was insufficient (positive control), the expression of FUS1 protein was also for an analysis of the reason why FUS1 protein expression is either very low or completely absent from the sarcoma cell lines lacking from these tissues. tested, although FUS1 mRNA expression was detected in all In conclusion, this study showed the frequent lack of these cell lines (Figure 1D). In addition, the immuno- FUS1 protein expression in bone and soft tissue sarcomas histochemical analysis detected FUS1 protein expression in only and in benign BSTs, as well as in healthy mesenchymal 3 out of 70 sarcoma samples (Table III and Figure 2). However, tissues, despite their high FUS1 mRNA expression. FUS1 is the results of this study regarding FUS1 protein expression in a potential tumor suppressor of sarcoma cells and may play normal cells, healthy tissues and benign tumor tissues differ an important role in the pathogenesis of sarcomas. Further from previous studies. This is because FUS1 protein expression studies of FUS1 gene and protein expression in a larger was relatively low in two normal fibroblasts that originated from number of BSTs and healthy bone and soft tissues, and mesenchymal tissues (Figure 1D) and was also absent from all studies of its biological activity in vitro and in vivo are of the benign BST samples and of the various healthy human required to confirm its precise role in these tumors. bone and soft tissues assayed (Table III and Figure 2). Thus, whereas previous observations showed a high level of FUS1 References protein expression in healthy lung bronchial epithelia, but a reduced or lost expression in preneoplastic respiratory epithelia 1 Jemal A, Siegel R, Ward E, Hao Y, Xu J and Thun MJ: Cancer and malignant lung tumors (9), in the present statistics, 2009. CA Cancer J Clin 59: 225-249, 2009. 2 Goorin AM, Schwartzentruber DJ, Devidas M, Gebhardt MC, immunohistochemical study, the expression of the FUS1 protein Ayala AG, Harris MB, Helman LJ, Grier HE and Link MP: was undetectable in both healthy and benign mesenchymal Pediatric Oncology Group: Presurgical chemotherapy compared tumor tissues. There are two possible explanations for these with immediate surgery and adjuvant chemotherapy for results. Firstly, FUS1 may function differently as a tumor nonmetastatic osteosarcoma: Pediatric Oncology Group Study suppressor in different tissues. Secondly, FUS1 may not be a POG-8651. J Clin Oncol 21: 1574-1580, 2003. major contributor to the tumorigenesis of sarcomas. 3 Lewis IJ, Nooij MA, Whelan J, Sydes MR, Grimer R, The exact mechanism of the inactivation of FUS1 protein Hogendoorn PC, Memon MA, Weeden S, Uscinska BM, van Glabbeke M, Kirkpatrick A, Hauben EI, Craft AW and Taminiau expression in primary human carcinomas, including lung and AH: MRC BO06 and EORTC 80931 collaborators; European breast cancer, remains unknown. The haploinsufficiency that Osteosarcoma Intergroup: Improvement in histologic response occurs at the 3p21.3 region where the FUS1 gene resides but not survival in osteosarcoma patients treated with intensified may result in a decrease in, or loss of, FUS1 protein chemotherapy: a randomized phase III trial of the European expression (6, 8, 26). A reduction or loss of FUS1 protein Osteosarcoma intergroup. J Natl Cancer Inst 99: 112-118, 2007. may also result from its rapid proteasome-dependent 4 Pisters PW: Clinical evaluation and treatment of soft tissue tumors. degradation due to a deficiency in post-translational In: Enzinger and Weiss’s Soft Tissue Tumors. 5th edn. Weiss SW and Goldblum JR (eds.). St Louis, Mosby, pp. 15-31, 2001. myristoylation of FUS1 proteins. Alternatively, FUS1 protein 5 Lerman MI and Minna JD: The 630-kb lung cancer homozygous translation may be down-regulated after transcription by deletion region on human chromosome 3p21.3: identification micro-RNAs that target the 3’UTR of FUS1 (27, 28). and evaluation of the resident candidate tumor suppressor genes. Although the exact mechanism by which FUS1 protein The International Lung Cancer Chromosome 3p21.3 Tumor expression is inhibited in soft tissue sarcomas is unknown, Suppressor Gene Consortium. Cancer Res 60: 6116-6133, 2000.

20 Li et al: FUS1 Expression in Sarcomas

6 Kondo M, Ji L, Kamibayashi C, Tomizawa Y, Randle D, Sekido 17 Sonobe H, Manabe Y, Furihata M, Iwata J, Oka T, Ohtsuki Y, Y, Yokota J, Kashuba V, Zabarovsky E, Kuzmin I, Lerman M, Mizobuchi H, Yamamoto H, Kumano O and Abe S: Roth JA and Minna JD: Overexpression of candidate tumor Establishment and characterization of a new human synovial suppressor gene FUS1 isolated from the 3p21.3 homozygous sarcoma cell line, HS-SY-II. Lab Invest 67: 498-505, 1992. deletion region leads to G1 arrest and growth inhibition of lung 18 Iwasaki H, Ohjimi Y, Ishiguro M, Isayama T, Kaneko Y, Yoh S, cancer cells. 20: 6258-6262, 2001. Emoto G and Kikuchi M: Epithelioid sarcoma with an 18q 7 Ji L, Nishizaki M, Gao B, Burbee D, Kondo M, Kamibayashi C, aberration. Cancer Genet Cytogenet 91: 46-52, 1996. Xu K, Yen N, Atkinson EN, Fang B, Lerman MI, Roth JA and 19 Aman P, Ron D, Mandahl N, Fioretos T, Heim S, Arheden K, Minna JD: Expression of several genes in the human Willén H, Rydholm A and Mitelman F: Rearrangement of the chromosome 3p21.3 homozygous deletion region by an gene CHOP in myxoid liposarcomas with adenovirus vector results in tumor suppressor activities in vitro t(12;16)(q13;p11). Genes Chromosomes Cancer 5: 278-285, 1992. and in vivo. Cancer Res 62: 2715-2720, 2002. 20 Fletcher CDM, Unni KK and Mertens F (eds.): World Health 8 Uno F, Sasaki J, Nishizaki M, Carboni G, Xu K, Atkinson EN, Organization Classification of Tumours. Pathology and Genetics of Kondo M, Minna JD, Roth JA and Ji L: Myristoylation of the Tumours of Soft Tissue and Bone. Lyon, France: IARC Press, 2002. FUS1 protein is required for tumor suppression in human lung 21 Asaoka K, Tada M, Sawamura Y, Ikeda J and Abe H: cancer cells. Cancer Res 64: 2969-2976, 2004. Dependence of efficient adenoviral gene delivery in malignant 9 Prudkin L, Behrens C, Liu DD, Zhou X, Ozburn NC, Bekele BN, glioma cells on the expression levels of the Coxsackievirus and Minna JD, Moran C, Roth JA, Ji L and Wistuba II: Loss and adenovirus receptor. J Neurosurg 92: 1002-1008, 2000. reduction of FUS1 protein expression is a frequent phenomenon in 22 Fernandes-Alnemri T, Litwack G and Alnemri ES: CPP32, a the pathogenesis of lung cancer. Clin Cancer Res 14: 41-47, 2008. novel human apoptotic protein with homology to Caenorhabditis 10 Deng WG, Kawashima H, Wu G, Jayachandran G, Xu K, Minna elegans cell death protein Ced-3 and mammalian interleukin-1 JD, Roth JA and Ji L: Synergistic tumor suppression by beta-converting . J Biol Chem 269: 30761-30764, 1994. coexpression of FUS1 and p53 is associated with down-regulation 23 Oliver FJ, de la Rubia G, Rolli V, Ruiz-Ruiz MC, de Murcia G of murine double minute-2 and activation of the apoptotic protease- and Murcia JM: Importance of poly(ADP-ribose) polymerase activating factor 1-dependent apoptotic pathway in human non- and its cleavage in apoptosis. Lesson from an uncleavable small cell lung cancer cells. Cancer Res 67: 709-717, 2007. mutant. J Biol Chem 273: 33533-33539, 1998. 11 Deng WG, Wu G, Ueda K, Xu K, Roth JA and Ji L: 24 Germain M, Affar EB, D’Amours D, Dixit VM, Salvesen GS Enhancement of antitumor activity of cisplatin in human lung and Poirier GG: Cleavage of automodified poly(ADP-ribose) cancer cells by tumor suppressor FUS1. Cancer Gene Ther 15: polymerase during apoptosis. Evidence for involvement of 29-39, 2008. caspase-7. J Biol Chem 274: 28379-28384, 1999. 12 Hotta T, Motoyama T and Watanabe H: Three human 25 Ivanova AV, Ivanov SV, Pascal V, Lumsden JM, Ward JM, Morris osteosarcoma cell lines exhibiting different phenotypic N, Tessarolo L, Anderson SK and Lerman MI: Autoimmunity, expressions. Acta Pathol Jpn 42: 595-603, 1992. spontaneous tumourigenesis and IL-15 insufficiency in mice with 13 Kawashima H, Ogose A, Gu W, Nishio J, Kudo N, Kondo N, Hotta a targeted disruption of the tumour suppressor gene FUS1. J T, Umezu H, Tohyama T, Nishijima H, Iwasaki H and Endo N: Pathol 211: 591-601, 2007. Establishment and characterization of a novel myxofibrosarcoma 26 Zabarovsky ER, Lerman MI and Minna JD: Tumor suppressor cell line. Cancer Gent Cytogenet 161: 28-35, 2005. genes on chromosome 3p involved in the pathogenesis of lung 14 Gu W, Ogose A, Kawashima H, Ito M, Ito T, Matsuba A, and other cancers. Oncogene 21: 6915-6935, 2002. Kitahara H, Hotta T, Tokunaga K, Hatano H, Morita T, Urakawa 27 Lee DY, Deng Z, Wang CH and Yang BB: MicroRNA-378 S, Yoshizawa T, Kawashima H, Kuwano R and Endo N: High- promotes cell survival, tumor growth, and by level expression of coxsackievirus and adenovirus receptor targeting SuFu and Fus-1 expression. Proc Natl Acad Sci USA messenger RNA in osteosarcoma, Ewing’s sarcoma, and benign 104: 20350-20355, 2007. neurogenic tumors among musculoskeletal tumors. Clin Cancer 28 Du L, Schageman JJ, Subauste MC, Saber B, Hammond SM, Res 10: 3831-3838, 2004. Prudkin L, Wistuba II, Ji L, Roth JA, Minna JD and Pertsemlidis 15 Ariizumi T, Ogose A, Kawashima H, Hotta T, Umezu H and A: miR-93, miR-98, and miR-197 regulate expression of tumor Endo N: Multinucleation followed by an acytokinetic cell suppressor gene FUS1. Mol Cancer Res 7: 1234-1243, 2009. division in myxofibrosarcoma with giant cell proliferation. J Exp Clin Cancer Res 28: 44, 2009. 16 Imaizumi S, Motoyama T, Ogose A, Hotta T and Takahashi HE: Characterization and chemosensitivity of tow human malignant Received November 17, 2010 peripheral nerve sheath tumour cell lines derived from a patient Revised December 15, 2010 with neurofibromatosis type 1. Virchows Arch 433: 435-441, 1998. Accepted December 16, 2010

21