Tohoku J. Exp. Med., 2006,Expression 209, 169-180 of p53/FXR2 Fusion in Leukemic Cells 169

Cloning and Characterization of the Novel Chimeric p53/FXR2 in the Acute Megakaryoblastic Leukemia Cell Line CMK11-5

1 1 1, 2 3 RIKA KANEZAKI, TSUTOMU TOKI, GANG XU, RAMSWAMY NARAYANAN 1 and ETSURO ITO

1Department of Pediatrics, Hirosaki University School of Medicine, Hirosaki, Japan, 2Department of Pediatrics, 2nd Affiliated Hospital of China Medical University, Shenyang, China, and 3Department of Biological Sciences, Florida Atlantic University, Boca Raton, FL, USA

KANEZAKI, R., TOKI, T., XU, G., NARAYANAN, R. and ITO, E. Cloning and Character- ization of the Novel Chimeric Gene p53/FXR2 in the Acute Megakaryoblastic Leukemia Cell Line CMK11-5. Tohoku J. Exp. Med., 2006, 209 (3), 169-180 ── The loss of p53 function is a key event in tumorigenesis. Inactivation of p53 in primary tumors and cell lines is mediated by several molecular mechanisms, including deletions and rearrange- ments. However, generation of a p53 fusion gene has not yet been reported. Here we report a novel p53/an autosomal homolog of the fragile X mental retardation (FXR2) chimeric gene generated by an interstitial deletion. Western blot analyses have shown that the p53/FXR2 protein is indeed expressed in a Down syndrome-related acute megakaryo- blastic leukemia cell line, CMK11-5 cells. To investigate the properties of the p53/FXR2 protein, we observed its subcellular localization. Flag-tagged expression vectors were transfected into COS-7 cells and the were stained with an anti-Flag antibody. The p53/FXR2 protein was expressed at high levels in the cytoplasm, whereas wild-type p53 and FXR2 were localized primarily in the nucleus and in the periphery of the nucleus, respectively. Treatment with a topoisomerase II inhibitor, VP16, failed to induce expres- sion of a p53 target gene, the cyclin-dependent kinase inhibitor p21WAF-1/CIP1, in CMK11-5 cells, and transient transfection analysis showed that the p53/FXR2 protein failed to trans- activate the p21WAF-1/CIP1 promoter. These results suggest that the p53/FXR2 fusion protein lacks the ability of wild-type p53 to function as a transcription factor. The p53/FXR2 gene is the first reported p53 fusion gene. ──── p53; FXR2; acute megakaryoblastic leuke- mia; Down syndrome; fusion protein © 2006 Tohoku University Medical Press

Received March 14, 2006; revision accepted for publication April 8, 2006. Correspondence: Etsuro Ito, M.D., Department of Pediatrics, Hirosaki University School of Medicine, Hirosaki 036-8563, Japan. e-mail: [email protected] 169 170 R. Kanezaki et al. Expression of p53/FXR2 Fusion Protein in Leukemic Cells 171

The p53 tumor suppressor gene plays a criti- Recently, acquired mutations of the GATA1 gene cal role in regulating cell proliferation, mainly have been detected in almost all cases of TMD through induction of growth arrest or apoptosis. and Down syndrome-related AMKL (DS-AMKL), p53 is a nuclear phosphoprotein induced in suggesting that other genetic changes contribute response to cellular stress such as DNA damage to the development of AMKL in TMD (Wechsler from radiation or alkylating agents. It binds DNA et al. 2002; Mundschau et al. 2003; Groet et al. in a sequence-specific manner to activate 2003; Hitzler et al. 2003; Rainis et al. 2003; Xu et transcription of a number of , including al. 2003). It was reported that p53 mutations p21WAF-1/CIP1, mouse double mitute 2 homolog might be involved in the evolution from TMD to (MDM2), and BCL2-associated X protein (BAX) AMKL, because two out of three AMKL patients (Kirsch and Kastan 1998). The loss of p53 func- harbored p53 mutations, but none of the seven tion is a key event in tumorigenesis and is associ- TMD harbored them (Malkin et al. 2000). ated with various characteristics of tumors, To understand the role of p53 mutations in including deregulation of the cell cycle, genomic the evolution of DS-AMKL, we analyzed p53 instability, and resistance to chemotherapy (Harris mutations in the DS-AMKL cell lines. Previously, 1996; Kirsch and Kastan 1998). The p53 gene is Sugimoto et al. (1992) reported the complete lack frequently mutated in solid tumors (Harris 1996; of p53 gene expression due to gross rearrange- Sakaguchi et al. 2005). p53 mutations are less ments and loss of normal p53 alleles in the DS- frequent in hematological malignancies. AMKL cell line CMK. However, in this study we Nevertheless, in these tumors a strong correlation report that a fusion gene containing p53 and an was found associating p53 mutations with unfa- autosomal homolog of the fragile X mental retar- vorable prognostic factors and resistance to che- dation gene (FXR2) (Zhang et al. 1995) was motherapy (Krug et al. 2002). expressed in CMK11-5 cells. The p53/FXR2 Inactivation of p53 in primary tumors and fusion protein lacks the ability of wild-type p53 to cell lines is mediated by several molecular mecha- function as a transcription factor. The p53/FXR2 nisms. The most frequent mechanisms by which gene is the first reportedp53 fusion gene. the wild-type p53 is inactivated in primary human tumors are point mutations. It has been recently MATERIALS AND METHODS observed that p53 gene alterations often emerge Rapid amplification of cDNA 3´ ends (3´RACE) in cell lines, even though the original tumor cells Poly A+ RNA was extracted from CMK11-5 with a expressed wild-type p53. A relapse specimen car- Quick prep micro mRNA purification kit (Amersham ried an identical mutation to that of the derived Biosciences, Piscataway, NJ, USA). A Marathon cDNA cell line. It was suggested that a p53 alteration in amplification kit (Clontech, Mountain View, CA, USA) a minor clone may confer a survival advantage to was used for the cDNA synthesis. 3´RACE was these malignant cells in vitro and presumably also performed with the following primers, p53 S1 (5´ in vivo (Drexler 2004). Another mechanism of GACACTTTGCGTTCGGGCTGGGAG 3´; for the first p53 inactivation is through deletions and rear- polymerase chair reaction [PCR]) and p53 S (5´ TCTGTCCCCCTTGCCGTCCCA 3´; for the nested rangements in the p53 gene (Sugimoto et al. PCR). 3´RACE products were ligated into pCR2.1 plas- 1992). In cell lines, p53 gene expression was mids (Invitrogen, Carlsbad, CA, USA) and their inserts completely abrogated due to gross rearrangements were sequenced with an ABI PRISM Cycle sequencing and loss of normal p53 alleles. However, genera- kit (Applied Biosystems, Foster City, CA, USA). tion of a p53 fusion gene due to DNA rearrange- ments has not yet been reported. Reverse transcriptase-polymerase chain reaction (RT- Acute megakaryoblastic leukemia (AMKL) PCR) develops in approximately 20 to 30% of Down Total RNA was isolated using the Isogen according syndrome patients with transient myeloprolifera- to the manufacturer’s recommendations (Nippon gene, tive disorder (TMD) (Zipursky et al. 1992). Tokyo). The first-strand cDNA was synthesized from 1 170 R. Kanezaki et al. Expression of p53/FXR2 Fusion Protein in Leukemic Cells 171

μ g of total RNA using M-MLV reverse transcriptase CMV2 (Sigma, St. Louis, MO, USA). Obtained clones (GIBCO BRL, Rockville, MD, USA) with random prim- were confirmed by sequencing. ers. cDNAs were amplified with Ex Taq DNA poly- For the purpose of constructing expression vectors merase (TaKaRa, Ohtsu). The study was approved by without Flag tags, the full coding regions were cut out the Ethics Committee of Hirosaki University School of from the Flag tagged expression vectors with EcoRI and Medicine. All clinical samples were obtained with XbaI, and inserted into the EcoRI-XbaI site of pcDNA3.1 informed consent. (Invitrogen), resulting in p53 pcDNA3.1, FXR2 pcD- NA3.1, and p53/FXR2 pcDNA3.1. Expression vectors PCR with genomic DNA carrying the human T24 or mutated Ha-ras (pHO6T1 and Genomic DNA was extracted from CMK11-5. pHO6N1, respectively) were described previously Amplification of the p53/FXR2 fusion gene was (Spandidos and Wilkie 1984). performed with the primers; p53S (5´- For promoter assays, the luciferase reporter plasmid PICA p21–2.3k was constructed as follows. A fragment TCTGTCCCCCTTGCCGTCCCA-3´) and FXR2 AS1 WAF-1/CIP1 (5´-CTCCCCATAGATGCGGAAAGTGCA-3´) and with containing 2.3 kb of the p21 promoter region was an LA PCR kit (TaKaRa). PCR products were sequenced amplified from CMK11-5 genomic DNA by PCR using and the p53 break point was confirmed. the primers 5´-AGGGTACCAGGAACATGCTTGGGC AGC-3´ and 5´- TGAAGCTTCCGGCTCCACAAGGA ACTGA-3´. The PCR products were digested with KpnI Southern blotting and HindIII, then subcloned into the PICA gene basic Genomic DNA samples from CMK11-5, MGS, and vector (Toyo Ink., Tokyo). normal healthy subjects were digested with BamHI or SacI. The DNA fragments were electrophoresed in a 0.8% Tris-borate-EDTA buffer (TBE) agarose gel and Cell culture transferred to Hybond-N (Amersham Biosciences, The human megakaryoblastic cell lines CMK11-5 Piscataway, NJ, USA). Hybridizations were performed (Sato et al. 1989; Adachi et al. 1991) and MGS were with the 32P-labelled entire p53 coding region or with a 0.4 established from Down syndrome-related acute myeloid kb fragment of the FXR2 gene. The FXR2 fragment was leukemia (AML) patients. CMK11-5 was obtained from prepared as follows. The amplification products of the the Institute for Fermentation (Osaka). MGS was a gift p53/FXR2 genomic DNA (see PCR with genomic DNA) from Dr. Mitsui (Yamagata University School of were digested with HhaI, and the 0.4 kb fragment con- Medicine). The human myeloblastic leukemia cell line taining FXR2 exon 8 and its upstream was extracted from ML-1 was obtained from Hayashibara Biochemical gels and purified. Laboratories (Okayama). These lines were maintained in RPMI1640 with 10% fetal bovine serum (FBS). COS-7, Construction of plasmids a green monkey renal epithelial cell line (Japanese Flag-tagged expression vectors (Flag p53, Flag Cancer Research Resources Bank, Tokyo) was grown in FXR2 and Flag p53/FXR2) were constructed as follows. Dulbecco’s modified Eagle’s medium (DMEM) contain- To amplify their entire coding regions, the PCR primers ing 10% FBS. The human osteosarcoma cell line MG-63 p53 5´-EcoRI (5´-AGGAATTCCATGGAGGAGCCGC (Japanese Cancer Research Resources Bank) was grown AGTCAG-3´), p53 3´-SalI (5´-TAGTCGACGTCAGTC in Eagle’s minimal essential medium (MEM) with non- TGAGTCAGGCCCT-3´), FXR2 5´-EcoRI (5´-TAGAAT essential amino acids and 10% FBS. NIH3T3-3-4, a TCCATGGGCGGCCTGGCCTCTG-3´), and FXR2 mouse fibroblast cell line (Riken Bio Source Center, 3´-SalI (5´-CCGTCGACTTTTATGAAACCCCATTCA Tsukuba), was maintained with DMEM 10% bovine se- C-3´) were synthesized. PCR reactions were carried out rum. All cells were cultured at 37˚C in 5% CO2. using CMK11-5 or normal human leukocyte cDNA as a template. The combinations of primers were as follows; Western blotting p53 5´-EcoRI and p53 3´-SalI were for Flag p53, FXR2 Whole cell lysates were extracted by a triple deter- 5´-EcoRI and FXR2 3´-SalI were for Flag FXR2, and gent lysis buffer, as described previously (Sambrook et p53 5´-EcoRI and FXR2 3´-SalI were for Flag p53/ al. 1989). In addition, nuclear extracts were prepared FXR2. PCR products were digested with EcoRI and from ML-1 by Nuclear and Cytoplasmic Extraction SalI, then cloned into the EcoRI-SalI site of pFLAG- Reagents (PIERCE, Rockford, IL, USA). Cell extracts 172 R. Kanezaki et al. Expression of p53/FXR2 Fusion Protein in Leukemic Cells 173 were separated on SDS-polyacrylamide gels and trans- Reporter assay ferred to Hybond-P membranes (Amersham Biosciences). MG-63 cells were seeded at a density of 4 × 104 Immunodetections were carried out with anti-p53 (clone cells/well (12 well plate). Following overnight culture, PAb1801: Zymed, Carlsbad, CA, USA), anti-FXR2 the cells were transfected using Lipofectamine reagent (Transduction Laboratories, San Jose, CA, USA) and (GIBCO BRL) with a total of 0.8 μg DNA per well: 0.3 anti-p21 (clone F-5: Santa Cruz, CA, USA) antibodies. μ g of luciferase reporter, 0.3 μ g of each expression Dilutions used were 1:2,000, 1:250, and 1:2,000, respec- vector, and 0.2 μ g of pcDNA lacZ β -galactosidase tively. The bands were visualized with anti-mouse HRP expression plasmid. Luciferase activities were measured conjugates (Santa Cruz, CA, USA) and ECL Western by the Picagene luciferase assay kit (Toyo Ink.) at 24 hrs blotting detection reagents (Amersham Biosciences). post transfection. The results were standardized with β-gal assays (Tropix, Bedford, MA, USA). Immunofluorescence staining COS-7 cells were seeded at a density of 1.5 × 105 RESULTS cells/well (in six-well plates). Following overnight cul- ture, cells were transfected with 2 μ g DNA using Isolation of cDNA for p53/FXR2 fusion tran- DMRIE-C reagent (GIBCO BRL) according to the man- scripts ufacturer’s procedure. After 24 hrs, transfected cells To find the p53 gene mutations in MGS and were fixed with a 10% formaldehyde neutral buffer solu- CMK11-5 cells, we attempted to amplify the tion (Nacalai tesque) for 15 min. In addition, they were entire coding regions by RT-PCR. In the case of permeabilized with 0.1% Triton-X 100 in phosphate- MGS cells, the amplification was successful, and buffered saline (PBS) for 10 min. They were then sequencing analysis showed one mutation causing incubated with an anti-FLAG antibody, M2 (Sigma) at an amino acid change, C176Y (data not shown). 1 μg/ml, and then anti-mouse IgG fluorescein isothiocya- However, none of the RT-PCR products were nate (FITC) (Dako) 1:20 at 37˚C for 30 min. The nuclei obtained from CMK11-5 cells. Next, we attempt- were then stained with 10 mM Hoechst 33342. ed to amplify five short fragments of p53. Only one fragment, which contained exons 1 through 4, Focus forming assay 5 was successfully amplified. We hypothesized that NIH3T3 cells were seeded at a density of 3 × 10 in CMK11-5 cells, one of the p53 gene alleles had cells/ 60 mm dish. Following overnight culture, cells been lost completely and the other was deleted were transfected with 4 μg DNA using Lipofectamine after exon 4. We then performed 3´RACE in or- reagent (GIBCO BRL). The medium was changed 24 hrs later. After an additional 24 hrs the transfected cells der to obtain the 3´ sequence of the transcript. As were diluted 1:20 and split into two aliquots that were a result, we unexpectedly found that the FXR2 added to 100 mm dishes. These cells were cultured for gene was fused in-frame to exon 4 of p53 (Fig. 14 days, changing the medium twice per week. The cells 1A). In addition, we found that codon 49 was were then fixed with 10% acetic acid and colonies were mutated from GAT to CAT, giving rise to the stained with 0.4% crystalviolet in 10% ethanol. To amino acid change D to H (data not shown). The observe the efficiency of transfection, G418 selection same alteration was previously reported in several was carried out at the same time. leukemia cases (Ahuja et al. 1991; Kawamura et al. 1995). Treatment with a DNA damaging agent To confirm expression of the p53 chimeric ML-1 and CMK11-5 cells were prepared in fresh transcript, we performed RT-PCR with the p53 5 medium at a density of 2 × 10 /ml and cultured over- sense primer (p53 S) and the FXR2 anti-sense night. The following day, cells were treated with 0-1 primer (FXR2 AS1 or FXR2AS2) using mM VP-16 (Myers Squibb, Santa Cruz, CA, USA) for CMK11-5 cDNA. A PCR product having the 24 hrs. The cells were then harvested and their proteins expected size was successfully amplified, and the were extracted (see Western blotting). p53/FXR2 mRNA product was confirmed by sequencing. The p53/FXR2 mRNA was in-frame, and thus we predicted that it encoded a protein 172 R. Kanezaki et al. Expression of p53/FXR2 Fusion Protein in Leukemic Cells 173

Fig. 1. Detection of the p53/FXR2 fusion gene in CMK11-5 cells. (A) The p53/FXR2 chimeric transcript detected by 3´RACE in CMK11-5 cells. The arrow indicates the fusion point. Amino acids derived from p53 are shown in black and those derived from FXR2 are in blue. (B) Genomic structure of the p53/FXR2 fusion gene. The sequence surrounding the fusion point is shown. Red arrows indicate the fusion point between p53 and FXR2. The sequences derived from p53 and FXR2 are shown in black and blue, respectively. (C) The p53/FXR2 gene was generated by an interstitial deletion on 17. About 80 kb of the sequences between p53 exon 5 and FXR2 intron 7 was deleted. Arrows indicate the 5´ to 3´ direction of each gene. 174 R. Kanezaki et al. Expression of p53/FXR2 Fusion Protein in Leukemic Cells 175 consisting of 578 amino acids: the N-terminal amino acids 1 to 125 of p53 followed by 453 amino acids, from 221 to 673, of the FXR2 C ter- minal region. The same fusion gene was detected in CMK86, the parent cell line of CMK11-5 (data not shown). We next examined other leukemic cells for expression of p53/FXR2. RT-PCR analysis was carried out for human hematopoietic cell lines: HL-60 (myeloid), U937 (monocytic), Raji (B-lymphoid), CEM (T-lymphoid), K562, HEL (erythro-megakaryocytic) and some patient sam- ples: eight patients with transient myeloprolifera- tive disorder (TMD), two AML patients, and three Down syndrome-related AML patients. However, we could not detect the p53/FXR2 fusion gene (data not shown).

Genomic structure of p53/FXR2 Normally, FXR2 is approximately 50 kb downstream of p53 on chromosome 17p13.1 (Fig. 1C). We confirmed the points where p53 joined Fig. 2. Southern blotting analysis demonstrating the with FXR2 by means of PCR with CMK11-5 existence of the p53/FXR2 fusion gene. genomic DNA as template, followed by sequenc- Genomic DNAs from CMK11-5, MGS, and normal healthy subjects were digested with ing. In comparing our PCR results with the BamHI or SacI. Hybridizations were per- genomic sequence of (accession formed with the 32P-labelled entire p53 coding number NT 010718), we calculated that about 80 region (lanes 1 through 6) or with a 0.4 kb kb between exon 5 of p53 and intron 7 of FXR2 fragment of the FXR2 gene (lanes 7 through was deleted. Furthermore, we found an insertion 12). CMK11-5: lanes 1, 4, 7 and 10, MGS: of the sequence AA at the fusion point (Fig. 1B). lanes 2, 5, 8, and 11: normal controls: lanes 3, 6, 9, 12. Bands having identical sizes were de- The fusion gene contains part of exon 5 from p53, tected by both the p53 and the FXR2 probes in but it is spliced out after being transcribed to CMK11-5 (lanes 1 and 7, or lanes 4 and 10, re- mRNA. spectively). These results were confirmed by Southern blot analysis using parts of the p53 and FXR2 genes as probes. Abnormal bands were detected Expression of the p53/FXR2 protein in only in CMK11-5 cells using both the FXR2 CMK11-5 cells probe as well as the p53 probe (Fig. 2). In addi- To examine whether the p53/FXR2 protein is tion, the bands detected by both the p53 and the indeed expressed in CMK11-5 cells, we next car- FXR2 probes in CMK11-5 cells migrated to the ried out a Western blot analysis. Wild-type p53 same position, indicating they had an identical was not expressed in CMK11-5 cells (Fig. 3 lane 3, size (Fig. 2 lanes 1 and 7, or lanes 4 and 10). and Fig. 5 upper panel), and FXR2 protein having These results strongly suggest there is a p53/FXR2 wild-type sizes (95 kDa) was not detected (Fig. 3 fusion gene in CMK11-5 cells. lane 7). However, bands having identical sizes (approximately 80 kDa) were detected in CMK11-5 cells using antibodies against both p53 and FXR2 (Fig. 3 lanes 3 and 7). These results 174 R. Kanezaki et al. Expression of p53/FXR2 Fusion Protein in Leukemic Cells 175

Fig. 3. Expression of the p53/FXR2 fusion protein in CMK11-5 cells. Western blot analysis was performed with an anti-p53 antibody (clone PAb1801: Zymed) or an anti- FXR2 antibody (Transduction Laboratories). Note that the epitope recognized by the p53 antibody is located in the N-terminal region (between amino acids 32-79 of p53). The FXR2 antibody recog- nizes the C-terminal amino acids 520-639 of FXR2. As controls, COS-7 cells transfected with pcDNA3.1 (Invitrogen) (lanes 1 and 5) or with the p53/FXR2 expression vector (lanes 2 and 6) were used. As additional controls, MGS cells expressing wild-type p53 and FXR2 proteins (lanes 4 and 8) were used. * Indicates non-specific band.

Fig. 4. Subcellular localization of the p53/FXR2 fusion protein. Flag-tagged expression vectors were transfected into COS-7 cells. The Flag-tagged proteins were stained with an anti-Flag antibody, M2 (Sigma), and then with anti-mouse IgG FITC (Dako). The nuclei were stained with 10 mM Hoechst 33342. The p53/FXR2 protein was extensively distributed throughout the cytoplasm, while wild-type p53 and FXR2 were mainly localized to the nucleus and the periphery of the nucleus, respectively. 176 R. Kanezaki et al. Expression of p53/FXR2 Fusion Protein in Leukemic Cells 177 clearly demonstrate that the p53/FXR2 fusion Treatment with a topoisomerase II inhibitor, protein was expressed in CMK11-5 cells. VP16, failed to induce expression of a p53 target gene, p21WAF-1/CIP1 in CMK11-5 cells Subcellular localization of p53/FXR2 protein DNA-damaging agents trigger an increase in The p53/FXR2 protein lacks the DNA bind- p53 expression leading to activation of particular ing and C-terminal domains of p53. To investi- target genes, most notably the cyclin-dependent gate the function of the p53/FXR2 protein, we kinase inhibitor gene, p21WAF-1/CIP1 (El-Deiry et al. next examined its subcellular localization. Flag- 1993, 1994). This transcriptional activation of tagged p53, FXR2, and p53/FXR2 were expressed p21WAF-1/CIP1 expression is mediated by the interac- in COS-7 cells, and the proteins were stained with tion of p53 with two response elements located in an anti-Flag antibody (M2). p53 is diffusely dis- the p21WAF-1/CIP1 promoter (El-Deiry et al. 1995). tributed in the cytoplasm of normal, unstressed To study the function of the p53/FXR2 fusion cells. However, p53 is localized in the nuclear protein, p21WAF-1/CIP1 expression was examined compartment of transformed cells (Rotter et al. after VP16 treatment of CMK11-5 cells. We used 1983). In COS-7 cells, which express the SV40 nuclear extracts from ML-1 cells and whole cell large T antigen, p53 was localized exclusively in extracts from CMK11-5 cells to detect p53, the nucleus, as previously reported (Fig. 4) because p53 was not clearly detected when we (Shaulsky et al. 1990). Wild-type FXR2 was used whole cell extracts of ML-1 cells. In ML-1 localized mainly at the periphery of the nucleus, cells, which express normal p53 (Kastan et al. forming a punctuated pattern. However, p53/ 1991), p21 was induced in coordination with p53 FXR2 was extensively distributed throughout the after VP16 treatment. However, no induction of cytoplasm, suggesting functional differences dis- p21 expression was observed in CMK11-5 cells, tinguishing the p53/FXR2 fusion protein from although the p53/FXR2 protein was expressed at p53 and FXR2. a relatively high level before VP16 treatment, and

Fig. 5. Western blot analysis of a p53 target gene, p21WAF-1/CIP1, in CMK11-5 cells after treatment with VP16. Two cell lines, p53/FXR2 expressing CMK11-5 and normal p53 expressing ML-1, were treated with various concentrations of VP16 for 24 hours. Extracts were prepared from ML-1 cells (30 μg) and CMK11-5 (50 μg), and then separated by 10% SDS-PAGE. After electrophoresis, samples were incubated with an anti-p53 antibody (clone PAb1801: Zymed) and with an anti-p21WAF-1/CIP1 antibody (clone F-5: Santa Cruz). Western blot analysis was performed using an anti-β-actin anti- body as a control for the amount of protein loaded in each lane. Note that treatment with VP16 failed to induce the expression of p21WAF-1/CIP1 in CMK11-5 cells, whereas p21WAF-1/CIP1 was induced in ML-1 cells after VP16 treatment. 176 R. Kanezaki et al. Expression of p53/FXR2 Fusion Protein in Leukemic Cells 177

promoter activity (Fig. 6). These results suggest that the p53/FXR2 protein is an inactive form of the p53 transcription factor.

p53/FXR2 failed to transform NIH3T3 cells Multiple experiments have shown that p53 mutations have gain-of-function or dominant- negative properties. To examine whether p53/ FXR2 had a dominant-acting transformation activity, focus formation assays using NIH3T3 cells were performed. The efficiencies of trans- fection did not differ among the expression plas- mids, since the number of G418 resistant colonies was almost the same for each expression vector. As expected, NIH3T3-3-4 cells were highly trans- formed by mutated Ras. However, no significant focus formation was observed after transfection with p53/FXR2 expression plasmids, as was seen with pcDNA3.1, p53, FXR2, and normal Ras Fig. 6. A p53/FXR2 expression plasmid failed to expression plasmids (Fig. 7). These results sug- induce p21WAF-1/CIP1 promoter activity. gest that the p53/FXR2 protein has no dominant- A luciferase reporter construct containing a 2.3 acting oncogenetic activity. kb fragment that included the transcription WAF-1/CIP1 initiation site for the p21 gene was DISCUSSION transiently co-transfected with a p53, FXR2, or In the present study, we describe the isola- p53/FXR2 expression vector into p53-negative MG-63 cells by Lipofectamine. The assays tion and characterization of a novel p53/FXR2 were repeated three times and representative fusion gene, which is expressed in an acute mega- data (mean ± S.D.) are shown. Each activity karyoblastic cell line CMK11-5. The p53/FXR2 was standardized by reference to the β-galacto- gene was generated by an interstitial deletion on sidase activity from the co-transfected expres- chromosome 17. The p53/FXR2 gene encodes a sion plasmid, which was driven by the CMV promoter. protein, consisting of 578 amino acids, which lacks the DNA-binding domain of p53. In trans- fected cells, the p53/FXR2 fusion protein was was induced after VP16 treatment of CMK11-5 localized mainly to the cytoplasm, whereas p53 cells (Fig. 5). accumulated in the nucleus. To our knowledge, These results suggest that the p53/FXR2 the p53/FXR2 gene is the first reported p53 fusion fusion protein lacks the ability of wild-type p53 to gene. function as a transcription factor. To test this Fragile X syndrome is caused by absence of notion directly, p53 negative MG-63 cells the fragile X mental retardation protein (FMRP). (Masuda et al. 1987) were transfected with a FMRP and its structural homologues FXR1P and luciferase reporter construct containing 2.3 kb of FXR2P form a family of RNA-binding proteins the human p21WAF-1/CIP1 promoter and with a p53, (FXR proteins) (Siomi et al. 1995; Zhang et al. FXR2, or p53/FXR2 expression vector. As 1995). The three proteins associate with polyri- expected, the p53 expression plasmid induced the bosomes as cytoplasmic mRNP particles. FXR2P, reporter gene activity driven by the p21WAF-1/CIP1 like the other members of the FXR family, con- promoter. However, p53/FXR2 as well as the tains two KH domains and an RGG box, which FXR2 expression plasmid failed to induce this are common among RNA-binding proteins, and it 178 R. Kanezaki et al. Expression of p53/FXR2 Fusion Protein in Leukemic Cells 179

Fig. 7. Focus forming assay shows p53-FXR2 fails to transform NIH3T3 cells. NIH3T3 cells were transfected with 4 μg DNA using Lipofectamine reagent (GIBCO BRL). After an additional 48 hours of culture, transfected cells were diluted 1:20, and split into two aliquots that were added to 100 mm dishes. These cells were cultured for 14 days, changing the medium twice per week (upper and middle panels, respectively). To observe the efficiency of transfection, G418 selection was carried out at the same time (lower panel). binds RNA with some degree of sequence speci- cells that express only the p53/FXR2 protein, ficity. In addition, FXR2P contains a functional although p53/FXR2 protein expression itself was nuclear localization signal (NLS), a nuclear induced after VP16 treatment. Furthermore, export signal (NES), and a nucleolar-targeting transient transfection analysis demonstrated that signal (NoS) (Eberhart et al. 1996; Tamanini et al. the p53/FXR2 protein failed to activate the 1999, 2000). Its subcellular localization may be p21WAF-1/CIP1 promoter. Together with the findings determined by the net balance of NLS signals and that the p53/FXR2 protein lacks the DNA binding NES activities. FXR2 is localized mainly in the domain and is distributed extensively throughout periphery of the nucleus, while p53/FXR2 was the cytoplasm, these results suggest that the found to be extensively distributed throughout the p53/FXR2 protein is an inactive form of the p53 cytoplasm. Because the p53/FXR2 fusion protein transcription factor. lacks an NLS, the NES activity should be pre- Although most of the effects of p53 are dominant. ascribed to its function as a transcription factor, The tumor suppressor gene p53 encodes a recent reports suggest that activation of the pro- sequence-specific transcription factor that can apoptotic Bcl-2 protein Bax by p53 occurs in a mediate many downstream effects, such as growth transcription-independent manner (Chipuk et al. arrest and apoptosis, through activation or repres- 2003, 2004). p53 directly activates Bax in the sion of its target genes (Kirsch and Kastan 1998). absence of other proteins, causing it to permeabi- The p53 protein directly transactivates a cell cycle lize mitochondria and engage the apoptotic pro- control protein, the p21WAF-1/CIP1, which is a spe- gram from the cytoplasm in the absence of p53- cific inhibitor that blocks CDK2 kinase activity. induced transcription (Chipuk et al. 2004). DNA damaging agents such as UV irradiation and Although the focus formation assay using anticancer drugs were shown to increase cellular NIH3T3 cells showed no oncogenic activity of levels of p53 (Maltzman and Czyzyk 1984; p53/FXR2, it remains to be clarified whether there Kastan et al. 1991; Fan et al. 1994). We have is a dominant negative effect of the p53/FXR2 demonstrated here that VP16 treatment did not protein on the apoptotic program in the cyto- induce the expression of p21WAF-1/CIP1 in CMK11-5 plasm. Further study will be required to deter- 178 R. Kanezaki et al. Expression of p53/FXR2 Fusion Protein in Leukemic Cells 179

mine if and how p53/FXR2 contributes to leuke- suppressor gene: clues for rational cancer therapeutic strat- egies. J. Natl. Cancer Inst., 8, 1442-1455. mogenesis. Hitzler, J.K., Cheung, J., Li, Y., Scherer, S.W. & Zipursky, A. (2003) GATA1 mutations in transient leukemia and acute Acknowledgments megakaryoblastic leukemia of Down syndrome. Blood, 101, 4301-4304. This work was supported in part by Grants-in- Kastan, M.B., Onyekwere, O., Sindransky, D., Vogelstein, B. & Aid for Scientific Research, Grants-in-Aid for Sci- Craig, R.W. (1991) Participation of p53 protein in the entific Research on Priority Areas from the Ministry cellular response to DNA damage. Cancer Res., 51, of Education, Science, Sports and Technology. 6304-6311. Kawamura, M., Kikuchi, A., Kobayashi, S., Hanada, R., Yamamoto, K., Horibe, K., Shikano, T., Ueda, K., Hayashi, References K., Sekiya, T. & Hayashi, Y. (1995) Mutations of the p53 Adachi, M., Ryo, R., Sato, T. & Yamaguchi, N. (1991) Platelet and ras genes in childhood t(1;19)-acute lymphoblastic leu- factor 4 gene expression in a human megakaryocytic leuke- kemia. Blood, 85, 2546-2552. mia cell line (CMK) and its differentiated subclone Kirsch, D. & Kastan, M. (1998) Tumor-suppressor p53: impli- (CMK11-5). Exp. Hematol., 19, 923-927. cations for tumor development and prognosis. J. Clin. Ahuja, H., Bar-Eli, M., Arlin, Z., Advani, S., Allen, S.L., Oncol., 16, 3158-3168. Goldman, J., Snyder, D., Foti, A. & Cline, M. (1991) The Krug, U., Ganser, A. & Koeffler, H.P. (2002) Tumor suppressor spectrum of molecular alterations in the evolution of genes in normal and malignant hematopoiesis. Oncogene, chronic myelocytic leukemia. J. Clin. Invest., 87, 21, 3475-3495. 2042-2047. Malkin, D., Brown, E.J. & Zipursky, A. (2000) The role of p53 Chipuk, J.E., Maurer, U., Green, D.R. & Schuler, M. (2003) in megakaryocyte differentiation and the megakaryocytic Pharmacologic activation of p53 elicits Bax-dependent leukemias of Down syndrome. Cancer Genet. Cytogenet., apoptosis in the absence of transcription. Cancer Cell, 4, 116, 1-5. 371-481. Maltzman, W. & Czyzyk, L. (1984) UV irradiation stimulates Chipuk, J.E., Kuwana, T., Bouchier-Hayes, L., Droin, N.M., levels of p53 cellular tumor antigen in nontransformed Newmeyer, D.D., Schuler, M. & Green, D.R. (2004) mouse cells. Mol. Cell. Biol., 4, 189-1594. Direct activation of Bax by p53 mediates mitochondrial Masuda, H., Miller, C., Koeffler, H.P., Battifora, H. & Cline, M.J. membrane permeabilization and apoptosis. Science, 303, (1987) Rearrangement of the p53 gene in human osteogen- 1010-1014. ic sarcomas. Proc. Natl. Acad. USA, 84, 7716-7719. Drexler, H.G. (2004) Malignant hematopoietic cell lines: in Mundschau, G., Gurbuxani, S., Gamis, A.S., Greene, M.E., vitro models for the study of myelodysplastic syndromes. Arceci, R.J. & Crispino, J.D. (2003) Mutagenesis of Leuk. Res., 24, 109-115. GATA1 is an initiating event in Down syndrome leukemo- Eberhart, D.E., Malter, H.E., Feng, Y. & Warren, T. (1996) The genesis. Blood, 101, 4298-4300. fragile X mental retardation protein is a ribonucleoprotein Rainis, L., Bercovich, D., Strehl, S., Teigler-Schlegel, A., Stark, containing both nuclear localization and nuclear export B., Trka, J., Amariglio, N., Biondi, A., Muler, I., Rechavi, signals. Hum. Mol. Genet., 5, 1083-1091. G., Kempski, H., Haas O.A. & Izraeli, S. (2003) Mutations El-Deiry, W.S., Tokino, T., Velculescu, V.E., Levy, D.B., in exon 2 of GATA1 are early events in megakaryocytic Parsons, R., Trent, J.M., Lin, D., Mercer, W.E., Kinzler, K. malignancies associated with trisomy 21. Blood, 102, & Vogelstein, B. (1993) WAF1, a potential mediator of 981-986. p53 tumor suppression. Cell, 75, 817-825. Rotter, V., Abutbul, H. & Ben-Ze’ev, A. (1983) P53 transfor- El-Deiry, W.S., Harper, J.W., O’Connor, P.M., Velculescu, V.E., mation-related protein accumulates in the nucleus of trans- Canman, C.E., Jackman, J., Pietenpol, J.A., Burrell, M., formed fibroblasts in association with the chromatin and is Hill, D.E., Wang, Y., Wiman, K.G., Mercer, W.E., Kastan, found in the cytoplasm of non-transformed fibroblasts. M.B., Kohn, K.W, Elledge, S.J., Kinzler, K.W. & EMBO J., 2, 1041-1047. Vogelstein, B. (1994) WAF1/CIP1 is induced in p53-medi- Sakaguchi, S., Yokokawa, Y., Hou, J., Zhang, X.L., Li, X.P., Li, ated G1 arrest and apoptosis. Cancer Res., 54, 1169-1174. S.S., Li, X.X., Zhu, D.C., Kamijima, M., Yamanoshita, O. El-Deiry, W.S., Tokino, T., Waldman, T., Oliner, J.D., & Nakajima, T. (2005) Environmental exposure and p53 Velculescu, V.E., Burrell, M., Hill, D.E., Healy, E., Rees, mutations in esophageal cancer patients in areas of low and J.L., Hamilton, S.R., Kinzler, K.W. & Vogelstein, B. (1995) high incidence of esophageal cancer in China. Tohoku J. Topological control of p21WAF1/CIP1 expression in nor- Exp. Med., 207, 313-324. mal and neoplastic tissues. Cancer Res., 55, 2910-2919. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989) Molecular Fan, S., El-Deiry, W.S., Bae, I., Freeman, J., Jondle, D., Bhatia, Cloning: A Laboratory Manual, 2nd ed., Cold Spring K., Fornace, A.J., Jr., Magrath, I., Kohn, K.W. & Harbor Laboratory, Cold Spring Harbor, NY. O’Connor, P.M. (1994) p53 gene mutations are associated Sato, T., Fuse, A., Eguchi, M., Hayashi, Y., Ryo, R., Adachi, M., with decreased sensitivity of human lymphoma cells to Kishimoto, Y., Teramura, M., Mizoguchi, H., Shima, Y., DNA damaging agents. Cancer Res., 54, 5824-5830. Komori, I., Sunami, S., Okimoto, Y. & Nakajima, H. (1989) Groet, J., McElwaine, S., Spinelli, M., Rinaldi, A., Burtscher, I., Establishment of a human leukaemic cell line (CMK) with Mulligan, C., Mensah, A., Cavani, S., Dagna-Bricarelli, F., megakaryocytic characteristics from a Down’s syndrome Basso, G., Cotter, F.E. & Nizetic, D. (2003) Acquired patient with acute megakaryoblastic leukaemia. Br. J. mutations in GATA1 in neonates with Down’s syndrome Haematol., 72, 184-190. with transient myeloid disorder. Lancet, 361, 1617-1620. Shaulsky, G., Goldfinger, N., Ben-Ze’ev, A. & Rotter, V. (1990) Harris, C.C. (1996) Structure and function of the p53 tumor Nuclear accumulation of p53 protein is mediated by several 180 R. Kanezaki et al.

nuclear localization signals and plays a role in tumorigene- related proteins FXR1P and FXR2P contain a functional sis. Mol. Cell. Biol., 10, 6565-6577. nucleolar-targeting signal equivalent to the HIV-1 regulato- Siomi, M.C., Siomi, H., Sauer, W.H., Srinivasan, S., Nussbaum, ry proteins. Hum. Mol. Genet., 9, 1487-1493. R.L. & Dreyfuss, G. (1995) FXR1, an autosomal homolog Wechsler, J., Greene, M., McDevitt, M.A., Anastasi, J., Karp, of the fragile X mental retardation gene. EMBO J., 14, J.E., Le Beau, M.M. & Crispino, J.D. (2002) Acquired 2401-2408. mutations in GATA1 in the megakaryoblastic leukemia of Spandidos, D.A. & Wilkie, N.M. (1984) Malignant transforma- Down syndrome. Nat. Genet., 32, 148-152. tion of early passage rodent cells by a single mutated Xu, G., Nagano, M., Kanezaki, R., Toki, T., Hayashi, Y., human oncogene. Nature, 310, 469-475. Taketani, T., Taki, T., Mitui, T., Koike, K., Kato, K., Sugimoto, K., Toyoshima, H., Sakai, R., Miyagawa, K., Imaizumi, M., Sekine, I., Ikeda, Y., Hanada, R., Sako, M., Hagiwara, K., Ishikawa, F., Takaku, F., Yazaki, Y. & Hirai, Kudo, K., Kojima, S., Ohneda, O., Yamamoto, M. & Ito, E. H. (1992) Frequent mutations in the p53 gene in human (2003) Frequent mutations in the GATA-1 gene in the myeloid leukemia cell lines. Blood, 79, 2378-2383. transient myeloproliferative disorder of Down’s syndrome. Tamanini, F., Bontekoe, C., Bakker, C.E., van Unen, L., Anar, B., Blood, 15, 2960-2968. Willemsen, R., Yoshida, M., Galjaard, H., Oostra, B.A. & Zhang, Y., O’Connor, J.P., Siomi, M.C., Srinivasan, S., Dutra, A., Hoogeveen, A.T. (1999) Different targets for the fragile Nussbaum, R.L. & Dreyfuss, G. (1995) The fragile X X-related proteins revealed by their distinct nuclear local- mental retardation syndrome protein interacts with novel izations. Hum. Mol. Genet., 8, 863-869. homologs FXR1 and FXR2. EMBO J., 14, 5358-5366. Tamanini, F., Kirkpatrick, L.L., Schonkeren, J., van Unen, L., Zipursky, A., Poon, A. & Doyle, J. (1992) Leukemia in Down Bontekoe, C., Bakker, C., Nelson, D.L., Galjaard, H., syndrome: a review. Ped. Hemat. Oncol., 9, 139-149. Oostra, B.A. & Hoogeveen, A.T. (2000) The fragile X-