We have cloned a novel TEL/protein tyrosine phosphatase receptor-type R (PTPRR) chimeric gene generated by inv(12)(p13q13). PTPRR is the first protein tyrosine phosphatase identified as a fusion partner of TEL. The chimeric gene fused exon 4 of the TEL gene with exon 7 of the PTPRR gene, and produced 10 isoforms through alternative splicing. Two isoforms that were expressed at the highest level in the leukemic cells could have been translated into COOH-terminally truncated TEL protein possessing the helix-loop-helix domain (tTEL) and TEL/PTPRR chimeric protein linking the helix-loop-helix domain of TEL to the catalytic domain of PTPRR. These two mutant proteins exerted a dominant-negative effect over transcriptional repression mediated by wild-type TEL, although they themselves did not show any transcriptional activity. Heterodimerization with wild-type TEL might be an underlying mechanism in this effect. TEL/PTPRR did not exhibit any tyrosine phosphatase activity. Importantly, overexpression of TEL/PTPRR in granulocyte macrophage colony-stimulating factor–dependent UT7/GM cells resulted in their factor-independent proliferation, whereas overexpression of tTEL did not. After cytokine depletion, phosphorylated signal transducers and activators of transcription 3 (STAT3) significantly declined in mock cells, but remained in both tTEL- and TEL/PTPRR-overexpressing cells. Loss of tumor suppressive function of wild-type TEL and maintenance of STAT3-mediated signal could at least partly contribute to the leukemogenesis caused by inv(12)(p13q13).
The 12p13 translocations are one of the most commonly observed chromosomal abnormalities in human leukemia and myelodysplastic syndrome and fuse the TEL gene on 12p13 with various partner genes. The TEL gene was originally cloned as a gene that was rearranged by t(5;12)(q33;p13) in chronic myelomonocytic leukemia, and encodes a member of the ETS family transcription factors (1). TEL shares with other ETS proteins an evolutionarily conserved ETS domain at the COOH terminus that is responsible for DNA binding to the ETS-binding consensus site (EBS; ref. 2). TEL also contains an NH2-terminal domain that is referred to as the helix-loop-helix or pointed domain. The helix-loop-helix domain in TEL has the unique property of inducing its stable homodimerization or heterodimerization with other ETS family members (3–6). Associating with the relevant corepressors mSin3A and N-CoR as well as histone deacetylase-3 (7), TEL works as a transcriptional repressor. Known target genes of TEL are FLI-1 (8), Id1 (9), stromelysin-1 (10), and Bcl-XL (11).
Gene-engineered mice have highlighted critical roles of this transcription factor in the embryonic development and hematopoietic regulation. The ablation of the TEL gene by homologous recombination causes death in uteri between E10.5 and E11.5 (12). These knock-out embryos show defect in yolk sac angiogenesis and intraembryonic apoptosis of mesenchymal and neural cells, although they present normal yolk sac hematopoiesis. Generating chimeric mice with TEL−/− embryonic stem cells, the pivotal function of TEL in establishing hematopoiesis of all lineages in neonatal bone marrow has been clarified, whereas TEL−/− embryonic stem cells contributed to both primary and definitive fetal hematopoiesis (13). Moreover, a recent study shows that inactivation of TEL in adult mice leads to decrease of hematopoietic stem cells in bone marrow (14). On the other hand, TEL is believed to function as a tumor suppressor because its overexpression in NIH3T3 fibroblasts results in reduced cell growth in liquid and soft agar cultures (10, 15).
Molecular dissecting of the TEL-related chimeric genes has provided interesting clues to the pathogenesis of 12p13 translocation-type leukemia. In some translocations, receptor-type and non-receptor-type tyrosine kinases are fused to the NH2-terminal portion of TEL and are thus catalytically activated by homodimerization through the helix-loop-helix domain in the TEL moiety. Examples for the former include platelet-derived growth factor receptor β in t(5;12)(q33;p13) (refs. 1, 16, 17) and tyrosine kinase C in t(12;15)(p13;q25) (ref. 18), and those for the latter Abl in t(9;12)(q34;p13) (refs. 19, 20), Janus-activated kinase (JAK)-2 in t(9;12)(p24;p13) (refs. 21, 22), Syk in t(9;12)(q22;p13) (ref. 23), and Abl-related gene in t(1;12)(q25;p13) (ref. 24). In other translocations, transcription factors are structurally and functionally modified by fusing with the NH2- or COOH-terminal part of TEL. Examples include acute myelogenous leukemia (AML)-1 in t(12;21)(p13;q22) (refs. 25–29), MN1 in t(12;22)(p13;q11) (refs. 30, 31), Evi-1 in t(3;12)(q26;p13) (ref. 32), PAX5 in t(9;12)(q11;p13) (ref. 33), and CDX2 in t(12;13)(p13;q12) (ref. 34). Thus, perturbation of original functions of the partner genes could be a mechanism in causing leukemia in patients with such translocations. Furthermore, disruption of tumor-suppressive function of wild-type TEL itself seems to be another cause of leukemogenesis because some chimeric molecules such as TEL/AML1 are shown to dominantly interfere with function of wild-type TEL (35).
To obtain a new insight into the molecular mechanism in leukemogenesis by the 12p13 translocations, we cloned several species of novel chimeric cDNAs generated by inv(12)(p13q13) found in a patient with acute myelogenous leukemia [M2 according to the French-American-British (FAB) classification]. These cDNAs contained the NH2-terminal TEL sequence followed by the COOH-terminal sequence from protein tyrosine phosphatase receptor-type R (PTPRR) and were expected to produce either truncated TEL or chimeric TEL/PTPRR protein. Both molecules lost DNA binding to and trans-repression through EBS, but blocked the molecular function of wild-type TEL probably by heterodimerizing with it. TEL/PTPRR showed no tyrosine phosphatase activity. Notably, overexpression of TEL/PTPRR in factor-dependent human leukemia cell line UT7/GM led to factor-independent growth, suggesting the oncogenic potential of this chimeric molecule.
Materials and Methods
Case presentation. A 24-year-old woman presented with slight fever in April 2001. Her hemoglobin level was 9.1 g/dL, WBC count 4,200/μL with 29% blasts, and platelet count 191,000/μL. The bone marrow aspirate disclosed 79% blasts positive for myeloperoxidase, and a diagnosis of acute myelogenous leukemia (M2 according to the FAB criteria) was made. Flow cytometric assay revealed that the blasts were positive for CD13, CD15, CD33, and CD34. Cytogenetic analysis showed 46, XX, inv(12)(p13q13) (12) /46, XX (8). The patient achieved complete remission after one course of induction chemotherapy and underwent bone marrow transplantation in the first remission from her HLA-matched sibling in November 2001. She has thus far been free from relapse. All the following experiments were done under the written informed consent of the patient.
Fluorescence in situ hybridization analysis. The metaphase samples that were subjected to conventional cytogenetic studies were also applied to fluorescence in situ hybridization (FISH) analysis. LL12NCO1 cosmid probes (2G8, 163E7, and 184C4) which are located within the TEL gene were used in the assay (36). The probes were labeled with biotin-11-dUTP or digoxigenin-11-dUTP using PCR labeling after sequence-independent amplification, and were hybridized to metaphase samples as previously described (37, 38). The hybridization was detected with avium fluorescein or anti–digoxigenin rhodamine and the metaphase cells were subsequently counterstained with 4′,6-diamidino-2-phenylindole dihydrochloride. Images of the hybridized signals were captured under fluorescence microscopy.
Cloning of TEL/PTPRR. Total RNA was extracted from cryopreserved leukemic cells with inv(12)(p13q13) using RNeasy RNA miniprep system (Qiagen, Valencia, CA). We purified mRNA by oligo-dT column. The first-strand cDNA was synthesized from 2 μg of RNA using the Superscript first-strand synthesis system (Invitrogen, Carlsbad, CA) with random hexamers and Moloney murine leukemia virus reverse transcriptase. For 3′-rapid amplification of cDNA ends (RACE) procedure, we adopted Marathon cDNA amplification system (Clontech, Palo Alto, CA). Synthesis of the second-strand cDNA and adaptor ligation were carried out according to the instructions of the manufacturer. RACE-PCR was done for 40 cycles with primers TELf2 and activator protein (AP)-1, followed by nested PCR for 30 cycles with primers TELf4 and AP-2 (refer to the instructions of the manufacturer for the sequences of primers AP-1 and AP-2). PCR products were subcloned into the pCR2.1-TOPO cloning vector (Invitrogen) and nucleotide sequences were determined by ABI Prism 3100 Genetic Analyzer (Applied Biosystems, Foster City, CA).
PCR amplification. To amplify TEL/PTPRR, PTPRR/TEL, wild-type TEL, and wild-type PTPRR cDNAs, we used primer sets TELf2 and PTPRRr7a, PTPRRf6 and TELr5, TELf2 and TELr5, and PTPRRf6 and PTPRRr7a, respectively. PCR was done for 40 cycles of 94°C for 30 seconds, 55°C for 30 seconds, and 72°C for 1 minute. Extension time was elongated to 2 minutes when we amplified full-length TEL/PTPRR cDNA with a set of primers TELf1 and PTPRRr14. Sequences of all PCR primers are given in the next section.
Oligonucleotides. TELf1 [nucleotide (nt) 25-47 of TEL]: 5′-ATGTCTGAGACTCCTGCTCAGTG-3′; TELf2 (nt 88-111 of TEL): 5′-AGCCCAGTGCCGAGTTACGCTTCC-3′; TELf3 (nt 328-347 of TEL): 5′-TTTCGCTATCGATCTCCTCA-3′; TELf4 (nt 376-405 of TEL): 5′-CAGCATATTCTGAAGCAGAGGAAACCTCGG-3′; TELr5 (complement, nt 637-660 of TEL): 5′-GAGGCGGCGGATCATGTTGTCCAG-3′; PTPRRf6 (nt 971-992 of PTPRR): 5′-ACCAGGAGATCCACCTATCGCC-3′; PTPRRr7a (complement, nt 1,313-1,334 of PTPRR): 5′-ACGACGTCCCTCAGCTGAGACC-3′; PTPRRr7b (complement, nt 1,193-1,216 of PTPRR): 5′-TCATGTCCAATGTAAGAGATACGT-3′; PTPRRf10 (nt 1,565-1,580 of PTPRR): 5′-CCACGCAGGGCCCCAT-3′; PTPRRr10 (complement, nt 1,565-1,584 of PTPRR): 5′-GATCATGGGGCCCTGCGTGG-3′; PTPRRr14 (complement, nt 2,126-2,148 of PTPRR): 5′-TCACTGGACAGTCTCTGCTGAAA-3′.
Plasmid construction. Constructions of pME18S-HA-TEL, pME18S-FLAG-TEL, pcDNA3-HA-TEL, (EBS)3tkLuc, and pSRαMSVtkneo-FLAG-wild-type-TEL were previously described (39, 40). Hemagglutinin or FLAG tag was inserted at both ends of TEL/PTPRR chimeric cDNAs by the PCR amplification method. The resultant cDNAs were subcloned into the EcoRI sites of pME18S, pcDNA3, and pCXN2 expression plasmids in the sense orientation. To obtain hemagglutinin-tagged wild-type PTPRR cDNA, a 5′ fragment spanning the initiation codon (nt 175 according to GenBank accession number U42361) and the ApaI site (nt 1,567) was amplified from a human brain total cDNA library (Clontech). Together with a 3′ ApaI (nt 1,567 in PTPRR cDNA)/XhoI (nt 355 in pCR2.1-TOPO vector) fragment derived from pCR2.1-TOPO-TEL/PTPRR-HA, it was subcloned into the XhoI site of pME18S. FLAG-tagged human SUMO-1 cDNA was obtained by the PCR amplification method, and was subsequently cloned into the XbaI site of pME18S.
Cell culture. COS-7 cells were cultured in DMEM (Sigma, St. Louis, MO) containing 10% FCS and transfected with various expression plasmids by the DEAE-Dextran method. NIH3T3 fibroblasts were grown in DMEM with 10% bovine serum. HeLa cells were cultured in Eagle's MEM (Sigma) supplemented with 10% FCS and 1% nonessential amino acids. Human leukemia cells UT7/GM were maintained in Iscove's modified Dulbecco's medium (Invitrogen) containing 10% FCS and 1 ng/mL of human recombinant granulocyte macrophage-colony stimulating factor (GM-CSF; Kirin, Gunma, Japan). KASUMI-1 cells were cultured in RPMI 1640 (Sigma) with 10% FCS.
Establishing bulk transfectants of NIH3T3 cells. NIH3T3 bulk population constitutively overexpressing FLAG-tagged wild-type TEL and its control were obtained by retroviral infection and G418 (Sigma) resistance as previously described (41).
Generating bulk and stable transfectants of UT7/GM cells. To obtain bulk transfectants for experiments, 1 × 107 of UT7/GM cells suspended in 500 μL PBS were electroporated at 380 V and 975 μF with pCXN2 plasmid carrying the aimed gene. Forty-eight hours after the electroporation, cells were selected with 0.8 mg/mL of G418. Stable UT7/GM clones overexpressing TEL/PTPRR were established by the limiting dilution method. For factor deprivation, UT7/GM cells were washed with PBS at least thrice and were finally suspended in factor-free media.
Metabolic labeling. Forty-eight hours after transfection, COS-7 cells were washed with phosphate-free DMEM and then cultured for 2 hours in DMEM supplemented with 400 μCi [32P]orthophosphate/mL (Amersham Biosciences, Piscataway, NJ) and 10% dialyzed FCS.
Subcellular localization. NIH3T3 cells were transiently transfected with expression plasmids using TransFast (Promega). Forty-eight hours later, cells were suspended in hypotonic suspension buffer [10 mmol/L sodium phosphate (pH 7.0), 5 mmol/L EDTA, 1 mmol/L sodium orthovanadate, 1 mmol/L DTT, and 1 mmol/L phenylmethylsulfonyl fluoride], and were separated into nucleic and cytoplasmic fractions using Dounce homogenizer (Wheaton, Millville, NJ). Equal volumes of aliquots were applied to Western blot analysis.
Western blot analysis and immunoprecipitation. Western blot analysis and immunoprecipitation were done as previously described (42). Anti-FLAG (M2) and anti-hemagglutinin (CH-7 or rabbit) antibodies were purchased from Sigma, and antibodies against TEL (N-19 and C-20), histone H1, actin, Bcl-XL, signal transducers and activators of transcription (STAT)-3, and phospho-STAT3 were from Santa Cruz Biotechnology (Santa Cruz, CA).
Electrophoretic mobility shift assay. The procedures for electrophoretic mobility shift assays (EMSA) were previously described (43). Lysates were in vitro prepared by TNT-Coupled Wheat Germ Extract System (Promega). The EBS oligonucleotide used as a probe or a specific competitor and its mutant used as a nonspecific competitor were previously described (2). End labeling of the double-stranded oligonucleotide was carried out with [α-32P]dCTP (Amersham Biosciences) and Klenow enzyme (Takara, Shiga, Japan) at room temperature for 30 minutes. Unincorporated nucleotides were removed by G-50 Sephadex columns (Amersham Biosciences). Protein-DNA complexes were separated on a 4% polyacrylamide gel and visualized by autoradiography. In competition studies, a 300-fold molar excess of unlabeled oligonucleotide was added to the reaction. In supershift assays, we employed anti-TEL (N-19) supershift antibody of which epitope is encoded by NH2-terminal TEL sequence in tTEL and TEL/PTPRR cDNAs. The antibody was preincubated with the lysates at 4°C for 1 hour.
Luciferase assay. HeLa cells in 24-well tissue culture plates were transfected with 1 μg of (EBS)3tkLuc or (mEBS)3tkLuc along with 1 μg of expression plasmids by using Tfx-20 (Promega). To equalize transfection efficiencies, total amounts of expression plasmids were kept constant in terms of weight by adding empty pME18S vector. Luciferase assays were done with Dual-Luciferase reporter assay system (Promega) as previously described (39, 42).
In vitro phosphatase assay. Lysates of COS-7 cells overexpressing hemagglutinin-tagged wild-type PTPRR, tTEL, or TEL/PTPRR were immunoprecipitated with anti-hemagglutinin (CH-7) antibody conjugated with Sepharose A beads (Amersham Biosciences). The immunoprecipitates were washed and finally suspended in 150 μL of assay buffer [20 mmol/L Tris-HCl (pH 7.4), 150 mmol/L NaCl, 1 mmol/L EDTA, 1 mmol/L DTT, and 1 mmol/L phenylmethylsulfonyl fluoride]. One third of the mixture (50 μL) was added to 100 μL of assay buffer containing 15 mmol/L p-nitrophenyl phosphate (Sigma-Aldrich), followed by incubation at 37°C for 10 minutes. The reaction was stopped by adding 25 μL of 2.5 N NaOH. The absorption at 405 nm was read on a 96-well microtiter plate.
Cloning of the TEL/PTPRR chimeric gene. Because the TEL gene on 12p13 is fused to various partners in a variety of hematologic malignancies, we suspected that inv(12)(p13q13) implicated it. To look whether the TEL gene is rearranged by the chromosomal abnormality, we first did FISH analysis using some cosmid probes spanning the TEL gene (Fig. 1A; ref. 36). Split signals were observed in bone marrow metaphases with inv(12)(p13q13) when we used cosmid 184C4 (Fig. 1B) or 163E7 (data not shown) as a probe. Because the procedure with cosmid 2G8 did not yield split signals (data not shown), it was likely that the inversion breakage occurred between introns 5 and 6 of the TEL gene. Subsequently, we adopted the 3′-RACE method to identify a fusion partner of the TEL gene. RACE-PCR products shown in Fig. 1C were subcloned into the pCR2.1-TOPO vector and resultant five clones were sequenced. Among them, four clones contained the wild-type TEL sequence only, but the last one included TEL exon 4–derived sequence followed by an unknown sequence. BLAST database searching revealed that the chimeric cDNA joined TEL exon 4 and PTPRR exon 7 with frameshift (human TEL was referred to GenBank accession number U11732 and human PTPRR to U42361; Fig. 2A).
Then, we did reverse transcription-PCR (RT-PCR) analysis to examine whether TEL/PTPRR, PTPRR/TEL, wild-type TEL, and wild-type PTPRR transcripts were expressed in the leukemic cells with inv(12)(p13q13). To detect each transcript, we used primer sets TELf2 (in TEL exon 2) and PTPRRr7a (in PTPRR exon 7) for TEL/PTPRR, PTPRRf6 (in PTPRR exon 6) and TELr5 (in TEL exon 5) for PTPRR/TEL, TELf2 and TELr5 for wild-type TEL, and PTPRRf6 and PTPRRr7a for wild-type PTPRR. As for TEL/PTPRR, two smaller bands (416 and 251 bp) besides a product of the expected size (551 bp) were observed (Fig. 1D). Sequencing revealed that these smaller cDNAs lacked exon 4 or exons 3 and 4 of the TEL gene, strongly suggesting the presence of alternative splicing mechanisms. On the other hand, reciprocal PTPRR/TEL mRNA was not detected in spite of our efforts with several sets of primers including the one described above. Thus, we conclude that TEL/PTPRR is expressed in the leukemic cells of this patient and might therefore contribute to leukemogenesis. Wild-type TEL mRNA probably derived from the intact TEL allele was expressed in the leukemic cells, whereas wild-type PTPRR mRNA was not.
Alternative splicing leads to generation of 10 TEL/PTPRR chimeric cDNAs. Because the TEL gene was fused out-of-frame to the PTPRR gene, the resultant full-length TEL/PTPRR cDNA (type 1) represented an open reading frame encoding exons 1 to 4 of the TEL gene (154 amino acids) with additional 11 amino acids (Fig. 2B and C). This isoform expresses truncated TEL including the intact helix-loop-helix domain, but lacks the COOH-terminal ETS domain of TEL and any functional domains of PTPRR. We thus refer to it as “truncated TEL (tTEL)” in the following sections. To seek for other TEL/PTPRR isoforms in the inv(12)(p13q13)-carrying leukemic cells, we further did RT-PCR analysis with other combinations of primers. When we used a set of primers TELf1 (in TEL exon 1) and PTPRRr7b (in PTPRR exon 7), various TEL/PTPRR cDNAs of smaller sizes were amplified as well as a full-length cDNA (data not shown). Sequencing analysis showed that exon-skipping mechanisms in the TEL gene produced seven isoforms (types 2-8). All these isoforms were also out-of-frame and should express only the NH2-terminal portion of TEL in which a part or the entire of the helix-loop-helix domain is spliced out. Moreover, using another set of primers TELf3 (in TEL exon 3) and PTPRRr10 (in PTPRR exon 10), we identified two other isoforms that lacked exon 7 (type 9) or exons 7 and 8 (TEL/PTPRR) of the PTPRR gene. Although type 9 isoform again contained an out-of-frame junction, TEL/PTPRR was the unique in-frame isoform with an open reading frame of 1,158 nucleotides coding for 385 amino acid residues that linked the helix-loop-helix domain of TEL and almost the entire protein tyrosine phosphatase domain of PTPRR (Fig. 2B and C). RT-PCR with a combination of primers PTPRRf10 (in PTPRR exon 10) and PTPRRr14 (in PTPRR exon 14) amplified only one kind of cDNA that contained exons 10 to 14 of the PTPRR gene without any deletions (data not shown), indicating that alternative splicing did not occur in this region. To examine which of these 10 isoforms were dominantly expressed in the leukemic cells, we then made PCR amplification with a set of primers TELf1 and PTPRRr14. Interestingly, two major bands which turned out to be derived from tTEL and TEL/PTPRR by sequencing analysis were observed (Fig. 2D). Thus, we decided to investigate molecular and biological functions of these two isoforms in the following experiments to establish the underlying mechanisms in inv(12)-type leukemia.
Truncated TEL and TEL/PTPRR affect nuclear localization of wild-type TEL. We first induced wild-type TEL, tTEL, or TEL/PTPRR expression in COS-7 cells by transfecting the corresponding cDNAs into them. As previously reported (42), slow-migrating bands were detected in addition to that of the expected size, when wild-type TEL was expressed (Fig. 3A,, left). tTEL and TEL/PTPRR proteins also showed similar slow-migrating bands. When these proteins were metabolically labeled with [32P]orthophosphate, all these size-shifted bands for wild-type TEL, tTEL, or TEL/PTPRR turned out to be hyperphosphorylated forms (Fig. 3A , right). The lowest band of tTEL was likely to be derived from a degradation product. Given that Ser22 in wild-type TEL is a constitutive phosphorylation site (42), these aberrant TEL proteins could have been also phosphorylated at least on the same residue and showed larger-sized bands.
To get some insights into molecular functions of the aberrant TEL proteins, we next examined subcellular localization of tTEL and TEL/PTPRR by Western blot analysis with fractionated lysates overexpressing each protein. We confirmed that specific marker for nuclear or cytoplasmic fraction, histone H1 or actin, was exclusively located in the corresponding fraction (Fig. 3B,, left). Because both tTEL and TEL/PTPRR lack ETS DNA-binding domain of TEL containing nuclear localization signal, and ETS-lacking mutant or isoforms of wild-type TEL have been reported to reside in the cytoplasm (41, 44), it is quite plausible that they show different distribution patterns from that of wild-type TEL. When wild-type TEL cDNA was transiently transfected to NIH3T3 cells, overexpressed wild-type TEL protein was predominantly localized in the nucleus. On the other hand, overexpressed tTEL protein was exclusively distributed in the cytoplasm and overexpressed TEL/PTPRR chimeric protein also chiefly resided in the cytoplasm. Surprisingly, when tTEL or TEL/PTPRR was coexpressed with wild-type TEL, all these proteins were almost equally expressed in both fractions. These data suggest that both tTEL and TEL/PTPRR prevent de novo translated wild-type TEL from entering the nucleus, whereas wild-type TEL draws tTEL and TEL/PTPRR into the nucleus. We further compared subcellular localization of endogenous wild-type TEL between the inv(12)-carrying M2 leukemic cells and t(8;21)-carrying M2 KASUMI-1 cells. Interestingly, whereas endogenous TEL proteins were predominantly distributed to the nucleus in KASUMI-1 cells, they were exclusively cytoplasmic in the inv(12)-carrying cells (Fig. 3B , right).
It has been reported that wild-type TEL is sumoylated on Lys99 and that sumoylated wild-type TEL is a target of CRM1-mediated nuclear export (44, 45). Because tTEL and TEL/PTPRR encompassed the acceptor site and mainly resided in the cytoplasm, we examined whether they were modified by SUMO-1. COS-7 cells were transfected with hemagglutinin-tagged wild-type TEL, tTEL, or TEL/PTPRR expression plasmid with or without FLAG-tagged SUMO-1 plasmid. As reported on wild-type TEL, bands corresponding to SUMO-1–modified proteins (wild-type TEL, 75 and 85 kDa; tTEL, 45 kDa; TEL/PTPRR, 65 kDa) were detected in Western blot analysis with anti-hemagglutinin antibody only when SUMO-1 plasmid was cotransfected (Fig. 3C,, left). To confirm that these larger-sized bands were indeed derived from SUMO-1–modified proteins, the total cell lysates were immunoprecipitated with anti-hemagglutinin antibody. The precipitated proteins were visualized with anti-FLAG antibody at the expected sizes of each SUMO-1–conjugated protein (Fig. 3C , right). Therefore, we conclude that both tTEL and TEL/PTPRR are subjected to modification with SUMO-1, which may support their cytoplasmic localization. As an additional remark, several proteins of higher molecular weights than sumoylated tTEL were concomitantly immunoprecipitated with anti-hemagglutinin antibody in the presence of exogenous SUMO-1. Because they were not detected with anti-hemagglutinin antibody but with anti-FLAG antibody, a likely scenario is that some other SUMO-1–modified proteins interacting with tTEL were coimmunoprecipitated.
Truncated TEL and TEL/PTPRR heterodimerize with wild-type TEL. Because tTEL and TEL/PTPRR contain the entire helix-loop-helix domain of TEL, they conceivably heterodimerize with wild-type TEL. To test this hypothesis, we transiently transfected into COS-7 cells hemagglutinin-tagged tTEL or TEL/PTPRR expression plasmid, alone or along with FLAG-tagged wild-type TEL expression plasmid. Cell lysates were subjected to immunoprecipitation with anti-FLAG antibody followed by Western blot analysis with anti-hemagglutinin antibody. As shown in Fig. 3D, both tTEL and TEL/PTPRR were coimmunoprecipitated with anti-FLAG antibody when they were coexpressed with FLAG-tagged wild-type TEL. This coimmunoprecipitation completely disappeared when anti-FLAG antibody was removed from the reaction or when FLAG-tagged wild-type TEL was not coexpressed in COS-7 cells. Moreover, when we applied anti-hemagglutinin antibody for immunoprecipitation, FLAG-tagged wild-type TEL was coimmunoprecipitated with hemagglutinin-tagged tTEL or TEL/PTPRR (data not shown). These data indicate that these aberrant proteins heterodimerize with wild-type TEL in vivo.
TEL/PTPRR prevents wild-type TEL from binding to ETS-binding consensus site. Because the breakpoint occurs between exons 4 and 5 of the TEL gene, neither tTEL nor TEL/PTPRR encodes the ETS domain of TEL. It is therefore quite likely that these isoforms do not possess EBS-specific DNA-binding property as wild-type TEL does. To clarify this point, wheat germ extracts expressing wild-type TEL, tTEL, and TEL/PTPRR at comparable levels (Fig. 4A) were applied to EMSA in which radioactive EBS oligonucleotide was used as a probe. As previously observed (2, 40), wild-type TEL formed a specific DNA-protein complex that emerged as a somewhat broad band possibly due to weak binding (Fig. 4B , left). This band represented a specific association of wild-type TEL and the EBS probe because it was completely canceled by cold specific competitor but not by nonspecific competitor. Furthermore, this band was supershifted when anti-TEL antibody was preincubated with the lysate. In contrast, neither tTEL nor TEL/PTPRR formed a specific DNA-protein compound. Taken together, we conclude that both isoforms lose EBS-specific DNA-binding capacity.
We next examined whether tTEL and TEL/PTPRR alter the capacity of wild-type TEL to bind with EBS. Whereas mixture of wild-type TEL and tTEL did not change the intensity of the band derived from possible EBS-TEL complex, mixture of wild-type TEL and TEL/PTPRR obviously diminished it (Fig. 4B , right). We confirmed that the shifted bands contained wild-type TEL protein by the supershift experiment with anti–COOH-terminal TEL antibody (data not shown). This result indicates that TEL/PTPRR prevents wild-type TEL molecule from binding to EBS, possibly by heterodimerizing with it.
Truncated TEL and TEL/PTPRR themselves do not repress transcription through ETS-binding consensus site, but confer a dominant-negative effect over wild-type TEL. TEL is a member of ETS family transcription factors and represses transcription of target genes through EBS. Because both tTEL and TEL/PTPRR lack the ETS domain and thus do not bind to EBS, they could lose activity as a transcription factor. We thus did luciferase reporter assays with (EBS)3tkLuc and examined the effects of tTEL and TEL/PTPRR on transcription through EBS in HeLa cells. When increasing amounts of wild-type TEL expression plasmid were cotransfected with (EBS)3tkLuc, the luciferase activities were repressed in a dose-dependent manner (Fig. 4C , top). This transcriptional regulation was EBS specific because no repression was observed when (EBS)3tkLuc was substituted with a mutant reporter plasmid, (mEBS)3tkLuc (data not shown). On the other hand, transfection of tTEL or TEL/PTPRR expression plasmid along with (EBS)3tkLuc yielded no significant changes in luciferase activities even at the highest dose.
Because they were found to associate with wild-type TEL, we then speculated that tTEL and TEL/PTPRR could affect the functions of wild-type TEL by heterodimerizing with it. Thus, increasing amounts of tTEL or TEL/PTPRR expression plasmid were transfected with (EBS)3tkLuc and wild-type TEL expression plasmid. Interestingly, this led to a decrease in fold repression induced by wild-type TEL (Fig. 4C , bottom). These data suggest that both tTEL and TEL/PTPRR exert a dominant-negative effect on wild-typeTEL–mediated transcriptional repression and that tumor-suppressive functions of wild-type TEL may be abolished in the leukemic cells.
To confirm the dominant-negative effect of tTEL and TEL/PTPRR over wild-type TEL in vivo, we evaluated by Western blot analysis the endogenous expression of Bcl-XL, a well-known transcriptional target of TEL. For this purpose, NIH3T3 cells constitutively expressing FLAG-tagged wild-type TEL were transiently transfected with tTEL or TEL/PTPRR expression plasmid, and were subjected to Western blot analysis with anti–Bcl-XL antibody. As previously reported (11), expression of endogenous Bcl-XL was found repressed by wild-type TEL (Fig. 4D). Importantly, coexpression of tTEL and TEL/PTPRR recovered suppressed Bcl-XL expression in the wild-type TEL–expressing cells.
TEL/PTPRR lacks protein tyrosine phosphatase activity. PTPRR, a human homologue of PTPSL/PTPBr7 in mouse and PC12-PTP1/PCPTP1 in rat, is a receptor-type tyrosine phosphatase that contains one cytoplasmic protein tyrosine phosphatase domain at its COOH terminus. Because TEL/PTPRR possesses almost full length of the protein tyrosine phosphatase domain, we analyzed if it retained catalytic activity by in vitro phosphatase assay using p-nitrophenyl phosphate as a substrate. Lysates from COS-7 cells overexpressing hemagglutinin-tagged wild-type PTPRR, tTEL, or TEL/PTPRR were immunoprecipitated with anti-hemagglutinin antibody. To confirm that each protein was successfully collected, a portion of immunoprecipitates and total cell lysates was subjected to Western blot analysis (Fig. 5A). All these proteins were gathered at comparable levels. Wild-type PTPRR showed catalytic activity as a protein tyrosine phosphatase (Fig. 5B). In contrast, tTEL completely lacking the protein tyrosine phosphatase domain was catalytically inactive, as expected. Surprisingly enough, TEL/PTPRR also lost its activity, probably because it did not express the 13 NH2-terminal amino acid residues of the protein tyrosine phosphatase domain. We speculate that phosphatase activity is not a requisite for the development of inv(12)-carrying leukemia.
TEL/PTPRR endows UT7/GM cells with factor independence. To search for transforming activity of tTEL and TEL/PTPRR, we employed the human megakaryocytic leukemia cell line UT7/GM that requires GM-CSF for proliferation and survival, and analyzed the effects of overexpressed tTEL or TEL/PTPRR on the factor-dependent growth. By using electroporation, we introduced tTEL or TEL/PTPRR expression plasmid into UT7/GM cells. As indicated in Fig. 6A, Western blot analysis with anti-hemagglutinin antibody showed the comparable expression of tTEL and TEL/PTPRR proteins in the bulk populations and we used these uncloned cells in the following experiments. When cultured under the presence of GM-CSF, both tTEL- and TEL/PTPRR-overexpressing cells proliferated slightly faster than mock cells (Fig. 6B). Mock cells expectedly starved to death after depletion of GM-CSF from the media. The tTEL-overexpressing cells did not survive after cytokine removal either. In contrast, the TEL/PTPRR-overexpressing cells continued to proliferate slowly in factor-free media at least beyond day 60 of culture. We also established by limiting dilution three stable lines overexpressing TEL/PTPRR (T/P-4, T/P-5, and T/P-7), and confirmed the factor independence beyond 2 months of culture. The calculated doubling times at day 28 were 45, 34, and 76 hours, respectively. These observations suggest that the GM-CSF–independent survival is not due to selection and expansion of mutated clones, but results from the expression of TEL/PTPRR itself.
We hypothesized that some growth signalings were constitutively activated in the TEL/PTPRR-overexpressing cells, leading to the factor-independent proliferation. To identify the exaggerated signals in these cells, we next compared the phosphorylation levels of molecules in various signal transduction systems including mitogen-activated protein (MAP) kinase signals, JAK/STAT signals, and phosphatidylinositol 3-kinase signals among mock, tTEL-, and TEL/PTPRR-overexpressing cells. We observed that c-jun-NH2-kinase (JNK) in the MAP kinase signalings, STAT1 and STAT5 in the JAK/STAT signalings, and Akt1/2/3 in the phosphatidylinositol 3-kinase signalings were almost equally phosphorylated among them and their phosphorylation levels remained unchanged before and after the cytokine removal (data not shown). Extracellular signal-regulated kinase (ERK)-1/2 in the MAP kinase signalings alone became almost thoroughly dephosphorylated within 12 hours after the factor deprivation in all the bulk populations. On the other hand, in spite of the steady expression of STAT3 proteins, phosphorylated STAT3 rapidly declined after GM-CSF withdrawal in the mock cells (Fig. 6C). Notably, however, the levels of phosphorylated STAT3 were unaltered in the tTEL- and TEL/PTPRR-overexpressing cells. These data suggest that overexpressed mutant TEL proteins, tTEL and TEL/PTPRR, might stimulate the JAK/STAT signal via STAT3, but that TEL/PTPRR might activate other pathways to render UT7/GM cells factor independent.
We have shown in this study that inv(12)(p13q13) results in the formation of the TEL/PTPRR chimeric gene, fusing exon 4 in the TEL gene with exon 7 in the PTPRR gene. Ten isoforms of the TEL/PTPRR cDNAs that are generated possibly through alternative splicing have been cloned in RT-PCR assay. Among the isoforms, only one isoform encodes a chimeric product (TEL/PTPRR) connecting the helix-loop-helix domain of TEL and the protein tyrosine phosphatase domain of PTPRR, and the rest express COOH-terminally truncated TEL due to frameshift. As judged from higher levels of amplification in RT-PCR analysis, two isoforms, tTEL (type 1) and TEL/PTPRR, are likely to be dominantly expressed in the leukemic cells. Generation of truncated form of TEL from the TEL-related chimeric allele is also reported in the TEL/JAK2 chimeric gene (22). Importantly, whereas wild-type TEL mRNA is detected in the leukemic cells, neither reciprocal PTPRR/TEL nor wild-type PTPRR mRNA is detected in the sample. This suggests that promoter of the PTPRR gene is inactivated in these leukemic cells. These data indicate that tTEL and TEL/PTPRR are implicated in the pathogenesis of inv(12)-type leukemia.
To get some insight into the underpinning mechanisms in leukemogenesis by inv(12)(p13q13), we evaluated molecular functions of tTEL and TEL/PTPRR. Because TEL is a tumor suppressor, functional effects of the mutant proteins on wild-type TEL may be one cause of leukemia. We reported that ΔETS isoforms, which lack the ETS DNA-binding domain and are frequently expressed in acute myelogenous leukemia transformed from myelodysplastic syndrome, molecularly and biologically show dominant-negative effects over functions of wild-type TEL (41). As expected from their structural similarity in TEL part to ΔETS isoforms, both tTEL and TEL/PTPRR dominantly interfere with transcriptional repression mediated by wild-type TEL of both artificial EBS reporter and endogenous Bcl-XL gene. Thus, losing the ETS DNA-binding domain but retaining the helix-loop-helix domain seems critical to bolster the dominant inhibitory ability. Considering that both tTEL and TEL/PTPRR are found to heterodimerize with wild-type TEL, there are three plausible explanations for molecular basis of this effect. First, tTEL and TEL/PTPRR seem to prevent de novo translated wild-type TEL from entering the nucleus because wild-type TEL molecule resides mainly in the nucleus on its own, but almost equally in both the nucleus and the cytoplasm when coexpressed with the TEL mutants. Importantly, we observed the cytoplasmic localization of endogenous wild-type TEL in primary leukemic cells carrying inv(12). Second, TEL/PTPRR impairs DNA-binding property of wild-type TEL although tTEL does not. Third, tTEL and TEL/PTPRR may take corepressor mSin3A away from wild-type TEL because they associate with endogenous mSin3A in COS-7 cells (data not shown). The same dominant interfering functions of chimeric protein over authentic TEL have also been reported in TEL/AML1 by t(12;21)(p13;q22) in pediatric acute lymphoblastic leukemia (35) and provide one molecular mechanism (loss of tumor suppressive function of wild-type TEL) in leukemogenesis by 12p13 translocations. Given that Bcl-XL is an antiapoptotic molecule, its derepressed expression by the aberrant TEL proteins could be one of the molecular pathogeneses in inv(12)-type leukemia.
PTPRR is the first protein tyrosine phosphatase that was identified as a fusion partner for TEL. It is a receptor-type phosphatase possessing an intracellular catalytic domain at the COOH terminus (46–48). Due to alternative splicing or promoter switch, PTPRR exists as various isoforms in different organs such as brain, placenta, uterus, and colon (49, 50). However, there is no report that PTPRR is expressed in hematopoietic tissues. Notably, wild-type PTPRR is not expressed in the leukemic cells, and TEL/PTPRR does not show any phosphatase activity probably due to incomplete structure of the catalytic domain with loss of its NH2-terminal part. These observations indicate that alteration in phosphatase activity is not causally related to leukemogenic mechanism in these cells. Despite this indication, one interesting hypothesis is that aberrant expression of a portion of catalytically inactive PTPRR in the chimeric form may protect some phosphorylated tyrosine residues that transmit growth-stimulating signals from dephosphorylation. As a result, this may constitutively activate proliferation-inducing pathways.
Employing factor-dependent UT7/GM cells, we explored the oncogenic potential of tTEL and TEL/PTPRR. Overexpression of TEL/PTPRR renders these cells factor independent, whereas overexpression of tTEL does not prevent apoptotic induction after factor deprivation. Therefore, TEL/PTPRR could exhibit transforming activity in myeloid cells, but tTEL alone could not, although tTEL may support the TEL/PTPRR activity. This suggests that the dominant-negative effects over wild-type TEL are not sufficient in the development of leukemia. Interestingly, phosphorylated STAT3 does not fade away after factor withdrawal in both tTEL- and TEL/PTPRR-expressing cells, which is in sharp contrast to mock cells. All the other signaling molecules examined, such as ERK1/2 and JNK in MAP kinase pathway, Akt1/2/3 in phosphatidylinositol 3-kinase pathway, and STAT1 and STAT5 in JAK/STAT pathway, show no difference in phosphorylation levels among mock, tTEL-, and TEL/PTPRR-expressing cells both before and after factor removal. Because STAT3 is one of the well-known signal transducers involved in UT7/GM cell growth (51), the maintenance of phosphorylated STAT3 seems to contribute to the leukemic cell growth without GM-CSF. How these mutants preserve phosphorylated STAT3 remains unknown, but it could be possible for TEL/PTPRR to bind STAT3 through the helix-loop-helix domain of TEL and protect its phosphorylated tyrosine residues through the catalytically inactive domain of PTPRR, because it has been reported that wild-type TEL associates with STAT3 through the helix-loop-helix domain (52). Furthermore, wild-type TEL is found to repress STAT3-mediated transcriptional activity in the literature. Although we could not show the association between endogenous STAT3 and overexpressed tTEL or TEL/PTPRR in UT7/GM cells (data not shown), tTEL and TEL/PTPRR may stimulate the STAT3 signals by heterodimerizing with wild-type TEL and thus blocking its inhibitory functions on STAT3.
In summary, tTEL and TEL/PTPRR produced from the rearranged TEL allele could be important players in the development of leukemia carrying inv(12)(p13q13). They could block two functions of tumor-suppressive wild-type TEL: transcriptional repression through EBS and inhibition of STAT3-mediated signal. However, common functions of these two molecules may not be sufficient, if any, in the leukemogenesis because overexpression of tTEL does not induce autonomous cell growth of the factor-dependent cells by itself. The PTPRR part in TEL/PTPRR fusion protein might provide additional unknown functions for growth advantage. Further experiments with mouse modeling are needed to prove their leukemogenic roles in vivo.
Grant support: Grants-in-Aid from the Ministry of Education, Culture, Sports, Science, and Technology of Japan; Ministry of Health, Labour, and Welfare of Japan; Japanese Society for the Promotion of Science; and the Japan Health Sciences Foundation.
We thank Dr. N. Komatsu (University of Yamanashi, Yamanashi, Japan) for the generous gift of UT7/GM human leukemia cells; Dr. J. Miyazaki (Osaka University, Osaka, Japan) for presenting us with pCXN2 plasmid; and KIRIN Brewery Co. Ltd. for the kind gift of human recombinant GM-CSF.