- Email: [email protected]
The TEL gene and human leukemia
BiBB och~Pic~a
wjtAc: -
ELSEVIER
et Biophysica A~ta Biochimica et Biophysica Acta 1288 (1996) M7-M 10
Mini Review
The TEL gene and human leukemia Todd R. Golub *, Thomas McLean, Kimberly Stegmaier, Martin Carroll, Michael Tomasson, D. Gary Gilliland Division of Hematology~ Oncology, Brigham and Women's Hospital and the Dana-Farber Cancer Institute, Harvard Medical School, Boston, MA 02115, USA Received 6 May 1996; accepted 8 May 1996
1. TEL gene structure TEL (translocation, ets, leukemia) is a new member of the ETS family of transcription factors that was cloned by virtue of its fusion to the platelet-derived growth factor 13 receptor (PDGF13R) in chronic myelomonocytic leukemia (CMML) associated with t(5;12)(q33;p13). ETS proteins were first discovered as part of a fusion with gag and myb in the E26 avian erythroblastosis virus, and now constitute a large family of DNA binding proteins that recognize the core motif C / A GGA A / T [1,2]. TEL contains the phylogenetically conserved carboxy terminal DNA binding domain which defines the ETS family of transcription factors. In addition, TEL belongs to a subclass of ETS proteins which contain a highly conserved amino terminal domain. The 3D structure of the amino terminal conserved domain has not yet been solved, and its function is not fully understood. What is known is that it is essential for full transactivating function, but does not contact DNA directly [3,4]. Because of weak homology to the helixloop-helix (HLH) unit of proteins such as MYC and MYO-D [5-7], it has been suggested that the N-terminal region of ETS may be involved in protein-protein interactions which facilitate transcriptional activation [8]. On the other hand, direct evidence that the N-terminal HLH-like sequence mediates protein-protein interaction is, thus far, lacking. As described below, the presumed involvement of TEL in human leukemias has led to the first direct evidence that the TEL HLH sequence mediates one or more protein-protein interactions.
2. Involvement of TEL in h u m a n leukemia
TEL is notable for its apparent involvement in a broad spectrum of human leukemias, including acute myeloid
* Corresponding author. 0304-419X/96/$15.00 © 1996 Elsevier Science B.V. All rights reserved PII S 0 3 0 4 - 4 1 9 X ( 9 6 ) 0 0 0 1 5 - 7
leukemia (AML), chronic myelomonocytic leukemia (CMML), acute lymphocytic leukemia, and myelodysplastic syndrome (Table 1). In all cases of leukemia involving the TEL gene, a novel transcript is created by fusion of TEL to another gene as a consequence of chromosomal translocation. There is diversity of fusion partners, which, thus far, includes the platelet-derived growth factor 13 receptor (PDGFflR), the protooncogene tyrosine kinase ABL, the transcription factor AML1, and the MN1 gene, the function of which is not known. A remarkable feature of TEL involvement in human leukemias is that different functional domains of TEL are implicated in different leukemias. Thus, in the TEL-PDGFflR, TEL-ABL and TEL-AML1 fusions, a major functional role for the TEL HLH domain seems likely, while, in the MN1-TEL fusion, the focus is on the TEL DNA binding domain. Another interesting feature of TEL gene translocations in leukemia is frequent deletion of the TEL allele not involved in the translocation. For example, in patients with the TEL-AML1 fusion, the non-interrupted TEL allele appears to be invariably deleted [9-11]. Since one TEL allele is disrupted by translocation and the other is deleted, there is no functional TEL in these leukemic cells. This observation raises the possibility that loss of TEL function contributes to pathogenesis of disease a n d / o r that wildtype TEL interferes with the oncogenic properties of the TELrelated fusion protein.
3. TEL is fused to PDGFI3R in C M M L with t(5;12) CMML is a myelodysplastic syndrome characterized by monocytosis and dysplastic hematopoiesis, t(5;12) is a recurring translocation which occurs in patients with CMML, and is frequently associated with myelofibrosis and eosinophilia [12-16]. The consequence of the t(5;12) is the in-frame fusion of the TEL HLH sequence to the transmembrane and tyrosine kinase domains of PDGFt3R
M8
T.R. Golub et al. /Biochimica et Biophysica Acta 1288 (1996) M7-MIO
[17]. Wildtype PDGFI3R requires dimerization by PDGF BB ligand for tyrosine kinase activation and mitogenesis, which has led to the hypothesis that transformation by TEL-PDGFI3R is due to constitutive dimerization and activation of the PDGFI3R kinase mediated by the TEL HLH sequence [17]. Consistent with this hypothesis, we have shown that TEL-PDGFI3R confers factor independent growth to the murine hematopoietic cell line B a / F 3 and is constitutively tyrosine phosphorylated in these cells (M.C., M.T., T.R,G. and D.G.G., unpublished observation). Furthermore, deletion of the TEL HLH sequence or inactivation of the PDGFI~R kinase activity by point mutations abrogates transforming activity of the TEL-PDGFI~R fusion protein. These data are consistent with the possibility that constitutive dimerization and activation of the PDGFI3R kinase are mediated, at least in part, by the TEL HLH domain. Further experiments are needed to learn whether that the TEL HLH domain mediates TELPDGFI3R dimerization and to identify targets of the activated TEL-PDGFI3R fusion protein.
4. TEL-ABL fusion is associated with AML, ALL and
atypical CML with t(9;12) A TEL-ABL fusion protein has recently been reported in a pediatric patient with ALL and t(9;12) [18], in a patient with a complex t(9;12;14) translocation and acute undifferentiated leukemia with myeloid markers [19], and in a patient with atypical CML (F. Birg et al., personal communication). The consequence of t(9;12) in each case is in-frame fusion of the TEL HLH sequence ito exon 2 of the protooncogene ABL. TEL-ABL is similar in structure to the well characterized BCR-ABL fusion associated with chronic myelogenous leukemia (CML) and t(9;22). TEL is the only fusion partner that has been identified for ABL other than BCR, and reveals important similarities to and differences from BCR. For example, BCR contains a coiled-coil domain which mediates oligomerization, and is necessary for tyrosine kinase and transforming activity of BCR-ABL. The coiled-coil motif of BCR and the HLH domain of TEL may both therefore mediate dimerization or multimerization leading to constitutive activation of the ABL tyrosine kinase. As detailed below, we have obtained direct evi-
Table 1 Involvementof TEL Translocation t(5;12) t(9;12) (12;22) t(12;21)
in human leukemia Phenotype CMML ALL AML AML, MDS pediatric ALL
dence from analysis of in vitro translated TEL-ABL that the TEL HLH domain mediates self-association. TEL-ABL transforms B a / F 3 cells, ratl fibroblasts and primary murine bone marrow cells. Inactivation of the ABL kinase by point mutation abrogates transforming activity of TEL-ABL, as does deletion of the TEL HLH domain [19], indicating an important role for these sequence units in the transforming mechanism. Interestingly, deletion of the TEL HLH sequence results in loss of ABL kinase activity, suggesting that the former sustains the latter, possibly by promoting self-association (e.g. dimerization). In addition, TEL-ABL, like BCR-ABL, localizes to the cytoskeleton of transformed cells. TEL-ABL kinase inactive mutants retain normal cytoskeletal localization properties, but deletion of the TEL HLH domain resulted in loss of cytoskeletal association [19]. This implies that the HLH sequence also contributes to proper intracellular TEL-ABL localization, which is part of the key to the expression of TEL-ABL kinase function, which, in turn, is essential for TEL-ABL transforming activity. Given the similarities between TEL and BCR, further analysis of signal transduction and transformation by the TEL-ABL fusion protein may well enhance one's understanding of transformation by BCR-ABL.
5. The MN1-TEL fusion associated with t(12;22) and AML incorporates the TEL DNA binding domain T h e M N 1 - T E L fusion associated with AML and t(12;22) further demonstrates the pleiotropic involvement of TEL in hematologic malignancy. The T E L - M N 1 fusion (Table 1) contains exon 1 of the M N 1 gene fused in frame to the TEL DNA binding domain [20]. In contrast with the TEL-PDGF[3R and TEL-ABL fusions, the TEL HLH unit is not incorporated into the fusion. The structure of the fusion is thus similar to the EWS-FLI1 fusion associated with t(11;22) and Ewing's sarcoma and to the TLS-ERG fusion associated with AML and t(16;22), which incorporate the DNA binding domain of the ETS family members, ERG and FLII. The mechanism of transformation of fusion proteins which incorporate ETS family DNA binding domains is not known, but may include (i) aberrant expression of ETS DNA binding domains driven by the promoter of the fusion partner (MN 1, TLS or EWS), and (ii) abnor-
Fusion gene TEL-PDGFI~R TEL-ABLatypicalCML
Functionaldomains TEL HLH PDGFI3Rtyrosinekinase TEL HLH ABL kinase
MNI-TEL TEL-AML1
TEL DNA bindingMN1 exon 1 TEL HLH AML1 DNA binding and transactivationdomains
Cloning ref. [17] ([18,19] and F. Birg, pers. commun.) [20] [9,10]
T.R. Golub et al. / Biochimica et Biophysica Acta 1288 (1996) M7-MIO
real transcriptional activation activity of an ETS DNA binding domain when fused to a heterologous partner. The function of the MNI portion of the fusion is not known [21], but MNI has structural similarity to EWS. MN1 may therefore serve a similar function in transformation mediated by the TEL DNA binding domain as EWS does in the EWS-FLI! fusion [20].
6. TEL-AML1 is the most c o m m o n gene rearrangement in childhood acute lymphoblastic leukemia
The TEL-AML1 fusion gene was first cloned from pediatric patients with pre-B cell ALL [9,10]. These first four cases of TEL-AML1 fusion shared several interesting characteristics. First, the TEL-AMLI fusion occurred as a consequence of a cryptic (12;21) chromosomal translocation which was not apparent at the cytogenetic level. Second, these cases were the first example of AML1 involvement in lymphoid leukemia. Other translocations involving AML1 such as the AML1-ETO fusion associated with t(8;21) and the AML1-EVI1 fusion in t(3;21) all occurred in myeloid leukemias. Third, the structure of the TEL-AML1 fusion was different from other fusions involving AMLI, in that both the DNA binding domain (RHD) and the transactivating domain (TA) of AML1 were incorporated into the fusion protein. Finally, an intriguing observation in all four initial cases was that one TEL allele was involved in the translocation, and the other TEL allele was deleted. There was thus no wildtype TEL in these leukemic cells. Subsequent analysis of larger numbers of patients with TEL-AML1 using FISH and TEL specific cosmid probes suggests that, in the vast majority of leukemias carrying a TEL-AML1 fusion gene, the residual TEL allele is deleted [11]. Since the t(12;21) was cryptic in the first cases analyzed, several groups have analyzed larger numbers of cases of pediatric ALL and have determined that the frequency of the translocation is extraordinarily high [11,22-24]. For example, in our series of 81 cases of childhood ALL analyzed from the Dana-Farber Cancer Institute, 22 patients (27%) were TEL-AML1 positive [23]. Since ALL is the most common malignancy of childhood, the TEL-AML1 fusion is the most common gene rearrangement observed in any pediatric malignancy. The mechanism by which TEL-AML1 transforms cells is not known. It is possible that the fusion of these two transcription factors leads to a protein with abnormally high transactivation activity compared to either wildtype TEL or AML1 gene products. Alternatively the fusion protein may have qualitatively different behavior from either of the parents. Consistent with this notion, wildtype AMLI is an activator of the TCRI3 enhancer, but TELAMLI is a repressor of this element [25]. Further analysis will be necessary to characterize the structure-function relationships of the TEL-AMLI fusion protein and the
M9
effect of the fusion protein on the expression of all known AMLI target genes.
7. The TEL HLH-like sequence mediates self-association As suggested earlier, one function for the TEL HLH-like sequence might be self-association leading to activation of TEL-tyrosine kinase fusion proteins such as TEL-ABL and TEL-PDGFR. To determine whether the TEL HLH unit mediates self-association, we have analyzed in vitro translated TEL-ABL, wiidtype TEL, and mutants of these species lacking the TEL HLH sequence. The data indicate that the TEL HLH sequence mediates TEL-ABL::TELABL homodimerization, TEL-ABL::TEL heterodimerization, and TEL::TEL homodimerization in this setting [19]. Similar results were obtained from an analysis of in vitro translated TEL-PDGF[3R and TEL-AMLI (T.R.G. and D.G.G., unpublished observation). More work is needed to prove that the TEL HLH-like sequence mediates self-association, and to determine the in vivo stoichiometry of wildtype and TEL fusion proteins. These studies would be enriched, in time, by solution of the 3D structure of the TEL HLH unit.
8. Loss of the residual TEL allele in leukemic cells carrying a TEL fusion gene Raynaud et al. have provided convincing evidence that the residual TEL allele was deleted in leukemic cells of 14/16 patients with the TEL-AMLI fusion gene [11]. Furthermore, all patients with TEL deletion by FISH analysis also had TEL-AML1 gene rearrangements, suggesting an invariant association between TEL deletion and TELAML1 fusion. One explanation for these data would be that TEL is a tumor suppressor gene, and that loss of TEL function is a primary determinant of transforming activity. In this model, one allele of TEL is deleted, and chromosomal translocation serves only to disrupt the function of the residual TEL gene. Translocation as a mechanism for loss of function has recently been reported for the p l6 gene [26]. If this model is correct, one would predict that as for p16 and other tumor suppressor genes, some patients with one TEL allele deleted should have point mutations in the other allele which cause loss of function. To address this possibility, we analyzed 33 patients childhood acute lymphoblastic leukemia and loss of one TEL allele to determine if there were point mutations in the other TEL allele. We found no evidence for mutation of the residual TEL allele by either SSCP or direct sequence analysis [27]. In fact, based on the data of Raynaud et al. [11], it seems likely that all cases of LOH of the TEL gene locus in childhood ALL are associated with a TEL-AMLI gene rearrangement.
MI0
T.R. Golub et al. / Biochimica et Biophysica Acta 1288 (1996) M7-MIO
The invariant association of loss of one TEL allele with a TEL-AML1 gene rearrangement suggests that loss of TEL function alone may not be sufficient for the malignant phenotype. The phenotypic consequences of TEL loss of function can be probed by deletion of the gene through homologous recombination in murine embryonic stem cells, and assessing the effect of the loss of TEL on hematopoietic differentiation of ES ceils, and in TEL deficient mice. This approach is currently under investigation.
9. Does TEL loss of function represent a new paradigm for pathogenesis of malignancy in humans? The tumor suppressor paradigm, notwithstanding, the question remains: why is loss of one TEL allele so frequently associated with translocation and fusion of the other TEL allele? The association appears to be invariant in TEL-AML1 fusions, and occurred in the single TEL-ABL fusion patient we have described [19]. The explanation that we favor represents a new paradigm for loss of function in malignancy, in which loss of TEL function alone does not transform cells. Rather, one might argue that TEL interferes with the transforming activity of TEL fusion proteins, so that its loss is necessary to unleash the transforming potential of the former. In support of this hypothesis, we have shown that TEL can bind to TEL-AML1, TELPDGFI3R and TEL-ABL, and that homodimerization of each of the fusions is required for transforming activity. Additional experiments are needed to test the above-noted hypothesis, such as determining the effect of overexpression of TEL on TEL-ABL kinase activity and transforming activity. Experiments of this type may provide insight into the mechanism of transformation by TEL fusion proteins, and identify strategies for reverting the malignant phenotype which can be tested in vitro and in animal models of leukemia.
References [1] Nunn, M., Seeburg, P.H., Moscovici, C. and Duesberg, P.H. (1983) Nature 306, 391-395. [2] Nye, J.A., Petersen, J.M., Gunther, C.V., Jonsen, M.D. and Graves, B.J. (1990) Genes Dev. 6, 975-990. [3] Siddique, H.R., Rao, V.N., Lee, L. and Reddy, S.P. (1993) Oncogene 8, 1751-1755. [4] Rao, V.N., Ohno, T., Prasad, D.D.K., Bhattacharya, G. and Reddy, E.S.P. (1993) Oncogene 8, 2167-2173. [5] Wasylyk, B., Hahn, S.L. and Giovane, A. (1993) The Ets family of transcription factors. Eur. J. Biochem. 211, 7-18. [6] The, S.M., Xie, X., Smyth, F., Papas, T.S., Watson, D.K. and Schulz, R.A. (1992) Oncogene 7, 2471-2478.
[7] Watanabe, H., Sawada, J.-U., Yano, K.-I., Yamaguchi, K., Goto, M. and Handa, H. (1993) Mol. Cell. Biol. 13, 1385-1391. [8] Seth, A. and Papas, T.S. (1990) Oncogene 5, 1761-1767. [9] Golub, T.R., Barker, G.F., Bohlander, S.K., Hiebert, S.W., Ward, D.C., Bray-Ward, P., Morgan, E., Raimondi, S.C., Rowley, J.D. and D.G.G. (1995) Proc. Natl. Acad. Sci., in press. [10] Romana, S.P., Mauchauffe, M., Leconiat, M., Chumakov, I., Le Paslier, D., Berger, R. and Bernard, O.A. (1995) Blood 85, 36623670. [11] Raynaud, S.D., Cave, H., Baens, M., Bastard, C., Cacheux, V., Grosgeorge, J., Guidal-Giroux, C., Guo, C., Vilmer, E., Marynen, P. and Grandchamp, B. (1996) Blood 87, 2891-2899. [12] Keene, P., Mendelow, B., Pinto, M.R., Bezwoda, W., MacDougall, L., Falkson, G., Ruff, P. and Bernstein, R. (1987) Br. J. Haematol. 67, 25-31. [13] Srivastava, A., Boswell, H., Heerema, N.A., Nahreini, P., Lauer, R.D., Antony, A.S., Hoffman, R., Tricot, G.J. (1988) Cancer Genet. Cytogenet. 35, 61-71. [14] Berkowicz, M., Rosner, E., Rechavi, C., Mamon, Z., Neuman, Y., Ben-Bassat, I. and Ramot, B. (1991) Atypical chronic myelomonocytic leukemia with eosinophilia and translocation (5;12): A new assocation? Cancer Genet. Cytogenet. 51,277-278. [15] Lerza, R., Castello, G., Sessarego, M., Cavallini, D. and Pannacciulli, I. (1992) Br. J. Haematol. 82, 476-477. [16] Wessels, J.W., Fibbe, W.E., van der Keur, D., Landegent, J.E., van der Plas, D.C., den Ottolander, G.J., Roozendaal, K.J., Beverstock, G.D. (1993) t(5;12)(q31;p12) A clinical entity with features of both myeloid leukemia and chronic myelomonocytic leukemia. Cancer Genet. Cytogenet. 65, 7-11. [17] Golub, T.R., Barker, G.F., LoveU, M., Gilliland, D.G. (1994) Fusion of PDGF receptor beta to a novel ets-like gene, tel, in chronic myelomonocytic leukemia with t(5;12) chromosomal translocation. Cell 77, 307-316. [18] Papadopoulos, P., Ridge, S.A., Boucher, C.A., Stocking, C. and Wiedemann, L.M. (1995) The novel activation of ABL by fusion to an ets-related gene, TEL. Cancer Res. 55, 34-38. [19] Golub, T.R., Goga, A., Barker, G., Afar, D., McLaughlin, J., Boh|ander, S., Rowley, J., Witte, O., Gilliland, D.G. (1996) Oligomerization of the ABL tyrosine kinase by the ETS protein TEL in human leukemia. Mol. Cell Biol., in press. [20] Buijs, A., Sherr, S., van Baal, S., van Bezouw, S., van der Plas, D., van Kessel, A.D., Riegman, P., Deprez, R.L., Zwarthoff, E., Hagemeijer, A. and Grosveld, G. (1995) Oncogene 10, 1511-1519. [21] Deprez, R.H.L., Riegman, P.H.J., Groen, N.A., Warringa, U.L., van Biezen, N.A., Molijn, A.C., Bootsma, D., de Jong, P.J., Menon, A.G., Kley, N.A, Seizinger, B.R. and Zwarthoff, E.C. (1995) Oncogene 10, 1521-1528. [22] Romana, S.P., Piorel, H., Leconiat, M., Flexor, M.A., Mauchaffe, M., Jonveaux, P., Berger, R. and Bernard, O.A. (1995) Blood 86, 4263-4269. [23] Golub, T., McLean, T., Stegmaier, K., Ritz, J., Sallan, S., Neuberg, D., Gilliland, D.G. (1995) Blood 86 (Suppl 1), 597a. [24] Shurtleff, S., Buijs, A., Behm, F., Rubnitz, J., Raimondi, S., Hancock, M., Chan, G., Pui, C., Grosveld, G. and Downing, J. (1995) Leukemia 9, 1985-1989. [25] Hiebert, S.W., Sun, W., Davis, J.N., Golub, T.R., Shurtleff, S., Buijs, A., J.R.D., Grosveld, G., Roussel, M., Gilliland, D,G., Lenny, N. and Meyers, S. (1996) Mol. Cell. Biol., in press. [26] Duro, D., Bernard, O., Della Valle, V., Berger, R. and Larsen, C.J. (1995) Oncogene 11, 21-29. [27] Stegmaier, K., Takeuchi, S., Golub, T.R., Bohlander SK, C.R. B, Koeffler HP, Gilliland DG. (1996) Cancer Res. 56, 1413-1417.
wjtAc: -
ELSEVIER
et Biophysica A~ta Biochimica et Biophysica Acta 1288 (1996) M7-M 10
Mini Review
The TEL gene and human leukemia Todd R. Golub *, Thomas McLean, Kimberly Stegmaier, Martin Carroll, Michael Tomasson, D. Gary Gilliland Division of Hematology~ Oncology, Brigham and Women's Hospital and the Dana-Farber Cancer Institute, Harvard Medical School, Boston, MA 02115, USA Received 6 May 1996; accepted 8 May 1996
1. TEL gene structure TEL (translocation, ets, leukemia) is a new member of the ETS family of transcription factors that was cloned by virtue of its fusion to the platelet-derived growth factor 13 receptor (PDGF13R) in chronic myelomonocytic leukemia (CMML) associated with t(5;12)(q33;p13). ETS proteins were first discovered as part of a fusion with gag and myb in the E26 avian erythroblastosis virus, and now constitute a large family of DNA binding proteins that recognize the core motif C / A GGA A / T [1,2]. TEL contains the phylogenetically conserved carboxy terminal DNA binding domain which defines the ETS family of transcription factors. In addition, TEL belongs to a subclass of ETS proteins which contain a highly conserved amino terminal domain. The 3D structure of the amino terminal conserved domain has not yet been solved, and its function is not fully understood. What is known is that it is essential for full transactivating function, but does not contact DNA directly [3,4]. Because of weak homology to the helixloop-helix (HLH) unit of proteins such as MYC and MYO-D [5-7], it has been suggested that the N-terminal region of ETS may be involved in protein-protein interactions which facilitate transcriptional activation [8]. On the other hand, direct evidence that the N-terminal HLH-like sequence mediates protein-protein interaction is, thus far, lacking. As described below, the presumed involvement of TEL in human leukemias has led to the first direct evidence that the TEL HLH sequence mediates one or more protein-protein interactions.
2. Involvement of TEL in h u m a n leukemia
TEL is notable for its apparent involvement in a broad spectrum of human leukemias, including acute myeloid
* Corresponding author. 0304-419X/96/$15.00 © 1996 Elsevier Science B.V. All rights reserved PII S 0 3 0 4 - 4 1 9 X ( 9 6 ) 0 0 0 1 5 - 7
leukemia (AML), chronic myelomonocytic leukemia (CMML), acute lymphocytic leukemia, and myelodysplastic syndrome (Table 1). In all cases of leukemia involving the TEL gene, a novel transcript is created by fusion of TEL to another gene as a consequence of chromosomal translocation. There is diversity of fusion partners, which, thus far, includes the platelet-derived growth factor 13 receptor (PDGFflR), the protooncogene tyrosine kinase ABL, the transcription factor AML1, and the MN1 gene, the function of which is not known. A remarkable feature of TEL involvement in human leukemias is that different functional domains of TEL are implicated in different leukemias. Thus, in the TEL-PDGFflR, TEL-ABL and TEL-AML1 fusions, a major functional role for the TEL HLH domain seems likely, while, in the MN1-TEL fusion, the focus is on the TEL DNA binding domain. Another interesting feature of TEL gene translocations in leukemia is frequent deletion of the TEL allele not involved in the translocation. For example, in patients with the TEL-AML1 fusion, the non-interrupted TEL allele appears to be invariably deleted [9-11]. Since one TEL allele is disrupted by translocation and the other is deleted, there is no functional TEL in these leukemic cells. This observation raises the possibility that loss of TEL function contributes to pathogenesis of disease a n d / o r that wildtype TEL interferes with the oncogenic properties of the TELrelated fusion protein.
3. TEL is fused to PDGFI3R in C M M L with t(5;12) CMML is a myelodysplastic syndrome characterized by monocytosis and dysplastic hematopoiesis, t(5;12) is a recurring translocation which occurs in patients with CMML, and is frequently associated with myelofibrosis and eosinophilia [12-16]. The consequence of the t(5;12) is the in-frame fusion of the TEL HLH sequence to the transmembrane and tyrosine kinase domains of PDGFt3R
M8
T.R. Golub et al. /Biochimica et Biophysica Acta 1288 (1996) M7-MIO
[17]. Wildtype PDGFI3R requires dimerization by PDGF BB ligand for tyrosine kinase activation and mitogenesis, which has led to the hypothesis that transformation by TEL-PDGFI3R is due to constitutive dimerization and activation of the PDGFI3R kinase mediated by the TEL HLH sequence [17]. Consistent with this hypothesis, we have shown that TEL-PDGFI3R confers factor independent growth to the murine hematopoietic cell line B a / F 3 and is constitutively tyrosine phosphorylated in these cells (M.C., M.T., T.R,G. and D.G.G., unpublished observation). Furthermore, deletion of the TEL HLH sequence or inactivation of the PDGFI~R kinase activity by point mutations abrogates transforming activity of the TEL-PDGFI~R fusion protein. These data are consistent with the possibility that constitutive dimerization and activation of the PDGFI3R kinase are mediated, at least in part, by the TEL HLH domain. Further experiments are needed to learn whether that the TEL HLH domain mediates TELPDGFI3R dimerization and to identify targets of the activated TEL-PDGFI3R fusion protein.
4. TEL-ABL fusion is associated with AML, ALL and
atypical CML with t(9;12) A TEL-ABL fusion protein has recently been reported in a pediatric patient with ALL and t(9;12) [18], in a patient with a complex t(9;12;14) translocation and acute undifferentiated leukemia with myeloid markers [19], and in a patient with atypical CML (F. Birg et al., personal communication). The consequence of t(9;12) in each case is in-frame fusion of the TEL HLH sequence ito exon 2 of the protooncogene ABL. TEL-ABL is similar in structure to the well characterized BCR-ABL fusion associated with chronic myelogenous leukemia (CML) and t(9;22). TEL is the only fusion partner that has been identified for ABL other than BCR, and reveals important similarities to and differences from BCR. For example, BCR contains a coiled-coil domain which mediates oligomerization, and is necessary for tyrosine kinase and transforming activity of BCR-ABL. The coiled-coil motif of BCR and the HLH domain of TEL may both therefore mediate dimerization or multimerization leading to constitutive activation of the ABL tyrosine kinase. As detailed below, we have obtained direct evi-
Table 1 Involvementof TEL Translocation t(5;12) t(9;12) (12;22) t(12;21)
in human leukemia Phenotype CMML ALL AML AML, MDS pediatric ALL
dence from analysis of in vitro translated TEL-ABL that the TEL HLH domain mediates self-association. TEL-ABL transforms B a / F 3 cells, ratl fibroblasts and primary murine bone marrow cells. Inactivation of the ABL kinase by point mutation abrogates transforming activity of TEL-ABL, as does deletion of the TEL HLH domain [19], indicating an important role for these sequence units in the transforming mechanism. Interestingly, deletion of the TEL HLH sequence results in loss of ABL kinase activity, suggesting that the former sustains the latter, possibly by promoting self-association (e.g. dimerization). In addition, TEL-ABL, like BCR-ABL, localizes to the cytoskeleton of transformed cells. TEL-ABL kinase inactive mutants retain normal cytoskeletal localization properties, but deletion of the TEL HLH domain resulted in loss of cytoskeletal association [19]. This implies that the HLH sequence also contributes to proper intracellular TEL-ABL localization, which is part of the key to the expression of TEL-ABL kinase function, which, in turn, is essential for TEL-ABL transforming activity. Given the similarities between TEL and BCR, further analysis of signal transduction and transformation by the TEL-ABL fusion protein may well enhance one's understanding of transformation by BCR-ABL.
5. The MN1-TEL fusion associated with t(12;22) and AML incorporates the TEL DNA binding domain T h e M N 1 - T E L fusion associated with AML and t(12;22) further demonstrates the pleiotropic involvement of TEL in hematologic malignancy. The T E L - M N 1 fusion (Table 1) contains exon 1 of the M N 1 gene fused in frame to the TEL DNA binding domain [20]. In contrast with the TEL-PDGF[3R and TEL-ABL fusions, the TEL HLH unit is not incorporated into the fusion. The structure of the fusion is thus similar to the EWS-FLI1 fusion associated with t(11;22) and Ewing's sarcoma and to the TLS-ERG fusion associated with AML and t(16;22), which incorporate the DNA binding domain of the ETS family members, ERG and FLII. The mechanism of transformation of fusion proteins which incorporate ETS family DNA binding domains is not known, but may include (i) aberrant expression of ETS DNA binding domains driven by the promoter of the fusion partner (MN 1, TLS or EWS), and (ii) abnor-
Fusion gene TEL-PDGFI~R TEL-ABLatypicalCML
Functionaldomains TEL HLH PDGFI3Rtyrosinekinase TEL HLH ABL kinase
MNI-TEL TEL-AML1
TEL DNA bindingMN1 exon 1 TEL HLH AML1 DNA binding and transactivationdomains
Cloning ref. [17] ([18,19] and F. Birg, pers. commun.) [20] [9,10]
T.R. Golub et al. / Biochimica et Biophysica Acta 1288 (1996) M7-MIO
real transcriptional activation activity of an ETS DNA binding domain when fused to a heterologous partner. The function of the MNI portion of the fusion is not known [21], but MNI has structural similarity to EWS. MN1 may therefore serve a similar function in transformation mediated by the TEL DNA binding domain as EWS does in the EWS-FLI! fusion [20].
6. TEL-AML1 is the most c o m m o n gene rearrangement in childhood acute lymphoblastic leukemia
The TEL-AML1 fusion gene was first cloned from pediatric patients with pre-B cell ALL [9,10]. These first four cases of TEL-AML1 fusion shared several interesting characteristics. First, the TEL-AMLI fusion occurred as a consequence of a cryptic (12;21) chromosomal translocation which was not apparent at the cytogenetic level. Second, these cases were the first example of AML1 involvement in lymphoid leukemia. Other translocations involving AML1 such as the AML1-ETO fusion associated with t(8;21) and the AML1-EVI1 fusion in t(3;21) all occurred in myeloid leukemias. Third, the structure of the TEL-AML1 fusion was different from other fusions involving AMLI, in that both the DNA binding domain (RHD) and the transactivating domain (TA) of AML1 were incorporated into the fusion protein. Finally, an intriguing observation in all four initial cases was that one TEL allele was involved in the translocation, and the other TEL allele was deleted. There was thus no wildtype TEL in these leukemic cells. Subsequent analysis of larger numbers of patients with TEL-AML1 using FISH and TEL specific cosmid probes suggests that, in the vast majority of leukemias carrying a TEL-AML1 fusion gene, the residual TEL allele is deleted [11]. Since the t(12;21) was cryptic in the first cases analyzed, several groups have analyzed larger numbers of cases of pediatric ALL and have determined that the frequency of the translocation is extraordinarily high [11,22-24]. For example, in our series of 81 cases of childhood ALL analyzed from the Dana-Farber Cancer Institute, 22 patients (27%) were TEL-AML1 positive [23]. Since ALL is the most common malignancy of childhood, the TEL-AML1 fusion is the most common gene rearrangement observed in any pediatric malignancy. The mechanism by which TEL-AML1 transforms cells is not known. It is possible that the fusion of these two transcription factors leads to a protein with abnormally high transactivation activity compared to either wildtype TEL or AML1 gene products. Alternatively the fusion protein may have qualitatively different behavior from either of the parents. Consistent with this notion, wildtype AMLI is an activator of the TCRI3 enhancer, but TELAMLI is a repressor of this element [25]. Further analysis will be necessary to characterize the structure-function relationships of the TEL-AMLI fusion protein and the
M9
effect of the fusion protein on the expression of all known AMLI target genes.
7. The TEL HLH-like sequence mediates self-association As suggested earlier, one function for the TEL HLH-like sequence might be self-association leading to activation of TEL-tyrosine kinase fusion proteins such as TEL-ABL and TEL-PDGFR. To determine whether the TEL HLH unit mediates self-association, we have analyzed in vitro translated TEL-ABL, wiidtype TEL, and mutants of these species lacking the TEL HLH sequence. The data indicate that the TEL HLH sequence mediates TEL-ABL::TELABL homodimerization, TEL-ABL::TEL heterodimerization, and TEL::TEL homodimerization in this setting [19]. Similar results were obtained from an analysis of in vitro translated TEL-PDGF[3R and TEL-AMLI (T.R.G. and D.G.G., unpublished observation). More work is needed to prove that the TEL HLH-like sequence mediates self-association, and to determine the in vivo stoichiometry of wildtype and TEL fusion proteins. These studies would be enriched, in time, by solution of the 3D structure of the TEL HLH unit.
8. Loss of the residual TEL allele in leukemic cells carrying a TEL fusion gene Raynaud et al. have provided convincing evidence that the residual TEL allele was deleted in leukemic cells of 14/16 patients with the TEL-AMLI fusion gene [11]. Furthermore, all patients with TEL deletion by FISH analysis also had TEL-AML1 gene rearrangements, suggesting an invariant association between TEL deletion and TELAML1 fusion. One explanation for these data would be that TEL is a tumor suppressor gene, and that loss of TEL function is a primary determinant of transforming activity. In this model, one allele of TEL is deleted, and chromosomal translocation serves only to disrupt the function of the residual TEL gene. Translocation as a mechanism for loss of function has recently been reported for the p l6 gene [26]. If this model is correct, one would predict that as for p16 and other tumor suppressor genes, some patients with one TEL allele deleted should have point mutations in the other allele which cause loss of function. To address this possibility, we analyzed 33 patients childhood acute lymphoblastic leukemia and loss of one TEL allele to determine if there were point mutations in the other TEL allele. We found no evidence for mutation of the residual TEL allele by either SSCP or direct sequence analysis [27]. In fact, based on the data of Raynaud et al. [11], it seems likely that all cases of LOH of the TEL gene locus in childhood ALL are associated with a TEL-AMLI gene rearrangement.
MI0
T.R. Golub et al. / Biochimica et Biophysica Acta 1288 (1996) M7-MIO
The invariant association of loss of one TEL allele with a TEL-AML1 gene rearrangement suggests that loss of TEL function alone may not be sufficient for the malignant phenotype. The phenotypic consequences of TEL loss of function can be probed by deletion of the gene through homologous recombination in murine embryonic stem cells, and assessing the effect of the loss of TEL on hematopoietic differentiation of ES ceils, and in TEL deficient mice. This approach is currently under investigation.
9. Does TEL loss of function represent a new paradigm for pathogenesis of malignancy in humans? The tumor suppressor paradigm, notwithstanding, the question remains: why is loss of one TEL allele so frequently associated with translocation and fusion of the other TEL allele? The association appears to be invariant in TEL-AML1 fusions, and occurred in the single TEL-ABL fusion patient we have described [19]. The explanation that we favor represents a new paradigm for loss of function in malignancy, in which loss of TEL function alone does not transform cells. Rather, one might argue that TEL interferes with the transforming activity of TEL fusion proteins, so that its loss is necessary to unleash the transforming potential of the former. In support of this hypothesis, we have shown that TEL can bind to TEL-AML1, TELPDGFI3R and TEL-ABL, and that homodimerization of each of the fusions is required for transforming activity. Additional experiments are needed to test the above-noted hypothesis, such as determining the effect of overexpression of TEL on TEL-ABL kinase activity and transforming activity. Experiments of this type may provide insight into the mechanism of transformation by TEL fusion proteins, and identify strategies for reverting the malignant phenotype which can be tested in vitro and in animal models of leukemia.
References [1] Nunn, M., Seeburg, P.H., Moscovici, C. and Duesberg, P.H. (1983) Nature 306, 391-395. [2] Nye, J.A., Petersen, J.M., Gunther, C.V., Jonsen, M.D. and Graves, B.J. (1990) Genes Dev. 6, 975-990. [3] Siddique, H.R., Rao, V.N., Lee, L. and Reddy, S.P. (1993) Oncogene 8, 1751-1755. [4] Rao, V.N., Ohno, T., Prasad, D.D.K., Bhattacharya, G. and Reddy, E.S.P. (1993) Oncogene 8, 2167-2173. [5] Wasylyk, B., Hahn, S.L. and Giovane, A. (1993) The Ets family of transcription factors. Eur. J. Biochem. 211, 7-18. [6] The, S.M., Xie, X., Smyth, F., Papas, T.S., Watson, D.K. and Schulz, R.A. (1992) Oncogene 7, 2471-2478.
[7] Watanabe, H., Sawada, J.-U., Yano, K.-I., Yamaguchi, K., Goto, M. and Handa, H. (1993) Mol. Cell. Biol. 13, 1385-1391. [8] Seth, A. and Papas, T.S. (1990) Oncogene 5, 1761-1767. [9] Golub, T.R., Barker, G.F., Bohlander, S.K., Hiebert, S.W., Ward, D.C., Bray-Ward, P., Morgan, E., Raimondi, S.C., Rowley, J.D. and D.G.G. (1995) Proc. Natl. Acad. Sci., in press. [10] Romana, S.P., Mauchauffe, M., Leconiat, M., Chumakov, I., Le Paslier, D., Berger, R. and Bernard, O.A. (1995) Blood 85, 36623670. [11] Raynaud, S.D., Cave, H., Baens, M., Bastard, C., Cacheux, V., Grosgeorge, J., Guidal-Giroux, C., Guo, C., Vilmer, E., Marynen, P. and Grandchamp, B. (1996) Blood 87, 2891-2899. [12] Keene, P., Mendelow, B., Pinto, M.R., Bezwoda, W., MacDougall, L., Falkson, G., Ruff, P. and Bernstein, R. (1987) Br. J. Haematol. 67, 25-31. [13] Srivastava, A., Boswell, H., Heerema, N.A., Nahreini, P., Lauer, R.D., Antony, A.S., Hoffman, R., Tricot, G.J. (1988) Cancer Genet. Cytogenet. 35, 61-71. [14] Berkowicz, M., Rosner, E., Rechavi, C., Mamon, Z., Neuman, Y., Ben-Bassat, I. and Ramot, B. (1991) Atypical chronic myelomonocytic leukemia with eosinophilia and translocation (5;12): A new assocation? Cancer Genet. Cytogenet. 51,277-278. [15] Lerza, R., Castello, G., Sessarego, M., Cavallini, D. and Pannacciulli, I. (1992) Br. J. Haematol. 82, 476-477. [16] Wessels, J.W., Fibbe, W.E., van der Keur, D., Landegent, J.E., van der Plas, D.C., den Ottolander, G.J., Roozendaal, K.J., Beverstock, G.D. (1993) t(5;12)(q31;p12) A clinical entity with features of both myeloid leukemia and chronic myelomonocytic leukemia. Cancer Genet. Cytogenet. 65, 7-11. [17] Golub, T.R., Barker, G.F., LoveU, M., Gilliland, D.G. (1994) Fusion of PDGF receptor beta to a novel ets-like gene, tel, in chronic myelomonocytic leukemia with t(5;12) chromosomal translocation. Cell 77, 307-316. [18] Papadopoulos, P., Ridge, S.A., Boucher, C.A., Stocking, C. and Wiedemann, L.M. (1995) The novel activation of ABL by fusion to an ets-related gene, TEL. Cancer Res. 55, 34-38. [19] Golub, T.R., Goga, A., Barker, G., Afar, D., McLaughlin, J., Boh|ander, S., Rowley, J., Witte, O., Gilliland, D.G. (1996) Oligomerization of the ABL tyrosine kinase by the ETS protein TEL in human leukemia. Mol. Cell Biol., in press. [20] Buijs, A., Sherr, S., van Baal, S., van Bezouw, S., van der Plas, D., van Kessel, A.D., Riegman, P., Deprez, R.L., Zwarthoff, E., Hagemeijer, A. and Grosveld, G. (1995) Oncogene 10, 1511-1519. [21] Deprez, R.H.L., Riegman, P.H.J., Groen, N.A., Warringa, U.L., van Biezen, N.A., Molijn, A.C., Bootsma, D., de Jong, P.J., Menon, A.G., Kley, N.A, Seizinger, B.R. and Zwarthoff, E.C. (1995) Oncogene 10, 1521-1528. [22] Romana, S.P., Piorel, H., Leconiat, M., Flexor, M.A., Mauchaffe, M., Jonveaux, P., Berger, R. and Bernard, O.A. (1995) Blood 86, 4263-4269. [23] Golub, T., McLean, T., Stegmaier, K., Ritz, J., Sallan, S., Neuberg, D., Gilliland, D.G. (1995) Blood 86 (Suppl 1), 597a. [24] Shurtleff, S., Buijs, A., Behm, F., Rubnitz, J., Raimondi, S., Hancock, M., Chan, G., Pui, C., Grosveld, G. and Downing, J. (1995) Leukemia 9, 1985-1989. [25] Hiebert, S.W., Sun, W., Davis, J.N., Golub, T.R., Shurtleff, S., Buijs, A., J.R.D., Grosveld, G., Roussel, M., Gilliland, D,G., Lenny, N. and Meyers, S. (1996) Mol. Cell. Biol., in press. [26] Duro, D., Bernard, O., Della Valle, V., Berger, R. and Larsen, C.J. (1995) Oncogene 11, 21-29. [27] Stegmaier, K., Takeuchi, S., Golub, T.R., Bohlander SK, C.R. B, Koeffler HP, Gilliland DG. (1996) Cancer Res. 56, 1413-1417.