Transcriptional activation is a key function encoded by MLL ... - Nature

Leukemia (2003) 17, 359–365  2003 Nature Publishing Group All rights reserved 0887-6924/03 $25.00 www.nature.com/leu

Transcriptional activation is a key function encoded by MLL fusion partners BB Zeisig, S Schreiner, M-P Garcıá-Cue´llar and RK Slany Department of Genetics, University of Erlangen, Staudtstrasse 5, 91058 Erlangen, Germany

Chromosomal translocations that fuse the mixed lineage leukemia gene (MLL) to a variety of unrelated partner genes are frequent in pediatric leukemias. The novel combination of genetic material leads to the production of active oncoproteins that depend on the contributions of both constituents. In a search for a common function amongst the diverse group of MLL fusion partners we constructed artificial fusions joining MLL with generic transactivator and repressor domains (acidic blob, GAL4 transactivator domain, Herpes simplex VP16 activation domain, KRAB repressor domain). Of all constructs tested, only MLL-VP16 was able to transform primary bone marrow cells and to induce a block of early myeloid differentiation like an authentic MLL fusion. Interestingly, the transformation capability of the artificial MLL fusions was correlated with the transcriptional potential of the resulting chimeric protein but it was not related to the strength of the isolated transactivation domain that was joined to MLL. These results prove for the first time that a general biological function – transactivation – might be the common denominator of many MLL fusion partners. Leukemia (2003) 17, 359–365. doi:10.1038/sj.leu.2402804 Keywords: leukemia; MLL; transcription; transactivator domain

Introduction Chromosomal translocations of the locus 11q23 that join the MLL gene (mixed lineage leukemia) with a variety of different fusion partners are the hallmark of many acute leukemias in infants (for a review see Refs 1 and 2). In contrast to the majority of other fusion oncoproteins, MLL forms chimeras with a puzzling number of non-related proteins. To date, the MLL locus has been found to be translocated to approximately 50 different genetic loci and at least 25 of the partner genes have been cloned and sequenced.3 MLL partners can be divided into two classes. The first group is mainly cytoplasmic and frequently associated with cytoskeleton-dependent signal transduction. Although the largest group by diversity, these translocations are rarely seen in patients and for each example only a few cases are known. The second group represents the majority of all 11q23 translocations. There, MLL is fused to nuclear proteins that resemble transcription factors. Three translocations within this group, the t(4;11), t(9;11) and the t(11;19) joining MLL with AF4, AF9 and ENL, respectively, make up for more than 75% of all observed cases. Despite the heterogeneity of the fusion partners, all known 11q23 translocations delete a large 3⬘ portion of the MLL gene and connect the remaining 5⬘ part in frame with the corresponding partner gene. This leads to the expression of a chimeric protein that has actively transforming properties. MLL itself is a human homologue of the epigenetic transcriptional regulator TRITHORAX (TRX) in Drosophila melanogaster.4–7 TRX is the prototype member of a genetically defined group of proteins that are part of the ‘cellular memory’. The TRX group is necessary for the maintenance of the

Correspondence: RK Slany, Department of Genetics, University of Erlangen, Staudtstrasse 5, 91058 Erlangen, Germany; Fax: 49 9131 8528526 Received 5 August 2002; accepted 2 October 2002

transcription of certain targets, eg the Hox cluster of homeobox-containing regulator genes. TRX group proteins act by altering the chromatin architecture, establishing a transcriptionally permissive, ‘open’ state of chromatin.8 Knockout studies corroborate a similar function also for mammalian MLL, as mice heterozygous for Mll have homeotic transformations of the skeleton and homozygous Mll-null embryos progressively lose Hox gene expression before they die around dpc 9.5.9 Supposedly, because Hox genes also play a key role in the regulation of hematopoietic development, Mll−/− embryos show a severe defect in fetal liver hematopoiesis and Mll−/− ES cells do not efficiently form blood cells upon in vitro differentiation.10,11 There is good evidence that the acquisition of novel properties by the combination of MLL with the fusion partners, rather than the loss of wild-type MLL function, leads to the generation of an active oncoprotein. In a series of knock-in studies it was clearly demonstrated that MLL needs a fusion partner to become leukemogenic, whereas a simple truncation did not cause leukemias in mice.12–14 Moreover, several structure– function studies testing the transforming capacity of MLL fusions in a bone marrow transformation assay found a critical contribution of the fusion partner.15–17 Whereas all studies agreed on the necessity of two critical DNA binding motifs in the MLL moiety, namely the AT hooks and the methyltransferase homology (MT) domain, the crucial function of the fusion partner is still unknown. In a knock-in experiment a fusion of Mll with the bacterial LacZ protein was sufficient to elicit leukemias in mice, albeit only after an extended latency period and with low penetrance.14 The authors of this report speculated that an intrinsic multimerization domain within LacZ supplies an essential function and they suggested that a dimerization of the truncated Mll is the basic oncogenic trigger. In other studies, it has been postulated that transcriptional transactivation might be the common denominator of several fusion partners and transactivator domains are indeed found in the three most frequent fusion partners AF4, AF9 and ENL.18 For ENL the 90 C-terminal amino acids not only constitute a transcriptional activator but they are also the minimal component necessary to convert MLL into a leukemogenic oncoprotein.17 The crucial domain of another MLL partner, ELL, is a protein–protein interaction domain that binds to EAF1 (ELL associated factor).16,19 EAF1, in turn, comprises a stretch of amino acids with homology to the transactivator domain of AF4.20 Finally, the MLL partners CBP and p300 are general cofactors in transcriptional activation.21,22 To clarify the general contribution of the MLL fusion partners for the function of the MLL chimeras we decided to test the influence of transcriptionally active motifs on the transformation capacity of the corresponding MLL fusion proteins. To this end we juxtaposed MLL with different generic transactivator and repressor domains. Here, we show that a strong transactivator can replace the function of ENL in the MLL fusion and that transformation is correlated with transactivation when measured in the context of the whole MLL chimeric protein.

Transactivation by MLL fusion partners BB Zeisig et al

360

Materials and methods

Plasmid construction Fusions of MLL with various generic transcriptional regulator domains were generated in the pMSCVneo retroviral vector backbone23 containing the 5⬘ MLL portion that is conserved in all known fusion genes. The respective transactivator/ repressor sequences were inserted in frame either at the unique PflMI (aa 1397 of MLL) or the first ApaLI (aa 1444 of MLL) restriction site. The acidic activation domain AD42 was released from the commercial two-hybrid vector pBAD42 (Invitrogen/Clontech, Palo Alto, CA, USA). The GAL4 activation domain was transferred from the vector pACT2 (Invitrogen/Clontech). The VP16 domain and ENL were taken from laboratory stocks and the KRAB repressor domain was a gift from F Rauscher III (The Wistar Institute, Philadelphia, PA, USA) (pAX3-KRAB). The same domains were also fused to the GAL4 DNA binding domain (amino acids 1 to 147) in the vector pSG424.24 The amino acids of the respective protein motifs fused to MLL and GAL4 are given in Figure 1a. Cloning details are available on request. All plasmid inserts were verified by sequencing.

Tissue culture and transfection procedures Human embryonic kidney cells (293T) were cultured in high glucose DMEM (Invitrogen/Life Technologies, Karlsruhe, Germany), pre-B REH cells were kept in RPMI1640 (Life Technologies). For protein expression the cells were transfected with a standard Ca-phosphate method. The ecotropic packaging cell line Phoenix was obtained from Gary Nolan (Stanford, CA, USA). The maintenance of Phoenix cells is outlined on the world wide web under: http://www.stanford.edu/group/nolan/. For luciferase reporter studies 293T cells were transfected by Fugene-6 liposomes (Roche, Penzberg, Germany) in 24well plates and REH cells were electroporated in RPMI containing 5 ␮g/ml DEAE dextran at 300 V and 925 ␮F in 4-mm cuvettes without pulse controller in a BioRad electroporator. In total, 1 ␮g of DNA was transfected in a ratio of expression plasmid to reporter of 9:1. Luciferase activity was determined with the luciferase assay system (Promega, Madison, WI, USA) according to the instructions of the manufacturer. Transduced bone marrow cells were kept either in MethoCult (M3234) methylcellulose medium (Stem Cell Technologies, Vancouver, Canada) or in RPMI. Recombinant mouse cytokines (Strathmann Biotech, Hannover, Germany) were added in the following concentrations: IL-3, IL-6, GM-CSF 10 ng/ml; SCF 100 ng/ml. All liquid media were supplemented with 10% bovine fetal calf serum (FCS Gold, PAA Laboratories, Austria) and penicillin/streptomycin.

Western blot and antibodies For the detection of MLL and GAL4 fusion proteins nuclear extracts were prepared from transfected 293T cells in a high salt elution buffer (500 mM NaCl, 20 mM Hepes pH 7.5, 0.5 mM EDTA, 0.1% Triton X-100, 0.5 mM sodium vanadate, 2 mM NaF, 2 mM DTT, 0.2 mM PMSF, 20 ␮g/ml leupeptin and 40 ␮g/ml pepstatin A). After separation on standard SDS-PAGE gels the proteins were blotted on to nitrocellulose in 10 mM Leukemia

Figure 1 Construction and properties of artificial MLL and GAL4 fusion proteins. (a) Schematic representation of the constructs used in this study. The indicated transcriptional regulators were fused either to MLL or to the GAL4 DNA binding domain in the vectors pMSCVneo (retroviral backbone) and pSG424. The numbers indicate the amino acid positions of the respective proteins. ENL, minimal transactivation domain of the authentic MLL fusion partner ENL; AD42, acidic transactivator sequence; GAL4a, GAL4 activation domain; VP16, transactivation domain of the Herpes simplex transactivator protein; KRAB, repressor domain of a KRAB type Zn-finger protein. (b) Expression of MLL and GAL4 fusions. The correct expression of the respective fusion proteins was verified by immunoblot in nuclear extracts of transfected 293T cells. Blots were probed either with a monoclonal antibody directed against MLL, or with an antibody against the GAL4 DNA binding domain. The position of molecular size markers as well as the transfected construct is indicated. MLL, MLL portion without fused partner protein. (c) Transcriptional activity of the GAL4 fusion proteins. To prove the functionality of the GAL4 fusion constructs 0.9 ␮g of the respective plasmid was cotransfected into 293T cells (left panel) or REH cells (right panel) together with 0.1 ␮g of a reporter plasmid. A luciferase gene under the control of the SV40 minimal promoter preceded by a triple repeat of GAL4 recognition sites was used as a reporter. Luciferase activities were normalized to protein content and are given as relative activities compared to the background level obtained with cotransfection of empty vector alone. Means and standard deviations of triplicates are given.

CAPS, 0.01% SDS, 1% methanol for MLL fusion proteins and 10 mM CAPS, 20% methanol for GAL4 fusion proteins. The detection was done with monoclonal antibodies against MLL25 or the GAL4 DNA binding domain (Santa Cruz Biotech, Santa Cruz, CA, USA) in a standard ECL protocol. Antibodies for FACS analysis (isotype control, c-kit, Gr-1, Mac-1/CD11b) were purchased from BD Biosciences (Heidelberg, Germany) and used according to the recommendations of the manufacturer.

Retroviral transduction of mouse primary hematopoietic cells High-titer retrovirus supernatants were produced by transient transfection of the packaging cell line Phoenix-E by a standard Ca-phosphate precipitation method. The viral titers were all


in the range of 5 × 106 cfu/ml. The retroviral transduction of primary hematopoietic cells was done according to Ref. 26. In short, bone marrow cells (BMCs) enriched in non-cycling hematopoietic precursors were recovered from 5-fluorouracil (150 mg/kg) primed Balb/C mice. Previous to infection the BMCs were activated for 24 h by a cytokine cocktail (IL-3, IL6, SCF). After infection by spinoculation (3 h, 35°C, 2500 g) the infected cells were returned to activation medium as above. Following overnight activation duplicates of 1 × 104 infected cells each were plated in 24-well plates under G418 drug selection (1mg/ml) in 1 ml of MethoCult murine methylcellulose medium with the addition of SCF, IL-3, IL-6 and GMCSF. For visualization colonies were stained by addition of 100 ␮l of 1 mg/ml p-iodonitrotetrazolium violet in PBS (INT; Sigma). INT is converted by living cells to a brown–violet insoluble tetrazolium salt. The colonies in the remaining well were resuspended in medium, counted and replated as above for two more rounds.

RT-PCR, Southern blot Total RNA was isolated by ion-exchange chromatography with kits from Qiagen (Hilden, Germany) according to the recommended protocols. MLL fusion-specific RNA was detected by RT-PCR after reverse transcription of total RNA with a random hexamer primer. The primers for amplification were designed to span the breakpoint between the fusion partners: MLL common primer: gcaaacagaaaaaagtggctccccg; ENL reverse primer: tcatgtggccacggcctccag; VP16 reverse primer: tccgctcgagctacccaccgtactcgtcaattc. Primers for amplification of actin: cctgggcatggagtcctg and ggagcaatgatcttgagcttc. Southern blotting was done on DNA isolated by phenol/chloroform extraction and blotted under alkaline conditions according to standard procedures.

allowing for an efficient transfer into primary bone marrow cells. To enable an assessment of the transcriptional properties when directly bound to promoter DNA the respective domains were also fused to the DNA binding moiety of the GAL4 protein (aa 1 to 147). The structure of the plasmid constructs is depicted schematically in Figure 1a. The production of proteins with the appropriate molecular weight was verified by immunoblotting of nuclear extracts from transiently transfected human embryonic kidney 293T cells. Either a monoclonal IgM type antibody directed against an N-terminal MLL epitope, or a monoclonal IgG recognizing the GAL4 DNA binding moiety was used to reveal the presence of fusion proteins displaying, within the resolution limits of SDS-PAGE, the expected size (Figure 1b). The functionality of the cloned transcriptional regulator domains was controlled in a transient reporter gene assay. A luciferase gene under control of the minimal SV40 promoter, preceded by a triplicate of the GAL4 recognition sequences was cotransfected with the individual GAL4 fusion plasmid into 293T (human embryonic kidney cells) and REH (pre-B cells). Twenty-four hours after transfection the luciferase levels were determined and normalized to protein content. As Figure 1c shows, a strong transactivation activity could be detected for all four domains tested. In 293T cells GAL4-VP16 achieved a 290-fold increase in luciferase levels above background. The other domains accomplished a 50-fold (AD42), 16-fold (GAL4) and 10-fold transactivation (ENL). The GAL4-KRAB protein repressed luciferase activities five-fold to about 20% of the base level obtained after cotransfection of empty expression vector. Similar results were obtained in REH cells with the exception that in these cells the ENL transactivaton domain was slightly more potent than the GAL4 transactivation domain.

361

Transforming potential of artificial MLL fusion proteins Results

Construction of artificial MLL fusion proteins In order to investigate the role of transcriptionally active domains in MLL fusion partners we constructed a series of artificial MLL fusion proteins. To this end the cDNA corresponding to the conserved N-terminal portion of MLL was joined in frame with three different transactivator domains and one repressor motif. A heterologous sequence from E. coli (AD42, aa 1 to 88) encoding an ‘acidic domain’ and the transactivator domain of the yeast regulator GAL4 (aa 768 to 881) were transferred from commercial two-hybrid vectors to MLL. The sequence encoding the amino acids 402 to 479 of the alpha trans-inducing protein of human herpes virus 1 (accession No. IXBE1F) known as VP16 was taken from a plasmid laboratory stock. Care was taken to include the complete VP16 transactivation domain, as it was noted that several available ‘VP16 plasmids’ contained only a subdomain. Fusions of MLL with these shorter versions of VP16 have been described to be non-transforming.27 Additionally, MLL was fused to the KRAB domain (aa 1 to 74) of the human zinc finger protein 10 (accession No. p21506), a general repressor and as a positive control to the minimal transactivation domain of the authentic MLL fusion partner ENL (accession No. L04285, aa 430 to 590). A similar vector expressing only the N-terminal portion of MLL without any added sequence served as negative control. All constructs were engineered in the retroviral backbone pMSCVneo

The transforming capability of MLL fusion proteins can be tested in a bone marrow colony formation assay. This test exploits the fact that oncogenic MLL fusion proteins are able to block hematopoietic differentiation.28 Hematopoietic precursor cells can thus be arrested in a maturation state where the cells are still self-renewing. Upon serial replating in methylcellulose medium supplemented with cytokines (IL-3, IL-6, SCF, GM-CSF), these cells will continuously form colonies.26 Normal cells or cells expressing non-functional MLL fusions under the same conditions undergo a program of differentiation accompanied by a progressive loss of their proliferative capacity. Therefore, they will form colonies only in the initial but not in subsequent rounds of plating. It has been previously shown that this test is also a good predictor of the in vivo transforming capacity of MLL fusion proteins.15,16,29 To assess the oncogenic properties of the artificial MLL fusion proteins bone marrow cells from 5-fluorouracil conditioned Balb/C mice were retrovirally transduced with the respective constructs and plated in methocel. Selection for neomycin resistance during the first round of plating ensured that only transduced cells survived. As can be seen in Figure 2a (upper panel) similar numbers of colonies were obtained in the first round of plating for all retroviral constructs used. This confirmed that approximately equal numbers of infected cells were plated and that the proteins expressed were not toxic. In the third round of plating, however, only cells transduced with MLL-ENL and MLL-VP16 showed a positive readout. The cells expressing MLL-GAL4AD, MLL-AD42, MLLLeukemia


362

Figure 2 Bone marrow transformation assay of MLL fusion proteins. (a) Primary bone marrow cells from 5-fluorouracil primed mice were retrovirally transduced with vectors expressing the individual MLL fusion proteins. After infection the cells were plated in methocel medium in the presence of IL-3, IL-6, SCF and GM-CSF. The upper panel shows stained colonies that developed in the first plating round. The lower panel depicts the situation after two rounds of replating. (b) MLL-ENL (upper panel) and MLL-VP16 (lower panel) transduced cells, after cultivation in liquid medium. A May–Gru¨nwald–Giemsa stained cytospin preparation is shown to the left. The original magnification was × 63. The surface marker expression was analyzed by FACS analysis and is shown to the right. The staining results for the c-kit, Gr-1 and Mac-1 (=CD11b) markers are shown (filled graph) in comparison to an isotype control (open).

KRAB or the N-terminus of MLL no longer formed colonies (Figure 2a, lower panel). This experiment was independently repeated five times with essentially identical results. The MLL-ENL and MLL-VP16 transduced cells could be cultured in liquid medium supplemented with the same cytokines as present in the methocel medium (IL-3, IL-6, SCF, GM-CSF). Approximately 2 to 3 weeks after explantation from methocel the respective cells were analyzed for morphology and surface marker expression (Figure 2b). As has been previously seen with cells transformed by other MLL derivatives,15,16,19,26 MLLENL transduced cells were early myeloblasts as judged by phenotype. The myeloid nature of the cell population was corroborated by the presence of the surface markers Mac-1 (CD11b) and Gr-1 (Figure 2b, upper panel). MLL-VP16 transduced cells had comparable properties, with the difference that a significant proportion of the cells showed signs of a more advanced maturation state like ‘donut’-shaped nuclei and higher surface expression levels of Gr-1. Additionally a side population of Gr-1neg, Mac-1neg cells was consistently Leukemia

present in the MLL-VP16 populations (Figure 2b, lower panel). It was also repeatedly noted that MLL-VP16 cells proliferated more slowly than the MLL-ENL transduced population (not shown). To test for the presence of the respective retroviruses Southern blot experiments were performed. DNA from three different, independently derived MLL-ENL or MLL-VP16 infected cell populations was isolated, digested with BamHI and analyzed with a probe specific for the N-terminus of MLL. DNA from the cell line M1 was used as a control for the hybridization to the endogenous mouse Mll locus. The localization of the probe and the restriction sites within the retroviral construct is schematically depicted in Figure 3a. Consistent with previous results, the hybridization pattern indicated multiple integration events and as inferred from band intensity the populations were either mono- or pauciclonal. Of note, each cell population displayed a unique configuration of integration sites with no discernable preference for particular loci. The presence of unrearranged virus and the identity of the


with relaxed specificity and therefore can recruit the fusion protein to different DNA templates.30,31 Also a SV40 minimal promoter can be activated by MLL-ENL. Therefore it was possible to investigate further the apparent discrepancy between the strength of the individual transactivation domains (see Figure 1c) and their transformation potential as an MLL fusion by checking the overall transcriptional capability of the chimeric protein on the same reporter construct as used before. The use of the same promoter–reporter combination ensured that differences in the observed transactivation potency were only due to the different nature of the transactivator protein. The pre-B cell line REH was chosen for these experiments because it is permissive for the transactivation by MLL fusion proteins.31 Furthermore, REH represents an early lymphatic cell type and it is therefore more related to the actual situation in the leukemias. When assayed in this system (Figure 4a), a perfect corre-

363

Figure 3 Presence of integrated retroviral DNA and expression of fusion gene RNA in transduced cells. (a) Genomic DNA was isolated from three independently derived cell populations transduced either with MLL-ENL or MLL-VP16 as indicated. As a control for endogenous hybridization patterns genomic DNA from the mouse cell line M1 was applied (designated ‘C’). The DNA was digested with BamHI and hybridized with an MLL-specific probe as schematically depicted (probe = black bar). This digest/probe combination is able to distinguish different integration sites. The positions of molecular size markers is indicated. (b) Southern blot of NheI digested DNA hybridized with an ENL-specific probe. (c) The same blot as in (b) reprobed without previous stripping for VP16. (d) Results of an RTPCR experiment to detect the expression of fusion gene-specific transcripts. The result of only one representative experiment is shown. In the top panel breakpoint spanning primer pairs were used to amplify target sequences in reverse transcribed cDNA samples either from MLL-ENL or MLL-VP16 transduced cells as indicated. A template without addition of cDNA (H2O) served as a control. Specific amplification products are only seen in the cognate samples. The integrity of the RNA and possible DNA contaminations were checked in the lower panel with a primer pair amplifying actin cDNA sequences. No amplification can be seen in the absence of reverse transcriptase indicating a pure RNA without traces of DNA.

encoded MLL fusion was ascertained by probing an NheI digest of the DNA for ENL (Figure 3b) and then reprobing the same blot for VP16 (Figure 3c). Finally, the expression of the mRNA encoding the fusion transcripts could be verified by RT-PCR in all tested samples. Figure 3d shows one representative result. Primer pairs spanning the breakpoint were designed for the MLL-ENL and MLLVP16 constructs. When tested on cDNA reverse transcribed from MLL-ENL or MLL-VP16 cells, only cognate primer cDNA pairs yielded the expected amplification product (Figure 3d, upper panel). An RT-PCR with actin-specific primers performed on samples in the presence and absence of reverse transcriptase served as a control for RNA integrity and lack of DNA contamination (Figure 3d, lower panel).

Transactivation potential of artificial MLL fusion proteins It has been shown that MLL fusion proteins are themselves pleiotropic activators of transcription that transactivate a variety of unrelated promoters. This activity is dependent on DNA binding motifs within the MLL moiety that bind to DNA

Figure 4 Transcriptional activity of artificial MLL fusion proteins. Expression constructs for proteins that fuse MLL to the indicated moieties were electroporated into the pre-B cell line REH together with either a luciferase reporter controlled by the SV40 minimal promoter (a) or by the murine Hoxa7 promoter (b). Both promoters have been previously shown to be responsive to MLL-ENL-mediated transactivation (see text). The results show the means and standard deviations of three independent experiments and are plotted on a logarithmic scale. Leukemia


364

lation between transactivation and transformation (as measured by colony formation) was revealed. Only MLL-ENL and MLL-VP16 were capable of transactivating the plasmid template (31-fold and 25-fold, respectively). Neither the MLLAD42 protein, nor the MLL-KRAB fusion showed any significant influence on luciferase levels and MLL-GAL4AD was only a very weak activator (two-fold transactivation). A similar result was obtained by testing the series of artificial MLL fusions on a luciferase reporter driven by the murine Hoxa7 promoter (Figure 4b). As described previously31 this particular construct is MLL-ENL responsive and constitutes the most physiologically relevant MLL target that is available to date. Also for this promoter, MLL-ENL displayed by far the strongest transactivation capacity (69-fold) followed by the MLL-VP16 protein (20-fold transactivation). MLL-GAL4AD and MLL-AD42 had a minor effect (five-fold and six-fold, respectively) and the coexpression of MLL-KRAB did not alter measurable luciferase levels. Discussion In this report we provide the first evidence that a generic transactivator function can replace the authentic MLL fusion partner ENL. However, only the strong VP16 transactivator with a transactivation potency an order of magnitude higher than that of ENL could achieve this fact. When juxtaposed to less powerful activation domains or a repressor, MLL did not acquire transforming properties. Consequently, only the transactivation potential of the complete fusion protein and not the strength of the isolated transactivation domain was correlated to a positive score in the transformation test. These results corroborate the hypothesis, that an intrinsic transactivation potential is the common denominator of many MLL fusion partners. Regions with a positive effect on transcription have been found for example in AF4, AF9, ENL and ELL.16,18,19 For ENL and ELL it was explicitly shown that the minimal portion of the protein that is necessary to produce a transforming MLL fusion coincides with the transactivator domain. Mutations that eliminated the transforming ability also affected the transactivation function and vice versa.16,17,19 Moreover the transactivator domain of ELL binds to ELL-associated factor (EAF-1), a protein that contains stretches of amino acids homologous to the transactivation domain of AF4.20 Further support for a decisive role of transcriptional activation in the mechanism of MLL-mediated transformation comes from cases of leukemia where MLL is fused directly to the transcriptional coactivators p300 and CBP.21,22 It could be demonstrated that the portions of CBP that are responsible for the chromatin-related coactivator properties also constitute the essential contributions towards a transforming MLL fusion protein.15 The oncogenic activity of MLL derivatives is a consequence of a combination of the N-terminal MLL moiety with a fusion partner. Therefore a heterologous chimera joining the GAL4 DNA binding domain with MLL partners does not truly reflect the situation of the MLL fusion protein and might be an unsuitable model to judge the correlation between transcription and transformation in the corresponding MLL fusion. As our results demonstrate, the absolute strength of a transcriptional activator when measured in a GAL4 fusion is not necessarily predictive for the overall effect on transcription of the analogous MLL fusion. Nevertheless, it remains remarkable that a 20-fold to 30fold more potent transactivator is necessary to functionally

Leukemia

replace ENL in conjunction with MLL. Although it cannot be excluded that this is due to the fact that the bulky MLL protein interferes with the function of some of the transactivators tested, it seems more likely that there is a more fundamental difference in the way in which ENL achieves transactivation in combination with MLL. It is known that the minimal essential domains of ENL as well as of AF9 recruit the novel polycomb protein Pc3.32,33 Two more fusion partners, ABI1 and AF10, are also connected to ENL either by a direct interaction34 or via the GAS41 protein, an ENL homologue.35 Polycomb proteins in turn are known to be members of chromatin remodelling machines that have been connected mostly, but not exclusively, with transcriptional silencing.36 It might be possible that, by unknown mechanisms, the assembly of an MLLENL/Polycomb compound complex leads to chromatin structure changes that cause a persistent and strong activation of target genes. This would fit with the fact that chromatin structure changes are also the underlying mechanism in CBP and p300 fusion partners achieving their transactivation effect. To our knowledge, none of the other transactivation domains used in this study has been connected to the Polycomb machinery and the available data suggest that they primarily work by directly recruiting the RNA polymerase II complex. Therefore these domains might have to be considerably more potent to accomplish the same outcome as ENL in connection with MLL. Acknowledgements We are grateful to Frank Rauscher III for the gift of the KRAB repressor plasmid. The authors wish to thank Renate Zimmermann for technical assistance and Georg Fey for continuous support. This work was supported by grant SFB466/C7 and partially by grants SFB473/B10 and SL27/4–1 from the DFG. RKS is a recipient of a Ria Freifrau-von-Fritsch Stiftung career development award. References 1 Dimartino JF, Cleary ML. MLL rearrangements in haematological malignancies: lessons from clinical and biological studies. Br J Haematol 1999; 106: 614–626. 2 Ayton PM, Cleary ML. Molecular mechanisms of leukemogenesis mediated by MLL fusion proteins. Oncogene 2001; 20: 5695– 5707. 3 Huret JL, Dessen P, Bernheim A. An atlas of chromosomes in hematological malignancies. Example: 11q23 and MLL partners. Leukemia 2001; 15: 987–989. 4 Djabali M, Selleri L, Parry P, Bower M, Young BD, Evans GA. A trithorax-like gene is interrupted by chromosome 11q23 translocations in acute leukaemias. Nat Genet 1992; 2: 113–118. 5 Gu Y, Nakamura T, Alder H, Prasad R, Canaani O, Cimino G, Croce CM, Canaani E. The t(4;11) chromosome translocation of human acute leukemias fuses the ALL-1 gene, related to Drosophila trithorax, to the AF-4 gene. Cell 1992; 71: 701–708. 6 Tkachuk DC, Kohler S, Cleary ML. Involvement of a homolog of Drosophila trithorax by 11q23 chromosomal translocations in acute leukemias. Cell 1992; 71: 691–700. 7 Ziemin-van der Poel S, McCabe NR, Gill HJ, Espinosa R III, Patel Y, Harden A, Rubinelli P, Smith SD, LeBeau MM, Rowley JD. Identification of a gene, MLL, that spans the breakpoint in 11q23 translocations associated with human leukemias. Proc Natl Acad Sci USA 1991; 88: 10735–10739. 8 van Lohuizen M. The trithorax-group and polycomb-group chromatin modifiers: implications for disease. Curr Opin Genet Dev 1999; 9: 355–361.


9 Yu BD, Hess JL, Horning SE, Brown GA, Korsmeyer SJ. Altered Hox expression and segmental identity in Mll-mutant mice. Nature 1995; 378: 505–508. 10 Yagi H, Deguchi K, Aono A, Tani Y, Kishimoto T, Komori T. Growth disturbance in fetal liver hematopoiesis of Mll-mutant mice. Blood 1998; 92: 108–117. 11 Hess JL, Yu BD, Li B, Hanson R, Korsmeyer SJ. Defects in yolk sac hematopoiesis in Mll-null embryos. Blood 1997; 90: 1799–1806. 12 Corral J, Lavenir I, Impey H, Warren AJ, Forster A, Larson TA, Bell S, McKenzie AN, King G, Rabbitts TH. An Mll-AF9 fusion gene made by homologous recombination causes acute leukemia in chimeric mice: a method to create fusion oncogenes. Cell 1996; 85: 853–861. 13 Dobson CL, Warren AJ, Pannell R, Forster A, Lavenir I, Corral J, Smith AJ, Rabbitts TH. The Mll-AF9 gene fusion in mice controls myeloproliferation and specifies acute myeloid leukaemogenesis. EMBO J 1999; 18: 3564–3574. 14 Dobson CL, Warren AJ, Pannell R, Forster A, Rabbitts TH. Tumorigenesis in mice with a fusion of the leukaemia oncogene Mll and the bacterial lacZ gene. EMBO J 2000; 19: 843–851. 15 Lavau C, Du C, Thirman M, Zeleznik-Le N. Chromatin-related properties of CBP fused to MLL generate a myelodysplastic-like syndrome that evolves into myeloid leukemia. EMBO J 2000; 19: 4655–4664. 16 Luo RT, Lavau C, Du C, Simone F, Polak PE, Kawamata S, Thirman MJ. The elongation domain of ELL is dispensable but its ELL-associated factor 1 interaction domain is essential for MLL-ELLinduced leukemogenesis. Mol Cell Biol 2001; 21: 5678–5687. 17 Slany RK, Lavau C, Cleary ML. The oncogenic capacity of HRXENL requires the transcriptional transactivation activity of ENL and the DNA binding motifs of HRX. Mol Cell Biol 1998; 18: 122–129. 18 Prasad R, Yano T, Sorio C, Nakamura T, Rallapalli R, Gu Y, Leshkowitz D, Croce CM, Canaani E. Domains with transcriptional regulatory activity within the ALL1 and AF4 proteins involved in acute leukemia. Proc Natl Acad Sci USA 1995; 92: 12160–12164. 19 DiMartino JF, Miller T, Ayton PM, Landewe T, Hess JL, Cleary ML, Shilatifard A. A carboxy-terminal domain of ELL is required and sufficient for immortalization of myeloid progenitors by MLL-ELL. Blood 2000; 96: 3887–3893. 20 Simone F, Polak PE, Kaberlein JJ, Luo RT, Levitan DA, Thirman MJ. EAF1, a novel ELL-associated factor that is delocalized by expression of the MLL-ELL fusion protein. Blood 2001; 98: 201– 209. 21 Ida K, Kitabayashi I, Taki T, Taniwaki M, Noro K, Yamamoto M, Ohki M, Hayashi Y. Adenoviral E1A-associated protein p300 is involved in acute myeloid leukemia with t(11;22)(q23;q13). Blood 1997; 90: 4699–4704. 22 Sobulo OM, Borrow J, Tomek R, Reshmi S, Harden A, Schlegelberger B, Housman D, Doggett NA, Rowley JD, Zeleznik-Le NJ. MLL is fused to CBP, a histone acetyltransferase, in therapy-related

23 24 25 26 27

28

29

30

31 32

33

34

35

36

acute myeloid leukemia with a t(11;16)(q23;p13.3). Proc Natl Acad Sci USA 1997; 94: 8732–8737. Hawley RG, Lieu FH, Fong AZ, Hawley TS. Versatile retroviral vectors for potential use in gene therapy. Gene Ther 1994; 1: 136–138. Sadowski I, Ptashne M. A vector for expressing GAL4(1–147) fusions in mammalian cells. Nucleic Acids Res 1989; 17: 7539. Butler LH, Slany R, Cui X, Cleary ML, Mason DY. The HRX protooncogene product is widely expressed in human tissues and localizes to nuclear structures. Blood 1997; 89: 3361–3370. Lavau C, Szilvassy SJ, Slany R, Cleary ML. Immortalization and leukemic transformation of a myelomonocytic precursor by retrovirally transduced HRX-ENL. EMBO J 1997; 16: 4226–4237. So CW, Cleary ML. MLL-AFX requires the transcriptional effector domains of AFX to transform myeloid progenitors and transdominantly interfere with forkhead protein function. Mol Cell Biol 2002; 22: 6542–6552. Schreiner S, Birke M, Garcia-Cuellar MP, Zilles O, Greil J, Slany RK. MLL-ENL causes a reversible and myc-dependent block of myelomonocytic cell differentiation. Cancer Res 2001; 61: 6480–6486. Lavau C, Luo RT, Du C, Thirman MJ. Retrovirus-mediated gene transfer of MLL-ELL transforms primary myeloid progenitors and causes acute myeloid leukemias in mice. Proc Natl Acad Sci USA 2000; 97: 10984–10989. Birke M, Schreiner S, Garcia-Cuellar MP, Mahr K, Titgemeyer F, Slany RK. The MT domain of the proto-oncoprotein MLL binds to CpG-containing DNA and discriminates against methylation. Nucleic Acids Res 2002; 30: 958–965. Schreiner SA, Garcia-Cuellar MP, Fey GH, Slany RK. The leukemogenic fusion of MLL with ENL creates a novel transcriptional transactivator. Leukemia 1999; 13: 1525–1533. Garcia-Cuellar MP, Zilles O, Schreiner SA, Birke M, Winkler TH, Slany RK. The ENL moiety of the childhood leukemia-associated MLL-ENL oncoprotein recruits human Polycomb 3. Oncogene 2001; 20: 411–419. Hemenway CS, de Erkenez AC, Gould GC. The polycomb protein MPc3 interacts with AF9, an MLL fusion partner in t(9;11)(p22;q23) acute leukemias. Oncogene 2001; 20: 3798– 3805. Garcia-Cuellar MP, Schreiner SA, Birke M, Hamacher M, Fey GH, Slany RK. ENL, the MLL fusion partner in t(11;19), binds to the cAbl interactor protein 1 (ABI1) that is fused to MLL in t(10;11). Oncogene 2000; 19: 1744–1751. Debernardi S, Bassini A, Jones LK, Chaplin T, Linder B, de Bruijn DR, Meese E, Young BD. The MLL fusion partner AF10 binds GAS41, a protein that interacts with the human SWI/SNF complex. Blood 2002; 99: 275–281. Brock HW, van Lohuizen M. The Polycomb group-no longer an exclusive club? Curr Opin Genet Dev 2001; 11: 175–181.

365

Leukemia