Molecular pathways driven by ETO2-GLIS2 in aggressive pediatric leukemia

ABSTRACT

The ETO2-GLIS2 fusion oncoprotein is associated with poor prognosis pediatric acute megakaryoblastic leukemia. Recently, we observed that ETO2-GLIS2 controls enhancers activity at genes regulating haematopoietic progenitor self-renewal and differentiation toward the megakaryocytic lineage. We also showed that targeting ETO2-GLIS2 complex stability inhibits these properties and may represent a novel therapeutic strategy.

Acute leukemia are defined by the accumulation in the bone marrow of highly proliferative cells that are blocked in differentiation leading to haematopoietic failure. Acute myeloid leukemia (AML) regroup heterogeneous cases classified according to the phenotype and the molecular alterations found in the blasts. AML with blasts presenting megakaryocytic morphology and markers such as the cell surface glycoproteins GPIIb (a.k.a. CD41) and/or GPIb (a.k.a. CD42) are named acute megakaryoblastic leukemia (AMKL). AMKL more frequently affect young children (4–15% of newly diagnosed AML) than adults (1% of AML). In pediatric cases, 2 distinct entities have been described: 1- the AMKL arising in the context of Down syndrome (DS-AMKL) and 2- the de novo AMKL. DS-AMKL result from a multi-steps process starting from the predisposing role of the Trisomy 21 followed by acquired mutations in the erythroid/megakaryocyte transcription factor GATA binding protein 1 (GATA1), and in epigenetic or signaling pathways regulators.¹ Genomic profiling of de novo AMKL cases identified, in most cases, chromosomal rearrangements leading to the expression of fusion oncoproteins. The most frequent is the inversion of the chromosome 16 [inv(16)(p13.3q24.3)] present in 18–27% of patients, and leads to the fusion of the CBFA2/RUNX1 translocation partner 3 (CBFA2T3 a.k.a. ETO2) and the GLIS family zinc finger 2 (GLIS2) genes resulting in the expression of the ETO2-GLIS2 fusion oncoprotein. De novo AMKL with ETO2-GLIS2 are associated with the worst prognosis among all pediatric AMKL. Only few additional mutations have been identified in this subgroup suggesting that ETO2-GLIS2 is most likely the main driver oncogene.

ETO2-GLIS2 fusions encode most of ETO2 and GLIS2 proteins. In other AML subgroups, ETO2 and its paralog RUNX1 translocation partner 1 (RUNX1T1 a.k.a. ETO) are involved in t(16;21)(q24;q22) and t(8;21)(q22;q22) translocations encoding RUNX1-ETO2 and RUNX1-ETO respectively, indicative of the wide implication of ETO proteins in leukemia development. ETO proteins are key components of large transcription factor complexes that include GATA and RUNX factors to control haematopoietic development, including the erythroid and megakaryocytic differentiation. They share several Nervy Homologous Regions (NHR) domains involved in oligomerization and protein interactions including the NHR2 domain that remains present in all fusion proteins. ETO proteins are generally associated with transcriptional repression through the recruitment of nuclear corepressors such as the SIN3 transcription regulator family member A (SIN3A) or the Nuclear receptor co-repressors (NCOR), or numerous histones deacetylases (HDAC).² GLIS2 is a member of the GLI- family of transcription factors that bind DNA through zinc finger domains. Glis2 requirement for haematopoietic stem cell repopulation in mice has been recently suggested,³ however it is not expressed in differentiating haematopoietic cells suggesting that its fusion with ETO2 leads to ectopic GLIS2 activity.

We have recently uncoupled ETO2 and GLIS2 contribution to leukemic transformation. Our data indicate that GLIS2 drives the megakaryocytic identity of transformed cells in a DNA-binding dependent fashion, while both ETO2 and GLIS2 impose aberrant self-renewal properties (see F1).⁴ Among the specific transcriptional signature of ETO2-GLIS2, we identify a functional imbalance between E-26 (ETS) and GATA master regulators of normal hematopoiesis. Indeed, ETO2-GLIS2 leads to the strong downregulation of the GATA binding protein 1 (GATA1) and the overexpression of the ETS-related gene (ERG). GATA1 is hematopoiesis-specific, positively controls megakaryocytic differentiation and limits proliferation of erythro-megakaryocytic progenitors.⁴ ERG is essential for haematopoietic stem cell maintenance as demonstrated by knockout mice that died in utero due to a defect in hematopoiesis. Moreover ERG is associated with poor prognosis in numerous cancers including leukemia where it is found translocated or overexpressed.⁵ In ETO2-GLIS2 AMKL, ERG is essential for self-renewal gene expression (e.g. KIT) and for cell survival.¹⁰ Interestingly, both genes are also altered in DS-AMKL through different mechanisms. Indeed, trisomy of ERG, located on chromosome 21, cooperates with acquired mutations in GATA1 that are found in nearly all DS-AMKL cases. Therefore, an imbalance in ETS/GATA factor activity may represent a common mechanism of transformation in genetically distinct AMKL subtypes.

Through ChIP-seq experiments, we show that ETO2-GLIS2 is located at loci known to be bound by ETO2 complexes in normal megakaryocytes and down-regulates associated genes expression (e.g., GATA1). The fusion also binds novel loci that are rather associated with transcriptional upregulation (e.g., KIT, ERG). Importantly, ETO2-GLIS2, along with ERG, binds many super-enhancers (SE) specifically identified in ETO2-GLIS2 AMKL patient cells. Super-enhancers are complex regulatory circuitries defined by large chromatin regions that have high occupancy of transcription factors and control the expression of master regulators or oncogenes.⁶ Most SE-associated genes (e.g., ERG and KIT) are strongly upregulated by the fusion. As ERG is both upregulated by the fusion and part of super-enhancers complexes, we proposed that this creates a feed-forward loop reinforcing the ETO2-GLIS2 transcriptional network. As seen in prostate cancer, ERG may act as a pioneer factor that open chromatin at essential self-renewal and megakaryocytic genes and impose leukemic transformation. How ETO2-GLIS2 and ERG precisely reorganize the chromatin landscape and recruit transcriptional complexes at enhancer elements remains to be investigated. Importantly, several molecular mechanisms could be shared with other poor prognosis fusions found in AML involving RUNX⁷ or the mixed lineage leukemia gene (MLL).^8,9 For ETO2-GLIS2, our data indicate that the strong transcriptional program results from its dimerization and its interaction with wild-type ETO2 complexes. Indeed, interference with these complexes through the expression of a small peptide (homologous to the NHR2 domain of ETO2) corrects the ERG/GATA1 imbalance, globally reverses enhancers activity, induces megakaryocytic differentiation and abrogates in vivo maintenance of human leukemic blasts in xenograft models. Therefore, targeting ETO2-GLIS2 complex integrity may be a new era of investigation to develop targeted therapeutic strategies for this poor prognosis pediatric leukemia.

Figure 1.

The ETO2-GLIS2 fusion drives the acquisition of leukemic properties. The ETO2-GLIS2 fusion protein is part of multimeric transcription factor complex that binds through both GLIS2 and/or ETO2 complexes at promoters and enhancers to control gene expression (KIT and ERG upregulation and GATA1 downregulation) driving leukemic properties.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Acknowledgments

We are grateful to team members for scientific discussions and apologize to our many colleagues whose work could not be cited due to space constraints.

This work was supported by Institut National Du Cancer (PLBIO-2014–176), Fédération Enfants et Santé and Société Française de lutte contre les Cancers et les Leucémies de l'Enfant et l'Adolescent (SFCE : CAMELIAT project), Association Laurette Fugain (ALF-2015/13), a José Carreras EHA award (DJCLS-EHA-2009-F-09/02), Fondation ARC, Fondation Gustave Roussy, SIRIC-SOCRATE (INCa-DGOS-INSERM 6043), Fondation pour la Recherche Médicale (FRM-ING20150532273), Cancéropôle Ile de France (to CéLo and Emergence 2015), Fondation de France (FdF-00057925), Lady Tata Foundation (R14120LL), Gustave Roussy Genomic Core Facility – TA2014. TM is a Equipe Labellisée LIGUE principal investigator.