Transcriptome Sequence Analysis of Pediatric Acute Megakaryoblastic Leukemia Identifies an Inv(16)(p13.3;q24.3)-Encoded CBFA2T3-GLIS2 Fusion Protein As a Recurrent Lesion in 39% of Non-Infant Cases: A Report from the St. Jude Children’s Research Hospital – Washington University Pediatric Cancer Genome Project
Tanja A. Gruber,Amanda Larson Gedman,Huy Q. Ta,Jinghui Zhang,Cary Koss,Shann-Ching Chen,Xiaoping Su,Vedant Gupta,Stacey Ogden,Anna K. Andersson,John Easton,Jianmin Wang,Michael Rusch,Li Ding,Giovanni Cazzaniga,Andrea Biondi,Steven M. Kornblau,Farhad Ravandi,Hagop M. Kantarjian,Stephen D. Nimer,Konstanze Doehner,Hartmut Doehner,Timothy J. Ley,Sheila Shurtleff,Jeffrey Rubnitz,Ching-Hon Pui,Elaine R. Mardis,Richard K. Wilson,James R. Downing
DOI: https://doi.org/10.1182/blood.v118.21.757.757
IF: 20.3
2011-01-01
Blood
Abstract:Abstract Abstract 757 Acute Megakaryoblastic Leukemia (AMKL) accounts for ∼10% of childhood acute myeloid leukemia (AML). Although AMKL patients with down syndrome (DS-AMKL) have an excellent 5 year event-free survival (EFS), non-DS-AMKL patients have an extremely poor outcome with a 3 year EFS of less than 40%. With the exception of the t(1;22) translocation seen in infant non-DS-AMKL, little is known about the molecular genetic lesions that underlie this leukemia subtype. To define the landscape of mutations that occur in non-DS-AMKL, we performed transcriptome sequencing on diagnostic blasts from 14 cases (discovery cohort) using the illumina platform. Our results identified chromosomal rearrangements resulting in the expression of novel fusion transcripts in 12/14 cases. Remarkably, in 7/14 cases we detected an inversion on chromosome 16 [inv(16)(p13.3;q24.3)] that resulted in the juxtaposition of the CBFA2T3, a member of the ETO family of transcription factors, next to GLIS2 resulting in a CBFA2T3-GLIS2 chimeric gene encoding an in frame fusion protein. 6 cases in the discovery cohort fused exon 10 of CBFA2T3 to exon 3 of GLIS2, while 1 case carried a larger product that fused exon 11 of CBFA2T3 to exon 1 of GLIS2. Both products retain the 3 CBFA2T3 N-terminal nervy homology regions that mediate protein interactions, and the 5 GLIS2 C-terminal zinc finger domains that bind the Glis DNA consensus sequence, along with one of its N-terminal transcriptional regulatory domains. GLIS2 is a member of the GLI super family of transcription factors and has been demonstrated to play a role in regulating expression of GLI target genes as well as inhibiting WNT signaling through the binding of beta catenin. Although GLIS2 is not normally expressed in hematopoietic cells, the translocation results in high level expression of the CBFA2T3-GLIS2 fusion protein. In addition to CBFA2T3-GLIS2, chimeric transcripts were detected in 6/7 cases that lacked evidence of the inv(16)(p13.3;q24.3). Specifically, we detected GATA2-HOXA9, MN1-FLI1, NIPBL-HOXB9, NUP98-KDM5A, GRB10-SDK1 and C8orf76-HOXA11AS, each in an individual case. Importantly, several of the genes involved in these translocations either play a direct role in normal megakaryocytic differentiation (GATA2 and FLI1), or have been previously shown to be involved in leukemogenesis (HOXA9, MN1, HOXB9). Evaluation of a recurrency cohort of 42 samples including 14 additional pediatric cases and 28 adult cases by RT-PCR revealed 4 additional pediatric samples carrying CBFA2T3-GLIS2 for an overall frequency of 39% in pediatric AMKL. In addition to these somatic structural variations, we also identified mutations in genes previously shown to play a role in megakaryoblastic leukemia including activating mutations in JAK2 and MPL (36%). To gain insight into the mechanism whereby CBFA2T3-GLIS2 promotes leukemogenesis, we introduced the fusion into murine hematopoietic cells and assessed its effect on in vitro colony replating as a surrogate measure of self-renewal. Hematopoietic cells transduced with a mCherry expressing retroviral vector failed to form colonies after the second replating. By contrast, expression of either wild-type GLIS2 or the CBFA2T3-GLIS2 fusion resulted in a marked increase in the self-renewal capacity, with colony formation persisting through eight replatings. Immunophenotypic analysis of the CBFA2T3-GLIS2 expressing colonies revealed evidence of megakaryocytic differentiation. Importantly, the CBFA2T3-GLIS2 cells remained growth factor dependent suggesting that cooperating mutations in growth factor signaling pathways are required for full leukemic transformation. Taken together these data identify a novel cryptic inv(16)-encoded CBFA2T3-GLIS2 fusion protein as a recurrent driver mutation in approximately 40% of non-infant pediatric non-DS-AMKLs. Moreover, the majority of pediatric cases that lacked this lesion were shown by transcriptome sequence analysis to contain other chromosomal rearrangements that encoded fusion proteins that directly alter megakaryocytic differentiation and/or myeloid cell growth. The alteration of a key transcriptional regulator within the hedgehog signaling pathways in a substantial percentage of pediatric AMKL raises the possibility that inhibition of this pathway may have a therapeutic benefit in this aggressive form of AML. *TAG and ALG contributed equally to this work. Disclosures: Biondi: BMS, Novartis, Micromed: Consultancy, Membership on an entity's Board of Directors or advisory committees. Ravandi:Bristol Myers Squibb: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Novartis: Honoraria. Kantarjian:Novartis: Consultancy, Research Funding; Pfizer: Research Funding; BMS: Research Funding. Doehner:Hoffmann La Roche: Honoraria.