Co-Expression of NUP98-HOXD13 and Mutant IDH2 Triggers an Early T-Cell Precursor-like Leukemia in Mice

Background and hypothesis: Acute lymphoblastic leukemia (ALL) is the most common pediatric malignancy, comprising of B lineage ALL and T-cell lineage ALL (about 85% and 15%, respectively). In most cases, treatment regimens for ALL patients have proved efficient and prognosis is usually encouraging. However, a recently described subgroup of T-cell ALL patients have been found to have a poor prognosis when treated with conventional therapy. These higher risk T-cell leukemias have a cell of origin thought to resemble an early T-cell precursor (ETP), which retains the capability to differentiate into both myeloid or T-lineage cells, and are therefore termed ETP ALL. ETP ALL is characterized by lack of or low expression of typical T-cell surface markers, aberrant expression of myeloid and hematopoietic stem cell (HSC) surface markers and a gene expression signature resembling that of murine ETP cells. Recently, ETP ALL patients were found to have a unique mutational landscape that includes acquired mutations in genes typically associated with myeloid differentiation. The ambiguous lineage attributes of ETP ALL may explain the ineffectiveness of current therapies, stressing the need for establishing pre-clinical animal models. Here we show that co-expression of Nup98-Hoxd13 (NHD13) fusion gene and a mutant IDH2 gene (IDH2 R140Q) triggers an ETP-like leukemia in mice. Expression of NHD13 has been reported in MDS and AML patients, and NHD13 mice have been previously shown to develop MDS, AML, and T-cell ALL. Mutations in IDH1/2 genes occur in several cancers, including AML; the most common IDH mutation in AML patients is the IDH2 R140Q. However, expression of mutant IDH1/2 in mice has been reported to induce an expansion of the HSC compartment but was not sufficient for establishment of leukemia. Targeted sequencing of NHD13 mice has identified recurrent mutations in IDH1, suggesting that these two aberrations may collaborate to cause leukemia.

Study Design and Methods: IDH2 R140Q mice were crossed with NHD13 mice and the progeny were monitored for survival. Classification of leukemic samples was done using flow cytometry and PCR analysis of TCRβ gene rearrangements. Global expression and mutational patterns of IDH2/NHD13 leukemias were analyzed using gene chip arrays and whole exome sequencing (WES).

Results and conclusions: Transgenic IDH2/NHD13 mice showed decreased survival compared to all control groups. Detailed analysis of surface marker expression revealed that most double Tg cases (10/11 tested) developed leukemias resembling ETP (cKit+CD44+CD25-) or double negative (DN) 2 cells (cKit+CD44+CD25+). PCR analysis of the TCRβ locus revealed that unlike NHD13 T-cell leukemias, IDH2/NHD13 leukemias do not have clonal VDJ rearrangements. However, most of IDH2/NHD13 leukemias have clonal or oligoclonal DJ rearrangements of the TCRβ locus, suggesting that these leukemias emerged during an early stage in T cell development. Expression analysis revealed that the IDH2/NHD13 leukemia expression signature is enriched in genes that are upregulated in ETP or DN2 cells. In addition, we found that the IDH2/NHD13 leukemia expression signature is also enriched in genes uniquely expressed in ETP ALL patients. Finally, WES showed correspondence between genes mutated in the mouse ETP-like leukemia and genes recurrently mutated in ETP ALL patients. These results indicate that IDH2/NHD13 leukemias resemble ETP ALL in terms of gene expression, immunophenotype, and cooperative mutations.

Relevance and importance: The poor prognosis of ETP ALL may be related to incomplete effectiveness of treatment strategies, emphasizing the need for novel therapies. IDH2/NHD13 mice can serve as an excellent pre-clinical model to study ETP ALL and to test potential treatments.