Mouse models for core binding factor leukemia.

RUNX1 and CBFB are among the most frequently mutated genes in human leukemias. Genetic alterations such as chromosomal translocations, copy number variations and point mutations have been widely reported to result in the malfunction of RUNX transcription factors. Leukemias arising from such alterations in RUNX family genes are collectively termed core binding factor (CBF) leukemias. Although adult CBF leukemias generally are considered a favorable risk group as compared with other forms of acute myeloid leukemia, the 5-year survival rate remains low. An improved understanding of the molecular mechanism for CBF leukemia is imperative to uncover novel treatment options. Over the years, retroviral transduction-transplantation assays and transgenic, knockin and knockout mouse models alone or in combination with mutagenesis have been used to study the roles of RUNX alterations in leukemogenesis. Although successful in inducing leukemia, the existing assays and models possess many inherent limitations. A CBF leukemia model which induces leukemia with complete penetrance and short latency would be ideal as a platform for drug discovery. Here, we summarize the currently available mouse models which have been utilized to study CBF leukemias, discuss the advantages and limitations of individual experimental systems, and propose suggestions for improvements of mouse models.

Aberrant expression of RasGRP1 cooperates with gain-of-function NOTCH1 mutations in T-cell leukemogenesis.

Ras guanyl nucleotide-releasing proteins (RasGRPs) are activators of Ras. Previous studies have indicated the possible involvement of RasGRP1 and RasGRP4 in leukemogenesis. Here, the predominant role of RasGRP1 in T-cell leukemogenesis is clarified. Notably, increased expression of RasGRP1, but not RasGRP4, was frequently observed in human T-cell malignancies. In a mouse bone marrow transplantation model, RasGRP1 exclusively induced T-cell acute lymphoblastic leukemia/lymphoma (T-ALL) after a shorter latency when compared with RasGRP4. Accordingly, Ba/F3 cells transduced with RasGRP1 survived longer under growth factor withdrawal or phorbol ester stimulation than those transduced with RasGRP4, presumably due to the efficient activation of Ras. Intriguingly, NOTCH1 mutations resulting in a gain of function were found in 77% of the RasGRP1-mediated mouse T-ALL samples. In addition, gain-of-function NOTCH1 mutation was found in human T-cell malignancy with elevated expression of RasGRP1. Importantly, RasGRP1 and NOTCH1 signaling cooperated in the progression of T-ALL in the murine model. The leukemogenic advantage of RasGRP1 over RasGRP4 was attenuated by the disruption of a protein kinase C phosphorylation site (RasGRP1(Thr184)) not present on RasGRP4. In conclusion, cooperation between aberrant expression of RasGRP1, a strong activator of Ras, and secondary gain-of-function mutations of NOTCH1 have an important role in T-cell leukemogenesis.

AID-induced T-lymphoma or B-leukemia/lymphoma in a mouse BMT model.

Activation-induced cytidine deaminase (AID) diversifies immunoglobulin through somatic hypermutation (SHM) and class-switch recombination (CSR). AID-transgenic mice develop T-lymphoma, indicating that constitutive expression of AID leads to tumorigenesis. Here, we transplanted mouse bone marrow cells transduced with AID. Twenty-four of the 32 recipient mice developed T-lymphoma 2-4 months after the transplantation. Surprisingly, unlike AID-transgenic mice, seven recipients developed B-leukemia/lymphoma with longer latencies. None of the mice suffered from myeloid leukemia. When we used nude mice as recipients, they developed only B-leukemia/lymphoma, presumably due to lack of thymus. Analysis of AID mutants suggested that an intact form with SHM activity is required for maximum ability of AID to induce lymphoma. Except for a K-ras active mutant in one case, specific mutations could not be identified in T-lymphoma; however, Notch1 was constitutively activated in most cases. Importantly, truncations of Ebf1 or Pax5 were observed in B-leukemia/lymphoma. In conclusion, this is the first report on the potential of AID overexpression to promote B-cell lymphomagenesis in a mouse model. Aberrant expression of AID in bone marrow cells induced leukemia/lymphoma in a cell-lineage-dependent manner, mainly through its function as a mutator.

EVI-1 interacts with histone methyltransferases SUV39H1 and G9a for transcriptional repression and bone marrow immortalization.

The ecotropic viral integration site-1 (EVI-1) is a nuclear transcription factor and has an essential function in the proliferation/maintenance of haematopoietic stem cells. Aberrant expression of EVI-1 has been frequently found in myeloid leukaemia as well as in several solid tumours, and is associated with a poor patient survival. It was recently shown that EVI-1 associates with two different histone methyltransferases (HMTs), SUV39H1 and G9a. However, the functional roles of these HMTs in EVI-1-mediated leukemogenesis remain unclear. In this study, we showed that EVI-1 physically interacts with SUV39H1 and G9a, but not with Set9. Immunofluorescence analysis revealed that EVI-1 colocalizes with these HMTs in nuclei. We also found that the catalytically inactive form of SUV39H1 abrogates the transcriptional repression mediated by EVI-1, suggesting that SUV39H1 is actively involved in EVI-1-mediated transcriptional repression. Furthermore, RNAi-based knockdown of SUV39H1 or G9a in Evi-1-expressing progenitors significantly reduced their colony-forming activity. In contrast, knockdown of these HMTs did not impair bone marrow immortalization by E2A/HLF. These results indicate that EVI-1 forms higher-order complexes with HMTs, and this association has a role in the transcription repression and bone marrow immortalization. Targeting these HMTs may be of therapeutic benefit in the treatment for EVI-1-related haematological malignancies.

Pbx1 is a downstream target of Evi-1 in hematopoietic stem/progenitors and leukemic cells.

Ecotropic viral integration site-1 (Evi-1) is a nuclear transcription factor, which is essential for the proliferation/maintenance of hematopoietic stem cells (HSCs). Aberrant expression of Evi-1 has been frequently found in myeloid leukemia, and is associated with a poor patient survival. Recently, we reported candidate target genes of Evi-1 shared in HSCs and leukemic cells using gene expression profiling analysis. In this study, we identified Pbx1, a proto-oncogene in hematopoietic malignancy, as a target gene of Evi-1. Overexpression of Evi-1 increased Pbx1 expression in hematopoietic stem/progenitor cells. An analysis of the Pbx1 promoter region revealed that Evi-1 upregulates Pbx1 transcription. Furthermore, reduction of Pbx1 levels through RNAi-mediated knockdown significantly inhibited Evi-1-induced transformation. In contrast, knockdown of Pbx1 did not impair bone marrow transformation by E2A/HLF or AML1/ETO, suggesting that Pbx1 is specifically required for the maintenance of bone marrow transformation mediated by Evi-1. These results indicate that Pbx1 is a target gene of Evi-1 involved in Evi-1-mediated leukemogenesis.