Using a loss-of-function approach, we have previously demonstrated that Zeb2 (Sip1/Zfhx1b) is highly expressed in the hematopoietic system1 and previous evidence from mouse retroviral mutagenesis studies points to a putative role of Zeb2 in initiation and/or progression of leukemia/lymphoma. Here we show that overexpression of Zeb2 from the Rosa26 locus resulted in altered T cell development. In addition, Zeb2 overexpressing mice spontaneously develop thymic lymphomas starting at 6 months of age, indicating that Zeb2 acts as a driver in T cell malignancies. To further investigate the role of Zeb2 in T-ALL progression, we have bred these mice onto a tumor-prone background (p53flox/flox) and observed a significant decrease in T-cell lymphoma/leukemia latency. Using a minimal dilution series of tumour cells into NOD/SCID mice we could demonstrate a 100 fold increase in leukemia initiating cells in the Zeb2 overexpressing tumors. Screening a cohort of T-ALL patients we found increased expression of ZEB2 predominantly associated within immature/ETP-ALL patients. Compared to other T-ALL subgroups (cortical or mature), ETP-ALLs are defined by a gene-expression profile similar to Early Thymic Precursor (ETP) cells, a subset of thymocytes with stem cell-like features, and a striking feature is that these patients have poor response to conventional therapies with a high risk of remission failure or hematological relapse. Importantly, we could identify two ETP-ALL patients who presented a unique chromosomal translocation t(2;14)(q22;q32) on karyotype. This new translocation involves the ZEB2 locus and the BCL11B locus as confirmed by FISH analysis.
In conclusion, we have shown that Zeb2 overexpression affects early T cell development and predisposes mice to develop an aggressive form of T-ALL with increased leukemic stem cell properties. This mimics patients with aggressive ETP-ALL driven by ZEB2 translocations, demonstrating that the EMT regulator ZEB2 is an oncogene for T-ALL.