For decades, T-ALL models were based on retroviral overexpression of Notch1 in mouse hematopoietic stem cells (HSCs). This induces aggressive T-ALL but does not reflect the clinical heterogeneity of the disease, making it questionable to what extent the insights gained from these models are generalizable. Models that reflect T-ALL's genetic complexity based on high-throughput sequencing studies are lacking. To address this, we developed a mouse strain that expresses a Cre-recombinase and Cas9 in the thymus (LSL.Cas9xLck.Cre) along with a lentiviral vector system in which T-ALL specific oncogenes are cloned in antisense flanked by LOX66/71 sites. Expression of gain-of-function mutations is induced upon Cre-mediated inversion together with loss-of-function mutations via the CRISPR-Cas9 system only in developing thymocytes.

Transplantation of HSCs from LSL.Cas9xLck.Cre mice transduced with our vector system into sublethally irradiated mice led to T-ALL development with different morphology and disease phenotypes. Leukemias ranged from immature arrested CD25⁺CD44⁺ (Tlx1_NRas) phenotype, towards more developed intermediate CD4⁺CD8⁺CD3⁻ phenotype (Tlx1_Pik3cd), up to fully mature single positive CD4⁺CD3⁺ or CD8⁺CD3⁺ blasts (Tal1.Lmo2+Pik3cd). Importantly, this in-vivo-multiplex CRISPR screening approach allowed deciphering oncogenic cooperation by probing more than 2000 possible mutational combinations in vivo. Our findings revealed a strong oncogenic synergy between overexpression of Tlx1 and loss-of-function mutations restraining PI3K-mTOR signaling. Tumor-derived cell lines could be easily established for further study and keep their phenotypic characteristics and mutation patterns. Thus, our system exemplifies how functional genetics can faithfully model hematological malignancies in vivo.