We established several lines of mice that expressed a NUP98-TOP1 fusion from Vav regulatory elements; more than 5% of these mice developed AML during a 2-year study period. We attempted to increase the penetrance of the disease by using retroviral insertional mutagenesis, but noted no acceleration of disease, nor any unique common insertion sites when compared to wild-type littermates infected with the same retrovirus. We also considered that the penetrance of the disease might be increased by the stress of hematopoietic stem cell repopulation, and performed bone marrow transplant experiments. However, 6 months after bone marrow transplant, there was no difference in complete blood counts, % engraftment, or other indication of leukemic transformation. We concluded that the NUP98-TOP1 fusion was at best weakly oncogenic in this system. We cloned a t(14;21) chromosomal translocation breakpoint which led to mis-expression of a gene that encodes a basic helix-loop-helix (bHLH) domain protein. We noted that expression of this gene, originally designated BHLHB1 but since renamed OLIG2, is normally restricted to neural tissues. Interestingly, overexpression of OLIG2 was identified not only in oligodendroglioma samples, but also in a wide spectrum of malignant cell lines including leukemia, non-small cell lung carcinoma, melanoma, and breast cancer cell lines. To determine whether enforced expression of OLIG2 is oncogenic, we generated mice that overexpressed OLIG2 in the thymus. Ectopic OLIG2 expression was weakly oncogenic as only 2% of the transgenic mice developed pre-T lymphoblastic leukemia/lymphoma (pre-T LBL). However, almost 60% of transgenic mice that overexpressed both OLIG2 and LMO1 developed pre-T LBL. In addition, gene expression profiling demonstrated upregulation of Notch1 and Notch1 downstream genes, and growth of leukemic cell lines established from OLIG2/LMO1 mice was suppressed by a gamma-secretase inhibitor, suggesting that Notch1 up-regulation is important for the proliferation of OLIG2-LMO1 leukemic cells. CALM-AF10 fusions are expressed as a result of a t(10;11)(p13;q22) chromosomal translocation associated with gamma/delta preT-LBL and AML. To develop a model for gamma/delta preT-LBL or AML, we generated mice that expressed a CALM-AF10 fusion gene in hematopoietic tissues, using Vav regulatory elements to direct pan-hematopoietic transgene expression. Mice were clinically healthy for the first 9 months of life, and had normal peripheral blood hemograms, but showed impaired thymocyte differentiation, manifested by decreased CD4+/CD8+ cells and increased immature CD4-/CD8- cells in the thymus. 40-50% of the F1 mice developed acute leukemia, at a median age of 12 months. Leukemic mice typically had enlarged spleens and invasion of parenchymal organs with cells positive for myeloid markers such as myeloperoxidase, Mac1, and Gr1. Interestingly, although most leukemias were AML based on immunophenotype, many showed lymphoid features, such as clonal Tcrb or Igh gene rearrangements. In addition, AMLs that developed showed a variable proportion of B220+/Mac1+/Gr1+/Kit1+ cells, a finding that is notable since B220 has been regarded as a B-lineage marker in mice36. Hematopoietic tissues from both clinically healthy as well as leukemic CALM-AF10 mice showed up-regulation of Meis1, as well as Hoxa5, Hoxa7, Hoxa9, and Hoxa10, but not Hoxb4, a finding of note since Hoxa9 and Hoxb4 have been associated with increased hematopoietic stem cell self renewal. In this context, it is important to note that patients with AML and CALM-AF10 translocations show clonal IGH gene rearrangements and an independent cohort of patients showed upregulation of MEIS1, HOXA5, HOXA9, and HOXA10, highlighting the similarity of the mouse model to the human disease. The long latency period and incomplete penetrance suggest additional genetic events are needed to complement CALM-AF10 and produce a frank leukemia. We developed a mouse model for preT-LBL by simultaneously expressing SCL and LMO1 in the thymus. SCL and LMO1 form a complex with E2A; this complex is more stable than E2A-E2A homodimers, and it seems likely that SCL exerts its oncogenic effect via E2A sequestration and inhibition of E2A function. We used SIL regulatory elements to direct the expression of SCL, since this mimics the most common mechanism leading to SCL misexpression in human preT-LBL. SCL expression alone was not leukemogenic in these studies, and LMO1 expression was weakly leukemogenic. In addition, other investigators have generated mice that overexpress SCL in the thymus and develop preT-LBL, albeit with incomplete penetrance. One explanation for these findings is that the SIL-SCL founder lines expressed levels of the SCL transgene which were insufficient to initiate pre-T LBL. We screened eight founder lines, but were unable to identify founders with levels of SCL expression comparable to those of pre-T LBL patients. Since SIL is normally expressed in proliferating cells, it is possible that high levels of SCL expression directed by SIL regulatory elements are toxic in utero. Therefore, to refine the SIL-SCL mouse model and obtain higher level, conditional expression of SCL in the thymus, loxP sites were inserted into the SIL and SCL genes at sites where recombination occurs in pre-T LBL patients. The resultant SILloxloxSCL mice were bred to Lck-Cre mice that express Cre recombinase in the thymus. Recombination between the two loxP sites was identified in more than 50% of the thymocytes from SILloxloxSCL/Lck-Cre mice, bringing the SCL gene under the direct control of SIL regulatory elements. SCL expression in the thymus of SILloxloxSCL/Lck-Cre mice was approximately 10-fold higher than that of the SIL-SCL mice described above39. However, although the SILloxloxSCL/Lck-Cre mice showed increased numbers of DN thymocytes and decreased numbers of mature CD4+ and CD8+ T-cells, the mice did not develop pre-T LBL. These results indicate that although conditional activation of SCL impaired T-cell differentiation, expression of SCL from SIL regulatory elements, even at levels 10-fold higher than that of the SIL-SCL mice, did not lead to pre-T LBL.
Showing the most recent 10 out of 27 publications
