Modeling Human T Cell Acute Lymphoblastic Leukemia In Vitro and In Vivo with LMO2.

Abstract 3648

The studies identifying gene translocations and mutations in T-ALL cell lines and/or in patients have contributed significantly to the understanding of the genetic abnormalities involved in T-ALL. However, studies on the biology of these genes, the targeted cells, the sequence and the number of hits required to convert a primary human hematopoietic stem cell (HSC)/progenitor cell into a fully transformed leukemic cell require good experimental models of human T cell development both in vivo and in vitro. The only in vivo model of human T cell leukemogenesis came unexpectedly from the gene therapy trial on patients with X-linked severe combined immunodeficiency (SCID-X1). Three to five years after gene therapy, 4 out of 10 patients in the trial developed clonal T-ALL. In these patients, retroviral integrations were found in proximity to the LMO2 promoter in the malignant clones, leading to aberrant expression of the oncogene. However, little is known on the effect of LMO2 overexpression in human cells and how it facilitates the development of T-ALL.

We have developed in vivo and in vitro models to study the role of T cell oncogenes in human cells. Using the OP9-DL1 co-culture system to differentiate human HSC into mature T cells in vitro, we culture human HSC transduced with lentiviruses expressing LMO2. LMO2 overexpressing cells are blocked at the double negative stage (CD4-CD8-) of differentiation when co-cultured on OP9-Delta-Like1 stroma and proliferate 50 to 100 times more than control cells. However, these cells are not immortalized and cultures lasted approximately 80 days. LMO2 overexpression have no effect on myeloid differentiation in vitro.

In vivo, LMO2 transduced human HSC/progenitor cells engraft the bone marrow of immunodeficient mice to levels comparable to control cells, while normal myeloid and B cell populations 20–24 weeks post-transplantation. LMO2 transduced cells have an increased capacity to generate T cells in the thymus in comparison to control cells (42% engraftment vs 8%, p<0.0001). Surprisingly, thymic and peripheral LMO2 cells are not blocked in their differentiation. LMO2 cells did not engraft secondary mice, confirming that LMO2 doesn't induce self-renewal of human HSC. However, the increase in thymic repopulation by LMO2 cells and the lack of differentiation block in vivo suggest that LMO2 overexpression generates an abnormal T cell population with an increase repopulation advantage (increase proliferation or decrease apoptosis) in the thymus which becomes the substrate for additional genetic/epigenetic events.

To test this hypothesis, we tried to immortalize LMO2 cells in vitro with secondary hits. Our preliminary results show that insertional mutagenesis can immortalized LMO2 cells in vitro. However these cells are not able to engraft immunodeficient mice or generate leukemia in vivo. The addition of intracellular NOTCH to one immortalized LMO2 cell line allows these cells to engraft and generate human T-ALL in vivo.

Globally, these results show that T cell oncogenes can be studied in primary human hematopoietic cells both in vitro and in vivo. Also, at least three hits are required to transform a human primary HSC/progenitor cell into a leukemic cell able to engraft and generate leukemia in vivo. It also suggests that a non-engrafting cell can be turned into a leukemic cell generating leukemia in vivo, implying that a cell can regain self-renewing properties.