Generation of a Pre-Leukemic, Transplantable Cell Line from the AML-Associated NUP98-TOP1 Fusion Gene as a New Model To Test Potential Collaborating Genes.

Over 15 distinct chromosomal translocations have been documented in human leukemias that involve rearrangement of the nucleoporin gene, NUP98. While most NUP98 fusion partners are HOX transcription factors, the remaining 7 have no obvious common unifying function or unique role in hematopoiesis. The translocation t(11;20) identified the novel NUP98 fusion partner Topoisomerase I (TOP1), a catalytic enzyme known for its role in relaxing supercoiled DNA. We recently demonstrated that over-expression of NUP98-TOP1 induces a lethal, transplantable AML in a murine bone marrow (BM) transplantation model (Gurevich et al, Blood 2004). To further explore the mechanistic process of leukemic transformation we sought to establish BM cell lines that retain the preleukemic-inducing properties of NUP98-TOP1. Murine BM was transduced with GFP-linked NUP98-TOP1 retrovirus. Following 4 weeks of culture, there was a striking out-growth of GFP⁺ cells (expansion of 1% to >80% GFP⁺ cells) and generation of poly-clonal cell lines (herein called NT 12.1 and NT 12.2) exhibiting features of primitive cells (blast-like morphology, lin^−/lo, increased levels of c-Kit, Sca-1 and CD34). NT cell lines had limited differentiation ability when plated in GM-CSF, G-CSF or M-CSF but in methylcellulose colony assays generated large granulo-monocytic colonies at high frequency (1 in 12) when plated in IL-3, IL-6, SCF and EPO. Strikingly, NT cell lines demonstrated significant levels of short-term in vivo repopulating ability when transplanted into lethally irradiated recipients. Transplant doses of 10⁴-10⁶ cells yielded >50% engraftment at 1 month post transplant (tx) (54±13% GFP⁺ WBC; n=12) which diminished to levels of <20% by 2 months. The frequency of cells with repopulating potential was estimated at 1 in 1609 following limit dilution assay which yielded repopulation of 3 of 4 recipients 1 month post-tx (range; 5–46%). While the level of engraftment diminished in NT 12.1 mice, two NT 12.2 mice developed fatal MPD/AML after a latency of 103 and 166 days. Thus these NUP98-TOP1 cell lines demonstrate a differentiation block in vitro, high level short term repopulating capacity and the potential to give rise to AML with a long latency that is consistent with a preleukemic phenotype. These lines now provide a novel platform to test candidate genes for their ability to co-operate with NUP98-TOP1 to induce leukemia. As an initial candidate, we tested the collaborating potential of Meis1, a HOX co-factor known to co-operate with several NUP98-HOX fusions to induce a rapid leukemia and may thus represent a common collaborating gene in NUP98 associated leukemias. When a representative cell line (NT12.2) was infected with a YFP-linked Meis1 retrovirus and transplanted into mice, the mice contained only GFP⁺/YFP⁺ circulating WBC by 1 months post-tx. However, despite co-expression of both genes in all transplanted mice, there was no apparent acceleration of induction of AML with 5 of 6 recipients succumbing to leukemia by 9 months post-tx, (median survival 164 days; range 101–193). This long latency argues that Meis1 does not collaborate with NUP98-TOP1 to induce leukemia and that NUP98-TOP1 has distinct mechanisms of leukemogenesis compared to NUP98-HOX fusions. These novel pre-leukemic, transplantable cell lines generated by NUP98-TOP1 will facilitate further large scale screens for potential cooperative genes and delineation of the mechanistic basis of NUP98-TOP1 induced transformation.