The Leukemogenic Potential of the NUP98-PMX1 Fusion Protein Is Independent of the Known Binding Properties of PMX1 to the Serum Response Factor and the Serum Response Element and Requires the NUP98 Sequences.

The translocation t(1;11)(q23;p15) encoding the NUP98-PMX1(N-PMX1) fusion gene combines the N-terminus of nucleoporin NUP98 with the homeodomain-containing C-terminus of PMX1(Nakamura et al., Blood. 1999 ). A large number of clustered Hox genes (Hox) have been reported to be fused to NUP98, however, N-PMX1 is the only fusion containing a non-clustered homeodomain, raising the intriguing possibility of unique mechanisms underlying its role in leukemia. Indeed, PMX1 was first identified for its role in enhancing the binding of the serum response factor (SRF) to the serum responsive element (SRE). Whether these PMX1 functions are involved in the leukemogenic role of N-PMX1 is not clear. Furthermore, in contrast to Hox genes, PMX1 is not expressed in hematopoietic cells, suggesting that the upregulation of its homeodomain, under the control of the ubiquitous NUP98 promoter, and/or the transcriptional activity of the NUP98 portion may be relevant for the leukemogenicity of N-PMX1. In order to gain insight into the possible unique functions of N-PMX1, we exploited retroviral gene transfer techniques and the murine bone marrow (BM) transplantation model. MSCV based retroviral vectors carrying an IRES linked GFP reporter gene were generated for the full length PMX1 and the N-PMX1 fusion. Lethally irradiated mice were transplanted, two days following transduction, with post 5-FU treated BM cells without preselection. Surprisingly, at 7 months post transplant, PMX1 recipients remained healthy and had normal blood cell parameters, whereas 2 out of 5 mice transplanted with N-PMX1 cells succumbed to AML, as evidenced by elevated white blood counts (202±154 × 10⁶ /ml), anemia (1.39±0.5 × 10⁹ /ml), splenomegaly (0.45±0.01 g) and high myeloid content in BM (>95% Gr1 and/or Mac1+). In all, no hematological abnormalities were observed in mice transplanted with PMX1 cells over a 9 month observation period. These results suggest that the ectopic expression of PMX1 alone in BM cells is not enough to perturb hematopoiesis, and support an essential role for the NUP98 portion of the fusion protein in the leukemogenic activity. In order to further clarify the mechanisms underlying the leukemogenic role of N-PMX1, we first examined its effects in vitro. N-PMX1 transduced BM cells expanded 50 fold more than PMX1 or GFP transduced BM cells after 21 days in liquid culture (n=5), and formed large granulo-macrophage colonies over at least 4 serial replatings in methylcellulose (n=5). Furthermore, as determined in CFU-S assays after 1 week in liquid culture, N-PMX1 increased the number of progenitor cells over 100-fold, whereas PMX1 transduced BM cells behaved as control cells (n=5). These results suggest that the leukemogenic potential of N-PMX1 is mediated by its abilities to block differentiation and promote proliferation and, further, that these effects are unique to the fusion protein. To examine if these abilities were mediated by SRF and SRE, which are known to regulate genes involved in oncogenic processes (e.g c-fos), we engineered strategic mutations reported to selectively eliminate the direct binding of PMX1 to the SRE or its transcription complex forming ability with SRF. Unexpectedly, these mutants had almost equivalent potency to block differentiation and promote proliferation with native N-PMX1. Together, these results indicate that different, as yet unknown PMX1 functions are crucial for N-PMX-mediated AML.