Retroviral Insertion Tagging Identifies Mn1 as a Specific Collaborator of NUP98-HOXD13 in a Murine Model of Leukemogenesis.

The NUP98-HOXD13 (NHD13) fusion gene is formed by the t(2;11)(q31;p15) in patients with myelodysplastic syndrome (MDS) or acute myeloid leukemia (AML). This fusion gene encodes a protein that fuses the N-terminal portion of NUP98, a nucleoporin involved in mediating RNA and protein transport in and out of the nucleus, with the C-terminal portion of HOXD13, a homeodomain protein not expressed during normal haematopoietic development. We have previously demonstrated that expression of the NHD13 fusion gene in a transgenic mouse model results in an invariably fatal MDS. These mice either die of complications of severe pancytopenia or progress to a fatal acute leukemia; greater than 90% of mice die by 14 months of age. The prolonged latency suggests that additional genetic events are required for leukemic transformation. We used a MOL4070LTR retroviral mutagenesis screen to identify candidate genes that might collaborate with the NHD13 transgene. All transgenic mice infected with the MOL4070LTR virus developed an acute leukemia, and the median survival of these mice was only four months, with none surviving beyond seven months of age. These survival figures are significantly decreased compared to either the wild type infected group or the transgenic uninfected group, suggesting a true synergistic effect between the NUP98-HOXD13 transgene and the genes affected by retroviral insertion events. Cloning of the insertion sites allowed identification of candidate collaborating genes. The most frequently identified gene was Meis1, a gene known to collaborate with HOX genes during leukemic transformation. This result supports the contention that the retroviral infection screen identifies specific collaboration between retrovirus and transgene. The second most common insertion site was near Mn1. These insertions clustered in two regions: four occurred within the final intron of Mn1, and four occurred between 80 kb and 160 kb 3′ of the 3′ end of the gene. Of note, MN1 is involved in human leukemia by fusion to the TEL oncogene via the t(12;22)(p13;q11) in myeloid leukemia. This translocation occurs within the intron of the human MN1 gene equivalent to that in which the four retroviral insertions occur in the mouse Mn1 gene. Moreover, this fusion transcript has recently been shown to collaborate with HOXA9 overexpression in leukemogenesis (Kawagoe and Grosveld, Blood 106(13):4269–77). We identified a fusion transcript formed in these four mice by RNA splicing between the 5′ part of Mn1 and the retroviral env gene. In addition, all eight of the mice with Mn1 insertions (both intronic and 3′) showed elevated wild-type Mn1 expression, suggesting that overexpression of wild-type Mn1 as well as a retroviral Mn1-env fusion can collaborate with the NHD13 transgene during leukemic transformation. Further studies are underway to investigate the oncogenic potential of an Mn1-env fusion as well as overexpression of the wild-type Mn1.