Molecular Dissection of Nup98-HoxA9 Oncogenic Activity Using Drosophila

Roughly three-quarter of the genes associated with human diseases have fly counterparts. This high degree of conservation, combined to a wide range of genetic tools, makes Drosophila an attractive model to study basic mechanisms lying at the heart of various human disorders. Several oncogenes mediate their effects by interfering with specific cell machinery components common to all eukaryotes. The systematic identification of cell components influencing the activity of oncogenes should therefore accelerate the characterization of those oncogenes. Toward this goal, we take advantage of Drosophila molecular genetics to identify conserved genes that functionally interact with oncogenes. Our effort currently focuses on the t(7;11)(p15;p15) translocation associated with acute myeloid leukemia (AML) and which fuses the N-terminal part of Nucleoporin 98 (NUP98) to the C-terminal part of the transcription factor HOXA9. As homologues of NUP98 and HOXA9 are present in flies, we hypothesized that expression of NUP98-HOXA9 during development will affect some of the same protein networks that are perturbed in human hematopoietic cells. We successfully conducted several modifier screens in the past by exploiting dosage-sensitive phenotypes specifically induced in the eyes. To that end, we expressed NUP98-HOXA9 during eye development, which interestingly phenocopied Homothorax (HTH) overexpression in its ability to block eye development and promoted head cuticle formation. HTH is the homologue of MEIS1; a DNA-binding co-factor for HOXA9 that functions with a third partner, PBX, and which together form a ternary complex that regulates gene expression. Importantly, we found that the NUP98-HOXA9 eye phenotype was suppressed by mutations in the hth and exd (Drosophila pbx) genes, thus lending support to the specificity of the phenotype. In agreement with this, a structure/function analysis of NUP98-HOXA9 conducted in the fly eye narrowed down the same functional domains/motifs as those that had been identified using mouse models, namely, the HOXA9 homeodomain, the HOXA9 ANW motif (a PBX-interaction site) and the NUP98 portion. Remarkably, we also found that NUP98-HOXA9 and HTH/MEIS synergistically induced cell proliferation when coexpressed in the developing eye. As a result, large tissue overgrowths were produced. The cooperation observed in this experimental setting is reminiscent of the ability of MEIS1 to accelerate AML onset when co-expressed with NUP98-HOXA9 in mouse models. Moreover, we found that the collaboration strictly depends on endogenous EXD/PBX as its depletion by RNAi completely prevents overgrowth formation. Together, these findings provide compelling evidence that the NUP98-HOXA9 fly model recapitulates several of the key functional features that had been established in mammals for this oncogene and thus should prove useful to further delineate the immediate events disturbed by NUP98-HOXA9 expression. Based on these premises, we conducted a genetic screen to isolate dominant modifiers of the NUP98-HOXA9 eye phenotype. Approximately 100,000 fly progeny have been screened, which led to the isolation of a few hundred mutations acting either as suppressors or enhancers. Several complementation groups have now been uncovered and their molecular identification is currently underway. Validation of relevant genes in mouse leukemia models will be conducted to confirm their significance with respect to HOX-dependent leukemia. The NUP98-HOXA9 fly model as well as the early findings of the screen will be presented.