Several Abd-B HOX genes have been found in translocations with the nucleoporin gene NUP98 in patients with acute myeloid leukemia (AML) or myelodysplastic syndrome (MDS). We have previously tested both known and engineered NUP98-HOX fusions in the murine bone marrow transplantation model (

N. Pineault et al.,
MCB
24
:
1907
,
2004
). Strikingly, an engineered NUP98-HOXA10 (NA10) fusion, not observed in patients, and the AML-associated fusion gene NUP98-HOXD13 (ND13) have a virtually indistinguishable ability to transform myeloid progenitor cells and to induce leukemia in collaboration with MEIS1. Importantly, their transforming ability is lost when the DNA-binding homeodomain is mutated. This functional overlap provides a potentially powerful strategy to identify key genes/pathways mediating HOX-induced leukemias by looking for overlapping gene expression changes induced by different NUP98-HOX fusion genes. 5-FU bone marrow cells were transduced with retroviral vectors encoding for the leukemogenic ND13 or NA10 fusion genes or a non-leukemogenic ND13 gene with a N51S homeodomain mutation or the empty MIG vector. RNA was extracted from transduced GFP+ Sca1+ Lin- cells and linear RNA amplification was performed before the analysis on the Affymetrix GeneChip MOE430. Three independent experiments were conducted and analyzed. Correlation analysis showed a high degree of similarity between ND13 and NA10 in their overall gene expression profiles, compared to the N51S mutant or the MIG control. Validation with real-time quantitative RT-PCR on non-amplified RNA revealed good agreement between the gene array and the PCR, with a tendency for bigger fold-changes with the PCR method. Close to 500 genes were found differentially expressed (changed ≥2-fold vs. MIG ctrl and t-test p-value <0.05) and some 100 of these were changed by both ND13 and NA10 but not by the N51S homeodomain mutant. These genes are strong candidates as direct and/or immediate downstream targets involved in leukemic transformation. Remarkably, among these were genes previously identified as a NUP98 fusion partner in human leukemia (DEAD-box protein, Ddx10), or part of the same family of genes found in NUP98-fusions (Ddx4 and the paired mesoderm homeobox gene, Pmx2). This suggests a possible molecular link in leukemogenicity between HOX- and non-HOX-NUP98 fusions. Other interesting genes that were induced by ND13 and NA10, but not by the N51S homeodomain mutant, were genes previously implicated in leukemia (e.g. Flt3, Evi1) as well as Hox-related genes, such as the Hox cofactor Pbx3 and several Hox A cluster members. Furthermore, approximately one third were ESTs or genes with unknown function. In conclusion, our results document similar changes in gene expression induced by functionally redundant but different NUP98-HOX fusions and should facilitate the identification of common target genes involved in leukemic transformation.

Topics:
bone marrow cells, bone marrow transplantation, candidate disease gene, dna, dna microarrays, fluorouracil, fusion genes, gene expression, gene expression profiling, leukemia

Author notes

Corresponding author

2005, The American Society of Hematology
2004
Sign in via your Institution