A Novel Strategy for Expanding Primitive Leukemic Cells from Chronic Phase CML Patients by Forced Overexpression of a NUP98-HOXA10 Homeodomain Fusion Gene.

Analysis of the leukemic stem cell compartment in CML patients with chronic phase disease remains a major challenge. This is due to the usually low frequency of these cells in the bone marrow and blood of most patients regardless of the WBC count and the fact that they are typically outnumbered by normal hematopoietic stem cells from which they cannot be currently separated. Moreover, thus far it has not been possible to identify conditions for their selective expansion in vitro or in vivo. To pursue this goal, we have begun to explore the effects of certain HOX gene-containing constructs on primitive chronic phase CML cells based on previous evidence that these genes markedly enhance the expansion of primitive normal murine and human cord blood cell numbers without inducing leukemia. Lineage-negative peripheral blood or bone marrow cells from 3 chronic phase CML patients (with >93%, <20% and <6% Ph+ LTC-ICs by G-banding karyotyping) were pre-stimulated overnight in a medium containing a serum substitute and 100 ng/ml hSteel Factor (SF), 100 ng/ml hFlt3-ligand and 20 ng/ml each of hIL-3, hIL-6 and hG-CSF. Cells were then exposed to a lenti-PGK-GFP virus with or without an upstream MDUS-NUP98-HOXA10 homeodomain (HD) element for 5 hours in the same medium. After removal of the virus, the cells were maintained in culture under the same conditions for 2 more days to allow full expression of the transduced genes. At this point, both cultures contained the same number of total cells, GFP+ cells and clonogenic progenitors (BFU-E + CFU-GM + CFU-GEMM); i.e., 2.2±0.5 x10⁵ vs 2.2±0.6 x10⁵ total cells, 1.0±0.2 x10⁵ vs 1.3±0.3 x10⁵ GFP+ cells, 3.6±1.7 x10⁴ vs 3.4±1.7 x10⁴ total CFCs and 1.7±0.9 x10⁴ vs 2.4±1.3 x10⁴ GFP+ CFCs per 10⁵ starting lin- cells. However, after the 2-day post-transduction, cells had been maintained for 6 weeks in longterm cultures (LTCs) containing murine stromal cells producing hIL-3, hSF and hG-CSF, we noted a markedly higher (4 to 74-fold) output of CFCs from the NUP98-HOXA10HD-transduced cells. Moreover, whereas the proportion of GFP+ CFCs in the 2-day post-transduction cultures was on average only 31% and 48 % for the control and tested cells respectively, this increased to >98% in the 6-week LTCs initiated with cells that were overexpressing NUP98-HOXA10HD but remained constant at 39% in the control LTCs - suggesting a significant growth advantage conferred by the NUP98A10HD transgene. Importantly, RT-PCR genotyping of the colonies in these assays showed the majority of LTC-IC-derived CFCs from the NUP98-HOXA10HD-transduced cells to be BCR-ABL+, indicative of an even greater output of CFCs by the NUP98-HOXA10HD transduced BCR-ABL+ vs normal cells. These results highlight the potential of NUP98-HOXA10HD to selectively expand primitive CML cells isolated directly from chronic phase patients which will facilitate their further investigation and use to screen and validate new therapeutic agents.