Stem Cells in T-ALL Have a Primitive CD133⁺/CD34⁺/CD7^- Phenotype.

The cell of origin of childhood acute lymphoblastic leukaemia (ALL) has been the subject of conflicting reports in recent years. One model suggests that many haemopoietic cell types are susceptible to transformation and the level of commitment of the target cell influences the characteristics of the resulting blast cell population. A second model suggests that primitive haemopoietic cells are the targets for transformation, with some differentiation occurring subsequent to the transformation event. This model suggests a hierarchy of progenitors may exist in ALL. In support of this latter model, we have demonstrated that leukaemic stem cells in B-ALL have a primitive CD34⁺/CD10⁻/CD19⁻ phenotype and T-ALL cells with NOD/SCID engrafting capacity are CD34⁺/CD4⁻. In this investigation we have attempted to further purify and characterise leukaemic stem cells from children with T-ALL. Cells from 7 patients were sorted for expression of CD34 and CD7 and the sorted subfractions evaluated for long-term proliferative ability in vitro using a serum free suspension culture assay and in the NOD/SCID mouse model. In this group of patients, the CD34⁺/CD7⁺ fraction represented 7±6% of cells at sorting, 6±4% were CD34⁺/CD7⁻ and the majority were CD34⁻/CD7⁺ (60±12%). After 3 weeks in culture, the majority of proliferating cells were derived from the CD34⁺/CD7⁻ subfraction (53±16%). By week 6, >70% of proliferating cells were derived from the CD34⁺/CD7⁻ subfraction. Unsorted ALL cells and the sorted subfractions from 4 of these patients, were evaluated for their ability to engraft sublethally irradiated NOD/SCID mice. In each case, engraftment was achieved using 10⁵–10⁶ unsorted cells (25–80% CD45⁺) and with the CD34⁺/CD7⁻ subfraction only (4–84% CD45⁺ with 3x10³–8x10⁴ cells). There was no engraftment with the other subfractions despite injecting up to 100 fold more cells. The engrafted cells had the same karyotype as the patient at diagnosis and expressed high levels of CD2, CD4 and CD7 implying they had differentiated in vivo. The self-renewal capacity of the CD34⁺/CD7⁻ cells was evaluated by secondary transplantation. CD45⁺ cells from NOD/SCIDs engrafted with CD34⁺/CD7⁻ cells successfully engrafted secondary recipients with equivalent levels of human cell engraftment, demonstrating these cells were capable of self-renewal. These findings suggest that cells with a more primitive phenotype may be the targets for transformation in T-ALL, rather than committed lymphocytes. To further investigate this hypothesis, we sorted cells from 4 of these patients for expression of CD133 and CD7 and evaluated their proliferative ability as described above. Results to date indicate that the CD133⁺/CD7⁻ fraction represents only 0.35% of nucleated cells at sorting. However, after 3 weeks in culture, 48±9% of proliferating cells were derived from this subfraction and by week 6, 58±20% of cells were derived from the CD133⁺/CD7⁻ subfraction. In vivo analyses completed in 2 patients to date have shown that only the CD133⁺/CD7⁻ subfraction was capable of engrafting NOD/SCID mice (0.5–54% CD45⁺ using 3x10³–10⁵ cells). These results demonstrate that T-ALL cells with long-term proliferative and NOD/SCID repopulating capacity express the primitive haemopoietic cell antigens CD133 and CD34 and lack expression of T-lineage markers. These findings add further support to the concept of a common cell of origin for acute leukaemias.