The immortalizing function of HOXA9 and NUP98-HOXA9 is increased in Smad4−/− primitive hematopoietic cells. (A) Lin− cells transduced with HOXA9, NUP98-HOXA9, or the empty vector (Mig) were assessed for immortalization by counting the number of CFU-GM colonies per 10 000 cells. Results shown are from 3 rounds of plating using Wt cells (blue dots) or Smad4−/− cells (red dots), with medium containing 15% of FBS (data show mean ± SEM, n = 4, *P < .01 is measured by Student unpaired t test). The cloning efficiency (number of cells that make colonies) was calculated for the third round of plating (below graphs). (B) Decrease in size of the CFU-GM colonies observed for the third round of plating, by visualization of the plate on a UV-transilluminator and microscopy. Representative pictures of 4 experiments observed under forced expression of NUP98-HOXA9 (left panel). Statistical analysis for HOXA9 and NUP98-HOXA9 (data show mean ± SEM, n = 3, *P < .01 measured by Student unpaired t test; right panel). (C) Increase in the myeloblastic cell population in the Smad4−/− samples detected by May-Grünwald Giemsa staining of cytospun cells that constitute the CFU-GM colonies. Representative picture of the staining observed under forced expression of NUP98-HOXA9 (left panel). Statistical comparison between HOXA9 and NUP98-HOXA9 samples (mean ± SEM, n = 3, *P < .01 measured by Student unpaired t test; right panel). (D) Increase of the c-Kit+ primitive subpopulation in Smad4−/− bone marrow (BM) by FACS analysis. Representative FACS data observed under forced expression of NUP98-HOXA9 (left panel). Statistical comparison between HOXA9 and NUP98-HOXA9 samples (mean ± SEM, n = 3, *P < .01 measured by Student unpaired t test; right panel). (E) BrdU and 7-AAD staining showing increased percentage of cells in S phase among c-Kit+ primitive subpopulation in Smad4−/− BM. Representative FACS data observed under forced expression of NUP98-HOXA9 (left panel). Statistical comparison (right panel) between HOXA9 and NUP98-HOXA9 samples (mean ± SEM, n = 3, *P < .01 measured by Student unpaired t test). (F) CFU assay with medium supplemented with 10 ng/mL of the recombinant TGF-β (rTGF-β), with a TGF-β–free medium (M3236) supplemented with TGF-β receptor kinase inhibitor (TGF-β R inh) or both rTGF-β and TGF-β R inh (mean ± SEM, n = 3, *P < .01 measured by Student unpaired t test). The cloning efficiency (number of cells that make colonies) was calculated for the third round of plating (below graphs). (G) The growth capacity of HOXA9- and NUP98-HOXA9–transduced Smad4−/− cells is increased compared with Wt cells. Limiting dilution assay was done by plating HOXA9 or NUP98-HOXA9–transduced cells in 96-well culture plates, cells were maintained in SFEM medium, and supplemented with the TGF-β receptor kinase inhibitor (+ TGF-β R inh).