Impact of PBX binding on transformation. (A) Deletion mutants removing the tryptophan and adjacent amino acids essential for PBX binding were developed as schematically depicted. Like the parental constructs, all deletion derivatives carried an N-terminal HA-epitope. (B) Anti-HA immunoblot testing the expression of PBX-deleted HOX proteins compared with the original version. (C) Coimmunoprecipitation of PBX2 with HOXA9 and HOXA9ΔPBX. Flag-tagged PBX2 or empty vector as control was coexpressed together with HA-labeled HOXA9 or the respective PBX binding site mutant. Immunoprecipitation was done with anti-flag antibodies and the precipitates were analyzed by anti-HA immunoblot for the presence of HOXA9 protein. As control, antiflag and anti-HA blots of the input are shown alongside. (D) Colony numbers obtained in replating assays performed with primary bone marrow cells transduced either with HOX or a HOXΔPBX construct as indicated. Data are mean ± SD of a biologic triplicate. (E) Kaplan-Meier survival plot of mice transplanted either with cells cotransduced with HOXA9/Meis (n = 3) or HOXA9ΔPBX/Meis (n = 5) as indicated. (F) Engraftment of HOXA9ΔPBX/Meis cells. Genomic DNA was isolated from a peripheral blood sample drawn at day 66 after transplantation from HOXA9ΔPBX/Meis recipients. The HOX transgene was detected by PCR with primers spanning the PBX deletion site. As controls, DNA from a HOXA9/Meis graft (lane HOXA9) and a plasmid containing HOXA9ΔPBX (lane plasmid) were used. Expected amplicon sizes are 190 bp for HOXA9 and 172 bp for HOXA9ΔPBX. (G) White blood cell (WBC) count at day 66 after transplantation in peripheral blood of the HOXA9ΔPBX/Meis recipients and 2 nontransplanted animals (control). Spleen weight of normal animals and of HOXA9ΔPBX/Meis recipients at day of death is listed in the right column.