BRG is essential for the proliferation of leukemic cells. (A) Survival curves of secondary leukemic Mx1-Cre⁺;Brg^fl/fl and Mx1-Cre;Brg^wt/wt (control) A9M mice treated or not with polyI:polyC (pIpC). (B) Cytological preparations of peripheral blood leukocytes (PBL), bone marrow (BM), and liver (Li) of representative secondary leukemic mice at time of death. (C) Southern blot analysis of Brg locus configuration in primary and secondary Mx1-Cre⁺;Brg^fl/fl and control A9M leukemias at time of death. The vertical line indicates a repositioned gel lane. (D) Brg locus configuration of leukemic colony-forming cells (L-CFCs) derived from 3 independent pIpC-treated Mx1-Cre⁺;Brg^fl/fl (group D) primary leukemias at time of death as determined by polymerase chain reaction. (E) Frequency of L-CFCs derived from Cre-GFP and GFP (control) transduced Brg^fl/fl primary leukemia isolated at time of death (n = 50 to 100 GFP⁺ L-CFC per leukemia were scored; n = 2 to 3 leukemias per group). Mean ± SD. *P < .05. (F) Frequency of highly proliferative clones (HPCs) derived from Brg^fl/fl L-CFCs infected with Cre-GFP or GFP (control) (n = 15-75 HPCs analyzed per leukemia; n = 2 primary leukemias per group). Mean ± SD. *P < .05. (G) Cell cycle analysis of donor-derived (Ly5.2⁺) Mx1-Cre⁺;Brg^fl/fl primary leukemias isolated on days 2 to 4 after pIpC or mock treatment using propidium iodide (PI) staining (n = 3 to 5 mice per group). Mean ± SD. **P < .005. (H) Apoptotic index of donor-derived (Ly5.2⁺) Mx1-Cre⁺;Brg^fl/fl primary leukemias isolated on days 2 to 4 after pIpC or mock treatment as determined by terminal deoxynucleotidyltransferase-mediated deoxyuridine triphosphate nick end labeling (TUNEL) assay (n = 3 to 5 mice per group). Mean ± SD. **P < .005. 2nd, secondary leukemia; L-CFC, leukemic colony-forming-cell; HPC, highly proliferative clone. See also supplemental Figures 4-8.