PF-04691502 induces the intrinsic apoptosis pathway after PI3K/mTOR pathway inhibition. (A) Immunoblotting was used to show expression of the Bcl-2 family of proapoptotic proteins, namely Noxa and Puma, in resting nonactivated cells. pAKT^S473 and pS6K^T389 were used to confirm that PF-04691502 treatment (18 hours) had inhibited the PI3K/mTOR pathway. HSC70 was used as a loading control. The figure is representative of 9 independent experiments. (B) BAX conformational change was investigated in the presence or absence of PF-04691502; immunoprecipitation (IP) with the BAX 6A7 antibody was performed as described in supplemental Methods). IgH and IgL were included to confirm that identical quantities of antibody were used for the immunoprecipitation. Blot is representative of 4 independent experiments. (C) Cleavage of caspase 3 and its substrate PARP (marker of apoptosis). Blot is representative of 10 independent experiments. (D) CLL samples (n = 7) were treated with PF-04691502 as above but with or without the pan-caspase inhibitor ZVAD.fmk and analyzed by flow cytometry for annexin V/PI negativity. (E) Stromal fibroblasts (HFFF2) were cocultured with CLL cells and the whole well was treated in the presence or absence of continuous PF-04691502 (2.5 μM) for 24 hours. CLL cells were removed by scraping (or by pipetting; see supplemental Figure 3D), and cell viability was analyzed using annexin V/PI analysis by flow cytometry (n = 7). (F) The assay was performed as in panel E, except that the CLL cells were pretreated for 1 hour with PF-04691502 prior to washing out the drug, and the treated CLL cells were plated into wells containing stromal fibroblasts (n = 7) (described in supplemental Methods). Error bars represent SEM. No difference was observed between scraping all cells (panel E) or washing off only the CLL cells (see supplemental Figure 3D). *P < .05, **P < .01. NS, not significant.