Prosurvival Bcl-2 proteins differ greatly in steady-state expression due to proteasomal turnover. (A) Schematic overview of the procedure to make J16 cell lines stably expressing N-terminal GFP fusions of the prosurvival Bcl-2 proteins at equal levels. (B) J16 cell lines transduced with constructs encoding N-terminal GFP fusions of the individual prosurvival Bcl-2 proteins as depicted in (A) were selected with antibiotics for 2 weeks, flow cytometrically sorted on equal GFP fluorescence intensity, and cultured on antibiotics for another week. Mean fluorescence intensity (MFI) of GFP was determined just before sorting (left) or 1 week after sorting (right) and normalized to untransduced cells (control). Data are shown as mean values ± SD of 3 independent transductions. (C) J16 cell lines expressing GFP fusions of the indicated prosurvival Bcl-2 proteins or empty vector (E.V.) were sorted on equal GFP fluorescence intensity and treated with the protein-synthesis inhibitor cycloheximide (50 µg/mL) or proteasome inhibitor MG132 (50 µM). At the indicated time points, GFP fluorescence intensity was assessed for each cell line by flow cytometry and expressed relative to its fluorescence intensity at the 0-hour time point, which was set at 100%. GFP expressed from the empty vector (E.V.) was a target for proteasomal degradation (C). This was due to a cloning artifact, only present for the E.V. but not for the fusion proteins, that endowed GFP-E.V. with a destabilizing C-terminal amino acid stretch (results not shown). Data are from 3 independent experiments and show mean values ± SD.