Erythroid differentiation sensitizes K562 leukemia cells to TRAIL-induced apoptosis by downregulation of c-FLIP

Abstract

Regulation of the apoptotic threshold is of great importance in the homeostasis of both differentiating and fully developed organ systems. Triggering differentiation has been employed as a strategy to inhibit cell proliferation and accelerate apoptosis in malignant cells, in which the apoptotic threshold is often characteristically elevated. To better understand the mechanisms underlying differentiation-mediated regulation of apoptosis, we have studied death receptor responses during erythroid differentiation of K562 erythroleukemia cells, which normally are highly resistant to tumor necrosis factor (TNF) alpha-, FasL-, and TRAIL-induced apoptosis. However, upon hemin-mediated erythroid differentiation, K562 cells specifically lost their resistance to TNF-related apoptosis-inducing ligand (TRAIL), which efficiently killed the differentiating cells independently of mitochondrial apoptotic signaling. Concomitantly with the increased sensitivity, the expression of both c-FLIP splicing variants, c-FLIP(L) and c-FLIP(S), was downregulated, resulting in an altered caspase 8 recruitment and cleavage in the death-inducing signaling complex (DISC). Stable overexpression of both c-FLIP(L) and c-FLIP(S) rescued the cells from TRAIL-mediated apoptosis with isoform-specific effects on DISC-recruited caspase 8. Our results show that c-FLIP(L) and c-FLIP(S) potently control TRAIL responses, both by distinct regulatory features, and further imply that the differentiation state of malignant cells determines their sensitivity to death receptor signals.

FIG. 1.

Hemin-mediated differentiation sensitizes K562 cells to TRAIL-induced apoptosis. (A) K562 cells pretreated for 16 h with or without hemin were treated or left untreated for 3 h with 100 ng of TRAIL/ml together with 2 μg of M2 cross-linking antibody/ml. The apoptotic morphology was detected by DIC microscopy. Ctrl, control. (B) K562 cells were treated for the indicated time periods with TRAIL and M2 antibody after a 16-h incubation with or without 30 μM hemin. For procaspase 8 cleavage detection, cell lysates were analyzed by Western blotting as indicated in Materials and Methods. The appearance of cleaved p18 subunit indicates activation of caspase 8; p43/41 represents the intermediate cleaved forms of caspase 8. Unspecific bands are marked with asterisks. (C) K562 cells were treated as described for panel B, and DEVDase activity was analyzed from cell lysates with a time-resolved fluorometry-based caspase 3 assay. (D) Western blot analysis of PARP to detect a cleaved fragment of PARP (85 kDa), resulting from caspase 3 activity, was performed from the same samples as those used for panel B. Equal loading is shown by Hsc70 blotting.

FIG. 2.

Erythroid differentiation of HEL cells and granulocytic differentiation of HL-60 cells, but not megakaryocytic differentiation of K562 cells, sensitizes to TRAIL. (A) Analysis of apoptosis in HEL cells after combined hemin (60 μM, 24 h) and TRAIL (3 h) treatments by using PE-conjugated antibody recognizing active caspase 3. The bars show mean values (mean ± standard error of the mean) from two independent flow cytometry analyses with four replicas altogether. (B) Analysis of apoptosis in HL-60 after treatments with DMSO (1%, 24 h) and TRAIL (3 h). The data show mean values (mean ± standard error of the mean) from a single experiment with two replicas, and it is a representative of three independent experiments. (C) K562 cells were pretreated with TPA (20 nM) or hemin (30 μM) for 24 h and were stimulated with TRAIL for 3 h. The data showing mean values (mean ± standard error of the mean) from a single experiment with duplicates are representative of three independent experiments.

FIG. 3.

Hemin-treated K562 cells retain their mitochondrial membrane potential during TRAIL-induced apoptosis. (A) Western blot analysis of Bcl-X_L in K562 cells treated with hemin for indicated time periods. Hsc70 was blotted to show equal loading. (B) K562 cells treated with hemin for 16 h were loaded with TMRM to measure mitochondrial membrane potential. After treatment with TRAIL for 3 h, the red fluorescence ofTMRM was analyzed using confocal microscopy. Jurkat cells were used as a positive control for the mitochondrial depolarization. The overlay images of transmission light and TMRM fluorescence are shown. Cells displaying characteristic apoptotic morphology, with abundant membrane blebs, are marked with arrows. The confocal images shown are representative of three independent experiments. (C) To analyze the subcellular localization of cytochrome c, K562 cells were treated with hemin (16 h) in combination with TRAIL and M2 (3 h). After treatments the cells were fixed and centrifuged on a coverslip. Cells were labeled with antibodies specific to cytochrome c combined with fluorescent secondary antibody (green). DNA (red) was visualized by DAPI staining. Immunolabeling was analyzed by confocal microscopy. Maximum projection of a stack of 32 optical sections from a cell undergoing nuclear fragmentation is shown separately.

FIG. 4.

Surface expression of DR4 and DR5 is not increased during erythroid differentiation of K562 cells. (A) For surface analysis of TRAIL receptors, K562 cells were treated for 16 h with 30 μM hemin or left untreated. After treatments the cells were incubated on ice in the absence or presence of DR4- and DR5-specific antibodies or with TRAIL followed by an incubation with M2 antibody. After PBS washes the cells were incubated with fluorescent secondary antibodies. The fluorescence intensity was analyzed with flow cytometry. The thin line indicates cells only incubated with the secondary antibody (secondary antibody control); the thick line indicates the specific labeling. (B) Cell surface proteins from untreated (C) and hemin-treated (He) (16 h) cells were labeled with biotin and purified with streptavidin-coated agarose beads. The amount of DR4 and DR5 was detected by Western blotting. TNF-R1 is shown as a loading control.

FIG. 5.

Sensitization corresponds to downregulation of c-FLIP. (A) A schematic comparison of c-FLIP isoforms and procaspase 8 (modified from reference 37). (B and C) Western blot analysis of c-FLIP_L and c-FLIP_S, respectively, in K562 cells treated with 30 μM hemin for indicated time periods. Hsc70 was blotted to show equal loading. (D) To analyze hemin-induced apoptotic sensitivity, K562 cells were pretreated with 30 μM hemin for indicated time periods, followed by a 3-h TRAIL treatment to induce apoptosis. The amount of apoptosis was determined by Western blot analysis of PARP to detect the cleaved fragment of PARP (85 kDa). Equal loading is shown by Hsc70.

FIG. 6.

c-FLIP mRNA is modestly downregulated in hemin-treated K562 cells. RNase protection assay analysis of various apoptosis-related genes, including c-FLIP, in cells left untreated (C) or treated with hemin for 16 h (H16). Each signal was normalized by using the L32 housekeeping gene as an internal control. The normalized value corresponding to the amount of mRNA after hemin treatment was divided by the normalized control value shown as fold induction. Results represent the mean ± standard deviation from three independent experiments with a total five pairs of samples.

FIG. 7.

Downregulation of c-FLIP alters the composition of TRAIL receptor (TRAIL-R) DISC. (A) For TRAIL receptor DISC analysis, K562 cells treated with hemin (16 h) or left untreated were stimulated with TRAIL and M2 for 20 min and then lysed. For unstimulated samples, TRAIL and M2 were added after lysis. TRAIL receptor DISC was immunoprecipitated (IP) using Protein G-Sepharose. Coimmunoprecipitated DISC proteins (FADD, caspase 8, and c-FLIP) were analyzed by Western blotting. DR5 was probed to control successful immunoprecipitation. The presence of the proteins in cell lysates before immunoprecipitation is shown on the right. The migration positions of proteins are indicated. Treatments were as follows: C, control; H, hemin; T, TRAIL; HT, hemin and TRAIL. (B) To study the effect of caspase 8 inhibition on TRAIL-mediated apoptosis in K562 cells, the cells were treated with hemin (16 h) or left untreated followed by a 3-h TRAIL induction in the presence or absence of 10 μM caspase 8-specific inhibitor Z-IETD-fmk. The amount of apoptosis was determined by labeling the cells with PE-conjugated antibody specific to active caspase 3. The fluorescence of the cells was analyzed by flow cytometry.

FIG. 8.

Stable overexpression of both c-FLIP isoforms prevents differentiation-mediated sensitization. (A) Western blot analysis of c-FLIP from parental K562 cells, a mock-transfected cell pool, and cell lines stably overexpressing c-FLIP_L (1F6, 2E10, and 2G11) or c-FLIP_S (1E5 and 2E11). Equal loading is shown by Hsc70. (B and C) Analysis of apoptosis in c-FLIP_L and c-FLIP_S cell lines, respectively, in response to hemin (16 h) and TRAIL (3 h) treatments by using PE-conjugated antibody recognizing active caspase 3. The bars show mean values (mean ± standard error of the mean) from three independent flow cytometry analyses.

FIG. 9.

DISC analysis of cell lines overexpressing c-FLIP isoforms. G, sample devoid of TRAIL and M2; C, control; H, hemin; T, TRAIL; HT, hemin and TRAIL. Parental K562 cells and lines stably overexpressing c-FLIP_L (1F6) or c-FLIP_S (1E5) were treated with 30 μM hemin (16 h), stimulated with TRAIL and M2 for 20 min, and then lysed. For the control samples, TRAIL and M2 were added after cell lysis. TRAIL receptor DISC was immunoprecipitated (IP) using Protein G-Sepharose. The specificity of immunoprecipitation was controlled with the sample devoid of TRAIL and M2. The immunoprecipitates were analyzed by Western blotting using anti-FADD, anti-caspase 8, anti-c-FLIP, and anti-DR5 antibodies. The presence of the proteins in cell lysates before immunoprecipitation is shown below. The migration positions of proteins are indicated. For c-FLIP_L and c-FLIP_S, the upper band represents the exogenous protein and the lower band represents the endogenous protein.