Glycogen Synthase Kinase-3β (GSK-3β) Inhibition Induces Apoptosis In Leukemic Cells through Mitochondria-Dependent Pathway

Abstract 2889

GSK-3β is a multifunctional kinase that plays a role in several signaling pathways. Due to the contradictory roles of GSK-3β as a mediator of both cell survival and apoptosis, we have examined the role of GSK-3β for proliferation and apoptosis in leukemic cell lines KG1a, K562 and CMK. GSK-3β was selectively inhibited by the small-molecule SB-415286.

Treatment of leukemia cells with SB-415286 (40 μM) for 72 hr approximately halved cell growth in all three cell lines. SB-415286 also showed a concentration-dependent stabilization of intracellular β-catenin: In KG1a cells the mean fluorescence intensity (MFI) [± 95% CI] was 3.1 [± 1.7] in untreated cells vs. 423 [± 24] in treated cell. The figures for the K562 and CMK cell lines were: 2.8 [± 1.6] vs. 353.2 [± 11.1], and 6.8 [± 4.0] vs. 320.2 [± 23.7], respectively. Cell cycle analysis was carried out to examine if the growth inhibition was caused by arrest in cell cycle and/or induction of apoptosis. We found that SB-415286 caused cell cycle arrest in the G₂/M phase and accumulation of events corresponding to the subG₁ phase, indicative of DNA fragmentation. The subG₁ population was 45%, 34% and 17% in KG1a, K562 and CMK cells, respectively. To confirm that the increase of the subG₁ fraction represented an apoptotic effect of the GSK-3β inhibition, we analyzed phosphatidylserine (PS) externalization and plasma membrane integrity. We found that SB-415286 caused a considerable increase of the proportion of early apoptotic cells, i.e. cells that were annexin V-positive and 7-AAD-negative: Mean [± 95% CI] in KG1a cells increased from 6.2% [± 1.2%] in untreated cells to 38% [± 3.1%] in treated cells. The figures for the K562 and CMK cell lines were: 3.0% [± 1.2%] vs. 29% [± 3.3%], and 3.9% [± 1.0%] vs. 16.0% [± 1.1%], respectively.

Apoptosis signaling can be initiated by extracellular (death receptor) and/or intracellular (mitochondrial) signals. Flow cytometric analysis of cells stained by a dual-fluorescent mitochondrial dye JC-1 showed that 5–11% of untreated leukemic cells had low mitochondrial membrane potential. After 72 hr exposure to SB-415286 the mean [±95% CI] loss of the mitochondrial potential was found in 23% [± 2.0%], 33% [± 3.5%] and 42% [± 3.8%], in CMK, K562 and KG1a cells, respectively.

Since drug treatment in some cell types may result in activation of both the intrinsic or extrinsic cell-death pathway in a parallel manner, we investigated if the external pathway is involved in SB-415286-induced apoptosis. For this purpose we assessed caspase-8 activation by flow cytometry. After 72 hr of treatment of CMK, K562 and KG1a cells the caspase-8 activities compared, to untreated cells, had increased 3.7-fold, 3.9-fold, and 4.4-fold, respectively.

In some cell types, the extrinsic cell-death pathway leads to the cleavage of Bid (pro-apoptotic member of the Bcl-2 family) by caspase-8, generating a truncated version of the protein (tBid) which in turn activates the mitochondrial apoptotic pathway. Therefore, we determined whether depolarization of the mitochondrial membrane in the leukemic cell lines was an effect of activated caspase-8 or a direct effect of SB-415286. For this purpose Z-IETD-FMK (25 μM), a specific inhibitor of caspase-8, was applied to the cells for 2 hr. We found that inhibition of caspase-8 did not prevent SB-415286-induced apoptosis assessed by PS externalization. This indicates that activation of caspase-8 is part of the intrinsic apoptotic pathway and occurs downstream of mitochondria membrane potential depolarization mediated by other caspases.

Taken together, our observations suggest that inhibition of GSK-3β induces apoptosis of leukemic cells by depolarizing the mitochondria membrane. Thus, inhibition of GSK-3β could be an attractive target for treatment of leukemia.