Abstract
Glycogen Synthase Kinase 3β (GSK3β) is a key regulator of cell metabolism, proliferation, survival, and differentiation. The kinase has abundant substrates including many proteins in the canonical WNT pathway. Considering that GSK3β phosphorylation of many pro-survival proteins result in their degradation (e.g MYC, MCL-1), it is not surprising that GSK3β activation by stress challenge leads to cell cycle arrest and/or apoptosis. GSK3β is negatively regulated by serine 9 phosphorylation mediated by Protein Kinase B (AKT). Since AKT activation supports survival of AML cells and inactivation of GSK3β could suppress stress signaling events, we hypothesize that serine 9 phosphorylation of GSK3β (p-GSK3β ) will be detrimental for AML patients. In the current study, we analyzed GSK3β expression by Reverse Phase Protein Analysis (RPPA) in a cohort of 511 AML patients. GSK3β expression was correlated with patient survival data and disease characteristics such as French-American-British (FAB) classification, cytogenetics, and mutational status. High levels of p-GSK3β were found to correlate with adverse outcome for survival and complete remission duration (CR) in patients with intermediate cytogenetics but not in those with unfavorable cytogenetics. CR was only 45 weeks in the third of patients with highest p-GSK3β levels compared to 98 weeks for patients with low levels (p = 0.008; N = 121). Even intermediate cytogenetic patients with FLT3 mutation fared better when p-GSK3β levels were lower (50 versus 24 weeks; p = 0.009; N = 35). Expression of GSK3β and its phosphorylated form was compared with expression of 229 other proteins using RPPA in the AML patient cohort. Consistent with p-GSK3β as an indicator of AKT activation, RPPA revealed that p-GSK3β is positively correlated with phosphorylation of AKT (S473), BAD (S136), and P70S6K. In addition, p-GSK3β negatively correlated with FOXO3A (which is degraded after phosphorylation by AKT). Bone marrow mesenchymal stem cells (BM MSCs) are a critical component of the leukemic microenvironment but how these cells modulate the survival of leukemia cells is not clear. RPPA analysis was performed on BM MSC from healthy donor (N = 71) and BM MSCs from AML patients (N = 106). Interestingly, both total and phosphorylated GSK3 were found to be elevated in the AML samples suggesting that AKT is also activated in the leukemic MSCs. In vitro models of the BM microenvironment suggest that the AKT pathway is activated in both the leukemic and supporting stromal cells so this finding is consistent with these models (Konopleva and Andreeff, Curr Drug Targets. 2007; 8: 685). Unlike in AML blast cells, FOXO3A was not correlated with GSK3β phosphorylation in the MSCs. Examination of miRs in normal versus AML BM MSCs using microarray analysis and validated by qRT-PCR indicate that miR-21 is elevated in the MSC of the normal individuals. As miR-21 is suppressed by FOXO3A, this finding supports the notion that AKT is active in the AML BM MSCs but FOXO3A may not be functional. This possibility is plausible as induction of FOXO3A results in apoptosis in MSC (Djouad Cloning Stem Cells. 2009; 11:407).
Conclusion
These findings suggest that AKT mediated phosphorylation of GSK3β may be detrimental to AML patients and p-GSK3β may serve as an important prognostic factor for at least a subset of AML patients. The results also suggest that activation of AKT can occur in both the malignant cells and MSC cells in the leukemic microenvironment.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.
This icon denotes a clinically relevant abstract