Glycogen synthase kinase-3 and Leukemia: Restoring the balance

Abstract

The role of GSK-3 in oncogenesis is paradoxical, acting as a tumor suppressor in some cancers and potentiating growth in others. In this issue of Cancer Cell, Wang et al. provide some mechanistic insight into GSK-3 activity’s role in potentiating leukemias which are dependent on homeobox (HOX) gene misregulation.

Glycogen synthase kinase-3 (GSK-3) was first identified for its role in regulating glycogen metabolism but has since been shown to be involved in the regulation of a variety of processes including signal transduction, gene expression, and cell fate determination (Jope and Johnson, 2004). These critical roles have become increasingly more appreciated as misregulated GSK-3 has been implicated in neurological disorders (i.e. Alzheimer’s disease and bipolar disorder), non-insulin-dependent diabetes mellitus, stroke, and neoplasias (Cohen and Frame, 2001; Rayasam et al. 2009). With over 40 proteins identified as potential GSK-3 substrates, it is not surprising that a complicated network of GSK-3 functioning has emerged (Jope and Johnson, 2004). This complexity is particularly evident in cancer where GSK-3 can take on seemingly opposing roles in tumor suppression or promotion (Ougolkov and Billadeau, 2006; Luo, 2009).

Decreased expression or activity of GSK-3 has been associated with skin and breast tumors. In vitro and in vivo rescue experiments in which active GSK-3 is restored to transformed tumors cells leads to suppression of cell proliferation. Tumor suppressive functioning of GSK-3 has been shown to involve the WNT signaling pathway in which active GSK-3 negatively regulates β-catenin through inhibitory phosphorylation and prevents transcription of β-catenin target genes involved in cell cycle progression (Figure 1). Decreased expression or activity of GSK-3 could, then, activate the WNT signaling pathway through stabilization of β-catenin and contribute to tumorigenesis (Luo, 2009; Rayasam et al. 2009). In contrast, overexpression of active GSK-3 is associated with increased proliferation and decreased patient survival of some cancers through pathways which are thought to involve cyclin D1 and NFκB (Luo, 2009).

Figure 1. The seemingly contrary roles of GSK-3 help to maintain a fine balance between normal cell cycle progression and tumorigenesis.

Constitutively active GSK-3 acts to destabilize β-catenin through phosphorylation, thus, functioning as a tumor suppressor to limit proliferation (unshaded region). In contrast, GSK-3 promotes expansion of leukemic cells transformed by HOX overexpression through phosphorylation of CREB and subsequent formation of a transcriptional co-activator complex (shaded region).

This opposing function of GSK-3 as a tumor promoter has also been suggested for acute leukemia in a report by Wang and colleagues in which MLL-associated leukemia was shown to depend on GSK-3 for sustained proliferation of transformed cells (Wang et al., 2008). Pharmacologic inhibition of GSK-3 caused decreased proliferation, reduced cell cycle progression, and increased myeloid differentiation of leukemia cells which had been transformed with chimeric MLL oncoproteins. Decreased GSK-3 activity in MLL leukemia cells was associated with increased levels of β-catenin, decreased cell proliferation in vitro, and enhanced survival of mice with these leukemias in vivo. An increase in the CDK inhibitor, p27^Kip1, expression was observed specifically in MLL leukemia cells upon treatment with GSK-3 inhibitors suggesting that GSK-3 may override cell cycle regulators in MLL cells to enhance proliferation, though the exact mechanism remained unclear.

In the current issue of Cancer Cell, Wang et al. present a follow-up of their previous study where they begin to delineate the mechanism whereby MLL leukemias depend on GSK-3 for maintenance of proliferation and transformation (Wang et al., 2010). Their data demonstrate that GSK-3 activity promotes the formation of a HOX/MEIS1/CREB complex that recruits co-activators CBP and TORC to maintain the MLL leukemia stem cell transcription program. It has been well established that homeobox (HOX) genes become misregulated in aggressive leukemias involving MLL translocations (Shah and Sukumar, 2010). During normal hematopoiesis, MLL maintains appropriate HOX gene expression in hematopoietic stem and progenitor cells. As progenitor cells mature into differentiated cells of a committed lineage, the expression of HOX cluster genes become downregulated. MLL fusion proteins, on the other hand, cause the aberrant expression of target HOX transcription factors which function in concert with MEIS1 to maintain a transformed capacity in leukemia cells. Wang and colleagues now show that GSK-3 functions downstream of MLL fusion proteins to facilitate HOX-mediated transcription (Figure 1). Pharmacologic inhibition of GSK-3 in HOX-transformed cells (or functional inhibition via conditional expression of AKT) shows decreased proliferation and decreased HOX/MEIS1 transcriptional activity. This inhibitor sensitivity is dependent on the presence of CREB which is shown to recruit co-activators, CBP and TORC, to activate transcription of HOX/MEIS1 target genes. The association of this transcriptional activating complex, dependent on GSK-3 serine/threonine kinase activity, ultimately allows for the maintenance of a transformation program which promotes proliferation of the leukemia stem cells.

The identification of a functional role for GSK-3 in MLL-associated leukemia suggests that GSK-3 inhibitors may be a promising therapy which is selective for transformed cells that are dependent on HOX overexpression. It has also been reported that 90% of AML cases have aberrant expression of caudal-type homeobox transcription factor (CDX2), another regulator of HOX genes, further suggesting that proper HOX functioning is critical for maintaining a balance between normal and pathologic states (Rice and Licht, 2007). GSK-3 inhibitors might prove to be particularly efficacious in these CDX2-overexpressing leukemias and those with MLL translocations which exhibit HOX misregulation. In vivo pre-clinical studies of GSK-3 inhibition with lithium chloride have already shown promise in prolonging survival of mice with HOX-transformed leukemias (Wang et al. 2008; Wang et al. 2010). The lithium ion, though, is a relatively non-specific inhibitor which competes with magnesium for protein binding, so off-target effects are likely to occur as may also be the case with GSK-3 inhibitors which competitively inhibit ATP binding. Now that a critical complex has been identified as a target for GSK-3 activity in MLL leukemias, more specific substrate inhibitors may be developed which selectively target the critical formation of the HOX/MEIS1/CREB complex. It is also possible that monotherapy with a GSK-3 inhibitor may deplete normal hematopoietic stem cells (HSCs) in addition to targeting leukemic stem cells dependent on GSK-3 activity. A recent report has shown that GSK-3 is necessary for self-renewal of normal HSCs and that knockdown of both α and β GSK-3 isoforms may significantly deplete these normal HSC pools (Huang et al., 2009). Combination therapy with rapamycin may maintain the HSC pool while allowing the GSK-3 inhibitor to deplete the leukemia stem cells for an effective therapy against this aggressive form of leukemia.

It will be important to understand how the tumor promoting function of GSK-3 is counterbalanced by its tumor suppressor functions in leukemic cells, particularly, if GSK-3 inhibitors are used therapeutically which may ultimately block activity in either capacity. Cells which are dependent on the tumor suppressor actions of GSK-3 could be adversely affected by GSK-3 inhibition. What factors ultimately determine which function of GSK-3 function will predominate in a cell? Can the opposing activities of GSK-3 be utilized to override the HOX-mediated oncogenic program proposed by Wang et al.? A better appreciation of GSK-3 functions in signal transduction, cell cycle regulation, and cell fate determination may eventually allow for the effective application of GSK-3 inhibitors in the clinic. Wang et al. provide a novel mechanism by which GSK-3 promotes transformation, and in doing so, the authors provide a new therapeutic strategy which may lead to better clinical outcomes for aggressive MLL-associated leukemias.