The CDC20-APC/SOX2 signaling axis: An achilles' heel for glioblastoma

ABSTRACT

Glioblastoma stem-like cells (GSCs) play a critical role in glioblastoma progression and recurrence. We discuss recent results on the role of the mitotic ubiquitin ligase cell division cycle 20–anaphase-promoting complex (CDC20-APC) in the governance of cardinal GSC functions through a mechanism involving the transcription factor sex-determining region Y-box 2 (SOX2). These findings expand the non-mitotic roles of CDC20-APC with implications for stem cell biology.

Glioblastoma (GBM) is the most common brain tumor in adults and is almost invariably fatal. The current standard-of-care treatment for GBM patients consists of maximal safe resection followed by temozolomide (TMZ) chemotherapy with radiotherapy, and is associated with a median overall survival of 15 months. There has been an increasing appreciation of intratumoral cellular and molecular heterogeneity in GBM.¹ Similar to cancers in other organ systems, the idea that a limited population of glioblastoma cells operates at the summit of a tumor cell hierarchy led to the identification of human glioblastoma stem-like cells (GSCs, also called cancer stem cells or tumor-initiating cells).²

Although definitive markers for GSCs remain elusive, GSCs can be operationally defined by their capacity for in vivo tumorigenicity, self-renewal, and differentiation along restricted lineages. GSCs also appear to display higher invasive potential compared to non stem-like cancer cells.³ Human GSCs are resistant to radiation therapy through increased activation of the DNA damage response and concomitant DNA repair.⁴ Using a transgenic model of GBM, Parada and colleagues demonstrated that GSCs propagate the regrowth of tumors following TMZ treatment,⁵ although the intrinsic resistance of human GSCs to TMZ remains controversial. These observations have led to the concept that targeting the GSC subpopulation in addition to non-GSCs may be required to achieve a durable treatment response. Accumulating evidence suggests that, rather than representing a static state, GSCs in fact represent a dynamic cell state influenced by cell-intrinsic and -extrinsic events.¹

The anaphase-promoting complex (APC) is a multisubunit E3 ubiquitin ligase that regulates timely cell cycle progression through 2 functionally distinct subcomplexes, cell division cycle 20-APC (CDC20-APC) and CDC20 homolog 1-APC (CDH1-APC). In mitosis, APC activation requires binding of CDC20, which recognizes specific substrates and thereby drives the metaphase-anaphase transition and mitotic exit.⁶ Increasing evidence suggests an oncogenic role for CDC20 in several human cancers.⁶ Until recently, little was known about the functional significance of CDC20-APC in GBM. Recent studies in developmental neurobiology have revealed surprising non-mitotic roles for CDC20-APC in postmitotic neurons, suggesting the possibility that CDC20-APC controls functions beyond the cell cycle.⁷

Recent work by Kim and colleagues has demonstrated a requirement for CDC20-APC in the maintenance of GSC functions.⁸ Using patient-derived GSCs, the authors found elevated CDC20 protein levels in GSCs compared to human astrocytes and serum-differentiated glioblastoma cells, suggesting GSC-specific roles for CDC20 in addition to its function in mitotic regulation. To investigate the role of CDC20 in GSCs, the authors used an RNA-interference (RNAi) approach and found that CDC20 is essential for GSC self-renewal and invasiveness in vitro. Consistent with these findings, the pharmacologic APC inhibitor ProTAME also inhibited these GSC phenotypes. In complementary experiments, CDC20 overexpression augmented GSC self-renewal capacity and invasiveness. Importantly, these manipulations of CDC20-APC did not substantially affect cell cycle parameters or cell counts, indicating that CDC20-APC has biological functions in GSCs that are separable from obvious cell cycle regulation. The gold-standard assay for human GSCs is their ability to generate tumors when injected orthotopically into immunocompromised mice. Remarkably, CDC20 knockdown greatly diminished the in vivo tumorigenicity of GSCs, and CDC20 overexpression elevated their tumorigenic potential. Thus, CDC20 is both required and sufficient for GSC self-renewal, invasiveness, and in vivo tumor initiation.

In studies to understand the mechanism of CDC20-APC action in GSCs, the authors found that expression of nuclear-targeted CDC20 increased GSC invasive potential. Consideration of nuclear proteins previously implicated in GSC invasion and self-renewal led to the identification of the pluripotency-associated transcription factor sex-determining region Y-box 2 (SOX2). CDC20 directly interacted with SOX2 in vitro, and the 2 proteins were found in an endogenous complex in human GSCs. Genetic and pharmacologic inhibition of CDC20-APC in GSCs decreased the levels of SOX2 protein but not mRNA levels, a phenomenon that was reversed by proteasome inhibitors. These experiments indicated that CDC20-APC maintains SOX2 protein stability by inhibiting degradation of SOX2 mediated by the ubiquitin-proteasome system. To examine the consequences of this molecular link, the authors demonstrated that CDC20 was required for SOX2-dependent transcriptional activity in GSCs. Structure-function experiments performed in the setting of SOX2 RNAi in GSCs revealed that a SOX2 deletion mutant lacking the CDC20-binding region could not rescue the SOX2 RNAi-triggered invasion phenotype, indicating that binding of CDC20 to SOX2 is necessary for invasiveness. Additionally, through epistasis experiments the authors found that SOX2 functions downstream of CDC20 to regulate GSC self-renewal and invasion.

Finally, examination of The Cancer Genome Atlas revealed that high CDC20 expression specifically in the GBM proneural subtype is associated with shorter overall survival. A subgroup of proneural GBMs characterized by mutations in the isocitrate dehydrogenase 1 (IDH1) gene is associated with a more favorable prognosis.¹ When IDH1-mutant tumors were excluded, CDC20 mRNA levels were no longer prognostic, suggesting an interaction between CDC20 expression and IDH1 mutant status. Examination of additional datasets with larger patient numbers will be important to follow up on these observations.

Together, these results demonstrate a requirement for the CDC20-APC/SOX2 pathway in the maintenance of key GSC functions. A recent study by Rich and colleagues also demonstrated an important role for CDC20-APC in maintaining human GSCs.⁹ These findings raise several intriguing questions for future investigation. An important remaining question is precisely how CDC20-APC regulates SOX2 in GSCs (Fig. 1). Since CDC20-APC promotes SOX2 protein stability, potential mechanisms of regulation may be that CDC20-APC ubiquitinates and thereby degrades an intervening E3 ligase that targets SOX2 or that CDC20-APC indirectly promotes the activity of a deubiquitinating enzyme responsible for removing ubiquitin from SOX2. It is also possible that CDC20-APC influences post-translational modifications of SOX2, which have been shown to impact SOX2 protein stability in embryonic stem cells.¹⁰ The role of SOX2 in mediating the effects of CDC20-APC in glioblastoma in vivo also remains to be determined. Can control of CDC20-APC activity dynamically affect the GSC pool in vivo, and does CDC20 promote GSC tumorigenicity through SOX2? It will also be interesting to examine whether CDC20-APC dictates the responsiveness of glioblastoma cells—both GSCs and non-GSCs—to chemoradiation and what role SOX2 might have in mediating this effect. Further downstream, the specific gene targets of SOX2 that control self-renewal and invasion in human GSCs remain to be identified. Since multiple substrates of CDC20-APC have been identified, other SOX2-independent mechanisms are likely to mediate CDC20-APC-dependent processes in GSCs; for instance, it has been suggested that CDC20-APC might act through CDKN1A (also known as p21^WAF1/Cip1) in human GSCs.⁹

Figure 1.

The CDC20-APC/SOX2 pathway in human GSCs. (A) Cell division cycle 20–anaphase-promoting complex (CDC20-APC) regulates key functions of human glioblastoma stem-like cells (GSCs). (B) Depicted are mechanisms of CDC20-APC regulation of sex-determining region Y-box 2 (SOX2) that have been demonstrated to mediate GSC self-renewal and invasiveness.

In addition to GBM, these findings have important implications for cancers in other organ systems as well as stem cell biology in non-cancer contexts. SOX2 has been shown to regulate invasion in several cancers, raising the possibility that CDC20-APC might also govern invasion in these cancers through the regulation of SOX2. Given the prominent role of SOX2 in both embryonic and induced pluripotent stem cells as well as neural stem cells, it is tempting to speculate that CDC20-APC might additionally contribute to the maintenance of stem-like characteristics in these contexts. In summary, the CDC20-APC/SOX2 pathway in GSCs represents an intriguing therapeutic target for glioblastoma therapy.

Disclosure Of Potential Conflicts Of Interest

No potential conflicts of interest were disclosed