Combined inhibition of EZH2 and CDK4/6 perturbs endoplasmic reticulum-mitochondrial 1 homeostasis and increases antitumor activity against glioblastoma

Background: Few druggable targets have been discovered in glioblastoma (GBM); however, increasing knowledge has uncovered two hallmarks of this disease: epigenetic modifications and cell cycle dysregulation. Methods: Here, we investigated the effects of the EZH2 inhibitor GSK126 alone and in combination with the CDK4/6 inhibitor abemaciclib on morphology, growth, invasiveness, protein and gene expression, mitochondrial function, and induction of cell death in patient-derived GBM cells (GBM03, GBM06, GBM14, and GBM15). Results: EZH2 expression, CDK4 amplification, and CDKN2A deletion were first confirmed. The combined use of GSK126 and abemaciclib resulted in synergistic effects in all patient-derived GBM cell lines. After treatment, the cells appeared swollen with a large flat cytoplasm and cellular stress fibers due to HIF1 upregulation and CalR translocation. Notably, abemaciclib-induced cellular stemness (NANOG+, OCT3/4+, and SOX2+) was antagonized by this combination treatment. The MitoStress Test revealed a massive impairment of mitochondrial function. The basal oxygen consumption rate, ATP synthesis, and maximal respiration of the mitochondria decreased, confirming reduced cellular fitness and disrupted endoplasmic reticulum-mitochondrial homeostasis. This was paralleled by massive mitochondrial ROS production leading to mitochondrial depolarization and upregulation of the UPR sensors PERK, ATF6, and IRE1. Spheroid invasion was reduced in 3/4 cases, again with synergistic effects after dual EZH2 and CDK4/6 blockade. Differentially expressed genes involved mitotic aberrations/spindle assembly (Rb, PLK1, RRM2, PRC1, CENPF, TPX2), histone modification (HIST1H1B, HIST1H3G), DNA damage/replication stress events (TOP2A, ATF4), immuno-oncology (DEPDC1), and a shift in the stemness profile towards a more differentiated state. Notably, the antitumor effect was partially confirmed in ovo as well as in patient-derived organoids. Conclusions: We propose dual blocking of GBM hallmarks as a potential treatment strategy. Biomarker-driven targeted therapies will hopefully guide the development of more refined treatments. Triplicates of patient derived glioblastoma cells (GBM15), treated with the EZH2 inhibitor GSK126 and the CDK4/6 inhibitor abemaciclib were compared alone and in combination (kombi) with sham controls (DMSO).