IDH1R132H Causes Resistance to HDAC Inhibitors by Increasing NANOG in Glioblastoma Cells

The R132H mutation in isocitrate dehydrogenase 1 (IDH1R132H) is commonly observed and associated with better survival in glioblastoma multiforme (GBM), a malignant brain tumor. However, the functional role of IDH1R132H as a molecular target for GBM treatment is not completely understood. In this study, we found that the overexpression of IDH1R132H suppresses cell growth, cell cycle progression and motility in U87MG glioblastoma cells. Based on cell viability and apoptosis assays, we found that IDH1R132H-overexpressing U87MG and U373MG cells are resistant to the anti-cancer effect of histone deacetylase inhibitors (HDACi), such as trichostatin A (TSA), vorinostat (SAHA), and valproic acid. Octyl-(R)-2-hydroxyglutarate (Octyl-2HG), which is a membrane-permeable precursor form of the oncometabolite (R)-2-hydroxyglutarate (R-2HG) produced in IDH1-mutant tumor cells, significantly increased HDACi resistance in glioblastoma cells. Mechanistically, IDH1R132H and Octyl-2HG enhanced the promoter activation of NANOG via increased H3K4-3Me, consequently increasing NANOG mRNA and protein expression. Indeed, HDACi resistance was attenuated in IDH1R132H-expressing glioblastoma cells by the suppression of NANOG using small interfering RNAs. Furthermore, we found that AGI-5198, a selective inhibitor of IDH1R132H, significantly attenuates HDACi resistance and NANOG expression IDH1R132H-expressing glioblastoma cells. These results suggested that IDH1R132H is a potential molecular target for HDACi-based therapy for GBM.

Does IDH1 R132H -driven R-2HG promote or suppress tumor development and growth in culture and in vivo? Koivenen et al. have shown that increased R-2HG in IDH1 R132H -overexpressing human astrocytes enhances cellular proliferation and anchorage-independent growth by the activation of Egl nine homolog (EGLN) and inactivation of hypoxia-inducible factor (HIF) [19]. Consistent with these findings, several reports have shown that R-2HG promotes distant metastasis via epithelial-mesenchymal transition (EMT) in colorectal cancer and tumor progression via NF-kB activation in AML cells [20,21]. In contrast, a tumor suppressive effect of R-2HG in glioblastoma cells by the inhibition of ATP synthase and mTOR signaling has been reported [22]. Su et al. have shown that R-2HG exerts anti-leukemic activity by suppressing leukemia cell proliferation and cell cycle progression and increasing apoptosis [23].
Given that IDH1/2 mutations are early observed events in several cancers such as glioma, chondrosarcoma and AML, selective inhibitors targeting IDH 1/2 mutation are attractive strategies for cancer therapy [24]. For example, selective inhibitors, AGI-5198 and MRK-A, targeting IDH1 R132H have been identified by a high-throughput screen, and found to suppress production of R-2-hydroxyglutarate (R-2HG) and IDH1 R132H -overexpressing glioma growth in a dose-dependent manner [25,26]. In addition, Enasidenib, which is an IDH2-mutant inhibitor, decreased production of R-2-hydroxyglutarate (R-2HG), and can be used for conventional therapy in AML. Nevertheless, many reports have shown that IDH1/2 mutant inhibitors could protect IDH1/2-muated glioma, chondrosarcoma, AML and colorectal carcinoma cells to multiple types of anti-cancer therapy, such as irradiation, daunorubicin, and PARP inhibitors [24,[27][28][29]. For example, pretreatment of AGI-5198 attenuated anti-cancer efficacy of irradiation due to decreased 2-HG, restored NADPH production and decreased ROS levels [27]. Additionally, protective role of AGI-5198 against anti-cancer effect of poly (ADP-ribose) polymerase (PARP) inhibitor has been reported [28]. Based on these backgrounds, it is becoming clear that development of therapeutic application of IDH1/2 mutant inhibitors would be useful for treatment of IDH1/2-mutated cancer.
Increased predicted survival and better outcomes have been observed in low-grade glioma with temozolomide (TMZ), the most common anti-cancer drug used in patients with glioma [30,31]. Several reports have shown that glioma cells with IDH1 R132H are more sensitive than wild-type cells to radiotherapy and various chemotherapies, such as metformin and cisplatin [13,[32][33][34]. It is becoming clear that histone deacetylase inhibitors (HDACi), such as trichostatin A (TSA), SAHA (also known vorinostat), and valproic acid, are promising anti-cancer drugs in multiple types of cancer [35]. Indeed, several pre-clinical studies of HDACs have shown that they target uncontrolled cellular proliferation, invasion, angiogenesis, and resistance to apoptosis in glioblastoma cells [36]. Based on these results from pre-clinical studies, vorinostat, the most advanced HDACi, has entered clinical trials in glioblastoma [36]. However, the IDH1 R132H mutation has not been evaluated for HDACi-based glioblastoma treatment.
The homeobox protein NANOG is an essential transcription factor that mainly regulates the self-renewal of embryonic stem cells and pluripotency [37]. NANOG regulates stem-like traits in cancer cells, such as proliferation, self-renewal, anchorage-independent growth, motility, epithelial-mesenchymal transition (EMT), immune evasion, and chemoresistance [38]. Indeed, several reports have shown that NANOG is abnormally increased and is closely associated with poor clinical outcomes in various types of cancer [39]. NANOG promotes multidrug resistance and immune evasion by increasing HDAC1; consequently, the inhibition of HDAC improves antigen-specific adoptive T-cell therapy as well as anti-cancer drug sensitivity [40]. In addition, suppressive effects of HDACi, such as TSA, vorinostat, and valproic acid, on the maintenance of cancer stem-like traits in pancreatic cancer cells and glioblastoma-derived stem cells have been reported [41].
Based on these previous results, we investigated whether IDH1 R132H overexpression alters the anti-cancer efficacy of HDACi in glioblastoma cells. In the present study, we showed that IDH1 R132H and its synthetic oncometabolite, R-2HG, significantly increase NANOG expression by activating its proximal promoter region, resulting in increased HDACi resistance. Furthermore, we found that the pharmacological and genetic suppression of IDH1 R132H and NANOG sufficiently attenuate HDACi resistance in IDH1 R132H -expressing glioblastoma cells. Consequently, our results reveal that increased NANOG expression in glioblastoma with IDH1 R132H is a potential molecular target for therapeutic strategies.

Overexpression of IDH1 R132H Suppresses Viability, Motility, and Cell Cycle Progression in U87MG Glioblastoma Cells
To determine whether the overexpression of IDH1 R132H has tumor suppressive or promoting effects in glioblastoma cells, we initially generated IDH1-wild type (WT) or IDH1 R132H -mutant stably expressing U87MG cells. The overexpression of IDH1 R132H was confirmed ( Figure 1A) and decreased cell viability in IDH1 R132H -overexpressing U87MG cells was observed ( Figure 1B). In addition, suppressed cell motility by IDH1 R132H overexpression was also observed ( Figure 1C). To determine the mechanism by which IDH1 R132H overexpression suppresses cell viability in U87MG glioblastoma cells, we investigated the alteration of cell cycle progression. We found that the overexpression of IDH1 R132H increases the G2/M population by approximately 18% compared to those of the empty vector or IDH1-WT-overexpressing U87MG cells ( Figure 1D,E). Indeed, decreased levels of genes that promote the cell cycle, such as Tubulin Beta 3 (TUBB3), cell-division cycle protein 20 (CDC20), minichromosome maintenance complex component 7 (MCM7), Baculoviral IAP Repeat Containing 5 (BIRC5), and Aurora Kinase A (AURKA), were observed in IDH1 R132H -overexpressing U87MG cells ( Figure 1F). Consistent with our results, previous work has shown that IDH1 R132H suppresses cell cycle progression and promotes apoptosis by inhibiting the Wnt/β-catenin signaling pathway [11]. These results suggest that IDH1 R132H acts as a tumor growth suppressive factor in U87MG glioblastoma cells. Overexpression of IDH1 R132H suppresses cell cycle promoting genes expression in U87MG cells. The stable cells were incubated for 24 h, and then cell cycle promoting genes expression was measured by using qRT-PCR. The values represent the mean ± SD of three independent experiments performed in triplicate; *p < 0.05.

Overexpression of IDH1 R132H Abolishes The Anti-Cancer Effect of HDAC Inhibitors
Increasing data from preclinical and clinical studies of HDACi have shown that they are promising chemotherapeutics for the treatment of multiple types of cancer, including glioblastoma [42]; accordingly, we tested whether the overexpression of IDH1 R132H affects HDACi-based glioblastoma. Surprisingly, we found that the decreased cell viability by trichostatin A (TSA), vorinostat, or valproic acid, which are major Class I and II HDAC inhibitors, was significantly abolished in IDH1 R132H -overexpressing U87MG glioblastoma cells ( Figure. 2A). In addition, the decreased anti-cancer effect of TSA was also observed in IDH1 R132H -overexpressing U373MG cells ( Figure. 2B). To confirm this functional role of IDH1 R132H on HDACi resistance, the apoptotic cell population in the absence or presence of TSA was quantitatively analyzed in IDH1-WT or IDH1 R132H -

Overexpression of IDH1 R132H Abolishes the Anti-Cancer Effect of HDAC Inhibitors
Increasing data from preclinical and clinical studies of HDACi have shown that they are promising chemotherapeutics for the treatment of multiple types of cancer, including glioblastoma [42]; accordingly, we tested whether the overexpression of IDH1 R132H affects HDACi-based glioblastoma. Surprisingly, we found that the decreased cell viability by trichostatin A (TSA), vorinostat, or valproic acid, which are major Class I and II HDAC inhibitors, was significantly abolished in IDH1 R132H -overexpressing U87MG glioblastoma cells (Figure 2A). In addition, the decreased anti-cancer effect of TSA was also observed in IDH1 R132H -overexpressing U373MG cells ( Figure 2B). To confirm this functional role of IDH1 R132H on HDACi resistance, the apoptotic cell population in the absence or presence of TSA was quantitatively analyzed in IDH1-WT or IDH1 R132H -overexpressing U87MG cells. Figure 2C shows that the increased apoptotic cell population upon TSA treatment was significantly decreased by approximately 40% in IDH1 R132H -overexpressing U87MG cells. These results revealed that the IDH1 R132H mutation might cause chemoresistance to HDACi-based glioblastoma therapy.
overexpressing U87MG cells. Figure 2C shows that the increased apoptotic cell population upon TSA treatment was significantly decreased by approximately 40% in IDH1 R132H -overexpressing U87MG cells. These results revealed that the IDH1 R132H mutation might cause chemoresistance to HDACibased glioblastoma therapy. The values represent the mean ± SD of three independent experiments performed in duplicate; *p < 0.05 and **p < 0.01. (B) IDH1-WT or IDH1 R132H overexpressing stable U373MG cells were incubated with TSA for 48 h. Cell viability were measured by using crystal violet staining. Cell viability was measured by using crystal violet staining. The values represent the mean ± SD of three independent experiments performed in duplicate; *p < 0.05. (C) U87MG cells stably expressing IDH1-WT or IDH1 R132H were incubated with TSA for 72 h. Apoptotic cell population was measured by using Annexin-V staining. The values are presented as the mean ± SD of three independent experiments performed in duplicate; *p < 0.05 and **p < 0.01.

NANOG is Increased in IDH1 R132H -Overexpressing U87MG and U373MG Glioblastoma Cells
Increased gene expression, including the sex determining region Y-box 2 (SOX2), CD133, CD44, and NANOG in multiple types of cancer are closely associated with malignant phenotypes, such as angiogenesis, metastasis, and chemoresistance [38,43,44]. In addition, previous report has shown that embryonic stem (ES)-like gene signature, such as NANOG, Oct4, SOX2 and c-Myc, was highly increased in grade 4 glioblastomas, which the most aggressive subtype of glioma, whereas lower or opposite ES-like signature was observed in lower grade glioblastoma and normal brain tissue [45]. Zbinden et al. also reported that NANOG modulates glioblastoma development and growth by acquisition of CD133+ stem cell behavior, gliomasphere clonogenicity and proliferation [46]. Thus, we further measured gene expression related to chemoresistance in IDH1 R132H -overexpressing glioblastoma cells. We found that NANOG mRNA levels were predominantly increased by IDH1 R132H overexpression in U87MG cells ( Figure. 3A). In parallel with these findings, increased NANOG The values represent the mean ± SD of three independent experiments performed in duplicate; * p < 0.05 and ** p < 0.01. (B) IDH1-WT or IDH1 R132H overexpressing stable U373MG cells were incubated with TSA for 48 h. Cell viability were measured by using crystal violet staining. Cell viability was measured by using crystal violet staining. The values represent the mean ± SD of three independent experiments performed in duplicate; * p < 0.05. (C) U87MG cells stably expressing IDH1-WT or IDH1 R132H were incubated with TSA for 72 h. Apoptotic cell population was measured by using Annexin-V staining. The values are presented as the mean ± SD of three independent experiments performed in duplicate; * p < 0.05 and ** p < 0.01.

NANOG Is Increased in IDH1 R132H -Overexpressing U87MG and U373MG Glioblastoma Cells
Increased gene expression, including the sex determining region Y-box 2 (SOX2), CD133, CD44, and NANOG in multiple types of cancer are closely associated with malignant phenotypes, such as angiogenesis, metastasis, and chemoresistance [38,43,44]. In addition, previous report has shown that embryonic stem (ES)-like gene signature, such as NANOG, Oct4, SOX2 and c-Myc, was highly increased in grade 4 glioblastomas, which the most aggressive subtype of glioma, whereas lower or opposite ES-like signature was observed in lower grade glioblastoma and normal brain tissue [45]. Zbinden et al. also reported that NANOG modulates glioblastoma development and growth by acquisition of CD133+ stem cell behavior, gliomasphere clonogenicity and proliferation [46]. Thus, we further measured gene expression related to chemoresistance in IDH1 R132H -overexpressing glioblastoma cells. We found that NANOG mRNA levels were predominantly increased by IDH1 R132H overexpression in U87MG cells ( Figure 3A). In parallel with these findings, increased NANOG protein levels were also observed in IDH1 R132H -expressing U87MG and U373MG glioblastoma cells ( Figure 3B).
Previous reports have shown that NANOG promotes chemoresistance by increasing multidrug resistance protein 1 (MDR1) also known as P-glycoprotein 1, which pumps anticancer agents out of the cells results in reduced intracellular drug concentration and cytotoxicity, expression [39,47]. Thus, we investigated whether increased NANOG by IDH1 R132H overexpression could functionally increase MDR1 expression. Increased MDR1 mRNA and protein expression was observed in IDH1 R132H -overexpressing U87MG and U373MG cells ( Figure 3C,D). These results suggest that NANOG and MDR1 levels are increased in IDH1 R132H -expressing glioblastoma cells.
Previous reports have shown that NANOG promotes chemoresistance by increasing multidrug resistance protein 1 (MDR1) also known as P-glycoprotein 1, which pumps anticancer agents out of the cells results in reduced intracellular drug concentration and cytotoxicity, expression [39,47]. Thus, we investigated whether increased NANOG by IDH1 R132H overexpression could functionally increase MDR1 expression. Increased MDR1 mRNA and protein expression was observed in IDH1 R132Hoverexpressing U87MG and U373MG cells ( Figure. 3C and 3D). These results suggest that NANOG and MDR1 levels are increased in IDH1 R132H -expressing glioblastoma cells. Gene expression levels were measured by using qRT-PCR. The values represent the mean ± SD of three independent experiments performed in duplicate; *p < 0.05. (B) IDH1 R132H and NANOG protein levels in IDH1-WT or IDH1 R132H expressing U87MG and U373MG cells were measured by using Western blotting. (C) IDH1-WT or IDH1 R132H overexpressing U87MG cells were incubated for 24 h. Gene expression levels were measured by using qRT-PCR. The values represent the mean ± SD of three independent experiments performed in duplicate; *p < 0.05. (D) MDR1 protein levels in IDH1-WT or IDH1 R132H expressing U87MG and U373MG cells were measured by using Western blotting.

Knock-Down of NANOG Attenuates HDACi Resistance in IDH1 R132H -Expressing U87MG and U373MG Glioblastoma Cells
Because increased NANOG is associated with chemoresistance to multiple types of anti-cancer drugs, including HDACi [40], we further investigated whether NANOG is closely linked to HDACi resistance caused by IDH1 R132H expression. Figure 5A shows that the knock-down of NANOG by small interfering RNA (siRNA) is sufficient to attenuate TSA resistance in IDH1 R132H -expressing U87MG and U373MG glioblastoma cells. Indeed, a reduction of cell viability by approximately 50% in both IDH1 R132H -overexpressing cells upon TSA treatment was observed ( Figure. 5A). In addition, strong knock-down of NANOG was also confirmed in both U87MG and U373MG cells (Figure. 5B). Quantitation of the apoptotic cell population by annexin-V staining showed that the suppression of NANOG abolishes TSA resistance in IDH1 R132H -expressing U87MG cells ( Figure. 5C and 5D). To confirm whether increased NANOG is functionally sufficient to cause TSA resistance in glioblastoma cells, cytotoxic effect of TSA was measured in NANOG overexpressed or parental U87MG cells. Increased cell viability upon both vehicle and TSA treatment was observed in NANOG overexpressed U87MG cells, suggesting that increased NANOG could be functionally associated with HDACi resistance (Figure. 5E). In addition, increased MDR1 mRNA levels were observed in NANOG-overexpressing U87MG cells, suggesting that increased MDR1 by NANOG could promote TSA resistance (Figure. 5F). These results indicate that increased NANOG is critical for HDACi resistance in IDH1 R132H -expressing glioblastoma cells.

Knock-Down of NANOG Attenuates HDACi Resistance in IDH1 R132H -Expressing U87MG and U373MG Glioblastoma Cells
Because increased NANOG is associated with chemoresistance to multiple types of anti-cancer drugs, including HDACi [40], we further investigated whether NANOG is closely linked to HDACi resistance caused by IDH1 R132H expression. Figure 5A shows that the knock-down of NANOG by small interfering RNA (siRNA) is sufficient to attenuate TSA resistance in IDH1 R132H -expressing U87MG and U373MG glioblastoma cells. Indeed, a reduction of cell viability by approximately 50% in both IDH1 R132H -overexpressing cells upon TSA treatment was observed ( Figure 5A). In addition, strong knock-down of NANOG was also confirmed in both U87MG and U373MG cells ( Figure 5B). Quantitation of the apoptotic cell population by annexin-V staining showed that the suppression of NANOG abolishes TSA resistance in IDH1 R132H -expressing U87MG cells ( Figure 5C,D). To confirm whether increased NANOG is functionally sufficient to cause TSA resistance in glioblastoma cells, cytotoxic effect of TSA was measured in NANOG overexpressed or parental U87MG cells. Increased cell viability upon both vehicle and TSA treatment was observed in NANOG overexpressed U87MG cells, suggesting that increased NANOG could be functionally associated with HDACi resistance ( Figure 5E). In addition, increased MDR1 mRNA levels were observed in NANOG-overexpressing U87MG cells, suggesting that increased MDR1 by NANOG could promote TSA resistance ( Figure 5F). These results indicate that increased NANOG is critical for HDACi resistance in IDH1 R132H -expressing glioblastoma cells. Transfected cells were incubated with or without 5 μM of TSA for 48 h. Cell viability was analyzed by using crystal violet staining and NANOG expression was measured by using western blotting. The values represent the mean ± SD of two independent experiments performed in triplicate; *p < 0.05 and **p < 0.01. (F) MDR1 mRNA levels were measured by using qRT-PCR. The values represent the mean ± SD of three independent experiments performed in duplicate; *p < 0.05.

Octyl-2-Hydroxyglutarate (Octyl-2HG) Increases NANOG and HDACi Resistance in Glioblastoma Cells
The heterozygous mutation of IDH1 to IDH1 R132H produces R-2HG by disrupting the conversion of isocitrate to α-ketoglutarate [1,3]. Thus, we investigated whether R-2HG is sufficient to mimic IDH1 R132H overexpression to increase NANOG expression and HDACi resistance in glioblastoma cells. We found that cellular-permeable octyl-2HG is sufficient to increase NANOG mRNA and protein levels in U87MG and U373MG glioblastoma cells ( Figure. 6A and 6B). Consistent with Transfected cells were incubated with or without 5 µM of TSA for 48 h. Cell viability was analyzed by using crystal violet staining and NANOG expression was measured by using western blotting. The values represent the mean ± SD of two independent experiments performed in triplicate; * p < 0.05 and ** p < 0.01. (F) MDR1 mRNA levels were measured by using qRT-PCR. The values represent the mean ± SD of three independent experiments performed in duplicate; * p < 0.05.

Octyl-2-Hydroxyglutarate (Octyl-2HG) Increases NANOG and HDACi Resistance in Glioblastoma Cells
The heterozygous mutation of IDH1 to IDH1 R132H produces R-2HG by disrupting the conversion of isocitrate to α-ketoglutarate [1,3]. Thus, we investigated whether R-2HG is sufficient to mimic IDH1 R132H overexpression to increase NANOG expression and HDACi resistance in glioblastoma cells. We found that cellular-permeable octyl-2HG is sufficient to increase NANOG mRNA and protein levels in U87MG and U373MG glioblastoma cells ( Figure 6A,B). Consistent with previous results suggesting that NF-κB (p65) is increased by IDH1 R132H overexpression in AML cells [21], increased p65 levels were also observed in octyl-2HG-treated glioblastoma cells ( Figure 6B). In addition, high enrichment for H3K4-3Me, a marker for active promoters, on the proximal NANOG promoter region was observed in octyl-2HG-treated U87MG and U373MG glioblastoma cells ( Figure 6C). Surprisingly, we further found that octyl-2HG abolishes the anti-cancer effect of TSA in U87MG and U373MG cells ( Figure 6D). These results reveal that R-2HG, as a crucial oncometabolite, might cause HDACi resistance in IDH1 R132H -expressing glioblastoma cells.
previous results suggesting that NF-κB (p65) is increased by IDH1 R132H overexpression in AML cells [21], increased p65 levels were also observed in octyl-2HG-treated glioblastoma cells (Figure. 6B). In addition, high enrichment for H3K4-3Me, a marker for active promoters, on the proximal NANOG promoter region was observed in octyl-2HG-treated U87MG and U373MG glioblastoma cells ( Figure.  6C). Surprisingly, we further found that octyl-2HG abolishes the anti-cancer effect of TSA in U87MG and U373MG cells (Figure. 6D). These results reveal that R-2HG, as a crucial oncometabolite, might cause HDACi resistance in IDH1 R132H -expressing glioblastoma cells. Figure 6. Octyl-2HG increases NANOG expression and causes TSA resistance in U87MG and U373MG glioblastoma cells. (A) U87MG and U373MG cells were incubated in the absence or presence of octyl-2HG (1 mM) for 24 h. Gene expression was measured by using qRT-PCR. The values represent the mean ± SD of three independent experiments performed in duplicate; *p < 0.05. (B) Protein levels were measured by using western blotting. (C) U87MG and U373MG cells were incubated with or without octyl-2HG (1 mM) for 24 h. Chromatin immunoprecipitation was performed by using antibody against for H3K4-3Me. Purified DNA samples were used to analyze the promoter region in NANOG gene. The values represent the mean ± SD of three independent experiments performed in duplicate; *p < 0.05. (D) U87MG and U373MG cells were incubated with or without TSA (5 μM) or octyl-2HG (1 mM) for 48 h. Cell viability was analyzed by using crystal violet staining. The values represent the mean ± SD of three independent experiments performed in duplicate; *p < 0.05 and **p < 0.01.

Pharmacological Inhibition of IDH1 R132H Attenuates HDACi Resistance in U87MG and U373MG Glioblastoma Cells
A selective inhibitor of IDH1 R132H , AGI-5198, with an anti-cancer effect in IDH1 R132H -expressing glioma has been reported [25]. To provide a possible therapeutic strategy to overcome HDACi resistance in glioblastoma, we tested whether AGI-5198 is sufficient to attenuate HDACi resistance in IDH1 R132H -overexpressing U87MG and U373MG cells. Initially, we found that AGI-5198 significantly decreases NANOG, but not PAX6 or Notch4, in IDH1 R132H -expressing U87MG cells (Figure. 7A). Decreased NANOG protein levels were also observed in AGI-5198-treated IDH1 R132Hexpressing U87MG cells (Figure. 7B). Figure 7C shows that AGI-5198 strongly suppresses TSA resistance by approximately 50% in IDH1 R132H -overexpressing U87MG and U373MG glioblastoma cells. In addition, attenuated vorinostat resistance by AGI-5198 was also observed in IDH1 R132Hexpressing U87MG cells (Figure. 7D). These results suggest that AGI-5198 is sufficient to abolish HDACi resistance in IDH1 R132H -expressing glioblastoma cells. Given that selective inhibitors of Figure 6. Octyl-2HG increases NANOG expression and causes TSA resistance in U87MG and U373MG glioblastoma cells. (A) U87MG and U373MG cells were incubated in the absence or presence of octyl-2HG (1 mM) for 24 h. Gene expression was measured by using qRT-PCR. The values represent the mean ± SD of three independent experiments performed in duplicate; * p < 0.05. (B) Protein levels were measured by using western blotting. (C) U87MG and U373MG cells were incubated with or without octyl-2HG (1 mM) for 24 h. Chromatin immunoprecipitation was performed by using antibody against for H3K4-3Me. Purified DNA samples were used to analyze the promoter region in NANOG gene. The values represent the mean ± SD of three independent experiments performed in duplicate; * p < 0.05. (D) U87MG and U373MG cells were incubated with or without TSA (5 µM) or octyl-2HG (1 mM) for 48 h. Cell viability was analyzed by using crystal violet staining. The values represent the mean ± SD of three independent experiments performed in duplicate; * p < 0.05 and ** p < 0.01.

Pharmacological Inhibition of IDH1 R132H Attenuates HDACi Resistance in U87MG and U373MG Glioblastoma Cells
A selective inhibitor of IDH1 R132H , AGI-5198, with an anti-cancer effect in IDH1 R132H -expressing glioma has been reported [25]. To provide a possible therapeutic strategy to overcome HDACi resistance in glioblastoma, we tested whether AGI-5198 is sufficient to attenuate HDACi resistance in IDH1 R132H -overexpressing U87MG and U373MG cells. Initially, we found that AGI-5198 significantly decreases NANOG, but not PAX6 or Notch4, in IDH1 R132H -expressing U87MG cells ( Figure 7A). Decreased NANOG protein levels were also observed in AGI-5198-treated IDH1 R132H -expressing U87MG cells ( Figure 7B). Figure 7C shows that AGI-5198 strongly suppresses TSA resistance by approximately 50% in IDH1 R132H -overexpressing U87MG and U373MG glioblastoma cells. In addition, attenuated vorinostat resistance by AGI-5198 was also observed in IDH1 R132H -expressing U87MG cells ( Figure 7D). These results suggest that AGI-5198 is sufficient to abolish HDACi resistance in IDH1 R132H -expressing glioblastoma cells. Given that selective inhibitors of IDH1/2 mutant could attenuate anti-cancer efficacy of irradiation, daunorubicin and PARP inhibitor [24], it is becoming clear that extended understanding of the therapy response-modulating effects of IDH1/2 mutations would provide useful therapeutic strategy for IDH1/2-mutated cancer treatment. In the present study, we found that a selective inhibitor, AGI-5198, efficiently overcame HDACi resistance in IDH1 R132H -expressing glioblastoma cells, and our results provide that IDH1 R1332H targeting inhibitors, such as AGI-5198 and MRK-A, are promising strategy for improving the anti-cancer efficacy of HDACi-based treatment of IDH1 R132H -mutated glioblastoma. IDH1/2 mutant could attenuate anti-cancer efficacy of irradiation, daunorubicin and PARP inhibitor [24], it is becoming clear that extended understanding of the therapy response-modulating effects of IDH1/2 mutations would provide useful therapeutic strategy for IDH1/2-mutated cancer treatment.
In the present study, we found that a selective inhibitor, AGI-5198, efficiently overcame HDACi resistance in IDH1 R132H -expressing glioblastoma cells, and our results provide that IDH1 R1332H targeting inhibitors, such as AGI-5198 and MRK-A, are promising strategy for improving the anticancer efficacy of HDACi-based treatment of IDH1 R132H -mutated glioblastoma. The values represent the mean ± SD of three independent experiments performed in duplicate; *p < 0.05 and **p < 0.01. (D) IDH1-WT or IDH1 R132H expressing U87MG cells were incubated with or without vorinostat (5 μM) or AGI-5198 (5 μM) for 48 h, as indicated. Cell viability was analyzed by using crystal violet staining. The values represent the mean ± SD of three independent experiments performed in duplicate; *p < 0.05 and **p < 0.01.

Cell Culture, Cell Viability Assay and Generation of Stable Cell Lines
Glioblastoma cell lines, U87MG and U373MG, were obtained from the Korean Cell Line Bank (Seoul, Korea). Cell lines were cultured in Minimum Essential Medium Eagle alpha medium (MEM- Figure 7. The beneficial effect of AGI-5198 on overcoming to HDACi resistance in IDH1 R132H overexpressing U87MG and U373MG glioblastoma cells. (A) IDH1-WT or IDH1 R132H expressing cells were incubated in the absence or presence of 5 µM of AGI-5198 for 24 h. Gene expression was measured by using qRT-PCR. The values represent the mean ± SD of three independent experiments performed in duplicate; * p < 0.05. (B) Protein levels were measured by using western blotting. (C) IDH1-WT or IDH1 R132H expressing U87MG or U373MG cells were incubated with or without TSA or AGI-5198 (5 µM) for 48 h, as indicated. Cell viability was analyzed by using crystal violet staining. The values represent the mean ± SD of three independent experiments performed in duplicate; * p < 0.05 and ** p < 0.01. (D) IDH1-WT or IDH1 R132H expressing U87MG cells were incubated with or without vorinostat (5 µM) or AGI-5198 (5 µM) for 48 h, as indicated. Cell viability was analyzed by using crystal violet staining. The values represent the mean ± SD of three independent experiments performed in duplicate; * p < 0.05 and ** p < 0.01.

Cell Culture, Cell Viability Assay and Generation of Stable Cell Lines
Glioblastoma cell lines, U87MG and U373MG, were obtained from the Korean Cell Line Bank (Seoul, Korea). Cell lines were cultured in Minimum Essential Medium Eagle alpha medium (MEM-α) and Roswell Park Memorial Institute medium (RPMI1640) with 10% fetal bovine serum (FBS) and incubated in a humidified atmosphere at 37 • C under 20% O 2 and 5% CO 2 . Cell viability was measured by using crystal violet staining [50]. HA-tagged IDH1-wild type (WT) or IDH1 R132H overexpressing U87MG and U373MG cells were seeded into 24-well tissue culture dishes and further incubated with or without HDAC inhibitors for 48 h. Then, cells were washed and fixed with phosphate-buffered saline (PBS) and 4% paraformaldehyde for 15 min at room temperature. After fixation, the cells were stained by using 0.5% crystal violet solution for 20 min, and then stained cells were lysed in 1% SDS solution. Optical density was obtained by using an absorbance reader at 570 nm (BioTek, Winooski, VT, USA) (OD570).

Cell Cycle Analysis and Apoptosis Assays
Cell cycle analysis was performed using Muse™ cell cycle assay kit (Millipore, Burlington, MA, USA) and Mini Flow Cytometry Muse™ Cell Analyzer (Millipore, Burlington, MA, USA), as previously described [50]. Annexin-V staining was performed using Muse TM Annexin V and Dead Cell Kit (Millipore, Burlington, MA, USA) to measure the apoptotic cells population [50]. HA-tagged IDH1-WT or IDH1 R132H overexpressing U87MG cells (1 × 10 5 cells/well) were seeded onto a 6-well cell culture plate and further incubated with or without TSA for 72 h. Then, cells were washed and collected into fresh tubes. Collected cell pellets were mixed and stained with 100 µL Muse™ Annexin V and Dead Cell kit reagents (Millipore, Burlington, MA, USA) for 20 min at room temperature. Apoptotic cell numbers were analyzed by using Mini Flow Cytometry Muse™ Cell Analyzer (Millipore, Burlington, MA, USA).

Chromatin Immunoprecipitation (ChIP) Assay
HA-tagged-IDH1-WT or IDH1 R132H overexpressing U87MG cells were used for analyzing the epigenetic reprogramming on NANOG promoter region. The ChIP assay was performed by using the EZ-ChIP™ Chromatin immunoprecipitation kit (Millipore, Burlington, MA, USA). Briefly, cross-linked cells (1 × 10 7 ) by using 37% formaldehyde were sonicated in SDS lysis buffer to shear chromatin to an average length of 200 bp to 1 kb. After chromatin shearing, 100 µL of samples were mixed with 900 µL of dilution buffer, and then diluted samples were incubated with protein A/G agarose beads and antibodies recognizing H3K4-3Me, H3K27-3Me, H3K9-3Me, or H3K9-2Me for overnight at 4 • C. Immunocomplexes were washed using lithium chloride containing wash buffer, and then contaminated proteins and RNA were removed using proteinase K and RNase A. Finally, the purified DNA was used for measuring the enrichment of NANOG promoter region by using quantitative PCR. The primer sequences for ChIP-PCR are as follows: 5'-TGTGGCAGAAAGGATTGGA-3' and 5'-TTGCAGGGTCATCATCAACG-3' [52].

In Vitro Migration Assay
In vitro cell migration assay was performed using a Transwell chamber (Sigma-Aldrich, St. Louis, MO, USA) as previously described [53]. HA-tagged-IDH1-WT or IDH1 R132H overexpressing U87MG cells (3 × 10 3 ) in 0.1 mL of MEM-α medium without fetal bovine serum (FBS) were seeded into the upper side of the Transwell chamber, and the lower side of the Transwell chamber was filled with MEM-α medium contained 6% FBS, and then cells were incubated for 12 h at 37 • C to allow migration from upper to lower side. After incubation, the Transwell insert membranes were washed using phosphate-buffered saline (PBS), and then migrated cells were stained using hematoxylin and eosin. The stained cells were captured using a microscope (Olympus, Tokyo, Japan).

Statistical Analysis
The unpaired Student's t-test and one-way ANOVA with Tukey post hoc test were performed for statistical analysis, and the data are presented as the mean ± standard deviation (SD). A p value of < 0.05 was considered statistically significant.