Epigenetic regulation known for DNA methylation and histone modification is critical for securing proper gene expression and chromosomal function, and its aberration induces various pathologic conditions including cancer. Trimethylation of histone H3 on lysine 27 (H3K27me3) is known to suppress various genes related to cancer cell survival and the level of H3K27me3 may have an influence on tumor progression and malignancy. However, it remains unclear how histone methylation is regulated in response to genetic mutation and microenvironmental cues to facilitate the cancer cell survival. Here, we report a novel mechanism of the specific regulation of H3K27me3 by cooperatively two mTOR complexes, mTORC1 and mTORC2 in human glioblastoma (GBM). Integrated analyses revealed that mTORC1 upregulates the protein expression of enhancer of zeste homolog 2, a main component of polycomb repressive complex 2 which is known as H3K27-specific methyltransferase. The other mTOR complex, mTORC2, regulates production of S-adenosylmethionine, an essential substrate for histone methylation. This cooperative regulation causes H3K27 hypermethylation which subsequently promotes tumor cell survival both in vitro and in vivo xenografted mouse tumor model. These results indicate that activated mTORC1 and mTORC2 complexes cooperatively contribute to tumor progression through specific epigenetic regulation, nominating them as an exploitable therapeutic target against cancer.
A dynamic regulation of histone methylation by mTOR complexes promotes tumor growth in human GBM, but at the same time could be exploitable as a novel therapeutic target against this deadly tumor.
This article is featured in Highlights of This Issue, p. 1111
Glioblastoma (GBM) remains the most common and deadly human primary brain tumor, and efforts to develop therapies for this disease have been largely unsuccessful. To address this issue, The Cancer Genome Atlas project has produced a genetic catalog for GBM. The data in the catalog show that 60% of GBMs have genetic abnormalities in EGFR, a receptor-type tyrosine kinase (1–3), which drives abnormal activation of downstream signaling, including the PI3K–Akt–mTOR pathway (4, 5). We found previously that mTOR complexes play a key role in integrating signal transduction and metabolic pathways, which promote cell growth or proliferation downstream of EGFR in GBM (6–8). However, the detailed mechanism by which cellular proliferation is regulated by mTOR complexes in response to metabolic reprogramming remains unclear.
Metabolic reprogramming may exert some of its most important consequences by globally altering gene transcription through epigenetic mechanisms such as DNA methylation and histone modifications (9–12). Histone modifications such as methylation, acetylation, phosphorylation, and ubiquitylation are increasingly recognized to play an essential role in normal as well as abnormal conditions of the cell (13, 14). Furthermore, epigenetic changes, either alone or in combination with gene mutations, drive tumor initiation and progression (9, 14). Importantly, histone modifications may be a key player to integrate genetic mutation and metabolic reprogramming into the proper gene control for cancer cell survival through the regulation of histone-modifying enzymes as well as intermediary metabolite as the substrate for the modification (15). Here, we show that mTORC1 and mTORC2 cooperatively control histone H3 methylation to drive tumor proliferation, nominating mTOR kinase as critical epigenetic regulator of GBM through the coordinated activity of its multiprotein complexes.
Materials and Methods
Antibodies and reagents
Cell Signaling Technology antibodies: EGF Receptor vIII (catalog no. 64952), EGFR (catalog no. 4267), p-EGFR (Y1068; catalog no. 2234), Akt (catalog no. 9272), p-Akt (S473; catalog no. 4060), Raptor (catalog no. 2280), Rictor (catalog no. 9476), H3K4me2 (catalog no. 9725), H3K9me2 (catalog no. 4658), H3K27me2 (catalog no. 9755), H3K27me3 (catalog no. 9733), Histone H3 (catalog no. 4499), EZH2 (catalog no. 5246), HA-tag (catalog no. 3274), S6 (catalog no. 2317), p-S6 (S235/236; catalog no. 4857), S6K1 (catalog no. 9202), p-S6K1 (T389; catalog no. 9234), p21 (catalog no. 2947), β-actin (catalog no. 3700), GAPDH (catalog no. 5174), HRP-linked anti-rabbit IgG (catalog no. 7074), and HRP-linked anti-mouse IgG (catalog no. 7076).
Reagents used were Rapamycin (Sigma-Aldrich; catalog no. S5636), SAM (Cayman Chemical; catalog no. 13643), GSKJ4 (Sigma; catalog no. T1952), PP242 (MedChem Express; catalog no. HY-10474), DZNep (Abcam; catalog no. ab145628), GSK126 (MedChem Express; catalog no. HY-13470), Doxycycline (WAKO; catalog no. 018-20941), EGF (R&D Systems; catalog no. 236-EG), Akti-1/2 (Calbiochem; catalog no. 124018), Bisindolylmaleimide I (Bis-I; Santa Cruz Biotechnology; catalog no. sc-24003), and GSK 650394 (Tocris Bioscience; catalog no. 3572/10).
All glioma samples were obtained by surgery at Tokyo Women's Medical University Hospital (Tokyo, Japan). Physicians obtained written informed consent from the patients. All methods and experimental protocols related to human subjects were approved by institutional review board ethics committee, and the procedures related to human subjects were carried out in accordance with institutional review board–approved protocol and Declaration of Helsinki, 2013.
Cell lines and cell culture
All cell lines used in the experiments were purchased from ATCC or kind gifts from P.S. Mischel Laboratory (Ludwig Institute for Cancer Research, San Diego, CA), and cell authentication was not performed in our laboratory: U-87MG (RRID: CVCL_0022), LN229 (RRID: CVCL_0393), U373 (RRID: CVCL_2219), and H1975 (RRID: CVCL_1511). Microbiological testing for the cells including Mycoplasma testing was performed by RIKEN Yokohama Institute and ICLAS Monitoring Center with PCR testing. Human U87 GBM cells were grown in DMEM (Thermo Fisher Scientific) containing 10% FBS (Omega Scientific) and 1% penicillin and streptomycin (Thermo Fisher Scientific). LN229 and U373 GBM cells were grown in DMEM containing 10% tetracycline-free FBS and 1% penicillin and streptomycin. H1975 lung cancer cells were grown in RPMI1640 (Thermo Fisher Scientific) with 10% FBS and 1% penicillin and streptomycin. All cells were grown at 37°C in 5% CO2 incubator. The number of passages between thawing and use in each experiment were P2–3.
DNA plasmid and siRNA transfection
GFP-NTD, HA-EZH2, and Myc-Raptor DNA plasmids were obtained from Addgene. Lentiviral short hairpin RNA (shRNA) vectors targeting human Raptor, Rictor, and Scramble sequences were also obtained from Addgene. Transfection of DNA plasmids was performed using X-tremeGene HP (F. Hoffmann-La Roche) in full serum, with medium change after 24 hours, and cells were typically harvested 48 hours posttransfection. Transfection of siRNA into cell lines was carried out using Lipofectamine RNAiMAX (Thermo Fisher Scientific) in full serum, with medium change after 24 hours. siRNAs were used at 10 nmol/L, and cells were harvested 48 hours posttransfection.
Lentivirus-mediated delivery of shRNA was performed as described previously (16). The 293T packaging cell line was cotransfected with the lentiviral and target shRNA constructs by using Lipofectamine 2000 (Thermo Fisher Scientific), according to the manufacturer's instructions. Viral supernatants were harvested at 48 and 72 hours after transfection, filtered (0.45 μm), and then used for overnight infections of U87MG cells in the presence of 12.5 μg/mL polybrene. Cells were allowed to recover in fresh media and were then selected in media containing 1,100 ng/mL puromycin. Cells were analyzed on the seventh day after infection.
To extract whole-lysate protein, cultured cells were lysed and homogenized with a RIPA lysis buffer (50 mmol/L Tris-HCl, 150 mmol/L NaCl, 1% NP-40, 0.5% sodium deoxycholate, and 0.1% SDS) from Boston BioProducts. Histone protein was extracted by EpiXtract (R) Total Histone Extraction Kit (Abcam). Protein concentration of each sample was determined by the BCA Kit (Thermo Fisher Scientific) as per the manufacturer's instructions. Equal amounts of 15 μg whole lysate or 3 μg histone extract were separated by electrophoresis on 4% to 12% Miniprotean TGX Gel (Bio-Rad) in Tris/Glycine Buffer (Bio-Rad), and then transferred to a nitrocellulose membrane (Bio-Rad). Primary antibodies used in this study were described above. The immunoreactivity was detected with Super Signal West Pico Chemiluminescent Substrate or West Femto Trial Kit (Thermo Fisher Scientific), and quantitative densitometry analysis was performed with image analysis software (ImageJ version 1.49, NIH).
Quantitative reverse transcription-PCR
Total RNA was extracted with RNAeasy Extraction Kit (Qiagen). The concentration and quality of RNA was checked using a NanoDrop 2000 Spectrophotometer (Thermo Fisher Scientific). cDNA was synthesized from 1 μg of total RNA by using the iScript cDNA Synthesis Kit (Bio-Rad), and qPCR analysis was performed using the Thermal Cycler DiceR Real Time System T-800 (Takara). Each PCR was performed in triplicate. For each sample, the expression of each gene was normalized by β-actin or GAPDH expression as an internal control. The primer sequences are available upon request.
Chromatin immunoprecipitation (ChIP) experiment was performed using SimpleChIP Enzymatic Chromatin IP Kit (Cell Signaling Technology) according to the manufacturer's instruction. H3K27me3 ChIP was performed from 5.0 × 106 U87-EGFRvIII cells treated with siRNAs against control (no siRNA), scramble, siRaptor, and siRictor at 10 nmol/L (n = 3). Purified DNA was quantified by SYBR-Green real-time quantitative PCR as described above. Recoveries were calculated as percent of input.
IHC and immunofluorescence staining
Immunostaining was performed as described previously (7). Slides were counterstained with hematoxylin or DAPI (Invitrogen) to visualize nuclei. Negative control staining was also performed for each section without primary antibodies to determine the threshold for immunopositivity. Immunostained sections underwent IHC analysis in which the results were evaluated independently by two pathologists who were unaware of the findings of the molecular analyses. Immunofluorescence samples were analyzed with a fluorescent microscope (Olympus BX53 Digital Fluorescence Microscope). Nuclear scoring was performed by ImageJ-based scoring systems (ImageJ version 1.49, NIH). Cytoplasmic staining in each case was semiquantitatively evaluated as 1+ (a few weakly positive cells), 2+ (positive cells < 50%), and 3+ (positive cells ≥ 50%).
Metabolome analysis was performed as described previously (17). Briefly, metabolome analysis was conducted by C-SCOPE package of Human Metabolome Technologies (HMT) using capillary electrophoresis (CE) time-of-flight mass spectrometry for cation analysis and CE-tandem mass spectrometry for anion analysis. Hierarchical cluster analysis was performed by HMT proprietary software, PeakStat. Detected metabolites were plotted on metabolic pathway maps using VANTED software.
ELISA-based measurement of S-adenosylmethionine
S-adenosylmethionine (SAM) concentration was measured using S-adenosylmethionine ELISA Kit according to the manufacturer's instruction (Cell Biolabs). Five micrograms of each sample (n = 4) was applied to each well of the assay. The absorbance was read on a microplate reader (Multiskan GO Microplate Spectrophotometer; Thermo Fisher Scientific) at 450 nm. After correcting background from all readings, values for each sample were determined and normalized by whole-protein concentration of each sample.
Cell death and cell proliferation assay
Cell death was assessed by 0.4% Trypan blue staining (BioSystems). U87-EGFRvIII cells with siRNA against Scramble, Raptor, or Rictor (2.0 × 105 cells per well) were seeded in 10% FBS and 1% penicillin and streptomycin medium for 24 hours and then transferred to medium with DMSO or GSKJ4 (5 μmol/L; n = 6). Cell counting was performed by automated cell counter (TC20, Bio-Rad). For the cell proliferation assay with EZH2 inhibitors, we used DZNep (5 μmol/L) or GSK126 (10 μmol/L) in U87 and U87-EGFRvIII cells.
Tumor xenograft study
A total of 1.0 × 106 U87-EGFRvIII cells were resuspended in 100 μL PBS and inoculated subcutaneously into the right side of 8-week-old nude BALB/c mice-nu/nu (SLC Company). A week after the tumor graft, mice were treated with PP242 (MedChem Express), dual inhibitor of mTOR complexes for 10 mg/kg in 20% DMSO and 40% PEG-300, and 40% saline. Mice were monitored daily for tumor formation and size, and sacrificed 10 days after PP242 drug administration. For the combinatorial treatment with PP242 and GSKJ4, we used PP242 at 10 mg/kg and GSKJ4 at 50 mg/kg for 8 days. Tumors were extracted for IHC staining and fixed with 4% PFA overnight. All the animal experiments were carried out in strict accordance with the approved protocols by Tokyo Women's medical University animal research committee (AE19-056) and Notification of the Japanese Government.
Statistical significance was determined using unpaired Student t tests or one-way ANOVA (comparing all the experimental groups to the same basal control group). Error bars represent SD unless otherwise noted, and statistical significance was indicated as *, P < 0.05; **, P < 0.01; and ***, P < 0.001.
EGFR mutation and amplification are correlated with trimethylation at 27 lysine on histone H3 in GBM
The trimethylation of histone H3 on lysine 27 (H3K27me3) is one of the most important histone modifications related to cell survival in cancer, including GBM (18). We hypothesized that the aberrant signaling activated in GBM may, therefore, control cell survival through the regulation of H3K27me3. We first examined whether activating EGFR signal transduction is related to the status of H3K27me3. In support of this hypothesis, constitutively activating EGFR variant III (EGFRvIII) was strongly correlated with the expression of H3K27me3 in human GBM tissues (Fig. 1A). We further found the positive correlation between EGFR amplification and the expression of H3K27me3 in human GBM cases (Fig. 1B), suggesting that histone methylation is not necessarily dependent on specific types of the mutation, but on the activation status of EGFR signaling.
Two mTOR complexes regulate H3K27me3 in GBM cells
EGFR amplification and the activated mutation EGFRvIII enhance the activity of downstream effectors including mTOR complexes (4, 7). mTOR activation markers [p-S6 (S235/S236) for mTOR complex1 (mTORC1) and p-Akt (Ser473) for mTOR complex 2 (mTORC2)] were more expressed in human EGFR–amplified GBM samples and cell lines in accordance with the previous reports (refs. 5, 19; Supplementary Fig. S1A and S1B). We next examined the effectors to regulate H3K27me3 downstream of EGFR in GBM cells. Knockdown-type of analyses were mainly performed in this study because the basal activity of mTORC1 and mTORC2 is relatively high in U87 (due to PTEN loss) and U87-EGFRvIII cells. We found that inhibition of mTORC1 and mTORC2, using siRNAs against Raptor (a main component of mTORC1) and Rictor (a main component of mTORC2) significantly suppressed H3K27me2 and H3K27me3 in EGFR-activated GBM cell lines (Fig. 2A). We then established EGFR-activated GBM cell lines stably transfected with lentiviral vectors encoding shRaptors (shRaptor-1 and shRaptor-2) or shRictor, and found that the protein level of H3K27me3 was also repressed in Raptor- or Rictor-knockdown cells, detected by immunoblot and immunofluorescence analyses (Fig. 2B and C). Of note, H3K27me3 was also regulated by mTOR complexes in another glioma cell line with EGFRvIII (LN229 with tetracycline-on EGFRvIII system), and U87 cell line without any known EGFR mutation (Supplementary Fig. S2A and S2B) as well as H1975, a lung cancer cell line with EGFR mutation (T790M; Supplementary Fig. S2D). These results indicate that H3K27me3 is regulated by both mTORC1 and mTORC2 downstream of abnormal EGFR signaling in cancer including GBM.
mTORC1 regulates translation of EZH2, a specific methyltransferase for H3K27
To investigate the mechanism by which mTOR complexes regulate H3K27me3, we focused on the enzyme, enhancer of zeste homolog 2 (EZH2), the main component of polycomb repressive complex 2 (PRC2), H3K27-specific methyltransferase (20), because the expression level of EZH2 was associated with EGFR expression in human GBM tissues (Fig. 3A). In vitro, the EGFR activation by EGF stimulation also induced EZH2 protein expression in U87 cells (Supplementary Fig. S3A and S3B), but mRNA expression of EZH2 was not increased by EGF stimulation (Supplementary Fig. S3C). In accordance with the findings on EZH2 expression, EGFR-mutated GBM cells (U87-EGFRvIII) were more responsive to EZH2 inhibitors over EGFR wild-type GBM cells (U87) in terms of H3K27me3 level and cell proliferation (Supplementary Fig. S4).
Considering the findings on EZH2 increase by EGFR activation, we performed the inhibition of main effectors downstream of EGFR to regulate EZH2 protein level, and found that EZH2 protein expression was significantly suppressed by inhibition of mTORC1 using siRNA against Raptor, but the effect of Rictor knockdown (the main component of mTORC2) on the expression level of EZH2 was modest (Fig. 3B and C). On the other hand, Raptor or Rictor knockdown did not affect mRNA expression level of EZH2 (Supplementary Fig. S5A). Pharmacologic inhibition of mTORC1 with rapamycin (a specific mTORC1 inhibitor) decreased the protein level of EZH2 in a dose- and time-dependent manner (Fig. 3D; Supplementary Fig. S5B). mTORC1 activation with overexpression of Raptor cDNA in U87 cells conversely increased EZH2 protein expression (Fig. 3E). Furthermore, targeting S6K1, a main effector of translation downstream of mTORC1 (21, 22), significantly affected the protein expression of EZH2 (Supplementary Fig. S5C). Our findings that mTORC1 did not regulate EZH2 transcript expression but its protein expression are consistent with the idea that mTORC1 is the main regulator of protein translation in the cell (23).
Functionally, a decrease in the expression level of H3K27me3 by mTORC1 inhibition with Raptor knockdown was significantly restored by the concurrent overexpression of EZH2 cDNA plasmid (Fig. 3F). A similar trend of EZH2 regulation by mTORC1 was observed in other types of human GBM cell lines (Supplementary Fig. S2A–S2D). Taken together, these data show that mTORC1 promotes H3K27me3 expression through the translational regulation of a histone-modifying enzyme, EZH2.
mTORC2 promotes H3K27 hypermethylation through the production of SAM
Having found that the regulation of H3K27me3 expression by mTORC1 was mediated through promotion of EZH2 translation, we next examined the impact of altered mTORC2 signaling on H3K27me3 expression in GBM. In addition to mTORC1, mTORC2 also had a significant effect on the expression level of H3K27me2 and H3K27me3 although mTORC2 did not affect both mRNA and protein level of EZH2 (Fig. 3B and C; Supplementary Fig. S2A–S2C). These results indicate that H3K27me3 is regulated by both mTORC1 and mTORC2 downstream of abnormal EGFR signaling, but their regulatory mode could be different.
We previously reported that mTORC2 activation contributes to the aggressiveness of GBM cells by promoting glycolytic metabolism (6, 8, 17, 24). This led us to focus on the metabolic factors involved in mTORC2-dependent histone methylation. Comprehensive metabolomic analyses demonstrated the drastic change in intermediary metabolites by mTORC2 inhibition (Rictor knockdown; Fig. 4A). Of note, the intracellular concentration of SAM, the substrate for histone methylation (25–28), was significantly reduced in Rictor-knockdown GBM cells (Fig. 4B). ELISA analyses further confirmed that the concentration of SAM was decreased in Rictor-knockdown (mTORC2 suppressed) GBM cells, but not in Raptor-knockdown (mTORC1 suppressed) GBM cells. (Fig. 4C). Notably, the supplementation with exogenous SAM restored H3K27me3 levels in Rictor-knockdown U87 cells (Fig. 4D), although addition of SAM to the media had no effect on H3K27me3 as well as EZH2 levels in U87 cells with intact Rictor (i.e., functional mTORC2; Supplementary Fig. S6A). Of interest, pharmacologic study demonstrated that mTORC2's effect on histone methylation may be Akt (canonical effector) independent (Supplementary Fig. S6B). These results indicate that mTORC2 regulates H3K27me3 expression in GBM cells via the production of a substrate for histone methylation, SAM.
mTOR complexes promote GBM cell proliferation through the induction of H3K27 trimethylation
Having shown that mTOR complexes regulate signaling cascades related to H3K27-specific histone methylation through two independent and coordinated pathways, we determined the functional consequences of the shift in histone methylation in GBM cells. ChIP and qPCR analyses revealed that mTORC1 and mTORC2 regulate H3K27me3 in the promoter region of the tumor-suppressor gene CDKN1A as well as its transcript (Supplementary Fig. S7A–S7C), suggesting that mTOR complexes may contribute to tumor cell proliferation by inhibiting the tumor-suppressor gene CDKN1A through the promotion of histone methylation in the promoter region. To explore the effect of mTOR-dependent histone methylation on cell proliferation, we performed the cell counting assay using U87-EGFRvIII with siRNA against Raptor (mTORC1) or Rictor (mTORC2) in combination with DMSO (control) or GSKJ4 (5 μmol/L), a pharmacologic inhibitor of H3K27-specific demethylase. We found that the cell growth was recovered by GSKJ4 treatment in U87-EGFRvIII with siRNA against Raptor or Rictor, along with restored H3K27me3 expression by immunoblot (Fig. 5A–D). Of interest, the effect on cell phenotypes by GSKJ4 treatment did not contribute to cell death (Fig. 5E).
These results in vitro show that H3K27me3 shift by mTOR complexes could accelerate cell proliferation in GBM cells. We finally aimed to confirm our findings in vivo and establish a xenografted mouse tumor model for the potential therapeutic application to interfere with these epigenetic pathways. Our findings with in vitro and human samples that histone methylation was dually regulated by two mTOR complexes prompted us to apply a dual mTOR inhibitor (PP242) for the therapeutic purpose (Fig. 6A). Indeed, double knockdown of Raptor (mTORC1) and Rictor (mTORC2) was more potent in decreasing H3K27me3 than knockdown of each gene (Raptor or Rictor) or EZH2 (Supplementary Fig. S7D and S7E). Moreover, in vitro PP242 treatment significantly deterred cell proliferation in GBM cells, which was partially rescued by the addition of GSKJ4 (Fig. 5F). We then promisingly found that PP242 treatment effectively suppressed the GBM tumor growth in vivo in association with significant decrease in mTOR marker, H3K27me3 and EZH2 expression as well as an increase in p21 expression (Fig. 6B–F). Importantly, the effect of PP242 on the xenografted tumor growth was partially dependent on the level of H3K27me3, demonstrated by the concurrent administration of GSKJ4 (Fig. 6G). These results indicate that mTOR complexes promote GBM cell proliferation through the induction of H3K27 trimethylation, which is dually regulated by mTORC1 and mTORC2 (Fig. 6H).
Recent advances in the understanding of molecular–genetic aspects in cancer have revealed the critical role of epigenetics including histone modifications in the biology as well as classification of various types of cancers (17, 29, 30). H3K27 di- and trimethylation are characteristic of PcG (Polycomb group) target genes, associated with transcriptional repression for the types of developmental features like X-chromosome inactivation (13, 18). PRC1 and PRC2 are involved in chromatin condensation, and PRC2 complexes containing EZH1 or EZH2 play a major catalytic role in suppressing gene expression (13, 31–33). Of note, excessive EZH2 expression is a marker of aggressive properties of cancer, including invasion and cancer progression (13, 34, 35), indicating the importance of EZH2 overexpression in cancer. Here, we report the novel finding that aberrant EGFR signaling can engage the expression of EZH2 through translational capacity of mTORC1, which strengthens the previously reported connection between EGFR and EZH2 in GBM cells (36). mTORC1 is activated by growth factors such as insulin, and nutrients such as amino acids, which eventually promotes cell growth and proliferation by regulating anabolic and catabolic processes, as well as by driving cell-cycle progression through phosphorylating its substrates (19, 37–39). Thus, regulation of EZH2 may be one of the key downstream effectors of mTORC1 in cancer cells, suggesting a novel link between oncogenic signaling, metabolic reprogramming, and changes in epigenetic status.
To our surprise, the other mTOR complex, mTORC2, regulates the production of SAM upon aberrantly activated EGFR signaling, leading to the possible upregulation of H3K27me3. Changes in methylation status occur because of differences in the enzymatic activity of methyltransferases, including EZH2 which we found to be regulated by mTORC1. It has also been long established that SAM is the universal methyl donor for the enzymes that transfer its methyl group to yield S-adenosylhomocysteine and a methylated substrate (40). Although it has been reported that the change of methionine metabolism could regulate the methylation status of various histone modification (41), our data showed specific regulation of H3 methylation at the K27 locus by mTORC2. However, SAM is the methyl donor for various methyltransferase enzymes which can shift non-H3K27 histone methylations, and mTORC2′s specific effects on H3K27me3 might also be mediated through the other modes of methyltransferase (EZH2) regulation in addition to SAM production (42, 43). Our finding could link abnormal cancer cell metabolism, a reemerging hallmark of cancer, with epigenetic shifts essential for cancer cell survival, but the mechanisms how mTORC2 can specifically regulate H3K27 methylation through the control of methionine metabolism awaits further investigations in the future.
Recent molecular genetic studies revealed that the level of H3K27me3 has been globally shifted in high-grade pediatric glioma or other types of cancers (14, 44–47), suggesting its critical role in diagnostic markers as well as druggable targets. Indeed, a series of studies demonstrate that pharmacologic interference with H3K27 methylation was shown to be a valid and promising therapeutic strategy for treating brainstem glioma (48). From a therapeutic standpoint, this study provides the critical evidence that H3K27me3, essential for various cellular functions, is dually regulated by two mTOR complexes, suggesting that cancer cells may display the resistance to therapeutics by bypassing the block of either route to increase H3 methylation, and the combinatorial targeting of both mTOR complexes would be necessary to effectively suppress histone methylation in cancer cells. In addition, we identified CDKN1A (known as the cell-cycle arrest gene) as a potential target gene regulated by mTOR-dependent histone methylation. Notably, although the benefits of using the dual mTOR inhibitor (PP242) to treat GBM including EGFR-mutant tumors have previously been shown in multiple studies (49–51), this study demonstrated that the antiproliferative effects of PP242 are at least partially driven by H3K27me3 inhibition. The limitation of our therapeutic results derives from the speculation that treatment with PP242 will impact numerous pathways, most of which will not be related to EZH2/H3K27me3 (49). Furthermore, in vivo studies for a longer period of time in the additional experiments will be necessary to deduce more accurately the effect of PP242 treatment on tumor growth. Future studies are necessary to untangle a critical and specific regulatory network of the gene expression epigenetically controlled by both mTOR complexes via metabolic reprogramming as well as to develop more specific targeted therapies, which should lead to a novel therapeutic strategy against cancer including GBM (45, 52).
In conclusion, we demonstrate a dynamic epigenetic status of histone methylation regulated by two mTOR complexes, which enables GBM to sustain cell survival. This study provides a better understanding of how oncogenic signaling is translated into cancer-specific epigenetic status through metabolic reprogramming and provides the potential of leading to novel therapeutic strategies aimed at disrupting mTOR–histone methylation pathways, which could have inherent specificity for, at least, the most deadly type of EGFR-mutant (isocitrate dehydrogenase wild-type) GBM tumors.
Disclosure of Potential Conflicts of Interest
P.S. Mischel is chair, scientific advisory board member and co-founder, and has ownership interest (including patents) in Boundless Bio, Inc. No potential conflicts of interest were disclosed by the other authors.
Conception and design: M. Harachi, K. Masui, N. Shibata
Development of methodology: M. Harachi
Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): M. Harachi, K. Masui, H. Honda, Y. Muragaki
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): M. Harachi, K. Masui, W.K. Cavenee, P.S. Mischel, N. Shibata
Writing, review, and/or revision of the manuscript: M. Harachi, K. Masui, Y. Muragaki, W.K. Cavenee, P.S. Mischel, N. Shibata
Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): M. Harachi, T. Kawamata
Study supervision: M. Harachi, K. Masui, T. Kawamata, N. Shibata
We thank Noriko Sakayori, Shuichi Iwasaki, Hideyuki Takeiri, and Mizuho Karita (Division of Pathological Neuroscience, Department of Pathology, Tokyo Women's Medical University, Tokyo, Japan) for their IHC technical assistance, and Yoshihiko Miyakawa and Miho Koizumi (Field of Human Disease Models, Major in Advanced Life Sciences and Medicine, Institute of Laboratory Animals, Tokyo Women's Medical University, Tokyo, Japan) for their animal handling technical assistance. We are grateful to Drs. Yuichi Takakuwa, Fumio Nakamura, Ichiro Koshino, Shotaro Tanaka, and Nobuto Arashiki for helpful discussions (Department of Biochemistry, Tokyo Women's Medical University, Tokyo, Japan). We thank Dr. Sharif Jafar (Laboratory for Developmental Genetics, RIKEN Center for Integrative Medical Sciences) for help with ChIP analysis. This work was supported by a Grant-in-Aid from Takeda Science Foundation (to K. Masui), Japan Society for the Promotion of Science KAKENHI Grant 19K07649 (to K. Masui), and the NIH NS73831 (to P.S. Mischel).
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.