Oncometabolite 2-hydroxyglutarate destabilizes CDH1 to regulate quiescence and enhance chemotherapy sensitivity Authors: Hai-Hui Zhang, Xiang-Tian Chen, Si-Qi Li, Xin-Lei Yan, Meng Wang, Yan Lin, Yi-Yuan Yuan, Yun-Zi Mao, Jie Liu, Wei Xu Published: 2026 ABSTRACT Isocitrate dehydrogenase 1 mutation IDH1R132H generates the oncometabolite (R)-2-hydroxyglutarate (2-HG). However, the mechanisms underlying the better clinical outcomes in patients with IDH1 mutation remain elusive. Here, we report that 2-HG sensitizes chemotherapy by destabilizing Fizzy And Cell Division Cycle 20 Related 1(CDH1) and reducing the number of quiescent cells (G0 phase cells). IDH1R132H-harboring glioblastoma tissues have decreased CDH1, which can be induced by IDH1R132H overexpression or 2-HG treatment in U87MG cells. Mechanistically, 2-HG inhibits Prolyl Hydroxylase Domain-containing protein 1 (PHD1), which hydroxylates proline 431 of CDH1 to maintain its stability. 2-HG induces hypo-hydroxylation of CDH1, rendering it susceptible to CDK1-mediated phosphorylation at serine 151/163, followed by β-TrCP-mediated ubiquitination and proteasomal degradation. Notably, in vitro and in vivo experiments demonstrate that 2-HG induces CDH1 downregulation and reduces quiescent cell populations not only in glioblastoma cells but also in lung cancer and colorectal cancer cells, and consistently sensitizes tumor cells to chemotherapy. Our findings shed light on why IDH1 mutations correlate with better prognosis and highlight the translational potential of 2-HG as a chemotherapy sensitizer.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12964-026-02988-9. Introduction Mutations in isocitrate dehydrogenase (IDH), most frequently an arginine 132 to histidine change in IDH1 (IDH1R132H) and an arginine 172 to histidine change in IDH2 (IDH2R172H), confer a novel and aberrant enzymatic activity. Specifically, the IDH1R132H and IDH2R172H mutants gain the ability to convert α-ketoglutarate (α-KG) into (R)-2-hydroxyglutarate (R-2-HG) [1], a biochemical reaction otherwise typically observed in bacteria. IDH mutations-derived R-2-HG is an oncometabolite, which promotes oncogenesis by either activating oncogenic signaling pathways (e.g., HIF1α [2, 3]) or altering epigenetic pathways including histone and DNA methylation [4–6] thereby promoting the development of glioblastoma and acute myeloid leukemia (AML). Intriguingly, the oncogenic role of R-2-HG appears paradoxical: patients with IDH1-mutant gliomas often exhibit better clinical outcomes [7]. IDH-mutant gliomas render tumor cells exquisitely sensitive to PARP inhibitors [8] and the ferroptosis inducer erastin [9]. Similarly, primary human AML cells harboring IDH1/2 mutations show increased sensitivity to ABT-199, a highly specific BCL-2 inhibitor [10]. These observations suggest that IDH mutations sensitize tumors bearing them to chemotherapy-induced apoptosis. However, whether this sensitization is associated with R-2-HG remains elusive. Apoptosis is closely linked to cell cycle arrest. Prolonged arrest at the G1/S phase induces either terminal growth arrest (i.e., apoptosis) [11, 12], or withdrawal from the proliferative cycle into a quiescent (G0) state. The G0 phase denotes cellular quiescence, which is critical for tissue homeostasis, stem cell maintenance, and tumor dormancy. Tumors with a high proportion of G0-phase cells frequently exhibit therapy resistance and recurrence [13–15]. Quiescent (G0) cells are tightly regulated by a network of transcriptional, post-translational, and epigenetic mechanisms [16–18]. Critical regulators, such as FOXO3a, p53, Rb, CDKN1B and CDH1, play essential roles in maintaining quiescence, preventing uncontrolled proliferation, and contributing to cancer stem cell dormancy and therapy resistance [19–21]. Targeting these quiescence regulators has been shown to sensitize tumor cells to chemotherapy [22, 23] and immunotherapy [24]; furthermore, sequential administration of CDK inhibitors following taxane-based chemotherapy suppresses quiescence escape and improves therapeutic efficacy in pancreatic ductal adenocarcinoma (PDAC) models [25], underscoring the therapeutic potential of targeting cell cycle-regulatory kinases. CDH1 (also known as FZR1), a co-activator of the anaphase-promoting complex/cyclosome (APC/C), is a master regulator of cell cycle exit and quiescence maintenance [26, 27]. CDH1-driven activation of APC/C promotes the proteasomal degradation of mitotic regulators, thereby enforcing G0 entry and sustaining a non-proliferative state. CDH1 overexpression has been shown to reinforce G0-G1 arrest and confer gemcitabine resistance in PDAC, with elevated CDH1 expression correlating with poor clinical responses [28]. Conversely, Fzr1 deletion enhances DNA damage sensitivity and prolongs survival in murine models of B-cell acute lymphoblastic leukemia (B-ALL), with analogous clinical associations observed in human patients [29]. These findings demonstrate that CDH1 upregulation promotes drug resistance by maintaining cellular quiescence, while simultaneously identifying CDH1 as a promising therapeutic target for overcoming treatment resistance. The oncogenic role of R-2-HG was first established through the identification of IDH1/2 mutations in gliomas and AML [2, 30–33]. As a potent oncometabolite, R-2-HG drives tumorigenesis and malignant progression through epigenetic reprogramming, metabolic dysregulation, and signaling pathway dysregulation. Mechanistically, R-2-HG acts as a competitive inhibitor of α-KG-dependent dioxygenases, including prolyl hydroxylase domain-containing proteins (PHDs), ten-eleven translocation (TET) methylcytosine dioxygenases, and Jumonji C-domain-containing histone demethylases (JMJDs) [4–6, 34]. The suppression of TET family enzymes, particularly TET2, induces genome-wide DNA hypermethylation, leading to epigenetic silencing of tumor suppressor genes [4, 35, 36]. Inhibition of JMJDs results in the accumulation of repressive histone marks, thereby establishing a transcriptionally repressive chromatin state [5, 35]. PHD inhibition exerts dual effects on substrates: it stabilizes proline-hydroxylated proteins such as HIF-1α [36, 37] and aminoacyl-tRNA synthetase 2 (AARS2) [38], while destabilizing other substrates including the pro-apoptotic factor BIM-EL [39–41]. In the present study, we identify CDH1 as a novel substrate of PHD1. We demonstrate that hydroxylation at proline 431 is essential for CDH1 stabilization and G0-phase maintenance. Collectively, our findings delineate a previously unrecognized 2-HG–PHD1–CDH1 axis that integrates oncometabolite signaling with quiescence control and therapeutic response. Methods and materials Cell lines and culture U87MG, A549, HCT116, HEK293T, U251MG and HS683 cell lines were obtained from American Type Culture Collection (ATCC) and cultured in DMEM supplemented with 10% fetal bovine serum (FBS) and 1% penicillin–streptomycin at 37 °C in a 5% CO₂ incubator. For stable cell line generation, lentiviral constructs expressing sgRNAs or cDNAs were packaged in HEK293T cells and used to infect target cells in the presence of 8 µg/mL polybrene. Cells were treated with pharmacological agents under the following conditions: cycloheximide (CHX, 15–75 min) to inhibit protein synthesis, MG132 (10 µM, 6 h) to block proteasomal degradation, AGI5198 (10 µM) to inhibit mutant IDH1 activity, or R-2-hydroxyglutarate (2-HG, 1–20 mM) for indicated time periods depending on the experimental design. Constructs and stable cell lines Gene knockout was conducted using the lentiCRISPRv2 lentiviral system as follows: The LentiCRISPRv2 vector containing target-specific sgRNAs was co-transfected with packaging plasmids (psPAX2 and pMD2.G) into HEK293T cells using Lipofectamine 8000 transfection reagent (Beyotime Biotechnology, Shanghai, China). Viral supernatants were collected at 48 and 72 h post-transfection, centrifuged at 1,000×g for 10 min to remove cellular debris, and then used to infect target cells in the presence of 8 µg/ml polybrene. After 48 h of infection, target cells were selected with 2 µg/ml puromycin-containing medium for at least 72 h. The guide sequences are as follows:sgCDH1:F: 5'-CACCGCTCGGGGGGCAACGACAACA-3'R: 5'-AAACTGTTGTCGTTGCCCCCCGAGC-3';sgPHD1:F: 5'-CACCGGTACTACGGCATCTGCGTCA-3'R: 5'-AAACTGACGCAGATGCCGTAGTACC-3';sgCDK1:F: 5'-CACCGTATCCCTCCTGGTCAGTACA-3'R: 5'-AAACTGTACTGACCAGGAGGGATAC-3'. sgCDH1:F: 5'-CACCGCTCGGGGGGCAACGACAACA-3'R: 5'-AAACTGTTGTCGTTGCCCCCCGAGC-3'; sgCDH1:F: 5'-CACCGCTCGGGGGGCAACGACAACA-3'R: 5'-AAACTGTTGTCGTTGCCCCCCGAGC-3'; F: 5'-CACCGCTCGGGGGGCAACGACAACA-3' F: 5'-CACCGCTCGGGGGGCAACGACAACA-3' R: 5'-AAACTGTTGTCGTTGCCCCCCGAGC-3'; R: 5'-AAACTGTTGTCGTTGCCCCCCGAGC-3'; sgPHD1:F: 5'-CACCGGTACTACGGCATCTGCGTCA-3'R: 5'-AAACTGACGCAGATGCCGTAGTACC-3'; sgPHD1:F: 5'-CACCGGTACTACGGCATCTGCGTCA-3'R: 5'-AAACTGACGCAGATGCCGTAGTACC-3'; F: 5'-CACCGGTACTACGGCATCTGCGTCA-3' F: 5'-CACCGGTACTACGGCATCTGCGTCA-3' R: 5'-AAACTGACGCAGATGCCGTAGTACC-3'; R: 5'-AAACTGACGCAGATGCCGTAGTACC-3'; sgCDK1:F: 5'-CACCGTATCCCTCCTGGTCAGTACA-3'R: 5'-AAACTGTACTGACCAGGAGGGATAC-3'. sgCDK1:F: 5'-CACCGTATCCCTCCTGGTCAGTACA-3'R: 5'-AAACTGTACTGACCAGGAGGGATAC-3'. F: 5'-CACCGTATCCCTCCTGGTCAGTACA-3' F: 5'-CACCGTATCCCTCCTGGTCAGTACA-3' R: 5'-AAACTGTACTGACCAGGAGGGATAC-3'. R: 5'-AAACTGTACTGACCAGGAGGGATAC-3'. The CDH1P431L knock-in cell lines were constructed using CRISPR/Cas9-mediated genome editing technology. The sgRNA-pX458 vector and single-stranded oligonucleotide repair templates (ssODNs) were co-transfected into U87MG cells at 70–80% confluency using Lipofectamine 8000, with pX458-only transfection serving as the negative control. Twenty-four hours post-transfection, GFP-positive single cells were sorted by fluorescence-activated cell sorting (FACS) into 96-well plates and expanded for 2–3 weeks under standard culture conditions. When cells reached 60–70% confluency, PCR analysis was performed, and successful knock-in events were verified by both Sanger sequencing of genomic DNA and Western blot analysis of cell lysates. The guide sequences targeting the CDH1P431L were:F: 5′-CACCGGCTACGTGACCACCTACGAC-3′R: 5′-AAA CGTCGTAGGTGGTCACGTAGCC-3′. F: 5′-CACCGGCTACGTGACCACCTACGAC-3′ F: 5′-CACCGGCTACGTGACCACCTACGAC-3′ R: 5′-AAA CGTCGTAGGTGGTCACGTAGCC-3′. R: 5′-AAA CGTCGTAGGTGGTCACGTAGCC-3′. Stable CDK1 knockdown cell lines were generated using a lentiviral shRNA system. The pLKO.1 vector encoding CDK1-specific shRNA (shCDK1) or a non-targeting control shRNA (shScram) was co-transfected with the packaging plasmids psPAX2 and pMD2.G into HEK293T cells using Lipofectamine 8000 (Beyotime Biotechnology, Shanghai, China). Viral supernatants were collected at 48 and 72 h post-transfection, centrifuged at 1,000×g for 10 min to remove cellular debris, and used to infect target cells in the presence of 8 µg/ml polybrene. 48 h after infection, cells were selected with 2 µg/ml puromycin for at least 72 h to establish stable knockdown cell lines. Knockdown efficiency was verified by immunoblotting. The target sequences were as follows:shCDK1: GTGGAATCTTTACAGGACTAT;shScram: CCTAAGGTTAAGTCGCCCTCG. shCDK1: GTGGAATCTTTACAGGACTAT; shCDK1: GTGGAATCTTTACAGGACTAT; shScram: CCTAAGGTTAAGTCGCCCTCG. shScram: CCTAAGGTTAAGTCGCCCTCG. Antibodies anti-CDH1(GeneTex, GTX111200; Abcam, Ab217038), anti-Actin(Abcam, ab1440), anti-PHD1(Proteintech, 12984-1-AP), anti-Flag(Abmart, M20008), anti-HA(Abmart, M20003), anti-Myc (Abmart, M20002), anti-cyclin A2(Abcam, ab181591), anti-cyclin B1(Cell Signaling Technology, 4138 S), anti-cyclin D(Cell Signaling Technology, 2922 S), anti-CDK1(Cell Signaling Technology, 28439 S; HUABIO, ET1602-51), anti-CDK2(Abcam, ab235941), anti-CDK4(HUABIO, ET1612-23), anti-cyclin E(Cell Signaling Technology, 20808 S), anti-β-Trcp(HUABIO, HA500358), Goat anti-rabbit IgG(Jackson, 111-035-003), Goat anti-mouse IgG(Jackson, 115-035-003), anti-Ki-67(Biolegend, 350508). Plasmid mutagenesis Point mutation plasmids were constructed using the QuikChange II Site-Directed Mutagenesis Kit (Agilent Technologies, Santa Clara, California, USA). Briefly, mutagenic primers containing the target mutation were designed and synthesized. PCR amplification was performed using wild-type (WT) plasmids as templates, with PfuTurbo DNA polymerase to generate circular mutated plasmids. Residual WT templates were digested with Dpn I (specific for methylated DNA) at 37 °C for 1 h. The resulting PCR products were transformed into competent E. coli DH5α competent cells, followed by selection on ampicillin-resistant Luria–Bertani (LB) agar plates. Positive clones were verified by Sanger sequencing to confirm the presence of the desired mutation and absence of unintended mutations. Validated plasmids were extracted using an EndoFree plasmid kit and quantified by spectrophotometry before use. Western blot assays Cells or tumor tissues were lysed for 30 min at 4 °C in pre-chilled lysis buffer (50 mmol/L Tris-HCl pH 7.5, 150 mmol/L NaCl, 0.5% NP-40, heparin, aprotinin, pepstatin, and phenylmethylsulfonyl fluoride [PMSF])). Lysate supernatants were collected by centrifugation (4 °C, 12,000 rpm, 15 min). Equal amounts of cell lysates were subjected to SDS-PAGE and transferred onto nitrocellulose membranes (Cytiva, Marlborough, Massachusetts, USA). The membranes were blocked with 5% skim milk (BD Biosciences, Franklin Lakes, New Jersey, USA) and sequentially incubated with primary and secondary antibodies for immunoblotting. Western blot signals were detected using the Typhoon FLA 9500 imaging system (GE Healthcare, Chicago, Illinois, USA). Small RNA interference The gene-silencing small interfering RNAs (siRNAs) targeting PHD1, P4HB, P4HA2, P4HA1, and FTO used in this study were synthesized by Tsingke Biotechnology Co., Ltd (Beijing, China). siRNA transfection was performed using Lipo8000 (Life Technologies, Shanghai, China) according to the manufacturer’s protocol. Cells were transfected with siRNAs 36 h prior to harvest, and knockdown efficiency was verified by Western blot analysis. All siRNA sequences are listed as follows:si-PHD1: 5’-ACUCAUUGGUUCCUUUAAGGG-3’;si-P4HB: 5′-CAGGACGGTCATTGATTACAA-3′;si-P4HA1: 5′-GGAAUUACAGGUAGCAAAU-3′;si-P4HA2: 5’-CGTCAGGTACTACGATGTCAT-3’;si-FTO: 5’- GCACCUACAAGUACUUGAAdTdT-3’. si-PHD1: 5’-ACUCAUUGGUUCCUUUAAGGG-3’; si-PHD1: 5’-ACUCAUUGGUUCCUUUAAGGG-3’; si-P4HB: 5′-CAGGACGGTCATTGATTACAA-3′; si-P4HB: 5′-CAGGACGGTCATTGATTACAA-3′; si-P4HA1: 5′-GGAAUUACAGGUAGCAAAU-3′; si-P4HA1: 5′-GGAAUUACAGGUAGCAAAU-3′; si-P4HA2: 5’-CGTCAGGTACTACGATGTCAT-3’; si-P4HA2: 5’-CGTCAGGTACTACGATGTCAT-3’; si-FTO: 5’- GCACCUACAAGUACUUGAAdTdT-3’. si-FTO: 5’- GCACCUACAAGUACUUGAAdTdT-3’. Cell cycle analysis The collected cells were washed twice with PBS buffer. The pelleted cells were then resuspended in 1 mL of pre-chilled 75% ethanol and fixed overnight at -20 °C. After fixation, the cells were centrifuged at 1,000 rpm for 5 min (4 °C), followed by two additional PBS washes. The cells were stained with anti-Ki-67 antibody at room temperature for 30 min. Following another PBS wash, the cell suspension was labeled with 50 µg/mL propidium iodide (PI; Yeasen Biotechnology, Shanghai, China). After staining, the samples were filtered through a 300-mesh sieve to remove aggregates and analyzed using the FL1/FL2 channels of a BD FACSCalibur™ flow cytometer (BD Biosciences, USA). Experimental data were processed using FlowJo software (BD Biosciences, USA). RT‒qPCR RNA was extracted from cells using the FastPure Cell/Tissue Total RNA Isolation Kit V2 (Vazyme Biotech, Nanjing, Jiangsu, China) and reverse-transcribed into cDNA. A ChamQ SYBR qPCR Master Mix kit (enzyme) was used for quantitative RT-PCR. β-Actin served as an internal reference standard. The sequences of primers used are provided below.β-Actin: 5’-CCTCTCCCAAGTCCACACAG-3’ (forward primer) and5’-GGGCACGAAGGCTCATCATT-3’ (reverse primer).CDH1: 5’-CCGTTCGACAAAGGAGTCTGT-3’ (forward primer) and5’-GGAAGGCAGAGTCCATGCAA-3’ (reverse primer). β-Actin: 5’-CCTCTCCCAAGTCCACACAG-3’ (forward primer) and β-Actin: 5’-CCTCTCCCAAGTCCACACAG-3’ (forward primer) and 5’-GGGCACGAAGGCTCATCATT-3’ (reverse primer). 5’-GGGCACGAAGGCTCATCATT-3’ (reverse primer). CDH1: 5’-CCGTTCGACAAAGGAGTCTGT-3’ (forward primer) and CDH1: 5’-CCGTTCGACAAAGGAGTCTGT-3’ (forward primer) and 5’-GGAAGGCAGAGTCCATGCAA-3’ (reverse primer). 5’-GGAAGGCAGAGTCCATGCAA-3’ (reverse primer). In vivo ubiquitination assay Cells were co-transfected with plasmids encoding the target gene(s) for 36 h. MG132 was then added to a final concentration of 10 µM, and cells were incubated for an additional 4–6 h prior to harvest. Cells were lysed in ubiquitination lysis buffer supplemented with 1 mM DTT. Lysates were mixed thoroughly by pipetting, boiled at 99 °C for 5 min to disrupt aggregates, and centrifuged at 12,000 rpm for 10 min at 4 °C. The clarified supernatants were diluted in Tris-HCl buffer and subsequently subjected to immunoprecipitation. Pull-down HEK293T cells were transfected to express Flag- or HA-tagged proteins and lysed 48 h later in 0.5% NP-40 buffer. Flag- and HA-tagged proteins were immunoprecipitated with anti-Flag or anti-HA magnetic beads, respectively. HA-tagged proteins were competitively eluted with Flag peptide and incubated with Flag-tagged proteins bound to Flag beads to assess protein–protein interactions. Protein complexes were solubilized in SDS-PAGE loading buffer and analyzed by Western blotting. TAP assay U87MG cells were transfected with either pcDNA3.1-Flag or pcDNA3.1-CDH1-Flag plasmids. After 48 h, Flag-tagged proteins were purified by immunoprecipitation and washed three times with 50 mM ammonium bicarbonate buffer (pH 8.0). Proteins were digested with sequencing-grade trypsin (1:50, enzyme-to-substrate ratio) at 37 °C for 3 h. The resulting peptide mixtures were centrifuged at 12,000 × g for 10 min at 4 °C to remove debris, and the supernatants were lyophilized. Dried peptides were reconstituted in 0.1% formic acid and subjected to analysis by ultra-performance liquid chromatography coupled with Q-Exactive tandem mass spectrometry (UPLC–QE MS/MS). MS/MS identification of CDH1-OH CDH1 protein was overexpressed in U87MG cells and subjected to affinity purification followed by tryptic digestion, with hydroxylated peptides analyzed through MS/MS sequencing using an Orbitrap Fusion mass spectrometer (Thermo Fisher Scientific, Waltham, Massachusetts, USA) for LC peaks co-eluting with synthetic standard peptides, enabling quantification by comparing peptide counts and/or peak areas; the instrument operated in data-dependent acquisition mode with automatic switching between MS and MS/MS scans, acquiring full-scan MS spectra (m/z 350–1600) at 60,000 resolution (at m/z 200) with a 300,000 AGC target and 50 ms maximum injection time, while MS/MS scans utilized a 3s cycle time at 15,000 resolution (m/z 200) with a 50,000 intensity threshold, 200 ms maximum injection time, 200,000 AGC target, and an isolation window of 2 m/z, fragmenting 2+/3+/4 + charged ions via higher-energy collisional dissociation (HCD) at 30% normalized collision energy with a fixed first mass of 120 m/z, all under dynamic exclusion with 30 s single micro scan duration. Cell counting kit-8 assay Cells were seeded at an appropriate density (1 × 10⁴ cells/well) in a 96-well plate with 100 µL of culture medium per well. Experimental groups: blank control (medium only), negative control (untreated cells), and drug-treated groups. Incubate at 37 °C with 5% CO₂ for 24–72 h, then add 10 µL CCK-8 reagent to each well and incubate for an additional 1 h. Measure the absorbance at 450 nm using a microplate reader. For data analysis, first subtract the blank control values, then calculate relative cell viability using the formula: (Absorbance of treatment group - Blank) / (Absorbance of control group - Blank) × 100%. In vitro phosphorylation of CDH1 CDH1 was phosphorylated by the CDK1-cyclin A2 complex in vitro. Briefly, CDH1-HA or its mutant CDH1P431L-HA was purified from HEK293T cells via immunoprecipitation. Cells were transfected with 6 µg of plasmid DNA, followed by treatment with or without 2-HG 36 h post-transfection to generate differentially hydroxylated CDH1-HA proteins. The proteins were then immunoprecipitated using anti-HA magnetic beads and washed twice with CDK1 kinase reaction buffer prior to the assay. The CDK1-cyclin A2 complex was similarly purified from HEK293T cells. Specifically, 10 µg each of CDK1-Flag and cyclin A2-HA plasmids were co-transfected into HEK293T cells cultured in 15 cm dishes, followed by immunoprecipitation using anti-HA beads. The complex was eluted from the beads with 50 µg/mL HA peptide in kinase reaction buffer and kept on ice for immediate use. For the kinase assay, CDH1-HA or CDH1P431L-HA was incubated with or without the CDK1-cyclin A2 complex in 200 µL of CDK1 kinase reaction buffer (containing 100 µM ATP, 50 mM HEPES [pH 7.5], 1 mM dithiothreitol, 10 mM MgCl₂, 0.1 mM Na₃VO₄, and 1 mM NaF) at 37 °C for 1 h. Reactions were terminated by adding 5× SDS loading buffer and boiling for 10 min. Phosphorylation status was detected by Phos-tag Western blot. IHC All glioma samples were acquired from Affiliated Huashan Hospital of Fudan University. A physician or nurse practitioner obtained informed consent from the patients. The paraffin-embedded biopsy tissues used for immunohistochemical (IHC) analysis were obtained from patients with stage II glioma. The expression levels of CDH1 were determined via IHC staining with an anti-CDH1 antibody. Briefly, tissue sections underwent deparaffinization and rehydration, followed by heat-mediated antigen retrieval and incubation with H2O2 to block endogenous peroxidase activity. The sections were then blocked with 5% goat blocking serum at room temperature for 30 min, followed by incubation with the primary anti-CDH1 antibody at 4 °C overnight. After washing, sections were incubated with a biotinylated secondary antibody at room temperature for 30 min, then with ABC reagent (Agilent Technologies, Santa Clara, California, USA) for another 30 min. Staining was developed using 3,3’-diaminobenzidine (DAB), and sections were counterstained with hematoxylin. Staining was scored based on two parameters:Staining intensity: 0 (negative); 1 (weak); 2 (moderate); or 3 (strong).Percentage of positive cells: 0 (≤ 5%), 1 (6%–25%), 2 (26%–50%), 3 (51%–75%), or 4 (> 75%). Staining intensity: 0 (negative); 1 (weak); 2 (moderate); or 3 (strong). Staining intensity: 0 (negative); 1 (weak); 2 (moderate); or 3 (strong). Percentage of positive cells: 0 (≤ 5%), 1 (6%–25%), 2 (26%–50%), 3 (51%–75%), or 4 (> 75%). Percentage of positive cells: 0 (≤ 5%), 1 (6%–25%), 2 (26%–50%), 3 (51%–75%), or 4 (> 75%). The final IHC score was calculated by multiplying the staining intensity score by the positive cell percentage score. Both parameters were independently evaluated by three pathologists who were blinded to the patients’ clinical data. Phos-tag SDS-PAGE and immunoblotting Protein phosphorylation was analyzed using Phos-tag SDS-PAGE. Briefly, protein lysates were prepared in RIPA buffer supplemented with protease and phosphatase inhibitors (Roche). Equal amounts of protein (20–40 µg) were separated on 8% SDS-PAGE gels containing 50 µM Phos-tag acrylamide (Wako Pure Chemical Industries) and 100 µM MnCl₂. After electrophoresis, gels were washed twice (10 min each) in transfer buffer containing 1 mM EDTA to remove Mn²⁺, followed by one wash in transfer buffer without EDTA. Proteins were then transferred onto PVDF membranes (Millipore) using a semi-dry blotting system. Membranes were blocked with 5% non-fat milk in TBST for 1 h at room temperature and incubated with primary antibodies overnight at 4 °C. After washing, membranes were incubated with HRP-conjugated secondary antibodies for 1 h at room temperature, developed using enhanced chemiluminescence (ECL, Thermo), and visualized by an imaging system (Bio-Rad ChemiDoc Imaging System, Hercules, California, USA). For comparison, equal aliquots of the same samples were analyzed in parallel by conventional SDS-PAGE. Animal models All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC). Male BALB/c nude mice (5–6 weeks old) were subcutaneously injected with 2 × 10⁶ cancer cells suspended in 100 µL PBS. Specifically, A549 cells stably expressing either the control vector or sgPHD1 were subcutaneously inoculated into BALB/c nude mice. After tumor formation (tumor volume 50–100 mm³), the mice were randomly assigned to groups and treated with: PBS (vehicle control), 2-HG (20 mg/kg, intraperitoneal injection), cisplatin (5 mg/kg, intraperitoneal injection), or 2-HG plus cisplatin (combination therapy). Tumor volume was measured every 3 days and calculated using the formula: Volume (mm³) = (length × width²)/2. Mice were euthanized when the tumor reached 1.5 cm in any dimension, in accordance with animal welfare guidelines. Bioinformatics TCGA-LGG clinical and transcriptomic data were obtained from The Cancer Genome Atlas (TCGA) database. Gene expression data were processed as normalized RNA-seq values. Survival analyses were performed using Kaplan–Meier curves with log-rank tests. Cox proportional hazards regression analysis was used to evaluate the prognostic significance of CDH1 expression after adjusting for clinical covariates including age, sex, and tumor grade. Statistical analysis and figure construction All in vitro and in vivo experiments were performed with at least three independent biological replicates. Data are presented as mean ± SEM. Statistical analyses were performed using GraphPad Prism (v9). Comparisons between two groups were performed using two-tailed Student’s t-test. One-way ANOVA was used for multiple group comparisons when appropriate. A p-value < 0.05 was considered statistically significant. For all quantitative analyses, comparisons were performed relative to the corresponding control group (vehicle-treated, empty vector, or wild-type), unless otherwise specified. All graphs were generated using GraphPad Prism (v9) and compiled into final figures using Microsoft PowerPoint. Results IDH1R132H downregulates CDH1 in U87MG cells and glioma tissues To investigate the effect of IDH1R132H on cell cycle, we overexpressed IDH1R132H in human glioblastoma U87MG cells. IDH1R132H overexpression resulted in reduced percentage of G0/G1 cells (Fig. 1A) and facilitated cell proliferation (Fig. 1B), consistent with that IDH1R132H produces oncogenic R-2-HG that is known to promote proliferation of glial cells [2, 34].Consistent with previous reports, we observed that overexpression of IDH1R132H significantly reduced the proportion of G0-phase cells in U87MG cells (Fig. 1C). This finding suggests that IDH1R132H suppresses cellular quiescence, a well-recognized driver of chemoresistance and tumor recurrence [42, 43]. Fig. 1IDH1R132H and 2-HG reduce CDH1 levels and promote cell cycle progression in U87MG cells. A Cell cycle distribution of U87MG cells transduced with Vector, IDH1-Flag or IDH1R132H-Flag analyzed by flow cytometry. The percentages of cells in G0/G1, S, and G2/M phases are shown (n = 3 independent biological replicates). B Proliferation of U87MG cells expressing transduced with Vector, IDH1-Flag or IDH1R132H-Flag monitored over 5 days using the CCK-8 assay (n = 3). C G0-phase proportion was determined by Ki-67/PI staining in U87MG cells under indicated conditions (Vector control, IDH1R132H-Flag, or IDH1R132H-Flag + AGI-5198). Representative plots (top) and quantification (bottom) are shown (n = 3). D, E Cell cycle distribution (D) and proliferation curves (E) of U87MG cells treated with 2-HG (20 mM) (n = 3). F G0-phase proportion was evaluated by Ki-67/PI dual staining in U87MG cells treated with or without 2-HG (n = 3). G Western blot analysis of CDH1 and cell-cycle regulatory proteins in U87MG cells transfected with IDH1R132H-Myc or control vector. Expression of Cyclin A, Cyclin B, Cyclin D, Cyclin E, CDK1, CDK2 and CDK4 was examined. Actin served as the loading control, and Myc confirmed IDH1R132H expression. H Western blot analysis of CDH1 and the indicated cell-cycle proteins in U87MG cells treated with 2-HG (20 mM). I Immunohistochemistry showing CDH1 expression in IDH1 WT and IDH1-mutant glioma specimens (n = 13). Scale bar = 50 μm. J The effects of CDH1 knockout on G0-phase proportion were determined by Ki-67/PI dual staining in U87MG cells treated with or without 2-HG (n = 3). Data are presented as mean ± SEM: *P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant IDH1R132H and 2-HG reduce CDH1 levels and promote cell cycle progression in U87MG cells. A Cell cycle distribution of U87MG cells transduced with Vector, IDH1-Flag or IDH1R132H-Flag analyzed by flow cytometry. The percentages of cells in G0/G1, S, and G2/M phases are shown (n = 3 independent biological replicates). B Proliferation of U87MG cells expressing transduced with Vector, IDH1-Flag or IDH1R132H-Flag monitored over 5 days using the CCK-8 assay (n = 3). C G0-phase proportion was determined by Ki-67/PI staining in U87MG cells under indicated conditions (Vector control, IDH1R132H-Flag, or IDH1R132H-Flag + AGI-5198). Representative plots (top) and quantification (bottom) are shown (n = 3). D, E Cell cycle distribution (D) and proliferation curves (E) of U87MG cells treated with 2-HG (20 mM) (n = 3). F G0-phase proportion was evaluated by Ki-67/PI dual staining in U87MG cells treated with or without 2-HG (n = 3). G Western blot analysis of CDH1 and cell-cycle regulatory proteins in U87MG cells transfected with IDH1R132H-Myc or control vector. Expression of Cyclin A, Cyclin B, Cyclin D, Cyclin E, CDK1, CDK2 and CDK4 was examined. Actin served as the loading control, and Myc confirmed IDH1R132H expression. H Western blot analysis of CDH1 and the indicated cell-cycle proteins in U87MG cells treated with 2-HG (20 mM). I Immunohistochemistry showing CDH1 expression in IDH1 WT and IDH1-mutant glioma specimens (n = 13). Scale bar = 50 μm. J The effects of CDH1 knockout on G0-phase proportion were determined by Ki-67/PI dual staining in U87MG cells treated with or without 2-HG (n = 3). Data are presented as mean ± SEM: *P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant Treatment with 20 mM 2-HG in the culture media of U87MG cells recapitulated the effects of IDH1R132H expression, including a reduced proportion of G0/G1-phase cells (Fig. 1D), facilitated cell proliferation (Fig. 1E), and decreased the G0 phase content in U87MG cells (Fig. 1F). Consistently, low-concentration, prolonged 2-HG treatment also led to a dose-dependent reduction in CDH1 protein expression in U87MG cells (Fig. S1A). These results suggest that IDH1R132H expression induces these effects through 2-HG production. Furthermore, inhibiting 2-HG production by IDH1R132H with AGI5198 partially abrogated the ability of IDH1R132H overexpression to reduce G0/G1 cell content (Fig. S1B), facilitate cell proliferation (Fig. S1C), and decrease the G0 phase population in U87MG cells (Fig. 1C), confirming that 2-HG produced by IDH1R132H is responsible for these effects. Western blot analysis revealed that CDH1, a key regulator of G1 and G0 phases [26], was downregulated by both IDH1R132H overexpression (Fig. 1G) and 2-HG treatment (Fig. 1H), whereas other cell-cycle regulators such as cyclins and CDKs remained largely unaffected, suggesting that 2-HG regulates CDH1 to influence G1/G0 phase fate decisions. To further validate the robustness of these findings across glioblastoma models, we extended our analysis to additional glioblastoma cell lines, including U251MG and HS683. Consistently, 2-HG treatment reduced CDH1 protein levels and decreased the proportion of G0-phase cells in both cell lines (Fig. S1D-G), indicating that 2-HG–induced CDH1 downregulation and suppression of cellular quiescence are conserved across multiple glioblastoma models. The ability of IDH1R132H to regulate CDH1 was further validated in a cohort of 26 human glioma specimens, including 13 with wild-type IDH1 and 13 with IDH1R132H mutations. CDH1 levels were significantly lower in IDH1-mutant gliomas compared to grade-matched IDH1-wild-type tumors (Fig. 1I, Fig. S1H).Kaplan–Meier survival analysis of the TCGA-LGG cohort showed that patients with low CDH1 expression had significantly longer disease-specific survival (DSS, log-rank p = 0.025)(Fig. S1I) and overall survival (OS, log-rank p = 0.034) (Fig. S1J). Multivariate Cox regression analysis further demonstrated that CDH1 remained an independent prognostic factor after adjustment for age, sex and tumor grade (Fig. S1K). Consistent with these findings, CDH1 knockout (Fig. S1L), which diminished the G0 phase content and promoted proliferation (Fig. S1M), abrogated the effect of 2-HG on G0 phase reduction (Fig. 1J). These results identify CDH1 as a downstream effector of IDH1R132H/2-HG in regulating cell-cycle quiescence. 2-HG regulates CDH1 protein stability We next investigated the underlying mechanisms of by which IDH1R132H induces CDH1 suppression. Neither IDH1R132H overexpression nor 2-HG treatment altered the mRNA level of CDH1 in U87MG cells (Fig. 2A, B). In addition, Cycloheximide (CHX), which inhibits protein translation, failed to prevent the CDH1 reduction induced by either IDH1R132H overexpression or 2-HG treatment (Fig. 2C, D). These results suggest that decreased CDH1 protein stability accounts for IDH1R132H- and 2-HG–induced CDH1 degradation. Moreover, proteasomal inhibition with MG132 (Fig. 2E, F), but not treatment with the lysosome inhibitor chloroquine (Fig. 2G, H), restored CDH1 protein levels under conditions of IDH1R132H expression or 2-HG treatment. This indicates that both IDH1R132H expression and 2-HG treatment promote CDH1 proteasomal degradation. This hypothesis was further confirmed by the observation that both IDH1R132H expression and 2-HG treatment enhanced CDH1 ubiquitination (Fig. 2I, J). Fig. 22-HG or IDH1R132H-expression promotes CDH1 protein degradation. A, B CDH1 mRNA levels were measured by quantitative RT–PCR in U87MG cells treated with 2-HG (A) or transduced with IDH1R132H-Flag (B) (n = 3). C, D Cycloheximide (CHX) chase assay. Western blot showing CDH1 protein levels in U87MG cells transduced with IDH1R132H-Myc (C) or treated with 2-HG (20 mM, D) and treated with 100 µg/mL CHX at indicated time points (0, 2, 4, 6, 8 h). The graph quantifies the relative CDH1 protein levels over time (normalized to 0 h). E, F U87MG cells transduced with IDH1R132H-Myc (E) or treated with 2-HG (F) were incubated with or without the proteasome inhibitor MG132 (10 µM) for 6 h. CDH1 protein levels were examined by Western blot. G, H U87MG cells transduced with IDH1R132H-Myc (G) or treated with 2-HG (H) were incubated with or without chloroquine (CQ, 50 µM) for 6 h. CDH1 protein levels were analyzed by Western blot. I, J HEK293T cells were co-transfected with CDH1-Flag, IDH1R132H-Myc, and ubiquitin-HA, and then treated with MG132. Ubiquitinated CDH1 was assessed by immunoprecipitation followed by immunoblotting with anti-HA (I), and input Flag-CDH1 expression was verified (J). Data are shown as mean ± SEM: ns = not significant 2-HG or IDH1R132H-expression promotes CDH1 protein degradation. A, B CDH1 mRNA levels were measured by quantitative RT–PCR in U87MG cells treated with 2-HG (A) or transduced with IDH1R132H-Flag (B) (n = 3). C, D Cycloheximide (CHX) chase assay. Western blot showing CDH1 protein levels in U87MG cells transduced with IDH1R132H-Myc (C) or treated with 2-HG (20 mM, D) and treated with 100 µg/mL CHX at indicated time points (0, 2, 4, 6, 8 h). The graph quantifies the relative CDH1 protein levels over time (normalized to 0 h). E, F U87MG cells transduced with IDH1R132H-Myc (E) or treated with 2-HG (F) were incubated with or without the proteasome inhibitor MG132 (10 µM) for 6 h. CDH1 protein levels were examined by Western blot. G, H U87MG cells transduced with IDH1R132H-Myc (G) or treated with 2-HG (H) were incubated with or without chloroquine (CQ, 50 µM) for 6 h. CDH1 protein levels were analyzed by Western blot. I, J HEK293T cells were co-transfected with CDH1-Flag, IDH1R132H-Myc, and ubiquitin-HA, and then treated with MG132. Ubiquitinated CDH1 was assessed by immunoprecipitation followed by immunoblotting with anti-HA (I), and input Flag-CDH1 expression was verified (J). Data are shown as mean ± SEM: ns = not significant PHD1 mediates 2-HG–Induced CDH1 degradation Supplementation of 20 mM α-KG to the culture media of U87MG cells partially prevented the 2-HG-induced reduction of CDH1 protein levels (Fig. 3A), suggesting that dioxygenases, which utilize α-KG as a substrate and are inhibited by its structural analog 2-HG [2, 3], mediate 2-HG-induced CDH1 degradation. To identify the specific dioxygenase involved, we performed a proteomic screen using CDH1 as the bait. Among 715 proteins potentially interacting with CDH1, several dioxygenases were identified, including Egl-9 Family Hypoxia Inducible Factor 2 (EGLN2/PHD1), Prolyl 4-Hydroxylase Beta Polypeptide (P4HB), Prolyl 4-Hydroxylase Subunit Alpha 1 (P4HA1), Prolyl 4-Hydroxylase Subunit Alpha 2 (P4HA2) and Fat Mass and Obesity-associated Protein (FTO) (Fig. 3B, Fig. S2A). However, only silencing EGLN2(PHD1), but not other identified dioxygenase genes, with small interfering RNAs decreased CDH1 levels (Fig. S2B, C), suggesting that PHD1 mediates CDH1 proteasomal degradation. Fig. 3PHD1 interacts with CDH1 and regulates its ubiquitination. A CDH1 levels in U87MG cells treated with 2-HG (20 mM) and/or α-KG (10 mM) were analyzed by Western blot. B Mass spectrometry analysis identified CDH1-associated proteins from anti-Flag immunoprecipitates in U87MG cells overexpressing CDH1-Flag. C In vitro pull-down assays demonstrated the interaction between CDH1-Flag and PHD1-HA in the presence or absence of 2-HG(20 mM). D Co-immunoprecipitation (Co-IP) assays confirmed the interaction between CDH1-Flag and PHD1-HA in the presence or absence of 2-HG(20 mM). E CDH1 expression levels in wild-type (WT) and PHD1 knockout (KO) U87MG cells analyzed by Western blot. F U87MG cells transfected with PHD1-Myc were analyzed for CDH1 protein levels by Western blot. G CDH1 protein levels in U87MG cells with WT or PHD1 KO background, with or without IDH1R132H-Myc expression. H The effect of 2-HG treatment at varying concentrations (0, 10, 20 mM) on CDH1 expression was assessed by Western blot in WT and PHD1 KO U87MG cells. I, J CDH1 ubiquitination assays. HEK293T cells were co-transfected with CDH1-Flag and ubiquitin-HA, with or without PHD1 overexpression (I) or PHD1 KO (J). Ubiquitinated CDH1 was detected by anti-Flag IP followed by anti-HA immunoblotting. K, L CDH1 ubiquitination in response to IDH1R132H expression (K) or 2-HG treatment (L) in WT or PHD1 KO HEK293T cells. M Cell cycle distribution (Ki-67/PI staining) in WT and PHD1 KO U87MG cells treated with or without 2-HG (20 mM). Quantification of G0-phase proportion is shown (n = 3). Data are presented as mean ± SEM: **P < 0.01, ***P < 0.001, ns: not significant PHD1 interacts with CDH1 and regulates its ubiquitination. A CDH1 levels in U87MG cells treated with 2-HG (20 mM) and/or α-KG (10 mM) were analyzed by Western blot. B Mass spectrometry analysis identified CDH1-associated proteins from anti-Flag immunoprecipitates in U87MG cells overexpressing CDH1-Flag. C In vitro pull-down assays demonstrated the interaction between CDH1-Flag and PHD1-HA in the presence or absence of 2-HG(20 mM). D Co-immunoprecipitation (Co-IP) assays confirmed the interaction between CDH1-Flag and PHD1-HA in the presence or absence of 2-HG(20 mM). E CDH1 expression levels in wild-type (WT) and PHD1 knockout (KO) U87MG cells analyzed by Western blot. F U87MG cells transfected with PHD1-Myc were analyzed for CDH1 protein levels by Western blot. G CDH1 protein levels in U87MG cells with WT or PHD1 KO background, with or without IDH1R132H-Myc expression. H The effect of 2-HG treatment at varying concentrations (0, 10, 20 mM) on CDH1 expression was assessed by Western blot in WT and PHD1 KO U87MG cells. I, J CDH1 ubiquitination assays. HEK293T cells were co-transfected with CDH1-Flag and ubiquitin-HA, with or without PHD1 overexpression (I) or PHD1 KO (J). Ubiquitinated CDH1 was detected by anti-Flag IP followed by anti-HA immunoblotting. K, L CDH1 ubiquitination in response to IDH1R132H expression (K) or 2-HG treatment (L) in WT or PHD1 KO HEK293T cells. M Cell cycle distribution (Ki-67/PI staining) in WT and PHD1 KO U87MG cells treated with or without 2-HG (20 mM). Quantification of G0-phase proportion is shown (n = 3). Data are presented as mean ± SEM: **P < 0.01, ***P < 0.001, ns: not significant Both in vitro pull-down (Fig. 3C) and co-immunoprecipitation (Co-IP) assays (Fig. 3D) confirmed the interaction between CDH1 and PHD1. PHD1 knockdown resulted in a decrease in endogenous CDH1 levels (Fig. 3E), whereas PHD1 overexpression increased endogenous CDH1 levels (Fig. 3F), supporting that PHD1 regulates protein levels of CDH1. Moreover, PHD1 knockdown abrogated the reduction in CDH1 protein levels induced by either IDH1R132H overexpression (Fig. 3G) or 2-HG treatment (Fig. 3H), demonstrating that IDH1R132H or 2-HG decreases CDH1 through PHD1. PHD1 overexpression reduced CDH1 ubiquitination (Fig. 3I), whereas CRISPR-mediated knockout of PHD1 enhanced CDH1 ubiquitination (Fig. 3J), indicating that PHD1 normally protects CDH1 from proteasomal degradation. Importantly, loss of PHD1 abolished the increase in CDH1 ubiquitination induced by IDH1R132H overexpression or 2-HG treatment (Fig. 3K, L), demonstrating that PHD1 is required for the 2-HG-dependent regulation of CDH1 stability. These results suggest that PHD1 mediates CDH1 proteasomal degradation, likely via its proline hydroxylation activity, as it does for proteins such as HIF-1α [44]. Functionally, PHD1 deletion also impaired G0-phase maintenance and prevented 2-HG from further reducing quiescent cell populations (Fig. 3M), placing PHD1 downstream of 2-HG in this pathway. In addition, 2-HG weakened the interaction between CDH1 and PHD1 (Fig. 3C, D), further supporting the idea that 2-HG promotes CDH1 destabilization by disrupting PHD1-mediated hydroxylation. Hydroxylation of proline 431 regulates CDH1 stability and G0 cell content To further investigate whether PHD1 regulates CDH1 proteasomal degradation through its hydroxylase activity, we overexpressed a loss-of-function mutant PHD1R367A in U87MG cells. Notably, unlike wild-type PHD1, this mutant failed to elevate CDH1 protein levels (Fig. 4A), confirming that PHD1’s hydroxylase activity is required for CDH1 stabilization. To further strengthen this conclusion, we performed a rescue experiment in PHD1-knockout U87MG cells. Re-expression of wild-type PHD1 restored CDH1 protein levels, whereas treatment with 2-HG markedly reduced CDH1 abundance. In contrast, PHD1R367A mutant was unable to rescue CDH1 expression, and CDH1 levels remained low regardless of 2-HG treatment, indicating that PHD1 catalytic activity is indispensable for maintaining CDH1 stability and mediating its sensitivity to 2-HG regulation(Fig. S3A). Mass spectrometry (MS) analysis was used to identify potential hydroxylated proline residues in CDH1. In an affinity-purified CDH1 tryptic library, seven proline residues were found to be hydroxylated (Fig. 4B, Fig. S3B). Among these, proline 431 (P431) resides within a canonical PHD-recognizing sequence (Fig. 4C), suggesting it may be a key target of PHD1-mediated hydroxylation. To validate this, we screened CDH1 proline mutants and found that only mutation of P431 prevented the 2-HG-induced reduction of CDH1 protein levels in U87MG cells (Fig. S3C). This indicates that 2-HG reduces CDH1 protein stability by inhibiting its hydroxylation at P431. Fig. 4CDH1 is hydroxylated at Pro431 by PHD1, which promotes its ubiquitination and degradation. A Western blot analysis was performed to assess CDH1 protein levels in U87MG cells transfected with PHD1-Myc, PHD1R376A-Myc (catalytically inactive mutant), or empty Myc vector. B MS/MS spectrum shows the identification results of the tryptic peptide of CDH1 (upper panel) and the synthetic hydroxylated peptide (lower panel), indicating that there is a hydroxylation modification at proline 431 (Pro431). C A schematic representation was created to demonstrate the conserved LXXLAP motif in CDH1 and its putative PHD1 hydroxylation site at Pro431. D, E CDH1 and CDH1P431L were overexpressed in U87MG cells. Western blot analysis was performed to detect the effects of IDH1R132H overexpression (D) and 2-HG treatment (E) on the expression of these two CDH1 proteins. F, G Ubiquitination of WT CDH1 and CDH1P431L mutant in response to IDH1 R132H overexpression (F) or 2-HG treatment (20 mM, G). HEK293T cells were co-transfected with CDH1-Flag/CDH1P431L-Flag and ubiquitin-HA. Ubiquitinated CDH1 was detected by anti-Flag IP followed by anti-HA immunoblotting. H, I Effect of PHD1 overexpression on CDH1P431L stability (H) and ubiquitination (I) in U87MG cells. J Cell cycle distribution (Ki-67/PI staining) in U87MG cells expressing WT CDH1 or CDH1P431L mutant, treated with or without 2-HG (20 mM). Quantification of G0-phase proportion is shown (n = 3). Data are presented as mean ± SEM: **P < 0.01, ***P < 0.001, ns: not significant CDH1 is hydroxylated at Pro431 by PHD1, which promotes its ubiquitination and degradation. A Western blot analysis was performed to assess CDH1 protein levels in U87MG cells transfected with PHD1-Myc, PHD1R376A-Myc (catalytically inactive mutant), or empty Myc vector. B MS/MS spectrum shows the identification results of the tryptic peptide of CDH1 (upper panel) and the synthetic hydroxylated peptide (lower panel), indicating that there is a hydroxylation modification at proline 431 (Pro431). C A schematic representation was created to demonstrate the conserved LXXLAP motif in CDH1 and its putative PHD1 hydroxylation site at Pro431. D, E CDH1 and CDH1P431L were overexpressed in U87MG cells. Western blot analysis was performed to detect the effects of IDH1R132H overexpression (D) and 2-HG treatment (E) on the expression of these two CDH1 proteins. F, G Ubiquitination of WT CDH1 and CDH1P431L mutant in response to IDH1 R132H overexpression (F) or 2-HG treatment (20 mM, G). HEK293T cells were co-transfected with CDH1-Flag/CDH1P431L-Flag and ubiquitin-HA. Ubiquitinated CDH1 was detected by anti-Flag IP followed by anti-HA immunoblotting. H, I Effect of PHD1 overexpression on CDH1P431L stability (H) and ubiquitination (I) in U87MG cells. J Cell cycle distribution (Ki-67/PI staining) in U87MG cells expressing WT CDH1 or CDH1P431L mutant, treated with or without 2-HG (20 mM). Quantification of G0-phase proportion is shown (n = 3). Data are presented as mean ± SEM: **P < 0.01, ***P < 0.001, ns: not significant This hypothesis was further supported by the observation that replacing P431 with leucine (CDH1P431L) to mimic a non-hydroxylated state not only destabilized CDH1 protein but also prevented CDH1 degradation induced by IDH1R132H expression or 2-HG treatment (Fig. 4D, E). Additionally, the CDH1P431L mutation increased ubiquitination of CDH1 and abolished the enhancement of CDH1 ubiquitination induced by IDH1R132H overexpression or 2-HG treatment (Fig. 4F, G). Furthermore, overexpression of PHD1 failed to elevate the protein level or reduce the ubiquitination level of CDH1P431L, as it did with wild-type CDH1 (Fig. 4H, I). To confirm the critical role of P431 hydroxylation in the 2-HG-induced reduction of G0-phase cell proportion, we established a CDH1P431L knock-in U87MG cell line. CDH1P431L knock-in cells exhibited a significantly lower proportion of G0-phase cells compared to CDH1WT cells, and 2-HG treatment no longer further reduced the G0-phase cell population (Fig. 4J). 2-HG inhibits P431 hydroxylation and enhances CDH1-CDK1 and CDH1-Trcp1 interactions CDH1 is targeted for degradation via phosphorylation by cyclin-dependent kinases (CDKs) and subsequent recognition by the SCF β-Trcp E3 ubiquitin ligase complex [45, 46]. To investigate how P431 hydroxylation regulates CDH1 stability, we performed semi-endogenous immunoprecipitation (IP) assays to detect interactions between CDH1 and intracellular β-Trcp, CDK1, CDK2, and Cyclin A2. Results showed that both IDH1R132H overexpression and 2-HG treatment enhanced the binding of CDH1 to CDK1 and β-Trcp1 (Fig. 5A, B). In vitro pull-down assays further confirmed that CDH1 purified from 2-HG-treated cells or IDH1R132H-overexpressing cells exhibited stronger CDK1 binding compared to CDH1 from control cells(Fig. 5C, D). Notably, PHD1 overexpression resulted in a weakened interaction between CDH1 and CDK1 (Fig. 5E). PHD1 deficiency enhanced this interaction, which was not affected by 2-HG (Fig. 5F). Consistently, the CDH1P431L mutant exhibited enhanced binding to CDK1 and β-Trcp1 regardless of 2-HG treatment (Fig. 5G). These results suggest that hydroxylation at P431 may reduce the interaction between CDH1 and CDK1/β-Trcp1. To further validate this, we compared the CDK1 binding affinity of CDH1 from PHD1 knockout (KO) cells (where CDH1 hydroxylation is reduced) and control cells: in vitro assays showed that CDH1 from PHD1 KO cells had greater binding affinity to CDK1 (Fig. 5H). At the same time, the CDH1P431L mutant also exhibited enhanced binding to CDK1 (Fig. 5I), which remained unaffected by 2-HG treatment. (Fig. 5J). Collectively, these results support a model in which hydroxylation at P431 impairs CDK1 recognition, thereby stabilizing CDH1. Fig. 5Hydroxylation of CDH1 at P431 impairs CDK1 binding. A, B Co-IP analysis of the interaction between CDH1-Flag and endogenous CDK1, CDK2, Cyclin A, and Trcp1 in U87MG cells treated with or without 2-HG (20 mM, A) or co-transfected with IDH1R132H-Myc (B). C, D In vitro pull-down assays assessing the interaction between CDK1-HA and CDH1-Flag immunoprecipitated from U87MG cells treated with or without 2-HG (20 mM, C) or co-transfected with IDH1R132H-Myc (D). E Co-IP analysis of the interaction between CDH1-Flag and endogenous CDK1, CDK2, Cyclin A, and Trcp1 in U87MG cells co-transfected with PHD1-Myc. F Co-IP analysis of the effect of 2-HG (20 mM) on the interaction between CDH1-Flag and endogenous CDK1 in WT and PHD1 KO U87MG cells. G Effect of 2-HG treatment (20 mM) on the interaction between CDH1P431L-Flag and endogenous CDK1 in U87MG cells. H, I In vitro pull-down assays assessing the interaction between CDK1-HA and CDH1 immunoprecipitated from WT vs. PHD1 KO U87MG cells (H) or between CDK1-HA and WT CDH1 vs. CDH1P431L mutant (I). J CDH1P431L-Flag or CDH1P431L-Flag(2-HG) was purified from U87MG cells with or without 2-HG treatment through immunoprecipitation. In vitro pull-down assays were used to determine the ability of CDK1-HA to pull down CDH1P431L-Flag with or without 2-HG treatment Hydroxylation of CDH1 at P431 impairs CDK1 binding. A, B Co-IP analysis of the interaction between CDH1-Flag and endogenous CDK1, CDK2, Cyclin A, and Trcp1 in U87MG cells treated with or without 2-HG (20 mM, A) or co-transfected with IDH1R132H-Myc (B). C, D In vitro pull-down assays assessing the interaction between CDK1-HA and CDH1-Flag immunoprecipitated from U87MG cells treated with or without 2-HG (20 mM, C) or co-transfected with IDH1R132H-Myc (D). E Co-IP analysis of the interaction between CDH1-Flag and endogenous CDK1, CDK2, Cyclin A, and Trcp1 in U87MG cells co-transfected with PHD1-Myc. F Co-IP analysis of the effect of 2-HG (20 mM) on the interaction between CDH1-Flag and endogenous CDK1 in WT and PHD1 KO U87MG cells. G Effect of 2-HG treatment (20 mM) on the interaction between CDH1P431L-Flag and endogenous CDK1 in U87MG cells. H, I In vitro pull-down assays assessing the interaction between CDK1-HA and CDH1 immunoprecipitated from WT vs. PHD1 KO U87MG cells (H) or between CDK1-HA and WT CDH1 vs. CDH1P431L mutant (I). J CDH1P431L-Flag or CDH1P431L-Flag(2-HG) was purified from U87MG cells with or without 2-HG treatment through immunoprecipitation. In vitro pull-down assays were used to determine the ability of CDK1-HA to pull down CDH1P431L-Flag with or without 2-HG treatment 2-HG enhances CDK1 to phosphorylate CDH1 and induces CDH1 degradation We performed Phos-tag SDS-PAGE to analyze CDH1 phosphorylation. 2-HG treatment markedly increased CDH1 phosphorylation, while PHD1 overexpression suppressed it (Fig. 6A, B). Conversely, CDH1P431L exhibited constitutive hyperphosphorylation that was insensitive to modulation by 2-HG or PHD1 (Fig. 6C, D). These results suggest that P431 hydroxylation regulates CDH1 phosphorylation. In vitro kinase assays revealed that CDH1 isolated from 2-HG-treated U87MG cells exhibited enhanced susceptibility to CDK1-mediated phosphorylation (Fig. 6E). By contrast, CDH1P431L mutants isolated from cells displayed constitutive phosphorylation susceptibility, which remained unaltered by 2-HG treatment (Fig. 6F, G). These findings indicate that hydroxylation at P431 confers protection to CDH1 against CDK1-driven phosphorylation. Fig. 6CDK1-mediated phosphorylation is required for PHD1-induced CDH1 degradation. A-D Phos-tag SDS-PAGE analysis of CDH1 phosphorylation. U87MG cells expressing CDH1-Flag or CDH1P431L-Flag were treated with or without 2-HG (20 mM, A, C) or co-transfected with PHD1-Myc (B, D). Total CDH1 was used as a loading control. E CDH1-HA was purified by immunoprecipitation from U87MG cells treated with or without 2-hydroxyglutarate (2-HG). In vitro phosphorylation assays were carried out to determine the ability of CDK1 to phosphorylate them. F In vitro phosphorylation assays were undertaken to investigate the ability of CDK1 to phosphorylate CDH1-HA and CDH1P431L-HA immunoprecipitated from U87MG cells. G CDH1P431L-HA was purified by immunoprecipitation from U87MG cells treated with or without 2-hydroxyglutarate (2-HG). In vitro phosphorylation assays were performed to verify the ability of CDK1 to phosphorylate them. H, I Western blot analysis was carried out on CDK1-knockdown and control U87MG cells after 2-HG treatment (H) and transfection with IDH1R132H-Myc (I) to evaluate the expression level of CDH1. J, K Ubiquitination assays in CDK1-knockdown HEK293T cells. Cells were treated with 2-HG (20 mM, J) or co-transfected with IDH1R132H-Myc (K), together with CDH1-Flag and ubiquitin-HA. Ubiquitinated CDH1 was detected by anti-Flag IP followed by anti-HA immunoblotting. L, M Effect of PHD1 overexpression on CDH1 stability (L) and ubiquitination (M) in CDK1-knockdown U87MG cells. N, O Western blot analysis of CDH1 protein levels in U87MG cells expressing WT CDH1 or various phospho-mutant versions(CDH1S40A, CDH1T121A, CDH1S40A+T121A, CDH1S151A, CDH1S163A, or CDH1S151A+S163A) with or without 2-HG (20 mM) CDK1-mediated phosphorylation is required for PHD1-induced CDH1 degradation. A-D Phos-tag SDS-PAGE analysis of CDH1 phosphorylation. U87MG cells expressing CDH1-Flag or CDH1P431L-Flag were treated with or without 2-HG (20 mM, A, C) or co-transfected with PHD1-Myc (B, D). Total CDH1 was used as a loading control. E CDH1-HA was purified by immunoprecipitation from U87MG cells treated with or without 2-hydroxyglutarate (2-HG). In vitro phosphorylation assays were carried out to determine the ability of CDK1 to phosphorylate them. F In vitro phosphorylation assays were undertaken to investigate the ability of CDK1 to phosphorylate CDH1-HA and CDH1P431L-HA immunoprecipitated from U87MG cells. G CDH1P431L-HA was purified by immunoprecipitation from U87MG cells treated with or without 2-hydroxyglutarate (2-HG). In vitro phosphorylation assays were performed to verify the ability of CDK1 to phosphorylate them. H, I Western blot analysis was carried out on CDK1-knockdown and control U87MG cells after 2-HG treatment (H) and transfection with IDH1R132H-Myc (I) to evaluate the expression level of CDH1. J, K Ubiquitination assays in CDK1-knockdown HEK293T cells. Cells were treated with 2-HG (20 mM, J) or co-transfected with IDH1R132H-Myc (K), together with CDH1-Flag and ubiquitin-HA. Ubiquitinated CDH1 was detected by anti-Flag IP followed by anti-HA immunoblotting. L, M Effect of PHD1 overexpression on CDH1 stability (L) and ubiquitination (M) in CDK1-knockdown U87MG cells. N, O Western blot analysis of CDH1 protein levels in U87MG cells expressing WT CDH1 or various phospho-mutant versions(CDH1S40A, CDH1T121A, CDH1S40A+T121A, CDH1S151A, CDH1S163A, or CDH1S151A+S163A) with or without 2-HG (20 mM) As expected, CDK1 knockout rescued CDH1 protein levels and reduced its ubiquitination; it also abrogated the effects of 2-HG and IDH1R132H overexpression on CDH1 protein levels and ubiquitination (Fig. 6H-K). Additionally, CDK1 knockdown counteracted the stabilizing effect induced by PHD1 overexpression (Fig. 6L, M). These data further confirm that CDK1 activity is necessary for the degradation of unhydroxylated CDH1. To identify the phosphorylation sites critical for 2-HG-mediated regulation of CDH1, we constructed CDH1 mutants (CDH13A-40 S, CDH13A-121T, CDH13A-151 S, and CDH13A-163 S) in which three of the four known phosphorylation sites (S40, T121, S151, S163) were substituted with alanine, leaving only one intact site. The study revealed that when CDH1 retained only the S40 or T121 site, neither its phosphorylation level nor its ubiquitination level responded to 2-HG treatment (Fig. S4A-D), indicating that these two sites are likely not 2-HG regulatory targets. In contrast, when either the S151 or S163 site of CDH1 was retained, both phosphorylation and ubiquitination modifications of CDH1 responded to 2-HG signaling (Fig. S4E, H), suggesting that 2-HG promotes ubiquitination by regulating phosphorylation at S151 and S163. Notably, CDH1 mutants with single substitutions (S40A or T121A) or the double substitution (S40A/T121A) still exhibited enhanced phosphorylation and ubiquitination upon 2-HG treatment (Fig. S4I, J). However, the S151A/S163A double mutant no longer responded to 2-HG treatment (Fig. S4K, L). These results collectively confirm that S151 and S163 are the phosphorylation sites that respond to 2-HG treatment, a notion further supported by the finding that the protein levels of S151A/S163A double mutant no longer decreased in response to 2-HG treatment (Fig. 6N, O). Collectively, these findings establish that inhibition of PHD1 by 2-HG promotes CDK1-dependent phosphorylation and ubiquitin-mediated degradation of CDH1. Given the established role of CDH1 in maintaining cellular quiescence, these results suggest that 2-HG–induced CDH1 destabilization may reduce the quiescent tumor cell population and thereby increase susceptibility to cytotoxic chemotherapy. 2-HG is a sensitizer of cancer chemotherapy Based on these mechanistic findings, we next asked whether 2-HG could function as a chemotherapy sensitizer, given that depletion of quiescent tumor cell populations is known to enhance chemotherapeutic efficacy [42, 47]. To evaluate the functional relevance of CDH1 destabilization beyond glioma and leukemia, we examined the effects of 2-HG in A549 lung cancer and HCT116 colorectal cancer cells. In both cell lines, 2-HG decreased CDH1 protein levels and reduced the G0-phase cell population (Fig. 7A, Fig. S5A-C). CDH1 knockout phenocopied these effects in A549 cells (Fig. 7B), as did PHD1 silencing (Fig. 7C, D), indicating that the PHD1–CDH1 axis is conserved across diverse tumor types to regulate cell quiescence. Fig. 72-HG treatment induces G0 phase arrest and enhances chemosensitivity in A549 cells. A Western blot analysis of CDH1 expression in A549 cells treated with 0 or 20 mM 2-HG. B Cell-cycle distribution (Ki-67/PI staining) in WT and CDH1 KO A549 cells (n = 3). CDH1 knockout was verified by Western blot. C Western blot analysis of CDH1 and PHD1 expression in WT and PHD1 KO A549 cells. D Cell-cycle distribution (Ki-67/PI staining) in WT and PHD1 KO A549 cells (n = 3). E Cell proliferation of WT and PHD1 KO A549 cells under different treatments (Control, 2-HG 5 mM, Cisplatin 5 µM, 2-HG + Cisplatin) measured by CCK-8 assay (n = 3). F Tumor growth curve of WT and PHD1 KO A549 xenografts in nude mice treated with vehicle, 2-HG, Cisplatin, or 2-HG + Cisplatin (n = 6 mice per group). G Western blot analysis of CDH1 expression in xenograft tumor tissues. H Cell-cycle distribution (Ki-67/PI staining) in xenograft tumor tissues analyzed by flow cytometry (n = 3). I Schematic model illustrating the mechanism by which (R)‑2‑hydroxyglutarate sensitizes tumor cells to chemotherapy via destabilization of CDH1. Data are presented as mean ± SEM: *P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant 2-HG treatment induces G0 phase arrest and enhances chemosensitivity in A549 cells. A Western blot analysis of CDH1 expression in A549 cells treated with 0 or 20 mM 2-HG. B Cell-cycle distribution (Ki-67/PI staining) in WT and CDH1 KO A549 cells (n = 3). CDH1 knockout was verified by Western blot. C Western blot analysis of CDH1 and PHD1 expression in WT and PHD1 KO A549 cells. D Cell-cycle distribution (Ki-67/PI staining) in WT and PHD1 KO A549 cells (n = 3). E Cell proliferation of WT and PHD1 KO A549 cells under different treatments (Control, 2-HG 5 mM, Cisplatin 5 µM, 2-HG + Cisplatin) measured by CCK-8 assay (n = 3). F Tumor growth curve of WT and PHD1 KO A549 xenografts in nude mice treated with vehicle, 2-HG, Cisplatin, or 2-HG + Cisplatin (n = 6 mice per group). G Western blot analysis of CDH1 expression in xenograft tumor tissues. H Cell-cycle distribution (Ki-67/PI staining) in xenograft tumor tissues analyzed by flow cytometry (n = 3). I Schematic model illustrating the mechanism by which (R)‑2‑hydroxyglutarate sensitizes tumor cells to chemotherapy via destabilization of CDH1. Data are presented as mean ± SEM: *P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant While 2-HG alone had no inhibitory effect on cell proliferation, it significantly enhanced cisplatin-induced growth inhibition in A549, and HCT116 cells (Fig. 7E, Fig. S5D). Notably, co-treatment with low concentrations of 2-HG (1, 2.5, and 5 mM) further potentiated cisplatin-mediated growth suppression in a dose-dependent manner (Fig. S5E), supporting a pharmacologically relevant chemosensitizing effect. Importantly, this sensitization was abrogated by PHD1 knockout (Fig. 7E), confirming that 2-HG acts through inhibition of PHD1 activity. Consistent with these cross-cancer observations, we next evaluated whether this chemosensitizing effect also occurs in glioma cells. In U87MG cells, 2-HG significantly enhanced cisplatin sensitivity in wild-type cells, whereas this effect was completely abolished in CDH1P431L mutant cells (Fig. S5F), demonstrating that CDH1 destabilization is required for 2-HG–mediated chemosensitization. Xenograft assays further validated these findings: treatment with 2-HG enhanced the antitumor effects of cisplatin, accompanied by a decrease in CDH1 expression and a reduction in G0 phase occupancy (Fig. 7F-I). Taken together, these results demonstrate that 2-HG enhances chemosensitivity, at least in part by disrupting the PHD1-CDH1 axis and depleting quiescent cell populations, thereby rendering cancer cells more susceptible to cytotoxic chemotherapy. Discussion In addressing why patients with IDH1-mutant gliomas often experience more favorable clinical outcomes, our study identifies a previously unrecognized role of the oncometabolite 2-HG in regulating both cell cycle progression and cell fate. Mechanistically, 2-HG competitively inhibits PHD1-mediated hydroxylation of CDH1 at proline 431, thereby facilitating CDK1-dependent phosphorylation and subsequent ubiquitin-mediated degradation of CDH1. As a result, 2-HG-induced CDH1 destabilization compromises G0-phase maintenance and diminishes the quiescent cell pool, consistent with the oncogenic properties of 2-HG. Notably, the PHD1-CDK1-CDH1 axis (Fig. 7J) emerges as an intrinsic mechanism that integrates environmental cues, such as nutrient and oxygen availability, into cell cycle and fate determination, given that PHD1 requires both α-KG and molecular oxygen as substrates. PHD1 belongs to the α-KG and oxygen-dependent dioxygenase family. 2-HG has been widely recognized as an oncometabolite that is associated with metabolic dysregulation and alterations in α-KG–dependent dioxygenase activity, including enzymes involved in DNA and histone demethylation. These studies provide a general framework for understanding IDH1 mutation–associated metabolic changes. Our findings extend this framework by revealing a non-epigenetic layer of 2-HG-mediated regulation through PHD1-dependent CDH1 hydroxylation, thereby linking metabolic perturbation to protein stability and cell cycle control. Together, these observations suggest that this pathway may represent part of a broader metabolic regulatory network. Future studies should expand the investigation to other dioxygenases, including distinct isoforms and substrates, to delineate whether additional hydroxylase-dependent or -independent pathways converge on cell fate control during metabolic reprogramming. Since several α-KG analog metabolites—such as succinate, fumarate, and itaconate [48, 49]—accumulate in pathological contexts, it will be important to determine whether their dysregulation also perturbs cell cycle control and contributes to disease states, including cancer. Given the established links between these metabolites and tumorigenesis [50], clarifying whether they drive malignant transformation by modulating cell cycle and fate decisions is of considerable interest. Approaches integrating spatial metabolomics with transcriptomic profiling of primary tumors will be particularly valuable for validating the relevance of this pathway in clinical samples. Finally, our findings carry important translational implications. By inhibiting PHD1-dependent hydroxylation, 2-HG destabilizes CDH1, reduces tumor cell quiescence, and thereby enhances chemotherapeutic efficacy. CDH1 may thus serve as a potential biomarker of chemoresistance. Importantly, given the well-established oncogenic role of IDH1 mutations, 2-HG is not suitable as a therapeutic agent, and its clinical implications should be interpreted with caution. The primary value of this study lies in identifying a druggable regulatory axis linking metabolic state to cell cycle control, rather than proposing direct clinical application of 2-HG. From a translational perspective, targeting components of the PHD1–CDH1 axis or modulating α-ketoglutarate–dependent enzymes may represent more feasible strategies to recapitulate the chemosensitivity phenotype observed in IDH1-mutant tumors. However, given the broad substrate spectrum of α-KG–dependent dioxygenases, the potential pleiotropic effects of targeting this enzyme family should be carefully evaluated in future studies. More broadly, each component of this axis represents a potential pharmacological target to enhance tumor vulnerability during chemotherapy and related treatments. Supplementary Information Supplementary Material 1: Fig. S1: IDH1R132Hreduces CDH1 levels and promote cell cycle progression in U87MG cells. A Western blot analysis of CDH1 protein levels in U87MG cells treated with increasing concentrations of 2-HG (0, 0.5, 1, 2.5, 5, and 10 mM) B Proliferation of U87MG cells expressing IDH1R132H-Flag with or without AGI-5198 treatment was monitored over 5 days (n = 3). C Cell cycle distribution of U87MG cells transduced with Vector, IDH1R132H-Flag, or IDH1R132H-Flag + AGI-5198 (10 µM). D Western blot analysis of CDH1 protein levels in U251MG cells treated with increasing concentrations of 2-HG (0, 5, 10, and 20 mM). E G0-phase proportion was evaluated by Ki-67/PI dual staining in U251MG cells treated with or without 2-HG (n = 3). F Western blot analysis of CDH1 protein levels in HS683 cells treated with increasing concentrations of 2-HG (0, 5, 10, and 20 mM). G G0-phase proportion was evaluated by Ki-67/PI dual staining in HS683 cells treated with or without 2-HG (n = 3). H IHC staining of CDH1 in IDH1 WT and mutant clinical specimens. Scale bars = 50 μm. I, J Kaplan-Meier survival analysis of TCGA-LGG patients stratified by CDH1 expression. Low CDH1 expression was associated with significantly prolonged disease-specific survival (I) and overall survival (J). K Multivariate Cox proportional hazards regression analysis showing CDH1 as an independent prognostic factor in LGG patients after adjustment for age, gender, and tumor grade. Hazard ratios (HR) and 95% confidence intervals (CI) are presented. L Western blot confirmation of CDH1 knockout in U87MG cells. M Proliferation assay of WT and CDH1 KO U87MG cells over 5 days (n = 3). Data are presented as mean ± SEM: *P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant. Fig. S2. Identification of CDH1-associated prolyl hydroxylases. A Venn diagram showing the overlap of proteins identified in two independent CDH1 pull-down mass spectrometry analyses. Five prolyl hydroxylase family members (EGLN2, P4HB, P4HA2, P4HA1, and FTO) were detected in the intersection. B Quantitative RT‑PCR analysis of CDH1 mRNA levels in U87MG cells following knockdown of candidate prolyl hydroxylases (si-EGLN2, si-P4HB, si-P4HA1, si-P4HA2, and si-FTO)(n = 3). C Western blot analysis of CDH1 protein levels in U87MG cells after knockdown of the indicated candidate genes. Data are presented as mean ± SEM: *P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant. Fig. S3. Identification and validation of hydroxylation sites on CDH1. A Western blot analysis of CDH1 expression in PHD1-KO U87MG cells. Cells were transduced with empty vector, wild-type PHD1-Myc, or catalytically inactive PHD1R367A-Myc, with or without 2-HG treatment. B Representative LC-MS/MS analysis was performed to identify multiple proline hydroxylation sites on CDH1 in U87MG cells. The annotated sequences, PEP values, and site-specific modifications (including P431, P453, and others) are listed with corresponding ion scores. C Western blot analysis of wild‑type CDH1‑Flag and its proline‑to‑leucine mutants (CDH1P431L-Flag, CDH1P147L-Flag, CDH1P347L-Flag, CDH1P365L-Flag, and CDH1P453L-Flag) in U87MG cells treated with or without 2‑HG. CDH1 and mutant variants were detected with anti‑Flag antibody. Fig. S4. Mapping the phosphorylation sites required for 2-HG-induced CDH1 degradation. A U87MG cells expressing CDH13A-40 S (CDH1T121A/S151A/S163A-S40)-Flag were treated with or without 2-HG. Co-immunoprecipitation with Flag antibody followed by Phos-tag SDS-PAGE was used to assess CDH1 phosphorylation. B HEK293T cells were transfected with CDH13A-40 S-Flag and Ub-HA, with or without 2-HG treatment. After Co-IP with Flag antibody, Western blotting was carried out to detect CDH1 ubiquitination. C U87MG cells expressing (CDH1S40A/S151A/S163A-T121)-Flag were analyzed by Co-IP and Phos-tag SDS-PAGE to determine 2-HG–induced phosphorylation changes. D HEK293T cells were co-transfected with CDH13A-121T-Flag and Ub-HA, and ubiquitination of CDH1 was examined after 2-HG treatment by Co-IP with Flag antibody. E U87MG cells expressing(CDH1S40A/T121A/S163A-S151)-Flag were subjected to Co-IP and Phos-tag SDS-PAGE to detect phosphorylation changes in the presence or absence of 2-HG. F HEK293T cells were co-transfected with CDH13A-151 S-Flag and Ub-HA. Ubiquitination of CDH1 after 2-HG treatment was evaluated by Flag Co-IP and immunoblotting. G U87MG cells expressing (CDH1S40A/T121A/S151A-S163)-Flag were analyzed by Co-IP and Phos-tag SDS-PAGE to assess phosphorylation under 2-HG treatment. H HEK293T cells transfected with CDH13A-163 S-Flag and Ub-HA were subjected to Co-IP and Western blotting to evaluate ubiquitination in response to 2-HG. I U87MG cells expressing CDH1-Flag were treated with 2-HG. Phosphorylation changes were analyzed by Co-IP and Phos-tag SDS-PAGE. J HEK293T cells were transfected with CDH1WT-Flag, CDH1T121A-Flag, CDH1S40A-Flag, or CDH1S40A+T121A-Flag together with Ub-HA. After 2-HG treatment, CDH1 ubiquitination was examined by Flag Co-IP and Western blot. K U87MG cells were transfected with CDH1WT-Flag, CDH1S151A-Flag, CDH1S163A-Flag, or CDH1S151A+S163A-Flag. Phosphorylation changes in response to 2-HG were analyzed by Co-IP and Phos-tag SDS-PAGE. L HEK293T cells expressing CDH1WT-Flag, CDH1S151A-Flag, CDH1S163A-Flag, or CDH1S151A+S163A-Flag together with Ub-HA were treated with or without 2-HG. CDH1 ubiquitination was assessed by Flag Co-IP and immunoblotting. Fig. S5. 2-HG treatment induces G0 phase arrest and enhances chemosensitivity across multiple cell lines. A Cell cycle distribution (Ki‑67/PI staining) in A549 cells treated with 0 or 10 mM 2‑HG. Representative flow plots (left) and quantification of cell cycle phases (right) are shown (n = 3). B Western blot analysis of CDH1 protein levels in HCT116 cells treated with 0 or 20 mM 2‑HG. C Cell cycle distribution (Ki‑67/PI staining) in HCT116 cells treated with 0 or 10 mM 2‑HG. Representative flow plots (left) and quantification (right) are shown (n = 3). D, E Cell proliferation (OD450) in HCT116 (D) and U87MG (E) cells under control, 2‑HG, cisplatin (Cis), or 2‑HG + cisplatin treatment (n = 3). F Proliferation curves comparing wild-type (WT) U87MG cells and CDH1P431L KIcells under different treatment conditions (± 2-HG, ±cisplatin). Data are presented as mean ± SEM: *P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant. Supplementary Material 1: Fig. S1: IDH1R132Hreduces CDH1 levels and promote cell cycle progression in U87MG cells. A Western blot analysis of CDH1 protein levels in U87MG cells treated with increasing concentrations of 2-HG (0, 0.5, 1, 2.5, 5, and 10 mM) B Proliferation of U87MG cells expressing IDH1R132H-Flag with or without AGI-5198 treatment was monitored over 5 days (n = 3). C Cell cycle distribution of U87MG cells transduced with Vector, IDH1R132H-Flag, or IDH1R132H-Flag + AGI-5198 (10 µM). D Western blot analysis of CDH1 protein levels in U251MG cells treated with increasing concentrations of 2-HG (0, 5, 10, and 20 mM). E G0-phase proportion was evaluated by Ki-67/PI dual staining in U251MG cells treated with or without 2-HG (n = 3). F Western blot analysis of CDH1 protein levels in HS683 cells treated with increasing concentrations of 2-HG (0, 5, 10, and 20 mM). G G0-phase proportion was evaluated by Ki-67/PI dual staining in HS683 cells treated with or without 2-HG (n = 3). H IHC staining of CDH1 in IDH1 WT and mutant clinical specimens. Scale bars = 50 μm. I, J Kaplan-Meier survival analysis of TCGA-LGG patients stratified by CDH1 expression. Low CDH1 expression was associated with significantly prolonged disease-specific survival (I) and overall survival (J). K Multivariate Cox proportional hazards regression analysis showing CDH1 as an independent prognostic factor in LGG patients after adjustment for age, gender, and tumor grade. Hazard ratios (HR) and 95% confidence intervals (CI) are presented. L Western blot confirmation of CDH1 knockout in U87MG cells. M Proliferation assay of WT and CDH1 KO U87MG cells over 5 days (n = 3). Data are presented as mean ± SEM: *P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant. Fig. S2. Identification of CDH1-associated prolyl hydroxylases. A Venn diagram showing the overlap of proteins identified in two independent CDH1 pull-down mass spectrometry analyses. Five prolyl hydroxylase family members (EGLN2, P4HB, P4HA2, P4HA1, and FTO) were detected in the intersection. B Quantitative RT‑PCR analysis of CDH1 mRNA levels in U87MG cells following knockdown of candidate prolyl hydroxylases (si-EGLN2, si-P4HB, si-P4HA1, si-P4HA2, and si-FTO)(n = 3). C Western blot analysis of CDH1 protein levels in U87MG cells after knockdown of the indicated candidate genes. Data are presented as mean ± SEM: *P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant. Fig. S3. Identification and validation of hydroxylation sites on CDH1. A Western blot analysis of CDH1 expression in PHD1-KO U87MG cells. Cells were transduced with empty vector, wild-type PHD1-Myc, or catalytically inactive PHD1R367A-Myc, with or without 2-HG treatment. B Representative LC-MS/MS analysis was performed to identify multiple proline hydroxylation sites on CDH1 in U87MG cells. The annotated sequences, PEP values, and site-specific modifications (including P431, P453, and others) are listed with corresponding ion scores. C Western blot analysis of wild‑type CDH1‑Flag and its proline‑to‑leucine mutants (CDH1P431L-Flag, CDH1P147L-Flag, CDH1P347L-Flag, CDH1P365L-Flag, and CDH1P453L-Flag) in U87MG cells treated with or without 2‑HG. CDH1 and mutant variants were detected with anti‑Flag antibody. Fig. S4. Mapping the phosphorylation sites required for 2-HG-induced CDH1 degradation. A U87MG cells expressing CDH13A-40 S (CDH1T121A/S151A/S163A-S40)-Flag were treated with or without 2-HG. Co-immunoprecipitation with Flag antibody followed by Phos-tag SDS-PAGE was used to assess CDH1 phosphorylation. B HEK293T cells were transfected with CDH13A-40 S-Flag and Ub-HA, with or without 2-HG treatment. After Co-IP with Flag antibody, Western blotting was carried out to detect CDH1 ubiquitination. C U87MG cells expressing (CDH1S40A/S151A/S163A-T121)-Flag were analyzed by Co-IP and Phos-tag SDS-PAGE to determine 2-HG–induced phosphorylation changes. D HEK293T cells were co-transfected with CDH13A-121T-Flag and Ub-HA, and ubiquitination of CDH1 was examined after 2-HG treatment by Co-IP with Flag antibody. E U87MG cells expressing(CDH1S40A/T121A/S163A-S151)-Flag were subjected to Co-IP and Phos-tag SDS-PAGE to detect phosphorylation changes in the presence or absence of 2-HG. F HEK293T cells were co-transfected with CDH13A-151 S-Flag and Ub-HA. Ubiquitination of CDH1 after 2-HG treatment was evaluated by Flag Co-IP and immunoblotting. G U87MG cells expressing (CDH1S40A/T121A/S151A-S163)-Flag were analyzed by Co-IP and Phos-tag SDS-PAGE to assess phosphorylation under 2-HG treatment. H HEK293T cells transfected with CDH13A-163 S-Flag and Ub-HA were subjected to Co-IP and Western blotting to evaluate ubiquitination in response to 2-HG. I U87MG cells expressing CDH1-Flag were treated with 2-HG. Phosphorylation changes were analyzed by Co-IP and Phos-tag SDS-PAGE. J HEK293T cells were transfected with CDH1WT-Flag, CDH1T121A-Flag, CDH1S40A-Flag, or CDH1S40A+T121A-Flag together with Ub-HA. After 2-HG treatment, CDH1 ubiquitination was examined by Flag Co-IP and Western blot. K U87MG cells were transfected with CDH1WT-Flag, CDH1S151A-Flag, CDH1S163A-Flag, or CDH1S151A+S163A-Flag. Phosphorylation changes in response to 2-HG were analyzed by Co-IP and Phos-tag SDS-PAGE. L HEK293T cells expressing CDH1WT-Flag, CDH1S151A-Flag, CDH1S163A-Flag, or CDH1S151A+S163A-Flag together with Ub-HA were treated with or without 2-HG. CDH1 ubiquitination was assessed by Flag Co-IP and immunoblotting. Fig. S5. 2-HG treatment induces G0 phase arrest and enhances chemosensitivity across multiple cell lines. A Cell cycle distribution (Ki‑67/PI staining) in A549 cells treated with 0 or 10 mM 2‑HG. Representative flow plots (left) and quantification of cell cycle phases (right) are shown (n = 3). B Western blot analysis of CDH1 protein levels in HCT116 cells treated with 0 or 20 mM 2‑HG. C Cell cycle distribution (Ki‑67/PI staining) in HCT116 cells treated with 0 or 10 mM 2‑HG. Representative flow plots (left) and quantification (right) are shown (n = 3). D, E Cell proliferation (OD450) in HCT116 (D) and U87MG (E) cells under control, 2‑HG, cisplatin (Cis), or 2‑HG + cisplatin treatment (n = 3). F Proliferation curves comparing wild-type (WT) U87MG cells and CDH1P431L KIcells under different treatment conditions (± 2-HG, ±cisplatin). Data are presented as mean ± SEM: *P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant. Supplementary Material 2. Supplementary Material 2.