Journal: Nature Communications

Article Title: α-KG conformationally stabilizes nuclear GLUD1-SNAIL complex to fuel trans-endothelial migration in metastasis

doi: 10.1038/s41467-026-77149-0

Figure Lengend Snippet: A VE-Cadherin protein and mRNA levels in NSCLC cells following GLUD1 knockdown or overexpression. B Representative immunoblotting analysis of GLUD1 distribution in nuclear and cytoplasmic fractions ( n = 3 independent experiments). C Confirmed interaction of GLUD1 and Importin-8 with co-IP assay. A representative immunoblot from three independent experiments is shown. D , E GLUD1 localization assessed by immunofluorescence D and cytoplasm/nuclear levels with immunoblot assay E in A549/DTX following IPO8 knockdown. Representative results from three independent experiments are shown. F Subcellular localization of FLAG-GLUD1 with or without its predicted nuclear localization sequence (NLS), together with VE-cadherin expression, in A549/DTX cells. The predicted NLS spans amino acids 517–550 and was identified using NLS Mapper ( https://nls-mapper.iab.keio.ac.jp/ ). Representative results from three independent experiments are shown. G Workflow of DNA pulldown-coupled mass spectrometry with different preparation methods. H Circular heatmap of proteins identified under stringent (red) or mild (blue) conditions. I JASPAR-predicted SNAIL binding motif in the CDH5 promoter and ChIP-qPCR analysis of GLUD1 occupancy at the predicted site in siNC- or si GLUD1 -treated A549/DTX cells ( n = 5 biologically independent samples). Data were normalized to input, Motif1: TGCAGGTGCT. J Dual-luciferase analysis of wild-type and mutant CDH5 promoters in vector- or GLUD1 -overexpressing A549 cells ( n = 6 biologically independent samples), normalized to the wild-type promoter/vector group. K Sequential ChIP using anti-GLUD1 followed by anti-SNAIL antibodies, with IgG controls, in siNC- or si GLUD1 -treated A549/DTX cells ( n = 6 biologically independent samples). L Complex levels in cytoplasmic and nuclear fractions (representative of three independent experiments). M Nuclear complex in A549/DTX with si GLUD1 treatment (left) or A549 with oe GLUD1 treatment (right). Representative results from three independent experiments are shown. N , O ChIP analysis of GLUD1 and SNAIL binding on CDH5 promoter following sequential si SNAI1 and oe GLUD1 treatments, normalized to input ( n = 5 biologically independent samples). P Representative immunoblotting analysis of VE-Cadherin protein levels after these manipulations ( n = 3 independent experiments). Q Proposed model of GLUD1/SNAIL complex binding to the CDH5 promoter. R , S Processing of paired tumor-adjacent and tumor tissues ( R ) and analysis of nuclear complex levels ( S , representative results from four biologically independent samples). T , U Sample-processing workflow T and analysis of nuclear complexes in leptomeningeal metastasis-competent A549/DTX cells ( U , representative results from five biologically independent samples). Data are represented as mean ± SEM. P values were calculated using two-way ANOVA followed by Šídák multiple-comparisons test ( I , K ); one-way ANOVA followed by Tukey’s multiple-comparisons test ( J ); one-way ANOVA followed by Dunnett’s multiple-comparisons test ( N , O ). Scale bars: D , 20 μm. Source data are provided as a Source Data file. For immunoblots, samples within each panel were derived from the same experiment, with separate gels processed in parallel where applicable; detailed gel assignments are provided in the Methods.

Article Snippet: The antibodies used were shown as followed: VE-Cadherin (ab33168; 1:1000, Abcam), Lamin B (ab16048; 1:5000, Abcam), Sodium/Potassium ATPase alpha 1 (ab308329; 1:1000, Abcam), FLAG (ab205606; 1:10,000, Abcam), ACTIN (ab8226, 1:2000, Abcam), GLUD1 (14299-1-AP; 1:1000, Proteintech), FAK (12636-1-AP; 1:1000, Proteintech), EPHA2 (25749-1-AP; 1:1000, Proteintech), SNAIL (13099-1-AP; 1:1000, Proteintech), β-Tubulin (80713-1-RR; 1:1000, Proteintech), Phospho-FAK (83933-1-RR, 1:1000, Proteintech), Phospho-EPHA2 (25749-1-AP, 1:1000, Proteintech), SOX2 (11064-1-AP, 1:500, Proteintech), SLUG (12129-1-AP, 1:500, Proteintech), ZEB1 (21544-1-AP, 1:500, Proteintech).

Techniques: Knockdown, Over Expression, Western Blot, Analysis, Co-Immunoprecipitation Assay, Immunofluorescence, Sequencing, Expressing, Mass Spectrometry, Binding Assay, ChIP-qPCR, Luciferase, Mutagenesis, Plasmid Preparation, Derivative Assay

Journal: Nature Communications

Article Title: α-KG conformationally stabilizes nuclear GLUD1-SNAIL complex to fuel trans-endothelial migration in metastasis

doi: 10.1038/s41467-026-77149-0

Figure Lengend Snippet: A – F Representative images A – C and quantification ( D – F ) of adhesion ( A , D , n = 4 biologically independent samples), barrier-destruction ( B , E , n = 3 biologically independent samples) and transmigration ( C , F , n = 3 biologically independent samples) in A549/DTX cells treated with DMSO or R162 (50 μM). Cells were counted; tight-junction fluorescence was normalized to 30 min. G , H Intravasation and extravasation under the same treatments ( n = 3 biologically independent samples), normalized to control. I – J GLUD1 distribution I and nuclear GLUD1/SNAIL complexes J after R162 treatment (representative of three independent experiments). K GLUD1 binding to predicted SNAIL sites following 50 μM R162 treatment, all data were normalized to input, n = 5 biologically independent samples. L Sequential ChIP using anti-GLUD1 then anti-SNAIL, with IgG controls, under both treatments, n = 6 biologically independent samples. M Nuclear α-KG measured by LC-MS/MS under both treatments ( n = 3 biologically independent samples), normalized to control. N , O Complex formation in purified proteins with 0–1000 μM α-KG (immunoblots, ( N ); quantification, ( O )). Representative results from three independent experiments are shown. Arrows indicate the estimated nuclear α-KG concentrations in the indicated cell lines, each calculated as the mean of three biologically independent samples. P Nuclear complexes in R162-treated A549 cells expressing GLUD1 R496S (representative of three independent experiments). Q , R Nuclear complexes ( Q ) and VE-Cadherin expression ( R ) after rescue with 10 μM dimethyl-α-KG (representative of three independent experiments). S , T Pull-down assays of GLUD1/SNAIL binding using GST-tagged ( S ) or FLAG-tagged ( T ) GLUD1 with 100 μM α-KG (representative of three independent experiments). U , V Promoter pull-down analysis of α-KG/GLUD1/SNAIL ternary-complex formation with 100 μM α-KG (scheme ( U ); quantification ( V ); n = 6 biologically independent samples). W Microscale thermophoresis of α-KG binding to GFP-GLUD1 or GFP-SNAIL (0.00023–7.5 mM; n = 3 biologically independent samples). X CETSA validation of α-KG binding to GLUD1 using 10 μM dimethyl-α-KG. Quantification is from one representative experiment ( n = 1); the experiment was independently repeated three times with similar results. The corresponding immunoblot is shown in Supplementary Fig. . All data are represented as mean ± SEM. P values were calculated using two-tailed unpaired Student’s t test ( D , F , H , K , M ); two-way ANOVA followed by Šídák multiple-comparisons test ( E , L ); one-way ANOVA followed by Tukey’s multiple-comparisons test ( V ). Scale bars: A , C 1000 μm; B 20 μm. Source data are provided as a Source Data file. For immunoblots, samples within each panel were derived from the same experiment, with separate gels processed in parallel where applicable; detailed gel assignments are provided in the Methods.

Article Snippet: The antibodies used were shown as followed: VE-Cadherin (ab33168; 1:1000, Abcam), Lamin B (ab16048; 1:5000, Abcam), Sodium/Potassium ATPase alpha 1 (ab308329; 1:1000, Abcam), FLAG (ab205606; 1:10,000, Abcam), ACTIN (ab8226, 1:2000, Abcam), GLUD1 (14299-1-AP; 1:1000, Proteintech), FAK (12636-1-AP; 1:1000, Proteintech), EPHA2 (25749-1-AP; 1:1000, Proteintech), SNAIL (13099-1-AP; 1:1000, Proteintech), β-Tubulin (80713-1-RR; 1:1000, Proteintech), Phospho-FAK (83933-1-RR, 1:1000, Proteintech), Phospho-EPHA2 (25749-1-AP, 1:1000, Proteintech), SOX2 (11064-1-AP, 1:500, Proteintech), SLUG (12129-1-AP, 1:500, Proteintech), ZEB1 (21544-1-AP, 1:500, Proteintech).

Techniques: Transmigration Assay, Fluorescence, Control, Binding Assay, Liquid Chromatography with Mass Spectroscopy, Purification, Western Blot, Expressing, Pull Down Assay, Analysis, Microscale Thermophoresis, Biomarker Discovery, Two Tailed Test, Derivative Assay

Journal: Nature Communications

Article Title: α-KG conformationally stabilizes nuclear GLUD1-SNAIL complex to fuel trans-endothelial migration in metastasis

doi: 10.1038/s41467-026-77149-0

Figure Lengend Snippet: A Predicted α-KG binding pocket in GLUD1 (AutoDock) and visualization using PyMOL. B MST confirming disrupted α-KG binding to K147A-GLUD1 mutant comparing with wildtype GLUD1, the concentration of α-KG ranges from 15 mM to 0.00046 mM, n = 3 biologically independent samples. C Metabolic activity of wild-type versus K147A-GLUD1, n = 5 biologically independent samples. D Conformational changes in GLUD1 upon α-KG binding. E Circular Dichroism result of GLUD1 with or without 100 μM α-KG. A representative result from three independent experiments is shown. F Molecular dynamics simulations revealing α-KG/GLUD1/SNAIL interaction interfaces visualizing with PyMOL, The displayed structural snapshot was obtained from one MD simulation trajectory. G , H Domain mapping: GLUD1 (amino acids 1–158) and SNAIL (amino 1–100) mediate complex formation (representative of three independent experiments). I Residue interaction maps for GLUD1-SNAIL complex formation with molecular dynamics assay, binding sites with significantly high contribution in GLUD1 (amino acids 1–158) and SNAIL (amino 1–100) were chosen. J Residue mapping: GLUD1 (amino acids Y2, Y4, R13, I132, V156) and SNAIL (amino acids P2, F5, R8) mediate complex formation (representative of three independent experiments). K Impaired binding between K147A-GLUD1 and SNAIL in cell nucleus (representative of three independent experiments). L Impaired binding between K147A-GLUD1 and SNAIL in purified protein system despite 100 μM α-KG treatment (representative of three independent experiments). M , N ChIP ( M ) and sequential ChIP ( N ) showing defective promoter recruitment with K147A-GLUD1, all data were normalized to input, n = 6 biologically independent samples. O Quantified intra/extravasation efficiency of wild-type versus K147A-GLUD1 in 3-lane vascular model, n = 3 biologically independent samples. P Relative quantification of migrating cells’ fluorescence intensity in GLUD1 variants using intravasation based zebrafish model, n = 3 embryo cohorts. Q , R In vivo metastasis assay using extravasation based mouse model: ( Q ) representative bioluminescence images; ( R ) weekly photon flux quantification (GFP-luciferase-tagged A549/DTX cells injected via tail vein in Balb/c nude mice), n = 4 mice. All data are represented as mean ± SEM. P values were calculated using two-tailed unpaired Student’s t test ( C , P ); one way ANOVA followed by Dunnett’s multiple-comparisons test ( M , O ); two way ANOVA followed by Šídák multiple-comparisons test ( N ); two way ANOVA ( R ). Source data are provided as a Source Data file. For immunoblots, samples within each panel were derived from the same experiment, with separate gels processed in parallel where applicable; detailed gel assignments are provided in the Methods.

Article Snippet: The antibodies used were shown as followed: VE-Cadherin (ab33168; 1:1000, Abcam), Lamin B (ab16048; 1:5000, Abcam), Sodium/Potassium ATPase alpha 1 (ab308329; 1:1000, Abcam), FLAG (ab205606; 1:10,000, Abcam), ACTIN (ab8226, 1:2000, Abcam), GLUD1 (14299-1-AP; 1:1000, Proteintech), FAK (12636-1-AP; 1:1000, Proteintech), EPHA2 (25749-1-AP; 1:1000, Proteintech), SNAIL (13099-1-AP; 1:1000, Proteintech), β-Tubulin (80713-1-RR; 1:1000, Proteintech), Phospho-FAK (83933-1-RR, 1:1000, Proteintech), Phospho-EPHA2 (25749-1-AP, 1:1000, Proteintech), SOX2 (11064-1-AP, 1:500, Proteintech), SLUG (12129-1-AP, 1:500, Proteintech), ZEB1 (21544-1-AP, 1:500, Proteintech).

Techniques: Binding Assay, Mutagenesis, Concentration Assay, Activity Assay, Circular Dichroism, Residue, Purification, Quantitative Proteomics, Fluorescence, In Vivo, Luciferase, Injection, Two Tailed Test, Western Blot, Derivative Assay

Journal: Nature Communications

Article Title: α-KG conformationally stabilizes nuclear GLUD1-SNAIL complex to fuel trans-endothelial migration in metastasis

doi: 10.1038/s41467-026-77149-0

Figure Lengend Snippet: A MST analysis of diethyl glutarate (DEG) binding to GFP-tagged wild-type (WT) or K147A mutant GLUD1 in A549 cell lysates, the concentration of DEG ranges from 15 mM to 0.00046 mM, n = 3 biologically independent samples. B Streptavidin pull-down assay showing competition between biotinylated α-KG and glutarate (100 μM each) for GST-GLUD1 binding (representative of three independent experiments). C , D CDH5 promoter pull-down analysis of GLUD1-SNAIL binding with α-KG or glutarate (100 μM each) (scheme, ( C ); quantification, ( D ); n = 6 biologically independent samples). E , F 3-lane vascular model images ( E ) and quantification ( F ) of TEM in A549/DTX cells expressing WT or mutant GLUD1 with DEG (10 μM) treatment, n = 3 biologically independent samples, normalized to pLV- GLUD1 . G Intravasation based zebrafish metastasis model showing fluorescence intensity of migrating A549/DTX cells microinjected with 1 nL 7 M DEG, n = 4 embryo cohorts. H Bioluminescence tracking of metastatic burden in nude mice intravenously injected with GFP-luciferase-tagged A549/DTX cells and treated with DEG (10 mg/kg, i.p., every other day), n = 6 mice. I – K H&E staining ( I ), metastatic lesion number ( J ) and metastatic area fraction ( K ) in lungs from ( H ) ( n = 6 mice). Metastatic area fraction was calculated as metastatic area divided by total lung area. L Bioluminescence imaging of nude mice implanted in situ with metastatic A549/DTX cells and treated with saline or DEG. Saline-treated mice developed brain metastases with hindlimb paralysis, whereas DEG-treated mice showed normal behavior. M Proportion of mice developing brain metastases ( n = 10 mice). N Survival and O body weight of the mice in ( L ) ( n = 10 mice). Data are represented as mean ± SEM. P values were calculated using one way ANOVA followed by Tukey’s multiple-comparisons test ( D , F , G , J , K ); two way ANOVA ( H , O ); two-sided Fisher’s exact test ( M ); two-sided log-rank (Mantel-Cox) test ( N ). Scale bars: I , 500 μm. Source data are provided as a Source Data file. For immunoblots, samples within each panel were derived from the same experiment, with separate gels processed in parallel where applicable; detailed gel assignments are provided in the Methods.

Article Snippet: The antibodies used were shown as followed: VE-Cadherin (ab33168; 1:1000, Abcam), Lamin B (ab16048; 1:5000, Abcam), Sodium/Potassium ATPase alpha 1 (ab308329; 1:1000, Abcam), FLAG (ab205606; 1:10,000, Abcam), ACTIN (ab8226, 1:2000, Abcam), GLUD1 (14299-1-AP; 1:1000, Proteintech), FAK (12636-1-AP; 1:1000, Proteintech), EPHA2 (25749-1-AP; 1:1000, Proteintech), SNAIL (13099-1-AP; 1:1000, Proteintech), β-Tubulin (80713-1-RR; 1:1000, Proteintech), Phospho-FAK (83933-1-RR, 1:1000, Proteintech), Phospho-EPHA2 (25749-1-AP, 1:1000, Proteintech), SOX2 (11064-1-AP, 1:500, Proteintech), SLUG (12129-1-AP, 1:500, Proteintech), ZEB1 (21544-1-AP, 1:500, Proteintech).

Techniques: Analysis, Binding Assay, Mutagenesis, Concentration Assay, Pull Down Assay, Expressing, Fluorescence, Injection, Luciferase, Staining, Imaging, In Situ, Saline, Western Blot, Derivative Assay

Journal: Nature Communications

Article Title: α-KG conformationally stabilizes nuclear GLUD1-SNAIL complex to fuel trans-endothelial migration in metastasis

doi: 10.1038/s41467-026-77149-0

Figure Lengend Snippet: In the cytoplasm, GLUD1 performs its canonical metabolic function in glutamate catabolism by generating α-KG. In addition to this enzymatic role, a fraction of GLUD1 undergoes Importin-8-dependent nuclear translocation and exerts a moonlighting function. Within the nucleus, GLUD1 associates with the transcription factor SNAIL. Nuclear α-KG binds to the non-catalytic Lys147 site of GLUD1 and acts as a conformational stabilizer, thereby promoting and stabilizing the interaction between GLUD1 and SNAIL. This metabolite-dependent interaction supports the assembly of a nuclear α-KG/GLUD1/SNAIL ternary complex. The resulting complex is recruited to the VE-cadherin promoter and enhances VE-cadherin transcription in tumour cells. Increased VE-cadherin expression strengthens trans-endothelial migration and facilitates tumor cell intravasation into and extravasation from the vasculature, two critical steps in hematogenous metastatic dissemination. The lower panel summarizes the biological consequence of this pathway, in which tumour cells with relatively high GLUD1 expression more readily interact with the vasculature, undergo trans-endothelial migration and contribute to the formation of distant metastases, whereas tumour cells with relatively low GLUD1 expression show a reduced capacity for these processes. Orange circles represent α-KG, yellow ovals represent GLUD1, pink ovals represent SNAIL and the purple structure represents Importin-8. The blue and grey tumour-cell symbols indicate relatively high and low GLUD1 expression, respectively. Dashed arrows indicate the proposed directional steps involved in α-KG production, GLUD1 nuclear translocation, ternary-complex assembly and transcriptional regulation. The schematic summarizes the mechanism supported by the experimental findings and is not drawn to scale.

Article Snippet: The antibodies used were shown as followed: VE-Cadherin (ab33168; 1:1000, Abcam), Lamin B (ab16048; 1:5000, Abcam), Sodium/Potassium ATPase alpha 1 (ab308329; 1:1000, Abcam), FLAG (ab205606; 1:10,000, Abcam), ACTIN (ab8226, 1:2000, Abcam), GLUD1 (14299-1-AP; 1:1000, Proteintech), FAK (12636-1-AP; 1:1000, Proteintech), EPHA2 (25749-1-AP; 1:1000, Proteintech), SNAIL (13099-1-AP; 1:1000, Proteintech), β-Tubulin (80713-1-RR; 1:1000, Proteintech), Phospho-FAK (83933-1-RR, 1:1000, Proteintech), Phospho-EPHA2 (25749-1-AP, 1:1000, Proteintech), SOX2 (11064-1-AP, 1:500, Proteintech), SLUG (12129-1-AP, 1:500, Proteintech), ZEB1 (21544-1-AP, 1:500, Proteintech).

Techniques: Translocation Assay, Expressing, Migration