Fig 2: Hepatic Gls2 but not Glud1 knockdown reduces hyperammonemia and mortality in a urea cycle deficiency animal model of OTC mice on high protein diet. Survival curves (A) in OTC and wildtype (WT) mice administered IV injection of AAV shRNA-Gls2 and shRNA-Glud1 at week 0 and placed on high protein diet at week 2. Plasma ammonia levels (B), body weights (C), spot blood glucose at week 2 (D), terminal plasma glutamine/glutamate levels (E, F), terminal plasma amino acid and orotic acid concentration heatmap (G) in the treatment groups. N = 8-13 per group. Data are mean ± SEM. Statistical analysis was conducted by two-way ANOVA. Comparisons were made between OTC sh-Gls2, OTC sh-Glud1 and OTC control (B, C), and between treatments within the same genotype (D-F). *P < .05, **P < .01, ***P < .001, ****P < .0001. Blue color indicates comparison to OTC control group and gold color indicates comparison to OTC sh-Glud1 group. (G) Heatmaps showing (left) standardized plasma concentration of each amino acid in individual mouse, grouped by genotype and treatment categories, and (right) signed significance of categorical comparisons of each amino acid. Significance is log10 transformed P-value from one-way ANOVA, and sign denotes direction of change. Only significances of P < .05 are colored

Fig 3: Gdh localization and influence of Gdh on Tet enzymes, gene expression and metabolism during neuronal stimulation.a Comparison of hmdC and fdC levels in HEK293T (n = 5 biologically independent samples) expressing GFP-Tet3 + Gdh-FLAG or GFP-Tet3 + Gdh-FLAG-NES for 24 h. A two-sided paired t-test was performed. b ICC in HEK239T showing FLAG and GFP signal after ectopic co-expression of GFP-Tet3 + Gdh-FLAG or GFP + Gdh-FLAG from different promoters. Scale bar is 50 µm. c ICC in HEK293T showing FLAG, Gdh, and CoxIV after expressing FLAG3-Tet3 and Gdh from a bicistronic vector compared to untransfected cells. Merged images: FLAG (magenta), Gdh (cyan), CoxIV (red) and Hoechst (yellow). Scale bar is 5 µm. d Hippocampal hmdC levels upon neuronal stimulation with 25 mM KCl for 6 h with and without 20 µM Gdh inhibitor R162 (n = 7 biologically independent animals) or 40 µM PDC inhibitor UK5099 (n = 6 biologically independent animals). Connected dots represent matched hmdC values from one individual mouse. RM one-way ANOVA combined with Tukey’s multiple comparisons test was performed. e aKG level and aKG/succinate ratio in hippocampus (n = 5 biologically independent animals) after depolarization with and without R162. Ordinary one-way ANOVA combined with Tukey’s multiple comparisons test was performed. A.U. arbitrary units. f Relative gene expression levels of Npas4 and Bdnf (n = 7 biologically independent animals) in hippocampus after depolarization with and without R162 normalized to expression levels in unstimulated control. An unpaired two-sided t-test with Welch’s correction was performed. g PLA of Tet3 + NPC and Gdh + NPC in the hippocampus. Images show Z-stacks. Scale bar is 5 µm. a, d–f p (a, f) or p(adj) (d, e) <0.05 (*), <0.01 (**), <0.001 (***); exact p values and statistical details are provided in the Statistics and Reproducibility sub-section within the “Methods” section. e, f Boxes display interquartile range, median is shown as center line, mean as + , whiskers represent the minimum and the maximum. Source data are provided as a Source Data file.

Fig 4: Interaction of Tet3 with Gdh in neurons.a Experimental set-up for Tet3-enriched coIP. b Volcano plot after Tet3-enriched coIP in nuclear brain lysate of adult mice (n = 4 biologically independent sample preparations). Interaction partners were analyzed using label-free quantification after LC–MS/MS. Enriched interaction partners (FDR < 0.05, log2 FC > 1) are highlighted in black. c Conversion of glutamate to aKG by Gdh and glutamate to aKG ratio determined by NMR-based metabolomics in murine hippocampus (n = 6 biologically independent animals, bar shows mean, error bars SD). d Western blots detecting Tet3, Ogt, and Gdh after endogenous Tet3-coIP in nuclear brain lysate. e PLA signal (white/orange dots) of Tet3 and Gdh in the dentate gyrus (DG) of the murine hippocampus. Nuclei (blue) were stained with Hoechst and NeuN staining indicates neuronal nuclei. Scale bar is 5 µm. Source data are provided as a Source Data file.

Fig 5: Hepatic Gls2 and Glud1 knockdown in mouse liver to target urea cycle disorders. Schematic of glutamine metabolism in the liver and mechanism of action of targeting GLS2 and GLUD1 to treat urea cycle disorders. For simplicity the conversion of glutamine to KG via the glutaminase II reaction 7 is not shown (A). Timeline of the study (B). Animals were injected AAV at week 0, were on high protein diet treatment at week 2 and were euthanized at week 5. Hepatic Gls2 and Glud1 knockdown at RNA and protein levels were confirmed by liver RNA-seq and western blot (C, D, F, and G). Deficient Otc expression in OTC mice were confirmed by liver RNA-seq and western blot (E, F, and G). N = 8-13 per group. Data are mean ± SEM. Statistical analysis was conducted by two-way ANOVA. Comparisons were made between sh-Gls2, sh-Glud1, and control within the same genotype (black asterisks), and comparisons were made between OTC and wildtype mice among the same treatment (red asterisks). OTC protein quantification was generated from a different set of blots with a loading control for normalization between blots. *P < .05, **P < .01, ***P < .001, ****P < .0001