Fig 2: The TS pathway does not contribute to GSH synthesis in gliomas. (A) 1H NMR spectra of media over time, including assignments for methyl-13C-methionine. Representative spectra shown as normalized intensities to the trimethylsilylpropanoic acid (TSP) signal (d = 0.00 p.p.m.) and cell number (experiment was done in triplicate for each cell line). (B) Methyl-13C-methionine levels in cells seeded in medium containing this tracer (data displayed as bar plots; mean ± SD, n = 3; ***P < 0.001, two-tailed Student’s t-test). (C) Representative western blot of CBS and CTH (quantitative data for n = 3 is displayed in Fig. S2A). (D) 13C-tracing experiments displaying [C3-13C]-cysteine and [C3,3'-13C]-cystine incorporation into cystathionine, glutathione (GSH), and hypotaurine (n = 3 samples per cell line and 13C probe, data displayed as bar plots; mean ± SD). These experiments were conducted separately for each tracer (13C-cysteine and 13C-cystine). (E) Percentage of m+1 isotopologue over the total pool of GSH from [C3-13C]-serine (data displayed as bar plots; mean ± SD, n = 3; *P < 0.5; ***P < 0.001, two-tailed Student’s t-test). (F) Decrease of GSH levels (data displayed as bar plots of Log2(Fold Change, FC), mean ± SD, n = 5; ***P < 0.001, two-tailed Student’s t-test) for both PAG treatment and cysteine/cystine deprivation displayed.

Fig 3: Thyroid Hormone Signaling through TRß Represses Hepatic H2S Production In Vivo(A–C) Analysis of TH-responsive and sulfur amino acid metabolism-associated mRNA levels (n = 4–6) (A), protein expression (n = 6) (B), and H2S production capacity (n = 5) (C) in livers of mice treated with GC-1 versus vehicle (saline) control. The asterisk indicates the significance of the difference between vehicle (saline) control (euthyroid) and +GC-1 groups (hyperthyroid); *p < 0.05.(D and E) Liver CBS and CGL protein expression (D) and H2S production capacity (E) in mice with indicated TRß status (WT, homozygous WT; TRßPV/+, Het; TRßPV/PV, homozygous mutant; n = 4–5/group). The asterisk indicates the significance of the difference between WT and TRßPV/PV, and the # sign indicates the significance of the difference between TRßPV/+ and TRßPV/PV; */#p < 0.05.(F) Fold enrichment of TRß binding to genetic regulator elements in sulfur amino acid metabolism and H2S producing genes in the livers of mice infected with Ad-GFP (control) or Ad-TRß while on PTU diets +/-T3 injection as indicated (n = 5/group). The asterisk indicates the significance of the difference between the Ad-GFP PTU and Ad-TRß PTU or Ad-TRß T3 groups, and the # sign indicates the significance of the difference between the Ad-TRß PTU and Ad-TRß T3 groups; */#p < 0.05.(G and H) Liver ATF4 protein expression in mice due to PTU/T3 administration (G) (n = 4/group) or TRß mutations (H) (n = 4–5/group). The asterisks indicate the significance of the difference between Hyper-T and Eu-T or Hypo-T (G), or WT and TRßPV/+ or TRßPV/PV (H), and the # sign indicates the significance of the difference between TRßPV/+ and TRßPV/PV (H), */#p < 0.05. Error bars are ± SEM. See also Figure S5.

Fig 4: Cysteine and cystine deprivation halt growth and reduce viability in glioma through reduced protein translation and glutathione synthesis. (A) Growth rate of IDH1-mutant glioma cell lines under cysteine/cystine (Cys/Cyss) deprivation conditions over 96 h. Data are shown as the ratio of viable cells to cells seeded (2-way ANOVA followed by Sidak’s multiple comparison test, *P < 0.05; **P < 0.005; ***P < 0.001; n = 4, data displayed as mean ± SD for each time point and condition). (B) Neurosphere formation after 96 h of incubation in the same medium. BT142 cells displayed as a representative picture (experiment was done in triplicate for each cell line); scale bar is 100 µm. (C) Viability after 96 h in medium lacking cysteine/cystine as the percentage over the viability of cells in control conditions (n = 3, data displayed as mean ± SD for each time point and condition). (D) Metabolic pathways related to cysteine (THF, tetrahydrofolate; MS, methionine synthase; CBS, cystathionine beta-synthase; CTH, cystathionine gamma-lyase; GCLM, glutamate–cysteine ligase modulatory subunits; GCLC, glutamate–cysteine ligase catalytic subunits). (E) Rescue experiments involving the addition of different metabolites and agents to cells grown without cysteine/cystine. Data displayed as control-normalized growth (top panel) and viability (bottom panel) as the mean ± SD for n = 3; *P < 0.05; **P < 0.005; ***P < 0.001; each condition was compared with the -Cys/Cyss values by using a two-tailed Student’s t-test. (F) Glutathione (GSH) levels in complete medium (+) and medium lacking cysteine/cystine (-) plus 0.1 mm homocysteine and cystathionine (2-way ANOVA followed by Dunnett’s multiple comparison test for -cysteine/cystine vs all groups, **P < 0.005; ***P < 0.001 was used and data are displayed as the mean ± SD for n = 3). (G) Representative western blot of p-GCN2 and p-eIF2a in both complete and cysteine/cystine-free media for the three cell lines investigated (quantitative data for n = 3 is displayed in Fig. S2C,D). (H) Intracellular ROS levels assessed by the DCFDA assay in glioma cell lines under cysteine/cystine deprivation and treated with 0.25 mm trolox for 96 h. BT142 cell line data are shown as a representative diagram and data from other cell lines are shown in the Fig. S1E.

Fig 5: TFEB stimulation of MAT1A transcription.a Male 10 weeks old C57BL/6 J mice were intravenously injected with Ad-Null or Ad-TFEB at 5 × 108 pfu/mouse. One week later, mice were fed chow (C) or Western diet (WD) for one additional week. Relative liver mRNA are shown with control arbitrarily set as 1. (n = 5). MAT1A Methionine adenosyltransferase 1A, CBS cystathionine ß-synthase, CSE cystathionine ? lyase. b, c Male 10 weeks old C57BL/6 J mice were intravenously injected with Ad-shCon (Ad-scramble) or Ad-shTFEB at 1 X 109 pfu/mouse. Mice were fed chow for 2 weeks and euthanized after 6 h fast (n = 8). Liver protein (b) and mRNA (c) were measured. Each band represents an individual mouse sample in (b). d Luciferase reporter constructs containing 1.9 kb or 0.9 kb human MAT1A promoter and TFEB expression plasmids were transfected in triplicates in AML12 cells for 48 h. Luciferase activity was normalized to ß-galactosidase activity. A representative of 2 independent experiments. e, f Chromatin immunoprecipitation assay with human liver nuclear fraction (e) or male 10 weeks old C57BL/6 J mouse liver nuclear fraction (f). NC negative control with IgG, IP immunoprecipitation with anti-TFEB antibody. Primer pair locations are relative to translational start site as +1. Results in a and c are expressed as mean ± SEM. Results in (d, e, and f) are expressed as mean ± SD of technical triplicate repeats. Two-way ANOVA and Tukey post hoc test are used for a and unpaired 2-tailed Student’s t-test is used for (c, d, e, f). A p value < 0.05 is considered statistically significant. Source data for (a–f) are provided as a Source Data file.