Tobacco Smoking Rewires Cell Metabolism by Inducing GAPDH Succinylation to Promote Lung Cancer Progression

Abstract

Patient behavior and physiology can directly affect cancer metabolism. Smoking is the leading risk factor for non-small cell lung cancer (NSCLC). In this study, we identified that smoking modulates lung cancer cell metabolism through altered protein post-translational modification. Proteomic analyses identified elevated K251 succinylation (K251-Su) of GAPDH, a key enzyme in glycolysis, in NSCLC samples, and GAPDH K251-Su was significantly higher in patients who smoke compared with nonsmokers. Exposure of lung cancer cells to cigarette smoke extract led to increased uptake of glutamine and enhanced GAPDH K251-Su. Glutamine uptake by cancer cells in hypoxic and nutrient-deficient microenvironments provided succinyl-CoA donors for GAPDH succinylation at K251, which was catalyzed by acyltransferase p300. K251-Su increased GAPDH stability by suppressing TRIM4-mediated K254 ubiquitination. GAPDH K251-Su enhanced glycolysis and glutamine reductive carboxylation to meet the demands for cell growth and to support survival in hypoxic and nutrient-depleted conditions, promoting tumor growth and metastasis. These findings indicate that tobacco smoking mediates metabolic reprogramming of cancer cells through succinylation of GAPDH to drive NSCLC progression.

Significance: Smoking-induced GAPDH succinylation coordinates glycolysis and glutamine metabolism and supports lung cancer cell survival in stressful microenvironments to promote tumor progression, highlighting quitting smoking as a potential strategy to target cancer metabolism.

Figure 1.

Identification of elevated GAPDH K251-Su in the tumor tissues of smoker patients with NSCLC. A, The workflow of identifying GAPDH K251-Su and its clinical significance in NSCLC. B, MS analysis of a tryptic fragment of GAPDH extracts from NSCLC tumor tissue proteomics at monoisotopic m/z 710.37714 Da (+5.21 mmu/+7.33 ppm) matched with the doubly charged peptide LEKPAKYDDIK with K3-succinyl (100.01604 Da), suggesting that GAPDH-K251 was succinylated. The vertical axis indicates the intensity of the ion peaks and the horizontal axis indicates the ratio of mass to charge. C, GAPDH protein was immunoprecipitated from A549 and NCI-H2170 cells using an anti-GAPDH antibody, and the level of succinylation of GAPDH was detected using lysine pan-succinylation modification antibody (Pan-Ksu). D, After site mutation of potential lysine succinylation modification sites of GAPDH, the level of succinylation was detected using the Pan-Ksu antibody in A549 and NCI-H2170 cells. Relative ratios of GAPDH Pan-Ksu were calculated from normalizing against Flag-GAPDH. E, The specificity of the prepared GAPDH K251-Su antibody was determined by immunoblotting after different doses of succinylated peptide and nonmodified peptide were fixed on the solid-phase membrane. F, The specificity and binding efficiency of the GAPDH K251-Su antibody were detected by Western blotting with Hela cells. G, The GAPDH K251-Su antibody was premixed with succinylated peptide-Su-2 and nonmodified peptide-NC to detect the GAPDH K251-Su modification in A549 and NCI-H2170 cells. H, WT and K251R-mutant GAPDH were expressed in A549 and NCI-H2170 cells, and the level of succinylation was detected using the GAPDH K251-Su antibody. I, Representative figures of IHC staining by GAPDH K251-Su antibody in the tumor tissues and adjacent normal tissues of NSCLC. Scale bar, 100 μm. J, Statistical analysis of the expression levels of GAPDH K251-Su in normal tissues and tumor tissues from different AJCC stages of NSCLC. K, Correlation between GAPDH K251-Su levels and overall survival of patients with NSCLC. According to the IHC staining score, tumor tissues were classified into low and high GAPDH K251-Su expression groups. L, Representative figures and comparison of the expression level of GAPDH K251-Su in the tumor tissues of smoker and nonsmoker patients with NSCLC by IHC staining. Scale bar, 50 μm. M, Representative figures and statistical analysis of expression level of GAPDH K251-Su in the tumor tissues that present mild, moderate, and severe severity anthracosis by H&E staining, and the area circled with yellow lines representing the anthracosis area. Scale bar, 100 μm. Data are presented as mean ± SD. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, nonsignificant. Statistical significance was determined by an unpaired t test (J, L, and M) and log-rank (Mantel–Cox) test (K). See also from Supplementary Figs. S1–S4. LUAD, lung adenocarcinoma; LUSC, lung squamous cell carcinoma.

Figure 2.

Smoke-induced increased cell uptake of glutamine and enhanced GAPDH K251-Su level. A, The levels of GAPDH K251-Su were detected after treating A549 cells with different doses of CSE dissolved in a cell culture medium for 24 hours. Hypoxia and nutrient deficiency (hypoxia + Nu-Def): cells were cultured in 1% FBS and 0.5 g/L glucose medium and an incubator at 2% O₂. Normoxia and nutrient sufficiency (normoxia + Nu-Suff): cells were cultured in 10% FBS and 4.5 g/L glucose medium and an incubator at 20% O₂. B, Bioinformatics analysis from TCGA and cBioPortal database to compare the gene (violin diagram) and protein (box diagram) expression of glutamine transporters (SLC1A5, SLC7A5, and SLC38A1) in the tumor tissues from smoker and nonsmoker patients with NSCLC. FPKM, fragments per kilobase of transcript per million mapped reads. C, Bioinformatics analysis from the Metabolights database to compare the glutamine abundance from lung tissue between mice treated with cigarette smoke or filtered air for 3 and 6 months (n = 6). D, The levels of cell uptake of glutamine were measured after treating A549 cells with 0% and 15% doses of CSE dissolved in cell culture medium for 24 hours. E, The levels of GAPDH K251-Su were detected after one of the amino acids—glutamine (Gln), glutamate (Glu), methionine (Met), proline (Pro), valine (Val), leucine (Leu), or isoleucine (Isoleu)—was removed from the cell culture medium. Cells were cultured in hypoxic and Nu-Def conditions. The control group indicates that the culture medium contains all the amino acid components. Relative ratios of GAPDH K251-Su were calculated from normalizing against tubulin. F and G, The levels of GAPDH K251-Su were detected after different doses of glutamine (F) or glutamate (G) were added to the cell culture medium for 24 hours. H and I, The concentration of succinyl-CoA was analyzed after cells were treated with 0.5 and 6 mmol/L of glutamine (H) or with 10 mg/L and 80 mg/L of glutamate (I) for 24 hours. J, The schematic diagram of smoke-induced increased cell uptake of glutamine enhances GAPDH K251-Su levels. Smoke leads to elevated expression of glutamine transporter on cancer cells, which makes cells prefer to uptake glutamine than other amino acids (e.g., methionine, proline, valine, leucine, and isoleucine) in a hypoxic Nu-Def microenvironment. Catabolism of glutamine generates glutamate and α-ketoglutarate (αKG) to participate in the TCA cycle and produce succinyl-CoA, which is a donor for succinylation and contributes to elevated levels of GAPDH K251-Su in NSCLC cells. Data are presented as mean ± SD. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, nonsignificant. Statistical significance was determined by an unpaired t test. See also Supplementary Fig. S5.

Figure 3.

GAPDH K251-Su increases stable expression of GAPDH by suppressing TRIM4-mediated K254 ubiquitination. A, Western blot assay of flag-GAPDH in A549 and NCI-H2170 cells with the flag fusion–expressed WT GAPDH, K251R-mutant, and K251E-mutant GAPDH proteins, using an anti-Flag antibody. B, qPCR assay of the GAPDH mRNA levels in A549 and NCI-H2170 cells expressing WT GAPDH, K251R-mutant, and K251E-mutant GAPDH. C, IP and Western blot assay of the ubiquitination level of GAPDH in A549 and NCI-H2170 cells expressing WT GAPDH, K251R-mutant, and K251E-mutant GAPDH proteins, using a pan-ubiquitination antibody (Pan-Ub). D and E, Western blot assay of GAPDH in WT and K251-mutant A549 (D) and NCI-H2170 (E) cells treated with the proteasome inhibitor MG132 for 0, 6, and 12 hours. F and G, Western blot assay of GAPDH in WT and K251-mutant A549 (F) and NCI-H2170 (G) cells treated with the protein synthesis inhibitor cycloheximide (CHX) for 0, 8, and 16 hours. H, MS analysis of a tryptic fragment of GAPDH extracts from A549 cells at monoisotopic m/z 521.29059 Da (−0.41 mmu/−0.78 ppm) matched with the triply charged peptide LEKPAKYDDIKK with K6-GG (114.04293 Da), suggesting that GAPDH-K254 was ubiquitinated. The vertical axis shows the intensity of the ion peak and the horizontal axis shows the ratio of mass to charge (m/z). I, The protein spatial structural simulation software Discovery Studio presents the spatial distribution of K251 (red) and K254 (black) in the GAPDH monomer (left) and GAPDH tetramer (right). J,  Immunoprecipitation and Western blot assay of the ubiquitination level of GAPDH in A549 and NCI-H2170 cells expressing WT GAPDH and K254R-mutant GAPDH using pan-ubiquitination antibody (Pan-Ub). K, Co-IP and Western blot assay to verify the interaction of TRIM4, TRIM21, and TRIM23 with GAPDH in A549 and NCI-H2170 cells, using an anti-GAPDH antibody for co-IP. WCL, whole-cell lysate. L, Immunoprecipitation and Western blot assay to detect the ubiquitination level of GAPDH in A549 cells expressing WT GAPDH, K251R-mutant, K251E-mutant, and K251R/K254R double-mutant GAPDH proteins, using a pan-Ub antibody (top). The interaction of TRIM4 with GAPDH in A549 cells expressing WT GAPDH, K251R-mutant, K251E-mutant, and K251R/K254R double-mutant GAPDH proteins is shown at the bottom. M and N, Enzyme catalytic activity of GAPDH in WT and K251-mutant cells was assayed using a GAPDH activity assay kit in A549 and NCI-H2170 cells. OD values indicate the amount of NADH catalyzed by GAPDH (M) and statistical analysis of relative GAPDH activity based on OD values was demonstrated (N). O, The schematic diagram of GAPDH K251-Su increasing the stability of GAPDH protein by suppressing TRIM4-mediated K254 ubiquitination. In NSCLC cells with a low level of GAPDH K251-Su, TRIM4 mediates the ubiquitination of GAPDH K254 and promotes the degradation of GAPDH through the proteasome pathway. In NSCLC cells with a high level of GAPDH K251 succinylation, GAPDH K251 succinylation suppresses TRIM4-mediated ubiquitination at K254 and prevents GAPDH degradation through the proteasome pathway, leading to enhanced protein stability of GAPDH. Data are presented as mean ± SD (n = 3). *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, nonsignificant. Statistical significance was determined by an unpaired t test. See also Supplementary Figs. S6 and S7.

Figure 4.

Acyltransferase p300 catalyzes GAPDH K251-Su. A, Expression of GAPDH K251-Su was detected after siRNA on the succinyltransferases GCN5, CPT1A, DLST, and p300/CBP. Designed three siRNAs for each gene were siRNA 1#, 2#, and 3#. The controls without added siRNA are labeled as “–." B, Confocal assay to detect the colocalization of GAPDH (green) and p300 (red) in A549 and NCI-H2170 cells. DAPI (blue), nucleus. Scale bar, 20 μm. C, Co-IP and immunoblotting assay to test the interaction of succinyltransferases GCN5, CPT1A, DLST, and p300/CBP with GAPDH in A549 cells. D, Co-IP and immunoblotting were used to analyze the interaction of expressed His-p300 and Flag-GAPDH. E and F, The levels of GAPDH K251-Su (E) and GAPDH enzyme activity (F) were detected after overexpression of p300 in GAPDH WT, K251R-mutant, and K251E-mutant A549 cells, respectively. The OD value indicates the amount of NADH catalyzed by GAPDH (n = 3). G, Cells were treated with 2 μmol/L acyltransferase inhibitor A485 and then added with 0.5 and 6 mmol/L glutamine for 24 hours. The effect of A485 on the expression of GAPDH K251-Su was detected. Hypoxic nutrient deficiency: cells were cultured in 1% FBS and 0.5 g/L glucose medium and an incubator at 2% O₂. H, Bioinformatics analysis from TCGA and cBioPortal database to compare the gene (violin diagram) and protein (box diagram) expression of p300 in the tumor tissues from smoker and nonsmoker patients with NSCLC. The comparison of p300 gene expression also includes the data of adjacent normal tissues. Data are presented as mean ± SD. *, P < 0.05; ****, P < 0.0001; ns, nonsignificant. Statistical significance was determined by an unpaired t test. See also Supplementary Fig. S8. WCL, whole-cell lysate.

Figure 5.

GAPDH K251-Su promotes cell glycolysis and enhances glutamine reductive carboxylation. A, Generation of the smoking-related Kras^G12D-driven NSCLC GAPDH WT/K251R mouse and the following absolute quantitative targeted metabolomics analysis. B, Kyoto Encyclopedia of Genes and Genomes enrichment of differential metabolic pathway between WT and K251R groups by targeted metabolomics in the tumor tissues of the Kras^G12D-driven NSCLC mouse. The blue arrows and text in red font represent the metabolic pathways with significant differences and enriched more metabolites. C, Histograms represent the absolute quantitative concentration of metabolites from tumor tissues of GAPDH-WT and GAPDH^K251R mice. The red dashed box indicates the metabolites that rank in the highest concentration and with the most significant difference among all differential metabolites. D, Seahorse glycolysis stress test to analyze glycolytic capacity in GAPDH WT and K251R/E-mutant cells. E, Seahorse glycolysis stress test to analyze the glycolytic capacity in A549 and NCI-H2170 cells by adding low (0.5 mmol/L) and high (6 mmol/L) doses of glutamine in hypoxic Nu-Def or Nu-Suff culture environments for 24 hours. F, Seahorse real-time ATP rate test to analyze the ATP production rate and percentage of ATP from mitochondrial oxidative phosphorylation and glycolysis pathway in GAPDH WT and K251R-mutant cells by adding 6 mmol/L glutamine in hypoxic Nu-Def or Nu-Suff condition for 24 hours. G, Seahorse mitochondria stress test to analyze the mitochondrial oxidative phosphorylation capacity in GAPDH WT and K251R-mutant cells by adding 6 mmol/L glutamine in hypoxic Nu-Def or Nu-Suff condition for 24 hours. H, Seahorse glutamine substrate oxidation stress test to analyze the glutamine oxidative phosphorylation capacity in GAPDH WT and K251R-mutant cells treated with BPTES (glutaminase 1 inhibitor) before adding 6 mmol/L glutamine for 24 hours. I, Metabolic tracing analysis of (U¹³-C₅)-labeled glutamine in GAPDH WT and K251R-mutant cells under hypoxic Nu-Def conditions. Glutamine metabolic flow data were classified into the glutamine reductive carboxylation (red arrow) and oxidative phosphorylation (blue arrow) pathways. J, Schematic diagram shows the balance of glutamine oxidative phosphorylation and reductive carboxylation capacity when the level of GAPDH K251-Su is high and low. Data are presented as mean ± SD (n = 3). *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, nonsignificant. Statistical significance was determined by an unpaired t test.

Figure 6.

GAPDH K251-Su promotes tumor growth and metastasis in NSCLC. A, Micro-CT scanning and 3D reconstruction image of the lung in Kras^G12D-driven NSCLC GAPDH WT/K251R mouse. The yellow arrows in CT scanning images and the red nodular lesions in 3D reconstructed images indicate tumor lesions. B, Statistical analysis of tumor volume and the number of lung tumor lesions between GAPDH WT and K251R-mutant groups detected by micro-CT in Kras^G12D-driven NSCLC mouse (n = 6). C, PET-CT scanning of the lung tissues in GAPDH WT and K251R mice. The green fluorescence is the isotope absorption signal of ¹⁸F-FLT. D, Photos of lung tissues dissected from GAPDH WT and K251R mouse. Black asterisks, mouse lung tumor lesions. E, H&E staining to compare the metastasis lesions in lung and liver tissues between GAPDH WT and K251R mouse. Black arrows, the tumor lesions. Scale bars, 500 μm (lung lobe and liver). F, IHC staining of eGFP, GAPDH K251-Su, Ki67, and CD31 was performed in the lung tissues of GAPDH WT and K251R mice. Scale bars, 100 μm. G, The graphical summary of this study. Tobacco smoking is related to increased uptake of glutamine by cancer cells under a hypoxic and nutrient deficiency tumor microenvironment. Enhanced glutamine catabolism provides more succinyl-CoA donors for the succinylation of a key glycolytic enzyme, GAPDH. Elevated GAPDH K251-Su readjusts cell glucose and glutamine metabolism to meet the demands for cancer cell growth, resulting in the progression and metastasis of NSCLC. Data are presented as mean ± SD. *, P < 0.05; ***, P < 0.001. Statistical significance was determined by an unpaired t test. See also Supplementary Fig. S8 and Supplementary Videos S1 and S2.