Fig 2: Depletion of LDHB results in decreased tumorigenesis and delayed tumor growth in a genetically engineered mouse model of NSCLC. a Schematic representation of the lung tumor induction experiment in LDHB knockout and wild-type mice. b Bodyweight was determined from LDHB+/+; K-rasLSL-G12D/+; p53fl/fl and LDHB-/-; K-rasLSL-G12D/+; p53fl/fl mice at different timepoints (n = 3). The error bars represent mean ± SD. c Representative axial CT scans from LDHB+/+; K-rasLSL-G12D/+; p53fl/fl and LDHB-/-; K-rasLSL-G12D/+; p53fl/fl mice after intratracheal instillation with AAV-Cre virus at different timepoints. d Tumor volume of LDHB+/+; K-rasLSL-G12D/+; p53fl/fl and LDHB-/-; K-rasLSL-G12D/+; p53fl/fl mice was quantified at 16 weeks of age by 3D Slicer software (n = 4). **P < 0.01 (two-tailed unpaired Student’s t-test). The error bars represent mean ± SD. e 3D view of lungs from LDHB+/+; K-rasLSL-G12D/+; p53fl/fl and LDHB-/-; K-rasLSL-G12D/+; p53fl/fl mice at 16 weeks of age (8 weeks after intratracheal instillation with AAV-Cre). f Lungs from LDHB+/+; K-rasLSL-G12D/+; p53fl/fl and LDHB-/-; K-rasLSL-G12D/+; p53fl/fl mice were harvested and weighed after 8 weeks of intratracheal instillation with AAV-Cre virus (n = 3). A two-tailed unpaired Student’s t-test was used for statistical analysis. The error bars represent mean ± SD. g Representative HE images are from the mice sacrificed above

Fig 3: OXPHOS deficiency induced by LDHB inhibition is related to persistent mtDNA damage. a Venn diagram showing the number of genes down-regulated in A549 and H358 after LDHB silencing. Comprehensive pathway analysis showing top 25 down-regulated pathways or gene ontology terms in A549 and H358 after LDHB silencing. b Relative RNA expression-based enrichment of Mitochondrial Respiratory Chain Complexes in response to mitochondrial complex activity in siCTRL compared to siLDHB. The y-axis represents the enrichment value, and the x-axis represents the rank of differential expression for all genes, with group-specific genes indicated by vertical black lines. Rank positions to the left indicate increased expression in siCTRL. c mtDNA copy number were determined by qRT-PCR analysis on isolated DNA from siCTRL and siLDHB after 48 h of transfection or shCTRL and shLDHB cells (n = 3–8). *P < 0.05, **P < 0.01, ****P < 0.0001 (two-tailed paired Student’s t-test). The error bars represent mean ± SD. d Mitochondrial mass was analysis by flow cytometer using MitoTracker™ Deep Red FM after 48 h of transfection. Mean fluorescence intensity (MFI) was used for quantitative analysis (n = 3). ns no significant difference (two-tailed unpaired Student’s t-test). Error bars represent mean ± SD. e mtDNA damage were determined by qRT-PCR analysis on isolated DNA from siCTRL and siLDHB after 48 h of transfection or shCTRL and shLDHB cells (n = 3–8). **P < 0.01 (two-tailed paired Student’s t-test). The error bars represent mean ± SD. f Maximal ADP-stimulated respiration was quantified by OROBOROS to evaluate mitochondrial respiration complex I (CI), II (CII), IV (CIV) activity in A549 siCTRL and A549 siLDHB after 48 h of transfection or A549 shCTRL and shLDHB cells. The respiration rate is presented in pmol/(sec × million cells) (n = 3). *P < 0.05, **P < 0.01 (two-tailed paired Student’s t-test). The error bars represent mean ± SD

Fig 4: LDHB mediated lactate utilization is critical to OXPHOS. a Immunoblot analysis of A549 cells transfected with control siRNA (siCTRL) or LDHB-specific siRNA (siLDHB) (10 nM) using Lipofectamine 2000 after 48 h. ß-actin was used as the loading control. b LDHB activity levels were determined in A549 cells 48 h after transfection with siCTRL or siLDHB using an enzymatic colorimetric kit. The MCF7 cell line, expressing LDHA but not LDHB protein, was used as a negative control. Mean OD values were normalized to the number of cell lysates and reaction time. ****P < 0.0001 (Ordinary one-way ANOVA). The LDHB protein expression of MCF7 was analysis by western blot. c siCTRL and siLDHB cells were plated at 3000 cells/well in 96-well plates for overnight culture. Cells then were changed into DMEM with 2.5 mM glucose, no pyruvate and no glutamine medium (low-GLC) containing 20 mM L-lactic acid or HCl and adjusted to pH 6.8. After 4 days, cell viability was determined by Acid Phosphatase (APH) Assay. Absorbance was quantified in a Tecan Infinite® M1000 Microplate Reader at 405 nm (n = 10). ****P < 0.0001 (Ordinary two-way ANOVA). The error bars represent mean ± SD. d siCTRL and siLDHB cells were plated at 1 × 105 cells/well in 6-well plates for overnight culture. Cells then were changed into DMEM medium containing 2.5 mM glucose and 20 mM 3-13C sodium L-lactate. After 72 h, supernatants were collected for NMR analysis. The measurement was previously described [36–39]. For each sample, the 13C lactate peak of interest was integrated. The absolute integral was then converted to an mM value using the integral of the initial condition with a known concentration of 20 mM lactate. Lactate consumption values were normalized to the corresponding cell number and then used for statistical analysis (n = 6). *P < 0.05 (two-tailed unpaired Student’s t-test). Data are shown with mean ± SD. e Representative plot showing mean ± SEM of the real-time oxygen consumption rate (OCR) of A549 siCTRL and A549 siLDHB cells using the Seahorse XFe96 Analyzer (n = 39–40 technical replicates with 3–5 readings). The procedure corresponds to the description in Fig. 1. The basal OCR of cell lines were used to perform the statistical analysis. The normalized OCR were plotted as bar graphs (n = 3–4 biological replicates). The error bars represent mean ± SD. ****P < 0.0001 (two-tailed unpaired Student’s t-test). f, g Immunoblot analysis of A549 shCTRL (scramble control) and shLDHB clones. ß-actin was used as a loading control. LDHB activity levels were determined as described above (n = 3). The error bars represent mean ± SD. ns no significant difference, ****P < 0.0001 (Ordinary one-way ANOVA). h Representative plot showing mean ± SEM of the real-time oxygen consumption rate (OCR) of A549 shCTRL and A549 shLDHB cells using the Seahorse XFe96 Analyzer (n = 6 technical replicates with 3–5 readings). The basal OCR were plotted as bar graphs (n = 4 biological replicates). The error bars represent mean ± SD. *P < 0.05, ***P < 0.001, ****P < 0.0001 (Ordinary one-way ANOVA)

Fig 5: LDHB inhibition downregulates mitochondrial-dependent pyrimidine and purine synthesis pathways. a, b Venn diagram showing the number of metabolites reduced in A549 and H358 after LDHB silencing. Balloon plot showing metabolite enrichment analysis (MSEA) of reduced metabolites in A549 and H358 after LDHB silencing. The top 10 dysregulated metabolic pathways are shown. The size and color of the balloon indicate the enrichment ratio and P-value, respectively. Heatmap showing the metabolomic comparison of siCTRL and siLDHB cells. Six replicates (including three biological replicates with two technical replicates) are shown as separate columns for each cell type. log2 of the ratio between the metabolite intensity of each sample to the average intensity of all samples. c, d For proliferation and sphere formation ability, rescue experiments were performed with the addition of the following nucleotide precursors: 100 µM hypoxanthine, 100 µM adenine, 400 µM uridine. The number of cells and spheres was counted after 5 and 7 days with nucleotide precursor rescue (n = 3). ns no significant difference, *P < 0.05, **P < 0.01, ****P < 0.0001 (Ordinary two-way ANOVA). The error bars represent mean ± SD