Fig 2: Pdk2 deficiency inhibits high-fat diet-induced fat accumulation.a Representative images and computed tomography (CT) scans revealed fat accumulation in the epididymal adipose tissue of WT and Pdk2-/- mice fed a LFD or a HFD for 4 weeks. b Body weight was monitored in WT and Pdk2-/- mice fed a LFD or a HFD (n = 5–6 per group). c Body composition (% of body weight) was analyzed in WT and Pdk2-/- mice fed a LFD (n = 5 per group) or a HFD (n = 6 per group) for 3 weeks. d Representative morphology and weight (% of body weight) of subcutaneous and epididymal adipose tissue were evaluated in WT and Pdk2-/- mice fed a LFD or a HFD (n = 5–6 per group). e H&E and perilipin staining of epididymal adipose tissue from WT and Pdk2-/- mice fed a HFD for 4 weeks (n = 6 per group). The adipocyte area was measured with ImageJ. The data are presented as the mean ± SEM values. *P < 0.05, **P < 0.01, ***P < 0.001 by 1-way ANOVA followed by the LSD test.

Fig 3: Combination treatment of 5-FU with DCA or miR-149-3p enhances chemotherapy effects in vivo. a Representative images of tumors in each group. b Xenograft tumor volumes were calculated in the intraperitoneal group. The mean ± SD are shown, n × 6. c Immunostainings of xenograft tumor sections with Ki-67 in the intraperitoneal group. Positive staining is in brown (Scale bars: 100 µm) (upper panel). The immunoreactive score of Ki67 was calculated (lower panel). The mean ± SD are shown, n × 4. d Xenograft tumor volumes were calculated up to 39 days post-tumor inoculation in the intratumoral group. e The expression of miR-149-3p was determined by quantitative real-time PCR in xenograft tumors after DCA intratumoral injection. The mean ± SD are shown, n × 6. f Representative TUNEL stained sections from the mice in the intratumoral group are shown (scale bars: 50 µm). g Representative image of tumors from mice that received an intratumoral injection of agomiR-NC or agomiR-149-3p for 3 weeks. h Xenograft tumor volumes were calculated up to 21 days post treatment. The mean ± SD are shown, n × 6. i The levels of miR-149-3p were analyzed by quantitative real-time PCR. The mean ± SD are shown, n × 5. *P < 0.05. j Apoptosis of tumor cells was determined by the TUNEL assay (Scale bars: 50 µm). k Xenograft tumors from three mice in each group were analyzed by immunoblot with the indicated antibodies. l Images to visualize the positive staining of PDK2 in xenograft tumors (Scale bars: 100 µm) (left panel). The immunoreactive score of PDK2 was calculated (right panel). The mean ± SD are shown, n × 3. m Representative 18F-FDG micro-PET/CT images of tumor-bearing mice. Tumors are indicated by red arrows and circles. The values of SUVmax and MTV were obtained in tumors. The mean ± SD are shown, n × 3. *P < 0.05; ns no significance

Fig 4: PDK1 and PDK2 expression levels are decreased in human visceral adipose tissue and are positively correlated with the PPARγ expression level.a Correlations between the mRNA expression of PDK1, PDK2, PDK3, and PDK4 and body mass index (BMI). b Correlations of PPARγ and SREBP-1c mRNA expression with BMI. c Correlations between the expression of PDK isoforms and that of PPARγ.

Fig 5: Inhibition of HIF1α expression or lactate production decreases adipogenesis in 3T3-L1 cells with stable Pdk1 or Pdk2 overexpression.a Oil Red O staining showed the effect of siHif1a on adipogenesis in 3T3-L1 cells stably overexpressing Pdk1 or Pdk2. b The mRNA expression levels of adipogenic genes were measured on day 6 after induction of differentiation (n = 3). c Extracellular lactate level in 3T3-L1 cells stably overexpressing Pdk1 or Pdk2 and treated with the LDH inhibitor GSK2837808A. d, e Lipid accumulation was evaluated by Oil Red O staining in 3T3-L1 cells stably overexpressing Pdk1 or Pdk2 with or without GSK2837808A treatment. The data are presented as the mean ± SD values. *P < 0.05, **P < 0.01, ***P < 0.001 by 2-tailed Student’s t test.