Fig. 5.

ADT or enzalutamide in hypoxia rewires glucose metabolism from PPP to glycolysis through GPI. a Schematic representation of glucose metabolism that can be directed to PPP by G6PD or to glycolysis by GPI. Androgen/AR is known to have positive effects on PPP, including the increase of glucose uptake, conversion to G6P, and flux to PPP via G6PD^39,40. b–d LNCaP-AdtHs cells were cultured in conditions of R1881 (1 nM), ADT (5% CSS), hypoxia (1% O₂), or ADT + hypoxia (A+H). The glucose metabolic flux was evaluated for glucose uptake and phosphorylation to G6P (b), for pentose phosphate pathway (c), and for glycolysis pathway (cytosolic glycolysis) and the mitochondrial TCA cycle (d). The glucose uptake was determined with a commercial kit (BioVision). The metabolic flux was determined by incubating cells with [U-¹³C]-labelled glucose, and measuring levels of ¹³C-glucose metabolites with electrophoresis-mass spec (EC-MS) or ¹H NMR. The level of glucose metabolites was adjusted by the viable cell number and expressed as % to the R1881 condition, mean ± s.d., n = 3, *,^#P < 0.05 ADT + hypoxia vs. R1881 + hypoxia, ^#P < 0.05 ADT vs R1881, ^hP < 0.05 hypoxia vs non-hypoxia, two-sided t-test. e–i The effect of GPI-siRNA on glucose metabolic pathways based on PPP-metabolite 6-PGA vs. glycolysis metabolite F6P, in AdtHs cells from LNCaP (LN) or LAPC4 (LA). Cells were cultured in media with CSS + 1 nM R1881, transfected with siRNA (siNT or siGPI), and then treated by solvent control (Ctl) or 10 µM enzalutamide in normoxia (20%) or hypoxia (1%) for 48 h. The enzymatic activity of GPI (e), PPP-metabolite 6-PGA (f), glycolysis/GPI-metabolite F6P (g), the upstream metabolite and substrate for both G6P (h), and glucose uptake (i) were all determined by commercially available kits from BioVision. All values were adjusted with viable cell numbers and expressed as mean + s.d. relative to siNT with solvent control in 20% O₂, n = 3, *P < 0.05 enza + hypoxia vs. Ctl + hypoxia, ^#,^ΔP < 0.05, NSD, two-sided t-test