Fig. 3: Model predicted association between gene states and metabolic pathway activities.

A Nullclines and steady states in the space of pAMPK and HIF-1 in normal cells. The red line represents the nullcline where the rate of change of HIF-1 is zero, and the blue line represents the nullcline where the rate of change of pAMPK is zero. Solid dots represent stable steady states while hollow dots represent unstable steady states. Each stable state is associated with a metabolic phenotype. Cancer cells can acquire an OXPHOS phenotype, referred to as the state ‘O’, characterized by high pAMPK/low HIF-1 activity. This phenotype represents a scenario where cells optimize energy production through oxidative phosphorylation, which can occur under specific physiological or mild stress conditions during cancer progression; the other state is the glycolytic phenotype that is when oxygen is limited, referred to as the state “W”. B The nullclines and steady states in the phase space of pAMPK and HIF-1 in cancer cells. Cancer cells can acquire two additional hybrid metabolic states—the ‘W/O’ state, characterized by intermediate HIF-1 and pAMPK activity, and the ‘Q’ state, characterized by low pAMPK/HIF-1 but high glutamine oxidation. C Net ATP production rates of different metabolic pathways for the “O”, “W/O”, “W”, and “Q” states. Positive rates represent catabolic processes where ATP is produced, while negative rates represent anabolic processes where ATP is consumed. Compared to the “W” state, the “O” state exhibits higher activities of glucose oxidation (G₁), fatty acid oxidation (F), glutamine oxidation (Q₁), and GSH synthesis (GSH), but lower activities of glycolysis (G₂), reductive glutamine carboxylation (Q_re), reductive glucose metabolism (G_re), and reductive fatty acid metabolism (F_re). The hybrid “W/O” state exhibits intermediate metabolic activities, while the “Q” state relies on high glutamine oxidation activity.