Asymmetric co-substrate usage at a metabolic branch point as a minimal network motif that can drive overflow metabolism

Abstract

Metabolic overflow is pervasive from microbes to tumors: as glucose consumption increases, cells switch from respiration to respirofermentation with excretion of organic acids or ethanol. How do cells regulate their fluxes to implement this switch? Using simplified models of a branch point, we found that three independent "built-in" mechanisms can each enable flux switching at a branch point: differential enzyme kinetics at the two branches, allosteric regulation of a branch enzyme, and asymmetric co-substrate (e.g. NADH/NAD+) usage across the two branch point re-actions. Notably, the co-substrate mechanism achieves several fold higher sensitivity, which can be further enhanced by the other two mechanisms. We then applied this theory to the yeast pyruvate branch point underpinning overflow metabolism by developing a compartmentalized model with or without NADH dynamics in cytoplasm and in mitochondria. Crucially, only models that included NADH dynamics recapitulated experimental data: (a) NADH level increases nonlinearly with increasing glucose influx; (b) this increase occurs at a lower glucose threshold than that for overflow; and (c) glucose threshold for overflow increases when NADH oxidase is expressed in mitochondria but not in cytosol. Thus, cosubstrates are not merely passive carriers of chemical groups; their asymmetric utilization at a branch point offers an intrinsic, network-embedded regulatory layer that synergises with enzyme-based control. Beyond explaining overflow metabolism in yeast, this theory offers a metabolic engineering strategy for re-routing flux through manipulating co-substrates.