Enzyme-constrained genome-scale modeling resolves growth-production trade-offs in fermentative biohydrogen production

Hydrogen is central to sustainable energy systems, with biological production from waste offering a low-energy, environmentally compatible route. Anaerobic dark fermentation by microbes converts organic substrates into hydrogen, yet yields remain limited by competing metabolic pathways and poor understanding of cellular resource allocation in hydrogen-producing strains. Conventional genome-scale models rely on stoichiometric constraints alone, often failing to capture realistic enzyme limitations or strain-specific biomass composition. Here we show that an enzyme-constrained genome-scale metabolic model (ecGEM) of the hydrogen-producing bacterium Ethanoligenens harbinense YUAN-3, built with experimentally measured biomass composition and predicted k _cat values, quantitatively captures the trade-off between growth and hydrogen production. Enzyme constraints eliminate unrealistic flux predictions of standard models, accurately matching experimental growth rates and yields, and reveal that diversion of carbon and NADH flux into glutamate and glutamine biosynthesis enhances hydrogen production by reducing ethanol formation. In silico single-gene knockouts identify targets such as phosphoglycerate kinase that increase hydrogen flux by up to 30% under low-carbon conditions. These findings elucidate system-level metabolic regulation in fermentative hydrogen production and provide a predictive framework for rational strain engineering. The approach offers a scalable platform for optimizing biohydrogen processes and advancing sustainable hydrogen economies.