Fuente: PubMed "industrial biotechnology" Front Microbiol. 2026 Jun 26;17:1827551. doi: 10.3389/fmicb.2026.1827551. eCollection 2026.ABSTRACTBiological hydrogen (H2) production via dark fermentation (DF) offers a renewable pathway for energy recovery from biomass, but process performance is highly strain dependent and often limited by suboptimal operating conditions. The present study characterizes fermentation physiology and optimizes small-scale H2 production by the recently discovered thermophilic anaerobe Thermoactinomyces mirandus. Batch cultivations at 52 °C in anoxic serum bottles were used to screen carbon sources (glucose, fructose, xylose, arabinose, and lactose), nitrogen sources (yeast extract, casein peptone, casamino acids, and ammonium chloride), carbon to nitrogen (C/N) ratios, and a targeted set of potential inhibitory factors. Yeast extract supported the highest H2 yields, while fructose, xylose, and lactose were the most effective carbon sources. Baseline lactose fermentation without pH control yielded 118.31 ± 19.91 mmol H2 mol hexose equivalent-1 after 11 days, with incomplete lactose conversion. Implementing a small-scale, closed-flask pH control at setpoint 7.2 increased the yield 4.2-fold to 466.6 ± 10.2 mmol H2 mol hexose equivalent-1 and achieved 99.4 ± 0.2% lactose consumption. Daily nitrogen (N2) gas sparging modestly improved conversion, while 10-fold supplementation of trace elements, ferrous iron, or formate had no effect. Under controlled pH, an optimum pH range of 7.5-8.0 for maximum H2 production rate (VHPR) was identified. Higher lactose loads increased the VHPR by 147% from 5.6 to 13.9 mmol lactose L-1, with a trade-off in yield (26% decrease). Fermentation product profiles shifted under pH control, with ethanol and acetate increasing relative to lactate, consistent with enzyme pH optima and upregulation of pflB (pyruvate formate lyase) after 4 days cultivation. These results demonstrate that pH control is a key factor for enhancing H2 yield by T. mirandus, define quantitative operating windows, and deliver a transferable workflow to identify and mitigate inhibitory factors in DF processes.PMID:42434554 | PMC:PMC13350025 | DOI:10.3389/fmicb.2026.1827551