Glutathione contributes to thiol-redox control and to extra-mitochondrial iron-sulphur cluster (ISC) maturation. To determine the physiological importance of these functions and sort out those that account for the GSH requirement for viability, we performed a comprehensive analysis of yeast cells depleted of or containing toxic levels of GSH. Both conditions triggered an intense iron starvation-like response and impaired the activity of extra-mitochondrial ISC enzymes but did not impact thiol-redox maintenance, except for high glutathione levels that altered oxidative protein folding in the endoplasmic reticulum. While iron partially rescued the ISC maturation and growth defects of GSH-depleted cells, genetic experiments indicated that unlike thioredoxin, glutathione could not support by itself the thiol-redox duties of the cell. We propose that glutathione is essential by its requirement in ISC assembly, but only serves as a thioredoxin backup in cytosolic thiol-redox maintenance. Glutathione-high physiological levels are thus meant to insulate its cytosolic function in iron metabolism from variations of its concentration during redox stresses, a model challenging the traditional view of it as prime actor in thiol-redox control.

Iron-sulfur (Fe-S) clusters are essential protein cofactors whose biosynthetic defects lead to severe diseases among which is Friedreich’s ataxia caused by impaired expression of frataxin (FXN). Fe-S clusters are biosynthesized on the scaffold protein ISCU, with cysteine desulfurase NFS1 providing sulfur as persulfide and ferredoxin FDX2 supplying electrons, in a process stimulated by FXN but not clearly understood. Here, we report the breakdown of this process, made possible by removing a zinc ion in ISCU that hinders iron insertion and promotes non-physiological Fe-S cluster synthesis from free sulfide in vitro. By binding zinc-free ISCU, iron drives persulfide uptake from NFS1 and allows persulfide reduction into sulfide by FDX2, thereby coordinating sulfide production with its availability to generate Fe-S clusters. FXN stimulates the whole process by accelerating persulfide transfer. We propose that this reconstitution recapitulates physiological conditions which provides a model for Fe-S cluster biosynthesis, clarifies the roles of FDX2 and FXN and may help develop Friedreich’s ataxia therapies.