Abstract

Friedreich’s ataxia (FA) is a rare genetic neurodegenerative and cardiac disease caused by the disrupted expression of frataxin (FXN), a mitochondrial protein that promotes the formation of iron-sulfur (Fe-S) clusters. Fe-S clusters play critical roles as protein cofactors in redox and non-redox catalysis, signalling, electron transfer, and sulfur donation. The building block of all Fe-S clusters is the [2Fe‑2S] cluster, whose synthesis is orchestrated by the mitochondrial iron-sulfur cluster (ISC) machinery. This machinery consists of several proteins that deliver iron, sulfur, and electrons to a scaffold protein, ISCU2. The first step of this process is binding of iron to the scaffold protein ISCU2 at its so-called assembly site. Sulfur is sourced from free L-cysteine, which is converted into a cysteine-bound persulfide by the cysteine desulferase NFS1. The persulfide is then transferred to a cysteine residue of ISCU2 located near the iron center of the assembly site. This transfer is accelerated by the regulatory protein FXN. Electrons are provided by NADPH through ferredoxin-2 (FDX2) and its redox-partner ferredoxin-reductase (FDXR) which cleaves the persulfide into sulfide. This reaction ultimately leads to the formation of a [2Fe‑2S] cluster. In the absence of FXN, the system has inefficient [2Fe‑2S] cluster synthesis and leads to cellular toxicity, iron accumulation, increased sensitivity to oxidative stress, and leads to the onset of FA. Several strategies are currently under development to find a treatment for FA. The most promising therapeutic strategies to treat FA have focused on increasing FXN levels via gene therapy. However, this approach has been challenged by the acute cellular toxicity of FXN, which has significantly hindered therapeutic progress.

In this work, we have investigated the mechanistic basis of FXN-induced toxicity in the context of [2Fe‑2S] cluster synthesis. Using a complete in vitro reconstitution of the human ISC system, we demonstrate that FXN and FDX2 compete for overlapping binding sites on the NFS1-ISCU2 complex. Through a combination of isothermal titration calorimetry (ITC) and flow-induced dispersion analysis (FIDA), we characterize the binding of proteins within the ISC complex. We reveal that FDX2 and FXN have similar binding affinities, can competitively replace the other, and describe a novel two-phase binding mechanism of FDX2. Using recombinant assays tracking the rate of [2Fe‑2S] cluster synthesis, persulfide generation, transfer, and reduction, we show that imbalanced FXN and FDX2 levels negatively impacts cluster synthesis at specific steps of the assembly pathway. Moreover, we uncover a previously undescribed regulatory role of FDX2 in decreasing the ability of NFS1 to both generate and transfer a persulfide to ISCU2. In a Drosophila model of FA, we demonstrate that interfering with FDX2 expression can significantly extend lifespan.

Furthermore, we explore potential therapeutic strategies by screening peptides and small molecules that can either functionally replace FXN or disrupt FDX2-mediated inhibition of persulfide transfer. In parallel to this work, we investigated the mechanism leading to the final formation of [2Fe‑2S] clusters following persulfide reduction by FDX2. Using the homologous bacterial ISC machinery, we showed that persulfide reduction by Fdx, the homolog of FDX2, leads to a [1Fe-1S] intermediate. IscU carrying this intermediate then dissociates from the cysteine desulferase IscS and dimerizes to generate a [2Fe‑2S] cluster. This work thus significantly enhances our understanding of the mechanism of [2Fe‑2S] cluster biosynthesis and of overexpression-mediated FXN toxicity, which opens a new therapeutic avenue for the treatment of FA.