1. Introduction

Friedreich’s ataxia (FRDA) is one of the most common rare neurodegenerative disorders, with an estimated incidence of 1 in 30,000 to 1 in 50,000 individuals with European ancestry. It is also the prevalent form of inherited ataxia, a group of diseases that lead to progressive muscle coordination issues. In particular, FRDA occurs in an autosomal recessive manner, and it can manifest with various phenotypic forms due to various epigenetic alterations^[1-4]. Indeed, this disease is caused by the deficiency of the mitochondrial protein frataxin (FXN), triggered by the expansion of the guanine-adenine-adenine (GAA) triplet repeat in the first intron of the FXN gene on chromosome 9q21, which leads to its transcriptional inhibition through heterochromatinization and subsequent gene silencing^[5-8].

FRDA exhibits a wide range of both clinical features and symptom severity, which directly correlate with the number of GAA repeats. This genetic disorder primarily results in injuries to the large neurons within the dorsal root ganglia (DRG) as well as damage in the spinocerebellar tract, leading to the distinctive ataxia features. In addition to coordination symptoms, despite FRDA being primarily a neurological condition, patients also exhibit diabetes and hypertrophic cardiomyopathy, which are the primary cause of premature death^[2,4]. With few exceptions, FRDA patients are compound heterozygotes for the GAA expansion, with most alleles presenting 600-900 GAA repeats, even if alleles with over 1,500 GAA repeats have also been documented^[9,10].

As a result of the FXN gene expansion, FRDA cells are characterized by reduced levels of FXN, a small mitochondrial protein synthesized as a precursor which undergoes a two-step maturation process in the mitochondria, to form the 14 kDa mature protein^[11]. The exact roles of FXN have not yet been completely elucidated, but scientific evidence indicates its role in iron-sulfur cluster (ISC) formation, iron homeostasis/binding, and heme biosynthesis^[12-15]. Specifically, FXN is widely considered to act as a crucial iron chaperone or allosteric regulator within the ISC assembly machinery^[16].

As a consequence, FXN deficiency leads to a pronounced oxidative stress^[17,18] since ISCs constitute prosthetic groups of essential proteins like respiratory chain complexes and energy metabolism enzymes like aconitase^[19]. In detail, a reduced activity of complexes I, II, and III of the mitochondrial respiratory chain as well as of aconitase in mouse models of FRDA, due to the impaired ISCs formation, was reported^[20].

Moreover, iron homeostasis is compromised due to FXN’s intrinsic feature of iron binding^[11,12], and this leads to free iron accumulation which in turn can result in mitochondrial impairment, reactive oxygen species (ROS) generation, and ISC machinery failure^[21-23]. Noteworthily, it is uncertain whether the initiation of mitochondrial failure is due to ISC impairment or iron accumulation and the related ROS production^[24,25]. Indeed, a vicious circle hypothesis has been proposed^[26], suggesting that an initial defect in ISC biogenesis leads to mitochondrial iron overload, which subsequently generates excessive ROS that oxidatively damage the few remaining intact ISCs, further exacerbating the pathology. Although it is not yet known unambiguously whether the FRDA cells’ condition upon FXN deficiency primarily depends on ISC impairment or iron accumulation, it is worth noting that an increment in oxidative stress sensitivity is a key feature of FRDA cells, especially in cells with high respiration rates like neurons and cardiomyocytes^[13,27]. Free iron accumulation in mitochondria is known to be responsible for an increment in ROS production, mainly due to the Fenton reaction, triggering the consequent lipid peroxidation radical reaction that ultimately executes ferroptosis^[28].

Historically, scientific literature has viewed FRDA primarily as a mitochondrial bioenergetic disorder^[29]. However, an emerging conceptual shift suggests that lipid peroxidation and ferroptosis significantly contribute to FRDA pathogenesis, acting in concert with other well-established disease mechanisms such as mitochondrial bioenergetic failure, apoptosis, and inflammatory signaling^[29,30]. Therefore, the aim of this review is to summarize and clarify the recent evidence linking FXN deficiency to ferroptotic cell death. We will systematically dissect the failure of multiple antioxidant defense systems, including the nuclear factor erythroid 2-related factor 2 (Nrf2)/GSH axis and GSH-independent rescue networks, and discuss the therapeutic potential of targeting the ferroptotic cascade.

2. The Core Mechanics of Ferroptosis

2.1 Defining ferroptosis: A distinct modality of regulated cell death (RCD)

Dixon and his team first outlined the concept of ferroptosis in 2012^[31] and the Nomenclature Committee on Cell Death subsequently gave it an official classification^[32]. We now understand it as a distinct iron-dependent form of RCD driven by the accumulation of lipid peroxidized toxic products at the plasma and intracellular membranes^[33,34]. What truly sets this process apart from well-known pathways like apoptosis, necrosis, and autophagy is its unique footprint, encompassing how the cell looks, how it reacts chemically, and the genes driving the collapse^[31,35].

A ferroptotic cell reveals the absence of the traditional apoptotic signs. Chromatin does not condense. Nuclei stay intact, and apoptotic bodies never form^[36]. The early membrane ruptures and swollen organelles of necrosis are missing as well, along with the prominent double-membrane vesicles that characterize classical autophagy^[37]. The physical damage is instead principally confined to one location: the mitochondria. Under an electron microscope, these organelles appear shrunken and severely distorted. Their membranes pack together with unusual density, cristae vanish completely, with frequent ruptures across the outer membrane^[31,38].

Caspases play no role in this biochemical cascade. The administration of pan-caspase inhibitors (e.g., z-VAD-FMK) fails to halt the execution of cell death^[39,40]. Rescue requires a completely different pharmacological approach. Only iron chelators (such as deferoxamine (DFO)) or lipophilic radical-trapping antioxidants (RTAs) (such as ferrostatin-1 (Fer-1) and liproxstatin-1 (Lip-1)) can successfully interrupt the process^[31,41,42]. This specific vulnerability proves that a broken iron balance and runaway oxidative stress drive the execution phase, bypassing standard apoptotic signaling entirely.

The underlying genetics confirm this independence. The regulatory genes controlling ferroptosis simply do not cross paths with the known mediators of apoptosis (such as the BCL-2 family) or autophagy (such as the ATG genes)^[43]. Survival instead hinges on networks managing lipid metabolism (e.g., Acyl-CoA synthetase long-chain family member 4 (ACSL4), lysophosphatidylcholine acyltransferase 3 (LPCAT3))^[44], iron transport and storage (e.g., transferrin receptor, ferritin heavy polypeptide 1, nuclear receptor coactivator 4 (NCOA4))^[45], and redox homeostasis (e.g., SLC7A11, glutathione peroxidase 4 (GPX4))^[46,47]. The transcription factor Nrf2 acts as the central hub here. By dictating the expression of the cystine/glutamate antiporter SLC7A11 alongside several core antioxidant enzymes, Nrf2 effectively sets the cell’s tolerance limit for ferroptotic stress^[48,49]. Artificially silencing or overexpressing these targets immediately shifts the cell’s vulnerability. Such genetic flexibility confirms that we are dealing with an entirely independent, genetically programmed death pathway rooted in cellular metabolism^[38].

2.2 Iron-catalyzed ROS generation and lipid peroxidation

Iron is an essential transition metal, but its highly reactive ferrous form (Fe²⁺) is potentially cytotoxic. To prevent damage, cells confine unbound redox-active iron to the labile iron pool (LIP)^[50]. During ferroptosis, the abnormal expansion of the LIP catalyzes the spontaneous generation of ROS. Specifically, Fe²⁺ reacts with intracellular hydrogen peroxide (H₂O₂), through the Fenton reaction, yielding the highly reactive hydroxyl radical (•OH)^[51]. Within cellular microenvironments, this process is frequently coupled with the Haber-Weiss reaction, wherein the superoxide anion (O₂•^-) from mitochondria continuously reduces Fe³⁺ back to Fe²⁺, creating a self-sustaining and vicious cycle of radical generation^[51,52]. Because the hydroxyl radical has an extremely short half-life and reacts non-specifically with molecules in its immediate proximity, its generation near lipid bilayers is particularly detrimental. In pathological conditions where the cellular iron-buffering capacity (mediated by ferritin) is overwhelmed or disrupted, the uncontrolled Fenton reaction becomes the proximal trigger for the ferroptotic cascade. While the LIP exists throughout the cytosol, the mitochondrial compartment is uniquely vulnerable in FRDA due to the simultaneous pathological iron overloading and localized H₂O₂ overproduction from impaired respiratory complexes^[53,54].

Hydroxyl radicals predominantly target polyunsaturated fatty acids (PUFAs), such as arachidonic acid and adrenic acid (AdA)^[55,56], which must first be esterified into membrane phospholipids (PUFA-PEs) by enzymes such as ACSL4 and LPCAT3^[57]. The abstraction of bis-allylic protons by •OH initiates a chain reaction of lipid peroxidation, forming lipid peroxyl radicals (PLOO•) and lipid hydroperoxides (PLOOH). This propagation is further accelerated by arachidonate lipoxygenases (ALOXs) and cytochrome P450 oxidoreductase^[55,58]. Under physiological conditions, GPx4 reduces toxic PLOOH molecules into harmless lipid alcohols. However, when GPx4 fails, unchecked lipid hydroperoxides undergo an iron-catalyzed breakdown into alkoxyl radicals (PLO•)^[47,59], significantly expanding the oxidized lipidome as highlighted by recent mass spectrometry analyses^[40,60]. The subsequent degradation of these diverse oxidized species yields reactive aldehydes, such as 4-hydroxynonenal (4-HNE) and malondialdehyde (MDA), which cross-link proteins and DNA^[61,62]. Ultimately, this unmitigated lipid peroxidation fundamentally alters membrane biophysics, disrupting the lipid bilayer and leading to rapid cell death^[59].

3. The Intersection: Molecular Mechanisms of Ferroptosis in FRDA

The susceptibility of FXN-deficient cells to ferroptosis cannot be attributed to a single biochemical failure; rather, it emerges from the concurrent impairment of multiple intersecting defense mechanisms. Current evidence suggests that FRDA pathogenesis creates a highly conducive environment for lipid peroxidation (Figure 1), driven by dysregulated iron handling, amino acid metabolic shifts, and the failure of both glutathione (GSH)-dependent and independent pathways^[29,63]. However, while the collapse of the GSH/GPx4 axis and iron dysregulation are well-documented hallmarks of FXN deficiency, the precise involvement of other canonical ferroptotic nodes, such as the ferroptosis suppressor protein 1 (FSP1)/dihydroorotate dehydrogenase (DHODH) rescue pathways, NCOA4-mediated ferritinophagy, and ACSL4-driven lipid remodeling, remains largely inferred. These latter mechanisms represent compelling translational hypotheses that currently lack extensive direct experimental validation in specific FRDA models.

Display Full Size

Download

Figure 1. The ferroptotic landscape in Friedreich’s ataxia. Frataxin deficiency in the mitochondrion leads to ISC failure and ROS generation. Simultaneously, impaired cellular defenses, characterized by Nrf2 downregulation, lead to decreased GSH levels and defective GPx4 activity. In the cytosol, aberrant IRP1 activation increases the labile iron pool through enhanced transferrin-mediated iron import and NCOA4-mediated ferritinophagy. At the membrane level, lipid remodeling (orange arrows) is driven by PLA2 which generates LPLs, while ACSL4 activates free PUFAs into PUFA-CoAs, which are subsequently esterified into the membrane by LPCAT3 to form highly vulnerable PUFA-PEs. The convergence of accumulated iron, ROS, and ACSL4-mediated incorporation of PUFA-PEs into the membrane results in unchecked LPO, ultimately driving ferroptotic cell death. Green upward arrows denote upregulation or accumulation; red downward arrows denote downregulation or functional impairment. Created in BioRender. Cozza, G. (2026) https://BioRender.com/9m8dus3. ISC: iron-sulfur cluster; ROS: reactive oxygen species; Nrf2: nuclear factor erythroid 2-related factor 2; GSH: glutathione; GPx4: glutathione peroxidase 4; IRP1: iron regulatory protein 1; NCOA4: nuclear receptor coactivator 4; PLA2: phospholipase A2; LPLs: lysophospholipids; ACSL4: Acyl-CoA synthetase long-chain family member 4; LPCAT3: lysophosphatidylcholine acyltransferase 3; PUFA-PE: polyunsaturated fatty acid-phosphatidylethanolamines; LPO: lipid peroxidation.

3.1 The downregulation of Nrf2/system X_c axis

Defending against ferroptosis starts right at the cell surface. Here, the System X_c^- antiporter, built from the catalytic subunit SLC7A11 and its regulatory chaperone SLC3A2, acts as the primary gatekeeper. It enforces a strict 1:1 trade: pushing intracellular glutamate out to pull extracellular cystine in. This specific exchange provides the absolute rate-limiting fuel for GSH biosynthesis^[64].

In FRDA, however, this defense system breaks down at the genetic level. The problem centers on the Nrf2 pathway^[65-67]. Under physiological conditions, Nrf2 is constitutively targeted for proteasomal degradation by the Kelch-like ECH-associated protein 1 (Keap1)-Cul3-Rbx1 E3 ubiquitin ligase complex. When cells face oxidative stress, electrophiles or ROS oxidize specific reactive cysteine residues on Keap1 (e.g., Cys151), inducing a conformational change that prevents Nrf2 ubiquitination. Newly synthesized Nrf2 then accumulates and translocates to the nucleus, where it binds to antioxidant response elements (AREs) to drive the transcription of cytoprotective genes, including SLC7A11 (which drives cystine import) as well as glutamate-cysteine ligase and glutathione synthetase, which collectively orchestrate the de novo synthesis of GSH^[68-70]. Yet, the loss of FXN somehow blunts this vital response. Despite the presence of chronic oxidative stress, the expected Nrf2-mediated compensatory response is markedly blunted^[71]. This impaired activation is hypothesized to stem from defects in the actin cytoskeleton, which is often disorganized in FXN-deficient cells, thereby hindering the efficient nuclear translocation of Nrf2. An overly stable Nrf2-Keap1 inhibitory complex makes matters even worse. If impaired autophagy prevents the cell from clearing Keap1, Nrf2 remains locked down, completely unable to trigger antioxidant defense genes^[71,72]. Without Nrf2 driving transcription, SLC7A11 expression drops sharply. Cystine import is, in turn, severely restricted^[73]. Deprived of this critical precursor, the cell loses its capacity to sustain de novo GSH synthesis. This metabolic failure neutralizes the primary barrier against ferroptosis, leaving sensitive neuronal and cardiac tissues highly vulnerable to extensive oxidative damage^[74,75].

3.2 The impairment of the glutathione/GPx4 axis

As already mentioned, GPx4 represents the central enzymatic shield against ferroptosis. Unlike other peroxidases, GPx4 possesses the unique ability to directly reduce toxic PLOOH within complex biological membranes into harmless lipid alcohols^[34,47]. However, the catalytic cycle of GPx4 is strictly dependent on the continuous availability of reduced GSH as a cofactor, which serves as the essential electron donor for the reduction of the enzyme’s active-site selenocysteine^[34,76].

In the context of FRDA, the intracellular GSH pool is subjected to a dual assault. First, the aforementioned Nrf2/SLC7A11 impairment limits GSH production^[74,77]. Second, the chronic mitochondrial ROS leakage, characteristic of FXN deficiency, continuously depletes existing GSH pools to neutralize superoxide and hydrogen peroxide radicals^[78,79]. This chronic oxidative stress shifts the cellular redox balance toward a pro-oxidant state, leading to a marked decrease in the GSH/GSSG ratio^[80-82]. The ensuing GSH depletion indirectly inactivates GPx4. Without functional GPx4, the cell loses its primary mechanism to clear lipid peroxides, leaving cellular membranes highly vulnerable to iron-catalyzed oxidative damage.

Crucially, comparative analyses across diverse FRDA models, ranging from patient-derived fibroblasts to induced pluripotent stem cell-derived neurons and YG8sR transgenic mice, confirm a unanimous consensus regarding the collapse of the GSH-GPx4 axis, albeit with critical tissue-specific nuances^[83,84]. Systemic GSH depletion is a ubiquitous hallmark, readily detectable even in patient blood and isolated peripheral fibroblasts^[74]. However, in these peripheral models, the GSH deficit typically translates into an enhanced susceptibility to induced ferroptosis (e.g., upon pharmacological GSH depletion or iron overload) rather than spontaneous cell death^[85]. In contrast, highly metabolically active cells like neurons and cardiomyocytes, which are inherently enriched in PUFAs, exhibit a dramatically lower ferroptotic threshold^[86,87]. In these specific targets, current experimental evidence points to a functional impairment of GPx4 due to the severe depletion of its obligate cofactor, GSH, exacerbated by the failure of the Nrf2 response^[49,75]. This loss of catalytic activity, combined with chronic iron-mediated ROS production, leaves these vulnerable cells unable to clear accumulating lipid peroxides, irrevocably predisposing them toward ferroptotic execution^[83,88].

3.3 GSH-independent rescue: FSP1, DHODH, and CoQ10 dynamics

The two most prominent GSH-independent rescue pathways are mediated by FSP1 at the plasma membrane^[89,90] and DHODH within the inner mitochondrial membrane^[91]. Both enzymes function by reducing ubiquinone (Coenzyme Q10, or CoQ10) to its biologically active, antioxidant form, ubiquinol (CoQH₂), which acts as a lipophilic RTA to halt the propagation of lipid peroxyl radicals^[89,90].

The reliance on these secondary pathways is particularly critical in FRDA due to the disease’s primary bioenergetic defect. In healthy mitochondria, the continuous reduction of the CoQ10 pool is predominantly driven by respiratory chain complexes (namely Complex I and II). However, because FXN deficiency impairs ISC biogenesis, these specific complexes suffer substantial structural and functional alterations^[20]. Consequently, the physiological reduction of mitochondrial CoQ10 is compromised. Compounding this enzymatic failure, clinical evidence has demonstrated a significant systemic reduction in the lipophilic antioxidant pool itself. As recently demonstrated, patients with FRDA exhibit markedly lower plasma levels of both Coenzyme Q10 and vitamin E compared to healthy controls^[92]. This quantitative depletion suggests that FSP1 and DHODH are not only deprived of their primary reducing partners (due to Complex I/II failure) but are also deprived of their essential antioxidant substrates^[92].

In this compromised state, these secondary defenses must operate in spatially distinct compartments to prevent extensive cellular dysfunction. While FSP1 acts primarily to halt lipid peroxidation at the plasma membrane, the DHODH pathway, which utilizes flavin mononucleotide rather than ISCs for its catalytic activity, has been proposed as a compensatory endogenous mechanism to maintain a reduced pool of CoQH₂ and defend against localized mitochondrial membrane peroxidation^[93]. However, it is important to note that the ferroptosis-protective role of DHODH remains a matter of ongoing debate and appears to be highly context- and cell-type-dependent^[94]. Therefore, its specific contribution to FRDA requires cautious interpretation and further direct experimental validation.

This novel anti-ferroptotic lens allows us to recontextualize historical therapeutic approaches in FRDA. Idebenone, a short-chain synthetic benzoquinone analog of CoQ10, was historically evaluated in clinical trials to bypass mitochondrial complex I/II/III defects and scavenge ROS^[95]. While clinical outcomes were mixed, likely due to pharmacokinetic limitations as exemplified by the failure of the Phase III MICONOS trial to meet its primary neurological endpoints, and the Phase III IONIA trial for cardiomyopathy, the molecular rationale can now be reinterpreted^[95-97]. The emerging anti-ferroptotic perspective suggests that while idebenone can theoretically act as a surrogate for the FSP1/DHODH axes^[98], its limited bioavailability and sub-optimal mitochondrial targeting may have prevented effective suppression of lipid peroxidation at the tested doses. Investigating how FXN deficiency impacts FSP1 and DHODH expression remains a critical frontier in FRDA research.

3.4 The iron paradox and NCOA4-mediated ferritinophagy

The hallmark of FRDA is a distinct iron paradox: a significant accumulation of mitochondrial iron deposits coexists with a perceived cytosolic iron deficiency (Figure 2). This cytosolic “pseudo-starvation” signal, caused by the sequestration of iron within the mitochondrial matrix and the concomitant failure of ISC biogenesis, triggers a compensatory but ultimately maladaptive response^[2,99,100]. The molecular sensor of this paradox is iron regulatory protein 1 (IRP1). In healthy cells, IRP1 functions as a cytosolic aconitase when bound to a 4Fe-4S cluster. In FRDA, the lack of FXN prevents cluster assembly, converting IRP1 into its active RNA-binding form regardless of actual cellular iron levels^[19,99]. This constitutive activation of IRP1 leads to a coordinated shift in the iron proteome: the stabilization of transferrin receptor 1 (TfR1) mRNA to increase cellular iron import and the translational repression of cytosolic ferritin to limit iron storage. Experimental data in FXN-deficient models confirm this specific upregulation of TfR1 and concomitant decrease in cytosolic ferritin, effectively driving continuous iron influx^[101].

Display Full Size

Download

Figure 2. The iron paradox and cytosolic pseudo-starvation signal. A hierarchical representation of the maladaptive iron-sensing cascade in Friedreich’s ataxia. Mitochondrial frataxin deficiency and subsequent ISC failure create a pathological compartmentalization of iron. While mitochondria become iron-overloaded, the cytosol registers a false “pseudo-starvation” signal. This mismatch constitutively activates IRP1, which orchestrates a tripartite response: upregulation of TfR1, translational repression of ferritin, and activation of NCOA4-mediated ferritinophagy. These coordinated events lead to a deleterious expansion of the cytosolic labile iron pool, providing the catalytic fuel for the Fenton reaction and subsequent ROS generation. Created in BioRender. Cozza, G. (2026) https://BioRender.com/9m8dus3. ISC: iron-sulfur cluster; IRP1: iron regulatory protein 1; TfR1: transferrin receptor 1; NCOA4: nuclear receptor coactivator 4; ROS: reactive oxygen species.

One of the most critical pathways in this context is ferritinophagy, the autophagic degradation of ferritin, the cell’s primary iron-storage protein, to release free iron into the cytosol^[102]. This process is mediated by the cargo receptor NCOA4, which targets ferritin to the lysosome for degradation^[102,103]. In FRDA, the persistent iron starvation signal and the lack of functional ferritin storage keep ferritinophagy upregulated^[84]. However, instead of restoring iron homeostasis, the continuous breakdown of the few remaining ferritin heteropolymers releases more iron into the LIP^[45,84]. This expanded LIP serves as a continuous catalyst for the Fenton reaction, significantly accelerating the generation of hydroxyl radicals and the initiation of lipid peroxidation. Thus, what begins as a survival mechanism to counteract perceived iron deficiency, driven by the “blind” activation of the IRP1-NCOA4 axis, becomes a primary driver of ferroptotic execution in FXN-deficient neurons and cardiomyocytes^[20,84,99].

3.5 Lipid remodeling: ACSL4 and membrane susceptibility

Ferroptosis cannot occur without the proper structural substrates: PUFAs. The integration of these highly vulnerable fatty acids into biological membranes is strictly governed by a deacylation-reacylation process known as the Lands cycle. Initially, phospholipases (such as PLA2) cleave saturated or monounsaturated fatty acids from the sn-2 position of membrane phospholipids, creating lysophospholipids. Subsequently, ACSL4 activates free PUFAs (particularly arachidonic and adrenic acids) into PUFA-CoAs. Finally, LPCAT3 specifically esterifies these activated PUFAs into the lysophospholipid acceptors, forming mature PUFA-PEs^[44].

Emerging lipidomic and metabolomic analyses of FRDA models have confirmed that FXN deficiency induces significant intracellular lipid remodeling. Recent profiling in patient-derived cells demonstrated an abnormal enrichment of PUFA-containing lipids, a dysfunction that directly correlates with disease severity and decreased FXN levels^[84,104]. These specific lipid species, most notably arachidonic acid-containing phosphatidylethanolamines, are the preferred substrates for ALOX enzymes and the subsequent peroxidative chain reactions, significantly lowering the threshold for ferroptotic execution^[55]. Because neurons and cardiomyocytes, the primary cell types affected in FRDA, rely on extensive, complex membrane networks for action potential propagation and contractility, their membranes are inherently rich in PUFAs^[86,87]. If ACSL4 activity is maintained or upregulated in FRDA, it continuously supplies the lipid substrates for ferroptosis. The intersection of this PUFA-enriched membrane landscape with an impaired GPx4 antioxidant shield and an expanded LIP makes the membranes of the central nervous system (CNS), and the heart highly susceptible to extensive lipid peroxidation^[84].

4. Therapeutic Perspectives Targeting Ferroptosis in FRDA

The emergence of ferroptosis as a pivotal mechanism in FRDA partially redefines the landscape for therapeutic intervention^[105]. For years, clinical efforts focused on generalized mitochondrial enhancement or broad-spectrum antioxidant support, which ultimately fell short. Modern strategies can instead target the exact molecular nodes of the ferroptotic cascade. Addressing the Nrf2/GSH axis, clearing compartmentalized iron, or halting lipid peroxidation directly allows the development of new pharmacological tools. Evidence from patient-derived cellular models, FXN-deficient rodents, and ongoing clinical trials (Figure 3) suggests that these specific interventions could provide a compelling alternative to mitigate the defining neurodegeneration and cardiomyopathy of FXN deficiency.

Display Full Size

Download

Figure 3. Pharmacological modulation of the ferroptotic cascade. Mechanistic mapping of current and experimental therapeutic interventions targeting specific nodes of ferroptosis. (1) Iron compartmentalization strategies contrast systemic chelators like DFO/DFX, which fail to cross the BBB, with membrane-permeable agents like DFP/TPP-chelators that specifically access mitochondrial iron deposits; (2) Enzymatic inhibition via vatiquinone (EPI-743) blocks the 15-LOX enzyme, mitigating PUFA oxidation; (3) Redox restoration using omaveloxolone targets the Keap1-Nrf2 molecular axis. Under uninhibited conditions (bottom pathway), the Keap1 sensor targets Nrf2 for proteolysis and degradation. Omaveloxolone inhibits Keap1, releasing Nrf2 and allowing its nuclear translocation. Inside the nucleus, Nrf2 binds to ARE, driving the transcription of crucial antioxidant genes (SLC7A11, GCL, GSS). This restores GSH production and reinforces the GPx4 enzymatic shield to neutralize toxic PLOOH; (4) RTAs directly scavenge and neutralize lipid radicals within the membrane structure. Flat-headed lines indicate enzymatic or receptor inhibition. Created in BioRender. Cozza, G. (2026) https://BioRender.com/9m8dus3. DFO: deferoxamine; DFX: deferasirox; BBB: blood-brain barrier; DFP: deferiprone; TPP: triphenylphosphonium; 15-LOX (ALOX15): 15-lipoxygenase; PUFA: polyunsaturated fatty acids; Keap1: Kelch-like ECH-associated protein 1; Nrf2: nuclear factor erythroid 2-related factor 2; ARE: antioxidant response element; SLC7A11: solute carrier family 7 member 11; GCL: glutamate-cysteine ligase; GSS: glutathione synthetase; GSH: glutathione; GPx4: glutathione peroxidase 4; PLOOH: phospholipid hydroperoxide; RTAs: radical-trapping antioxidants.

4.1 Restoring the redox balance: Nrf2 activators

A major milestone in FRDA management has been the development and regulatory approval of omaveloxolone (Skyclarys^®, RTA 408), which was authorized by the Food and Drug Administration in 2023 based on improvements in neurological function scores (mFARS) observed in the Phase II/III MOXIe trial^[106]. However, it is crucial to temper this milestone with a critical clinical perspective: the overall clinical efficacy of omaveloxolone remains relatively modest. Furthermore, definitive evidence regarding long-term survival benefits and true structural disease modification is still pending and requires extended follow-up. Traditionally classified as a synthetic oleanane triterpenoid and an Nrf2 activator, omaveloxolone acts as a highly potent, reversible Michael acceptor. It selectively binds to highly reactive cysteine residues (primarily Cys-151) on the Keap1 sensor protein^[107]. Omaveloxolone functions by inhibiting the Keap1-mediated degradation of Nrf2, thereby facilitating its nuclear translocation and the subsequent transcription of ARE-dependent genes. In the context of ferroptosis, this pharmacological activation directly counteracts the Nrf2 impairment characteristic of FRDA^[107].

Extensive preclinical evaluations have provided a robust mechanistic foundation for its anti-ferroptotic efficacy. As specifically demonstrated in recent studies on FXN-deficient DRG neurons and patient fibroblasts, treatment with omaveloxolone successfully drives the upregulation of Nrf2 and the GPx4 enzymatic shield. By reinstating cystine import and resolving the GSH biosynthetic bottleneck, omaveloxolone effectively restores metabolic and iron balance, significantly protecting these cells from lipid peroxidation execution^[108].

The clinical success of omaveloxolone must be contextualized against the historical failure of previous Nrf2 activators in FRDA. First-generation activators, including natural isothiocyanates like sulforaphane (SFN) and fumaric acid esters like dimethyl fumarate (DMF), demonstrated in vitro efficacy but failed to reach advanced clinical translation. The primary limitation of SFN and DMF is their nature as non-selective, broad-spectrum electrophiles. They bind non-specifically to numerous cellular targets beyond Keap1, resulting in narrow therapeutic windows and significant off-target toxicity^[107]. Furthermore, compounds like DMF can act as direct Michael acceptors for GSH itself; at higher concentrations, they can paradoxically deplete the free GSH pool, an effect that would exacerbate ferroptosis^[109]. Recently, this functional divergence was experimentally validated; while omaveloxolone successfully improved cardiac function in a severe FRDA mouse model, equivalent dosing with DMF did not manage to rescue the cardiomyopathy^[110]. However, despite the functional improvements, neither omaveloxolone nor DMF reversed the underlying structural derangements, such as interstitial fibrosis or cardiac hypertrophy^[110]. Furthermore, data from female FXN-cKO mice suggested that chronic activation of the Nrf2 signaling axis with omaveloxolone may inadvertently shift the cellular redox potential into a “too reduced” state, known as reductive stress. This overly reduced environment can paradoxically hinder cellular signaling and contribute to cardiac remodeling^[110].

Crucially, the clinical development of omaveloxolone was heavily influenced by the failures of its predecessor, Bardoxolone methyl (RTA 402). While RTA 402 is a highly potent triterpenoid Nrf2 activator, its Phase III clinical trial for chronic kidney disease (the BEACON trial) was prematurely terminated due to severe cardiovascular toxicity, specifically fluid retention and heart failure^[111]. Given that hypertrophic cardiomyopathy is the leading cause of mortality in FRDA, the use of RTA 402 was strictly contraindicated. Omaveloxolone was therefore specifically synthesized and pharmacologically optimized to eliminate this fluid-retaining liability while maintaining target selectivity, making it suited for the highly vulnerable cardiac and neurological landscape of FRDA^[108,112].

4.2 Iron chelators: Finding the right compartment

Given the profound iron dyshomeostasis in FRDA, iron chelation has long been explored as a therapeutic strategy. However, selecting the appropriate chelator is fundamentally a problem of compartmentalization. Classical iron chelators, such as the hexadentate DFO and the oral tridentate deferasirox (DFX), which are highly effective in systemic iron overload diseases (e.g., beta-thalassemia), are largely contraindicated in FRDA. Due to their large molecular size, hydrophilicity, and specific binding kinetics, DFO and DFX fail to efficiently cross the blood-brain barrier and the inner mitochondrial membrane^[113-116]. Consequently, they excessively deplete cytosolic iron without accessing the pathological mitochondrial deposits, further exacerbating the cellular “iron starvation” signal. Therefore, the primary candidate for FRDA clinical translation became deferiprone (DFP), a small, lipophilic bidentate chelator capable of crossing cellular and mitochondrial membranes to access and theoretically redistribute the toxic accumulation of mitochondrial iron to extracellular sinks^[117,118]. While DFP has demonstrated the capacity to reduce mitochondrial iron overload and mitigate oxidative stress in preclinical models, its clinical application highlights a critical challenge. The therapeutic window for systemic iron chelation in FRDA is notoriously narrow. Phase III clinical trials, most notably the 6-month randomized controlled study, revealed a significant dose-dependent adverse effect: while conservative dosing (20 mg/kg/day) stabilized cardiac indices, aggressive chelation (40 mg/kg/day) precipitated a significant worsening of ataxia and neurological symptoms^[119].

From a ferroptotic perspective, this clinical failure is highly predictable. Aggressive chelation even with a membrane-permeable agent like DFP risks exacerbating the pre-existing “cytosolic illusion” of iron starvation. As previously discussed, worsening the cytosolic iron deficit hyperactivates IRP1/2 and can precipitate further NCOA4-mediated ferritinophagy. This paradoxical response can inadvertently expand the cytosolic LIP through uncontrolled lysosomal iron release and trigger the very Fenton chemistry the therapy intends to prevent^[103]. Consequently, the future of iron-targeted therapeutics in FRDA relies on the engineering of highly compartmentalized tools that exclusively target the mitochondrial LIP without depleting cytosolic iron. The primary strategy to create these targeted tools involves conjugating a validated iron-chelating moiety (such as DFP derivatives) to a lipophilic cation, most notably triphenylphosphonium (TPP). Because the mitochondrial matrix is highly electronegative relative to the cytosol, these TPP-conjugated chelators selectively accumulate within the mitochondria driven by the membrane potential^[120]. Once inside, they can successfully sequester pathological iron deposits directly within the matrix, neutralizing ROS generation without disrupting systemic or cytosolic iron-sensing networks.

4.3 From quinone-based therapeutics to direct ferroptosis inhibitors

Direct pharmacological inhibition of ferroptosis represents a highly promising frontier for FRDA, an approach that has evolved significantly over the past two decades. Historically, attempts to halt lipid peroxidation relied on replenishing the depleted lipophilic antioxidant pool through the combined systemic administration of high-dose Coenzyme Q10 and vitamin E, or via short-chain synthetic analogs like idebenone^[121,122]. As previously discussed, while conceptually sound for bypassing respiratory chain defects and scavenging general ROS, these first-generation approaches largely failed to alter the neurological trajectory of the disease, primarily due to poor blood-brain barrier penetrance and an inability to reach the critical threshold required to arrest the ferroptotic cascade in vivo^[122]. The evolution of this therapeutic strategy led to the development of vatiquinone (EPI-743), a synthetic para-benzoquinone analog of vitamin E. Unlike general scavengers, vatiquinone marks a paradigm shift toward targeted anti-ferroptotic therapy. Its primary mechanism of action involves the direct inhibition of 15-Lipoxygenase, a key oxidoreductase enzyme that catalyzes the targeted peroxidation of PUFAs, thereby acting as a critical upstream driver of ferroptosis^[123]. However, the exact mechanism of action of vatiquinone may extend beyond enzymatic inhibition. As demonstrated for other lipoxygenase inhibitors^[124], and considering its structural similarity to quinones with long hydrophobic side chains, such as CoQ10 and vitamin K, the latter acting as a potent ferroptosis suppressor^[125], it is highly plausible that vatiquinone, upon conversion to its hydroquinone form, could also exert direct, LOX-independent RTA activity. This dual potential mechanism warrants further investigation to fully elucidate its protective effects in FRDA. The clinical relevance of targeting this specific node was recently evaluated in the global Phase III MOVE-FA trial. Although the study did not meet its primary endpoint (placebo-corrected change in the overall mFARS score at 72 weeks), vatiquinone treatment demonstrated nominally significant and clinically meaningful benefits in specific disease subscales, most notably preserving upright stability and reducing fatigue, with an excellent safety profile. These data support the biological hypothesis that inhibiting the enzymatic drivers of lipid peroxidation can positively modify aspects of disease progression in patients (ClinicalTrials.gov ID: NCT04577352). However, the failure to meet the primary clinical endpoint highlights a recurrent translational challenge in FRDA: the significant gap between a robust mechanistic rationale, halting lipid peroxidation at the source, and achieving comprehensive clinical benefit. While subscale improvements are encouraging, this discrepancy suggests that single-node ferroptosis inhibition may not be sufficient to overcome the multi-systemic metabolic collapse in FRDA. Building upon the rationale of lipoxygenase inhibition, the next preclinical frontier involves pure RTAs, such as Fer-1 and Lip-1^[41,42]. These highly lipophilic molecules localize directly to cellular membranes, where they act as surrogate scavengers of lipid radicals (such as PLOO• and PLO•)^[41,42]. Mechanistically, RTAs intercept the peroxidative chain reaction propagating through the highly vulnerable, PUFA-enriched phospholipid networks. By acting as potent kinetic traps for these radicals, RTAs effectively halt lipid peroxidation even in the absence of functional GPx4 and despite the continuous iron influx driven by the IRP1/NCOA4 axis^[41,42]. In vitro and in vivo studies utilizing FXN-knockdown models have consistently demonstrated that treatment with Fer-1 or optimized analogs like SRS11-92 prevents cell death, preserves mitochondrial morphology, and reduces terminal markers of lipid peroxidation^[88]. Despite their potent efficacy as experimental probes, the clinical translation of first-generation RTAs is currently hindered by poor in vivo stability, rapid metabolic clearance, and suboptimal pharmacokinetic profiles. However, these proof-of-concept studies and the clinical signals from vatiquinone validate the targeted inhibition of lipid peroxidation as a highly viable therapeutic strategy (summarized in Table 1).

4.4 The translational landscape and the rationale for combinatorial interventions

As summarized in Table 1, the translational landscape of FRDA therapeutics is marked by both milestones and significant setbacks. For discontinued or failed clinical trials, specific reasons for failure often highlight the complexity of the disease. For instance, the Phase III trials for idebenone (MICONOS/IONIA) failed to meet primary neurological endpoints, likely due to suboptimal bioavailability and failure to achieve critical tissue concentrations^[95-97]. Aggressive systemic iron chelation with DFP faced dose-limiting neurotoxicity^[119], and broad-spectrum Nrf2 activators (like DMF) were discontinued due to off-target effects and paradoxical GSH depletion^[107,109].

These historical setbacks emphasize that while single-agent therapies targeting specific ferroptotic nodes are valuable, they often fall short of halting the multi-systemic progression of FRDA. Consequently, the future of FRDA therapeutics inevitably points toward combinatorial strategies. However, combining targeted agents requires careful consideration of potential synergies and risks. For example, co-administering an Nrf2 activator (to boost antioxidant defenses) with a specifically mitochondria-targeted iron chelator could synergistically protect against ferroptotic execution. Conversely, poorly optimized combinations, such as coupling Nrf2 activators with aggressive systemic iron chelators, might inadvertently exacerbate the cytosolic iron “pseudo-starvation” signal, potentially overriding the benefits of either drug and precipitating toxicity. Therefore, rigorous preclinical profiling of potential drug-drug interactions is essential before advancing combinatorial strategies into clinical trials.

5. Conclusion and Future Directions

The integration of ferroptosis into the pathogenic model of FRDA redefines our understanding of the disease, providing a framework that bridges mitochondrial dysfunction with RCD. However, it is essential to acknowledge that ferroptosis is likely not an isolated, primary driver of the disease, but rather acts as a parallel contributor or a secondary consequence of the initial mitochondrial defect. Long-established pathogenic mechanisms, including defective bioenergetics, apoptosis, and inflammatory signaling, remain fundamental to FRDA pathology. Ferroptosis represents the culmination of these intersecting metabolic failures, emerging when the collapse of the lipid redox landscape exceeds the cellular threshold for survival. The transition from a strictly ATP-centric view to a model characterized by the severe dysregulation of lipid redox homeostasis presents novel, actionable pharmacological nodes. As this review has highlighted, FXN deficiency creates a unique cellular environment where multiple ferroptosis defense nodes, including the Nrf2-driven System X_c^-/GSH axis and the GPx4 enzymatic shield, are simultaneously compromised. This vulnerability is further exacerbated by the mitochondrial iron paradox and subsequent NCOA4-mediated ferritinophagy, which together supply the catalytic fuel for excessive lipid peroxidation. Moving forward, several critical knowledge gaps must be addressed to successfully translate these molecular insights into definitive disease-modifying therapies.

5.1 The biomarker bottleneck

Translating anti-ferroptotic therapies into the clinic faces one major roadblock: a complete lack of validated, non-invasive in vivo biomarkers. Basic research heavily relies on standard oxidative stress markers (like MDA, 4-HNE, or GSH/GSSG ratios) for mouse and cell models, but this approach frequently yielded inconclusive or contradictory results in a clinical setting due to methodological limits and sample sizes^[35,126]. To bridge this gap, the field needs to pursue advanced lipidomic profiling using patient biofluids (such as cerebrospinal fluid and plasma) to identify and quantify the specific oxidized PUFA species (e.g., oxPEs) that represent the ultimate executioners of the ferroptotic cascade^[55,104]. Importantly, recent clinical research has laid a solid foundation for this approach. Comprehensive lipidomic profiling of FRDA patient-derived fibroblasts has demonstrated a profound pathological remodeling of the lipidome that directly predicts disease severity^[104]. This dysfunction is characterized by an abnormal enrichment of PUFA-containing triglycerides and phospholipids, alongside elevated acylcarnitines and specific long-chain ceramides, all of which strongly correlate with GAA repeat length and decreased FXN levels^[104]. Furthermore, terminal by-products of this lipid peroxidation, specifically 4-HNE, have been recently found to be significantly elevated in FRDA patient cells, where they not only serve as a well-established hallmark of ferroptosis but also actively trigger neuro-inflammatory responses via the TLR4/NF-kB axis^[127]. Transitioning from these cellular findings to the robust identification of these circulating oxidized byproducts (such as 4-HNE, specific epoxy-derivatives, or altered ceramide ratios) in the bloodstream will solve two problems at once: providing a reliable, non-invasive tool to monitor FRDA progression and offering definitive proof of target engagement when testing new anti-ferroptotic drugs.

5.2 Overcoming translational hurdles

Historically, antioxidant and Nrf2-based therapies in FRDA have struggled to reach the clinic. Translating novel anti-ferroptotic drugs successfully requires a rigorous reassessment of why those past efforts failed. The primary culprit is model selection. Preclinical rodent models frequently fail to capture the true human disease timeline or systemic severity. The KIKO mouse serves as a prime example: without severe cardiac iron deposits or clear coordination defects, it generates misleading efficacy data that crashes once human trials begin^[128]. Beyond the choice of model, therapeutic timing remains a critical flaw. Pre-symptomatic neurodevelopmental issues and early Nrf2 deficits are now well documented. Delaying ferroptosis-targeted therapies until late symptomatic stages severely limits their potential efficacy^[128]. Overcoming these barriers depends entirely on fixing these two parameters. Translational research must rigorously transition toward pre-symptomatic pharmacological interventions using advanced humanized transgenic models, such as the YG8sR and the more recently developed YG8-800 lines^[129]. Unlike simple murine knockouts, these humanized models are generated using bacterial artificial chromosomes to harbor the entire human FXN gene, including its native promoter elements and pathologically expanded GAA repeats (over 800 repeats in the latter model). Consequently, they accurately recapitulate the human epigenetic silencing mechanisms (heterochromatinization) and present a much more clinically relevant, progressive, multisystemic phenotype.

5.3 Investigating GSH-independent defenses and next-generation therapeutics

While the collapse of the GSH/GPx4 axis is now well-documented in FRDA, the precise functional status of GSH-independent defense networks remains underexplored. Intensive investigation is required to determine how profound FXN deficiency alters the expression and specific activity of FSP1 and DHODH in critical target tissues. Given the extensive structural damage to the respiratory chain complexes (I-III) characteristic of FRDA, mapping the functional viability of DHODH, which relies directly on CoQ pool dynamics, is particularly urgent^[89,91].

Understanding these compensatory pathways provides the exact rationale for the next wave of pharmacological development. Future drug development efforts are pivoting toward synthesizing next-generation RTAs with enhanced bioavailability and CNS penetrance. Furthermore, moving beyond direct radical trapping, the targeted modulation of these GSH-independent rescue pathways holds theoretical potential. Specifically, developing small-molecule enhancers of FSP1 to augment endogenous CoQ10 regeneration at the plasma membrane offers a sophisticated strategy to bypass the impaired Nrf2-GPx4 axis. While enhancing mitochondrial DHODH presents a similarly attractive concept, the pursuit of DHODH modulators must be approached with caution, given the aforementioned debate regarding its context-dependent efficacy as a ferroptosis suppressor and the pre-existing respiratory chain defects in FRDA^[89,91].

5.4 Combinatorial therapeutics: The ultimate strategy

Blocking lipid peroxidation clearly holds great therapeutic promise. Yet, targeting ferroptosis alone only treats the late stages of FXN deficiency, leaving the root genetic cause untouched. The best path forward for FRDA relies on combinatorial “protect and restore” strategies (Figure 4^[130]). A fascinating two-way link between Nrf2 and FXN makes this approach even stronger. Preclinical studies utilizing FXN-silenced motor neurons and patient-derived cells have demonstrated that the FXN gene promoter contains highly conserved AREs, meaning Nrf2 directly drives FXN transcription in vitro^{[128,131,132]}. While these cell-level findings are highly promising, they require further rigorous in vivo animal validation to be considered clinically established facts. Activating Nrf2 thus offers a dual benefit: it strengthens anti-ferroptotic defenses while actively supporting therapies designed to raise native FXN levels. Combining potent antioxidants with definitive genetic treatments, such as adeno-associated virus-mediated FXN gene replacement therapy, CRISPR-based epigenetic editing, or FXN protein restoration, could produce a powerful combined effect^[133]. In this setup, anti-ferroptotic drugs essentially pause the disease. By keeping cellular lipids stable and halting ongoing neurodegeneration, they save post-mitotic cells from irreversible death. Preserving this tissue is crucial; it ensures there are enough healthy cells left for the genetic therapies to actually work. Understanding the links between iron dyshomeostasis, lipid remodeling, and RCD does more than just explain a complex disease. It actively paves the way for a new era of precision medicine in FRDA.

Display Full Size

Download

Figure 4. The combinatorial “protect and restore” therapeutic strategy. A schematic demonstrating the synergistic rationale for future FRDA treatments. Strategy 1 (Protect) utilizes ferroptosis modulators, such as omaveloxolone or RTAs, to halt lipid peroxidation and prevent the irreversible degeneration of post-mitotic cells. Strategy 2 (Restore) employs gene or protein replacement therapies to address the genetic root cause and restore native frataxin levels. The convergence of these two parallel approaches could provide a comprehensive disease-modifying synergy, ensuring the morphological preservation of the tissue while metabolic homeostasis is fundamentally repaired. Created in BioRender. Cozza, G. (2026) https://BioRender.com/9m8dus3. FRDA: Friedreich’s ataxia; RTAs: radical-trapping antioxidants.

Acknowledgements

The authors would like to thank Prof. Antonella Russo (Department of Molecular Medicine - DMM, University of Padua) for her precious insights and scientific support.

Authors contribution

Cravin G: Writing-original draft, writing-review & editing.

Cozza G: Conceptualization, supervision, writing-original draft, writing-review & editing.

Conflicts of interest

The authors declare no conflicts of interest.

Ethical approval

Not applicable.

Consent to participate

Not applicable.

Consent for publication

Not applicable.

Availability of data and materials

Not applicable.

Funding

This work was supported by Associazione “OGNI GIORNO” - per Emma ONLUS and Associazione Per Il Sorriso di Ilaria di Montebruno - ONLUS (Grant No. COZZ_PRIV_CONTRIBUTI25_02).

Copyright

© The Author(s) 2026.