The enrichment of PUFAs in membrane PLs provides the necessary substrates for lipid peroxidation (22,31) (Fig. 1). Central to this process is acyl-CoA synthetase long-chain family member 4 (ACSL4), which is key to PUFA activation. In collaboration with lysophosphatidylcholine acyltransferase 3, ACSL4 esterifies FAs into membrane PLs, thereby increasing membrane susceptibility to peroxidation. ACSL4 activity is a well-recognized determinant of ferroptosis sensitivity and its inhibition confers substantial protection, making it a potential preclinical target for MASLD therapy (32-35). This pro-ferroptotic program is amplified by the protein kinase C βII-ACSL4 axis (36). Beyond these direct regulators, recent genome-wide clustered regularly interspaced short palindromic repeats screening has identified the TATA box binding protein associated factor RNA polymerase I subunit C (TAF1C) as an upstream epigenetic determinant of ferroptosis in MASLD (37). TAF1C is upregulated during disease progression and directly interacts with SET domain-containing 1A to remodel histone methylation and acetylation, thereby enhancing ACSL4 expression. The resulting increase in ACSL4 promotes lipid peroxidation and sensitizes hepatocytes to ferroptosis (37). Conversely, incorporation of monounsaturated FAs (MUFAs) into PLs shifts the lipid composition toward ferroptosis resistance. This protective lipid remodeling is mediated by acyl-CoA synthetase long-chain family member 3 (ACSL3), which preferentially incorporates MUFAs into PLs and decreases susceptibility to lipid peroxidation (38,39). Membrane-bound O-acyltransferase domain-containing proteins 1 and 2 are key regulators of PL remodeling and modulate ferroptosis by increasing MUFA-PL while decreasing PUFA-PL levels (40). These findings have direct therapeutic implications, as non-specific manipulation of lipid metabolism may produce unintended effects. Future studies should include comprehensive lipidomic profiling to define the specific PL patterns that determine ferroptosis susceptibility in the liver in MASLD.

Iron homeostasis comprises three interconnected compartments: Systemic/intestinal handling, macrophage/Kupffer cell (KC) recycling and hepatocyte iron homeostasis (27). At the systemic level, dietary iron is absorbed by enterocytes and exported into the circulation via ferroportin (FPN). Hepcidin binds FPN and promotes its internalization and degradation, thereby decreasing both intestinal iron absorption and iron release from macrophages and hepatocytes (41,42). The hepatic hepcidin/intestinal hypoxia-inducible factor-2α pathway modulates iron absorption in response to systemic iron status (43). Hepcidin transcription is regulated by bone morphogenetic protein (BMP)/SMAD signaling in hepatocytes, with BMP6 serving as the key iron-responsive ligand and transmembrane serine protease 6 serving as a negative regulator (41,44-46). In inflammatory states, IL-6 activates STAT3, which directly stimulates hepcidin transcription (47,48). Elevated hepcidin promotes FPN degradation, thereby trapping iron within splenic and hepatic macrophages, duodenal enterocytes and hepatocytes (49). In MASLD, hepcidin levels are typically low relative to hepatic iron stores (50), although inflammatory signals may simultaneously stimulate hepcidin production. The relative contribution of hepcidin suppression vs. inflammation-driven hepcidin elevation may vary across disease stages and patient subgroups. For example, hepcidin suppression may predominate in earlier or less inflammatory disease, whereas inflammation-driven hepcidin induction may become more prominent in advanced disease and in patients with a higher inflammatory burden. Beyond the liver and intestine, the kidney contributes to systemic iron homeostasis by reabsorbing filtered iron and clearing hepcidin (51), although its specific role in MASLD remains to be defined.

KCs, liver-resident macrophages, recycle iron from senescent erythrocytes and export it back into the circulation via FPN. Under conditions of low hepcidin, this export is enhanced, contributing to the circulating iron pool. Conversely, when hepcidin is elevated, FPN is degraded and iron is retained within macrophages (49). In MASLD, dysregulated iron handling leads to iron accumulation in KCs. Iron-rich KCs exhibit phenotypical changes toward a pro-inflammatory and pro-fibrotic state during MASH progression (52). A recent study (53) identified neutrophil cytosolic factor (NCF)1 as a key regulator. Its upregulation in liver macrophages causes iron overload and ferroptosis in KCs. NCF1-induced oxidized PLs activate toll-like receptor (TLR)4-dependent hepatocyte hepcidin production, which increases iron deposition in KCs, forming a positive feedback loop that amplifies ferroptosis and MASH (53). Notably, not all hepatic macrophages experience iron overload in MASLD. Embryonic KCs exhibit decreased labile iron pools (LIPs) and mitochondrial dysfunction in MASH and KC-specific FPN knockout aggravates MASH (54). This highlights the heterogeneity of iron handling across macrophage populations.

Hepatocytes take up transferrin-bound iron via transferrin receptor 1. Within acidified endosomes, six transmembrane epithelial antigen of prostate 3 converts Fe³⁺ to Fe²⁺, which is transported into the cytosol by solute carrier family 11 member 2 (SLC11A2) (55,56). Under iron overload conditions, uptake of non-transferrin-bound iron via SLC39A14 increases (57). Intracellularly, excess iron is sequestered in ferritin, which serves both as a reservoir and as a protective buffer against the Fenton reaction (58). When iron demand increases or oxidative stress rises, ferritin is selectively degraded through nuclear receptor coactivator 4-dependent ferritinophagy, releasing Fe²⁺ into the LIP (59,60). Expansion of the LIP facilitates ferroptosis by supplying the redox-active iron required for the Fenton reaction and subsequent lipid peroxidation (61-63). The mitochondrial LIP acts as a major catalyst of Fenton reactions, generating reactive oxygen species (ROS) that trigger and sustain lipid peroxidation, thereby directly promoting ferroptosis (64-66). Iron efflux from hepatocytes is mediated by FPN, which is downregulated by hepcidin (41,42). These alterations in iron handling manifest as elevated serum ferritin concentrations and increased hepatic iron levels (67-69). Notably, a recent study (70) demonstrated that decreased FPN expression in hepatocytes leads to iron accumulation and systemic metabolic disturbance, identifying fetuin-A and leukocyte cell-derived chemotaxin 2 as key hepatokines. These findings are crucial for elucidating the association between MASLD and metabolic disorders and suggest therapeutic strategies such as hepatokine silencing, FPN overexpression or iron chelation (70). In addition, iron serves as a potent pro-oxidant, and its accumulation in the liver increases ROS production, lipid peroxidation and subsequent cell damage (71,72). The resulting oxidative stress and inflammation exacerbate the progression of liver fibrosis and cirrhosis, underscoring the key role of iron metabolism in MASLD pathogenesis (73).

Ferroptosis is controlled by antioxidant pathways, including the GSH-GPX4 axis, the FSP1/CoQ10 pathway and the GCH1-BH4 axis. These pathways are key for maintaining cell redox homeostasis and preventing the accumulation of harmful lipid peroxides (74). Central to the GSH-GPX4 axis is the cystine/glutamate antiporter SLC7A11, which imports cystine, a key component for synthesizing GSH (75). GPX4 uses GSH to transform PL hydroperoxides into non-toxic lipid alcohols, thereby preventing ferroptosis (76). However, in MASLD, GSH depletion and GPX4 inactivation are typically observed (31). In parallel, FSP1 reduces CoQ10 to ubiquinol at the plasma membrane and other compartments, trapping lipid radicals independently of GSH (77,78). The GCH-BH4 axis also serves a key role in ferroptosis regulation. GCH1 and its metabolic derivative BH4 are involved in lipid remodeling that maintains the pool of PLs containing PUFA tails, counteracting their loss during ferroptosis and thereby conferring resistance to ferroptosis (79). NRF2 is key for orchestrating the antioxidant response and preventing ferroptosis. Upon activation, NRF2 upregulates genes associated with GSH synthesis and iron metabolism such as SLC7A11 and GPX4, protecting against oxidative stress and ferroptosis (80,81).

Mitochondrial dysfunction is a key feature of MASLD. Iron overload causes mitochondrial damage, characterized by increased ROS production and impaired mitochondrial respiration, which may exacerbate liver injury (82,83). Mitochondria are also involved in controlling ferroptosis. Mitochondrial ROS, particularly those derived from the electron transport chain (ETC), are key contributors to lipid peroxidation and ferroptosis (84,85), which are associated with mitochondrial metabolism. The tricarboxylic acid cycle and ETC maintain mitochondrial membrane potential and redox balance, which are essential for ferroptosis (86). Mitochondria also employ a key defense network against ferroptosis. Central to this system is the transport of GSH into the mitochondrial matrix, mediated by specific carrier proteins such as the dicarboxylate carrier and the oxoglutarate carrier. This import is required for the proper functioning of mitochondrial GPX4. Blocking these carriers increases ferroptosis by elevating mitochondrial ROS and depleting GSH (87). In addition, dihydroorotate dehydrogenase operates within the inner mitochondrial membrane, where it catalyzes the reduction of ubiquinone to ubiquinol, which directly mitigates lipid peroxidation independently of the GSH-GPX4 axis (88,89).

Collectively, the metabolic environment in MASLD is characterized by a lipid composition that permits ferroptosis, excess active iron and weakened antioxidant defenses. These factors increase the likelihood of ferroptosis (Fig. 1).

In MASLD, hepatocyte ferroptosis is a key contributor to liver inflammation and fibrosis (15,83,90). Studies highlight its role in amplifying oxidative stress and activating fibrogenic pathways (15,83,90). In particular, the accumulation of free FAs (FFAs) in hepatocytes is a hallmark of MASLD. This accumulation triggers lipid peroxidation and ferroptosis, thereby intensifying liver damage (91) (Fig. 2). Notably, another study demonstrated the protective role of activating transcription factor 4 in normal hepatocytes, where it maintains GSH production and inhibits ferroptosis-dependent inflammatory cell death (92).

Emerging evidence (93) suggests ferroptotic hepatocytes communicate with HSCs through exosome-mediated signaling. A recent study (93) showed that dietary cholesterol-induced downregulation of PPARδ sensitizes hepatocytes to ferroptosis by modulating SLC7A11. Notably, exosomal double-stranded DNA (dsDNA) released from ferroptotic hepatocytes activates the stimulator of interferon genes (STING)/TANK-binding kinase 1/interferon regulatory factor 3 (IRF3) signaling pathway in HSCs, promoting the expression of profibrotic genes (93). This intercellular communication between hepatocytes and HSCs is reversed either by hepatocyte-specific overexpression of PPARδ or by treatment with the selective PPARδ agonist DN203316 (93). These findings underscore the role of PPARδ as a key regulator of cholesterol-induced ferroptosis and highlight exosomal dsDNA/STING signaling as a key mediator linking hepatocyte ferroptosis with HSC-driven fibrogenesis.

In MASLD, decreased GPX4 levels are associated with increased vulnerability to ferroptosis, consequently enhancing hepatic oxidative stress and promoting fibrogenesis (15). Moreover, the vulnerability of hepatocytes to ferroptosis is modulated by immune-derived signals. Under intermittent hypoxia, a prevalent comorbidity in MASLD, IL-6 produced by macrophages induces upregulation of the E3 ubiquitin ligase membrane-associated RING finger protein 3 (MARCH3) in hepatocytes. MARCH3 facilitates ubiquitination and degradation of GPX4, increasing hepatocyte susceptibility to ferroptosis (94). Similarly, in MASH with T2DM, advanced glycation end product receptor 1 (AGER1) deficiency initiates hepatocyte ferroptosis and promotes epithelial-mesenchymal transition, thereby promoting fibrogenesis. Conversely, restoring AGER1 mitigates these effects (95). Collectively, hepatocyte ferroptosis serves a key role in the development of MASLD as it promotes oxidative stress, inflammation and fibrogenesis, thereby contributing to disease progression and complications.

HSCs are the primary cause of liver fibrosis, as their transdifferentiation into myofibroblasts leads to excessive extracellular matrix production (96). The role of ferroptosis in HSCs is context-dependent, with evidence pointing to opposing outcomes (97,98). Ferroptosis in HSCs can inhibit HSC activation and reduce liver fibrosis. On the other hand, ferroptosis-related programs in HSCs can exacerbate fibrogenesis. Quercetagetin (QG), the active compound in the Tibetan medicine Huagan Tongluo Fang, induces ferroptosis in HSCs by promoting ubiquitin-proteasome-mediated degradation of GPX4. This is mediated by MARCHF8 and ultimately prevents HSC activation and reduces liver fibrosis (97). Conversely, other ferroptosis-related programs in HSCs promote fibrogenesis. Han et al (98) showed that acetyl-CoA acyltransferase 2 (ACAA2), a mitochondrial enzyme involved in FA β-oxidation, is markedly upregulated in HSCs during MASLD-associated fibrosis. ACAA2 undergoes palmitoylation, a modification that controls its subcellular localization and enhances ferroptosis-associated transcriptional programs, thereby exacerbating hepatic fibrosis. Inhibiting ACAA2 or blocking its palmitoylation attenuates both ferroptosis and fibrogenesis, an effect associated with AMP-activated protein kinase (AMPK) pathway activation and suppression of HSC activation (98). Rather than merely eliminating HSCs, ACAA2-driven ferroptosis triggers a pro-inflammatory and pro-fibrotic cascade, or occurs within a metabolic state that that enhances fibrogenic signaling following cellular death. These divergent outcomes may be determined by several contextual factors, including the HSC activation state, disease stage, experimental model and local microenvironment. These observations demonstrate that inducing HSC ferroptosis cannot yet be regarded as a generally applicable antifibrotic strategy until the biological determinants governing its net tissue effect are defined.

KCs are liver-resident macrophages that serve a key role in maintaining hepatic homeostasis and responding to liver injury (99). In MASLD, iron-rich KCs exhibit alterations associated with inflammation and fibrosis, such as upregulated CD11c and proinflammatory and profibrotic genes (Tnfa, Ccl3, and Timp1) and activation of MiT/TFE transcription factors (52). NCF1 is elevated in liver macrophages and dendritic cells (DCs) in patients with MASLD and in MASH mice. This upregulation promotes iron overload and ferroptosis in KCs, exacerbating MASH progression (53). Oxidized PLs released by macrophages activate TLR4-dependent production of hepcidin in hepatocytes, which in turn increases iron accumulation in KCs and initiates ferroptosis (53). These findings suggest targeting NCF1 may improve KC function and limit MASH progression.

KCs exhibit abnormal lipid metabolism in MASLD, leading to the accumulation of toxic lipids and a pro-inflammatory state (100). Lipid-induced stress promotes ferroptosis and further contributes to liver injury. Importantly, induction of transcription factor EB in KCs protects against ferroptosis by enhancing lysosomal lipolysis and mitochondrial FA oxidation, decreasing oxidative stress and maintaining KC function (101). This highlights the potential of targeting lipid metabolism in KCs to prevent ferroptosis in MASLD.

Beyond hepatocytes, HSCs and KCs, ferroptosis may also influence other liver cell populations; for example, accumulation of PUFAs in mucosal-associated invariant T (MAIT) cells causes ferroptosis-driven metabolic exhaustion, impairing their anti-tumor surveillance and accelerating disease progression (22). Similarly, oxidized low-density lipoprotein induces ferroptosis in double-negative regulatory T cells via CD36-mediated lipid uptake, decreasing their survival and immunosuppressive capacity and thereby exacerbating liver inflammation (102).

DCs are the main antigen-presenting cells and are particularly susceptible to ferroptosis in MASLD. Li et al (103) demonstrated that, in MASLD-related HCC, a lipid-rich environment promotes lipid peroxidation and ferroptosis in DCs via a T cell immunoglobulin and mucin domain-containing protein 3 (Tim-3)-dependent pathway. This depletes DCs, impairs CD8⁺ T cell response and facilitates tumor growth. Genetic or pharmacological inhibition of Tim-3 decreases lipid peroxidation, improves DC survival, reactivates CD8⁺ T cells and slows tumor growth. Tim-3 blockade enhances the efficacy of anti-PD-1 therapy and lenvatinib, achieving sustained tumor control in preclinical models (103,104). These findings identify Tim-3 as a key regulator of DC ferroptosis in metabolic liver disease.

A disease stage-specific framework may clarify when ferroptosis should be inhibited or induced. In early MASLD and in MASH with fibrosis, ferroptosis in hepatocytes and KCs is detrimental and should be inhibited. By contrast, selective induction of ferroptosis in activated HSCs may promote fibrosis resolution, although this strategy requires caution because HSC ferroptosis can be pro-fibrotic depending on context. During HCC progression, ferroptosis in immune cells such as DCs and MAIT and regulatory T cells impairs anti-tumor immunity. At this stage, protecting these immune cells or inducing ferroptosis specifically in malignant hepatocytes may be preferable. The consequences of ferroptosis vary by cell type and disease stage; therapeutic strategies should be designed accordingly.

Recent studies have increasingly focused on inhibiting ferroptosis in hepatocytes as a strategy to decrease liver injury in MASLD (15,26,67). Several natural products and herbal formulations attenuate ferroptosis in MASLD models (112-114), although the strength of evidence varies (Fig. 3). Direct evidence of ferroptosis has been provided for a limited number of compounds, including betaine, plant sterol ester of α-linolenic acid, protocatechuic acid and arbutin. These compounds decrease erastin- or RAS-selective lethal 3-induced ferroptosis and modulate NRF2/GPX4 or fat mass and obesity-associated protein/SLC7A11 signaling in hepatocytes (115-118). By contrast, other natural products and herbal formulations including Hu Gan Tang, diosgenin, puerarin and gastrodin have primarily been shown to attenuate ferroptosis-associated molecular changes (112,119-121). For these compounds, the evidence supports an association with ferroptosis-related pathways rather than establishing ferroptosis as the causal mechanism. A common limitation of most natural products is the lack of information on oral bioavailability, tissue distribution and long-term safety, which hinders clinical translation. Additionally, many studies use the methionine-choline-deficient (or high-fat diet models that do not fully recapitulate human MASLD (Table II) (113-114,122-127).

Clinically approved drugs with anti-ferroptotic properties have translation potential. Liraglutide improves MASLD in mouse models (128,129) by activating the AMPK/acetyl-CoA carboxylase (ACC) pathway and attenuating ferroptosis-associated markers (129). Similarly, metformin, a first-line antidiabetic agent, mitigates hepatic iron overload via the AMPK/FPN pathway, reducing iron availability for Fenton reactions and lipid peroxidation (130). To the best of our knowledge, however, their anti-ferroptotic effects in MASLD have been demonstrated only in animal models, and clinical validation is lacking.

Tao et al (67) identified ferroterminator 1 (FOT1), a novel iron chelator, as an effective treatment for MASH. Analysis of clinical cohort and multiple mouse models (67,131) shows that hepatic iron overload contributes to MASH progression via the c-Myc/ACSL4 pathway and FOT1 ameliorates MASH in preclinical settings. Although serum ferritin predicted the response to FOT1 in clinical cohorts, FOT1 has not yet entered clinical trials for MASH (67). Liproxstatin 1 exerts hepatoprotective effects in a murine MASLD model, decreasing markers of steatosis, inflammation and fibrosis (28). Collectively, these findings support iron chelation as a promising strategy for ferroptosis-targeted therapy in MASLD, although clinical validation remains necessary. These direct ferroptosis-targeting agents may be particularly relevant for patients who do not achieve adequate response to approved therapies such as resmetirom or semaglutide. If serum ferritin or other iron-related markers can be used to identify a subpopulation with high ferroptosis burden, combining a ferroptosis inhibitor or iron chelator with standard MASH therapy may be explored in future trials.

Although inhibiting ferroptosis in hepatocytes is key to limiting liver injury and MASLD progression, selectively inducing ferroptosis in activated HSCs may offer a strategy to reverse fibrosis. As aforementioned, QG induces ferroptosis in HSCs by promoting ubiquitination and degradation of GPX4, thereby decreasing collagen deposition in a MASLD mouse model (97). However, the effect of HSC ferroptosis depends on multiple factors, such as HSC activation state, disease stage and the local microenvironment. In addition, most preclinical studies on HSC ferroptosis in liver fibrosis use carbon tetrachloride or thioacetamide-induced models (132-138), which do not fully recapitulate the metabolic and inflammatory context of MASLD. Achieving cell type specificity also remains a challenge, and future work should focus on HSC-targeted delivery systems. Therefore, induction of HSC ferroptosis should not be regarded as a generally applicable antifibrotic strategy. Its therapeutic potential requires validation in more disease-relevant models before clinical translation.

In a mouse model of MASH, baicalein, a traditional Chinese medicine extracted from Scutellaria baicalensis, improves metabolic dysfunction and slows disease progression. Baicalein restores hepatic iron balance and decreases iron overload and lipid peroxidation (139,140). Its anti-ferroptotic effects have been confirmed in Raw264.7 macrophages exposed to FFAs (140). Mechanically, baicalein suppresses M1 macrophage polarization and promotes M2 polarization by activating the NRF2/FSP1 signaling pathway, thereby decreasing ferroptosis (140). Similarly, acacetin, a natural flavonoid, alleviates MASLD through mechanisms involving two distinct cell types: In hepatocytes, acacetin reduces Notch1-mediated lipogenesis and necroptosis; in macrophages, it blocks ferroptosis through the Keap1/NRF2 signaling pathway and promotes M2 polarization (141). These findings suggest targeting macrophage ferroptosis may represent a potential therapeutic strategy for MASLD.

The gut microbiome serves a key role in the onset and progression of MASLD and its complications. Changes in microbiome composition are associated with disease progression and interventions targeting the microbiome can influence disease outcomes through several mechanisms, such as increasing gut permeability, decreasing levels of endotoxin-producing bacteria and modulating bile acid metabolism (142-146).

Recent studies (147-149) suggest that targeting the gut microbiome and its associated ferroptosis pathways may offer a therapeutic strategy for MASLD; for example, Camellia japonica radix extract modulates gut microbiota and linoleic acid metabolism, specifically targeting the pro-ferroptotic metabolite 9(S)-hydroperoxy-10E,12Z-octadecadienoic acid. This metabolite exerts a dual effect on Keap1 and SLC7A11, thereby regulating GPX4-mediated ferroptosis-associated signaling (147). Similarly, Tricholoma mongolicum polysaccharides alter gut microbiota composition and promote liver health (148). In addition, modulation of the gut microbiota by ginsenoside Rd (G-Rd) contributes to its protective effects against MASLD. G-Rd decreases the relative abundance of Firmicutes and the Firmicutes/Bacteroidetes ratio, both of which are typically elevated in MASLD. Furthermore, G-Rd lowers the abundance of Faecalibaculum rodentium while increasing that of Muribaculum intestinale. These alterations in gut microbiota composition regulate lipid metabolism and alleviate oxidative stress in the liver (149). The findings support an association with ferroptosis-associated pathways rather than establishing ferroptosis as the causal mechanism.

Metabolites from certain gut microbial strains directly or indirectly inhibit hepatic ferroptosis. For example, Zhang et al (150) demonstrated that hexadecanedioic acid, a metabolite produced by Bacteroides uniformis, is transported to the liver via the portal vein, where it inhibits the inositol-requiring enzyme 1α/spliced X-box binding protein 1 signaling pathway. This inhibition alleviates liver fibrosis and suppresses ferroptosis-associated molecular changes (150). The aforementioned study provides novel insight into the links between gut microbiota, endoplasmic reticulum stress and hepatocyte ferroptosis. In addition, Lactobacillus plantarum 1-2-3 has beneficial effects on lipid metabolism and inflammation and protects against MASLD by modulating gut microbiota and metabolic pathways (151). Furthermore, acetate produced by Akkermansia muciniphila activates the hepatic AMPK/SIRT1/PPAR gamma coactivator 1 alpha (PGC-1α) signaling pathway, thereby suppressing ferroptosis-associated signaling and ameliorating MASLD (152).

Obeticholic acid is promising therapeutic candidate in MASH (153). However, its antifibrotic efficacy is decreased by microbiota-induced lipid peroxidation. This inhibitory effect largely originates from Bacteroides, which deconjugates tauro-conjugated bile acids, leading to aberrant production of chenodeoxycholic acid and subsequent accumulation of ROS in the liver (154). Collectively, these findings indicate that MASLD arises not only from hepatic dysfunction but also from gut and systemic metabolic perturbation, underscoring the importance of targeting the gut-liver axis.