Targeting Ferroptosis to Overcome Drug Resistance in Cancer: Molecular Mechanisms and Therapeutic Prospects

Abstract

Drug resistance in cancer cells remains a major obstacle limiting the clinical efficacy of current anticancer therapies. The induction of ferroptosis, an iron-dependent, regulated form of cell death, may offer an alternative therapeutic strategy to overcome such resistance. The generation of reactive oxygen species (ROS) has been implicated in this process, and depending on the cellular context, ROS can be either detrimental or beneficial. Ferroptosis can be effectively triggered in drug-resistant cancer cells in which ROS levels are often highly elevated. Key signaling pathways, including receptor tyrosine kinase (RTK), mitogen-activated protein kinase (MAPK), and nuclear factor erythroid 2-related factor 2 (NRF2), are promising targets for modulating ROS homeostasis and sensitizing cancer cells to ferroptosis. In this review, we discuss the molecular mechanisms governing ferroptosis, the interplay between ROS and ferroptosis resistance, and emerging therapeutic approaches designed to enhance ferroptosis induction in drug-resistant cancer cells. Altogether, a combination of ferroptosis inducers and conventional treatments may improve the therapeutic efficacy and help overcome resistance mechanisms.

INTRODUCTION

Cancer remains a major global health burden and causes a substantial number of deaths worldwide (Siegel et al., 2024). Despite the advent of targeted therapies, therapeutic resistance frequently occurs, posing a significant challenge to effective cancer treatment (Lei et al., 2023). Ferroptosis is an iron-dependent form of regulated cell death characterized by lipid peroxidation and oxidative stress (Fig. 1) (Dixon et al., 2012). Although ferroptosis may not completely account for chemotherapeutic agents-induced cell death, targeting this pathway could provide an alternative therapeutic strategy to overcome drug resistance in cancer. Drug-resistant cancer cells often exhibit elevated levels of reactive oxygen species (ROS) (Shah and Rogoff, 2021; Viswanathan et al., 2017), rendering them particularly vulnerable to ferroptosis. In this review, we discuss the molecular mechanisms underlying ferroptosis, the interplay between ROS and ferroptosis resistance, and emerging therapeutic approaches to sensitize drug-resistant cancer cells through ferroptosis activation.

Fig. 1.

Schematic diagram of ferroptosis. Ferroptosis occurs when the lipid peroxidation of cellular membranes leads to structural disruptions. Intracellular iron levels are regulated by the balance between the iron transporter transferrin receptor (from outside to inside) and the iron exporter ferroportin. Reactive oxygen species (ROS), generated from metabolism and the Fenton reaction of excessive iron, promote lipid peroxidation. Conversely, glutathione peroxidase 4 (GPX4) eliminates ROS and prevents lipid peroxidation. The supply of glutathione (GSH) and cystine via the system Xc^- antiporter helps to protect cells from ferroptosis.

SHORT OVERVIEW

Molecular mechanisms underlying ferroptosis

Ferroptosis is a type of cell deaths (the suffix “-ptosis” in ferroptosis), characterized by the involvement of ferrous iron (the prefix “ferro-“ in ferroptosis) (Dixon et al., 2012). Several compounds have been demonstrated to induce ferroptosis (Table 1). Iron is involved in the regulation of cell death processes, in addition to its multiple essential roles in cell physiology in maintaining and regulating mitochondrial respiration, oxygen transport and storage, and DNA synthesis and repair (Theil and Goss, 2009). Iron is involved in cell death, particularly apoptosis and ferroptosis. In cells, ROS generation is influenced by intracellular iron levels, which are tightly regulated by several proteins, including transferrin, ferritin, and ferroportin (Dixon and Stockwell, 2014). High levels of iron, beyond a critical threshold, cause oxidative stress and promote apoptosis. The Fenton reaction is responsible for generating hydroxyl radicals (⋅OH) and ferric iron (Fe³⁺) from the reaction between ferrous iron (Fe²⁺) and hydrogen peroxide (H₂O₂). Hydroxyl radicals are extremely reactive and cause significant damage to several cellular molecules. Although the function of iron in apoptosis is not critical, iron overload appears to be an essential and defining event in ferroptosis. Iron overload and oxidative stress, particularly lipid peroxidation, significantly contribute to ferroptosis progression. Ferroptosis can occur independently or concurrently with other types of cell death if the cell exhibits characteristic features of ferroptosis.

Table 1.

Induction of ferroptosis by selected compounds

Compound	Mechanism
Artesunate	Increases iron level by promoting ferritin degradation by ferritinophagy
Carnosic acid	Inactivates NRF2
Trigonelline	Inhibits NRF2
Erastin	Inhibits system Xc-
Sulfasalazine	Inhibits system Xc-
Sorafenib	Inhibits system Xc-
Vorinostat	Downregulates transporter xCT
FIN56	GPX4 degradation
RSL3	Inhibits GPX4
β-elemene	Downregulates GPX4

Iron level regulation: Cellular iron levels are primarily regulated by transferrin, ferritin, and ferroportin. Transferrin functions as a carrier of iron (Fe³⁺) in the extracellular space. The complex formed between transferrin and transferrin receptor 1 (TfR1) on the cell surface mediates the internalization of transferrin along with its iron cargo (Fe³⁺) (MacKenzie et al., 2008). Cellular iron is reduced from the ferric (Fe³⁺) to the ferrous (Fe²⁺) form and subsequently released into the cytoplasm. Instead of being reduced to the ferrous form, iron can be stored in the mitochondria or cytoplasm, which safely stores iron in the ferric form. Although the overall storage capacity is greater for cytoplasmic ferritin, mitochondrial ferritin (FtMt) is more critical in protecting against oxidative stress, as mitochondria are abundant in H₂O₂, which can react with iron (Campanella et al., 2009). The safely stored iron may be mobilized through ferritinophagy, a selective autophagic pathway that degrades ferritin and releases iron (Fe²⁺) (Wang et al., 2023a). Ferroportin, encoded by SLC40A1, functions as an iron exporter in mammalian cells, and is the only known protein that performs this function (Camaschella et al., 2002). It regulates iron homeostasis by facilitating the efflux of iron from the cells. Hepcidin, a peptide hormone, specifically binds to ferroportin, triggering its internalization and degradation, and leading to increased cellular iron levels (De Domenico et al., 2011).

Cystine/glutamate antiporter, the system Xc-: The cystine/glutamate antiporter, often known as system Xc^- functions as a supply of cysteine to the cytoplasm, as extracellular L-cystine and intracellular L-glutamate are exchanged through this antiporter system (Piani and Fontana, 1994). Once imported, L-cystine, a dimer of L-cysteine linked by a disulfide bond, is reduced within the cells to yield two cysteine molecules that are used for synthesis of proteins and glutathione (GSH), a tripeptide antioxidant (Burdo et al., 2006). GSH is responsible for one of the most significant antioxidant mechanisms in the cell, and low cellular cysteine levels due to the dysregulation of system Xc^- may cause depletion of intracellular GSH, eventually resulting in ferroptotic cell death (Liu et al., 2021). The transcription of the gene encoding SLC7A11, a component of the system Xc^-, is suppressed by p53, thereby inducing ferroptosis (Jiang et al., 2015). Erastin, a ferroptosis inducer, is an inhibitor of system Xc^- which blocks cystine uptake into cells, eventually leading to ferroptosis (Dixon et al., 2012). In addition, sulfasalazine (Roh et al., 2016) and sorafenib (Louandre et al., 2013) are well-known inducers of the ferroptosis blocking system Xc^-, whereas vorinostat downregulates xCT ((Miyamoto et al., 2020).

Lipid peroxidation: Lipid peroxidation is a hallmark of ferroptosis and is largely driven by elevated iron levels and the depletion of the antioxidant GSH. Elevated intracellular GSH levels support the activity of glutathione peroxidase 4 (GPX4), which detoxifies lipid hydroperoxides and prevents lipid peroxidation triggering ferroptosis (Feng et al., 2021; Koppula et al., 2021; Torrence et al., 2021). Downregulation of GPX4 levels can induce ferroptosis, as demonstrated by β-elemene, which reduces GPX4 expression potentially through modulation of degradation pathways (He et al., 2024). The supply of GSH depends on the system Xc^-, whereas GPX4 consumes GSH to remove lipid peroxidation. However, depletion of GSH or the inadequate function of GPX4 damages and ruptures the plasma membrane. Therefore, both system Xc^- and GPX4 are crucial regulatory factors that govern lipid peroxidation and ferroptosis (Fig. 2). Several key enzymes are involved with enzymatic lipid peroxidation. Acyl-CoA synthetase long-chain family member 4 (ACSL4) activates long-chain PUFAs by attaching to coenzyme A, thus promoting their incorporation into membrane phospholipids. ACSL4 preferentially supplies and promotes the availability of long-chain PUFAs such as arachidonic acid for peroxidation, making cells susceptible to ferroptotic death (Doll et al., 2017; Kagan et al., 2017). Lysophosphatidyl CoA acyltransferase 3 (LPCAT3) catalyzes the reacylation of lysophospholipids with CoA-attached fatty acids, including arachidonic acid, to generate PUFA-phospholipids. This process further enriches cell membranes with PUFAs that are susceptible to peroxidation. Increased LPCAT3 activity increases the fraction of phospholipids containing long-chain PUFAs, priming cells for ferroptosis (Dixon et al., 2015). Lipoxygenases (LOXs) directly catalyze the oxygenation of PUFAs within membrane phospholipids, triggering lipid peroxidation and membrane damage. LOXs function as principal executors of oxidative stress during ferroptosis by generating lipid hydroperoxides (Shah et al., 2018). In summary, ACSL4 and LPCAT3 work together to increase PUFA-rich phospholipid pools that are vulnerable to peroxidation, whereas LOXs catalyze their oxidation, causing membrane disruption during ferroptosis.

Fig. 2.

Key targets in inducing ferroptosis. The supply of cystine from outside the cell via system Xc^- and the removal of lipid peroxides by GPX4 are two critical targets inducing ferroptosis. NRF2 regulates ferroptosis by controlling several key genes involved in this process, including SLC7A11. glutathione peroxidase 4 (GPX); Nuclear factor erythroid 2-related factor 2 (NRF2); Soluble carrier family 7 member 11 (SLC7A11).

MAPK and ferroptosis

The mitogen-activated protein kinase (MAPK) pathway is a well-characterized signaling cascade involved in normal and pathological cellular functions (Dhillon et al., 2007). Although the direct relationship between the MAPK pathway and ferroptosis has not yet been well established (Wang et al., 2023b), the modulation of oxidative stress (Son et al., 2011) and iron metabolism (Munoz et al., 2006) is certainly connected to MAPK pathways, suggesting the involvement of these pathways in regulating ferroptosis. Multiple studies have reported a correlation between ferroptosis and the regulation of MAPK signaling. We have summarized the current understanding based on recent research (Table 2).

Table 2.

Correlation between ferroptosis and MAPK signaling

Drug/insult	Cellular target(s)	System
Ferroptosis accompanied with downregulated MAPK signaling
Heteronemin	ERK, GPX ↓	Hepatocellular carcinoma
Simvastatin	MAPK ↓	Endometrial cancer cell
Orlistat	MAPK, GPX4 expression ↓	Pancreatic neuroendocrine tumor cells
Dictamnine	ERK phosphorylation ↓	Colorectal cancer cells
β-elemene	MAPK ↓	Endometriosis epithelial cells
Jujuboside B	MAPK ↓	Colorectal cancer cells
	Musashi-2 (MSI2)	Colorectal cancer cells
	Rho GTPase-activating protein 6 (ARGHAP6)	Breast cancer mouse model
RNA interference	RON receptor tyrosine kinase (RTK)	Thyroid cancer cell
Ferroptosis accompanied with MAPK signaling activation
Theaflavin-3,3’-digallate	MAPK ↑	Osteosarcoma cells
Realgar	RAF, MAPK ↑	KRAS mutant lung cancer cells
Micheliolide	MAPK ↑	Pancreatic and colon tumor cell

Ferroptosis accompanied by downregulated MAPK signaling: We identified approximately a dozen studies reporting the relationship between ferroptosis and MAPK signaling pathways in several cancer and non-cancer models. Most studies indicated a negative correlation between ferroptosis progression and MAPK signaling activation. It is plausible that MAPK signaling plays a role in sustaining the cancer cell survival (Braicu et al., 2019). Several compounds have been reported to induce ferroptosis by modulating MAPK signaling. Heteronemin, a natural marine compound isolated from Hippospongia sp., inhibits the proliferation of hepatocellular carcinoma cells by downregulating extracellular signal-regulated kinase (ERK) and GPX4, while promoting ferroptosis (Chang et al., 2021). The downregulation of ERK suppresses the survival signaling of the MAPK pathway, whereas the downregulation of GPX4 results in oxidative stress and promotes lipid peroxidation. Similarly, simvastatin, a cholesterol-lowering drug, downregulates components of the MAPK signaling pathway and promotes ferroptosis in endometrial cancer cells (Zhou et al., 2022). In addition, orlistat, a weight loss agent, induces ferroptosis in pancreatic neuroendocrine tumor cells by inhibiting the MAPK pathways and suppressing GPX4 expression (Ye et al., 2023). The inhibition of ERK phosphorylation, an indication of MAPK signaling suppression, caused by dictamnine, results in ferroptosis in colorectal cancer (CRC) cells (Zuo et al., 2024). In addition, downregulation or inhibition of MAPK signaling is implicated in the cytotoxicity of β-elemene in endometriosis epithelial cells (Fu et al., 2025) and jujuboside B in CRC cells (Zhai et al., 2025).

Certain cellular targets are related to the downregulation of MAPK signaling and induction of ferroptosis. Musashi-2 (MSI2) is an RNA-binding protein that drives different cancers when ectopically expressed (Kharas et al., 2010). MSI2 directly interacts with p-ERK, and MSI2 deficiency results in downregulation of the MAPK signaling pathway and induces ferroptosis by modulating HSPB1 (Meng et al., 2023). Rho GTPase-activating protein 6 (ARHGAP6) correlates well with the survival of patients breast cancer (Chen et al., 2019). A mouse model study involving an ARHGAP6 knockout revealed that ARHGAP6 suppresses the p38 MAPK signaling and induces ferroptosis (Chen et al., 2024). RON is a receptor tyrosine kinase (RTK) that is highly expressed in several cancer types (Kang et al., 2014). A study using a thyroid cancer cell model with RNA interference demonstrated that the suppression of RON expression dowunregulates MAPK signaling and glycolysis, as well as induces ferroptosis (Jin et al., 2024). Although the precise relationships between MSI2, ARHGAP6, and RON in regulation of MAPK signaling and ferroptosis remain to be determined, these targets could provide alternative avenues for cancer treatment.

Ferroptosis accompanied by MAPK signaling activation: In contrast to different compounds inducing ferroptosis while suppressing MAPK signaling, cases exist in which ferroptosis occurs alongside the activation of MAPK signaling. Theaflavin-3,3’-digallate (TF3) is a natural phenol compound found in black tea with potent anticancer activity in different cancer cells (Hibasami et al., 1998). TF3 induces oxidative stress in osteosarcoma cells, triggering both apoptosis and ferroptosis, accompanied by the activation of MAPK signaling (He et al., 2022). Realgar is an arsenic compound used in traditional Chinese medicine, and several studies have reported its anticancer activity (Xiaoxia et al., 2020). A study with KRAS mutant lung cancer cells suggested that realgar induces ferroptosis by regulating RAF and MAPK signaling (Liu et al., 2022). Micheliolide, a sesquiterpene compound, exhibits anticancer properties (Viennois et al., 2014). Micheliolide activates the MAPK signaling pathway and promotes autophagy, paraptosis, and ferroptosis (Yang et al., 2024). All three ferroptosis-inducing compounds, TF3, realgar, and micheliolide, increase ROS generation, MAPK activation likely occurs as a consequence of ROS production rather than the opposite. However, further evidence is required to elucidate the relationship between MAPK activation and ferroptosis induction.

MAPK and energy metabolism-associated ferroptosis

Energy demand during cell proliferation: Cell division is intrinsically energy-intensive because each cycle requires complete genome duplication (large dNTP pools and ATP for polymerization/helicase activity), high-rate RNA transcription, ribosome biogenesis, and protein translation, all of which consume ATP/GTP at multiple steps of initiation, elongation, and protein processing (Buttgereit and Brand, 1995; Vander Heiden et al., 2009). Beyond ATP, proliferating cells require abundant carbon and reducing equivalents (e.g., NADPH) to expand biomass, namely, nucleotides, amino acids, and lipids for new membranes and organelles. The coupling of proliferation and metabolic throughput is the central principle of cancer metabolism (DeBerardinis and Chandel, 2016). Cancer cells meet these demands through aerobic glycolysis (the Warburg effect) even in the presence of oxygen. They favor glycolysis to rapidly generate ATP and, crucially, divert metabolic intermediates from glycolysis and the Krebs cycle into biosynthetic pathways for synthesizing nucleotides, non-essential amino acids, and fatty acids, which are necessary building blocks for cell division. This metabolic phenotype supports sustained proliferation despite the lower ATP yield per glucose (Liberti and Locasale, 2016; Warburg, 1956).

Oncogenic MAPK signaling couples mitogenic cues to metabolic rewiring. ERK1/2–p90RSK directly stimulates mTORC1 signaling by phosphorylating Raptor and TSC2, thereby enhancing cap-dependent translation and synthesis of metabolic enzymes required for biomass accumulation (Carriere et al., 2011). In addition, MAPK tunes glycolytic control points more proximally; for example, EGFR–ERK2 phosphorylates PKM2 (Ser37) to promote its nuclear translocation and transcriptional programs that reinforce the Warburg phenotype. Altogether, these mechanisms ensure that proliferative signaling is matched by the production of ATP, NADPH, and macromolecular precursors (Yang et al., 2012). Thus, the MAPK cascade functions as a pivotal signaling conduit that couples proliferative signaling to enhance energy metabolism. Continuous RAS–ERK–RSK activation not only promotes mitogenic gene expression but also ensures that sufficient ATP, NADPH, and biosynthetic precursors are available to fuel rapid cell cycle progression and sustained tumor growth (Moss et al., 2022; Yang and Wu, 2024).

Enhanced energy metabolism and ROS generation in cancer cells: Cancer cells require abundant energy to sustain uncontrolled proliferation, continuous macromolecule synthesis, and cellular redox balance. This high bioenergetic demand drives the activation of both glycolytic and mitochondrial oxidative metabolisms (DeBerardinis and Chandel, 2016; Vander Heiden et al., 2009). Although glycolysis provides rapid ATP and intermediates for anabolism, proliferating cells also depend on mitochondrial Krebs cycle and oxidative phosphorylation (OXPHOS) to generate the majority of cellular ATP and reducing equivalents, such as NADH and FADH₂ (Ward and Thompson, 2012).

During this process, the electron transport chain (ETC) becomes a major source of ROS, primarily through electron leakage from complexes I and III, which react with molecular oxygen to form superoxide anion (O₂•⁻) (Murphy, 2009). The accelerated flux of electrons under high metabolic load increases the probability of electron leakage, causing elevated mitochondrial ROS generation (Sabharwal and Schumacker, 2014). In parallel, oncogenic Ras–ERK signaling stimulates NADPH oxidase (NOX) activity and enhances cytosolic ROS production, further amplifying the oxidative pressure within cancer cells (Sena and Chandel, 2012; Weinberg et al., 2010). Although moderate ROS levels function as signaling molecules that promote proliferation through MAPK activation, excessive ROS accumulation can damage nucleic acids, proteins, and lipids. Particularly, ROS-driven lipid peroxidation of polyunsaturated fatty acids (PUFAs) in cellular membranes serves as a biochemical hallmark of ferroptosis (Dixon et al., 2012). Thus, the hypermetabolic state of cancer cells inevitably creates a paradox: the same pathways that provide energy for growth also generate oxidative stress, which can push cells toward ferroptotic death when antioxidant defenses fail (DeNicola et al., 2011; Stockwell et al., 2017).

Adaptive tolerance to ROS in cancer cells downstream of the RAS–RAF–MEK–ERK–RSK pathway: Oncogenic activation of the RAS–RAF–MEK–ERK–p90RSK signaling cascade plays a central role in enabling cancer cells to survive in chronic oxidative environments by inducing adaptive antioxidant programs. Persistent RAS or BRAF activation elevates basal ROS production through mitochondrial metabolism and NADPH oxidase; however, paradoxically, these oncogenes also promote tolerance to oxidative stress by upregulating transcriptional antioxidant networks. For example, KRAS^G12D or BRAF^V619E mutations increase the expression and nuclear accumulation of NRF2, an oxidative stress–responsive transcription factor that activates the expression of detoxifying enzymes such as heme oxygenase-1 (HO-1), NAD(P)H: quinone oxidoreductase 1 (NQO1), and the catalytic and modifier subunits of glutamate-cysteine ligase (GCLC and GCLM) (DeNicola et al., 2011; Rojo de la Vega et al., 2018). Activated ERK1/2 further enhances this response by phosphorylating NRF2 and promoting its transcriptional activity, thereby increasing the antioxidant capacity of cancer cells (Chen et al., 2017).

Downstream of ERK, p90RSK contributes to redox adaptation by maintaining mTORC1–ATF4 signaling, which coordinates amino-acid metabolism and redox homeostasis. A key target of this adaptive response is the cystine/glutamate antiporter SLC7A11 (xCT, a component of system Xc^-). SLC7A11 expression is transcriptionally induced through both NRF2-dependent antioxidant response elements (AREs) and ATF4-dependent amino-acid response elements (AAREs), leading to increased cystine uptake and GSH synthesis. In parallel, the flavoprotein ferroptosis suppressor protein 1 (FSP1), also known as AIFM2, functions as a GSH-independent ferroptosis suppressor by regenerating ubiquinol (CoQH₂) at cellular membranes, quenching lipid radicals and serving as a secondary defense line when the GPX4–GSH axis is impaired (Bersuker et al., 2019; Doll et al., 2019; Tang et al., 2021). Recent findings in melanoma and lung cancer models indicate FSP1 could be a promising target to combat tumor metastasis independent of GPX4-GSH (Palma et al., 2025; Wu et al., 2025).

Together, these layers of antioxidant adaptation constitute a robust redox-buffering network that enables cancer cells to tolerate elevated ROS generated by oncogenic signaling and hypermetabolism. However, this tolerance contributes to therapeutic resistance and the same MAPK-driven antioxidant circuitry that supports proliferation can also block the oxidative cell death pathways that would otherwise eliminate damaged cells. The inhibition of the MEK/ERK axis dismantles the NRF2–SLC7A11–GSH defense program and resensitizes tumor cells to ferroptosis, particularly when combined with GPX4 inhibitors or translation suppressors that limit SLC7A11 synthesis (Jiang et al., 2025; Yin et al., 2024). Therefore, the RAS–RAF–MEK–ERK–RSK pathway not only promotes survival in high-ROS microenvironments but also defines metabolic vulnerability, and disruption of this axis or its antioxidant downstream effectors can shift the balance from tolerance to ferroptotic cell death.

The role of MAPK signaling in cancer proliferation: The MAPK pathway consists of the RAS–RAF–MEK–ERK–p90RSK signaling cascade, plays a pivotal role in regulating cell proliferation, differentiation, and survival. In normal cells, this pathway transduces extracellular mitogenic signals, such as growth factors, cytokines, or stress stimuli—into nuclear transcriptional responses that promote controlled cell cycle progression. However, constitutive activation of this cascade in cancer through mutations in RAS (KRAS, NRAS, and HRAS), BRAF, or the loss of upstream regulatory control results in sustained proliferative signaling, one of the fundamental hallmarks of cancer (Hanahan and Weinberg, 2011). Upon activation, GTP-bound RAS recruits and activates RAF kinases (A-RAF, B-RAF, and C-RAF), which phosphorylate and activate MEK1/2, a dual-specificity kinase that phosphorylates ERK1/2. Activated ERK translocates to both the cytoplasmic and nuclear compartments, where it phosphorylates numerous substrates, including transcription factors (ELK1, CREB, c-Fos, and c-Myc) and downstream kinases such as p90 ribosomal S6 kinase (RSK) (Lee et al., 2025). The ERK–RSK module stimulates the transcription and translation of key cell cycle regulators such as cyclin D1, CDK4, and E2F1, driving the G1/S transition and promoting continuous cell division (Chambard et al., 2007; Lavoie et al., 2020).

In addition to controlling cell cycle, the MAPK pathway supports cellular growth by stimulating anabolic metabolism. ERK1/2 and RSK phosphorylate and activate components of the mTORC1 pathway, including Raptor and TSC2, thereby enhancing cap-dependent translation and ribosome biogenesis (Carriere et al., 2011; Roux et al., 2007). The coupling of MAPK signaling with mTORC1 ensures the production of proteins and macromolecules required for biomass accumulation during proliferation. Furthermore, ERK and RSK activation promotes the expression of metabolic enzymes, such as hexokinase II (HK2), phosphofructokinase (PFK), and pyruvate kinase M2 (PKM2), thereby enhancing glycolysis and mitochondrial metabolism to meet the energetic and biosynthetic demands of rapidly dividing cells (Yang et al., 2012). Collectively, the RAS–RAF–MEK–ERK–RSK pathway functions as a central conduit linking extracellular mitogenic signals to cell cycle progression, protein synthesis, and metabolic reprogramming. Dysregulation of this cascade confers an uncontrolled proliferative capacity, allowing cancer cells to sustain autonomous growth independent of normal regulatory constraints.

INTERCONNECTION BETWEEN ROS SIGNALING, CELL SURVIVAL, AND CANCER MALIGNANCY

ROS and carcinogenesis

ROS function as both signals and lesions that collectively accelerate carcinogenesis from initiation to promotion and progression. During initiation, ROS directly damage the genome and generate mutational substrates. Hydroxyl radical and related oxidants oxidize guanine to 8-oxo-7,8-dihydroguanine (8-oxo-dG), mispairing with adenine to yield G→T transversions; in addition, ROS also produce sites, single- and double-strand breaks, and clustered lesions that overwhelm base-excision repair (BER), fostering genome instability (Cadet et al., 2010; Cooke et al., 2003). Lipid peroxidation generates electrophilic aldehydes, such as 4-hydroxynonenal (4-HNE) and malondialdehyde (MDA), which form mutagenic DNA adducts (e.g., M1dG, etheno-adducts) further increasing the burden of promutagenic lesions (Ayala et al., 2014; Nair et al., 2007). These oxidative and aldehyde-derived adducts provide substrates for selecting oncogenic variants and the loss of tumor suppressors, which are core events that lead to malignant transformation (Klaunig et al., 2011).

During promotion, sublethal ROS function as secondary messengers that sustain growth and survival signaling and remodel tissue ecology in favor of tumor expansion. H₂O₂ transiently oxidizes catalytic cysteines in protein tyrosine phosphatases (PTPs), prolonging RTK activity and downstream MAPK/ERK signaling, thereby enhancing cell-cycle progression and anabolic metabolism (Rhee, 2006). In addition, ROS activate NF-κB/AP-1, inducing cytokines, COX-2, and MMPs that foster a tumor-promoting inflammatory niche (Klaunig et al., 2011; Mantovani et al., 2008). At the metabolic level, mitochondria-derived ROS stabilize HIF-1α (via inhibiting prolyl hydroxylases), shifting cells toward glycolysis, angiogenesis (VEGF), and survival under hypoxia—changes that consolidate clonal expansion (Semenza, 2013). Together, redox-dependent signaling and inflammation constitute a promotive milieu that favors the growth of the initiated clones (Reczek and Chandel, 2017).

As the disease progresses, chronic oxidative stress promotes invasion, metastasis, and therapeutic tolerance. ROS-driven lipid peroxidation alters membrane fluidity and receptor partitioning, rewiring growth-factor sensing, and aldehyde adducts on proteins perturbing enzyme activities and proteostasis, further remodeling metabolism (Ayala et al., 2014). Sustained ROS maintains NF-κB-dependent inflammatory circuits and stromal activation, supports angiogenesis via HIF-1α, and promotes epithelial–mesenchymal transition (EMT) programs that enhance motility and matrix degradation (Mantovani et al., 2008; Semenza, 2013). In parallel, selection for high antioxidant capacity (NRF2-GSH/GPX and Trx systems) yields redox-tolerant subclones that better withstand oxidative microenvironments and cytotoxic therapies, thereby coupling ROS exposure with tumor evolution (Klaunig et al., 2011; Reczek and Chandel, 2017). Conceptually, ROS operate on two intertwined fronts: (i) lesion formation (DNA and lipid-derived adducts) that seeds genetic diversity and (ii) redox signaling that stabilizes growth, inflammatory, metabolic, and angiogenic programs, together with driving initiation, promotion, and progression.

Modulating cancer cell growth through ferroptosis: targeting the MAPK pathway and ROS generation

Considering the roles of the MAPK pathway and ROS generation in different aspects of carcinogenesis and cell survival, it is not surprising that these pathways could serve as potential targets for controlling cancer cell growth. Considering ROS generation alone, both the suppression of ROS production and induction of excessive ROS levels could be exploited as anticancer strategies. Downregulation of ROS mitigates the microenvironmental stimuli that promote metabolic upregulation and carcinogenesis, ultimately creating conditions that are unfavorable for cancer cell survival. Certain antioxidants have been reported to induce apoptosis in cancer cells by downregulating ROS (Jeong and Joo, 2016). However, it would be difficult to predict the occurrence of ferroptosis under conditions of downregulated ROS levels, as this form of cell death depends on lipid peroxidation driven by ROS. Therefore, if ferroptosis is to be employed as an anticancer strategy, the level of ROS generation must exceed the antioxidant capacity of cancer cells. To achieve this, either the induction of ROS generation or the suppression or inhibition of antioxidant mechanisms can be utilized.

As shown above in part I-3, ferroptosis is generally accompanied by the inhibition of MAPK signaling, suggesting that targeting this pathway could be a promising strategy for anticancer therapy by inducing ferroptosis. Notably, to the best of our knowledge, compounds or stimuli that simultaneously activate MAPK signaling and serve as effective anticancer therapeutics have not yet been reported. Altogether, the control of cancer cell growth by targeting the MAPK pathway and ROS generation could leverage the ferroptotic cell death mechanism, particularly when MAPK signaling is inhibited and ROS levels exceed the antioxidant capacity of cancer cells.

CONTROL OF ROS GENERATION AS AN ANTICANCER THERAPEUTIC STRATEGY UTILIZING FERROPTOSIS

ROS: Friend or foe?

The cytotoxic potential of ROS is highly dependent on their intracellular concentration, duration, and subcellular localization. Under physiological conditions, basal intracellular H₂O₂ levels remain within the low-nanomolar range (1-50 nM), functioning as signaling molecules that modulate kinases and transcription factors (Lyublinskaya and Antunes, 2019). Minute increases of 3-10 nM are sufficient to alter phosphatase activity and trigger adaptive transcriptional responses. However, the same molecules become cytotoxic when the steady-state H₂O₂ level exceeds the buffering capacity of antioxidant systems. Experimental measurements using controlled steady-state delivery systems reveal that apoptosis can be initiated at extracellular H₂O₂ concentrations around 7 µM, corresponding to submicromolar intracellular levels (Antunes and Cadenas, 2001). For example, Jurkat T cells, grown in suspension undergo apoptosis when maintained in a steady-state oxidative environment. In contrast, most adherent cancer cells, such as A549 lung carcinoma, require higher doses—approximately 100 µM H₂O₂ exposure for 6-24 h—to reach a similar threshold for cell death (Vilema-Enriquez et al., 2016). These discrepancies reflect the antioxidant capacities and metabolic adaptations of different cancer cell types. Cells with elevated GSH or Trx activity can transiently withstand oxidative insults, shifting the lethal threshold upward.

Although H₂O₂ and superoxide (O₂•⁻) underlie classical oxidative cytotoxicity, lipid peroxidation–derived ROS represent another decisive trigger for regulated cell death, particularly ferroptosis. In this context, the quantitative determinant of death is not the total ROS concentration but the accumulation rate of lipid-ROS that exceeds the detoxifying ability of GPX4. Pharmacological inhibition of GPX4 by RSL3 has provided functional measures of this threshold. RSL3 induces irreversible cell death at submicromolar concentrations (<1 µM) in highly ferroptosis-sensitive cell lines such as MDA-MB-231 and HCC1954, whereas relatively resistant lines, such as MCF-7, SK-BR-3, and T47D, require 5-11 µM RSL3 to reach equivalent lipid-ROS levels (Park et al., 2019). Fluorescent reporters, such as BODIPY-C11, visualize this process, where a sharp increase in the oxidized/reduced ratio marks the lipid-ROS death threshold (Co et al., 2024). Altogether, these data indicate that the boundary between adaptive and lethal ROS is narrow and context-dependent. In most mammalian cells, intracellular H₂O₂ concentrations above several hundred nanomolar or sustained extracellular levels exceeding 5-10 µM begin to trigger apoptosis. In contrast, lipid-ROS–driven ferroptosis operates through membrane-localized oxidative chain reactions, often at total ROS levels lower than those required for necrosis or apoptosis but concentrated within phospholipid domains. These quantitative distinctions emphasize that the biological outcome of ROS is not absolute but relative, determined by the balance between generation and scavenging, the spatial distribution of oxidative events, and the intrinsic redox tolerance of the cell.

Therapeutically, these insights imply that manipulating ROS levels toward either side of the survival threshold can determine the fate of cancer. Sublethal ROS promote oncogenic signaling and resistance, whereas exceeding the oxidative limit, through mitochondrial overload, GPX4 inhibition, or radiotherapy, irreversibly commits cells to death. Thus, although ROS may serve as a “friend” to cancer during its growth, carefully controlled perturbation of its redox homeostasis transforms it into a potent “foe.”

Sensitization of anticancer therapy by controlling ROS level and ferroptosis induction

Sublethal levels of ROS may promote oncogenic signaling and resistance to anticancer therapy. The balance of ROS levels transforms ROS into a potent foe. Below are certain targets that regulate ROS levels and induce ferroptosis.

Transcription factors and related targets: Multiple transcription factors regulate ROS; among them, NRF2 (Rojo de la Vega et al., 2018; Sun et al., 2016) plays a prominent role in ferroptosis, particularly in drug-resistant cancer cells. Activation of MAPK signaling may potentiate NRF2 nuclear translocation and transcriptional activity, allowing antioxidant defense mechanisms to inhibit ferroptosis. In contrast, NRF2 downregulation results in the depletion of antioxidants, such as GSH, as well as an increase in lipid peroxidation to promote ferroptosis. Not surprisingly, one of the key molecular targets involved in the induction of ferroptosis, system Xc^-, is transcriptionally regulated by NRF2. Moreover, NRF2 regulates several key genes involved in GSH synthesis (SLC7A11, GCLC, GCLM, and GSR) and iron metabolism (FTH1) (Tang and Kang, 2024). Therefore, NRF2 is a critical therapeutic target for overcoming resistance to ferroptosis. Enhanced chemosensitivity has been reported in ovarian cancer cells treated with tripterygium glycosides, which downregulate of NRF2 (Ma et al., 2023). Carnosic acid inactivates NRF2 and sensitizes oral squamous cell carcinoma cells to cisplatin (Han et al., 2022). In addition, the suppression of NRF2 signaling by triptolide induces ferroptosis in doxorubicin-resistant leukemia cells (Wu et al., 2023), and the inhibition of NRF2 by trigonelline sensitizes head and neck cancer cells to cisplatin (Roh et al., 2017). In contrast, curcumol sensitizes cisplatin-resistant gastric cancer cells by regulating the p62/KEAP1/NRF2 pathway (Feng et al., 2024). The activation of the NRF2 pathway confers drug resistance, whereas its inhibition leads to ferroptosis (Roh et al., 2017). In addition to NRF2, other transcription factors could be utilized in inducing ferroptosis. RE1-silencing transcription factor (REST) is an inhibitory transcription factor (Plaisance et al., 2005). Erianin, a natural product, targets REST leading to the upregulation of LRSAM1 expression that ubiquitinates and degrades ferroportin, resulting in ferroptosis in TMZ-resistant glioma cells (Mansuer et al., 2024).

Although not a transcription factor, LINC00152 is involved in the transcriptional regulation of ferroptosis signaling. LINC00152 is an oncogenic long noncoding RNA that enhances the stability of PDE4D mRNA, which encodes a phosphodiesterase (Li et al., 2022). Inhibition of LINC00152 reduces PDE4D expression, thereby potentiating cAMP signaling through PKA and CREB, ultimately leading to increased ROS generation and lipid peroxidation (Saatci et al., 2024). The downregulation of a single enzyme can trigger the induction of multiple gene products and ferroptosis, possibly overcoming tamoxifen resistance.

Antioxidant systems: Drug-resistant cancer cells exhibit elevated ROS levels and sustain their viability by upregulating antioxidant systems. However, once these antioxidant systems are compromised, cancer cells can no longer protect themselves from high ROS levels, and this vulnerability can be leveraged to overcome resistance. GPX4 is one of the most commonly used targets for the induction of ferroptosis. Fin56 is a well-known inducer of ferroptosis that targets GPX4 (Sun et al., 2021), and siRNA treatment targeting GPX4 can induce ferroptosis in oxaliplatin-resistant CRC cells (Golbashirzadeh et al., 2023). Furthermore, the effect of GPX4 knockdown in lapatinib-resistant NSCLC cells indicates the central role of GPX4 in ferroptosis too (Ni et al., 2021).

ALDH5A1, the succinic semialdehyde dehydrogenase (SSADH), is involved in ROS regulation (Chambliss and Gibson, 1992). A xenograft mouse model demonstrated that ALDH5A1 expression is related to cisplatin-resistance and that ALDH5A1 silencing promotes ferroptosis signaling (Song et al., 2024).

Antioxidant systems are inhibited by different mechanisms, including the regulation of ubiquitination. USP2 is a deubiquitinating enzyme (DUB) belonging to the ubiquitin-specific protease family (Shin et al., 2017). It is involved in the deubiquitination of p53 at the K305R stabilizing p53 and promoting its nuclear translocation, which enhances ferroptosis and reduces cisplatin-resistance in NSCLC cells (Gong et al., 2025). The role of USP2 is somewhat complicated as the upregulation of USP2 promotes migration and invasion in certain other cancer cells (Qu et al., 2015). USP14 is another DUB (Wang et al., 2021). Recently, the curcumin derivative mitocur-1 was demonstrated to inhibit USP14, leading to GSH depletion and SLC7A11 downregulation, which ultimately induces ferroptosis. These findings suggest that mitocur-1 can be used to treat vemurafenib-resistant melanoma cells (Li et al., 2024).

Receptor tyrosine kinases (RTKs) and MAPK signaling: In Section II-2, we discuss the potential of modulating MAPK signaling and ROS generation to induce ferroptosis. As demonstrated in the case of the RON receptor tyrosine kinase, inhibition of RTK activity can downregulate MAPK signaling and induce of ferroptosis (Jin et al., 2024). Several growth factors, such as EGFR and MET, are RTKs, and their activation is closely related to carcinogenesis (Du and Lovly, 2018). FMS-like tyrosine kinase 3 (FLT) is highly expressed in breast cancer cells, and it functions as a ferroptosis inhibitor (Shen et al., 2024). NRF2, a pivotal target for ferroptosis induction, is a downstream signaling pathway for RTKs (Cirotti et al., 2024). Overall, RTKs may be good targets for inducing ferroptosis, especially in cells resistant to anticancer drugs (Viswanathan et al., 2017). It also should be noted that the activation of RTKs, rather than their inhibition, can sometimes increase cellular susceptibility to ferroptosis. The upregulation of ACSL4, a kind of ferroptosis promoter, can be induced by RTK activation through the RAS/RAF/c-Myc axis, implicating the dual function of RTKs in regulating ferroptosis (Fig. 3) (Sun et al., 2024).

Fig. 3.

Dual Regulation of Ferroptosis by RTK Signaling Pathways: Inhibition vs. sensitization. Activated RTK signaling modulates key regulators of ferroptosis. Activation of NRF2 induces the expression of SLC7A11, thereby suppressing ferroptosis. In contrast, activation of c-Myc upregulates ACSL4 expression, sensitizing cells to ferroptosis. receptor tyrosine kinase (RTK); nuclear factor erythroid 2-related factor 2 (NRF2); soluble carrier family 7 member 11 (SLC7A11); acyl-CoA synthetase long-chain family member 4 (ACSL4).

FUTURE DIRECTION AND CONCLUSION

We have provided a concise overview of ferroptosis, including the major regulators that drive the process, key signaling pathways such as the RTK, MAPK, and NRF2 pathways, and several molecular targets involved in ferroptosis induction. Molecules that selectively modulate key regulators of ferroptosis, such as GPX4, ferroportin, and other proteins involved in iron metabolism, need to be developed, in addition to the currently identified targets. These targets and modulators should exhibit tumor specificity, with a particular emphasis on drug-resistant cancer cells. We do not expect these approaches to replace current chemotherapies or targeted therapies. Combining ferroptosis inducers with conventional treatments may improve therapeutic outcomes and potentially overcome resistance mechanisms. To better utilize ferroptosis to overcome drug resistance in anticancer therapy, it is essential to understand the tumor microenvironment, particularly in relation to ferroptosis regulation involving immune cells and metabolic crosstalk. In the search for drug candidates that induce ferroptosis, drug repositioning, especially of inhibitors targeting the RTKs and MAPK signaling pathways, may provide valuable insights. Moreover, minimizing off-target effects and protecting normal tissues are critical for improving drug selectivity and facilitating clinical translation. In conclusion, inducing ferroptosis in drug-resistant cancer cells may provide a promising strategy for developing effective therapies against treatment-resistant malignancies.