System X c − (xCT) is a cystine/glutamate antiporter consisting of two subunits, including SLC3A2 and SLC7A11, which is responsible for exporting glutamate and importing cystine (12). Upon this transportation, cystine in the cytosol following an NADPH-consuming reduction reaction by thioredoxin reductase 1 (TXNRD1) is converted to cysteine, an antioxidant amino acid and precursor of tripeptide glutathione (GSH). Moreover, GSH is also a potent antioxidant that is used by glutathione peroxidase 4 (GPX4) for inhibition of ferroptosis by preventing the accumulation of lipid hydroperoxides (LOOHs) and converting them to lipid alcohols (LOHs) ( Figure 1 ). Therefore, system X c − as a membrane heterodimer is implicated in inhibition of ferroptosis under oxidative stress to provide cancer cell survival (13). Up to now, various studies have pointed to the role of this system in the regulation of ferroptosis in lung cancer, most of them representing SLC7A11 as the targets of various novel drugs and upstream regulators ( Figure 1 ). For example, a recent study has elucidated that there is a positive relationship between the level of transcription factor SOX2 and SLC7A11 in lung cancer stem-like cells (CSLC), and upregulation of SOX2 in lung tumors can highly inhibit ferroptosis and increase resistance to imidazole ketone erastin (IKE), a strong inducer of ferroptosis (14). Likewise, a similar relationship was observed between RNA binding protein RBMS1 and SLC7A11, in a way that depletion of RBMS1 sensitizes lung cancer cells to ferroptosis and radiotherapy (15). Moreover, Ma et al. have reported that YT521-B homology domain containing 2 (YTHDC2) is a downregulated tumor suppressor in lung adenocarcinoma (LUAD), which can suppress the activity of system X c − through targeting mRNA encoding SLC7A11 (16). However, their further investigations revealed that inhibition of SLC7A11 by YTHDC2 is not enough to induce ferroptosis in lung adenocarcinoma. Therefore, they proposed that SLC3A2 as another subunit of system X c − should be inhibited by YTHDC2 as well. High-level induction of YTHDC2 could inhibit the expression of SLC3A2 indirectly by inhibiting the expression of HOXA13. Further in vivo experiments also confirmed that the induction of YTHDC2 can highly contribute to ferroptosis. Therefore, elevating the levels of YTHDC2 was shown to be a promising strategy in LUAD to induce ferroptosis by selectively inhibiting the expression of both subunits of system X c − in tumor cells (17).

GPX4, a selenoperoxidase and a key upstream regulator of ferroptosis, plays two roles at the same time including, converting GSH to glutathione disulfide (GSSG) and reducing phospholipid hydroperoxides (PLOOHs) to their corresponding alcohol (PLOHs), thereby preventing the accumulation of lipid peroxides and leading to suppression of ferroptosis. In addition, GSSG can be recycled to GSH under the action of glutathione–disulfide reductase (GSR) using the electrons provided by NADPH/H⁺ (9) ( Figure 1 ). Various studies have reported that GPX4 is upregulated in different cancers, including LUAD, and is associated with poor prognosis of patients and chemotherapy resistance. Suppression of GPX4 by a small molecule inhibitor named RSL3 strongly enhances the anticancer effects of cisplatin in vivo. In other words, RSL3 combined with cisplatin could induce ferritinophagy/ferroptosis (18). Another study has also shown that the levels of GPX4 and mTORC1 are higher in NSCLC cells, and inhibition of these molecules provides a promising strategy to overcome lapatinib resistance in vivo (19). cAMP response element-binding protein (CREB) is also an upregulated oncogene in LUAD tissues that positively regulates the expression of GPX4, and their levels are closely related to tumor size and stage (20). A similar relationship was also observed between serine threonine tyrosine kinase 1 (STYK1) and GPX4 in NSCLC. In brain metastasis of lung adenocarcinoma, Glutathione S‐transferase M1 (GSTM1) is another protein that is stablized by GPX4 and concurrently result in inhibition of ferroptosis and subsequent resistance to platinum through increasing GSH consumption (21). We will further show that many drugs and upstream signaling pathways significantly affect the regulation of GPX4 in lung cancer (22)( Figure 1 and Table 2 ).

Recent studies have identified that GPX4 is not the only ferroptosis suppressor in human cancer cells, and ferroptosis suppressor protein 1 (FSP1, formerly known as AIFM2) independent of GSH suppresses lipid peroxidation and subsequent ferroptosis through converting ubiquinone (coenzyme Q10, CoQ10) to ubiquinol (CoQH2, reduced form of CoQ10) using an NADPH-consuming reduction reaction ( Figure 1 ). Therefore, this pathway is also referred to as the CoQ-dependent pathway (9). Bersuker et al. showed that FSP1 expression positively correlates with ferroptosis resistance in various cell lines, particularly in the xenograft mouse model of lung cancer ( 23).

Cysteine desulfurase nitrogen fixation 1 homolog (NFS1) is an enzyme that extracts sulfur from cysteine for the biosynthesis of iron-sulfur clusters (ISCs) and is expressed at higher levels in differentiated lung adenocarcinomas. A study has demonstrated that suppressing NFS1 as well as ISCs biosynthesis triggers an iron-starvation response and subsequent iron influx to cells by molecules such as transferrin receptor protein (TRFC), thereby promoting ferroptosis and inhibiting tumor growth in lung tumors (24, 25) ( Figure 1 ).

Transcription factor nuclear factor erythroid 2-related factor 2 (NFE2L2 or NRF2) is an overriding antioxidant transcription regulator of genes involved in lipid peroxidation and the accumulation of free iron (26). Takahashi et al. emphasized that hyperactivation of NRF2 is required for proliferation and survival of 3D culture models of lung tumors through preventing ferroptosis. However, NRF2 downregulation was not enough to suppress ferroptosis since GPX4 levels were increased, proposing that enhanced oxidative stress caused by NRF2 downregulation might activate other cytoprotective signaling pathways, including nuclear factor-κB (NF-κB), involved in activating downstream antioxidant enzymes. Notably, simultaneous inhibition of NRF2 and GPX4 could efficiently induce ferroptosis (27). The study by Liu et al. showed that Wnt signaling pathway through provoking NRF2 increases the activity of GPX4 in brain metastasis of lung adenocarcinoma cells, thereby resulting in chemoresistance to platinum (21). In addition, the activity of NRF2 positively reflects erastin resistance in isogenic lung cancer cell models, regardless of Kelch-like ECH-associated protein 1(KEAP1) mutation (28), a molecule that represses NRF2 via its ubiquitin proteasomal degradation in normal conditions (29). Likewise, KEAP1-mutant lung cancer cells were shown to have higher levels of NRF2 and its downstream target SLC7A11. In addition, KEAP1 deficient lung tumors were shown to be sensitive to inhibition of glucose transporter 1 (GLUT1) due to their glucose dependency (30). Similarly, Wohlhieter et al. (31) showed that LUAD tumors with concurrent mutations in serine/threonine kinase 11 (STK11) and KEAP1 were more resistant to ferroptosis since the activity of NRF2 pathway and targets involved in ferroptosis such as SLC7A11 and GPX4 were increased, leading to worse overall survival and enhanced tumor proliferation both in vivo and in vitro ( Figure 2 ). Moreover, their further investigations turned out that stearoyl-CoA desaturase (SCD1, SCD) was necessary for the proliferation of the cells mentioned above ( Figure 2 ). SCD1 was previously shown to participate in aberrant lipid metabolism and promoted cell growth in lung cancer (32); therefore, its genetic and pharmacological inhibition could sensitize STK11/KEAP1 co-mutant cells to ferroptosis induction even in vivo. Noteworthy to mention, the role of mitochondria and the tricarboxylic acid (TCA) cycle has been highlighted recently in regulating ferroptosis (33). NRF2 can also affect mitochondrial function. For example, in NSCLC, FOCAD-FAK signaling was shown to be involved in cysteine deprivation-induced ferroptosis, and NRF2 was shown to inhibit the FOCAD-FAK signaling axis and induce ferroptosis via increasing the activity of Complex I in the mitochondrial electron transport chain (ETC) and TCA cycle. However, inhibiting NRF2 is not enough for ferroptosis induction since the FOCAD-FAK axis does not affect GPX4. Therefore, adding an NRF2 inhibitor such as brusatol concurrent with erastin can promote ferroptosis induction better in NSCLC cells (34) ( Figure 2 ). Similarly, a novel study conducted by Kang et al. has highlighted the role of the NRF2 signaling pathway in cystine depletion conditions. NSCLC cells with higher expression levels of NRF2 produce many γ-glutamyl-peptides due to increased activity of glutamate-cysteine ligase catalytic subunit (GCLC). Following production of γ-glutamyl-peptide, glutamate is not accumulated in the cytosol anymore and this process leads to ferroptosis inhibition (35). Activating transcription factor 2 (ATF2) is another protein with an oncogenic role in lung cancer and increases NRF2 expression following treatments with a group of drugs named BET inhibitors (BETi, with the ability to induce ferroptosis in breast cancer) in LUAD, leading to ferroptosis resistance (36). In addition, a very recent study has shown that the expression of E3 ubiquitin ligase Mindbomb 1 (MIB1) is upregulated in a group of lung squamous and adenocarcinoma cells and correlates negatively with patients’ survival. Nevertheless, cells overexpressing MIB1 are more sensitive to ferroptosis due to proteasomal degradation of NRF2 by MIB1 (37). Taken together, due to the prominent role of NRF2 in regulating ferroptosis, targeting it by various treatments can significantly induce ferroptosis (see Table 2 ).

Acyl-CoA synthetase long-chain family member 4 (ACSL4) is a crucial enzyme responsible for lipid metabolism, which converts PUFAs to PUFA-CoAs, and following the action of LPCAT3, the products of this reaction are esterified into phospholipid containing polyunsaturated fatty acid chain (PUFA-PLs). Then, these PUFA-PLs are oxidized by another enzyme named ALOX15 into PL-PUFA-OOHs. Since PL-PUFA-OOHs can trigger ferroptotic cell death, the activity of the abovementioned enzymes contributes to the promotion of ferroptosis (38). Surprisingly, although ALOX15 and ACSL4 facilitate the ferroptosis process, recent studies showed that higher expression of these molecules was associated with increased cancer malignant features. Thereby, these molecules can act as double-edged swords either by promoting or inhibiting cancer progression (39). However in SCLC, Bebber et al. showed that ACSL4 and LPCAT3 are expressed at higher levels in non-neuroendocrine (non-NE-SCLC) than NE-SCLC, contributing to ferroptosis sensitivity and resistance, respectively in the abovementioned cells. Moreover, they proposed that TRX antioxidant pathway is overactivated in NE-SCLC cells, and its inhibition by auranofin alongside treatment with buthionine sulfoximine (BSO), as a GSH level reducer, could successfully induce ferroptosis and inhibit tumor progression in mice with xenograft NE-SCLC tumors (40). In addition, ACSL4 was found to act as a tumor suppressor and a favorable prognostic factor in patients with LUAD and promoted ferroptosis but inhibited tumor cell survival, invasion, and migration. Most interestingly, a high-fat diet could reverse these effects via downregulating ACSL4 both in vivo and in vitro ( 41).

The epidermal growth factor receptor (EGFR) pathway is involved in the progression of various cancers, and its mutation is frequently seen in lung adenocarcinomas (42). It is noteworthy to mention that the previously mentioned SCD1 is stabilized by EGFR via Y55 phosphorylation and contributes to cancer progression in lung cancer (32). Moreover, activation of mitogen-activated protein kinase (MAPK) as a downstream target of the EGFR pathway is required to induce ferroptosis ( Figure 3 ). Following cystine-deprived conditions, those NSCLC cells with the highest MAPK signaling activity significantly lose their viability through a ferroptosis process caused by inhibited and promoted expression of GPX4 and NOX4, respectively (43).

Hypoxia, as a hallmark of cancer, leads to excess tumor vascularization and progression (44). Activation of both MAPK and EGFR pathways can induce the expression of hypoxia-inducible factor 1-alpha (HIF-1α) (44, 45). In addition, induction of HIF-1 by EGFR can make A549 lung cancer cells resistant to another type of PCD called anoikis under the lipid raft-disrupting stress (45). Emerging evidence also supports the role of HIFs in the induction of ferroptosis (46, 47). Jiang et al. showed that iron-dependent enzymes Egl nine homolog (EGLNs) under hypoxia conditions degrades HIF-1α ( Figure 3 ), leading to increased lymphoid-specific helicase (LSH) expression, a chromatin remodeling factor that acts as an oncogene and ferroptosis inhibitor in lung cancer via increasing the expression of SCD1 (48, 49).

Tumor suppressor P53 has a dual role in regulating ferroptosis. Spermidine/spermine N1-acetyltransferase (SAT1) is one of those proteins which is upregulated by p53 and responsible for oxidative stress and ferroptosis (50, 51). However, SAT1 is under the control of factors other than p53. For instance, RNF113A is an oncogene RNA-binding protein whose deficiency provokes ferroptosis via promoting SAT1 expression ( Figure 3 ) and was upregulated and contributed to cisplatin resistance in lung adenocarcinomas (52). Recently, it has been discovered that ΔNp63α as a major isoform of p63 can inhibit ferroptosis and oxidative stress independent of p53 and NRF2 activity in lung cancer through transcriptional controlling the expression of genes involved in glutathione synthesis (53).

Ubiquitin–proteasome system (UPS) belongs to a degradation pathway that controls lipid peroxidation and iron accumulation via degradation of molecules involved in ferroptosis (8). A recent review has also highlighted the role of ubiquitination in ferroptosis (54). Ubiquitin‐specific protease 35 (USP35) is a deubiquitinase that is overexpressed in lung cancer, and its knockdown, in addition to promoting ferroptosis and chemotherapeutic sensitivity to cisplatin and paclitaxel, inhibits lung cancer cell growth, colony formation, and tumor progression. The mechanism by which USP35 overexpression led to ferroptosis was attributed to ferroportin (FPN) stabilization, a protein responsible for exporting iron to the outside of the cells ( Figure 3 ) ( 55). In the same way, a similar relationship between deubiquitinase USP11 and NRF2 was found in patients with NSCLC (SCC subtype), leading to ferroptosis resistance and cell proliferation ( Figure 3 ) ( 56). Noteworthy to mention, the effects of some drugs in inducing ferroptosis are also carried out by the proteasome degradation system. For instance, concurrent treatment with siramesine alongside lapatinib was shown to induce ferroptosis via proteasome degradation of heme oxygenase-1 (HO-1) (57).

Autophagy is another intracellular degradation pathway in which various molecules and organelles in the cell are engulfed and then degraded following the formation of lysosomal structures (8). There is close crosstalk between ferroptosis and autophagy, in a way that some of the recent reviews consider ferroptosis as a type of autophagy-dependent cell death (58, 59). The reason behind this assumption is that in the presence of some ferroptosis inducers such as erastin and RSL3, the formation of autophagosomes and components of the autophagy system is also provoked, and thereby ferroptosis is promoted (58). Some pieces of evidence support this relationship in lung cancer. Following inhibition of USP14 by 6-Gingerol (a natural product) in A549 lung cancer cells, the levels of ROS, iron concentration, and autophagosomes started to increase and contributed to increased expression of ferroptosis and autophagy-related proteins in vivo and in vitro ( Figure 3 ). Therefore, it can be concluded that USP proteins are a common executer among autophagy and ferroptosis process (60). Similarly, treatment with curcumin, a well-known natural derived product, could significantly induce cell death both in vivo and in vitro through an autophagy-dependent ferroptosis mechanism and inhibition of autophagy reversed ferroptosis caused by curcumin in NSCLC (61).

Hippo is an evolutionarily conserved signaling pathway with potent tumor suppressor activities involved in determining cell fate. Its dysregulation highly induces cancer progression and therapy resistance through aberrant activation of two transcriptional coactivators, including yes-associated protein (YAP) and transcriptional coactivator with PDZ-binding motif (TAZ). When hippo signaling is off, YAP and TAZ are translocated to the nucleus and interact with DNA binding transcription factors named TEA domain family member (TEAD) to provoke the expression of genes involved in cell proliferation (62). Recently the role of Hippo/YAP/TAZ axis in regulating ferroptosis was described as dependent on the distance between cells and their contact. In other words, when there is no contact between the cells (low-density conditions), due to the inactivation of Hippo signaling, YAP and TAZ trigger the expression of genes responsible for ferroptosis induction rather than apoptosis. While this process is reversed when cells get closer to each other and extracellular E-cadherin stimulates Hippo pathway activation, leading to increased apoptosis (62). In LUAD, Zhang et al. demonstrated that the magnitude of YAP suppression is a crucial determiner of ferroptosis sensitivity. While inhibiting system X c − by sorafenib could induce ferroptosis and decrease YAP levels, suppressing GPX4 could not, proposing that subsequent glutamate accumulation in the cytosol might sensitize cells to ferroptosis. Their further experiments showed that glutamate repletion conditions (e.g., system X c − inhibition) provokes Ca²⁺ ions accumulation in the cytosol and ADCY10 initiates a cascade of reactions leading to YAP destabilization and subsequent ferroptosis ( Figure 3 ). Since ADCY10 is expressed higher in advanced stage and therapy resistance LUAD cells, targeting ADCY10 as a molecule linked with YAP and ferroptosis was shown to have clinical significance (63).

The latest review on ferroptosis insists that ions other than iron (e.g., zinc) cannot induce ferroptosis (8); however, other ions can indirectly affect ferroptosis. As just mentioned earlier, at least in lung cancer, accumulation of Ca²⁺ subsequently triggers the reactions that lead to ferroptosis sensitivity (63). Likewise, treating lung cancer cells with erianin (a natural product) promoted ferroptosis via calcium/calmodulin (CaM) signaling activation, a pathway involved in increasing intracellular Ca²⁺ levels ( Figure 3 ) ( 64). Moreover, zinc intoxication in NSCLC cells upregulates the mTORC2/RICTOR pathway, resulting in system X c − phosphorylation and subsequent ferroptosis ( Figure 3 ). However, it should be noted that these effects of zinc are reversed after a specific time and are limited using iron chelator deferoxamine and vitamin E (65).

The mechanistic target of rapamycin (mTOR) is an essential negative regulator of autophagy (8). Moreover, GPX4 is also involved in regulating ferroptosis, and the levels of mTOR are positively correlated with GPX4 levels; therefore, mTOR inhibitors can induce autophagy-dependent ferroptosis (66). For this reason, “inhibition of GPX4 or mTORC1 overcomes resistance to Lapatinib via promoting ferroptosis in NSCLC cells” (19), while zinc intoxication increases mTORC2 activity and ferroptosis (65). This proposes the theory that mTORC1 and mTORC2 may have distinct impacts on regulating ferroptosis, at least in lung cancer. Moreover, it has been proposed that lung tumors, which in them NRF2 signaling pathway is activated more than usual, are more dependent on the mTOR pathway, and synergistic cooperation between NRF2 and mTOR signaling can enhance cell proliferation in 3D cultures (27).

miRNAs are a class of ncRNAs with 18-25 nucleotides long that can inhibit protein translation by affecting the expression of their target mRNAs (69). Tumor suppressor miRNAs are downregulated in cancers, and their targets are oncogenic proteins. miR-302a-3p is one of those downregulated tumor suppressor miRNAs that induces ferroptosis via targeting 3’-UTR of FPN mRNA in NSCLC. Ectopic induction of miR-302a-3p mimics provides a promising strategy against NSCLCs and sensitizes them to erastin, RSL3, cisplatin, and paclitaxel (70). Similarly, miR-324-3p expression is downregulated in cisplatin-resistant NSCLC cells, and its overexpression sensitizes resistant cells to cisplatin via targeting GPX4 ( Figure 1 ) ( 71). Moreover, tumor cells can release exosomes containing miRNAs to their surrounding microenvironment. miR-4443 is an oncogenic miRNA found abundantly in exosomes released from cisplatin-resistant NSCLCs and suppresses ferroptosis in vivo and in vitro through increasing FSP1 protein level indirectly by targeting methyltransferase-like 3 (METTL3), a molecule that causes N⁶-methyladenosine (m ⁶A) methylation of FSP1 (72) ( Figure 1 ).

lncRNAs are a group of non-coding RNAs which are more than 200 nucleotides in length. Similar to miRNAs, these molecules affect gene expression through an extensive range of mechanisms. Moreover, they may act as sponges for miRNAs and inhibit their activity (73). For example, LSH can increase the expression of lncRNA-LINC00336 in lung cancer and thereby inhibit ferroptosis through sponging the activity of MIR6852, a miRNA that targets mRNA of a ferroptosis suppressor named cystathionine-β-synthase (CBS) (74) ( Figure 3 ). Similarly, LSH is responsible for the downregulation of tumor suppressor lncRNA-P53RRA in NSCLC. P53RRA activates the p53 signaling pathway and induces ferroptosis via interacting with Ras GTPase-activating protein-binding protein 1 (G3BP1)( Figure 3 ) ( 75). Furthermore, lncRNA-NEAT1 was shown to induce ferroptosis in NSCLC by increasing the expression levels of ACSL4 mRNA (76). Recently, Sui et al. showed that SAT1 mRNA could be stabilized by an RNA binding protein named U2AF2 and consequently lead to ferroptosis in LUAD. However, lncRNA-ASMTL‐AS1, which is responsible for recruiting U2AF2 and promoting SAT1 expression, is downregulated in LUAD cells, and thereby recovering its expression can inhibit malignant features of the cancer cells (77). Targeted delivery of ncRNAs can be considered a promising strategy against cancer. Gai et al. adopted a novel strategy to induce ferroptosis in NSCLC tumors by deploying folate-modified liposomes containing lncRNA-MT1DP combined with erastin. Since MT1DP increases the activity of miR-365a-3p, a miRNA that targets NRF2, subsequent deactivation of NRF2 ( Figure 2 ) confers sensitivity to erastin-induced ferroptosis in vivo and in vitro ( 78).

circRNAs are more stable than miRNAs and lncRNA due to their covalently-closed structures, and their dysregulation is implicated in the progression of various cancer, including lung cancer (79). CircDTL is an upregulated oncogene circular RNA in NSCLC cells that inhibits ferroptosis via acting as a sponge for miR-1287-5p, a miRNA that targets GPX4. Moreover, inhibition of circDTL can increase the sensitivity of lung cancer cells to erastin in vivo ( 80).