(Anti-)oxidant Metabolism plays an important role in ferroptosis. Glutathione (GSH) metabolism and anti-oxidant capacity regulate sensitivity to ferroptosis. GSH is a tripeptide antioxidant that acts as a cofactor of Se-dependent GPX4 to reduce lipid hydroperoxides (Yant et al., 2003; Lu, 2009). Inhibition of cystine required for GSH synthesis eventually leads to depletion of intracellular GSH levels (Dixon et al., 2012; Dixon and STOCKWELL, 2014). GPX4 converts GSH between the reduced and oxidized states and converts lipid hydroperoxides to lipid alcohols. This process prevents the formation of Fe²⁺ dependent toxic lipid ROS (Labunskyy et al., 2014; Forcina and DIXON, 2019). GPX4 is the only reported enzyme that can directly reduce complex phospholipid peroxides and is the downstream target gene of NRF2 (Nuclear factor E2-related factor 2) (Forcina and DIXON, 2019; Friedmann Angeli et al., 2019). Erastin, a classical ferroptosis-inducing drug, depletes GSH and indirectly inactivates GPX4, leading to accumulation of toxic lipid ROS and subsequent lipid peroxidation (Dixon et al., 2012; Dixon and STOCKWELL, 2014), ultimately leading to ferroptosis.

At present, most studies on NRF2 in HCC involve the p62-Keap1 (Kelch-like ECH-associated protein 1)-NRF2 axis. The p62-Keap1-NRF2 signaling pathway is involved in the process of cell avoiding ferroptosis. NRF2 is a key regulator of the antioxidant response, including the expression of the Cystine/glutamate exchange system (system X^C−) (Hassannia et al., 2019). Inhibition or knockdown of NRF2 enhances erastin- or sorafenib-induced ferroptosis in HCC in vitro and in vivo (Hassannia et al., 2019). The System X^C− consists of solute carrier family 7 member 11 (SLC7A11, xCT) and solute carrier family 3 member 2 (SLC3A2, 4F2hc) by disulfide bonded, which import the extracellular oxidized form of cysteine and cystine, in exchange for intracellular glutamate. SLC7A11 indirectly inactivates GPX4 by reducing cysteine uptake, thereby limiting GSH synthesis, increasing lipid ROS, and ultimately leading to ferroptosis (Sato et al., 1999; Cao and DIXON, 2016). NRF2 has antioxidant elements and is regulated by Keap1. Its gene transcription is partially under the control of ROS. Sun et al. (2016a) found p62 expression prevents NRF2 degradation by Keap1 inactivation and enhances the subsequent nuclear accumulation of NRF2. They also demonstrate that NRF2-mediated anti-ferroptosis activity depends on the induction of NADPH (Reduced Nicotinamide Adenine Dinucleotide Phosphate) quinone oxidoreductase 1 (NQO1), heme oxygenase-1(HO-1), and ferritin heavy chain-1 (FTH1).

In morphology, ferroptosis mainly occurred in cells with reduced mitochondrial size, increased bilayer membrane density, and decreased or disappeared mitochondrial crest (Dixon et al., 2012; Yang and STOCKWELL, 2008; Yagoda et al., 2007). Mitochondria are the main source of ROS. Excessive ROS can cause significant oxidative stress and lead to cell and tissue damage (Czaja et al., 2013). Gao et al. (2019)showed that ROS derived from mitochondria are involved in cysteine deprivation induced ferroptosis. Li et al. (2021) found depletes cysteine can enhance sorafenib-induced ferroptosis and lipid ROS production, and increase oxidative stress and mitochondrial ROS accumulation. And they point out that sorafenib exerts its anti-HCC function partly by targeting the mitochondrial function. Huang et al. (2021a) found the use of ZZW-115 (Nuclear protein 1 inhibitor) induced ferroptosis and subsequent mitochondrial morphological changes, including the disintegration of mitochondrial network and severe mitochondrial metabolic disorders, which were compatible with the process of ferroptosis, and this process can be complementary to TFAM (a core mitochondrial transcription factor) (Zhao, 2019).

Iron is a redox-active metal that can participate in the formation of free radicals and the propagation of lipid peroxidation. Elevated iron levels increase susceptibility to ferroptosis. Iron overload or excessive activity of heme oxygenase 1 (HMOX1) increases the labile iron pool (LIP) that cause ferroptosis. Excessive iron increases ROS through Fenton reaction (through reaction with hydrogen peroxide (H₂O₂), ferrous iron (Fe²⁺) is oxidized into trivalent iron (Fe³⁺), forming highly active hydroxyl radical) (Hassannia et al., 2019), ROS is reversely neutralized by iron (Arefieva et al., 2021). Iron metabolism mainly involves the interaction between transferrin (TF) and its receptor (TFR), the input of iron through divalent metal transporter 1 (DMT1), the storage of iron as ferritin and iron-sulfur clusters (ISC), and the output of iron through iron transporter (FPN) (Abeyawardhane and LUCAS, 2019; Wang et al., 2019a).

The protection of the p62-Keap1-NRF2 signaling pathway on ferroptosis in HCC cells also involves the regulation of Fe homeostasis. An early study showed an increase in TFR1 and a decrease in ferritin (FTL and FTH1) expression in ferroptosis sensitive cells compared with iron-resistant cells (Yang and STOCKWELL, 2008). Sun et al. (2016a) showed that it was FTH1, not FTL or TFR1, that was regulated by NRF2 in ferroptosis. FTH1 inhibited ferroptosis by storing and transporting Fe²⁺ in HCC cells. In addition, excess iron in the liver may play a role in carcinogenesis by promoting tumor growth and altering the immune system (Kowdley, 2004). It is important to note that induction of ferroptosis in the liver may have different roles in tumorigenesis and cancer therapy.

Ferroptosis is iron-dependent regulatory necrosis induced by lipid peroxidation that occurs in cell membranes, a peroxidation reaction by polyunsaturated fatty acids catalyzed by the synthesis of acyl-CoA synthetase long-chain family member 4 (ACSL4) (Doll et al., 2017; Conrad and PRATT, 2019). Some polyunsaturated fatty acids (PUFAs) such as phosphatidylethanolamine (PE) and phosphatidylcholine (PC) are responsible for inducing ferroptosis by lipid peroxidation. Since de novo synthesis of PUFAs is strictly limited in mammals, various PUFAs are produced by the PUFAs biosynthesis pathway through the uptake of essential fatty acids from the blood and lymphatic fluid by cells. Free polyunsaturated fatty acids can be incorporated into cell membranes by various enzymes, such as ACLS4 and LPCAT3 (lysophosphatidylcholine acyltransferase 3), and lipid peroxidation can be induced by enzyme-induced and non-enzyme-induced mechanisms, resulting in ferroptosis (Lin et al., 2021). In this regard, knockdown of ACLS4, which preferably converts arachidonoyl (AA) to acylated AA, or loss of LPCAT3, which catalyzes the insertion of acylated AA into PLs (phospholipids), and make cells resistant to ferroptosis (Dixon et al., 2015; Yuan et al., 2016; Doll et al., 2017; Kagan et al., 2017). Magtanong et al. (2019) found that acyl-CoA synthetase long-chain family member 3 (ACSL3) converts monounsaturated fatty acids (MUFAs) into its acyl-CoA ester for incorporation into membrane phospholipids, thereby protecting cancer cells from ferroptosis. However, the levels of fatty acids (include MUFAs and PUFAs) in human serum are much higher than those in classical media containing fetal bovine serum (FBS), so how cells maintain the level of free fatty acid pools in cells is important to determine whether cells experience ferroptosis (Kamphorst et al., 2013; Magtanong et al., 2019).

Cellular energy metabolism is directly related to ferroptosis because it regulates antioxidant defense by mediating the synthesis of biological macromolecules and biological reductants such as NADPH (Zheng and CONRAD, 2020). Tumor cells typically exhibit upregulated glycolysis and PPP (pentose phosphate pathway) activity, which not only reduces ROS production by inhibiting mitochondrial respiration but also replenishes NADPH supply, thereby helps maintaining redox homeostasis to ensure cell survival. In energy metabolism, previous studies have reported that Cytochrome P450 oxidoreductase (POR) is a key mediator of ferroptosis, which promotes ferroptosis through the peroxidation of saturated phospholipids in cell membranes (Zou et al., 2020). Glucose 6-phosphate dehydrogenase (G6PD) is a key enzyme in PPP and plays a key role in NADPH production (Yang et al., 2019). G6PD may negatively regulate ferroptosis in HCC by regulating POR (Cao et al., 2021). Lu et al. (2018) pointed out that G6PD induces epithelial-mesenchymal transition (EMT) by activating the Signal Transducers and Activators of Transcription 3(STAT3) pathway, thereby promoting migration and invasion of HCC. Therefore, it can be concluded that disruption of tumor energy metabolism pathway not only changes the sensitivity of mutant tumor cells to ferroptosis, but also reduces their antioxidant defense ability to promote ferroptosis, and even affects tumor migration and invasion.

According to length and shapes, ncRNAs are divided into various types including microRNAs (miRNAs), PIWI-interacting RNAs (piRNAs), small nuclear RNAs (snRNAs), small nucleolar RNAs (snoRNAs), long ncRNAs (lncRNAs), circular RNAs (circRNAs), transfer RNAs (tRNAs), and ribosomal RNAs (rRNAs) (Wang et al., 2019b; Alzhrani et al., 2020). MiRNAs exhibit functions by binding to the 3′-untranslated regions of target mRNAs and suppressing their expression (Majidinia et al., 2020). MiRNA can regulate ferroptosis and control cancer progression by regulating GSH, iron levels, NRF2, and ROS. LncRNAs mainly act as the regulatory factors of transcription factors in the nucleus or as miRNAs of sponges in the cytoplasm to regulate ferroptosis (Wu et al., 2020). However, there were few studies on the relationship between ferroptosis and circRNA, tRNA, rRNA, piRNA, snRNA, and snoRNA. Studies have reported that the tRNA mutations in HCC leads to decreased expression of selenoproteins, except for GPX4 and GPX1 (glutathione peroxidase 1), and introduces some weak changes in ferroptosis (Kipp et al., 2013; Becker et al., 2014; De Spirt et al., 2016). The regulation of ferroptosis found in HCC about ncRNAs in recent years was sorted out in Table 1 and Figure 1. Wider and deeper studies are needed to explore the function of ncRNAs in ferroptosis.

Although the treatments have become more diversified in recent years, the average life expectancy of HCC was lagged far behind those of other cancers. The result of systemic chemotherapy has been particularly disappointing, not only because of the chemotherapeutic resistance of HCC, but also the severe results of major side effects, making the treatment of advanced HCC depends on the degree of underlying liver dysfunction, the burden of malignancy, and the patient’s general profile or expectations. Treatment options for advanced HCC are limited comparing to early HCC. In this context, several therapeutic agents have been developed over the past 50 years to provide better responses and improve the average life expectancy in patients with HCC. Some common chemotherapeutic agents in HCC are summarized in Table 2. However, In two randomized clinical trials of advanced HCC patients in stage III, Sorafenib, which is a commonly used chemotherapy drug, only increased overall survival by 2.8 and 2.3 months compared to the placebo, suggested limited effect to drug-resistant HCC in advanced HCC (Llovet et al., 2008; Cheng et al., 2009). Therefore, overcome the resistance of sorafenib and find more effective new drugs has become an urgency for advanced HCC patients and postoperative adjuvant chemotherapy patients. Different regulatory strategies and delivery routes have been proposed to enhance the antitumor activity of these drugs (Kodama et al., 2008; Hung et al., 2012; Li et al., 2013; Song et al., 2013). Although some ferroptosis inducers, for example, Erastin, are very effective in killing cancer cells in vitro, their pharmacokinetic properties, such as solubility and metabolic stability, are not suitable for the usage in vivo (Yang et al., 2014). It is now believed that sorafenib can induce a new type of regulated cell death-ferroptosis (Louandre et al., 2013), distinct from apoptosis, necrosis, and autophagy (Dixon et al., 2012), not only sorafenib, Guo et al. (2018) killed a variety of tumor cells with cisplatin, which can simultaneously cause apoptosis and ferroptosis. Wu et al. (2018) found that some ncRNAs affect the sensitivity of 5-Fu-resistant cells by regulating some key steps of ferroptosis.

In recent years, the adjustment of HCC-related chemotherapy resistance is shown in Table 3.

Radiation therapy is an important non-surgical treatment for cancer, but the clinical problems such as low efficacy and severe side effects remained unsolved. Gene therapy can synergistically increase the effect of radiation therapy through its antitumor mechanisms, which may reduce the dose. Radiotherapy induces ferroptosis by down-regulation of SLC7A11 and up-regulation of ACSL4, resulting in GSH production, increasing lipid synthesis, and subsequent oxidative damage (Lang et al., 2019; Lei et al., 2020). Studies have found that collectrin (CLTRN), as a target of radiation, is regulated by NRF1 (nuclear respiratory factor 1)/RAN (RAS oncogene family)/DLD (dihydrolipoamide dehydrogenase) protein complex and enhances the radiosensitivity of HCC cells through ferroptosis (Yuan et al., 2021). A combination of gene therapy and radiation therapy is one way forward, allowing the radiation doses to be reduced and the side effects to be reduced. It is worth considering whether the application of iron death inhibitors to non-tumor cells can increase their radiation tolerance to reduce the adverse effects of radiotherapy.

The use of nano drugs to induce ferroptosis will become a new anticancer strategy (Shen et al., 2018). More and more anticancer nano drugs have been approved by FDA, and the development of drugs with higher efficacy and safety will become an emerging road for future cancer treatment (Bobo et al., 2016). Tang et al. (2019) synthesized manganese-doped mesoporous silica nanoparticles (MMSNs) from manganese and silica. This reaction resulted in the inactivation of GPX4 and the increase of intracellular lipid peroxides through the consumption of intracellular GSH induced by the degradation of MMSNs. Ou et al. (2017) used natural omega-3 fatty acid docosahexaenoic acid (LDL-DHA) reconstructed into Low-density lipoprotein nanoparticles to selectively kill HCC cells. LDL-DHA induces ferroptosis by increasing tissue lipid hydroperoxide levels and inhibition of GPX4 expression. Tian et al. (2022) reported a novel cascade copper-based metal-organic framework (MOF) therapeutic nanocatalyst using HKUST-1 (a kind of metal organic framework) combining meloxicam (Mel), a cyclooxygenase-2 (COX-2) inhibitor, and sorafenib (Sol). Down-regulation of COX-2 induces PINK1/Parkin-mediated mitochondrial autophagy, chemodynamic Therapy (CDT) -mediated cytotoxic ROS, accumulated lipid peroxides (LPO) and Sol through inhibition system X^C−, the three interacted to activate ferroptosis and increase the sensitivity of HCC cells to chemotherapy. Liu et al. (2021a) constructed mil-101 (Fe) nanoparticles (NPs) loaded with sorafenib and iRGD (iRGD peptide (amino acid sequence: CRGDK/RGPD/EC) [MIL-101 (Fe) @ SOR], which co-administration significantly promoted the development of ferroptosis. Ma et al. (2017) enhanced the sensitivity of cancer cells to cisplatin by loading cisplatin prodrug onto iron oxide nanoparticles to increase ROS production. Du et al. (2021) designed an exosome with three parts, including surface functionalization of CD47, membrane loading of ferroptosis inducer Er (Erastin), and core of photosensitizer RB (Rose Bengal), and demonstrated potent antitumor therapeutic effects with surprisingly low toxicity.