HCC can be induced by OS-associated genetic changes, including both oxidative nuclear and mitochondrial DNA damage, DNA hypomethylation, and aberrant microRNA expression. ROS-induced DNA damage and genetic instability are crucial developments in HCC onset and advancement [27, 28]. 8-Hydroxydeoxyguanosine (8-OHdG), an oxidative DNA adduct, is a widely accepted marker for OS [29]. Ma-On et al. [30] studied the cellular response to OS and the level of DNA damage in human HCC tissues attributable to oxidation. They found that the 8-OHdG expression level was elevated in HCC tissues, particularly in HCC cell nuclei, and was linked to poor patient survival. In vitro experiments demonstrated that oxidative stress was induced by ROS, which substantially increased 8-OHdG expression levels in HCC cells. These outcomes indicate that ROS fosters an oxidative microenvironment in HCC cells, crucially promoting tumor progression.

Mitochondria can be impacted by various damaging factors, such as drugs, viruses, fat accumulation, and carcinogens [31]. When these factors exceed the mitochondria’s coping ability, abnormalities in mitochondrial structure and function occur, for example, ROS accumulation following a disruption in the ROS scavenging system [18, 32]. MtDNA, which lacks histone protection, is vulnerable to damage caused by ROS [33, 34]. Mitochondrial dysfunction contributes to ROS accumulation and mtDNA damage, potentially leading to HCC advancement. Shetty et al. [35, 36] reported that mitochondrial ROS levels doubled at the time of the disease progression of nitrosodiethylamine-induced HCC in mice, resulting in DNA damage and proto-oncogene activation, and ultimately promoting tumorigenesis.

A high-calorie diet can cause fat accumulation, leading to metabolic stress and abnormalities in liver mitochondria [16]. When the condition progresses to NASH, the adaptability and flexibility of the mitochondria are reduced. As previously described, OS, mitochondrial malfunction, and oxidative mtDNA may facilitate the progression of NASH to HCC. Xie et al. [37] introduced the hepatitis B virus x (HBx) gene into the human liver cell line HL7702 and discovered that, in the presence of OS conditions, HBx triggered the inflammasome NLRP3 in HL7702 cells, as well as promoted pyroptosis via the mtROS pathway under OS conditions. The binding of the HBV protein HBx to the mitochondrial outer membrane enhances its permeability, resulting in the breakdown of mitochondrial membrane potential and increased ROS generation [38]. In addition, HBx directly binds to cytochrome c oxidase, thereby disrupting the mitochondrial respiratory chain and inducing increased ROS production [39]. HCV induces OS in liver cells, resulting in increased levels of ROS, including O${}_2$${}^{\cdot -}$, OH${}^{\cdot }$, and H${}_2$O${}_2$ [40]. HCV core proteins inhibit the induction of ROS production through mitochondrial electron transfer, which further induces both mitochondrial damage and ROS formation by enhancing CYP2E1 expression [41].

DNA methylation represents a significant epigenetic process that modulates gene expression, and is linked to the onset and advancement of differing cancer types [42]. This mechanism involves the covalent bonding of a methyl group, catalyzed by DNA methyltransferases (DNMTs), to the carbon at the 5th position of cytosine residues within CpG dinucleotides [42]. ROS can hinder this mechanism resulting in global DNA hypomethylation [43, 44, 45]. A link between chromosomal region abnormalities and DNA hypomethylation at repetitive DNA sequences has been observed, suggesting that global DNA hypomethylation could trigger chromosomal instability, ultimately contributing to hepatocarcinogenesis [46]. OS can change DNA methylation status by influencing the function of other enzymes that maintain epigenetic status including histone methylases, and histone deacetylases (HDACs). Inducing OS in the HCC cells using hydrogen peroxide (H${}_2$O${}_2$) led to E-cadherin gene promoter hypermethylation stemming from enhanced Snail expression [47]. Snail triggers DNA methylation by recruiting HDAC1 and DNMT1 to sites of SNAIL occupancy. Human HCC also exhibit an association between ROS induction, E-cadherin downregulation, Snail upregulation, and E-cadherin promoter hypermethylation [47]. Recent findings revealed that increased ROS levels could upregulate Forkhead box C1 (FOXC1) expression via the ERK1/2-pELK1 pathway. FOXC1 subsequently elevates DNMT3B expression levels, resulting in DNA hypermethylation of the cystathionine $\gamma$-lyase (CTH) promoter and subsequent CTH gene silencing [48].

Micro RNAs (miRNAs) are responsible for regulating oncogenes and tumor suppressor genes in HCC, establishing a mechanistic link between epigenetics, inflammation, viral infection, and OS [49]. Both miRNAs and OS participate in the process of development of acute and chronic liver diseases, including viral hepatitis and oncogenesis, by affecting various signaling and metabolic pathways [50]. For example, miR483-3p may promote HCC metastasis via OS stimulation, uncovering a newly-discovered function of epigallocatechin-3-gallate for protection from miR483-3p–regulated HCC metastasis through the epigenetic modulation of miR483-3p [51]. Additionally, miR-33 deficiency improves mitochondrial function and reduces OS, and may represent an effective therapeutic approach for disease progression at different stages of HCC [52]. Overexpression of miR-92 and activation of telomerase activity are linked to the accumulation of ROS-mediated oxidative DNA damage during the development of chronic liver injury in HCC [53].

Numerous studies, including those on the Keap1/Nrf2/ARE, PI3K/AKT/mTOR, Wnt/$\beta$-catenin, and Notch pathways have demonstrated the significance of redox modifications of components of these signaling pathways in HCC pathogenesis.

2.2.1 Keap1/Nrf2/ARE Pathway

The Nrf2 protein is a critical regulator of antioxidant response [54]. Under normal physiological circumstances, Keap1 functions as a Nrf2 inhibitor, maintaining basal cytoplasmic Nrf2 levels through ubiquitin-proteasomal system degradation of Nrf2 [55]. Conversely, during stress conditions, and most notably OS, Nrf2 expression increases. Cysteine residues within the Keap1 intervening region domain serve as redox sensors, and upon their oxidation, Nrf2 is liberated from its regulatory complex [56]. Further, under OS, the Nrf2 protein stabilizes and initiates a multistep activation process, including nuclear translocation and heterodimerization with its partner, small v-maf avian musculoaponeurotic fibrosarcoma oncogene homolog (sMaf) proteins including MafG. Subsequently the complex recruits transcriptional co-activators and later attachment to target gene antioxidant response elements (AREs) [57]. The Nrf2-sMaf heterodimers occupy specific ARE sites within the genome, recruiting other transcriptional activators and inducing further Nrf2 transcription [58]. As Nrf2 shields cells from stress damage, it is formally classified as a tumor suppressor [59]; however, persistent oxidative stress during cancer development can result in Nrf2 hyperactivation, thus enhancing cancer cell survival [60] (Fig. 3).

Fig. 3.

Mechanism of the OS-mediated Keap1/Nrf2/ARE signaling axis in HCC. Under normal physiological circumstances, Keap1 binds to Nrf2 and the ubiquitin ligase culin3, and forms a cytoplasmic complex. Keap1 reduces the stability of the Nrf2 protein through the ubiquitin-proteasome degradation, thereby negatively regulating Nrf2 and maintaining low cellular levels of Nrf2. However, during OS, cysteine residues within Keap1 are oxidized, Nrf2 is freed from its regulatory complex, resulting in stabilization of the Nrf2 protein and translocation to the nucleus where it forms a heterodimer with sMaf. Subsequently, Nrf2-sMaf heterodimers occupy ARE in the genome and activate target genes which contribute to cancer development. For example, Nrf2-dependent activation of Bcl-xl and MMP-9 can enhance growth and invasiveness of cancerous hepatic cells. Further, increased expression of MRP1 and MRP2 mediated by Nrf2 can promote multidrug resistance in hepatocellular carcinoma. ARE, antioxidant response element; HCC, hepatocellular carcinoma; MMP, matrix metalloprotein; MRP, multidrug-associated protein; OS, oxidative stress.

In HCC, Nrf2 expression is associated with cell survival and various clinicopathological factors [61]. Nrf2 is abundantly expressed in cancerous cells and functions in tumor cell division and invasion. ROS significantly increases Nrf2 expression in liver cancer cells and creates an oxidative microenvironment in HCC, thus promoting further tumor development. The advancement of HCC due to ROS may be partially mediated through the activation of Nrf2 [62]. This suggests that stimulation of the Nrf2/Keap1/ARE signaling axis improves response to OS, which is advantageous for tumor cells in promoting liver cancer progression [30]. Thus, Nrf2 could be a potential prognostic indicator, and contribute to human HCC tumor cell proliferation and invasion partially by modulating Bcl-xL and matrix metalloproteinase-9 (MMP-9) expression [62] (Fig. 3).

The Keap1/Nrf2/ARE signaling axis is associated with chemotherapy resistance [62]. ROS-induced activation of Nrf2 controls sorafenib resistance in liver cancer cells [63]. Recent findings indicate that Nrf2 is overexpressed in chemoresistant HCC cells, thus contributing to the development of chemoresistance [64, 65, 66]. Ma et al. [67] demonstrated that Nrf2 expression was observed in three HCC cell lines, including HepG2, HepG3B, and SMMC-7721, and was associated with cisplatin (DDP) chemoresistance. Inhibiting dysregulated Nrf2 activation, particularly when combined with anticancer chemotherapeutics, is thought to be an effective strategy for overcoming multidrug resistance (MDR) and inhibiting tumor progression [68]. Nrf2 involvement in the MDR of HCC cells includes anti-apoptotic mechanisms and overexpression of Nrf2-regulated efflux transporters such as MDR-regulated proteins (MRPs), which enhance chemotherapy drug efflux and strengthen ROS defense systems. Niture et al. [69]reported that Nrf2 regulates Bcl-xl (an anti-apoptotic gene) expression by occupying a promoter in close proximity to Bcl-xl. This suggests that Nrf2 might contribute to anti-apoptotic protein expression and, thus, promote chemoresistance in HCC cells [69]. The expression of Nrf2-regulated MRPs, including MRP1 and MRP2, also increases, further contributing to chemoresistance in HCC cells [70] (Fig. 3). Gonzalez-Sanchez et al. [70] determined that Nrf2 activation reduced ROS levels stimulated by adriamycin, cisplatin, sorafenib, and SN-38 (an irinotecan metabolite with antineoplastic activity), leading to anti-apoptosis and reducing the effectiveness of chemotherapy against HCC. Nrf2 is upregulated in response to low doses of ROS, and Nrf2 has been shown to increase telomerase activity, decrease oxidative stress, and promote the survival of liver cancer cells through regulating human telomerase reverse transcriptase (hTERT) gene expression [71]. Mitochondrial hTERT promotes drug resistance of tumor cells by reducing ROS production and Mitochondrial DNA damage, and playing a protective role on mitochondrial respiratory chain [72]. Recent clinical immunohistochemistry analysis of hTERT expression in 135 liver cancer tissues found that hTERT is mostly expressed in the cytoplasm of these tissues and is directly connected to oxidative stress [73].

2.2.2 PI3K/AKT/mTOR Pathway

The PI3K/AKT/mTOR pathway is a key intracellular signaling pathway that participates in regulating various cellular mechanisms, including cell cycle progression, cell proliferation, death, metabolism, and angiogenesis. It is stimulated in different cancer types due to dysregulation of receptor tyrosine kinases (RTKs) [74]. RTK monomers serve as high-affinity cell surface receptors for growth factors, cytokines, and hormones. Upon ligand binding, RTK monomers activate and dimerize, triggering autophosphorylation and subsequent activation of the PI3K/AKT/mTOR signaling axis [75, 76]. Activated RTKs activate the lipid kinase PI3K, which catalyzes phosphorylation of phosphatidylinositol present within the internal leaflet of the plasma membrane. Direct activation of PI3K is achieved through binding to the regulatory region of RTKs or indirectly through adaptor molecules including insulin receptor substrates. Activated PI3K phosphorylates phosphatidylinositol-4,5-bisphosphate (PIP2) to form phosphatidylinositol-3,4,5-triphosphate (PIP3) [77, 78]. Subsequently, PIP3 interacts with phosphoinositide-dependent protein kinase 1 (PDK1) and AKT and recruits them to the plasma membrane. For AKT activation to be complete, Thr308 and Ser473 residues within the AKT regulatory domain must be phosphorylated by PDK1 and PDK2, respectively [79]. The tumor suppressor phosphate and tensin homolog (PTEN), is a phosphatase that catalyzes dephosphorylation of PIP3 to generate PIP2, which also regulates AKT. Two complexes, mTOR complex 1 (mTORC1) and mTOR complex 2 (mTORC2), are formed when several proteins bind to mTOR and activate it [80] (Fig. 4).

Fig. 4.

Transduction mechanism of OS-controlled PI3K/AKT/mTOR signaling in HCC. When ligands, including growth factors, cytokines, and hormones, bind RTKs the RTKs undergo autophosphorylation and form homodimers. The activated RTKs can subsequently recruit PI3K, undergo direct activation via the regulatory subunits that bind to the G-protein KRAS, or are indirectly activated by IRS. Activated PI3K can phosphorylate PIP2 to form PIP3 which can, in turn, recruit PDK1 and PDK2. PDK1 and PDK2 phosphorylate Thr308 and Ser473 of the AKT regulatory region, respectively, thereby phosphorylating AKT. PTEN can dephosphorylate PIP3 to block AKT activation, while HBx and HCV core proteins can downregulate PTEN through ROS-mediated pathways to activate the PI3K/AKT/mTOR signaling. AKT can promote the survival and drug resistance of liver cancer cells by activating mTOR1/2 and can enhance the growth and proliferation of cancerous hepatic cells by upregulating cyclin D1 and blocking BAD. Furthermore, AKT can also upregulate EMT-related proteins such as MMP-2 and 9, N-cadherin, and vimentin, promoting liver cancer cells migration and invasion. BAD, Bcl-2-associated death; EMT, epithelial-mesenchymal transition; HBx, hepatitis B virus x; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; MMP, matrix metalloproteins; mTOR, mammalian target of rapamycin; OS, oxidative stress; PTEN, phosphate and tensin homolog; RTK, receptor tyrosine kinase.

Stimulation of the PI3K/AKT/mTOR signaling pathway is linked proliferation, migration, invasion, and drug resistance of HCC cells [81, 82, 83]. The mTOR pathway is the major tumor-initiating pathway in HCC, and mTORC1 is upregulated in liver cancer, which is a possible pharmacological target for mitigating drug resistance in liver cancer [83]. Among the major factors for the abnormal stimulation of the AKT/mTOR signaling pathway is PTEN expression. Strong evidence for the crucial function of mTORC2 during liver tumor formation comes from how PTEN loss combines with c-Met to increase the development of HCC via the mTORC2 pathway [84].

The primary regulatory protein in HCC advancement is HBx, which is encoded by the HBV genome [85]. The oxidative inactivation of PTEN by HBx-stimulated ROS production activates the AKT pathway. Moreover, the positive regulatory loop that HBx and ROS maintain, which, through cyclin D1, contributes to hepatocellular carcinogenesis [86]. Hep3B cells and 293T cells overexpressing HBx stimulate AKT to phosphorylate I$\kappa$B kinase (IKK$\alpha$) and drive its nuclear translocation, encouraging tumor cells to invade and migrate [87]. HBx also activates PI3K/AKT, leading to phosphorylation and blockade of the Bcl-2-associated death promoter (BAD), a pro-apoptotic protein that induces an opening of the mitochondrial permeability transition pore (MPTP). Phosphorylation of BAD thus inhibits apoptosis and promotes the proliferation of liver cancer cells [88]. Moreover, HCV downregulates PTEN through ROS-mediated pathways and stimulates the PI3K/AKT signaling pathway to counteract the growth inhibition stemming from p53 activity, thereby enhancing tumor cell growth and proliferation [89]. Further, targeting PTEN to activate the PI3K/AKT/mTOR signaling axis can lead to sorafenib resistance in liver cancer cells [90]. Recent research by Bataller et al. [91] showed that HCV core protein expression promoted growth in hepatic stellate cells in a manner dependent on RAS/ERK and PI3K/AKT. Another HCC study determined that ROS mediates apoptosis via suppression of the PI3K/AKT/mTOR signaling pathway [92], which is associated with ROS concentration, confirming a dual function for OS in cancer (Fig. 4). In addition, ROS participates in the regulation of cancer telomerase activity through activation of AKT signaling by activating telomerase to enhance HCC malignant potential [93, 94].

HCC cell invasiveness induced by ROS is also believed to be mediated through the epithelial-mesenchymal transition (EMT), and is considered among the key regulatory factors for activating EMT [95]. Furthermore, EMT plays a significant part in HCC progression [96]. Stimulation of the PI3K/AKT/mTOR signaling pathway can induce EMT to promote HCC progression through MMP2, MMP9, N-cadherin, and vimentin [97] (Fig. 4).

2.2.3 Wnt/$\beta$-Catenin Signaling Pathway

The Wnt/$\beta$-catenin signaling pathway, which affects metastasis and treatment resistance, is another significant signaling pathway in HCC [98, 99, 100]. Tumor cell growth is caused by the ROS-dependent oxidation and inactivation of nucleoredoxin (NRX), a redox-sensitive regulator. This set of events is triggered by NADPH oxidase (NOX), stimulates the canonical Wnt/$\beta$-catenin pathway [101]. Low-density lipoprotein receptor-related protein 5/6 (LRP5/6) phosphorylation primes the signal, which activates and recruits dishevelled (Dvl), forming an assembly platform for the LRP6 signalosome. These events are followed by ligand binding to the heterodimeric receptor complex generated by Frizzled (Fz) and LRP5/6 located on the plasma membrane. Wnt, Fz, Axin, phosphorylated LRP5/6, GSK3, and CK1 are all signalosome components, and this complex eventually enhances the stabilization and cytoplasmic accumulation of $\beta$-catenin. Myc, cyclin D1, and Axin2 are examples of Wnt target genes that can be promoted by free $\beta$-catenin following its translocation into the nucleus. Once in the nucleus, $\beta$-catenin forms a complex with the lymphoid enhancer factor/T-cell factor (LEF/TCF) transcription factor complex [102, 103]. A destruction complex, which is made up of the scaffold protein Axin, APC, casein kinase 1 (CK1), and glycogen synthase kinase 3$\beta$ (GSK3$\beta$), closely modulates the level of $\beta$-catenin in the cytoplasm. When Wnt ligands are absent, the destruction complex binds to and phosphorylates $\beta$-catenin, which is then recognized by the E3 ubiquitin ligase $\beta$-transducing repeats-containing protein ($\beta$-TRCP) followed by proteasomal degradation of $\beta$-catenin [103] and inhibition of the Wnt/$\beta$-catenin pathway [104, 105]. Phosphorylation of GSK3$\beta$ is upregulated by ROS and this results in the inactivation of GSK3$\beta$ via stimulation of the PI3K/AKT signaling pathway (Fig. 5, Ref. [106]). Moreover, meningioma-associated protein 30 (MAC30) speeds up the growth and infiltration of HCC cells by modulating Wnt/GSK3$\beta$/$\beta$-catenin signaling [107]. Together, these findings demonstrate that the Wnt/$\beta$-catenin pathway and the redox status of HCC cells are subject to bidirectional modulation and interactions between these regulatory mechanisms function in HCC pathophysiology.

Fig. 5.

Transduction mechanism controlling OS-regulated Wnt/$\beta$-catenin signaling in HCC. The destruction complex is comprised of the proteins Axin, APC, CK1, and GSK3. When Wnt ligands are absent, the disruption complex captures and phosphorylates $\beta$-catenin present within the cytosol, which is subsequently identified by $\beta$-TRCP and targeted for proteasomal degradation. This prevents nuclear entry of $\beta$-catenin and thereby inhibits Wnt/$\beta$-catenin signaling. The presence of Wnt ligands can induce Fz and LRP5/6 to form heterodimer complexes, and phosphorylated LRP5/6 can activate and recruit Dvl. Phosphorylated Dvl, however, can recruit Axin leading to disruption of the complex, thus promoting $\beta$-catenin stabilization and cytoplasmic accumulation. Free $\beta$-catenin enters the nucleus to displace Groucho and /form a complex with the LEF/TCF transcription factor complex, promoting the activation of Wnt target genes such as c-Myc, cyclin D1, c-Met, MMPs. This also promotes EMT, cancer cell stemness, tumor growth, and drug resistance in HCC [106]. ROS can either cause phosphorylation and inactivation of GSK3$\beta$ through the PI3K/AKT pathway or inhibit Dvl by oxidizing and inactivating nucleoredoxin; thus, ROS can bidirectionally regulate the Wnt/$\beta$-catenin pathway and redox status in HCC cells. Fz, frizzled; Dvl, dishevelled; HCC, hepatocellular carcinoma; LEF, lymphoid enhancer factor; MMP, matrix metalloprotein; OS, oxidative stress; ROS, reactive oxygen species; TCF, T-cell factor; $\beta$-TRCP, $\beta$-transducin repeats-containing protein.

The activation of EMT is a crucial step for HCC cells to acquire a malignant phenotype [108]. Among the primary signaling pathways that control the onset of EMT is the Wnt/$\beta$-catenin pathway and this signaling axis is crucial for the modulation and maintenance of HCC stemness [109] (Fig. 5). A negative correlation exists between Wnt activity and T-cell signatures, specifically, by enhancing CD8+ T-cell infiltration in HCC, blocking Wnt/$\beta$-catenin can improve anti-PD-1 immunotherapy [110]. In this way, inhibiting the Wnt pathway can make HCC cells more sensitive to chemotherapy [106].

ROS stimulates several transcription factors such as hypoxia-inducible factor-1 alpha (HIF-1$\alpha$), forkhead box O (FOXO), heat shock factor 1 (HSF1), NF-$\kappa$B, Nrf2, and p53. Activation of these molecules subsequently regulate the cellular redox state [119]. HCC is regulated by FOXO, HSF1, NF-$\kappa$B, and HIF-1$\alpha$ under OS conditions.

FOXO family members including FOXO1, FOXO3, FOXO4, and FOXO6 are involved in supporting cellular homeostasis through various mechanisms [120]. These molecules enhance antioxidant status by inducing genes that eliminate ROS, improve mitochondrial redox, and inhibit levels of free transition metal ions by increasing metallothionein and ceruloplasmin abundance [121]. FOXOs respond to ROS, nutritional status, as well as genes implicated in apoptosis and cell cycle arrest like GADD45 [120]. ROS can activate AKT to directly phosphorylate FOXOs, which leads to the expulsion of FOXOs from the nucleus and suppression of their transcriptional activity. Inhibition of FOXO transcriptional activity could trigger apoptosis in HCC cells [122]. Since FOXOs cause cell cycle arrest and apoptosis, they were once characterized as tumor suppressors; however, mounting evidence indicates that FOXOs, particularly FOXO3, actually promote carcinogenesis. According to Lu et al. [123], FOXO3 expression accelerated carcinogenesis caused by hepatotoxin and the positive feedback loop between AKT and mTORC2 activation was activated when FOXO3 induced OS and DNA damage. It bears consideration that FOXO3, triggered by both ROS-promoting and ROS-eliminating systems, is linked to the activation of the pentose phosphate pathway (PPP).

Hypoxic response is largely regulated by hypoxia inducible factor-1 alpha (HIF-1$\alpha$), a transcription factor that is commonly expressed in numerous carcinomas, particularly HCC [124]. Low levels of oxygen result in an unbalanced flow of electrons in the electron transport chain which leads to the generation of ROS [125]. HIF-1$\alpha$ is triggered in response to ROS accumulation and permits the transcription of genes that improve glycolysis within anaerobic environments [126, 127]. For example, in hypoxic HCC cells, mitochondrial ubiquinol-cytochrome c reductase complex assembly factor 3 (UQCC3) creates a positive feedback loop with ROS generated in the mitochondria, which sustains UQCC3 expression and the production of ROS. As a result, mitochondrial structure and function is preserved and HIF-1$\alpha$ improves glycolysis [128]. The transcriptional factor and stem cell marker NANOG facilitates the maintenance of stem cell capacity for self-renewal, pluripotency, and an undifferentiated phenotype [129], and NANOG can modulate HIF-1$\alpha$. Of note, HIF-1$\alpha$ or Toll-like receptor 4 (TLR4) can activate NANOG [130, 131]. In HCC, NANOG activation through TLR4 increases fatty acid oxidation and suppresses mitochondrial oxidative phosphorylation. These processes result in the reduction of oxygen consumption rates and ROS formation [131]. This scenario further induces drug resistance and the maintenance of tumor-initiating stem-like cell capacity for self-renewal which can be resolved by restoring oxidative phosphorylation and suppressing fatty acid oxidation [131].

Heat Shock Transcription Factor 1 (HSF1) responds to protein misfolding–induced stressors through induction of genes encoding heat shock protein chaperones. HSF1 specifically responds to ROS via oxidation of internal Cys-35 and Cys-105 residues, and subsequent stimulation of antioxidant gene expression [132, 133]. Upregulated HSF1 expression commonly occurs to protect tumor cells against diverse types of stresses [134] and increased expression of HSF1 activates a metabolic phenotype switch and chemoresistance in cancer cells. For example, HSF1/AMP-activated protein kinase $\alpha$ (AMPK$\alpha$2)–mediated alterations in the metabolic phenotypes confer enhanced oxaliplatin resistance in HCC cells [135]. In HSF1-knockdown HCC cells, glucose intake, lactate generation rates, and intercellular ROS levels were found to be reduced [136]. These findings indicate a positive correlation between ROS levels and HSF1 expression. Furthermore, liver carcinogenesis has been linked to the overexpression of inducible nitric oxide (NO) synthase and HSF-1, thus promoting tumor advancement [137].

By triggering cytokines, chemokines, and receptors that promote tissue healing, NF-$\kappa$B modulates an adaptive immune response aimed at eliminating invasive infections and NF-$\kappa$B also controls the expression of antioxidant genes [138]. By upregulating anti-apoptotic genes, cyclins, MMPs, cell adhesion, and pro-angiogenesis genes, NF-$\kappa$B facilitates tumor cell survival, growth, and metastasis [139]. NF-$\kappa$B also helps shift metabolism towards glycolysis and controls the TME by modulating primary immune cell functions [140]. NF-$\kappa$B serves as an anti-apoptotic factor, and multiple cancer cell types frequently exhibit dysregulated NF-$\kappa$B activity [141]. In HCC, ROS produced by the mitochondrial respiratory chain activates NF-$\kappa$B, which promotes tumor development and metastasis [142]. Cellular ROS levels are influenced by the transcriptional activity of NF-$\kappa$B-dependent genes, and NF-$\kappa$B levels are likewise influenced by ROS [138]. Thus, ROS can not only activate NF-$\kappa$B signaling but also inhibit NF-$\kappa$B. For example, NF-$\kappa$B activity is both inhibited by activated by mitochondrial ROS, which induces apoptosis in HCC cells [143].

P53 increases the antioxidant status of cells via transactivation of genes which encode proteins that scavenge ROS, support GSH production, enhance NADPH synthesis, detoxify xenobiotics, and promote pro-oxidant enzymes such as NO synthase 2 (NOS2) and cyclooxygenase 2 (COX2) [144, 145]. Improvement in NADPH synthesis triggered by p53 occurs as a result of upregulation of the p53-induced glycolysis and apoptosis regulator TIGAR, which promotes DNA repair and reduces ROS to maintain redox balance [146]. Conversely, p53 also exhibits pro-oxidant activities through upregulation of p53-inducible genes (PIGs), such as PIG3 a quinone oxidoreductase/$\zeta$-crystallin synthesizing ROS via redox-cycling quinones, and p67phox which triggers formation of the NOX2 complex [147, 148]. P53 provides defense against cancer by regulating cell cycle arrest, senescence, and apoptosis [149].

P53 promotes antioxidant genes under moderate OS conditions to aid in adaptability; however, under harsher circumstances, p53 modulates cancer cell death by promoting the generation of ROS. PIG3 and PIG6 are two p53 target genes that promote apoptosis, and mitochondrial proline dehydrogenases generate ROS via a mechanism that supplies carbon to various other mitochondrial dehydrogenases, which in turn produce ROS [148, 150]. Genes encoding BAX, PUMA, and p66Shc are induced when p53 is activated by high levels of ROS. These genes disrupt the functioning of mitochondria, release cytochrome c, and enhance the generation of ROS [151]. Moreover, SOD2 and numerous Nrf2 target genes are repressed when p53 is active in proapoptotic scenarios [152]. OS induces changes in p53 activity and can upregulate p53 in HCC cells thereby inducing apoptosis of cancerous hepatic cells and inhibiting tumor proliferation and growth [153]. Furthermore, chronic infection with HBV and HCV contributes to ROS and RNS production, which can cause mutations in the p53 gene and subsequent promotion of liver cancer [154].

HCC is a form of cancer associated with inflammation, and most cases of HCC occur due to liver damage and chronic inflammation [155]. Prolonged and unresolved inflammation leads to the infiltration of immune cells into the liver which facilitates tissue remodeling [156]. The infiltration of immune cells results in a disturbance in the generation of chemokines and cytokines and a rise in the synthesis of ROS and RNS, which promote the development of fibrosis, cirrhosis, and, ultimately, the malignant transformation of liver cells [157]. The genesis, development, and therapeutic response in HCC tumors are all influenced by TME remodeling. By influencing various factors such as tumor-associated macrophages (TAMs), neutrophils, myeloid-derived suppressor cells (MDSCs), and Treg cells within the TME, OS plays a prominent role in modulating HCC development and progression.

TAMs within the TME mainly exert their function through secreted factors. Such factors include epidermal growth factor (EGF), cytokines (IL-6 and TNF-$\alpha$), chemokines (CXCL17 and CCL24) which directly regulate the growth and metastasis of tumors [158]. TAMs can also increase the expression levels of IL-10 and transforming growth factor $\beta$ (TGF-$\beta$) within the TME as well as producing stromal and tumor cell activators such as EGF, basic fibroblast, vascular endothelial, platelet-derived growth factors, and TGF-$\beta$, all of which can regulate tumor development [159]. These findings suggest an interaction between factors secreted by TAMs and the TME that influence tumor progression. In hepatocytes, ROS are primarily produced by the mitochondria and released by complex I and/or complex III of the respiratory chain in response to TNF-$\alpha$ [160]. TNF-$\alpha$–triggered synthesis of ROS stimulates NF-$\kappa$B, ultimately activating pro-survival signals, causing an increase in cell migration. OS induces TAMs to release cytokines such as TNF-$\alpha$, IL-6, IL-10, and TGF-$\beta$ which lead to HCC progression [104, 161, 162]. ROS production is critical for macrophage differentiation and continuous administration of the ROS inhibitor nanoliposome-loaded C6-ceremide (LipC6) in mouse models of hepatic malignancy effectively prevents TAM accumulation and significant inhibition of carcinogenesis [163].

Neutrophils are also becoming recognized as a significant player in the pathophysiology of HCC. Current evidence indicates that neutrophils are key mediators of the immunosuppressive environment as well as being drivers of tumor progression [164]. Growing evidence indicates that neutrophils function during the initial phases of carcinogenesis because they directly interact with HCC cells, mostly by producing ROS which leads to genomic instability [165]. S100A9, which belongs to a class of proteins termed damage-associated molecular patterns (DAMPs), displays upregulated expression when induced via HBV and has a direct impact on HCC cells by facilitating HCC progression and metastasis [166]. S100A9 affects neutrophil recruitment in acute and chronic hepatic damage in a mouse model, as well as promoting neutrophil stimulation and degranulation [166]. S100A9, which is also regulated by HBV, has an indirect impact on HCC malignancy by promoting the proliferation of neutrophil extracellular traps and activating neutrophils.

MDSCs induce the differentiation and expansion of Tregs during tumorigenesis, as well as producing OS [167]. The expansion of MDSCs can suppress immune function, as demonstrated by increased levels of arginase, NO, and ROS [168]. MDSCs also exhibit consistent alteration following a rise in the synthesis of ROS in response to therapy with Pam3CSK4 and Ro5-3335 [169]. Maintaining the immunological TME is crucial for MDSCs and promoting their polarization will enhance cancer immunotherapy. The immunosuppressive TME that enhances tumor onset and advancement can be alleviated by preventing the formation and recruitment of MDSCs within the liver. For example, long-term metformin therapy can prevent the accumulation of MDSCs in Ncoa5${}^{+/-}$ mouse livers displaying NASH, along with a decreased risk of developing HCC [170].

Factors involved in chronic liver disease influence Treg cells. Alcohol, high-fat diet, gut microbiota, and metabolites are some examples of environmental stressors that can promote hepatic damage [18]. OS considerably contributes to the downregulation of Treg stimulation and their population, resulting in steatosis and fibrosis [171]. The progression of NASH toward tumor development increases the population of Tregs to establish a pro-tumorigenic TME. The development of NAFLD-linked HCC is aided by Tregs and neutrophils facilitating communication between the innate and adaptive immune systems. In work conducted by Hang et al. [172], isoallo-LCA increases FOXP3 expression levels which in turn triggers mitochondrial ROS generation and encourages Treg development. Fe-MnO${}_2$/dihydroartemisinin (DHA) undergoes degradation within the TME, which induces ROS accumulation and subsequent triggering of ferroptosis and apoptosis of hepatoma cells [173]. DHA also acts as an immunomodulator to inhibit Tregs and systemic antitumor effects. Taken together, Fe-MnO2/DHA offers a multimodal treatment for HCC that is triggered by ferroptosis, apoptosis, and immune activation and which substantially promotes synergistic chemotherapy for cancer.

These results suggest that a rise in ROS triggers alterations in the TME that promote carcinogenesis by changing the roles of MDSCs and TAMs and, coordinately, generating alterations in Tregs that may inhibit immunological reactions against tumor cells.

Chemotherapeutics cause OS, inflammation, apoptosis, and aberrations in neurotransmitter metabolism, all of which contribute to their toxicity. Oxazoline anthracyclines have drawn interest as potential signaling mediators of proliferation, differentiation, and death of cancer cells because they produce ROS and RNS [229]. High amounts of ROS are produced by platinum-containing complexes, alkylating agents, camptothecins, and arsenic agents. In contrast, low quantities of ROS are produced during response to taxanes, vinca alkaloids, nucleotide analogs, and antimetabolites such as antifolates and nucleosides [230]. 1-Methylpropyl 2-imidazolyl disulfide (PX-12), a thioredoxin 1 (Trx1) inhibitor, has been studied in various cancer types [231, 232, 233]. PX-12 therapy upregulates Bax expression and downregulates Bcl-2 expression, indicating that PX-12–mediated apoptosis occurs in a mitochondrion-dependent manner. PX-12 also synergizes with 5-fluorouracil to significantly inhibit tumorigenicity both in vitro and in vivo [192]. The aminopeptidase N inhibitor 4cc synergistically acts with 5-fluorouracil to exert antitumor effects on cancerous human hepatocytes though ROS-regulated drug resistance, inhibition, as well as simultaneous triggering of the apoptotic mitochondrial pathways [193]. In addition, itraconazole exposure activates apoptosis, blocked cell cycle advance, and downregulated MMP in HepG2 cells [194]. Additionally, a link between intracellular ROS production and accelerated senescence triggered by cisplatin could be utilized as a prospective HCC target [234].

HCC has been treated with various natural chemical extracts and Chinese herbal medications. For example, alnustone increased the synthesis of ROS in BEL-7402 and HepG2 cells, and downregulated proteins linked to apoptosis and PI3K/AKT/mTOR/p70S6K signaling [92]. Coptisine promotes autophagic cell death via the downregulation of proteins linked to PI3K/AKT/mTOR signaling and upregulation of ROS-regulated mitochondrial dysfunction in Hep3B cells [181]. Koumine suppresses HCC proliferation and enhances apoptosis. NF-$\kappa$B and ERK/p38 MAPK pathways both improve koumine activity through a mechanism that is dependent on ROS [209]. In HCC cells, sanguinarine causes lysosomal dysfunction, ROS-mediated mitophagy, and death [210]. In HCC cells, artesunate elevated ROS levels which, in turn, enhanced Bax/Bcl-2 and the triggering of apoptosis [211]. A novel derivative of curcumin, 3,5-(E)-Bis(3methoxy-4-hydroxybenzal)-4-piperidinone hydrochloride (C0818), has shown potency in inhibiting Hsp90 and exhibits good antitumor effectivity. Through a mitochondrially-regulated pathway, C0818 caused HCC cells to undergo apoptosis in a manner dependent on ROS and caspases [208]. The HCC-derived cell lines HepG2 and BEL-7402 exhibited a substantial growth inhibitory response following exposure to giganteaside D (GD). Moreover, in HCC cells GD also triggered ROS-mediated apoptosis [204]. Co-incubation of apigenin with 5-fluorouracil synergistically elevated the levels of ROS in HCC cells [207]. However, some therapeutic drugs such as decalactone, daphnetin, and genistein inhibit HCC proliferation by reducing OS [221, 222, 223]. Most of these agents can prevent tumor development by reducing inflammatory cytokines and boosting antioxidant enzyme abundance.

Due to their distinctive characteristics, including its significant drug-loading capacity, inherent anticancer actions, integrated diagnostic and therapeutic functions, and ease of surface engineering with targeting ligands, nanomedicines have received significant attention when attempting to discover and develop effective therapies for HCC [235]. Bevacizumab and CCR2 inhibitor-containing nanoparticles sensitize doxycycline-treated Huh-7 cells by activating ROS-stimulated apoptosis [224]. Lipid-coated cisplatin/oleanolic acid calcium carbonate nanoparticles (CDDP/OA-LCC NPs) were reported to induce tumor cell apoptosis via upregulation of proapoptotic proteins, downregulation of proteins associated with the P13K/AKT/mTOR pathway, and upregulation of proteins linked to p53 proapoptotic functions [226]. Cisplatin and curcumin-loaded nano-liposomes (CDDP/CUR-Lip) act by elevating ROS levels in HCC cells [225] and can be used to co-deliver CDDP and CUR in vitro and in vivo to improve HCC treatment through synergistic effects with minimal toxicity.

Silver nanoparticles (AgNPs) have recently come to light as a cutting-edge method of treating tumors, particularly hepatocarcinoma. AgNP-induced apoptosis is dependent on ROS overproduction as well as the effects of MAPKs, AKT signaling, and DNA damage-mediated p53 phosphorylation to trigger HepG2 cell apoptosis [227]. ZnO nanoparticle–ferulic acid conjugate (ZnONPs-FAC) triggers cells to undergo apoptosis and results in the suppression of diethylnitrosamine (DEN)-induced HCC. ZnONPs-FAC can cause significant mitochondrial damage and generate ROS production, and it can also induce oxidative DNA damage in HCC cells [228]. There is little doubt that nanomaterials will play a significant role in treating HCC when moving forward, particularly when paired with therapeutic drugs.