Multifaceted roles of aerobic glycolysis and oxidative phosphorylation in hepatocellular carcinoma

Introduction

Cancer is a leading cause of death worldwide and an important barrier to increased life expectancy (Bray et al., 2021). Liver cancer is one of the most common malignant tumours, causing an estimated 906,000 new cases and 830,000 deaths annually worldwide. Primary hepatic carcinoma was the sixth most common cancer, third leading cause of cancer-related deaths worldwide, and leading cause of cancer-related deaths in China among men in 2020. Among primary hepatic carcinoma, hepatocellular carcinoma (HCC) accounts for 75–85% of all cases (Sung et al., 2021). The highest rates of HCC have been documented in East Asia and Africa, with more than one million annual cases projected by 2025 (Llovet et al., 2021). As a highly invasive tumour, HCC is associated with poor prognosis. Although immune checkpoint inhibitors have some therapeutic effect, patient survival time remains short and treatment options are limited.

Mitochondria participate in a variety of metabolic processes, including OXPHOS, calcium and iron homeostasis, apoptosis, and reactive oxygen species production (Andreyev & Starkov, 2005; Boese & Kang, 2021). Their dysfunction is closely related to the occurrence and development of human cancers (Kim, Maiti & Barrientos, 2017). Senescent cells exhibit tumour-friendly behaviour, as evidenced by an increase in mitochondrial calcium load, oxygen consumption rate (OCR), and OXPHOS. These traits help cells escape oncogene-induced senescence (Farfariello et al., 2022). Decades ago, Otto Warburg observed that cancer favoured glucose fermentation even in the presence of oxygen, suggesting that defects in mitochondrial respiration may be the underlying cause of cancer (Warburg, 1956a; Warburg, 1956b). To meet their demand for rapid growth, tumour biomacromolecules can be synthesised using intermediate metabolites generated during aerobic glycolysis (Vaupel, Schmidberger & Mayer, 2019). However, not all tumours share this property of aerobic glycolysis. Furthermore, it is now apparent that defects in mitochondrial respiration are generally not the cause of aerobic glycolysis, nor are they specifically selected during tumour evolution (Zong, Rabinowitz & White, 2016). Several studies suggest that the role of mitochondria in cancer is more complex than that envisioned by Warburg. The mitochondrial 1C enzymes serine hydroxymethyltransferase 2 (SHMT2) and methylenetetrahydrofolate dehydrogenase 2 (MTHFD2) are among the most transcriptionally upregulated genes in cancer (Nilsson et al., 2014; Vazquez, Tedeschi & Bertino, 2013). Tumorigenesis can be impaired by inhibiting the replication of mitochondrial DNA (Tan et al., 2015). Inactivation of the mitochondrial transcription factor Tfam, which depletes mitochondria from tumour cells, disrupts the growth of K-ras lung tumours (Weinberg et al., 2010). These studies emphasise the importance of mitochondrial function in tumour growth.

The liver is not only the largest detoxifying organ but is also critically involved in controlling energy metabolism (Piccinin, Villani & Moschetta, 2019), with the mitochondria as the primary sites of metabolism. Non-alcoholic fatty liver disease, alcoholic fatty liver disease, viral hepatitis, and HCC are closely associated with mitochondrial dysfunction (Zhang et al., 2022; Middleton & Vergis, 2021; Zhang et al., 2019). Reprogramming of mitochondrial metabolism is inextricably linked to the growth and development of HCC (Wang et al., 2021; Wang et al., 2016; Lee et al., 2021). In addition to aerobic glycolysis, OXPHOS plays a key role in HCC cell survival and progression. Unlike for glycolysis, the therapeutic targeting of mitochondrial OXPHOS has received little attention. Based on the most recent findings, this review analyses how aerobic glycolysis and OXPHOS affect the formation, progression, and drug resistance of HCC. Our study makes a significant contribution to the literature, given the global importance of liver cancers and the general lack of comprehensive and tenable theories regarding cancer metabolism. Further, we believe that this article will be of interest to the readership of oncologistsas it highlights the potential avenues for treatment of HCC based on crucial molecules involved in metabolic processes.

Survey Methodology

Relevant published articles were identified from Pubmed, Scopus, Web of Science, China National Knowledge Infrastructure using the search term of “hepatocellular carcinoma”, “mitochondrial metabolism”, “oxidative phosphorylation”, and“aerobic glycolysis.” We performed by crossing these descriptors using the boolean operators “OR” and “AND”. We mainly included relevant articles from 2015 to 2022. Articles related to pyrimidine, amino acid, and phospholipid metabolism with mitochondria and articles not related to hepatocellular carcinoma were excluded. The resulting articles were included where it considered aerobic glycolysis and oxidative phosphorylation in hepatocellular carcinoma. We made a comprehensive interpretation and analysis. To discuss the complex relationship between the two, we added the search term of “L-lactate”. In the process of summarizing the literature on the biological behavior of hepatocellular carcinoma, we added key search terms such as “autophagy” and “angiogenesis” Regarding drug resistance of HCC, we added the search term “sorafenib”. Briefly, the review identified 151 relevant research articles from research labs working on aerobic glycolysis and oxidative phosphorylation in hepatocellular carcinoma.

Aerobic glycolysis during HCC progression and metastasis

In the 1920s, Warburg, Posener & Negelein (1924) found that glycolysis occurred even in the presence of sufficient oxygen in rat liver cancer. Considerable research on aerobic glycolysis in HCC has provided further insights. The ATP produced by aerobic glycolysis may enable HCC to adapt to an energy-deficient tumour microenvironment (Xu & Herschman, 2019). Macromolecules essential for the proliferation of HCC can also be produced through aerobic glycolysis (Heiden, Cantley & Thompson, 2009; Ganapathy-Kanniappan, 2018). Hence, the onset and progression of HCC are associated with enhanced glycolysis (Feng et al., 2020b).

Enhanced expression of aerobic glycolysis enzymes in HCC

Aerobic glycolysis is primarily regulated by hexokinases (HKs), phosphofructokinases (PFKs), and pyruvate kinases (PKs), and the overexpression of these enzymes is associated with a poor prognosis in HCC (Feng et al., 2020b). In this section, we discuss regulators of key glycolytic enzymes to highlight novel therapeutic targets for the treatment of HCC.

The methodological principle is illustrated by a scheme shown in (Fig. 1).

HK2

Overexpression of HK2 leads to a poor prognosis of HCC (Zhang et al., 2016), and increased HK2 levels significantly enhance HCC propagation (Li et al., 2022). Several upstream factors, including Astragalin (Li et al., 2017c), Quercetin (Wu et al., 2019), Caveolin-1 (Chai et al., 2019), and Forkhead box 1(FOXK1) (Cui et al., 2018) regulate HK2 mRNA and protein levels in HCC. Furthermore, FOXK1 and Quercetin act on HK2-dependent glycolysis via the AKT/mTOR pathway (Wu et al., 2019; Cui et al., 2018). Moreover, HK2 binds to voltage-dependent anion-selective channel protein 1(VDAC1) to protect it from the inhibitory effects of downstream products, thereby enhancing glycolysis (Mathupala, Ko & Pedersen, 2009).

The androgen receptor potentiates protein kinase A/cyclic adenosine monophosphate response element-binding (CREB) protein signalling to promote HK2 expression, ultimately promoting HCC cell glycolysis and growth (Sun et al., 2021). This might explain why liver cancer ranks second in male mortality (Sung et al., 2021). Moreover, peroxisome proliferator-activated receptor gamma(PPAR γ) and hypoxia-inducible factor-1 α (HIF-1 α) might contribute to HK2 induction in fatty liver disease and its evolution towards cirrhosis and carcinogenesis (Panasyuk et al., 2012; Mesarwi et al., 2016). Thus, HK2 up-regulation correlates with hepatic diseases progression regardless of cause. (Xu & Herschman, 2019).

PFK1

PFK-1 catalyzes the conversion of fructose 6-phosphate(F-6-P) to fructose 1, 6-bisphosphate(F-1,6-P) and is inhibited by high ATP and phosphoenolpyruvate levels and activated by fructose-2,6-bisphosphate(F-2,6-P) (Okar et al., 2001). PFK-1 is also regulated by posttranslational modifications, such as glycosylation and acetylation (Yi et al., 2012; Feng et al., 2020c). Both hypoxia and glucose deficiency can induce PFK1 glycosylation. O-linked β-N-acetylglucosamine (O-GlcNAc) of PFK1 directly regulates glycolysis to increase cancer cell growth (Yi et al., 2012). Tumor cells express PFK-1 isoforms at high levels to attain a persistent change in glycolytic flux (Ros & Schulze, 2013). In addition, the E3 ubiquitin ligase A20 (A20) promotes the ubiquitination of liver-type phosphofructokinase 1 (PFKL) to downregulate aerobic glycolysis (Feng et al., 2020c). PFKL can mediate multienzyme assembly for glucose metabolism (Kohnhorst et al., 2017), with its tetrameric form affecting glycolysis more stably (Bartrons et al., 2018). The downstream products of PFK1 induce the dissociation of PFK1 tetramers to dimers, leading to the inhibition of PFK1 activity (Wu et al., 2020). In addition, the enzymes 6-phosphofructo-2kinase/fructose-2,6-bisphosphatase-3 (PFK2/PFKFB3) is involved in liver fibrogenesis (Mejias et al., 2020), activates PFK1, and stimulates high glycolytic flux in human cancers (Bando et al., 2005). Metformin inhibits the expression and activity of PFK1 by suppressing PFKFB3 (Zeng et al., 2019). Overall, PFK-1 plays a critical role in the glycolytic pathway.

PKM2

PKM2 is ubiquitously expressed in tumour tissues and regulates cancer cell metabolism (Yang & Lu, 2013; Mazurek, 2011). The expression of PKM2 is controlled by several signaling pathways and transcription factor. HIF-1 α regulates HCC cell glycolysis, and its high expression corroborates the Warburg effect (Feng et al., 2020b). HIF-1 α is the primary upstream regulator of PKM2 in HCC (Feng et al., 2020a), and induces the expression of L-lactate dehydrogenase(L-LDH) and pyruvate dehydrogenase kinase, which promote glycolytic conversion (Kelly & O’Neill, 2015). Heat shock protein 90 (HSP90) promotes the overexpression of HIF-1 α (Liu et al., 2016). In addition, HSP90 also directly regulates the expression of PKM2. Xu et al. (2017) suggested that HSP90 could induce the Thr-328 phosphorylation of PKM2 and increase the abundance of PKM2 in HCC. Thus, the HSP90/HIF-1 α/PKM2 axis has an important effect on aerobic glycolysis. The following reflects another way in which HIF-1 α affects PKM2. HIF-1 α knockdown attenuates the effect of LNCAROD (a newly identified Long non-coding RNA) (Ge et al., 2021). LNCAROD increases PKM2 levels by either combining with serine- and arginine-rich splicing factor 3 (SRSF3) to trigger PKM switching from PKM1 to PKM2, or by sponging miR-1455p to upregulate PKM2 (Jia et al., 2021). However, HCC chemoresistance is also strengthened by LNCAROD overexpression (Jia et al., 2021), although no clear mechanism has been suggested. Additionally, the c-myc-hnRNP A1 pathway has been shown to regulate PKM splicing, increasing expression of PKM2 (David et al., 2010). In HCC, E3 ligase zinc finger protein 91 (ZFP91)—a tumour suppressor gene—regulates alternative splicing of the glycolytic enzyme PKM pre-mRNA through hnRNP A1 to reduce the expression of PKM2 (Chen et al., 2020). PPAR γ also contributes to the pathogenesis or development of hepatic diseases through PKM2 (Panasyuk et al., 2012). Thus, PKM2 activity is regulated by numerous allosteric effectors. Accordingly, targeting PKM2 can be a new therapeutic approach for HCC.

Two primary effects of OXPHOS in HCC

The complex relationship between aerobic glycolysis and OXPHOS HCC

OXPHOS levels are strongly correlated with the levels of L-lactate and pyruvate. The p53 up-regulated modulator of apoptosis (PUMA)-mediated apoptotic response in hepatocytes is a direct cause of compensatory proliferation and ensuing carcinogenesis (Qiu et al., 2011). Wild-type p53 (WTp53) promotes glycolysis by activating PUMA and inhibiting pyruvate-driven OXPHOS in HCC patients (Kim et al., 2019). Moreover, overexpression of PTEN restores the inactivated PI3K/Akt pathway, significantly inhibits L-lactate production, and enhances access of pyruvate to mitochondria (Zhao et al., 2020). Hence, efficient glycolysis promotes L-lactate accumulation, thereby ensuring a reduced mitochondrial pyruvate intake and preventing normal OXPHOS processes.

However, glycolysis seems to play a different role in HCC cells that depend on OXPHOS for survival: L-lactate, but not pyruvate, is thought to be the end product of glycolysis and the main substrate of the mitochondrial TCA cycle (see Fig. 2) (deBari & Atlante, 2018; Schurr, 2018). L-lactate appears to be the favoured substrate during high metabolic activity in the heart and brain, where it is involved in mitochondrial oxidation processes (Schurr, 2018; Passarella et al., 2008). ATP production in HepG2 cells greatly depend on OXPHOS (Yang et al., 2020). OXPHOS may be promoted by an increase in the extracellular L-lactate output from neighbouring glycolytic tumour cells (Hong et al., 2019; Schurr & Passarella, 2022) Mitochondrial L-lactate dehydrogenase (mL-LDH) is highly expressed in Hep G2 cell, where pyruvate carrier function is diminished. This suggests that pyruvate metabolism may be not the main driver of OXPHOS activity in Hep G2 cells (Pizzuto et al., 2012). mL-LDH can oxidise L-lactate, leading to low OCR and membrane potential generation. Pyruvate is also formed in this process and continues to participate in the TCA cycle (deBari & Atlante, 2018; Passarella et al., 2008; Pizzuto et al., 2012; Passarella & Schurr, 2018). This L-lactate mechanism can restore the L-lactate/pyruvate shuttle and the key substances, oxaloacetate (OAA), malate (MAL), and citrate (CIT) in Hep G2 cells, also appear outside the mitochondria because of L-lactate uptake and metabolism (Pizzuto et al., 2012; Passarella & Schurr, 2018). The L-lactate/pyruvate shuttle can compensate for the limited ability of cancer cells to recycle NADH to NAD+ (deBari & Atlante, 2018).

The above shows that L-lactate participates in the transport of essential molecules from the cytosol to the mitochondria and promotes the OXPHOS process in Hep G2 cells. However, Lee et al. (2017) noted that defects in OXPHOS may be triggered by an increased extracellular L-lactate release from the surrounding glycolytic tumour cells. Taken together, the results of these studies are contradictory. Future studies should explore the mechanism of L-lactate/pyruvate transport into mitochondria as a target to disrupt the relationship between glycolysis and OXPHOS and suppress tumour growth.

The available information seems to indicate that key steps in both glycolysis and OXPHOS pathways must be simultaneously inhibited to effectively kill HCC cells through energy deprivation. However, a logical explanation for this thermodynamic resistance is lacking. The significance and consequences of fluctuations in lactate levels in HCC are poorly documented, requiring further research on the intricacies of HCC metabolism.

Links between mitochondrial metabolic reprogramming and viable biological processes in hepatocellular carcinoma

Roles of OXPHOS in HCC treatment

The regulation of OXPHOS through targeted drug treatments has had limited success in HCC. Recovery of OXPHOS can increase the sensitivity of HCC cells to drugs (Plecita-Hlavata, Jezek & Jezek, 2015; Schmidt et al., 2021; Skolik et al., 2021). Sorafenib is the first-line targeted therapy for advanced HCC (Qin et al., 2021; Cheng et al., 2020). The glycolytic activity in HCC cells is strengthened in the presence of sorafenib and promotes sorafenib resistance (Zhang et al., 2017). Spitz et al. (2009) demonstrated that acetylsalicylic acid can promote anti-tumour effects by inhibiting glycolysis and elevating OXPHOS levels. Further studies showed that acetylsalicylic acid reduced glycolysis rates by PFKFB3 in sorafenib-resistant HCC cells, corresponding to a change from glycolysis to OXPHOS. The combined use of sorafenib and acetylsalicylic acid can reduce the proliferation and increase the apoptosis of HCC cells (Li et al., 2017b). Similarly, simvastatin can inhibit PKM2-mediated glycolysis, re-sensitizing HCC cells to sorafenib (Feng et al., 2020a). Therefore, combinatorial therapies with sorafenib are promising and require further investigation.

In addition, OXPHOS also has an impact on the immunotherapy of HCC. Inhibition of OXPHOS in activated T cells inhibits their proliferation and upregulate the genes associated with T cell depletion (Vardhana et al., 2020). Up-regulation of OXPHOS promotes metabolic reprogramming, which restores T cell function and enhances the efficacy of cancer immunotherapy (Guo et al., 2021). Human leukocyte antigen class-I (HLA-I) promotes the immune response by presenting antigenic peptides to cytotoxic T cells (Garcia-Lora, Algarra & Garrido, 2003). HK2 can lead to the downregulation of HLA-I to achieve immune suppression in HCC (Liu et al., 2021b).

Contrary to the above-mentioned view, OXPHOS inhibitors, which lower the OCR and inhibit tumour growth, could be effective agents in anticancer therapies. Aglycemia increases fermentative glycolytic substrate-level phosphorylation following glucose refeeding as well as the responsiveness of both fermentation and OXPHOS to meet the energy demand in HepG2 cells. Hence, glycaemic OXPHOS HepG2 cells exhibit increased drug resistivity (Plecita-Hlavata, Jezek & Jezek, 2015). In addition, HepG2 cells increase their resistance to chloroethylnitrosourea (CENU) by boosting OXPHOS and ATP levels (Loiseau et al., 2009). This seems to contradict the idea that recovery of OXPHOS can increase the sensitivity of HCC cells to drugs. 3-Bromopyruvic acid (3-BrPA)—a potential anti-tumour agent—not only inhibits glycolysis in HepG2 cells but also inhibits the metabolism of glucose and glutamine in the TCA cycle (Li et al., 2018b). Its dual effects provide great value for future exploration of OXPHOS targeted therapy for HCC.

These findings highlight potential strategies to promote the sensitivity of HCC to drugs. We conclude that OXPHOS plays an important role in the sensitivity and resistance to HCC treatment. However, further in vivo studies are required to determine their efficacy, pharmacokinetics, dosing methods, and side effects of such strategies (Ashton et al., 2018).

Conclusions

Cancer was previously believed to be a proliferative disease. However, growing evidence suggests that it is a metabolic disease (Ashton et al., 2018; Roth et al., 2020; Lyssiotis & Kimmelman, 2017; Martinez-Reyes & Chandel, 2021). Dysregulation of cellular metabolism is as widespread in cancer cells as other hallmarks of cancer function (Roth et al., 2020). Metabolic reprogramming is a crucial microenvironmental adaptation in HCC cells (Shi et al., 2009). The occurrence of HCC is highly dependent on elevated levels of aerobic glycolysis. However, HCC, especially the lower differential and metastasis potential cells, are more dependent on OXPHOS. Elevated aerobic glycolysis and decreased OXPHOS promote the development of HCC, and elevated OXPHOS levels have a similar effect. The dynamic regulation mechanism of mitochondria is very complex. There are different regulatory mechanisms under different environmental pressures. Glycolysis and OXPHOS are complex metabolic processes that affect the development, growth, metastasis, drug resistance of liver cancer through a variety of signalling pathways (Table 1). Understanding the molecular mechanisms of mitochondrial activity/silencing in HCC cells and the limitations of this phenomenon is critical for HCC treatment and drug efficacy. Therefore, we still need to explore how OXPHOS and glycolysis can coordinate the metabolic adaptations of HCC cells in the future. We hope to enhance the therapeutic efficacy of HCC by strategically combining multiple drugs that target different metabolic pathways.