Tumor tissue is characterized by a microenvironment of hypoxia and low nutrients. Tumor cells undergo metabolic reprogramming to adapt to severe living conditions. “Metabolic reprogramming” has been recognized as one of the 10 markers of cancer [1]. In the last century, Otto Warburg, a German biologist, first described that compared with nonproliferating normal cells, tumor cells tend to choose to rapidly produce ATP by enhancing the conversion of glucose to pyruvate even when oxygen is sufficient [2]. Although the Warburg effect has been widely accepted as a common feature of metabolic reprogramming, increasing evidence has shown that tumor metabolic reprogramming is reflected not only in aerobic glycolysis, in which fatty acid metabolic reprogramming targeting tumor cells has gradually become the focus of tumor research. However, clinical research on tumor metabolism, especially fatty acid metabolism, failed to keep pace with the progress of basic research [3].

According to the length of the carbon chain, fatty acids (FAs) were divided into short-chain, medium-chain, and long-chain fatty acids. FAs play an important role in all stages of tumors [4, 5]. Generally, rapidly proliferating cells need a large number of fatty acids to promote membrane synthesis and form phospholipids to support replication. Simultaneously, fatty acids act as substrates for mitochondrial ATP synthesis [6], to regulate post-translational lipid modification and the function of signaling proteins [7, 8]. In conclusion, fatty acids show an important effect in tumors, and interfering with fatty acid metabolism may become a potential strategy for tumor treatment. Targeting fatty acid metabolism reprogramming of tumor cells has gradually become the focus of research [9]. Most of the current studies have focused on the therapeutic targets of de novo fatty acid synthesis and on FAs uptake to limit its use as a source of energy and cell membrane phospholipids [10, 11]. However, some studies have found that fatty acid $\beta$-oxidation (FAO) is the ultimate fate in the FAs energy generation cycle [12, 13]. It is worth noting that fatty acids are an important energy source of tumors in nutritional deficiency and even for some types of malignant tumors. ATP, NADPH, and intermediates produced by $\beta$-oxidation are vital for cell survival and the maintenance of a malignant phenotype. Therefore, clarifying the key role of FAO reprogramming in the tumor process may help to find potential targets for tumor therapy, which is of great research significance and clinical value. Herein, we summarize the important role of key enzymes and important metabolites involved in the process of FAO in promoting tumor progression and further explore the value and potential of targeted FAO in tumor therapy.

Although mitochondrial FAO is the main source of bioenergy, it is not generally considered a part of the cancer metabolic blueprint. However, in the past few years, people’s views on the relationship between FAO and tumors have changed. The proliferation, survival, stemness, drug resistance, and metastasis of cancer cells depend on FAO. FAO is also reprogrammed in cancer-related immune cells and other stromal cells, which may contribute to the immunosuppressive tumor microenvironment. FAO is the transformation of long-chain fatty acids into acetyl-CoA through the process of fatty acid activation, fatty acid transport and fatty acid oxidation under the action of a series of enzymes. The whole process produces a large number of reducing agents and ATP, which is more efficient than the tricarboxylic acid cycle. Recently, studies have provided vital evidence that cancer has a “Lipolytic phenotype” [14]. Similar to glycolysis or fatty acid synthesis in tumors, FAO shows abnormalities in a variety of tumors [15, 16, 17].

Some studies demonstrated that the expression of $\beta$-oxidation-related proteins in the mitochondria of liver cancer cells is higher than that of normal stem cells, and the oxidation rate of free fatty acids and other NAD-linked substrates by the mitochondria of liver cancer cells is faster, up to 6.6 times [18]. The expression of FAO-related enzymes (CPT1 and CD36) was significantly upregulated in the EMT model of breast cancer [19]. Lee CK et al. [20] proposed that FAO is an emerging factor for tumor lymph node metastasis. In lymph nodes, a lipid-rich microenvironment, tumor cells may preferentially use FAs as an energy source to enhance their metastatic potential [20, 21]. The above studies suggested that the survival, proliferation, stemness, drug resistance, and metastasis of tumor cells depend on FAO [22, 23, 24, 25]. Moreover, in addition to affecting tumor cells, FAO also plays a decisive role in the differentiation and function of stromal cells and immune cells. For example, the abnormal activity of FAO promoted the M2 differentiation of macrophages [26, 27], and FAO contributed to the proliferation of Treg cells in vitro [28, 29].

In addition, FAO produces a large amount of ATP, which provides the possibility for rapid tumor proliferation, invasion, and metastasis. Simultaneously, as one of the main sources of NADPH, the role of balancing ROS, maintaining redox balance, and promoting tumor cell survival cannot be ignored [30]. Moreover, a variety of intermediates are produced in the process of FAO as signal molecules or raw materials for the synthesis of other important substances, providing potential conditions for tumor progression [31]. FAO affects various aspects of cancer, including proliferation, metastasis, stemness, and the immune microenvironment (Fig. 1). The exploration of targeted FAO in tumor treatment is full of great potential and possibility.

Fig. 1.

Fatty acid $\beta$-oxidation plays a vital role in tumors. FAO affects all aspects of tumors, including not only the proliferation, metastasis, stemness, and drug resistance of tumors but also provides beneficial conditions for tumor progression by interfering with the function of immune cells and other stromal cells.

In the presence of ATP, CoA-SH and Mg${}^{2+}$, FAs are catalyzed by fatty acyl-CoA synthase to produce fatty acyl-CoA, which is then transported to mitochondria with the assistance of carnitine lipoacyltransferase. Based on the catalysis of a series of $\beta$-oxidases, long-chain fatty acids are gradually oxidized into multiple acetyl-CoAs, and different enzymes play a role in different links in the whole process [32]. Some studies have shown that abnormal expression of genes involved in FAO is associated with malignant phenotypes, including therapeutic resistance, metastatic potential, and recurrence [33, 34, 35]. Therefore, several key enzymes in the FAO process may be used as therapeutic targets with tumor therapeutic potential (Fig. 2).

Fig. 2.

Fatty acid $\beta$-oxidation process and various related key enzymes. ACSLs, acyl-CoA synthetases; CPT1, carnitine acyltransferases 1; CPT2, carnitine lipoxytransferases 2; ACDLs, lipoacyl-CoA dehydrogenases; FADH2, flavin adenine dinucleotide transmitter; ECHs, enoyl-CoA hydratases; NADH, nicotinamide adenine dinucleotide; HACDs, hydroxyacyl-CoA dehydrogenases; KACTs, ketoalkyl-CoA thiolases; FASNs, fatty acid synthetases; CD36, CD36 molecule.

Changes in tumor cell metabolic patterns inevitably affect cell redox homeostasis [64]. In most cases, the growth and survival potential of tumor cells is limited by the level of NADPH in cells. On the one hand, it provides redox ability to counteract oxidative stress; on the other hand, it is a coenzyme of anabolic enzymes to maintain cell growth and proliferation. During the occurrence and development of tumors, the level of intracellular ROS increases significantly [65]. Reduced glutathione is an important antioxidant in cells that counteracts the oxidative pressure brought by ROS. In tumor cells with elevated ROS levels, reduced glutathione could be oxidized to oxidized glutathione, followed by glutathione reductase and reduced NADPH. It is reduced again under the action of to maintain the redox balance in tumor cells [66, 67].

In addition to providing energy, FAO is also an important source of NADPH [68]. Numerous studies have shown that NADPH derived from FAO in tumor cells is a key factor in counteracting oxidative stress [14, 69, 70]. Acetyl-CoA produced by FAO entered tricarboxylic acid (TCA) cycle and generated citric acids with oxaloacetic acid. Citric acids were shuttled to the cytoplasm to generate NADPH [69] (Fig. 3). Previous studies have pointed out that the main purpose of FAO in rapidly proliferating endothelial cells is to carry out de novo dNTP synthesis. Compared with resting endothelial cells, the upregulation of FAO was three times or more higher than that of proliferating endothelial cells. Its main purpose is to maintain the tricarboxylic acid cycle through NADPH regeneration to maintain redox homeostasis [71]. Considering the adverse effects of a large number of ROS, cancer stem cell-like cells maintain ROS levels by coupling FOXM1-dependent PRX3 expression and fatty acid oxidation [30]. FAO was inhibited in glioma cells and showed a significant decrease in NADPH levels, resulting in an increase in ROS levels and cell death [70]. Nissm Hay et al. [72] also demonstrated the correlation between FAO and NADPH homeostasis. Nrf2, a transcription factor that regulates cellular redox status, has been shown to promote FAO and increase NADPH regeneration, thereby guiding metabolic reprogramming during stress [73]. FAO is an important component of metabolic reprogramming by providing ATP and maintaining redox balance to promote tumor progression.

Fig. 3.

FAO provides ATP and NADPH to promote tumor progression. On the one hand, NADH and FADH2 produced in the process of FAO could generate ATP through the electron transport chain (ETC). On the other hand, its metabolite acetyl-CoA enters the tricarboxylic acid (TCA) cycle to synthesize citric acid with oxaloacetic acid. Citric acid enters the cytoplasm through the citric acid shuttle to generate isocitrate. Under the action of isocitrate dehydrogenase (IDH), cytoplasmic NADPH is generated.

A variety of cells constitute a complex tumor microenvironment, including immune cells and stromal cells. Therefore, we should take the tumor as a whole into consideration. In addition to tumor cells, the existence, phenotype and function of other cells affect the progression of tumors, and the functional phenotype of these cells is closely related to their metabolic mode [74, 75]. Studies have shown that effector CD4+ T cells rely on glycolysis to provide energy and substances for biosynthesis. However, immunosuppressive T cells (Tregs) suggest a higher level of FAO [76]. Tregs combine glycolysis, fatty acid synthesis, oxidation, and other metabolic modes to defeat T cells that mainly rely on glycolysis to meet energy and material needs [77]. Several research groups have reported that M2 macrophages use FAO to promote mitochondrial oxidative phosphorylation, providing a survival advantage over M0 and M1 macrophages [78, 79]. Inhibition of FAO could prevent macrophages from polarizing to the M2 type [80]. Early studies have shown abnormal lipid accumulation in tumor-associated dendritic cells with a tolerance phenotype [81]. We investigated and summarized the role of FAO in different immune cells in the early stage and found that active FAO can cause a variety of immune cells, such as macrophages, dendritic cells, and NK cells, to change into an immune tolerance phenotype and contribute to the immunosuppressive microenvironment [82].

The dynamic crosstalk between stromal cells and tumor cells is also one of the potential mechanisms of malignant tumor progression. Adipocytes are an important component of the tumor microenvironment. Adipocytes in the tumor microenvironment secrete a large number of exosomes. These exosomes are absorbed by tumor cells, which lead to increased migration and invasion. Interestingly, it was found that the vesicles secreted by these adipocytes were rich in FAO-related proteins, which was one of their highly specific characteristics [83]. Further studies showed that in the presence of these exosomes, FAO in melanoma cells was also mobilized and became more active [83]. In addition to the abnormal FAO of adipocytes in the tumor microenvironment, studies have found that cancer related-fibroblasts actively oxidized FA and conducted minimal glycolysis by upregulating CPT1A to promote the proliferation, migration, and invasion of colon cancer cells [84]. Etomoxir directly blocks CPT1A-mediated FAO in fibroblasts, which could inhibit migration and invasion in vitro and reduce tumor growth and peritoneal metastasis in vivo [84].

Abnormally active FAO is one of the characteristics of carcinoma, which is a prerequisite for some tumors. Some studies have shown that FAO is the driving force of $\beta$-catenin induced hepatocellular carcinoma (HCC), and inhibiting FAO would prevent the progress of HCC [85]. Other studies have also pointed out that mutant KRAS promotes fatty acid uptake, accumulation and $\beta$-oxidation in lung cancer with an ACSL3-dependent manner. Therefore, ACSL3-mediated FAO is necessary for the occurrence of KRAS mutant lung cancer [86].

FAO was also regulated by multiple oncogenes or tumor suppressor genes (Table 1) [43, 61, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98]. c-Myc upregulated the main FAs production regulator sterol regulatory element binding protein 1 (SREBP1) in tumor cells and promoted the production of fatty acids and the process of FAO [43]. Meanwhile, c-Myc has also been shown to regulate FAO by inhibiting the expression of ACC2, and ACC2 suppressed the effect of CPT1A through targeting malonyl-CoA [87]. Significantly, Cyclin D1 is a cyclin, which is abnormally expressed in tumors as an oncogene. Studiese have evaluated that it not only play a key role in the process of cell cycle, but also inhibit the activity of PPAR$\alpha$ and block CPT1 expression to regulate FAO [88, 91]. CD147 is a key regulator of fatty acid metabolism and is overexpressed in a variety of cancers. At present, a drug named Licartin developed with CD147 antibody labeled with I${}^{131}$ has been approved by National Medical Products Administration (NMPA) for the treatment of HCC [89]. Previous studies have indicated that CD147 could not only upregulate SREBP1 by activating Akt/mTOR signaling pathway, and then activate FASN and ACC1 to promote fatty acids accumulation, but also inhibit PPAR$\alpha$ and CPT1A with activating p38/MAPK to disturb FAO [89]. FAO has showed protective factor in a variety of tumors [99, 100], so that cancer cells survived with facing severe challenges. Partial oncogenes could controll the fate of tumor cells by regulating the activity of FAO. And the deletion of some tumor suppressor genes caused fatty acid reprogramming to maintain the malignant phenotype of tumor cells.

At present, FAO and related regulatory genes as targets have been gradually recognized and tried as potential candidates for cancer therapy. The promising drugs (and related targets) targeting FAO in cancer mainly focused on CPT [101], we organized the main targeted drugs (Table 2 and Fig. 4). Among them, etomoxir plays an important role in the treatment of various tumors by targeting FAO. Etomoxir irreversibly inhibited CPT1A and CPT1B [102]. Etomoxir significantly reduced liver and lung metastatic nodules of colorectal cancer cells by promoting anoikis [56]. However, etomoxir has serious side effects. Long-term use of etomoxir could lead to cardiac hypertrophy by promoting oxidative stress and NF-$\kappa$B pathway [103]. The selective CPT1A inhibitor ST1326 (Teglicar) is safer than etomoxir, it does not cause cardiac hypertrophy, but still has some hepatotoxicity [104, 105]. This novel CPT1A inhibitor has antitumor activity in hematological malignancies such as acute myeloid leukemia and Burkitt lymphoma [53]. ST1326 combined with Bcl2 inhibitor ABT199 showed strong synergistic inhibitory effects on acute myeloid leukemia (AML) [106]. At present, ST1326 is still in the preclinical experimental research stage. In addition to inhibiting CPT1A, perhexiline can also inhibit CPT2, showing a similar but stronger anti-tumor effect than etomoxir [107]. As a partial $\beta$-oxidation inhibitor, ranolazine has showed anticancer effects in leukemia and breast cancer [108, 109]. Ranolazine increased the antitumor effect of prostate cancer cells by changing the activation status of the neighboring T-cells. 6-gingerol is known to have a potential anticancer agent by inducing apoptosis in cancer cells. Its apoptotic effect is to inhibit CPT1 by accumulating pathologically high concentrations of malonyl-CoA.

Fig. 4.

The mechanism of drugs targeting FAO in cancers. CPT in FAO as an important target in tumor therapy. There are also potential targets for $\beta$-oxidation and ACSLs.

In addition to above drugs, several potential FAO-related targeted drugs are in the experimental stage. Chemical inhibition of ACSLs activity by triacsin C (inhibits ACSL1, ACSL3, and ACSL4, but not ACSL5 or ACSL6) induces apoptosis in lung, colon, and brain cancer cells [110]. 2-bromopalmitate (2-BP) is an irreversible inhibitor of many membrane-associated enzymes. Reported as an inhibitor of $\beta$-oxidation, 2-BP can also alter lipid metabolism through CPT1, thereby affecting tumor growth [28, 29]. High-dose dexamethasone can also delay tumor growth and promote apoptosis by inhibiting CPT1A [111].

Notably, polyunsaturated fatty acids (PUFAs) have been shown to play a vital role in tumor treatment. According to the position and function of double bonds, PUFAs are divided into $\omega$-6 PUFAs, including linoleic acid and arachidonic acid and $\omega$-3 PUFAs, containing linolenic acid, docose hexaenoie acid (DHA), eicosa pentaenoic acid (EPA). Studies have showed that $\omega$-3 PUFAs are considered to be an immune nutrient to use in the nutritional treatment of cancer patients [112]. On the one hand, $\omega$-3 PUFAs showed anti-inflammatory and analgesic effects by inhibiting the expression of NF-$\kappa$B [113]. On the other hand, it can be used as an agonist of G protein-coupled receptors (free fatty acid receptor 1, FFA1 and free fatty acid receptor 4, FFA4) [114, 115]. Therefore, tumor patients accompanied by cachexia, pain and other complications could benefite from $\omega$-3 PUFAs administration.

In addition to the effect for tumor complications, recent researches have pointed that PUFAs also paly a vital role in suppression of tumor, including breast cancer, colorectal cancer, liver cancer. The influence of PUFAs on tumor could be summarized as follows: (1) Inhibition of tumor cell cycle. DHA disturbed cell cycle by suppressing DNA synthesis in liver cancer cells and melanoma cells [116]. Furhter, Istfan NW demonstrated the duration of S phase of tumor cells increased significantly when $\omega$-3 PUFAs were consumed [117]. (2) Induction of tumor cell apoptosis. Studies have showed PUFAs cause apoptosis in many ways, especially PUFAs inhibit the expression of Bcl2 and promote Bax expression [118, 119]. (3) Supression of tumor related angiogenesis. $\omega$-6 PUFAs reduced the content of leptin in adipocytes, resulting in the decrease of VEGF expression to prevent angiogenesis [120]. (4) The influence of immune system. PUFAs regulated the differentiation of directional thymocyte lineage and mediated inflammatory response by increasing the production of CD8+ T cells [121]. Furthermore, PUFAs in peripheral lymph regulated the specific proliferation of T cells and the secretion of cytokines, causing the inhibition of Th2 response [122]. (5) Lipid peroxidation mediated cell death. Lipid peroxidation is achieved by two main pathways, by enzymatic or by non-enzymatic oxidation, respectively [123]. Lipid peroxidation was accumulated in cancer cells inducing iron-dependent cell death, which was called ferroptosis [124]. In addition, the non-enzymatic peroxidation of PUFAs and of their 12/15-lipoxygenase-derived hydroperoxy metabolites caused the generation of the reactive aldehyde species 4-hydroxyalkenals, in particular 4-hydroxynonenal [125]. Those metabolites showed cytotoxic effect to kill tumor cells [126]. (6) Other machanisms. Some studies have found that PUFAs inhibit the expression of MMPs in gastric cancer and interferes with tumor cell invasion and metastasis [127]. Simultaneously, $\omega$-6 PUFAs was pointed out through PPAR$\gamma$ regulated downstream target genes to reduce synthesis of inflammatory mediators [128, 129]. In general, PUFAs showed an inhibitory effect on tumor cells, although this effect was thought to be related to the ratio of $\omega$-3 and $\omega$-6 [130, 131], however, this is still a potential strategy in tumor treatment, which is worthy of further discussion.

FAO plays an extremely important role in supporting tumor progression. Research on FAO has great potential in the diagnosis and treatment of tumors. This review focused on the key enzymes that may be used as tumor treatment targets in the process of fatty acid $\beta$-oxidation and described that the abnormal expression of enzymes in the process of fatty acid activation, transport, and $\beta$-oxidation mediates the progression of a variety of tumors. Inhibiting the expression of key enzyme-related genes or enzyme activity could hinder tumor cell proliferation, migration, and other malignant functions. Studies have shown that knocking down FAO-related genes induces tumor cell apoptosis; however, the specific mechanism of action is not particularly clear, and further research is needed. In addition to key enzymes, this article also summarized that FAO provided a large amount of ATP for tumors to support uncontrolled proliferation and maintained cell redox homeostasis by producing NAPDH. Abnormal fatty acid metabolism has attracted the attention and discussion of an increasing number of researchers. Compared with extensive research on de novo fatty acid synthesis and fatty acid absorption, FAO is equally important to tumor survival and progression, and it plays a role that cannot be ignored. Therefore, targeting abnormal FAO in tumors may be a new strategy and idea for tumor treatment.

At present, FAO inhibitors have been used in the clinic, but they are mainly used to treat heart disease [132, 133, 134]. Their application in tumors is still in preclinical research or encounters bottlenecks, such as drug toxicity, and the in vivo effect is not as good as the in vitro effect. On the other hand, FAO has been proven to be indirectly activated by PPAR activators [133], AMPK activators [72], or ACC inhibitors [135], so there are indirect strategies for these targets, but the various pathways activated by these methods complicate the interpretation of the results. Therefore, further. direct targeting of FAO has great significance and potential.

FA, fatty acids; FAO, fatty acid $\beta$-oxidation; ACSLs, long-chain fatty acyl-CoA synthetases; AR, androgen receptor; CPT1, carnitine palmitoyl transferase 1; ROS, reactive oxygen species; ACAD, acyl-CoA dehydrogenase; ECHS1, enoyl-CoA hydratase short-chain 1; PRDX3, peroxidase 3; DECR1, 2,4-dienoyl CoA reductase 1; PUFAs, polyunsaturated fatty acids; TCA cycle, tricarboxylic acid cycle; Treg, regulatory T cells; 3-KAT, 3-ketoacyl CoA thiolase; FASN, fatty acid synthetase; PGC1$\alpha$, PPAR coactivator-1$\alpha$; CEBPB, CCAAT/enhancer binding protein $\beta$; JAK/STAT3, janus kinase (JAK)/Signal transducer and activator of transcription 3 (STAT3); OXPHOS, oxidative phosphorylation; PPAR$\alpha$, peroxisome proliferator-activated receptor alpha; CCAT1, colon cancer associated transcript 1; SIK, salt-inducible kinase; PLA2, phospholipase A2; MCAD, medium-chain acyl-CoA dehydrogenase; LCAD, long-chain acyl-CoA dehydrogenase; HIF-1$\alpha$, hypoxia inducible factor 1 subunit alpha; AMPK, protein kinase AMP-activated catalytic subunit alpha 1; NDRG2, NDRG family member 2; RARRES1, retinoic acid receptor responder 1; REDD1, DNA-damage-inducible transcript 4.

HC and FZ designed the research study. HC performed the research. MT and YS provided help and advice on. ZY and SY analyzed the data. HC wrote the manuscript. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript.

Not applicable.

We apologize to colleagues whose work is not cited due to space constraints. And we also thank anonymous reviewers for excellent criticism of the article and all the authors in the reference list.

This research was funded by Natural Science Foundation of Chongqing, China, grant number cstc2020jcyj-msxmX0485 and Medical Science and Technology Innovation Fund of Chongqing General Hospital, grant number 2019ZDXM01 and Y2020ZDXM08.

The authors declare no conflict of interest.