Introduction

Our bodies utilize lipids for various fundamental processes, ranging from the major energy source to the synthesis of vital macromolecules such as cholesterol, membrane phospholipids, and hormones. Like most cells, T cells are heavily dependent on lipid consumption for energy needs, but their naiveté, activation, and effector functions are influenced by the subtype and quantity of lipid intake [1]. A number of surface and intracellular proteins, such as differentiation 36 (CD36), fatty acid-binding protein (FABP), fatty acid transporter protein (FATP), and sterol regulatory-element binding proteins (SREBPs), are responsible for processing lipids in T cells. Under normal circumstances, quiescent T cells process lipids into more energy-efficient oxidative phosphorylation (OXPHOS), which is replaced by aerobic glycolysis once T cells are activated. A complex cascade of co-stimulatory triggers channel T cells into T-regulatory cells (Tregs) and helper T cells (Th cells), each with differential metabolic shifts. As the immune response approaches an end, activated T cells undergo apoptosis or are converted to nondividing T memory cells that revert to OXPHOS [2, 3].

The tumor microenvironment (TME) is a complex space characterized by multiple cell types and their interwoven interactions, predominantly favoring cancerous growth. The interdependence of nutrients, vasculature, and metabolic demands actively shapes the cellular fate in TME. The specialized metabolic switches, nutrient preferences, cellular growth demands, and secretion of various intrinsic and extrinsic factors prime the particular kind of cell in TME to behave in a specific manner [4, 5]. For example, growth signals and upregulation of metabolic features in tumor cells make them more suitable for proliferation, while immune cells undergo tumor evasion after metabolic switch at the same time [6, 18,19,20], while naïve T cells and memory T cells (Tmem) maintain fatty acid oxidation (FAO) as the default metabolic program. FAS has to go through various steps mediated by key enzymes before lipid synthesis can occur. Some of these enzymes affect the rewiring of T cells. For instance, it has been found that CD8+ T cells cannot expand without SREBP signaling during viral infection; however, it was expendable for homeostatic growth. SREBPs in Teffs induce the expression of enzymes fatty acid synthetase (FASN), acetyl-CoA carboxylase (ACC) and hydroxy-methyl-glutaryl-CoA reductase (HMGCR) [17]. Seon Ah Lim et al. have reinforced these observations in a mouse model of TME, showing that inhibition of SREBP-dependent lipid synthesis and metabolic reprogramming in Tregs initiates a robust antitumor response without causing autoimmune disorders. It has been further proven that depletion of an obligatory SREBP factor SCAP curtails tumor growth and upgrades anti-PD-1 immunotherapy [21]. Similarly, Yang-An Wen et al. confirmed the tumor growth suppression after the knockdown of either SREBP1 or SREBP2 target genes required for lipid biosynthesis [22]. At least theoretically, depletion of SREBPs in cancer cells and Tregs yields anticancer benefits; however, deletion of SREBPs in Teffs can affect their blast, which may lead to immune evasion. These SREBPs might be context-dependent drug targets. Further research is required to ascertain them in all subsets of T cells and cancer cells in complex TME.

In addition to SREBPs, ACC is another widely studied lipogenic gene in this direction, which has led to similar results. For instance, Luciana Berod and colleagues found that TH17 cells, but not Tregs, rely on the de novo lipid synthesis mediated by ACC1. Tregs do not follow this scheme as they tend to utilize exogenous FFAs. T cell–specific deletion of ACC1 in mice is able to ameliorate autoimmune disease through anti-inflammatory actions [23]. Similarly, CD8 + T-cell expansion and proliferation are severely curtailed in the absence of ACC1 and subsequently restored by exogenous fatty acid supply [24]. Inhibition and activation of ACC1 favor peripheral Tregs and Th17 cell differentiation, respectively, under TME [25].

FAS required for proliferation is heavily based on acetyl-CoA produced from glycolysis. Inhibition of FAS-related enzymes such as SREBPs, ACC1, or downstream targets leads to defective effector T-cell responses [26]. Although modulation of SREBPs or downstream pathways is rarely attempted in cancer, they may be good therapeutic targets for the future.

FAO, a key bioenergetic pathway [27], displays a critical role in the development of tolerogenic dendritic cells (DCs) [28]. Interestingly, DCs of cancerous origin tend to accumulate oxidized lipids, thereby suppressing T-cell effector functions, which indirectly favors tumor progression [29]. It is speculated that accumulated fatty acids support FAO and therefore promote tolerogenicity in the cancer setting [30]. Furthermore, a carnitine palmitoyltransferase-1 A (CPT1A) inhibitor, etomoxir, has immunomodulatory actions on CD8+ Tmem cell differentiation [31]. However, Brenda Raud et al. have pointed out the flexible metabolic fuel choices of Tmem and found that CPT1A-mediated long chain fatty acid oxidation (LC-FAO) is expendable for the expansion of CD8+ T cell memory [32].

PD-L1/2 in T-cell metabolism in TME

Programmed death ligand-1/2 (PD-L1/2) belongs to CD28 family, and is predominantly expressed on tumors and tumor-infiltrating myeloid cells [33, 34]. PD-L1/2 has been suspected to play a negative role in TME by suppressing antitumor immunity by regulating inhibitory cascade on effector T cells [35, 36]. For example, a recent report has suggested that PD-1 participates in the metabolic reprogramming of activated T cells [37] by reducing Akt (protein kinase B) activation and subsequently inhibiting mammalian target rapamycin (mTOR) activity [37,38,39]. Indeed, reduced activation of mTOR in PD-1+ CD8 + T cells activates transcription factor forkhead box O1 (FoxO1), allowing the survival of exhausted CD8 T cells [39]. As glycolysis is necessary for T cells to obtain the required amount of energy for proliferation, the diversion from glycolysis to FAO likely sabotages antitumor immunity in TME. The preferential diversion from glycolysis to FAO leads to the longevity of Tmem cells, which possess substantial mitochondrial spare respiratory capacity (SRC) [40]. Consistently, inhibition of the PD-1 pathway in the early phases of a viral infection leads to raised mTOR signaling in virus-specific CD8 + T cells, resulting in quicker infection clearance [41]. Certainly, mTOR and Akt are key lipid regulators within the cell [42]. Similarly, inhibition of PD-1/PD-L1 leads to extensive cytotoxic T-cell infiltration into TME as reported in ex vivo, in vitro, and in vivo experimental models [43]. Blocking PD-1 in cancer patients leads to decreased tumor progression and improved survival [44, 45].

PD-1 has also been reported in T-cell exhaustion, which is counterproductive to immunity [46]. Exhaustion of PD-1–associated CD8 + T-cell existed in chronic viral infections in mice [47] and in clinical studies [48,49,50]. It is not surprising that, in 2014, FDA approved the first blocking antibody targeting PD-1 to treat metastatic melanoma, and up to August 2017, many drugs against the PD-1 pathway had been applied in various cancers [46]. However, a recent study has contradicted the holistic benefits of PD-1 blockade in TME as distinct responsiveness of T-cell subpopulations to PD-1 blockade was observed. The study observed that effector and central memory phenotypes were among the most affected T-cell subpopulations after PD-1 blockade but had different gene expression profiles with PD-L1 in comparison to PD-L2 [51].

T cells in TME of obese states

Although it is well established that obesity and lipid overload states are obvious causes of cancer and its progression [52,53,54,55], it has rarely been investigated how these conditions rewire the T-cell differentiation and metabolic switch in TME. Alison E Ringel et al. have recently addressed this important question and demonstrated how obesity shifts the metabolic program of TME to inhibit T-cell function and promote tumor growth. Researchers have systematically shown in a murine model that a high-fat diet (HFD) is differentially taken up by tumor cells in TME as compared to CD8 + T cells, which leads to modified fatty acid partitioning, diminished CD8+ T-cell infiltration, and promotion of tumorigenesis [56]. Of course, preferential fat consumption by tumor cells makes localized T cells less efficient as fat is required to raise the number and plasticity of T cells in TME.

There are several other genes and enzymes with limited evidence but with potential to be good drug targets. The consumption of fat is not only restricted to activate the T cells, but it is also evident that under normal circumstances de novo cardiolipin synthesis keeps the function of CD8 + T cells intact. Mauro Corrado and colleagues demonstrated poor T-cell antigenic responses in T cells deficient in the cardiolipin-synthesizing enzyme—protein tyrosine phosphatase, mitochondrial-1 (PTPMT1). PTPMT1-dependent cardiolipin synthesis is also important for mitochondrial fitness, especially during Tmem cell differentiation or nutrition scarcity [57]. Monoacylglycerol lipase (MGL) hydrolyzes monoglycerides into glycerol and fatty acids. It is abundantly present in tumor cells, and MGL knockout (KO) mice exhibit a reduced tumor size compared with control mice. Interestingly, the reduction in tumor progression is associated with a parallel upregulation in the number of CD8+ T cells. Furthermore, naïve CD8+ T cells exhibit enhanced tumoricidal activity in MGL KO mice [58]. P4HA2 is a metabolism-related gene that is upregulated in cervical cancer tissues, and negatively correlates with CD8 + T cells. Knockdown of P4HA2 suppresses lipid droplet storage in cancer cells [59]. Teresa Manzo and colleagues demonstrated a progressive accumulation of long-chain fatty acids (LCFAs), which, instead of providing an energy source, hamper the mitochondrial function and rewire the lipid metabolism pathways. In addition, intrapancreatic CD8 + T cells inhibit the very long-chain acyl-CoA dehydrogenase (VLCAD) enzyme, which worsens the accumulation of LCFAs and very-long-chain fatty acids (VLCFA), subsequently inducing lipotoxicity. In fact, recently obesity has also been described as a booster of antitumor pharmacotherapy in some cancers [60], but the mechanism remains unknown.

MDSCs in T-cell lipid (de)regulation

Myeloid-derived suppressive cells (MDSCs) manifest negative regulatory activity by promoting immunosuppression in immune-related diseases [61, 62]. In tumors, MDSCs accelerate tumor proliferation, tumor expansion, and immune escape, thereby further exacerbating the TME [63, 64]. MDSCs reshape TME by inhibiting T cells and natural killer (NKT) cells while inducing regulatory T cells (Tregs) and regulatory B cells (Bregs) [65, 66]. From recent studies, it is evident that lipid metabolism in tumor-infiltrating MDSCs (T-MDSCs) is rewired for raised fatty acid uptake, FAO upgrade, oxygen consumption rate (OCR), mitochondrial mass, and expression of core FAO enzymes [67]. It is interesting to note that only T-MDSCs, but not splenic MDSCs, raise lipid uptake [68], which implies that only infiltrating MDSCs undergo lipid metabolic reprogramming. This differential scheme of pro-tumor metabolic features sheds light on the complexity of TME.

In mammals, Liver X receptors (LXRs) are involved in lipid homeostasis. Previous studies have revealed that administration of LXR agonists initiates MDSC apoptosis and reduces tumor volume [69, 70]. In addition, lectin-type oxidized LDL receptor 1 (LOX-1) is present in PMN-MDSCs of cancer patients but is absent in healthy individuals [71]. Similarly, Caijun Wu et al. noticed the enhanced immunosuppressive role of monocytic MDSCs after administration of a multidose clinical regimen of gemcitabine (GEM). These authors have implicated that the deregulation of lipid metabolism in residual tumor cells is partially responsible for promoting immunosuppression [72]. It is plausible to conclude that tumor-derived MDSCs are forced to rewire lipid metabolism primarily because of robust lipid storage and related signaling activation.

CD36 and T-cell regulation in TME

CD36 is a scavenger receptor of oxidized lipids, and is expressed in multiple cell types, including T cells [73, 74]. Previous studies have highlighted that tumor-associated immune cells undergo CD36-oriented lipid metabolic reprogramming, which leads to immune evasion and cancer progression [75]. Shihao Xu and coworkers have reported that CD8+ tumor-infiltrating lymphocytes (TILs) are responsive to lipids in the TME, mediated by CD36, which is associated with progressive T-cell dysfunction. It has been explained that T-cell dysfunction occurs in a CD36-dependent manner, which leads to a raise in oxidized low-density lipoproteins (OxLDL) in T cells, promotion of lipid peroxidation downstream, and occurrence of ferroptosis. Interestingly, overexpression of glutathione peroxidase 4 reverses lipid peroxidation to improve the effector capacity of T cells [76]. Similar evidence has been provided by other research groups, associating overexpression of CD36 with shorter survival of melanoma patients with tumor-infiltrating CD8+ T cells, while CD36-depleted CD8+ T cells showed greater antitumor potential and survival compared with wild-type CD8+ T cells [77,78,79,80].

Importantly, there are very few identified metabolic drug targets that work in the same direction in both Tregs and Teffs. As presented above, metabolic drug targets are often context- and T-cell subtype–dependent, in which Tregs and Teffs promote or inhibit tumor, respectively. However, CD36 offers a rare opportunity because its deletion on both Tregs and Teffs results in enhanced antitumor activities [81]. For instance, Wang et al. stated that genetic knockdown of CD36 in Treg cells reduced tumor growth and intratumoral Treg cells, promoting the antitumor function of tumor-infiltrating lymphocytes [82]. Although only few studies have been reported in this direction, CD36 presents a viable common drug target that requires future research.

Effects of lipid metabolism on Tregs

The relative ratio of cytotoxic T cells and Tregs in TME plays a pivotal role in tumor progression and immune evasion [83]. Tregs contribute to immune evasion in TME [84,85,86]. Systematic ablation of Tregs in several cancer types has resulted in tumor suppression and cellular alterations within the TME [87, 88]. It is suspected that cytotoxic Teffs and Tregs follow different activation and proliferation pathways as Tregs are abundant even in the unfavorable metabolic states in TME [89]. Indeed, Weinberg et al. have stressed the necessity of mitochondrial metabolism in Tregs to maintain their immunosuppressive function [90]. A recent study has shown raised production of FFAs by RHOA Y42-mutated gastric cancer, modulated via the PI3K pathway, which favors the accumulation of Tregs in a low-glucose TME. Similarly, the expression levels of FAS, CPT-1, PPARα, and PPARɣ were also higher in gastric cancer with RHOA Y42 mutation [91]. It remains unknown what metabolic switch enables Tregs to expand and proliferate differently from Teffs in the same TME. However, one study confirmed that intratumoral Tregs indirectly promote M2-like TAMs by boosting SREBP1-dependent lipid metabolism and then limiting the CD8+ T-produced interferon-gamma (IFNγ), thereby leading to tumor progression and orchestrating tumor-associated immunosuppression [92]. In addition, inhibition of FABP5 on Tregs causes mitochondrial alterations characterized by impaired lipid metabolism, reduced OXPHOS, and loss of cristae structure. The authors concluded that FABP5 is a gatekeeper of mitochondrial integrity, which is necessary for normal functioning of Tregs [93]. An interesting observation regarding the complicity of Tregs and tumor cells to suppress the T-cell functioning has recently been highlighted. ** T-cell senescence in cancer models in vivo and in vitro [94].

Unconventional T cells in TME

γδ T cells are capable of differentiating into various subtypes of immune cells depending on the TME conditions [95]. Although the scientific knowledge on γδ T cells is underdeveloped and their proclivity as pro-tumorigenic or anticancer immune cells is still unclear [96,97,98], γδ T cells are potential agents against cancer cells [99]. Various studies have reported the chameleon-like nature of γδ T cells, pointing at the flexibility they exhibit in TME [100,101,102]. The phenomenon has been successfully explained in squamous cell carcinoma [103] and colorectal cancer [104], where it has been suggested that TME conditions can affect the proliferation and functional nature of γδ T-cells. For example, a recent study has highlighted two distinct subtypes of γδ T cells, namely, antitumoral IFN-γ–producing γδ T cells (γδIFN cells) and IL17-producing γδ T cells (γδ17 cells) [105]. Interestingly, it has been shown that Vδ2 cells are activated, independent of MHC, by small lipid molecules, phosphoantigens (pAgs), which are derived from the mevalonate pathway [106,107,108]. Furthermore, Emmanuel Scotet et al. identified two different lipid-related ligands of Vɣ9Vδ2 TCR in tumor cells, namely apolipoprotein A1 (Apo-A1) and ATP synthase/F1-ATPase (high-affinity apo A-I receptor). These authors revealed that Apo-A1, which is abundant in high-density lipoproteins (HDL), is needed for the activation of Vɣ9Vδ2 T cells by tumors expressing F1-ATPase [109]. Similarly, a related study by Rodrigues et al. showed that Vδ2 T cells express low-density lipoprotein (LDL) receptors when they are activated and their functions can be modified once LDL attaches to their activated receptors. It has also been demonstrated that expression levels of IFN, NKG2D, and DNAM-1 are downregulated when Vɣ9Vδ2 T cells are treated with LDL-cholesterol [110]. Furthermore, host-derived lipids from lung-infiltrating CD1d + B-1a cells are able to induce γδ T cells for the induction of IL-17 A [111].

As a specialized type of T lymphocytes, Natural killer T cells (NKT cells) recognize lipid antigens presented through CD1d [112, 113]. NKT cells are divided into two distinct types, including I and II NKT cells, which regulate the immune response in the development and progression of tumor [114,115,116]. Both type I and type II NKT cells show intermodulation, but type I NKT cells are known to increase antitumor responses, while type II NKT cells are inclined towards pro-cancer activities [117], with some contextual exceptions where type I can also suppress tumor immunity [115, 118, 119]. However, tumor growth in TME is bound to consume more lipids to support its rapid proliferation and meet excessive energy needs. De novo lipid synthesis, greater and preferential fatty acid uptake from surrounding tissues in TME, and altered equilibrium of polyunsaturated fatty acids (PUFAs) and saturated fatty acids (SFAs) change the lipid repertoire of tumor cells, which can affect membrane fluidity, cell–cell interaction, and membrane protein landscape, subsequently affecting the downstream signaling cascade [120, 121] in cancers [122, 123]. The changes in lipid repertoire are also linked to the altered structure of bio-in cancer cells [124]. In this context, HFD rich in SFA can negatively affect the capability of DCs to activate naïve T cells [125], which is critical for antitumor responses. The availability of lipids is necessary for the NKT cells development [126] as mice deficient in lysosomal lipid transfer enzyme Niemann Pick C (NPC) 2 have a decreased number of type I NKT cells [127]. There is no doubt that excess lipid states lead to the activation of type I NKT cells, which generates a proinflammatory environment in obese patients [128], while CD1d−/−mice show reduced inflammation under similar conditions [129]. Furthermore, the antitumor potential of NKT cells in obesity is reduced and does not inhibit tumor growth [130]. However, human studies were unable to show any changes in the number of NKT cells in a hepatocellular carcinoma (HCC) model [131]. It is interesting to note that higher lipids increase the NKT cell proliferation, leading to proinflammatory responses, but an obese state reduces NKT cells, causing hindrance in tumor immunity. Further research in this direction is vital and has the potential to unveil anticancer drug targets.

Conclusion

Lipid metabolic features differ in their ability to initiate antitumor or pro-tumor responses in glucose-diminished TME infiltrated by cytotoxic T cells, Tregs, Tmem, and NKT cells (Fig. 1). It may be oversimplified to argue that lipids proliferate Teffs and regress Tregs/Tmem in complicated TMEs where diversified forms of lipids exist. The predisposition of TME to attract and utilize excessive lipids leaves little energy for Teffs to expand. In a complex TME, the identification of lipid-based T-cell drug targets is context-dependent as the same genes or enzymes responsible for attracting lipids to T cells preferentially contribute to cancer cell fat intake. The type of lipid intake peculiarly affects different subsets of T cells. Formulation of a definitive hypothesis in this regard is too early. It is necessary to explore the expression difference of the same gene in different subtypes of T cells under TME to mark it as a drug target. Although CD36, SREBPs, PD-L1/2, FABP5, CPT-1, ACC1, GLUT1, and FAS has shown promising prospects to be potential metabolic drug targets of cancer, their context-dependence and varied implications in T-cell subtypes urge for more research. Identifying the differential expression of lipid-related genes, gatekeepers, and enzymes on T-cell subsets and cancer cells that can be manipulated to draw clinical gains presents an opportunity for future research.

Fig. 1
figure 1

Differential T-cell responses after metabolic rewiring in response to lipids within TME. Raised expression of CD36 leads to T-cell dysregulation and immune evasion in TME. Excess lipids generate differential responses in T-MDSCs and S-MDSCs, leading to the activation of Tregs and causing T-cell dysregulation. Similarly, PD-1/PD-L1 complex converts glycolysis to FAO, thereby affecting T-cell exhaustion and playing a role in tumor progression. Under the state of metabolic competition within TME, T cells may opt to consume more lipids to compensate for depleted glucose and in-process lead to T cells activation.TME: Tumor microenvironment; FAO: Fatty acid oxidation; T-MDSCs: Tumor MDSCs; S-MDSCs: Splenic MDSCs; oxLDL: Oxidized low-density lipoproteins; Tregs: T regulatory cells; SREBPs: Sterol-responsive element-binding proteins. Raised OxLDL intake: Raised intake of oxidized low-density lipoproteins (OxLDL) increases lipid peroxidation downstream and causes ferroptosis in T cells, leading to suppression of antitumor immunity. Lipids induce metabolic reprogramming: Myeloid-derived suppressive cells (MDSCs) rewiring lipid metabolism have a direct role in remodeling TME by repressing T cells as well as natural killer (NK) cells and generating regulatory T cells (Tregs) and regulatory B cells (Bregs), supporting immune escape. Metabolic reprogramming: As glycolysis is necessary for T cells to obtain the required amount of energy for proliferation, PD-1/PD-L1–involved metabolic reprogramming from glycolysis to fatty acid oxidation (FAO) causes T-cell exhaustion, which is likely to sabotage antitumor immunity in TME. Antitumor metabolic reprogramming: Sterol-responsive element-binding proteins (SREBPs) increase activated T cells with greater metabolic demands by enhancing fatty acid synthesis (FAS), therefore driving the antitumor immunity.

Full size image