Amino acid metabolism reprogramming: shedding new light on T cell anti-tumor immunity

Nutrient deficiency is a selective stress that shapes the evolution of most cellular processes [25]. Because amino acids play a crucial role in maintaining cellular homeostasis, different species have developed different mechanisms to detect amino acid abundance over the course of evolution. Eukaryotic cells are equipped with nutrient sensors such as mTOR, a conserved serine-threonine kinase that is activated when amino acids are abundant and regulates various anabolic processes required for growth [26]. Through binding with protein binding partners, mTOR can form mTOR complex 1 (mTORC1) or mTOR complex 2 (mTORC2) [27]. mTORC1 predominantly regulates metabolic reprogramming [28] and selectively regulates the differentiation of Th1, Th17 and cytotoxic CD8 + T cells [29]. mTORC2 mainly regulates actin polarization and endocytosis [30], and is closely associated with differentiation of Th2 and memory CD8 + T cell and migration of regulatory T cells (Tregs) [31, 32]. In conclusion, as the core component of mTORC1 and mTORC2, mTOR plays a key role in amino acid metabolism and immunity [33]. Not all amino acids can regulate mTOR activity; only leucine, arginine, lysine, glutamine, methionine, and tryptophan can regulate mTOR activity [34‐38]. Inhibition of mTOR affects protein translation, cell proliferation, differentiation, effector functions, and many other factors.

Protein translation is regulated by mTOR through two independent mechanisms: inactivation of the EIF4E-binding protein (4E-BP1) and activation of p70 S6 kinase (S6K1) [39]. Ornithine decarboxylase (ODC) translation is controlled by the mTOR/4E-BP1 axis [40]. Polyamine metabolism is affected by mTOR by influencing ODC translation, which plays an important role in T cell activation and differentiation [41].

T cell proliferation is regulated by mTOR through two pathways. On the one hand, inhibition of mTOR activity impairs the activation of the c-Myc signaling pathway, leading to metabolic stress and defective T cell proliferation [42]. On the other hand, mTOR forms an intracellular complex with the serine-threonine kinase aurora B and survivin from the costimulatory molecule CD28, which is responsible for allowing the G1-S transition in antigen-stimulated T cells [43].

Inhibition of mTOR promotes differentiation of T cells into immunosuppressive Tregs. One Study has shown that mTOR inhibition promotes Treg production via Rag and Rheb GTPases [35]. Another study demonstrated that mTOR inhibition promotes the differentiation of naive CD4 + T cells into Foxp3 + Treg cells, which have a suppressive function in vivo, whereas mTOR signaling activation supports the differentiation of naive cells into Th1 cells [44]. In conclusion, mTOR inhibition induced by amino acid deprivation shifts the balance between Th1 and Treg production toward the Treg phenotype.

GCN2 is another eukaryotic amino acid sensor that, unlike mTOR, directly senses the depletion of individual essential or nonessential amino acids in cells by binding to uncharged cognate tRNAs [45]. In eukaryotes, the GCN2 and mTOR pathways are major regulatory switches that determine protein synthesis in response to fluctuations in amino acid levels [26].

In contrast to the mTOR regulatory mechanism, the absence of any amino acids activates GCN2 kinase activity [36]. As eukaryotic amino acid sensors, GCN2 and mTOR have similar effects on T cell function. GCN2 also regulates protein translation, proliferation, differentiation, and effector functions of T cells.

The effect of GCN2 activation is mainly mediated by phosphorylation of the downstream target eukaryotic initiation factor 2 (eIF2α). Amino acid depletion triggers signaling through GCN2 kinase and inhibits cyclin D3 [46] and cyclin-dependent kinase 4 (CDK4) through eIF2α phosphorylation, leading to reduced Rb protein phosphorylation, low E2F1 expression, and cell cycle arrest [47, 48]. However, phosphorylated eIF2α leads to downregulation of T-cell receptor (TCR) CD3ζ in CD8 + T cells [49, 50], which further leads to downregulation of Jak-3 and decreased translocation of NFκB-p65, ultimately impairing the proliferation and interferon-γ (IFN-γ) production of T cells [51]. Furthermore, phosphorylated eIF2α fails to bind to methionyl tRNA, which blocks the translation initiation of most mRNAs [52], but selectively enhances the translation of a few transcripts, such as activating transcription factor 4 (ATF4) [36]. Specifically for T cells, ATF4 promotes metabolism reprogramming of T cells, including upregulation of glycolysis, oxidative phosphorylation, and glutaminolysis, thus providing substrates and energy for anabolism [53]. The increase in ATF4 translation has a protective effect on T cells and partially neutralizes the adverse effects of amino acid depletion on T cells to a certain extent.

In addition to cell cycle arrest and effector impairment, GCN2 activation blocks Th17 differentiation [54] and drives de novo differentiation of Foxp3 + Tregs [55]. Furthermore, GCN2 kinase directly activates mature Tregs, which in turn exert immunosuppressive effects in a PD-1/PD-L1-dependent manner [56]. In conclusion, GCN2 appears to promote formation and immunosuppressive activity of Tregs, as well as inhibit effector T cells.

PD-1 and PD-L1 are extensively studied immune checkpoints. PD-1 is expressed on T cells, and its cognate ligand PD-L1 is expressed on target cells such as cancer cells. Binding of PD-1 to PD-L1 leads to inhibition of TCR-related signaling molecules, resulting in T cell depletion and protection of target tissues from T cell-mediated damage [57]. A variety of metabolic processes have been found to regulate the expression of PD-1/PD-L1 in the TME, such as glucose [58] and prostaglandin E2 (PGE2) [59] metabolism. Amino acid metabolism also affects the expression of PD-1/PD-L1.

Different amino acids regulate PD-1/PD-L1 through different mechanisms. Under glutamine restriction, the level of intracellular glutathione (GSH) decreased, which led to the upregulation of PD-L1 in tumor cells, and then inhibited T cell activity [60]. The relationship between glutamine and PD-L1 is also demonstrated by the fact that upregulated PD-L1 can return to normal levels after glutamine recovery [61].

Dendritic cells (DCs) expressing IDO, a key enzyme in the tryptophan metabolism pathway, can activate Tregs, which in turn upregulates the expression of PD-L1 on DCs and enhances the immunosuppressive effect of Tregs through the PD-1/PD-L1 interaction [56, 62]. Indoleamine 2, 3-dioxygenase inhibitors were found to activate CD8 + T cells and downregulate PD-1 expression by increasing tryptophan levels and degrading the PD-1 transcriptional activator NFATc1 [63].

In addition, multiple amino acid deletions inhibit mTOR activity and Akt phosphorylation, leading to Forkhead box O (FOXO) transcription factor activation and upregulation of PD-1 expression in Tregs. Subsequently, PD-1 binds to the ligand, activates the lipid phosphatase phosphatase with tensin homology (PTEN) in Tregs, inhibits phosphatidylinositol 3-kinase (PI3K) activity, and blocks phosphorylation at another Akt activation site to maintain Akt inhibition, forming a feedback loop [64]. This initiates a stable state of self-sustaining inhibition in Tregs, which is maintained by the circulation between PD-1 and Akt, leading to the sustained suppression of antitumor immunity.

In conclusion, amino acid depletion can lead to the upregulation of PD-L1 and PD-1 thereby inhibiting antitumor immunity.

To prevent aberrant signaling that may adversely affect cell homeostasis, protein expression and activity must be strictly regulated. These regulatory mechanisms include epigenetic modifications of the genome and PTMs of proteins that determine the translational ability of transcripts and function of proteins, respectively. The core of epigenetics is the modification of histones and nucleic acids, which together regulate chromatin structure and gene expression, and produce genetic phenotypic changes without altering the DNA sequence [65]. Downstream of epigenetics is an additional level of regulation called PTM, which allows for the most refined and dynamic control of protein biology, including localization, conformation, interaction, and activation [66]. Amino acids are involved in a variety of epigenetic processes and PTMs associated with T-cell immunity (Table 1).

The main epigenetic modifications involved in amino acid metabolism are methylation, demethylation, and acetylation. The methylation of DNA and histones is dependent on methionine because its metabolite S-adenosyl methionine (SAM) is a universal methyl donor [37]. Loss of SAM in cells results in decreased Th17 polarization and death of CD8 + T cells [72]. Proline plays a role in epigenetic regulation by inducing specific histone methylation patterns [67]. DNA methylation can be reversed by removal of oxidized methylated bases by Tet proteins, a process that requires α-ketoglutaric acid (α-KG) [68]. α-KG is a TCA cycle intermediate, and glutamine is the main source of α-KG when glucose is scarce [73]. A recent study found that glutamine depletion is accompanied by a decrease in α-KG, which inhibits histone demethylation [74]. Histone acetyltransferases use acetyl-CoA to provide acetyl groups for acetylation. Glutamine and branch chain amino acids can be metabolized to produce acetyl-CoA [37]. The absence of these amino acids affects important cellular signaling pathways.

PTMs involving amino acids include hypusination, nitrosylation, O-GlcNAcylation, and glycosylation. These modifications modulate immune processes through various mechanisms, including regulating the activity of enzymes or signaling molecules, altering protein interactions, determining subcellular localization, and controlling protein translation.

Hypusine is a modified lysine formed by the reaction of lysine with spermidine [75]. SAM generated from methionine can be further transformed into spermidine, and hypusination mediated by it occurs only on the translation initiation factor eukaryotic initiation factor 5A (eIF5A) [69]. As a highly conserved protein, eIF5A is required for the elongation of the translation of specific mRNA transcripts and affects protein expression in a variety of immune cells [76]. Activation of eIF5A is required for the differentiation of CD4 + T cell subsets, and eIF5A inhibition leads to the selective reduction of Th1 cells [77].

Arginine metabolism produces nitric oxide (NO), which further reacts with superoxide to form peroxynitrite (ONOO⁻), which rapidly results in two PTMs on proteins: nitrosylation of thiol and nitration of tyrosine [59, 70]. One study has found that ONOO⁻ modifies chemokines such as CCL2 by nitrification or nitrosylation to inhibit T cell infiltration in tumor tissues [78]. In cells, ONOO⁻ has been found to regulate microsomal prostaglandin E synthetase 1 activity through post-translational nitration of tyrosine modification and positively regulate PGE2 production, which mediates the stimulation of immunosuppressive Tregs [59].

Glucose and glutamine can be metabolized to uridine diphosphate n-acetyl glucosamine (UDP-GlcNAc), a substrate for O-GlcNAcylation. O-GlcNAcylation is one of the most abundant PTMs [38]. Many signaling molecules regulated by O-GlcNAcylation are fundamental to T cell survival and biological function, such as c-Myc, NFAT, and NF-κB [38, 79, 80]. Among them, c-Myc plays a crucial role in the clonal expansion and effector function of T cells [81], whereas NFAT [82] and NF-κB [83] signal transduction are key regulators of T-cell activation.

Glutamine is required for N-linked glycosylation and is essential for protein stability and function. Aberrant glycosylation has been observed to affect cell proliferation and growth, as in the IL-3 receptor, where abnormal branching of the sugar chain leads to altered downstream signaling [71]. The depletion of amino acids leads to the weakening of glycosylation of proteins and lipids, thus affecting the function of many proteins.

There is increasing recognition that in addition to synthesizing proteins and peptides, some amino acids can act as signaling molecules and regulate key metabolic pathways that are necessary for immunity [84]. When amino acids are depleted, these biological processes can be directly inhibited, resulting in a disordered cell structure and function.

Amino acids are involved in the synthesis of many substances such as nucleotides, SAM, and GSH. Methionine is mainly involved in SAM synthesis and regulates epigenetic inheritance. Serine and glutamine are also involved in nucleotide synthesis. Serine is an important substrate for purine synthesis, and the entry of serine into one-carbon unit metabolism is a checkpoint for CD8 + T cell proliferation [85, 86]. Unlike serine, glutamine restriction mainly affects the CD4 + T cell subgroup. Glutamine depletion promotes differentiation of CD4 + T cells into Foxp3 + Treg cells by reducing nucleotide synthesis [87]. GSH is the most abundant antioxidant, and is synthesized from glycine, glutamate, and cysteine. Cysteine is a rate-limiting substrate of GSH synthesis [37]. GSH deficiency increases reactive oxygen species (ROS) and disrupts intracellular redox homeostasis, thereby affecting cell survival and function. Because of the different degrees of oxidative stress in Tregs and Th17 cells, GSH deficiency can promote Treg differentiation and inhibit Th17 differentiation, resulting in an imbalance between Treg and Th17 differentiation [88].

Amino acid deficiency can directly affect the expression levels of genes and downstream proteins that are closely related to the function of T cells. One previous study showed that T cells exhibit glutamine-dependent expression of cell surface activation markers CD25, CD45RO, and CD71, as well as IFN-γ and TNF-α production [89]. Besides this, glutamine deficiency in hepatocellular carcinoma upregulates the expression of LAG3 and induces functional failure of γδT cells, which are involved in mediating antitumor responses and are associated with positive prognosis [90].

Immune cell proliferation and function occur through activation of key enzymes and proteins. Amino acid deficiency or some amino acid metabolizing enzymes can also directly affect the activity of these proteins or enzymes. Glutamine depletion leads to decreased activity of extracellular signal-regulated kinase (ERK) and c-Jun amino-terminal kinase (JNK) kinases, which further results in inhibited transcription of proliferation-related genes [91]. Arginine deprivation reduces F-actin content and CD2 and CD3 accumulation in T cell immune synapses by impinging on cofilin dephosphorylation, ultimately reducing proliferation and cytokine synthesis [92]. In addition, the tryptophan-metabolizing enzyme IDO can selectively reduce the activity of electron transport chain complex I, limiting ATP production in CD8 + T cells and leading to the inhibition of effector functions [93].

In tumor cells, more than 95% of tryptophan is degraded via the Kyn pathway [94]. This ultimately leads to the biosynthesis of the cofactor nicotinamide adenine dinucleotide (NAD) [95]. This metabolic pathway produces metabolites, including Kyn, 3-hydroxykynurenine (3-HK), 3-hydroxyanthranilic acid (3-HAA), and quinolinic acid, all of which are collectively known as kynurenines. Among them, Kyn is the most widely studied protein associated with antitumor immunity. Tryptophan is catalyzed to Kyn by IDO1, IDO2, or tryptophan 2, 3-dioxygenase (TDO), and is the rate-limiting enzyme in this process [96]. Because IDO1 has the highest expression level and activity, many studies have identified IDO1 as a major cause of tryptophan depletion [97, 98].

On the one hand, Kyn can directly produce toxic effects on immune cells, inhibit the proliferation of T cells and induce their apoptosis. By arresting the cell cycle in the middle of the G1 phase, Kyn selectively inhibits proliferation of activated T cells, whereas resting T cells are unaffected and subsequently activate normally [99]. This inhibitory effect on T cell proliferation is concentration-dependent [100]. Simultaneously, Kyn can lead to changes in the intracellular redox balance and induce cell apoptosis through ROS production [101].

On the other hand, Kyn, as an endogenous aryl hydrocarbon receptor (AhR) agonist, contributes to immunosuppression of the TME and supports tumor immune escape. Kyn combined with AhR promotes the differentiation of CD4 + T cells into Tregs and inhibits the function of effector T cells [102, 103]. In Th17 cells, AhR activation promotes downstream IL-22 production and differentiation [104]. In addition, activated AhR controls the transcriptional program associated with tolerant DCs [105], which in turn inhibits T-cell immune activity. Activated AhR can also up-regulate the expression of PD-1 in CD8 + T cells [106, 107], inhibit T cell activity, and promote immune tolerance.

Of note, many in vitro experiments have used much higher concentrations of Kyn than the actual in vivo concentrations. The concentration of Kyn in human tumors is only in the low micromolar range, well below the 1 mM required to induce T cell apoptosis in vitro [108], which calls into question its clinical significance. However, treatment of mice with specific Kyn-depleting enzymes improved tumor growth and enhanced immunotherapy, suggesting that in vivo concentrations of Kyn-depleting enzymes may still be clinically relevant [98]. Further in vivo experiments are needed to verify the specific effects of Kyn.

Other products of the Kyn metabolic pathway, such as 3-HAA and 3-HK, also have immunosuppressive effects. 3-HAA inhibits T cell proliferation by suppressing the activation of pyruvate dehydrogenase and NF-κB [109], and impairs T cell activation by inhibiting TCR-mediated Ca²⁺ signaling in T cells [110]. In addition, 3-HAA stimulated TGF-β production and promoted Treg formation [111]. Notably, 3-HAA can induce selective apoptosis of Th1 cells by promoting the release of cytochrome C and activation of caspase 8 [112, 113] and mediate apoptosis via ROS production [114]. Moreover, 3-HK is also thought to reduce CD8 + T cell proliferation and maturation of naive T cells into CD8 + T cells [115].

There are two main arginine metabolic pathways: arginine/Arginase (ARG)/ODC/polyamine and arginine/ inducible nitric oxide synthase (iNOS)/NO [116]. On the one hand, arginine produces ornithine through the urea cycle, and ornithine is used to synthesize polyamines, which are important for T cell growth [37]. On the other hand, arginine forms NO through iNOS, which further combines with O₂⁻ to produce reactive nitrogen species, such as ONOO⁻ [117].

NO can block T cell function by interfering with the IL-2 receptor signaling pathway, thereby preventing the activation of multiple signaling molecules, including STAT5, Erk, and Akt [118]. In addition, NO can induce CD4 + T cells to differentiate into CD4 + Foxp3 + Treg cells through the NO/p53/IL-2/OX40/survivin signaling pathway, and iNOS inhibition completely inhibits NO-induced differentiation [119].

As mentioned in Sect. " Amino acid starvation regulates immune function through epigenetic and post-translational modifications (PTMs)", ONOO⁻ can regulate the immune response through PTMs. In addition, owing to its strong oxidizing effect, it can inhibit the activation-induced protein tyrosine phosphorylation or through nitration, inhibit a component of the mitochondrial permeability transition pore causing the release of cytochrome C and other death promoting factors, resulting in T cell apoptosis induction [120]. ONOO⁻ can also lead to changes in the expression of TCR, IL-2R, and CD8 molecules, thus damaging the T cell signaling pathway and inhibiting its effector function [78].

In addition to the above amino acid toxic metabolites, a recent study also found that the glycine derivative metabolite methylglyoxal can enter CD8 + T cells and combine with arginine. The consumption of free L-arginine and the destruction of the modification of proteins containing L-arginine can greatly inhibit the activation and function of CD8 + T cells [121]. In addition, cancer cells with abnormal glutamate decarboxylase 1 expression can use glutamine to synthesize gamma-aminobutyric acid (GABA) [122]. A main function of GABA is as an important neurotransmitter; in tumor tissues, GABA can activate the GABAB receptor and inhibit the activity of GSK-3β, resulting in enhanced β-catenin signaling [122]. Such GABA-mediated β-catenin activation can stimulate tumor cell proliferation and inhibit the intratumoral invasion of CD8 + T cells [122].

In conclusion, amino acid metabolism plays a significant negative regulatory role in T cell immunity. On the one hand, amino acid depletion inhibits immune cell function in many aspects; on the other hand, some toxic metabolites accumulate in the TME, further inhibiting T cell immunity and causing immune tolerance.

Leukemia cells are highly dependent on extracellular asparagine because of a deficiency in asparagine synthase, the only enzyme capable of asparagine synthesis [164]. L-asparaginase, which directly targets asparagine metabolism, has been successfully used in chemotherapy for ALL [72]. The successful application of asparaginase demonstrates the feasibility of targeting amino acid metabolism and is expected to promote the clinical transformation of metabolic enzyme-based amino acid deprivation therapy.

Similar to dependence of leukemia cells on exogenous asparagine, some solid tumors, including melanoma and hepatocellular carcinoma, depend on extracellular arginine for survival owing to the lack of de novo synthesis of arginine [165]. Therefore, two polyethylene glycol (PEG)-conjugated arginine decomposers have been developed to deplete the extracellular arginine pool. PEG functional modification can reduce its immunogenicity, prolong its blood circulation time and half-life in vivo, and improve its antitumor effect in vivo. PEG-Arg I (BCT-100) can convert arginine to ornithine, resulting in rapid depletion of extracellular and intracellular arginine libraries and reduced proliferation of tumor cells after monotherapy [166]. However, PEG-Arg I inhibited T cell proliferation and blocked T cell responses indirectly by inducing accumulation of bone marrow-derived suppressor cells (MDSCs). Therefore, L-arginine depletion therapy has a dual role in cancer therapy, with a risk of immunosuppression. ADI-PEG20, another drug that converts arginine to citcitine, has an unsatisfactory therapeutic effect because of the increased compensatory production of endogenous arginine caused by the overexpression of arginine succinic synthase 1 after monotherapy; however, it can achieve a certain effect in arginine succinic synthase 1-deficient tumor cells [136].

In addition, cystine or cysteine therapy leads to intracellular GSH depletion and ROS accumulation by increasing AMPK phosphorylation and decreasing mTOR phosphorylation, resulting in cell cycle arrest and cell death in various cancers [136].

Amino acid transmembrane transport is mediated by various amino acid transport systems within the solute carrier (SLC) superfamily [167]. More than 60 SLC proteins have been identified as amino acid transporters [168]. The relationship between amino acids and their transporters is complex. An amino acid can be transported by several different transporters, whereas a transporter passes through multiple amino acids [143]. In addition, tumor cells and immune cells have different expression levels of the same transporter [72]. In tumor cells, the imbalance of amino acid transporters leads to metabolism reprogramming, which changes the intracellular amino acid level and is an important mechanism leading to tumor development [168]. Therefore, amino acid transporters are reliable targets for tumor therapy.

Among the upregulated amino acid transporters in cancer cells, LAT1 (SLC7A5) is notable for cancer-specific expression. LAT1 transports almost all the neutral amino acids. BCH (2-aminobicyclo-(2,2,1)-heptane-2-carboxylic acid) was identified as an inhibitor of LAT1. It inhibited the proliferation of tumor cells in a dose-dependent manner. It also inhibited mTOR phosphorylation and induced cell cycle stasis in the G1 phase [137]. Another inhibitor of LAT1, JPH203 (KYT-0353), has a high affinity more than a thousand-fold higher than that of BCH [169]. JPH203 can lead to inhibition of the mTOR system, which leads to changes in downstream signaling pathways, in which cell cycle regulators such as cyclin-dependent kinase (CDK) 1–6 are considered to be the most downregulated kinases upstream [170]. JPH203 has shown encouraging results in Phase I clinical trials against advanced solid tumors and is currently being used in Phase II studies (UMIN000034080) [138, 171]. LAT1 is also unique in that its expression is tumor-specific and, therefore, can be used for the delivery of antitumor drugs. For example, because melphalan is transported by LAT1, antitumor L-phenylalanine mustard melphalan was designed to improve the cellular uptake of nitrogen mustard [169]. Similarly, the precursor of sesamol was designed by para-binding of sesamol to L-phenylalanine via a carbamate bond, which significantly enhanced its uptake and toxic effect on tumor cells [172]. In conclusion, the LAT1-mediated prodrug delivery strategy facilitates the selective uptake of drugs to increase their intracellular concentration and antiproliferative activity by targeting tumor cells that overexpress the LAT1 protein.

SLC1A5 (ASCT2) is the main glutamine transporter, and a variety of drugs have been developed for this transporter, such as γ-L-glutamyl-p-nitroanilide (GPNA), V-9302, and benzylserine. Selective pharmacological drugs have been developed based on the first-generation low-efficiency glutamine transport antagonist GPNA, and V-9302, a GPNA derivative, improved the ability to inhibit glutamine uptake in cells by approximately 100 times [173]. Blocking ASCT2 with V-9302 attenuates cancer cell growth and proliferation, and increases cell death and oxidative stress, which together promote antitumor responses [141]. Noteworthily, CD8 + T cells upregulate the glutamine transporter SLC6A14 via a compensatory pathway and maintain glutamine uptake and effector functions [28]. In other words, V-9302 selectively blocks the uptake of glutamine by tumor cells but not by CD8 + T cells and promotes the synthesis of the cellular antioxidant glutathione, thereby improving the efferent function of CD8 + T cells [174]. It has also been found that inhibiting glutamine metabolism with V-9302 may increase the expression of PD-L1 in tumor cells, thus inactivating T cells [60]. The ASCT2 inhibitor benzylserine can significantly reduce glutamine transport in tumor cells, inhibit the mTOR signaling pathway, and reduce the expression of cell cycle regulators, thus inhibiting cell cycle progression [142].

The amino acid transporter SLC6A14 transports all neutral amino acids, as well as the cationic amino acids lysine and arginine, and is a novel drug target. Alpha-methyltryptophan (α-MT), an inhibitor of this transporter, induces amino acid starvation and autophagy in tumor cells by blocking SLC6A14, inhibiting mTOR signal transduction, inducing amino acid starvation, and inhibiting tumor cell growth and proliferation [143].

SLC3A2 forms a heterodimer amino acid transporter with SLC7A5 or SLC7A11. Although SLC7A5 and SLC7A11 are actual transporters, they require SLC3A2 as a partner to recruit them into the plasma membrane [163]. Treatment with the humanized anti-SLC3A2 monoclonal antibody IGN523 showed antitumor efficacy in leukemia-derived and non-small cell lung cancer models [72]. IGN523 causes tumor cell death through NK cell-mediated cytotoxicity and inhibits the uptake of amino acids such as phenylalanine by tumor cells [175].

The cystine glutamate reverse transporter SLC7A11 (xCT) helps fight oxidative stress by promoting GSH-mediated antioxidant defenses. Therefore, xCT may be a promising target for cancer therapies. The xCT transport inhibitor sulfasalazine induced a decrease in cysteine and GSH and led to enhanced mitochondrial metabolism, resulting in increased ROS production, which triggered oxidative damage [176].

Glutamine antagonists have a long history of use. 6-Diazo-5-oxygen-L-deamine, a glutamine antagonist, inhibits glutamine-based enzymes [177]. In addition, two other compounds, acivicin and azaserine, are also glutamine antagonists [96]. However, owing to the important role of glutamine metabolism in normal tissue physiology, these compounds also cause varying levels of gastrointestinal toxicity, myelosuppression, and neurotoxicity, and were therefore deprecated [96].

The prodrug developed on this basis, JHU083, is a well-tolerated, brain-penetrating glutamine antagonist, and a promising new drug for treatment [178]. JHU083 can release 6-diazo-5-oxygen-L-deamine when cleaved by tumor cathepsin, thus playing a specific killing role in tumors [145]. Notably, this drug can differentially metabolize cancer cells and T cells, not only starving the cancer cells, but also making the TME a more suitable microenvironment for effector T cells, thus enhancing their attack on the tumors [179].

In addition to inhibiting IDO/TDO, the key enzyme for Kyn production, other strategies targeting the Kyn/AhR pathway have been proposed, including decomposition of Kyn and inhibition of AhR.

PEG-KYNase is a recombinant enzyme that degrades Kyn into an immunoinert metabolite. PEG-KYNase has been shown to be therapeutic in multiple mouse models for tumor when used alone or in combination with checkpoint blocking [160]. Degradation of Kyn inhibits AhR activation. The modified kynureninase can degrade extracellular Kyn and has shown remarkable efficacy in mouse tumor models [201].

Aryl hydrocarbon receptor antagonists also attenuate immunosuppression and inhibit tumor growth. IDB-AhRi is an AhR antagonist that blocks the nuclear transposition of AhR and increases the expression of IFN-γ and TNF-α. IDB-AhRi increases tumor infiltration by CD8 + T cells and decreases Tregs and tumor-associated macrophages in mouse models of colon cancer [161]. Selective AhR inhibitors (CH223191) can block the immunosuppressive effects of Tregs [162], inhibit Th17 differentiation [202], and reduce PD-1 expression [106].

The targeted therapy of amino acid metabolism combined with immunotherapy still mainly focuses on the three main amino acids, glutamine, arginine, and tryptophan. Glutamine transporter inhibitor V-9302 and GLS inhibitor BPTES, when used in combination with anti-PD-L1 antibodies, strongly promoted the effector function of T cells [60]. Another study found that the simultaneous transport of anti-PD-L1 antibodies and V9302 with molybdenum disulfide significantly promoted the infiltration of CD8 + T cells and strongly inhibited tumor growth [212]. Immunotherapy of the PD-1 checkpoint combined with glutamine-targeting JHU083 showed significantly increased response rates compared to those of PD-1 monotherapy [179]. These results demonstrate the correlation between tumor glutamine metabolism and antitumor immunity and suggest that the combined targeting of glutamine metabolism and PD-L1 is a promising therapeutic approach that can significantly enhance the antitumor effect.

CB-1158, an ARG1 inhibitor, showed a highly potent antitumor effect when used in combination with anti-PD-1 antibodies or anti-CTLA-4 antibodies [193]. In addition, the subtype of arginase expressed by T cells, mitochondrial ARG2, can regulate T cell activation, antitumor cytotoxicity, and memory formation in CD8 + T cells, independent of extracellular arginine availability. In addition, the specific loss of ARG2 in CD8 + T cells has a strong synergistic effect with PD-1 blocking in tumor growth control [213]. Inhibition of ARG2 in conjunction with PD-1 blocking therapy may improve the response to immunotherapy.

One Study has shown that plasma Kyn:Trp ratio increases in patients with tumors during pembrolizumab treatment [107]. Another preclinical study showed that the IDO inhibitor 1-MT enhanced the efficacy of anti-CTLA-4 and anti-PD-1/PD-L1 treatments by increasing the infiltration and activation of CD8 + T cells at the tumor site [19]. These findings suggest that the high expression level of IDO and the corresponding increase in Kyn may be the underlying factors that induce tolerance to ICB therapy. In addition, numerous experiments have demonstrated that tumor cells expressing IDO can suppress immune cells through tryptophan starvation and that AhR is involved in tumor immune escape. Therefore, combination therapy targeting the Trp-Kyn-AhR axis in immunotherapy has strong translational rationality and good preclinical effects; however, the clinical trial effect of combination therapy is not satisfactory [214] and has reignited the combinatorial approach debate. This may be because of the role of TDO in immune escape, although its role in tumors is not as important as that of IDO. Therefore, the dual inhibitors of IDO and TDO may be more effective. A subsequent study has demonstrated that navoximod, a dual IDO1/ TDO inhibitor, induces a powerful antitumor immune response and inhibits tumor progression when combined with anti-PD-L1 antibodies [151].

Inadequate persistence of CAR-T cells in vivo leads to poor therapeutic outcomes and disease recurrence [215]. To optimize the in vitro CAR-T/NK cell expansion process for better clinical efficacy, two main aspects of combination therapy may be considered: one is to enhance the adaptability of immune cells to the TME, and the other is to enhance the cytotoxicity of immune cells.

In the TME, nutrient deficiency, accumulation of large amounts of toxic metabolites, hypoxia, and low PH create a unique environment that promotes tumor growth and suppresses immunity. In combination with these two factors, CAR-T/NK cells have difficulty generating adaptive and durable antitumor responses. It has been found that poor proliferation and persistence of T cells is one of the main reasons why adoptive cell therapy has no or a weak response [216]. Therefore, in vitro cultured immune cells also require potent metabolic capacity to enhance their adaptability to harsh environments. The upregulation of amino acid transporters SLC1A5 and SLC7A5 may improve the function of CAR-NK and CAR-T cells [28]. In a recent study, authors found that re-engineering CAR-T cells to express SLC7A5 or SLC7A11 can promote CAR-T cells proliferation and IFN-γ release under low tryptophan or cystine conditions [217]. Thus, loading CAR-T/NK cells with amino acid transporters may enhance their resilience to amino acid depletion and improve their function. Another approach is to consider the loading of functional amino acid-metabolizing enzymes. Because of the low expression of argininosuccinate synthase and ornithine transcarbamylase, T cells are susceptible to arginine depletion [46]. Thus, T cells can be reengineered to express functional argininosuccinate synthase or ornithine transcarbamylase enzymes in conjunction with different CARs, which increases CAR-T cell proliferation without reducing cytotoxicity.

Another major strategy is to enhance immune cytotoxicity. This can be achieved through simultaneous or sequential application of drugs that directly target amino acid metabolism. For example, CD19 CAR-T had no effect on IDO-positive tumors but was restored in combination with 1-MT [20]. In contrast, CAR-T cell cytotoxicity can be enhanced by regulating T cell metabolism during in vitro expansion. In vitro-amplified CAR-T cells showed phenotypic heterogeneity, most of which were effector memory T cell or effector T cell subpopulations, and naïve T cell and central memory T cell populations, which showed stronger cytotoxicity, were very low. Transformation of differentiated subsets is closely related to the metabolic adaptability of T cells. One study demonstrated that pretreatment of CAR-T/NK cells with the glutamine antagonist 6-diazo-5-oxygen-L-deamine in vitro modulated the differentiation phenotype and enhanced metabolic adaptability [215]. Similarly, a recent study found that dynamic in vitro culture can enhance the antitumor activity of immune cells [218]. Dynamic culture can increase glutamine metabolic flux and promote ATP production. These cells are in a high metabolic state to produce increased amounts of energy. These findings provide new insights into the expansion of immune cells in vitro. However, it is worth noting that over-enhancing the cytotoxicity of CAR-T may produce toxic levels of cytokines and over-activation of immune system, leading to cytokine release syndrome or neurotoxicity [216]. Therefore, it is very important to have the right window of treatment. With the continuous optimization of CAR molecules and the development of combination therapy, it is believed that CAR-T cell therapy will be safer and more efficient.