1 Introduction

Oncometabolites usually refer to metabolic byproducts that are abnormally accumulated in cancer cells and play a crucial role in facilitating cancer progression [1]. R-2-hydroxyglutarate (R-2-HG) is the first well-defined oncometabolite characterized by its elevated levels in cancer cells and its cancer-promoting effects. Thereafter, 2-HG, succinate and fumarate have been generally recognized as bona fide oncometabolites due to their deregulated accumulation in certain types of cancer cells [1, 2]. Recently, increasing evidence has revealed that the functions of these oncometabolites extend beyond just accumulating in cancer cells. Most of oncometabolites can be secreted into the tumor microenvironment (TME) to influence cancer malignancy via various effects such as cancer immunity and angiogenesis. For example, 2-HG secreted by cancer cells can be imported into different cells including tumor infiltrated CD8+ T cells and vascular endothelial cells through specific transporters to exert its tumor-promoting effects [3, 4]. In addition, oncometabolites can also originate from non-tumor cells. For instance, lactate produced by cancer-associated fibroblasts (CAFs) can be utilized by cancer cells to fuel it energy and metabolic requirement [5]. These emerging evidence opens a new window to understand the biological significance of the rewired cancer metabolism and have catalyzed the conceptual advance in cancer therapy. Inhibitors specifically targeting mutant isocitrate dehydrogenase 1/2 (IDH1/2) to decrease 2-HG production, which mitigates 2-HG-mediated cancer-promoting roles in both cancer cells and non-cancer cells, have been approved for clinical use and represent a breakthrough in metabolism-targeted cancer therapies [6].

This review aims to discuss the recent advances in understanding functions of various oncometabolites generated by both cancer and non-cancer cells within the TME, and summarize the mechanistic understanding of oncometabolite-mediated cell crosstalk in the TME. With this, we hope to better understand the cancer pathology and inspire new therapeutic solutions for oncometabolite-driven cancer.

2 Oncometabolites generated by cancer cells

Reprogramming of metabolism is a hallmark of cancer [7]. As a consequence of the rewired metabolic network, oncometabolites are produced to guarantee brutal proliferation and invasion, as well as to modify a “friendly neighborhood” in the TME. Genetic abnormalities of metabolic enzymes in cancer cells constitute a primary mechanism leading to the aberrant accumulation of oncometabolites in cancer cells.

2.1 α-ketoglutarate (α-KG) structural analogs

Multiple genetic abnormalities in metabolic enzymes have been reported in human cancers. These mutations substantially alter the metabolic flux, such as disruption of the tricarboxylic acid (TCA) cycle, and result in the production of oncometabolites. Interestingly, accumulative evidence has shown that a set of metabolites exhibiting structural resemblance, such as R-2-HG, fumarate, succinate and itaconate (Fig. 1), share a common effect on inhibiting a variety of α-KG-dependent dioxygenases due to the competition with α-KG [2, 8]. This similarity also enables the interplay among different oncometabolites.

Fig. 1
figure 1

α-ketoglutarate structural analogs

Full size image

2.1.1 2-HG

2-HG consists of two enantiomers, R-2-HG and S-2-HG. Both enantiomers are detectable but at low levels in healthy mammalian cells, as controlled by 2-HG dehydrogenase (2HGDH). In a broad spectrum of human cancers, in particular low-grade glioma, acute myeloid leukemia (AML), chondrosarcoma and cholangiocarcinoma, mutations in IDH1/2 confers a neomorphic ability to convert α-KG to R-2-HG, leading to the accumulation of massive R-2-HG in cancer cells [8]. Unlike R-2-HG, generation of S-2-HG is mostly induced by hypoxia or acidic condition, where lactate dehydrogenase A (LDHA) and malate dehydrogenase 1/2 (MDH1/2) may acquire the ability to produce S-2-HG through their promiscuous enzymatic activity [9]. Reduced expression of 2HGDH in renal cell carcinoma also renders S-2-HG elevation [10]. In addition to cancer cells, CD8+ T-lymphocytes accumulate S-2-HG in response to T cell receptor (TCR) triggering [11].

Of note, functions of two 2-HG enantiomers are not fully overlapping. A shared characteristic is the ability to intervene the activity of α-KG-dependent dioxygenases, such as ten-eleven translocation (TET) family of 5-methylcytosine hydroxylases and prolyl hydroxylases, which is considered as the primary mechanism of oncogenic role of 2-HG in cancer [12]. The inhibition of TET enzymes and histone lysine demethylases (KDMs) by R-2-HG leads to altered gene expression, which contributes to the enhanced proliferation and impaired differentiation in erythroleukemia cells [13], and the promoted transformation in glioma [14]. Apart from the transcriptional change, inhibition of KDM4A and KDM4B by R-2-HG induces homologous recombination deficiency and sensitizes IDH1-mutant cancer cells to poly (ADP-ribose) polymerase (PARP) inhibitors [15]. R-2-HG also impairs the activity of AlkB homolog (ALKBH) DNA repair enzymes, sensitizing IDH-mutant cells to alkylating drugs like busulfan [16]. 2-HG accumulation also rewires tumor cell metabolism, through inhibiting α-KG-dependent metabolic enzymes like branched chain amino acids transaminases (BCATs). The inhibition of BCAT1 and BCAT2 leads to a decrease in glutamate level and a compensatory reliance on glutaminase, rendering a metabolic vulnerability[17]. It is worthy to mention that although both 2-HG enantiomers can bind to α-KG-dependent dioxygenases, the resulting effects can be enantiomer-specific. S-2-HG-mediated inhibition of prolyl hydroxylase domain-containing proteins (PHDs) stabilizes hypoxia-inducible factor-1α (HIF-1α), contributing to the upregulated expression of VEGF as well as other HIF-1α-targeted genes [18]. In contrast, R-2-HG diminishes HIF-1α level via stimulating EGLN/PHD hydroxylases activity [19].

Of note, the buildup of excessive 2-HG appears to have detrimental effects on cancer cells. For example, 2-HG interferes with the activity of the fat mass and obesity-associated protein, leading to a reduction in N6-methyladenosine RNA modification and decreased stability of MYC/CEBPA transcripts, exhibiting antitumor effects [20]. Likewise, inhibition of PHDs could lead to the accumulation of misfolded collagen proteins in the endoplasmic reticulum (ER), causing ER stress and increased apoptosis [21]. Hence, it is reasonable to assume that at a certain concentration, R-2-HG is expelled from cancer cells into the TME to extend its cancer-promoting capabilities. Lately, there is growing evidence that R-2-HG can be secreted into the TME and affect a variety of the non-cancer cells, including T cells, dendric cells (DCs), macrophages and endothelial cells, through mechanisms different from those identified in cancer cells (Fig. 2). In most findings, R-2-HG acts as a direct signaling mediator independent of the epigenome change, likely due to the limited uptake regulated by transporters [11].

Fig. 2
figure 2

Functions of R-2-HG in the TME. R-2-HG secreted by cancer cells, influences several cell types within the tumor microenvironment. In CD8+ T cells, R-2-HG diminishes cell viability and functional impairment via impairing lactate dehydrogenase (LDH) catalytic activity, inhibiting nuclear factor of activated T-cells (NFAT) signaling or downregulating STAT1 (signal transducer and activator of transcription 1) transcriptional activity. Endothelial cells also absorb R-2-HG, which invigorates mitochondrial respiration and fosters tumor angiogenesis. R-2-HG, R-2-hydroxyglutarate; TET, ten-eleven translocation enzymes; KDM, lysine-specific demethylase; PHDs, prolyl hydroxylase domain-containing proteins; BCATs, branched chain amino acids transaminases; NFAT, nuclear factor of activated T-cells; ATP5B, ATP synthase F1 β subunit; LDH, lactate dehydrogenase; TDO, tryptophan-2,3-dioxygenase; Kyn, kynurenine; L-Trp, L-tryptophan; AHR, Aryl hydrocarbon receptor

Full size image

T cell is the most investigated cell type due to the essential roles in antitumor immunity. Lines of evidence have shown that R-2-HG uptake by T cells leads to the reduced viability and dysfunction. While the functional impact is concrete, the molecular mechanism seems unclear. It has been shown that R-2-HG induces a perturbation of nuclear factor of activated T cells (NFAT) transcriptional activity and polyamine biosynthesis, resulting in suppression of T cell activity [3]. Migration of CD8+ T cells is impaired by R-2-HG through downregulated signal transducer and activator of transcription 1 (STAT1), which lowers the production of T cell-attracting chemokines CXCL9 and CXCL10 [22]. Very recently, R-2-HG has been shown to inhibit LDH activity in CD8+ T cells in a noncompetitive manner, resulting in the altered glycolytic flux, imbalanced NAD(H) and higher reactive oxygen species (ROS) production [23]. Besides T cells, R-2-HG reduces the expression of MHC-II and interleukin-6 (IL-6) in monocyte-derived DCs in IDH-mutant glioma [24]. R-2-HG also facilitates L-tryptophan uptake in macrophages via LAT-CD98, resulting in activation of AHR in IDH-mutant high-grade gliomas [25]. Classical and alternative pathways of complement activation are also hindered by R-2-HG through interfering with complement-mediated opsonization and phagocytosis [26]. Furthermore, our recent study reveals that R-2-HG fuels vascular endothelial cell migration to promote angiogenesis. The endothelial influx of R-2-HG triggers Na+ influx and subsequent Na+/Ca2+ exchange in the mitochondria, resulting the elevated mitochondrial respiration [4]. All these insights have ignited enthusiasm to develop inhibitors or vaccines targeting mutant IDH1/2. Several IDH1/2 inhibitors have been approved for the treatment of IDH1/2-mutant AML and cholangiocarcinoma, yet whether these treatments could boost immune responses or inhibit angiogenesis remained at preclinical stage and worthy clinical investigation.

2.1.2 Succinate

Succinate overproduction often results from succinate dehydrogenase (SDH) deficiency in cancer cells, which is characteristic of a rare subset of human cancers. Approximately half of these cancers carry primarily germline mutations in genes encoding SDH subunits, with SDHA mutations being the most prevalent, followed by subunits B, C, or D. In SDH-deficient cancers without SDH mutations, epigenetic silencing of the SDH complex likely plays a significant role [27]. Additionally, the function of SDH can also be affected by oncogenes. Mutation of the Von Hippel-Lindau (VHL) gene results in the upregulation of HIFs, leading to increased miR-210 levels and decreased SDHD expression [28], while activation of Myc deactivates SDHA through a mechanism involving SIRT3-mediated acetylation [29].

Succinate can cause posttranslational modification change via the production of succinyl-CoA (Fig. 3). Lysine succinyltransferases, like lysine acetyltransferase 2A (KAT2A) and carnitine palmitoyltransferase 1A (CPT1A), transfer the succinyl group from succinyl-CoA to specific lysine residual, resulting in succinylation that could affect protein stability, activity, and interactions, potentially facilitating cancer progression [30]. Succinylation of LDHA by CPT1A increases its stability by disrupting its ubiquitinated form binding to sequestosome 1, a ubiquitin-lysosome degradation pathway mediator, thus promoting migration and proliferation in gastric cancer cells [31]. Succinylation also affects the activity of key metabolic enzymes like glyceraldehyde-3-phosphate dehydrogenase (GAPDH), malic enzyme 2 (ME2), IDH2, and MDH2 by increasing succinyl-CoA synthetase GDP-forming subunit β (SUCLG2), dampening mitochondrial function in lung adenocarcinoma (LUAD) cells. High expression of SUCLG2 is associated with poor survival in LUAD patients, suggesting it as a promising therapeutic target [32]. Interestingly, R-2-HG competitively inhibits SDH, leading to increased succinylation and elevated levels of Bcl-2, thereby enhancing resistance to apoptosis in cancer cells [33], which underscores the intricate interplay among oncometabolites.

Fig. 3
figure 3

Functions of succinate in the TME. Succinate accumulation in cancer cells, often due to succinate dehydrogenase (SDH) deficiency, exerts multifaceted effects. It interferes with α-KG-dependent dioxygenases and succinylates lysine residues on proteins like LDHA and GAPDH, altering their stability or activity. Additionally, secreted succinate from cancer cells binds to succinate receptor (SUCNR1) through autocrine and paracrine mechanisms, promoting epithelial-mesenchymal transition (EMT), angiogenesis, and polarization of tumor-associated macrophages (TAMs). LDHA, lactate dehydrogenase A; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; ME2, malic enzymes 2; IDH2, isocitrate dehydrogenase 2; PI3K, phosphoinositide 3-kinases; HIF1-α, hypoxia inducible factor 1 subunit alpha; Arg1, arginase 1; Fizz1, resistin like beta; Mgl1, LLGL scribble cell polarity complex component 1; Mgl2, LLGL scribble cell polarity complex component 2; IL-1β, interleukin 1 beta; VEGF, vascular endothelial growth factor; STAT3, signal transducer and activator of transcription 3 ; ERK1/2, extracellular signal-regulated kinase

Full size image

Succinate is the natural ligand for SUCNR1 (also known as GPR91) and could activate SUCNR1 signaling through paracrine or autocrine mechanisms. Tumor-secreted succinate activates the SUCNR1-PI3K-HIF1-α axis in both cancer cells and macrophages. This activation leads to epithelial-mesenchymal transition (EMT) in cancer cells and polarization of tumor-associated macrophages (TAMs), ultimately promoting cancer metastasis [34]. While these findings suggest that succinate may contribute to an immune-suppressive microenvironment, a separate study revealed that upon lipopolysaccharide (LPS) stimulation, macrophages undergo metabolic shift characterized by elevated succinate levels, which drives mitochondrial oxidation, leading to heightened production of mitochondrial ROS and a pro-inflammatory gene expression profile [35]. Besides, extracellular succinate activates SUCNR1 and promotes CD8+ T cell cytotoxicity [36]. Succinate can also enhance tumor angiogenesis by upregulating VEGF expression through activation of extracellular regulated kinase (ERK) and STAT3 upon binding to SUCNR1 on vascular endothelial cells [37]. Considering the divergent roles of SUCNR1, it is prudent to evaluate its suitability as a therapeutic target.

It is worth mentioning that succinate may interplay with other metabolites beyond α-KG structural analogs. Exercise-induced low pH leads to protonation of succinate, allowing for pH-gated release via monocarboxylate transporter 1 (MCT1), which also facilitates lactate transport. This highlights the interplay between these two metabolites [38]. The acidic microenvironment present in the TME may allow this type of interaction.

2.1.3 Fumarate

Accumulation of fumarate in cancer cells usually results from loss-of-function mutations in fumarate hydratase (FH), which occurs in various types of cancer, including hereditary leiomyomatosis renal cell cancer (HLRCC), adrenocortical carcinoma and glioma [39]. Alike 2-HG and succinate, fumarate's structural similarity to α-KG allows it to inhibit α-KG-dependent enzymes, leading to cellular effects like EMT and DNA damage repair. Fumarate has been reported to impede TET-mediated demethylation of miR-200ba429, a regulatory region of antimetastatic miRNA cluster, promoting migratory capability and EMT [40]. Interestingly, fumarate demonstrates variable effects on DNA repair mechanisms. Excessive fumarate impairs KDM4B, leading to abnormal hypermethylation at loci surrounding DNA breaks, compromising homology-dependent repair [41]. Conversely, locally generated fumarate inhibits KDM2B and promotes non-homologous end joining (NHEJ) DNA repair [42]. Further exploration is warranted to determine whether targeting fumarate can be leveraged as a therapeutic strategy to interfere with DNA repair.

As an endogenous electrophile, fumarate is able to react with sulfhydryl groups to form S-(2-succino)-cysteine (2SC), termed as “succination” (Fig. 4). High levels of protein succination have been observed in tumors with loss-of-function mutation in FH gene [43]. The first reported succinated protein is Kelch-like ECH-associated protein 1 (KEAP1). Succination of KEAP1 disrupts its ability to repress the nuclear factor erythroid 2-related factor 2 (NRF2), consequently enhancing antioxidant capabilities and providing a growth advantage for cancer cells [44, 45]. Recently, it has been reported that fumarate-mediated succination of phosphatase and tensin homolog (PTEN) at cysteine 211 impairs its interaction with cellular membranes, leading to the activated PI3K/AKT signaling and sunitinib resistance in FH-deficient type 2 papillary renal cell carcinoma (PRCC2). Combination with an Akt inhibitor enhances the therapeutic efficacy of sunitinib in FH-deficient PRCC2 [46].

Fig. 4
figure 4

Functions of fumarate in the TME. Fumarate, structurally similar to α-KG, inhibits α-KG-dependent enzymes similarly to 2-HG and succinate. Its impact on DNA repair mechanisms is variable: inhibition of histone demethylase 4B (KDM4B) by fumarate impairs homology-dependent repair (HDR), while inhibition of histone demethylase 2B (KDM2B) promotes non-homologous end joining (NHEJ). Fumarate-mediated succination of proteins like phosphatase and tensin homolog (PTEN) and Kelch-like ECH-associated protein 1 (KEAP1) disrupts their interactions with other proteins, affecting cancer antioxidant responses and drug resistance. Fumarate uptake by CD8+ T cells results in succination of T cell receptor-associated protein kinase 70 (ZAP70), impairing T cell receptor (TCR) signaling and compromising T cell anticancer capabilities. HDR, homology-dependent repair; NHEJ, non-homologous end joining; KDM4B, histone demethylase 4B; KDM2B, histone demethylase 2B; PTEN, phosphatase and tensin homolog; TKI, tyrosine kinase inhibitor; NRF2, nuclear factor erythroid 2-related factor 2; ABL1, ABL proto-oncogene 1, non-receptor tyrosine kinase; ME2, malic enzyme 2; ZAP70, T cell receptor-associated protein kinase 70; TCR, T cell receptor

Full size image

Succination-independent mechanisms are increasingly recognized as well. Fumarate binds to ME2 to facilitate the formation of ME2 dimers, which leads to the activation ME2 and increases thymidine generation and mtDNA level to support the growth and progression of acute myeloid leukemia [47]. In FH-deficient HLRCC cells, excessive fumarate can indirectly activate tyrosine-protein kinase ABL1 to cope with impaired mitochondrial function and abnormal accumulation of intracellular fumarate [48]. The accumulation of fumarate can increase the uptake of cysteine and GSH biosynthesis to induce redox-dependent senescence, suggesting that fumarate-induced senescence needs to be overcome for the initiation of renal cancer [49]. Interestingly, accumulated fumarate in cancer cells also causes damage to the mitochondria, which in turn remodel their membranes to release mitochondria-derived vesicles (MDVs), thereby inducing the activation of cGAS and RIG-1 and triggers a chronic inflammatory response in renal cancer [50].

In addition to accumulating within tumor cells, excess fumarate produced by cancer cells can be secreted and play a functional role in the TME. Studies have shown that fumarate produced by FH-deficient tumor cells succinylates the zeta chain of T cell receptor-associated protein kinase 70 (ZAP70), thereby attenuating T cell receptor signaling and inhibiting CD8+ T cell activation [51]. The exploration of role of fumarate in TME remains limited and requires further investigation.

2.2 Lactate

Cancer cells exhibit a preference for aerobic glycolysis, known as the Warburg effect, to rapidly generate ATP, meet biosynthetic demands, and maintain redox balance, resulting in the abundant production of lactate [52]. This metabolic reprogramming is largely driven by oncogenic signaling. Oncogenes such as KRAS and FGFR enhance glycolytic flux by upregulating key glycolytic enzymes like LDHA and HK2 [53]. Additionally, transcription factors including HIF-1α, MYC, ATF4, and sine oculis homeobox 1 (SIX1), which are pivotal effectors in oncogenic signaling pathways, promote the expression of glycolysis-related genes [54,55,56].

For a long time, lactate was considered merely a byproduct of aerobic glycolysis. However, recent insights have revealed its underestimated significance. Lactate now emerges as a crucial molecule in energy production, oncogenic signaling maintenance, and epigenetic modifications in the TME (Fig. 5) [57, 58]. Lactate is one of the key energy sources for cancer cells. Lactate serves as a major energy source for the TCA cycle even in the presence of glucose [59]. Indeed, there exists a symbiotic relationship within tumors where lactate is exported from hypoxic cells to fuel nearby oxygenated cells, facilitated by MCT1 and MCT4 respectively [60]. Cancer cells also reprogramme CAFs to increase glycolysis and MCT4 expression, promoting lactate influx into cancer cells to fuel lipid metabolism for tumor invasion [61]. Additionally, lactate can be utilized by various other cells within the TME. Lactate exported from cancer cells can also be imported into regulatory T cells (Tregs), fueling their TCA cycle, supporting Tregs proliferation and immunosuppression [62]. Interestingly, although high lactate concentrations in the TME are typically thought to inhibit T cells, lactate at a concentration of around 3mM can efficiently enhance the effector function of CD8+ T cells under specific physical conditions [63]. These researches suggest that lactate functions as a universal fuel in the TME.

Fig. 5
figure 5

Functions of lactate in the TME. Lactate could function as an energy source, a signaling molecule, and an epigenetic modulator. It is transported between cells via monocarboxylate transporters MCT1 and MCT4. Lactate also acts as an autocrine or paracrine signaling effector by binding to its receptor, GPR81. This multifunctional role of lactate contributes to the complex interplay within the TME, influencing both cancer progression and the immune response. CAF, cancer-associated fibroblast; Kla, histone lactylation; Kac, histone acetylation; TCA, tricarboxylic acid; PKA, protein kinase A; MCT, monocarboxylate transporter; NFAT, nuclear factor of activated T-cells; LDH, lactate dehydrogenase

Full size image

Lactate serves as a signaling molecule through binding to GPR81, which is highly expressed in several types of cancer cells [64]. Activation of GPR81 facilitates the expression of genes responsible for lactate uptake and metabolism, thereby promoting lactate utilization under glucose deprivation. This process is critical for cancer growth and metastasis [64]. Additionally, lactate activation of GPR81 inhibits cAMP production, consequently suppressing protein kinase A (PKA) activity. This leads to TAZ-dependent PD-L1 upregulation, promoting immune evasion [65]. Activation of GPR81 promotes the expression of the pro-angiogenic mediator amphiregulin (AREG) through the PI3K-Akt-CREB axis in cancer cells. Overexpressed AREG is secreted into the TME, boosting angiogenesis and facilitating cancer growth [66]. Additionally, cancer cell-secreted lactate stimulates GPR81 in immune cells within the TME, contributing to an immune-suppressive environment. Activation of GPR81 in dendritic cells inhibits MHCII cell surface presentation and leads to the secretion of IL-6 and IL-12, further suppressing the anti-cancer immune response [67]. Lactate enhances the myeloid-derived suppressor cells (MDSCs) through the GPR81/mTOR/HIF-1α/STAT3 pathway, contributing to the radiation-induced immunosuppression [64]. Furthermore, lactate can function as a signaling molecule independent of GPR81. It can induce angiogenesis through the ROS- and NF-κB/IL-8-dependent pathway and by binding to N-Myc downstream-regulated protein 3, which in turn activates NADPH oxidase and inhibiting PHD [68]. It has also been shown that lactate induces the rapid depletion of endoplasmic reticulum Mg2+ stores, thereby enhancing the uptake of mitochondrial Mg2+ by Mrs2 This process may contribute to metabolism remodeling in tumor [69].

Lactate plays a significant role in inducing epigenetic remodeling through mechanisms such as lactylation and indirect effects on acetylation. The discovery of histone lactylation (Kla) has shed light on lactate's involvement in epigenetic regulation. This process induces the transcription of M2-like genes in M1 macrophages, potentially compromising their anti-tumor capabilities[70]. Moreover, ocular melanoma cells with elevated Kla levels exhibit increased transcription of YTHDF2, resulting in the degradation of tumor suppressor genes like PER1 and TP53, thereby promoting cancer progression [71]. Additionally, lactate influences other histone modifications by increasing acetyl-CoA levels through its contribution to the TCA cycle. This provides substrates for histone acetyltransferases, leading to enhanced acetylation at histone H3K9 and histone H3K27 [72]. Furthermore, lactate inhibits the activity of histone deacetylases, resulting in increased acetylation at the H3K27 of the Tcf7 super enhancer, which promotes stemness in CD8+ T cells and strengthens the anti-tumor immune response [73].

It is important to note that lactate accumulation is often accompanied by acidification, making it challenging to discern whether observed phenotypes are solely due to lactate or pH changes. Indeed, some effects of lactate have been attributed to its acidifying effect. For instance, the downregulation of IFN-γ in T cells and NK cells following lactate treatment has been associated with intracellular acidification-mediated NFAT translocation [74]. Similarly, lactate-induced mitochondrial dysfunction and apoptosis of liver-resident NK cells in colorectal liver metastasis can be replicated by lowering the intracellular pH of NK cells [75]. This may explain the contradictory impact of lactate on the immune system observed in various studies, and could be further investigated by using sodium lactate to eliminates the influence of acidification.

2.3 Amino acids

2.3.1 Kynurenine

The production of kynurenine (KYN) is limited by a specific step catalyzed by enzymes indoleamine 2,3-dioxygenase (IDO) and tryptophan 2,3-dioxygenase (TDO). This step generates N-formylkynurenine that is further converted into KYN. TDO is primarily expressed in the liver and its expression can be induced by tryptophan and hormones, while IDO (especially IDO1) is more widespread and can be activated by various stimuli, such as inflammatory signals [76]. KYN acts as a ligand for the aryl hydrocarbon receptor (AHR), and its binding to AHR leads to the translocation of AHR into the nucleus and increases the expression of genes related to stress response and cell fate determination [77]. For example, activation of AHR upregulates aquaporin-4 (AQP4) to increases cell mobility and glioma cell invasion [78], activates the mitogen-activated protein kinases (MAPK) pathway to promote the metastasis of gastric cancer [79], and activates PI3K-Akt signaling to promote nuclear translocation of β-catenin and colorectal cancer progression [80]. Of note, AHR activation is noted to induce the expression of the tumor metastasis suppressor gene KISS1 to inhibit neuroblastoma metastasis [81]. These opposing functions may suggest cell-context-dependent roles of KYN-AHR signaling. To date, how KYN-AHR pathway dysregulated in cancer is less understood. In colon cancer, MYC enhances the expression of tryptophan transporters as well as arylformamidase to boost KYN production and AHR activation [82]. Deficiency in adenomatous polyposis coli (APC) leads to TCF4/β-catenin-mediated upregulation of TDO2 to activate the KYN-AHR pathway [83].

KYN can be exported from cancer cells or IDO1-expressing non-cancer cells through SLC7A11, which leads to ROS scavenging and NRF2 pathway activation to protect the cancer cells from ferroptosis. KYN also competes with cysteine on SLC7A11, causing pseudo-starvation and activating the GCN2-ATF4 pathway to upregulate SLC7A11 expression, forming a positive feedback loop [84]. Transporter for KYN in TME appears to be cell-type dependent. KYN released from cancer-repopulating cells, upon IFN-γ stimulation, is transported through SLC7A8 and PAT4 into nearby CD8+ T cells to activate AHR and enhance PDCD1 expression [85].

Extracellular KYN could affect a wide range of cell types. KYN from AML binds to serotonin receptor 1B on osteoblast, reshaping bone marrow niche with proinflammatory signature. As a consequence, acute-phase protein serum amyloid A1 exported from osteoblast into AML, activating AHR and promoting IDO1 expression, further facilitating production of KYN [86]. KYN can stimulate cardiomyocyte proliferation and cardiac angiogenesis in neonatal heart, suggesting its ability to boost angiogenesis in the TME by acting on vascular epithelial cells [87]. TAMs can uptake KYN in TME as well, where KYN activates AHR and stimulates the expression of CCR2, KLF4 and ectonucleotidase CD39, resulting in TAM recruitment and CD8+ T cell dysfunction [88]. Nevertheless, it has been found that the immunosuppressive effect of KYN is weak in vivo, probably due to the limited concentration of KYN in some human cancers. It has also been suggested that T cell suppression by KYN can be mimiced by fatty acid depletion, which indicates culture media lack of lipids may yield misleading conclusions on KYN immunomodulatory functions [89].

2.3.2 Arginine

Arginine is deemed a conditional essential amino acid, as normal cells possess intrinsic synthesis pathways. However, cancer cells frequently exhibit the impaired biosynthesis of arginine, also known as arginine auxotrophy, resulting in a dependency on extrinsic arginine. This defect often stems from epigenetically silenced expression of arginosuccinate synthetase 1 (ASS1), a rate-limiting enzyme in arginine biosynthesis [90]. Our recent work also revealed that oncogenic activation of KRAS transcriptionally suppresses ASS1 to divert the utilization of aspartate to pyrimidine synthesis to meet the high demand for DNA replication, thereby creating a dependency on extracellular arginine [91].

Arginine serves multiple roles as a versatile metabolite, including polyamine biosynthesis, nitric oxide (NO) production, and cellular energy production. Deprivation of arginine significantly decreases the TCA cycle, urea cycle, and glycolysis. Arginine-derived ornithine fuels polyamine synthesis, which in turn mediates nucleotide synthesis and DNA structure. Chemotherapy disrupts the arginine-polyamine balance by impeding ornithine decarboxylase, sensitizing triple-negative breast cancer to cytotoxic chemotherapy [92]. Arginine provides NO production catalyzed by nitric oxide synthases (NOS), which serves as a critical metabolite in tumorigenesis and cancer progression, influencing processes such as proliferation, angiogenesis and EMT [93].

Arginine is one of the three amino acids directly activating mammalian target of rapamycin complex 1 (mTORC1), which is critical in cancer cell proliferation and survival. Lately, it has been shown that activation of mTOR by arginine enhances the production of acetyl-CoA, leading to global upregulation of oxidative phosphorylation (OXPHOS) genes in prostate cancer cells. In parallel, arginine promotes nucleus retention of transcriptional enhanced associate domain 4 (TEAD4) and its recruitment to the promoter/enhancer regions of OXPHOS genes in a mTOR-dependent manner [94]. In addition to the mTOR pathway, arginine can also specifically bind to RBM39, an RNA binding protein responsible for pre-mRNA splicing and transcription regulation of metabolic genes involved in glucose, pyruvate, fatty acid, and various metabolic pathways. The arginine-dependent effect of RBM39 is crucial for cancer progression in hepatocellular carcinoma and is associated with patient survival [95].

Although arginine in the TME primarily originates from the diet, it is largely depleted by cancer cells, thereby affecting the TME. Arginine depletion could affect multiple types of immune cells, such as macrophages, dendritic cells, and B cells in the TME, with T cells being particularly sensitive to arginine deficiency. Reduced arginine concentrations downregulate the CD3ζ subunit of the T cell receptor and decrease effector cytokine production like IL-2, impairing T cell function. Cancer-driven arginine deprivation in the TME may suppress T cell cytotoxic ability [96].

3 Oncometabolites generated by non-cancer cells

Cancers rely on the collaboration of various non-cancer cells in their microenvironment to thrive. This collaboration involves a bidirectional exchange of oncometabolites to maintain a delicate balance between cancer and non-cancer cells. Oncometabolites originating from non-cancer cells also significantly contribute to cancer progression.

3.1 Immune cell-derived itaconate

Itaconate is produced from the TCA cycle intermediate cis-aconitate under inflammatory stimuli, primarily in myeloid cells, through the induced immune-responsive gene 1 (IRG1) [97]. Itaconate has long been recognized as an antibacterial metabolite. Its accumulation in bacteria-containing vacuoles disrupts bacterial metabolism, effectively hindering the proliferation of bacteria [98]. Recently, it has been reported that IRG1 expression is associated with cancer progression, such as glioblastoma and ovarian carcinoma [99], yet the current suggest divergent conclusions itaconate in cancer. Cancer-induced IRG1 expression leads to elevated levels of itaconate in peritoneal tissue-resident macrophages. Itaconate is one of the most highly induced metabolites in activated macrophages. Itaconate production in macrophages upregulates OXPHOS and increases ROS production, which activates the MAPK pathway and promotes proliferation of cancer cells [99]. Additionally, itaconate shares common mechanism with other oncometabolites derived from TCA. Structurally similar to α-KG, itaconate competes with α-KG to bind TET2, impeding its activity and suppressing LPS-induced gene expression in macrophages [100]. Itaconate inhibits TAM polarization toward proinflammatory phenotypes by reducing the expression of chemokine genes such as Cxcl9 and Cxcl10 in a TET2-dependent manner. Inhibiting itaconate production significantly enhances anticancer immunity [101]. Itaconate also competitively inhibits SDH, likely due to structural similarity with succinate, in turn reducing mitochondrial respiration and exerting anti-inflammatory effects during macrophage activation [102]. CD8+ T cells within the TME could uptake itaconate secreted by MDSCs, which disrupts the biosynthesis of aspartate and serine/glycine in CD8+ T cells, leading to suppressed proliferation and diminished anti-cancer capabilities [103].

With its electrophilic α, β-unsaturated carboxylic acid, itaconate can directly modify proteins by alkylating cysteine residues, including KEAP1, transcription factor EB (TFEB) and NLR family pyrin domain containing 3 (NLRP3). Itaconate-mediated alkylation of KEAP1 activates Nrf2, increasing the expression of downstream genes with antioxidant and anti-inflammatory properties in macrophages [104]. Itaconate alkylates TFEB to promote nuclear localization and upregulating lysosomal and autophagic genes in macrophages, enhancing antibacterial innate immune defense [105]. Itaconate modifies a specific cysteine (C548) on NLRP3, inhibiting NLRP3 activation and interfering with its interaction with NEK7 [106]. Of note, the derivatives of itaconate such as dimethyl itaconate (DI), 4-octyl itaconate (4OI) and 4-menoethyl itaconate (4EI) exhibit distinct metabolic, electrophilic, and immunologic profiles compared with unmodified itaconate [107]. Although these findings were obtained in non-cancer models, they may offer useful insights for understanding the roles of itaconate in cancer.

3.2 Cancer-associated fibroblast-derived amino acids

3.2.1 Glutamine

Glutamine is known to be critical for cancer anabolism and plays a pivotal role in cancer growth and progression. Glutamine serves as a nitrogen source for amino acid and nucleotide biosynthesis, as well as a carbon source for TCA cycle metabolites and lipid biosynthesis. It acts as a precursor for multiple amino acids, including alanine, serine, and aspartate, via aminotransferase, and can be converted into asparagine by asparagine synthetase (ASNS) [108]. Inhibition of glutamine metabolism impairs asparagine production, thereby blocking the proliferation and metastasis of pancreatic ductal adenocarcinoma [109]. Glutamine can also be metabolized into glutamate by glutaminase (GLS), a process regulated by oncogene Myc and mTORC1 [110]. Glutamate fuels the TCA cycle with α-KG production. Impaired glutaminolysis decreases α-KG production, resulting in increased methylation at H3K27, and altered chromatin accessibility, thereby inhibiting progression of diffuse intrinsic pontine gliomas [111]. In cancer cells, glutamine metabolism also plays a pivotal role in maintaining redox balance. It supports the acquisition of environmental cystine via SLC7A11, influencing the biosynthesis of the antioxidant GSH. Active glutamine uptake, facilitated by upregulated glutamine transporter SLC1A1, drives cystine uptake and GSH synthesis in lung cancer, defending against oxidative stress and promoting malignant phenotypes [112]. Additionally, glutamine supports NADPH generation via ME1 through the glutamine-derived malate shuttle, crucial for cancer redox balance, especially under glucose deprivation conditions [113].

CAFs are pivotal producers of glutamine in the TME. The glutamine anabolic pathway in CAFs is enhanced through increased nitrogen supplementation from branched-chain amino acids and aspartate. Disrupting glutamine biosynthesis in CAFs significantly inhibits ovarian cancer growth [114]. Additionally, CAFs can convert collagen into glutamine, which is subsequently released and taken up by cancer cells. This process serves as a crucial fuel source for cancer cells, particularly under nutrient stress conditions [115].

Glutamine plays a crucial role in the metabolism of immune cells. Cancer cells and type-1 conventional dendritic cells (cDC1s) compete for glutamine uptake via the transporter SLC38A2. Intratumoral glutamine supplementation has been shown to inhibit cancer growth by enhancing cDC1-mediated CD8+ T cell immunity [116]. Additionally, adequate glutamine uptake is essential for the activation and maintenance of T cell function [117]. Therefore, strategies targeting glutamine metabolism should be developed with consideration for avoiding interference with immune cell metabolism, necessitating a more specific approach.

3.2.2 Alanine

Alanine plays a crucial role in maintaining amino acid homeostasis by acting as a major amino-nitrogen donor to transaminase networks. To data, our understanding of alanine's roles in cancer remains very limited. PDAC is known to exhibit high demand for alanine. PDAC cells overexpress the neutral amino acid transporter SLC38A1 to facilitate active alanine uptake from the TME. Inhibition of SLC38A2 leads to compartmentalized metabolic and redox crises, inhibiting cancer initiation and progression of PDAC [118]. Pancreatic CAFs could secrete alanine that competes with glucose and glutamine as a carbon source in PDAC cells, fueling the TCA cycle and facilitating the synthesis of non-essential amino acids and lipids. This metabolic shift reduces the PDAC’s reliance on glucose and serum-derived nutrients that are scarce in the pancreatic TME [119].

3.3 Microbial oncometabolites

The commensal microbiota, particularly that within the gut, is subject to modulation by a range of external factors, including dietary nutrients and medical treatments. These perturbations may disrupt the delicate equilibrium of microbial composition and functions, leading to a state of microbial dysbiosis [120]. Such pathogenic process is widely associated with cancer progression and immunotherapy [121,122,123]. However, it is only recently that the impact of microbial oncometabolites on these processes has been understood. Genotoxins may be one of the earliest inspirations of microbial oncometabolites. Colibactin, produced by organisms with the pks genomic island, including gut commensal Escherichia coli strains, can alkylate DNA and result in DNA damage, cell cycle arrest, and senescence in host cells, thereby promoting colon cancer growth [124, 125]. Moreover, indolamines produced by Morganella morganii, a gut bacterium commonly found in individuals with both inflammatory bowel disease and colorectal cancer, increase intestinal permeability and disrupt normal DNA replication, leading to the proliferation of intestinal epithelial cells [126]. In addition to genotoxins, microbial oncometabolites can interfere host cancer cells in various ways. Increased secretion of formate by F.nucleatum (Fn) in colon cancer activates AHR pathway in cancer cells, resulting in activation of MAPK and Wnt signaling to contribute to cancer stemness and invasion [127].

The impact of oncometabolites produced by microbiota is specific to organs and even regions, due to the unique localization of microbiota. Gallic acid derived by gut microbiota alone is sufficient to reverse the cancer suppressive function of mutant p53 by blocking its inhibition of WNT pathway-driven dysplasia in distal but not proximal gut, yielding the contrast effects of mutant p53 in distinct segments of gut [128]. Gut microbiome can also affect the biology of cancers in addition to CRC. Indole derived from tryptophan by lactobacillus activates AHR in TAM, impeding CD8+ T cell in PDAC microenvironment [129]. N, N, N-trimethyl-5-aminovaleric acid (TMAVA) derived from trimethyllysine and produced by gut microbiome, binds and inactivates the γ- butyrobetaine hydroxylase, causing a decrease in carnitine biosynthesis and promoting steatosis in the liver. This in turn, facilitates the development of liver cancer [130]. Secondary bile acids (SBAs), transformed from bile acids by gut gram-positive bacteria within the Clostridium cluster XIV, hinder the CCL6-mediated recruitment of NKT cells to the liver, which facilitates the progression of liver cancer [131]. Beyond gut microbiota, it has been shown that bacteria may present within certain types of cancer, such as pancreatic cancer, lung cancer, and breast cancer [132,133,134]. Intratumor bacteria have been found to be mostly intracellular, whose composition depends on cancer types [135]. Therefore, as an integral component in TME, the microbial metabolome is worthwhile to be explored.

4 Perspectives

Increasing number of oncometabolites produced from cancer cells exploit their cancer-promoting potential through reacting on non-cancer cells in TME, rather than simply accumulating in cancer cells. Oncometabolites mediate a dialogue instead of monologue in the context of TME. Oncometabolites not only serve as fuel for cancer cells but also carry out their functions through different mechanisms, such as exerting covalent modification, acting as signaling molecules, or interfering the function of metabolic enzymes or transporters (Fig. 6). Though current understanding about the roles of oncometabolites remain very limited, the study of oncometabolites could be interconnected. For example, the discovery of lactylation was inspired by other cellular metabolite-derived histone acylation [70]. Succinate and fumarate can yield succinylation and succination, respectively. It is not yet known whether excessive 2-HG and KYN in the TME will also result in certain modifications through covalent action. It is also important to be aware that oncometabolites do not function in isolation. For instance, R-2-HG has been found to competitively inhibit SDH, thus preferentially induced succinyl-CoA accumulation and hypersuccinylation in the mitochondria [33]. These interactions among oncometabolites in the TME suggest that they should be considered as an integral part of the system.

Fig. 6
figure 6

Mechanisms of oncometabolites in the TME. Oncometabolites orchestrate a complex intercellular communication, transcending their role as mere metabolic fuel. They engage in multifaceted mechanisms including covalent modifications of proteins, functioning as signaling entities, and modulating the activity of various metabolic enzymes and transporters. This multifunctionality enables oncometabolites to regulate a wide array of cellular processes, thereby influencing the TME dynamics and cancer progression. DC, dendric cell; CAF, cancer-associated fibroblast; KYN, kynurenine; AHR, aryl hydrocarbon receptor; TETs, ten-eleven translocation enzymes; PHDs, prolyl hydroxylases; KDM, lysine demethylases; SDH, succinate dehydrogenase; γ-BB, γ- butyrobetaine

Full size image

The field of oncometabolites is still expanding, with many metabolites yet to be discovered. One such category of metabolites that has yet to receive significant attention is lipids [136]. Although cancer cells often overexpress lipogenic enzymes such as fatty acid synthase (FASN) and ATP citrate lyase (ACLY) to meet their lipid needs, the effects of lipids on cancer growth are complex and contradictory [137]. Taking cholesterol as an example, high levels of cholesterol in T cells in TME can suppress the immune response and lead to T cell exhaustion [138]. However, cholesterol can also enhance T cell signaling and strengthen the immune response in some cases [139]. High cholesterol level has also been found to facilitate lipid raft formation and immune signaling in NK cells, boosting their anticancer capacity in liver cancer [140]. The contradictory effects of cholesterol are only the tip of iceberg of complex lipids regulation network. With the developing of cancer lipidome and relevant lipotypes, we may better understand how lipid modify cancer cells and TME.

The roles of oncometabolites can be variable according to cancer tissue origin, cancer grade, or different metastatic foci of the same cancer. For example, SDH and FH deficiency only trigger cancer in specific tissues, suggesting that susceptibility to oncometabolite is affected by the tissue of origin [141]. While early-stage PDAC shows increased BCAA uptake and catabolism, advanced PDAC becomes insensitive to BCAA inhibition [142]. Moreover, Moreover, primary breast cancer adapts glutamine metabolism for biomass production (anaplerosis), whereas breast-cancer-derived lung metastases exploit the pyruvate-rich lung environment and show pyruvate carboxylase-dependent anaplerosis, suggesting the organ-specific metabolic reliance [143]. It is important to understand how oncometabolite necessity is determined in different cancer context.

Given essential roles of oncometabolites, strategies for targeting these metabolites have gained increasing attention to enhance treatment efficacy. For oncometabolites generated by neomorphic enzyme activity, such as 2-HG, strategies to specifically disrupt their production is an effective approach. In fact, mutant IDH inhibitors to efficiently downregulate 2-HG production have demonstrated the therapeutic promise in patients [144]. Oncometabolite deprivation is another widely-applied strategy. For example, arginine is deprived utilizing recombinant arginine deiminase, such as pegylated arginine deiminase (ADI-PEG20), or bioengineered human arginases, both undergoing clinical trials [145]. Small-molecule inhibitors targeting glutamine and lactate metabolism are also evaluated under clinical trials alone or in combination with other targeted or radiotherapies across different cancer types [108, 146]. However, none of these approaches can differentiate the function of oncometabolite in different cell contexts, leaving uncertainty to the therapeutic benefit. To this end, targeting metabolite availability by inhibiting specific transporters could a more specific approach. For example, cancer cells and T cells utilize different transporters to import extracellular arginine, with cancer cells relying on transporters like SLC7A1, SLC6A14, and SLC7A3, while CD8+ T cells primarily use SLC7A1 and SLC7A2 [147], suggesting a possibility to disrupt arginine uptake of cancer cells without affecting that of CD8+ T cells. Furthermore, disrupting downstream signaling of oncometabolite could offer an alternative approach. In FH-deficient cancer, targeting AKT signaling using an AKT inhibitor could block oncogenic role of fumarate-mediated succination of PTEN [148].