Endothelial metabolic reprogramming influences tumor immune microenvironment

Abstract

The components of the tumor microenvironment are crucial in tumor growth, metastasis, immune evasion and therapeutic resistance. To adapt to the low-oxygen and nutrient-deficient conditions, cancer cells generate new blood vessels to promote tumor expansion and metastatic spread via tumor angiogenesis. Recent research has revealed that tumor endothelial cells reprogram their metabolic patterns during tumor progression. These metabolic changes influence the infiltration of cytotoxic T lymphocytes such as CD8+ T cells and recruit immune-suppressive cells, resulting in immune evasion and increased tumor progression. Therefore, targeting tumor endothelial metabolism alongside immunotherapies could offer a novel strategy for precise cancer treatment in clinical settings.

Key Words: Tumor microenvironment; Tumor endothelial cells; Metabolic reprogramming; Immunosuppression; Immunotherapy

Core Tip: In recent years, the tumor microenvironment (TME) has emerged as a critical target in cancer research due to its role in regulating tumor growth, metastasis and treatment response. Immune cells are core components of the TME, and the effects of tumor cell metabolic reprogramming on immune cells are well understood. However, research on the impact of metabolic reprogramming of tumor endothelial cells on the tumor immune microenvironment is limited. This review summarizes metabolic changes in endothelial cells within the TME and their effects on the tumor immune microenvironment. It also suggests combining anti-angiogenic therapy that targets endothelial metabolism with immunotherapy as a future treatment option.

INTRODUCTION

The constant and dynamic interactions between cancer cells and the tumor microenvironment (TME) are essential for tumor progression, metastasis, compromised immune surveillance and therapeutic response[1]. Within the TME, the main components are non-tumor cells, including fibroblasts, endothelial cells, pericytes, adipocytes, and immune cells, as well as non-cellular parts like the extracellular matrix and soluble molecules such as cytokines and growth factors[2]. The highly specific and complex TME can severely hinder the study of mechanisms and treatments for certain cancers. For instance, the efficacy of immunotherapy and anti-angiogenic therapy (AAT) in pancreatic neuroendocrine neoplasms remains limited due to their unique characteristics[3]. Therefore, analyzing the TME may provide crucial insights for identifying new targets in refractory tumors. Tumor endothelial cells (TECs) play a significant role in the TME and execute distinct functions that are stimulated by various inflammatory factors or angiogenesis-inducing cytokines, facilitating tumor growth and promoting distant metastasis[4]. Thus, the deadly tumor progression is heavily dependent on the tumor-vessel interactions. Under hypoxic conditions in the TME, quiescent endothelial cells become active, vascular endothelial growth factor (VEGF) expression is upregulated, and the angiogenic switch occurs when VEGF binds to the receptors on endothelial cells[5,6] (Figure 1). In addition, pericytes detachment and basement membrane degradation promote the migration of tip cells towards the tumor[7]. However, the new vessels from tumor angiogenesis are structurally and functionally abnormal. Twisted vascular structures and aberrant metabolic reprogramming of TECs could influence the infiltration and function of immune cells. For instance, TECs induce tumor-infiltrating T cells exhaustion by expressing glycoprotein nonmetastatic melanoma protein B[8]. Another research discovered that endothelial Toll-like receptor 2 promotes the recruitment of pro-tumorigenic immune cells, such as tumor-associated macrophages (TAMs) and tumor-associated neutrophils (TANs), by inducing the expression of pro-inflammatory cell adhesion molecules in endothelial cells[9].

Figure 1 Angiogenic processes in the tumor microenvironment. The hypoxic tumor microenvironment releases vascular endothelial growth factor (VEGF). When VEGF binds to VEGF receptors on quiescent endothelial cells, it activates the endothelial cells and initiates the angiogenic program. At the same time, pericytes detach and the basement membrane degrades. During angiogenesis, tip cell migrates toward the tumor and stalk cells proliferate to lengthen the vessels. Finally, structurally and functionally abnormal tumor vasculature is formed. VEGF-A: Vascular endothelial growth factor A; VEGFR2: Vascular endothelial growth factor receptor 2.

A defining feature of cancer is the alteration of metabolism[10]. In recent times, it is becoming clear that the components of the TME regulates its own metabolism, including TECs, to meet the high energy needs and fast growth of tumors. TECs survive and form new blood vessels depending on various metabolic pathways, including glycolysis, oxidation of fatty acids (FAs) and glutamine metabolism[11,12]. Despite extensive research into the metabolic alterations of cancer cells and individual cell types within the TME, the interactions between TECs and other TME components (such as immune cells) via metabolites are still unclear. Recently, research has reported that hypoxia induces endothelial metabolic reprogramming and DGKG-high expressing TECs promote angiogenesis and immune evasion by recruiting ubiquitin-specific peptidase 16 to facilitate zinc finger E-box binding homeobox 2 deubiquitination and then initiating the transforming growth factor-β1 positive feedback loop[13].

Immunotherapy has ushered in a new era of cancer therapy. It has offered long-lasting clinical responses for certain groups of cancer patients. It has been demonstrated that the TME significantly influences tumor growth and immunotherapy resistance[14-16]. Immunotherapy in combination with AAT is one of the classic cancer treatment strategies[17-19]. The metabolism of TECs is a key regulator of angiogenesis, and tumor vessel normalization has emerged as a novel therapeutic strategy[20,21]. Therefore, combining new agents that target the endothelial metabolism with immune modulators may be a promising therapeutic approach. This review highlights the metabolic reprogramming of TECs, the impact on the tumor immune microenvironment, and discusses a potential therapeutic possibility.

METABOLIC REPROGRAMMING IN TECS

Despite being in an oxygen-rich condition, energy production in resting endothelial cells is primarily dependent on anaerobic glycolysis. In addition, FA oxidation (FAO) in endothelial cells provides energy and a substrate for nucleotide synthesis[12,22]. In contrast to the normal state, the structural malformation and dysfunction of blood vessels in tumors, coupled with the vigorous proliferation of the cancer cells, lead to poor perfusion, hypoxic states and metabolic disorders in the TME[10]. TECs act as a metabolic interface between the blood flow and the tumor, and alterations of their metabolism will affect cancer cells and other components within the TME (Figure 2).

Figure 2 Metabolism of tumor endothelial cells. Tumor endothelial cell (TEC) exhibits higher levels of glycolysis than normal endothelial cell. 6-phosphofructo-2-kinase/fructose-2,6-bisphosph-atase 3, an activator of glycolysis, is overexpressed in TEC. Endothelial cells transport fatty acids into mitochondria via carnitine palmitoyltransferase 1A. In proliferating endothelial cells, carnitine palmitoyltransferase 1A is the rate-limiting enzyme for fatty acid oxidation and fatty acid oxidation provides a significant amount of substrate for deoxy nucleotide triphosphate synthesis. Enhanced glutamine metabolism in TEC provides amino acid substrates for nucleotide and biomass synthesis. Additionally, serine can be synthesized directly from 3-phosphoglycerate, which is produced by glycolysis, via serine synthesis pathway. Compared to normal blood vessels, tumor blood vessels have a disrupted barrier and distorted structure. They are more likely to develop hypoxia and metabolic abnormalities, which promote the dissemination of tumor cells. TEC: Tumor endothelial cell; PFKFB3: 6-phosphofructo-2-kinase/fructose-2,6-bisphosph-atase 3; HK2: Hexokinase 2; PPP: Pentose phosphate pathway; SSP: Serine synthesis pathway; PHGDH: Phosphoglycerate dehydrogenase; PSAT1: Phosphoserine aminotransferase 1; FASN: Fatty acid synthase; FAs: Fatty acids; FAO: Fatty acid oxidation; CPT1a: Carnitine palmitoyltransferase 1a; dNTP: Deoxy nucleotide triphosphate; TCA: Tricarboxylic acid; α-KG: Α-ketoglutarate; Glu: Glutamate; Gln: Glutamine; GLS1: Glutaminase 1; OXPHOS: Oxidative phosphorylation; ASNS: Asparagine synthetase.

Glycolysis

Compared to normal endothelial cells, TECs have higher levels of glycolysis. Some studies have found that TECs have higher expression levels of glycolytic genes, including 6-phosphofructo-2-kinase/fructose-2,6-bisphosph-atase 3 (PFKFB3), the glucose transporter glucose transporter 2 and rate-limiting enzymes of the glycolytic bypass pathway[23,24]. The glycolytic pathway is regulated by three enzymes: Phosphofructokinase-1 (PFK1), hexokinase 2 (HK2) and pyruvate kinase[25]. PFKFB3, a glycolytic activator, activates PFK1 and is overexpressed on TECs. Researchers have demonstrated that reducing or inhibiting PFKFB3 does not impact tumor growth, but instead hinders tumor angiogenesis, which in turn decreases cancer cell invasion, intravasation, and metastasis[24,26-28]. Therefore, targeting the PFKFB3-mediated glycolytic process of TECs could be an effective anti-angiogenic treatment approach. HK2 is another rate-limiting enzyme for glycolysis. Research into lung cancer has revealed the key role of the Foxp1-hypoxia-inducible factor 1-alpha-Hk2 pathway in regulating glycolysis in TECs, which in turn affects tumor angiogenesis[29]. Knockdown of Foxp1 results in attenuated inhibition of hypoxia-inducible factor 1-alpha-Hk2 pathway, which promotes glycolysis of TECs and neo-angiogenesis in lung tumor.

Research indicates that highly differentiated tumors exhibit lower vascularization compared to poorly differentiated tumors, and vascularization levels correlate with tumor prognosis and survival rates[3]. In highly vascularized tumors such as neuroendocrine tumors, VEGF levels in the microenvironment can reach up to 80%, thereby driving angiogenesis and tumorigenesis. The upregulation and sustained effects of angiogenesis inducers are crucial for boosting glycolysis in TECs. VEGF-A is a well-known molecule that encourages angiogenesis[10], and it increases the PFKFB3-driven glycolysis[30]. Fibroblast growth factor (FGF) is another element that can influence HK2 expression by controlling c-MYC expression[31]. The transcription factors forkhead box O (FOXO) are essential for cellular metabolism and tissue homeostasis[32]. FOXO1 maintains endothelial cells quiescence by inhibiting c-MYC signal[33,34]. Inhibition of FOXO1 expression in TECs is a driver of glycolysis and angiogenesis[35,36].

Another study in colorectal cancer (CRC) found that high glyceraldehyde-3-phosphate dehydrogenase expression in CRC patients was associated with enhanced glycolysis in TECs and the application of low-dose glyceraldehyde-3-phosphate dehydrogenase inhibitors normalized tumor vasculature and increased the infiltration of immune effector cells[37]. In addition to the high glycolytic metabolism, TECs have elevated transcript levels of the pentose phosphate pathway and the serine biosynthesis pathway related enzymes, representing an upregulation of the nucleotide synthesis pathway in TECs[12,24].

Fructose metabolism

Fructose is a common food additive in today’s society. Studies have shown that excessive fructose intake is strongly associated with obesity, metabolic syndromes and cancer[38]. Some other studies have found that reprogramming of fructose metabolism in the TME promotes the progression and metastasis of various cancers[39-41]. As components of the TME and important participants in tumor angiogenesis, TECs are also affected by fructose metabolism. Fang et al[42] and Cui et al[43] discovered that vascular endothelial cells can metabolize fructose, thereby activating the protein kinase B, Scr and AMP-activated protein kinase signaling pathways, which enhances their proliferation and migration. Additionally, fructose indirectly promotes angiogenesis by increasing VEGF expression in cancer cells.

Oxidative phosphorylation

In addition to glycolysis, activated endothelial cells increase their levels of oxidative phosphorylation (OXPHOS) to meet their energy needs. Hyperproliferating endothelial cells consume a lot of oxygen, and mitochondrial overuse makes them sensitive to uncoupling agents. When endothelial cells are treated with embelin, neoangiogenesis is effectively inhibited, which inhibits tumor growth[44]. Another genetic evidence similarly demonstrates the critical role of OXPHOS in the vascular endothelium for tumor angiogenesis and cancer progression. The researchers conducted experiments using mice with endothelium-specific deletion of cox10 and found that wound healing was slowed and tumor growth was impaired in mice under conditions of high energy consumption[45]. Elevated fructose metabolism in the TME can promote mitochondrial respiration and enhance endothelial cell migration by activating the AMP-activated protein kinase signaling pathway[42].

The above studies suggest that blocking the OXPHOS process in endothelial cells is a promising approach for treating cancer progression. However, not all mitochondrial respiration inhibitors target endothelial cells specifically. The β1 receptor inhibitor nebivolol, for example, can interfere with the OXPHOS process in cancer cells by inhibiting the activity of the OXPHOS complexes, yet it does not affect the mitochondrial respiratory function of TECs[46]. However, nebivolol interferes with endothelial cell proliferation because it blocks extracellular signal-regulated kinase activation and inhibits the glycolytic process of TECs[46].

FA metabolism

Endothelial cells metabolize FA to acetyl-CoA for maintenance of the tricarboxylic acid (TCA) cycle for energy. More importantly, FAO facilitates deoxy nucleotide triphosphate (dNTP) synthesis in proliferating endothelial cells[34,47,48]. Reduced FAO in endothelial cells doesn’t lead to energy depletion, but it impacts DNA synthesis. Studies of isotopically labelled endothelial cells have shown that the carbons in FA enter the Krebs cycle and become part of aspartate, uridine monophosphate and DNA[49]. Carnitine palmitoyltransferase 1A (CPT1A) is the enzyme that limits the rate of FAO and is responsible for transporting FAs to mitochondria for β-oxidation. It plays an important role in regulating the proliferation of endothelial cells. Pharmacological inhibition of CPT1 by using etomoxir as well as selective knockdown of the CPT1a gene in endothelial cells resulted in reduced endothelial cells proliferation, decreased dNTPs synthesis and impaired neo-angiogenesis[34].

FA synthase (FASN) is an important enzyme for synthesizing lipids endogenously. It catalyzes the production of palmitate, a long-chain FA, from acetyl-CoA and malonyl-CoA in the presence of NADPH[50,51]. Numerous past studies have shown that FASN is highly expressed in various tumors and is linked to a poor prognosis, including cervical[52], gastric[53], breast[54] and gallbladder cancer[55]. However, there are fewer studies of FASN in TECs. FASN has a promoting effect on tumor vessel sprouting[56]. Knocking down FASN in endothelial cells led to higher malonyl-CoA levels, resulting in mammalian target of rapamycin (mTOR) malonylation, impaired the activity of mTOR1 kinase and ultimately inhibited endothelial cells proliferation[56]. In addition, an interesting study in lymphatic endothelial cells (LECs) showed that FASN plays a role in supporting human LECs viability, proliferation and migration[57].

Amino acid metabolism

In the last ten years, numerous studies have emphasized the significance of glutamine metabolic reprogramming in cancer[58,59], establishing it as a potential target for treatment[60,61]. Nonetheless, the impact of endothelial glutamine metabolism reprogramming on tumor cells and TME have not been thoroughly investigated. Glutamine, which is the most prevalent non-essential amino acid in the somatic circulation, is absorbed by cells and catabolized by glutaminase (GLS) into glutamate and ammonia, with glutamate entering the TCA cycle as α-ketoglutarate[5]. One of the few studies focusing on endothelial cell metabolism found that endothelial cells have a high rate of glutamine metabolism[62]. Glutamine supplies carbon to the TCA cycle and nitrogen for the synthesis of pyrimidines and purines, and is metabolized to glutathione to main redox homeostasis[47,63,64]. Glutamine deprivation inhibits hypoxia- and hypoglycemia-induced up-regulation of VEGF-A expression, which may influence tumor angiogenesis[65]. Inhibiting GLS or selectively knocking down the GLS gene in endothelial cells led to reduced endothelial proliferation and migration[66,67], and even improved delivery of chemotherapeutic agents and therapeutic efficacy[67]. These findings revealed a novel feature of TECs glutamine metabolic reprogramming that promotes tumor angiogenesis, opening a new promising perspective for targeted tumor therapy.

Endothelial cells can also use glutaminolysis-derived glutamate to synthesize aspartate with the help of aspartate aminotransferase[68]. Through activation of the mTOR complex 1 pathway, aspartate regulates VEGF receptor 2 and FGF receptor 1 expression in endothelial cells and pyrimidine synthesis, thereby promoting tumor angiogenesis[68]. In addition to the above amino acids, proliferating endothelial cells also use serine to support intracellular biosynthesis. Serine can be taken up directly from extracellular sources or synthesized from 3-phosphoglycerate[5]. The phosphoglycerate dehydrogenase (PHGDH) is a key enzyme of the serine synthesis pathway and is essential for endothelial cell survival and proliferation. Knockdown of PHGDH in endothelial cells severely impairs endothelial cell survival and proliferation. This is due to mitochondrial dysfunction caused by insufficient heme synthesis[69]. Together, these findings emphasize the importance of endothelial metabolic reprogramming in solid tumor progression.

IMPACTS OF ENDOTHELIAL METABOLIC REPROGRAMMING ON TUMOR IMMUNE MICROENVIRONMENT

In tumors, endothelial cells show increased glycolysis, FA metabolism and glutamine metabolism to sustain their survival and proliferation. In addition to tumor cells, metabolic reprogramming of endothelial cells also help to create a metabolic environment conducive to immune evasion and tumor progression[70]. A study focusing on endothelial cell phenotypes in lung TME found that nucleotide metabolic pathways were significantly enhanced in TECs, and other pathways affected were OXPHOS and glycolysis[71]. However, genes related to immune activation and immune cell homing were expressed at lower levels in TECs[71]. DGKG is an enzyme associated with lipid metabolism, and hypoxia induces endothelial cell-specific high expression of DGKG, which would promote angiogenesis and immune evasion in hepatocellular carcinoma[13]. More importantly, tumor progression can be synergistically inhibited when targeting enzymes involved in glucolysis or lipid metabolism in TECs is combined with anti-programmed cell death-1 blockade[13,37]. Thus, TECs metabolism may differently affect anti-tumor immunity and contribute to immune evasion (Figure 3). The effects of TECs metabolic reprogramming on various immune cells will be discussed in detail below.

Figure 3 Endothelial metabolic reprogramming alters the tumor immune microenvironment. In the tumor microenvironment (TME), endothelial cells undergo metabolic reprogramming, including enhanced glycolysis, fatty acid oxidation and glutaminolysis. These alterations will not only affect the proliferation and migration of endothelial cells, but also have an impact on immune cells in the TME. The expression of programmed death-ligand 1 is increased on tumor endothelial cell (TEC), which in turn induces CD8+ T cell suppression. TECs induce the exhaustion of CD8+ T cells via glycoprotein nonmetastatic melanoma protein B. Moreover, they secrete C-X-C motif ligand 12 (CXCL12) to impede the differentiation of naïve CD8 T cells into cytotoxic T cells. TECs can also recruit regulatory T cells, tumor-associated macrophages, tumor-associated neutrophils and myeloid-derived suppressor cells to the TME by secreting factors including C-C motif ligand 2, CXCL1, CXCL2 and granulocyte colony-stimulating factor. This process contributes to the establishment of an immunosuppressive state. Furthermore, TECs secrete CXCL2 and CXCL4, which drive the polarization of macrophages towards the M2 phenotype. The accumulation of immunosuppressive cells within the TME leads to the further secretion of pro-angiogenic factors, thereby inducing abnormal angiogenesis. The structurally and functionally dysregulated blood vessels exacerbate the hypoxic condition of the TME, which in turn further promotes the metabolic reprogramming of TECs. MDSCs: Myeloid-derived suppressor cells; TME: Tumor microenvironment; TEC: Tumor endothelial cell; CCL2: C-C motif ligand 2; CXCL1: C-X-C motif ligand 1; G-CSF: Granulocyte colony-stimulating factor; VEGF: Vascular endothelial growth factor; JAK: Janus kinase; STAT: Signal transducer and activator of the transcription; NK: Natural killer; GPNMB: Glycoprotein non-metastatic melanoma protein B; PD-L1: Programmed death-ligand 1; Treg: Regulatory T; FAO: Fatty acid oxidation; ox-LDL: Oxidized low-density lipoprotein; LOX-1: Lectin-like oxidized low-density lipoprotein receptor-1; TAN: Tumor-associated neutrophil.

T cells

T cells not only interact with adjacent antigen-presenting cells but also with endothelial cells, which are regarded as the metabolic bridge between the blood and various tissues and organs in the body[4]. Researchers have identified an immunoregulatory endothelial cell subtype in triple-negative breast cancer (TNBC) that express high levels of tumor necrosis factor receptor 2 in chemoresistant TNBC patients[72]. Tumor necrosis factor receptor 2 activates the nuclear factor kappa B pathway to inhibit glycolysis and increases programmed death-ligand 1 (PD-L1) expression in endothelial cells, thereby inducing CD8+ T cells suppression[72]. In hepatocellular carcinoma, TECs induced CD8+ T cell depletion by upregulating the expression of the membrane glycoprotein nonmetastatic melanoma protein B[8]. This new target offers another potential treatment option for hepatocellular carcinoma. Lu et al[73] identified a novel TEC subpopulations that secretes C-X-C motif ligand 12 (CXCL12) and inhibits the differentiation of naïve CD8 T cells into cytotoxic CD8 T cells. Therefore, these TECs play an important role in shaping the immunosuppressive microenvironment.

It has been found that human umbilical vein endothelial cells (HUVECs) regulate the proliferation and activation of CD4+ T cells and regulatory T (Treg) cells[74]. Treg cells within the TME promote tumor development by inhibiting anti-tumor responses and interacting with tumor stromal cells such as fibroblasts and endothelial cells[75]. A high pro-angiogenic state is associated with the accumulation of tumor-infiltrating Treg cells in renal cell carcinoma, and the angiogenic network and immune tolerance cooperate with each other to ultimately promote tumor growth and progression[76]. In addition, in LECs, metabolic reprogramming affects immune cell stability and infiltration. Abnormal metabolism of LECs induced by a high cholesterol environment limits their ability to stabilize Treg cells[77]. High levels of cholesterol metabolites induced hyper-activation of the interleukin-18/nuclear factor kappa B/vascular cell adhesion molecule 1 axis in LECs, which inhibit Treg cells migration[77].

TAMs

TAMs are a primary type of immune cell found within tumors. They are usually polarized into two functionally contradictory subtypes, pro-inflammatory or immuno-suppressive[78,79]. To date, numerous studies have demonstrated that TAMs can promote tumor angiogenesis by secreting pro-angiogenic factors[78]. In pancreatic neuroendocrine neoplasms, the cellular density of TAMs is significantly correlated with invasive and malignant features, and these cells show markedly increased infiltration in liver metastases. This may be due to TAMs promoting the angiogenic switch[80,81]. Accordingly, angiogenic factors can shape the immuno-suppressive microenvironment by directly suppressing antigen-presenting cells and immune effector cells or by enhancing the effect of Treg cells, myeloid-derived suppressor cells (MDSCs) and TAMs[82]. Interactions between angiogenesis and TAMs polarization drive tumor initiation and progression. Some researchers have also found a close relationship between TECs and TAMs in tumorigenesis and progression. A senescence-associated secretory phenotype can be produced by senescent cells in the TME, and this can promote tumorigenesis. Researchers found that the most common types of senescent cells in KRAS-driven mouse lung cancer were macrophages and endothelial cells[83]. They also found that removing senescent macrophages reduced tumor load and improved survival. Although the potential mechanism of interaction between senescent endothelial cells and macrophages was not fully explored, this study provides ideas and rationale for future research. Further evidence of a strong link between TECs and TAMs in the TME is that reprogramming the tuberous sclerosis complex-mTOR complex 1 signaling pathway of TAM enhances its competitive advantage over endothelial progenitor cells, thereby inhibiting tumor neoangiogenesis[84].

A wide variety of metabolites in the TME are involved in shaping the TAMs phenotype[85,86], and the role of metabolic reprogramming of TECs in TAMs polarization has received extensive attention. In epithelial ovarian cancer, epithelial ovarian cancer cells induce endothelial cells to secrete CXCL2 via the recombination signal binding protein for immunoglobulin kappa J region-mediated classical Notch signaling pathway, which in turn promotes macrophage infiltration into the TME[87]. When CXCL2 binds to C-X-C motif chemokine receptor 2 on the surface of macrophages, it induces macrophages to upregulate the hyaluronan receptor CD44. Binding of CD44 to high-molecular-weight hyaluronic acid secreted by tumor cells induces cholesterol efflux in TAMs, making them more responsive to interleukin-4 signaling and thereby promoting their differentiation into immunosuppressive TAMs[87]. Additionally, higher levels of M2-type macrophage signature expression in bone metastasis samples from patients with bone-metastatic castration-resistant prostate cancer are associated with shorter overall survival[88]. Mechanistically, endothelial cells undergoing the endothelial cell-to-osteoblast transition secrete paracrine factors such as Wnts and CXCL14, which induce macrophage polarization toward the M2 type and aggregation toward the TME[88]. Research reveals that the process of mitochondrial transfer between adjacent cells in TME has a significant impact on tumor progression. In a xenograft animal model, transplanting endothelial mitochondria into melanoma cells promoted tumor growth and increased the number of M2-type macrophages via the nuclear factor erythroid 2-related factor 2/heme oxygenase-1 pathway[89].

TANs

TANs are also one of the immune cells that play a unique role in the TME. TANs not only stimulate tumor angiogenesis and epithelial-to-mesenchymal transition, but also specifically suppress the immune function of T cells[90]. Additionally, TANs interact with tumor cells. Under the stimulation of certain cytokines, such as transforming growth factor-β, they may transition from an antitumor (N1) phenotype to a pro-tumor (N2) phenotype[91,92]. The recruitment of these neutrophils requires close interaction with endothelial cells. Therefore, metabolic reprogramming of endothelial cells in the TME may affect the function of neutrophils in close contact with them. A multi-omics study focusing on the TME of intrahepatic cholangiocarcinoma found that intrahepatic cholangiocarcinoma patients with a high tumor-promoting propensity of TECs had an increased infiltration of neutrophils and a decreased number of natural killer cells and effector memory T cells in the TME[93]. More interestingly, TECs highly express lectin-like oxidized low-density lipoprotein receptor-1, which is recognized as a receptor for oxidized low-density lipoprotein, a key molecule in cardiovascular disease. The lectin-like oxidized low-density lipoprotein receptor-1/oxidized low-density lipoprotein axis induces TECs to secrete C-C motif ligand 2 to recruit neutrophils, thereby creating a microenvironment conducive to tumor metastasis[94].

MDSCs

MDSCs play a complicated role in regulating the tumor immune microenvironment. On the one hand, they can inhibit the killing effects of T lymphocytes and natural killer cells, and recruit Treg cells to further enhance the immunosuppressive effects; on the other hand, MDSCs can also induce tumor angiogenesis through the VEGF/Janus kinase/signal transducer and activator of the transcription signaling pathway[95]. Although a large number of previous studies have thoroughly explored the vital role of MDSCs in immunosuppression and angiogenesis, research on how metabolic alterations in TECs impact MDSCs status remains limited. A study by He et al[96] found that knocking out the Shb gene in mouse endothelial cells promoted the recruitment of MDSCs in breast cancer. The exact mechanism of this phenomenon is unclear, but elevated expression of C-C motif ligand 2, CXCL1, and granulocyte colony-stimulating factor was found in Shb-KO endothelial cell-derived CD45+ cells, which creates an excellent microenvironment for the aggregation of MDSCs.

In conclusion, these discoveries reveal a strong connection between TECs and tumor-infiltrating immune cells. TECs’ metabolic changes influence both the recruitment and the function of tumor-infiltrating lymphocytes and immune-regulatory cells, while the secretion of angiogenic factors by immuno-regulatory cells promotes abnormal neoangiogenesis of tumors, further aggravating the hypoxic state of TME.

FOCUSING ON THE METABOLISM OF TECS AS A NEW TREATMENT STRATEGY

By uncovering the features and processes of tumor angiogenesis, numerous AAT have been created, focusing on blocking blood vessel growth to deprive cancer cells of nutrients. Tumor angiogenesis is mainly induced by pro-angiogenic factors in the TME, and common classical pro-angiogenic factors include VEGF, FGF-2 and platelet derived growth factor[97]. Common drugs that selectively target VEGF are bevacizumab, aflibercept and ranibizumab[97]. However, some cancers that are initially effective often develop resistance during treatment, resulting in poor prognosis. Numerous studies have identified and summarized the mechanisms behind AAT resistance: Increased recruitment of bone marrow-derived cells[98], cancer cells achieve vasculogenic mimicry by upregulating Foxc2[99], microRNAs regulate VEGF expression[100], and long noncoding RNAs promote vessel co-option by inhibiting the process of autophagy degradation of alanine-serine cysteine-preferring transporter 2[101,102]. Additionally, tumor cells are also able to develop AAT resistance through metabolic symbiosis[103,104].

The emergence of these drug-resistant responses suggests that the paradigm of blocking angiogenesis in solid tumors must change. Therefore, in the following years, the concept of “tumor vascular normalization” was proposed. Rather than inhibiting neoangiogenesis, the concept was to promote the transformation of the tumor vasculature to a more mature phenotype, which would facilitate the delivery of therapeutic agents and the infiltration of immune effector cells[105-109]. Abnormal TECs metabolism drives angiogenesis and worsens hypoxia in the TME. Targeting TECs metabolism to interfere with pathological angiogenesis and thereby normalize the tumor blood vessels has become a new method[110-112] (Table 1).

Table 1 Specific inhibition of endothelial glycolysis, lipid metabolism, amino acid metabolism and oxidative phosphorylation as novel targets for antitumor therapy.

Metabolic type	Target	Agent	Conditions	Mechanism	Ref.
Glycolysis	PFKFB3	PFK15	Hepatocellular carcinoma	Suppressing glycolysis in TECs and tumor cells and inducing apoptosis. Inducing tumor vessel normalization	[28]
		Apatinib	Hepatocellular carcinoma	Suppressed glycolysis in ECs via blocking PI3K/AKT/PFKFB3 pathway	[113]
		3PO	Pathological neovascularization	Inhibiting glycolysis in ECs partially and reversibly. Enhancing the antiangiogenic efficacy of VEGF blockade	[114]
	PKM2	SAA	Melanoma and lung cancer	Reducing glycolysis in ECs and activating the PKM2-dependent β-catenin/claudin-5 signaling axis improves endothelial integrity	[115]
Fatty acid metabolism	FASN	Orlistat	Retinopathy of prematurity	Inhibiting pathologic ocular neovascularization by decreasing mTOR activity	[56]
		Orlistat	Melanoma	Reducing HDLEC cells proliferation and migration. Inhibiting the secretion of VEGF-C and promoted the secretion of VEGF-D by melanoma cells	[57]
	CPT1	Etomoxir	Lymphangiogenesis	Inhibiting injury-induced lymphangiogenesis	[119]
Glutamine metabolism	GLS	miR-367b-3p	Choroidal neovascularization	Inhibiting CEBPB leads to the inactivation of GLS1 transcription. Inhibiting ECs proliferation, migration and tubular structure formation while promoting apoptosis	[122]
	GS	MSO	Lung cancer	Inducing a shift from the M2 phenotype to the M1 phenotype in macrophages. Promoting tumor vessel normalization and inhibiting the metastasis of cancer cells	[124]
OXPHOS	Mitochondria	Embelin	Pathological neoangiogenesis in tumor and injury	Depleting the low respiratory reserve of proliferating ECs without deleterious effects on resting ECs	[44]

PFKFB3: 6-phosphofructo-2-kinase/fructose-2,6-bisphosph-atase 3; PFK15: Phosphofructokinase-15; TECs: Tumor endothelial cells; ECs: Endothelial cells; PI3K: Phosphoinositide 3-kinases; AKT: Protein kinase B; 3PO: 3-(3-pyridinyl)-1-(4-pyridinyl)-2-propen-1-one; VEGF: Vascular endothelial growth factor; PKM2: Pyruvate kinase M2; SAA: Salvianic acid A; FASN: Fatty acid synthase; mTOR: Mammalian target of rapamycin; HDLEC: Human lymphatic endothelial cell; CPT1: Carnitine palmitoyltransferase 1; GLS: Glutaminase; CEBPB: CCAAT/enhancer-binding protein beta; GS: Glutamine-synthetase; OXPHOS: Oxidative phosphorylation.

Compared to normal endothelial cells, TECs upregulated the expression of the glycolytic activator PFKFB3 and exhibited stronger glycolysis. In recent years, it has been discovered that inhibiting PFKFB3-driven glycolysis, directly or indirectly, induces tumor vascular normalization and inhibits tumor progression. For example, PFK15 specifically inhibit PFKFB3 in TECs to normalize the tumor vasculature[28]. Apatinib, the drug of choice for the treatment of hepatocellular carcinoma, reduces glycolysis in TECs by blocking the phosphoinositide 3-kinases/protein kinase B/PFKFB3 pathway[113]. Schoors et al[114] found that 3-(3-pyridinyl)-1-(4-pyridinyl)-2-propen-1-one, a small molecule inhibitor of PFKFB3, transiently reduced glycolysis in vivo and inhibited endothelial cell proliferation and migration, but this was sufficient to reduce pathological neovascularization. Meanwhile, blocking PFKFB3 with 3-(3-pyridinyl)-1-(4-pyridinyl)-2-propen-1-one did not impact tumor growth, but it did decrease cancer cell invasion, intravasation, and metastasis by promoting vascular maturation and improving blood perfusion[24].

PFKFB3 is not the only target that regulates the metabolism of TECs. Under hypoxic conditions, salvianic acid A binds to the glycolytic enzyme pyruvate kinase M2 in HUVECs and significantly reduces glycolysis[115]. In a study of TNBC, electroacupuncture was discovered to inhibit glyoxalase in TECs to attenuate endothelial cell glycolysis, inhibit cell proliferation, and promote vascular normalization[116]. It has been found that transient receptor potential melastatin 7 channels might be a regulator of glycolytic reprogramming in endothelial cells, and inhibiting transient receptor potential melastatin 7-dependent glycolysis could offer a new therapeutic approach for treating cancer[117].

In previous discussions, we have known that FAO can provide carbon for dNTP synthesis, which is essential for endothelial cell proliferation[48,49]. Targeting enzymes involved in FA metabolism in endothelial cells, such as FASN and CPT1, appears to be an ideal therapeutic approach. FASN inhibition with cerulenin and orlistat reduced LECs viability, proliferation and migration[57]. And in vivo experiments showed that FASN inhibitors reduced lymphatic metastasis of melanoma[57]. In a study of metastasis and angiogenesis in B16-F10 melanoma, FASN inhibitor treatment was found to reduce endothelial cell (rat aortic endothelial cell and HUVEC) viability, proliferation and formation of abnormal vascular structures[118]. Some researchers, in order to test the therapeutic potential of FASN blockade in pathologic angiogenesis, used a mouse model of retinopathy of prematurity and found that low-dose orlistat treatment inhibited ocular pathologic angiogenesis in mice by decreasing mTOR activity[56]. In addition, blocking CPT1 with etomoxir inhibits injury-induced lymphangiogenesis[119]. Thus, FASN and CPT1 may be potential targets for tumor vascular normalization. Targeting these two enzymes may be a novel way to stop tumor progression and metastasis.

The enzyme GLS1 converts glutamine into glutamate and ammonia. In endothelial cells, this GLS1-driven metabolism encourages their growth, movement, and longevity[120]. When HUVECs were treated with 6-diazo-5-oxo-L-norleucine, a specific inhibitor of GLS, the proliferative capacity of the cells was significantly reduced and shifted toward a senescent phenotype[121]. In addition, another GLS1 inhibitor, bis-2(5-phenylacetamido-1,2,4-thiadiazol-2-yl)ethylsulfide, also normalizes HUVECs metabolism by modulating glutamine metabolism and reduces proliferation and migration[66]. MicroRNAs have also been found to block pathologic angiogenesis by inhibiting glutamine metabolism in studies of choroidal neovascularization-related therapeutic targets[122]. Asparagine synthetase (ASNS) synthesizes asparagine using aspartic acid and ammonia from glutamine[123]. Since ASNS silencing impairs endothelial cell proliferation, not only GLS1 but also ASNS may be a potential target for regulating endothelial metabolism[123]. Interestingly, pharmacological or genetic targeting of glutamine metabolism may indirectly interfere with tumor angiogenesis and inhibit cancer metastasis. Glutamine-synthetase (GS) is the enzyme responsible for catalyzing the conversion of glutamate into glutamine. Pharmacological inhibition of GS induces macrophage polarization from the M2 phenotype to the M1 type, which promotes endothelial cell branching and deletion of the macrophage GS gene promotes normalization of the tumor vasculature[124].

Numerous previous studies have confirmed the effectiveness of combining conventional anti-angiogenic agents with immune checkpoint blockades for cancer treatment[17-19,125,126] (Table 2). In addition to targeting angiogenic factors, targeting endothelial metabolism may be another useful perspective in the study of tumor vasculature normalization in combination with immunotherapy. For example, it has been found that aerobic glycolysis mediates the upregulation of cellular PD-L1 expression[127], whereas TECs express PD-L1[72,128], and targeting glycolysis may be able to reduce PD-L1 expression on both tumor cells and TECs, which would be beneficial for enhancing the anti-tumor effects of immune checkpoint blockade. Additionally, enhanced endothelial glycolysis mediated by PHGDH leads to excessive proliferation of TECs, which is one of the factors that drive glioblastoma resistance to chimaeric antigen receptor T therapy[129]. Targeting endothelial PHGDH inhibits abnormal vascular proliferation, improves the hypoxic microenvironment, and promotes immune cell infiltration into tumors. This discovery reveals the potential of targeting endothelial metabolic reprogramming in combination with chimaeric antigen receptor T therapy. Although AAT has become the standard treatment, overcoming resistance to it remains a challenge. It is believed that metabolic reprogramming of TECs plays a significant role in treatment resistance[130]. A study on metastatic CRC revealed that, under anti-VEGFA treatment, the TEC-specific F3 protein inhibits the degradation of the CPT1A protein via the mitogen-activated protein kinase/c-Jun NH2-terminal kinase-mitogen-activated protein kinase/extracellular signal-regulated kinase-TP53/p53 signaling axis, thereby enhancing the FAO of TECs[131]. Further studies confirm that targeting FAO effectively overcomes resistance to anti-VEGFA therapy. Therefore, in addition to immunotherapy, combining anti-angiogenic drugs with endothelial metabolism targeting can further enhance tumor treatment efficacy.

Table 2 Ongoing or completed clinical trials of anti-angiogenic therapy in combination with immunotherapy.

Inhibitor	Clinical stage	Study status	Conditions	Intervention	Ref.
Anlotinib	II	Recruiting	NSCLC	In combination with penpulimab	NCT06341530
	II	Recruiting	Breast cancer	In combination with sintilimab and chemotherapy	NCT04877821
Apatinib	II	Ongoing	Solid tumor	In combination with adebrelimab and microwave ablation	NCT06537505
Axitinib	II	Completed	Advanced melanoma	In combination with nivolumab	NCT04493203
Bevacizumab	I/II	Completed	Malignant solid tumor	In combination with NK immunotherapy	NCT02857920
	II	Completed	Colorectal cancer	In combination with atezolizumab and chemotherapy	NCT02873195
	I/II	Recruiting	Ovarian cancer	In combination with oregovomab and chemotherapy	NCT04938583
	II/III	Recruiting	Colorectal cancer liver metastases	In combination with serplulimab and chemotherapy	NCT06280495
	II	Recruiting	Gastric cancer	In combination with neoadjuvant chemotherapy and PD-1 mAb	NCT06371586
	II	Recruiting	Recurrent TNBC	In combination with pembrolizumab and chemotherapy	NCT06976944
	IV	Ongoing	Advanced melanoma	In combination with iparomlimab and tuvonralimab	NCT07004335
Ramucirumab	II	Recruiting	Gastric cancer	In combination with pembrolizumab and chemotherapy	NCT04069273
Sorafenib	II	Completed	Hepatocellular carcinoma	In combination with PD-1 mAb	NCT04518852

NSCLC: Non-small cell lung cancer; NK: Natural killer; PD-1: Programmed cell death-1; mAb: Monoclonal antibody; TNBC: Triple negative breast cancer.

Research exploring targets for endothelial metabolic reprogramming is gaining momentum. An increasing number of studies have demonstrated that targeting endothelial metabolism can promote tumor vasculature normalization, thereby enhancing the efficacy of both immunotherapy and conventional anti-angiogenic treatments. However, further clinical validation is required to determine its translational value in the field of cancer therapy.

CONCLUSION

Currently, studies targeting the non-cancerous cellular components of the TME have become increasingly intensive. In this context, some progress has been made in the development of tumor vascular normalization therapy focusing on TECs. However, conventional antiangiogenic therapy has drawbacks, and there is now a need to develop a new approach that is different from previous antiangiogenic therapies. With the increasing research on endothelial cell metabolism and its effects on normalizing tumor vasculature, the application of modulating TECs metabolism in cancer therapy is promising. Furthermore, metabolic reprogramming of TECs plays an important role in chemotherapeutic drug delivery and immune evasion. In the future, in addition to the use of monotherapies targeting the metabolism of TECs, combinations of chemotherapy and immunotherapy should be explored to achieve greater therapeutic potential. Previous studies have demonstrated that risk models based on genes associated with lipid metabolism may predict the prognosis of CRC and response to immunotherapy[132]. This also provides us with a new perspective: Evaluating differences in TECs metabolic characteristics between tumor samples obtained before and after treatment may offer novel insights for AAT or immunotherapy. In addition, correlating these metabolic characteristics with clinical endpoints, such as response rates and progression-free survival, could indicate that TECs metabolic reprogramming serves as a new prognostic indicator.

ACKNOWLEDGEMENTS

The authors thank Chen C and Liao WX for critically reviewing the manuscript.