It is well established that the proliferation and aggressiveness of tumor cells are sustained by the prevalence of aerobic glycolysis, also known as the Warburg effect, over OXPHOS (Cortes Ballen et al., 2024; Zhao et al., 2024). In agreement with this aspect, several independent studies on GBM patients highlighted a significant correlation between upregulation of glycolytic genes, low survival and poor prognosis (Alfardus et al., 2017; Stanke et al., 2021; Cortes Ballen et al., 2024; Zhao et al., 2024).

Hypoxia is a significant activator of the metabolic reprogramming and the Warburg effect. GBM adapts to a poor oxygenated environment by promoting the expression of different transcription factors and epigenetic modifications to survive in these hostile conditions (Miranda-Galvis and Teng, 2020). Particularly, low oxygen levels stimulate the translocation of hypoxia-inducible factor 1 subunit alpha (HIF-1α) from the cytoplasm to the nucleus, where it is stabilized. HIF-1α activation promotes hypoxia response elements (HREs) transcription, and the expression of genes involved in glycolysis, including glucose transporter 1 (GLUT1) and GLUT3, favouring glucose uptake and lactate production (Trejo-Solis et al., 2024). Pyruvate is converted into lactate by lactate dehydrogenase (LDH) through the transfer of a hydride ion from NADH to the carbon C2 of pyruvate. This reaction is regulated by HIF-1α gene that, activating LDH transcription, increases the production and accumulation of lactate in the TME, thus promoting cell proliferation (Figure 1) (Valvona et al., 2016; Domenech et al., 2021).

Consequently, lactate supports the acidic pH of TME, stimulating immune cells recruitment but suppressing the anti-cancer immunity of innate and adaptive immune cells, facilitating tumor invasion (Wang et al., 2020). Lactate inhibits the function and proliferation of cytotoxic T cells, natural killer (NK) cells and dendritic cells, reshaping the polarization of macrophages toward a tumor-supporting M2 phenotype and increasing regulatory T cell (Treg) activity, further enhancing immune evasion (Figure 1) (Calcinotto et al., 2012; El-Kenawi et al., 2019). Mesenchymal GBM secretes higher levels of lactate as compared to other subtypes, contributing to an immunosuppressive microenvironment. Therefore, targeting lactate secretion could lead to a better GBM prognosis by preventing the PMT (Wang et al., 2023). Moreover, the inhibition of GLUT1 expression prevents lactate secretion, acting on glycolysis pathway, and leading to a reduction in immunosuppressive TAMs (Li et al., 2024). Additionally, lactate induces epigenetic reprogramming of immune cells through histone lactylation, altering gene expression to support tumor growth (Zhang et al., 2019). Particularly, Wang et al. demonstrated that such an epigenetic modification is correlated to CD47 upregulation and the phagocytosis alteration in GBM cells. These data suggest that the epigenetic reprogramming related to lactate secretion could be an effective target to increase the effectiveness of immuno-therapies against GBM (Wang et al., 2024).

GBM shows alterations also in glutamine metabolism, contributing to its increased tumor growth (Fu et al., 2024). Indeed, glutamine deprivation is correlated to reduced cell growth rate, while high levels are characteristic of GBM mesenchymal profile, contributing to treatment resistance (Ekici et al., 2022).

Recently, it has been demonstrated that epidermal growth factor receptor (EGFR), through the upregulation of glutamate dehydrogenase 1 (GDH1), supports glutamine metabolism in GBM, promoting cell proliferation (Yang et al., 2020).

Acetate is an astrocyte-specific bioenergetic substrate, exploited as an alternative source of energy in the brain. Recently, it has been demonstrated that acetate is also used by GBM as energetic metabolite; indeed, GBM cells show an increased acetate production, also leading to reactive astrogliosis in the surrounding microenvironment (Kim et al., 2024). GBM cells retain the glial capacity to oxidize acetate; however, they are able to co-oxidate both acetate and glucose to sustain a higher proliferative rate (Mashimo et al., 2014). The upregulation of acyl-CoA synthetase short chain family member 2 (ACSS2), which converts acetate to acetyl-CoA, is typical in GBM and it is correlated to a worst prognosis in GBM patients (Comerford et al., 2014; Mashimo et al., 2014).

It has been shown that GBM subtypes are characterized by different metabolic features, associated with a specific clinical outcome. For instance, from a metabolic perspective, mesenchymal GBM tumors show typical susceptibilities that are not observed in the other GBM subtypes (Wang et al., 2021). Through a proteogenomic study, Wang et al. characterized different types of GBMs from brain samples, dividing them into three clusters: proneural-like, mesenchymal-like and classical-like. This study reported that the proneural-like cluster is characterized by a higher abundance of acetylated proteins involved in OXPHOS, while the mesenchymal-like is enriched for glycolysis (Wang et al., 2021).

Metabolic differences between mesenchymal and proneural subtypes have been found also in glioma stem cells (GSCs) (Marziali et al., 2016). Interestingly, it has been demonstrated that GSCs share similarities with the respective GBM cells. Marziali et al. showed that GSCs with a proneural-like phenotype are characterized by high levels of N-acetylaspartate and glutamine, both of which are associated with neural metabolism and mitochondrial function, suggesting an oxidative phenotype. On the contrary, GSCs with a mesenchymal-like phenotype show a downregulation of N-acetylaspartate and glutamine, suggesting a shift towards glycolytic metabolism, which is often associated with aggressive and rapidly proliferating cells (Marziali et al., 2016). In addition, other studies revealed that mesenchymal subtype shows a higher glycolytic activity as compared to proneural one, with an increased expression of aldehyde dehydrogenase 1 family member A3 (ALDH1A3), a key enzyme associated with different metabolic processes, including glucose metabolism (Figure 1) (Mao et al., 2013; Xu et al., 2024).

In conclusion, GBM subtypes exhibit distinct metabolic profiles that reflect their aggressiveness. Mesenchymal subtype shows higher glycolytic activity and a more aggressive phenotype, while proneural subtype relies more on OXPHOS, correlating with a less aggressive behaviour and a better clinical outcome (Figure 1).

Lipids are structural components of cell membranes, nevertheless they are also involved in cellular signalling and energy reserve. Due to their rapid proliferation, cancer cells modulate their lipid metabolism adapting to increased energy request (Shakya et al., 2021).

Guo et al. demonstrated that GBM enhances fatty acid (FA) synthesis to foster tumor growth and proliferation (Guo et al., 2009a). Increased glucose and glutamine consumption promotes lipid production through sterol regulatory element-binding protein (SREBP)/SREBP-cleavage activating protein (SCAP) pathway (Figure 2) (Kou et al., 2022).

In addition, SREBP-1 leads to an increased glutamine consumption and lipogenesis through the upregulation of ASCT2 expression, a glutamine transporter. In turn, releasing ammonia, glutamine stimulates SREBP-1 activation, establishing a loop that improves glutamine metabolism and lipid biosynthesis (Zhong et al., 2024).

Recent studies revealed that the oncogenic pathway epidermal growth factor receptor (EGFR)/Phosphatidylinositol 3-Kinase (PI3K)/Protein kinase B (AKT) activates SREBP-1, that is directly involved in FA synthesis; indeed, SREBP-1 inactivation, through a pharmacological inhibitor or shRNA, reduces GBM proliferation in vitro and in vivo (Guo et al., 2009b; Geng et al., 2016). Interestingly, SREBP shows a key role in FA and cholesterol metabolism also under hypoxia, thus its inhibition impairs lipid biosynthesis in hypoxic cancer cells (Figure 2). Moreover, Lewis et al. demonstrated that stearoyl-CoA desaturase (SCD) and fatty acid-binding protein 7 (FABP7), involved in mono-unsaturated FA (MUFA) biosynthesis and FA uptake and trafficking, respectively, are correlated with SREBP function (Lewis et al., 2015). SCD1 is also crucial for GSCs lipid metabolism, and MUFAs synthesis inhibition impairs their stem cell features, reducing tumor growth in a GBM mouse model (Pinkham et al., 2019). Furthermore, mitogen-activated protein kinase (MEK)/extracellular signal-regulated kinase (ERK) signalling pathway is involved in GSCs vulnerability; indeed, MEK/ERK inhibits AMP-activated protein kinase (AMPK) activity that, through a modification in lipid composition, shields GSCs from lipotoxicity (Eyme et al., 2023).

Compared to healthy brain tissues, GBM stores more FA as an energy reservoir and to foster the de novo lipid synthesis supporting cell proliferation (Taib et al., 2019; Shakya et al., 2021). It has been shown that elongation of very long chain fatty acids protein 2 (ELOVL2), which catalyses the first reaction of long-chain FA elongation cycle, is upregulated in GSCs and it promotes their survival (Gimple et al., 2019). This supports the hypothesis that tumor growth is strengthen by the production of long-chain polyunsaturated acids (LC-PUFAs) (Figure 2).

The central nervous system (CNS), with about the 20% of total body cholesterol, is considered the most cholesterol-rich organ. Interestingly, GBM cells reduce their own cholesterol synthesis, while they exploit astrocyte-derived cholesterol to support their proliferation (Deshmukh et al., 2021). The strong responsiveness that GBM exhibits towards liver X receptor (LXR) agonists, can be explained by its reliance on cholesterol uptake (Figure 2). In particular, LXR regulates cholesterol homeostasis in the CNS and, its BBB-permeable agonist LXR-623 promotes cell death and blocks tumor growth in a GBM mouse model (Courtney and Landreth, 2016; Villa et al., 2016).

Interestingly, through metabolomic and lipidomic analyses, Wang et al. demonstrated that lipid composition is deeply different among GBM subtypes. Tumors with a mesenchymal subtype show a high level of triacylglycerols and decreased phosphatidylcholines, while proneural subtypes show a large amount of very long chain fatty acid lipids (VLCFAs) and LC-PUFAs. Moreover, proneural tumors exhibit an overexpression of Acyl-CoA synthetase long chain family member 6 (ACSL6) and of docosahexaenoic acid (DHA) as compared to mesenchymal tumors (Wang et al., 2021).

Notably, a GBM metabolic reprogramming involving the de novo synthesis of fatty acids promotes PMT. In particular, malic enzyme 2 (ME2), which catalyzes pyruvate formation from malic acid, upregulates mesenchymal markers, downregulating proneural genes. ME2 reprograms lipogenesis through AMPK/SREBP-1/ACSS2 signaling, by reducing mitochondrial ROS production and AMPK phosphorylation, leading to SREBP-1 maturation and ACSS2 lipogenesis pathway promotion (Yang et al., 2021; Wu et al., 2024).

Therefore, alterations in the de novo lipid synthesis seem to be a potential susceptibility target for GBM; however further studies are required to better understand the weak knot in lipid metabolic reprogramming that could be an effective therapeutic target.

Nucleotides are the principal structural components of nucleic acids, but they are also implicated in energy metabolism and cellular signalling. Their synthesis occurs through two main pathways: de novo nucleotide synthesis and nucleotide salvage (Zhou and Wahl, 2019). The de novo synthesis of purines and pyrimidines is extremely energy-consuming and requires amino acids and ribose; instead, the nucleotide salvage pathway needs less energy and uses purines and pyrimidines derived from nucleotide catabolism (Bernhard et al., 2023). Differentiated cells use the salvage pathway to produce nucleotides, while proliferative cells typically rely on de novo synthesis due to their increased purines and pyrimidines request. Thus, GBM cells reshape their nucleotide metabolism in order to increase progression, upregulating the de novo nucleotide synthesis to sustain their proliferation (Ghannad-Zadeh and Das, 2021). Using a large-scale targeted proteomics platform, Nakamizo et al. confirmed that the de novo pyrimidine synthesis is one of the most enriched pathways in GBM patients. Indeed, ribonucleotide reductase catalytic subunit M1 (RRM1) and nucleoside diphosphate kinase 1 (NME1), encoding for enzymes involved in nucleotide synthesis, are upregulated in GBM (Figure 3) (Nakamizo et al., 2022). Furthermore, GBM upregulates inosine monophosphate dehydrogenase-2 (IMPDH2), a rate-limiting enzyme involved in guanosine triphosphate (GTP) synthesis, improving RNA polymerase I and III transcription (Figure 3) (Kofuji and Sasaki, 2020).

Recently, Spina et al. demonstrated that the inhibition of dihydroorotate dehydrogenase (DHODH), involved in the de novo pyrimidine biosynthesis, reduces tumor growth, prolonging the overall survival in a GBM orthotopic mouse model (Spina et al., 2022). Notably, DHODH is required for ribosomal DNA (rDNA) transcription and its inhibition causes nucleolar stress in GBM cells, reducing their proliferation (Figure 3) (Lafita-Navarro et al., 2020).

Wang et al. demonstrated that the inhibition of purine synthesis, by altering the intracellular purine pool, reduces GSCs proliferation and tumor progression in vivo. However, the growth of differentiated GBM cells is not significantly influenced by the modulation of the enzymes involved in purine biosynthesis. The transcription factor MYC is implicated in GSCs de novo purine synthesis pathway and modulated by PI3K–AKT pathway, regulating the expression of enzymes involved in purine biosynthesis. Accordingly, GBM patients with higher expression of de novo nucleotide synthesis enzymes show a poor prognosis and dismal clinical outcome (Wang et al., 2017). Moreover, GBM could reprogram nucleotide salvage pathway to reuse purines and pyrimidines bases, as demonstrated by increased expression of related enzymes (Cortes Ballen et al., 2024). Metabolic reprogramming in GBM also affects nucleotide catabolic pathways, since the expression levels of enzymes involved in these pathways are dysregulated (Cortes Ballen et al., 2024).

In conclusion, GBM can reprogram nucleotide metabolism to increase its proliferation and progression. Therefore, the modulation of nucleotide metabolic pathways could be a therapeutic target or an adjuvant treatment for GBM patients. However, the de novo biosynthesis has been more studied and explored as compared to the salvage pathway, and few insights are available on purines and pyrimidines catabolism, hence new studies addressing these features are needed to gain a clearer understanding of nucleotide metabolism in GBM.

Iron serves as vital element for maintaining physiological processes, including DNA synthesis, cell proliferation, oxidative stress and energy production (Manz et al., 2016; Gao et al., 2019). Due to its multifaceted function, iron represents a pivotal element in the metabolic reprogramming of GBM, sustaining tumor aggressiveness and proliferation. Indeed, GBM cells exhibit an increased demand for iron to meet their metabolic needs (Schonberg et al., 2015). It has been observed that GSCs require a higher amount of iron as compared to non-stem cells due to their capacity of self-renewal. Knocking down ferritin, the protein responsible for storing iron in GSCs, impairs their proliferation capacity, indicating that iron is an essential element for their survival and growth (Schonberg et al., 2015).

It is reasonable to hypothesize that GBM cells modulate the expression of proteins involved in iron uptake to sustain tumor growth. Indeed, it has been shown that the expression of transferrin receptor 1 (TfR1), also known as cluster of differentiation 71 (CD71), is increased in GBM cells (Figure 4) (Rosager et al., 2017). As a matter of fact, TfR1 and ferritin, the main players involved in iron metabolism, are increased in correlation to the tumor grade, meaning that their increase is associated with poorer clinical outcomes (Liu et al., 2020). Interestingly, the differential expression of proteins involved in iron uptake between different GBM subtypes reflects the distinct prognoses associated with these phenotypes. Vo et al. discovered a commensal symbiosis between proneural and mesenchymal GSCs (Vo et al., 2022). In this scenario, proneural GSCs sustain mesenchymal GSCs releasing dopamine and transferrin, thus promoting the growth of neighbouring cells, enhancing iron uptake and increasing their proliferation capacity. Additionally, released dopamine increases TfR1 expression in mesenchymal GSCs, further promoting intracellular iron content (Vo et al., 2022).

Typical hallmarks of the TME, such as hypoxia and acidic environment, can influence iron metabolism, leading to an enhanced uptake and storage of iron, thereby facilitating uncontrolled cell proliferation and tumor growth (Wang et al., 2016). GBM cells also upregulate the levels of ferritin, the primary storage form of iron in the CNS, to increase iron sequestration (Liu et al., 2020). Li and collaborators observed increased levels of ferritin light chain (FTL) in TAMs within GBM microenvironment. The upregulation of FTL suppress calcium-independent phospholipase A2β (iPLA2β), whose activity, under physiological condition, prevents cells damage by inhibiting lipid peroxide accumulation (Li et al., 2023). This suppression promotes the polarization of TAMs towards the M2-like phenotype, which contributes to an immunosuppressive TME, thereby facilitating tumor progression and immune evasion (Li et al., 2023). Ferritin upregulation is associated with the epithelial-mesenchymal transition (EMT) and chemoresistance in glioma under hypoxic conditions, by activating the AKT/Glycogen synthase kinase-3 beta (GSK3β)/β-catenin signalling pathway (Figure 4) (Liu et al., 2020).

The maintenance of homeostatic iron levels is guaranteed by the iron regulatory protein 1 (IRP1, also known as ACO1), which regulates the expression of ferritin and transferrin at the post-transcription levels (Katsarou and Pantopoulos, 2020).

An additional piece of evidence, highlighting the involvement of iron metabolism in GBM metabolic reprogramming, analyses the role of epigenetic regulators that influence tumor growth. In an orthotopic GBM mouse model, mTOR2 complex (mTORC2) activation has been shown to induce H3 histone acetylation, promoting the transcription of iron-related genes, including TfR1, FTL and ferritin heavy chain (FTH1), which in turn enhances the survival of GBM cells (Figure 4) (Masui et al., 2019).

Iron also plays a significant role in maintaining the redox balance within GBM cells, participating in Fenton reaction in which ferrous iron (Fe²⁺) reacts with hydrogen peroxide (H₂O₂) to generate hydroxyl radicals. This means that excessive amount of free iron could lead to uncontrolled reactive oxygen species (ROS) production, the main responsible for DNA damage, contributing to genomic instability, an important hallmark of cancer (Torti and Torti, 2013). Moreover, these radicals could attack polyunsaturated fatty acids (PUFAs) in cell membranes, initiating lipid peroxidation, suggesting that the interplay between iron and lipid metabolism is another important aspect of GBM metabolic reprogramming.

The complex interplay between iron levels, metabolic demand and TME conditions underscores the importance of developing targeted therapies aimed at disrupting iron metabolism in GBM, by understanding the mechanism through which GBM cells manipulate iron levels.