Introduction

In recent years, substantial progress has been accomplished in the field of genetics and pathogenesis of pheochromocytoma/paraganglioma (PHEO/PGL) research. Advances in genetics and recognition of a high prevalence of PHEO/PGL in certain familial syndromes is now making it mandatory for routine screening of the tumor in patients with identified mutations, even in the absence of normally considered clinical signs and symptoms. Accumulating data also indicates that many more PHEOs/PGLs are caused by germline mutations than previously recognized, raising the importance of considering an underlying hereditary condition even when there is no obvious familial condition.

The number of PHEO/PGL susceptibility genes was recently increased to 21, a group that includes the von Hippel-Lindau (VHL) tumor suppressor gene [1], the rearranged during transfection (RET) proto-oncogene [2], the neurofibromatosis type 1 (NF1) tumor suppressor gene [3], the genes encoding the four succinate dehydrogenase complex (SDH) subunits (SDHA, -B, -C, -D) [47], and the gene encoding the enzyme responsible for flavination of the SDHA subunit (SDHAF2) [8]. Additionally, new susceptibility genes, transmembrane protein 127 (TMEM127) [9, 10], MYC-associated factor X (MAX) [11], and hypoxia-inducible factor 2α (HIF2A) [12, 13], have been identified. The kinesin family member 1B, transcript variant β (KIF1Bβ) [14, 15], prolyl hydroxylase 1 and 2 (PHD1/EGLN2 and PHD2/EGLN1) [16, 17], Harvey rat sarcoma viral oncogene (H-RAS) [18], Kirsten rat sarcoma viral oncogene (K-RAS) [19], isocitrate dehydrogenase 1 (IDH1) [20], fumarate hydratase (FH) [21, 22], and BRCA1-associated protein-1 (BAP1) [23] genes are also anecdotally reported. This year, germline mutations in malate dehydrogenase 2 (MDH2) [24] and somatic mutations in alpha thalassemia/mental retardation syndrome X-linked (ATRX) genes [25] were identified in PHEOs/PGLs.

Currently, there is a large effort to determine the similarities between PHEOs/PGLs resulting from different genetic mutations and to find the common signaling pathways and mechanisms involved in their pathogenesis. Microarray expression profile studies in hereditary PHEOs/PGLs showed two different signatures resulting from specific gene mutations thus dividing the neuroendocrine tumors into two clusters: cluster 1—pseudohypoxic (SDH, VHL, and lately FH and HIF2A mutations)—includes tumors presenting with impaired degradation and accumulation of HIF-1α/HIF-2α, which lead to changes in cell metabolism by pseudohypoxia (an aberrant activation of hypoxia response pathways in the absence of true oxygen deficiency [26]), angiogenesis, heightened reactive oxygen species (ROS) production, and diminished oxidation response. Cluster 2 gene mutations (RET, NF1, TMEM127, MAX, and KIF1Bβ) are connected by kinase signaling and protein translation pathway activation [2733]. Although these two clusters seem to show distinct cell signaling and new susceptibility genes were recently described, the available data suggests that the majority of PHEO/PGL susceptibility genes mutations are associated with dysregulation of several metabolic pathways, subsequently leading to defects in the hypoxia signaling pathway and adaptive responses to it [32, 34]. Mutations in genes encoding metabolic enzymes, such as SDHx subunits, IDH1, FH, or MDH2 (all were identified in PHEOs/PGLs as well as in a variety of other tumors, including acute myelogenous leukemia, gliomas, chondrosarcomas, and kidney cancer), disrupt the tricarboxylic acid (TCA) cycle and increase dependence on oxidative mitochondrial metabolism [24, 3540].

Recent evidence shows that pseudohypoxia and mitochondrial enzymes disruption may have direct oncogenic or tumor-suppressive effects by regulating and controlling diverse cellular processes [4143]. Alterations in cell metabolism have been shown to be associated with tumorigenesis and resistance of cancers to therapy in the past. Presence of pseudohypoxia has been found in a majority of cancers, and nowadays, it is considered to play an important role in the cancer pathogenesis. This present review outlines the ongoing story of PHEO/PGL pathogenesis, which shows an important role of the TCA cycle and hypoxia signaling in this process.

TCA Cycle Overview and Tumorigenesis

The TCA cycle, also known as the Krebs cycle, is a key metabolic pathway that unifies carbohydrate, lipid, and protein metabolism [44]. This cycle takes place in mitochondria and is the most important cellular metabolic network for oxidation of various energy sources, such as glucose, glutamine, and lipids [45]. Simply put, the TCA cycle is a cyclic route consisting of the oxidation of acetyl-coenzyme A (acetyl-CoA), deriving from glycolysis through pyruvate dehydrogenase (PDH) and from lipid β-oxidation to CO2, with the concomitant production of NADH and FADH2, whose electrons fuel the electron transport chain (ETC) for ATP generation. Besides being a central pathway for energetic metabolism, the TCA cycle provides metabolic intermediates for biosynthetic reactions leading to the de novo synthesis of proteins, lipids, and nucleic acids [46]. That means that most of the metabolic pathways in cells are directly or indirectly linked to mitochondria.

In proliferating cells, the TCA cycle operates in a different manner that is characterized by the exit of intermediates from the cycle to supply various biosynthetic pathways. Under these conditions, oxaloacetate (OAA) would become a limiting factor unless it was produced by another pathway that did not flow from mitochondrial citrate. OAA producing pathways, so-called anaplerosis, enable the TCA cycle to function as a biosynthetic pathway in addition to energy generation [47, 48].

For the TCA cycle to function properly, sufficient amounts of its substrates are needed and particular enzymes: citrate synthase (CS), aconitase (ACO), IDH, α-ketoglutarate dehydrogenase (α-KGDH), succinyl-CoA synthetase (SUCLG), SDH, FH, and MDH2. The most important source of carbon for energy-generating pathways that provide acetyl-CoA for oxidative metabolism in mitochondria is glucose [48]. The second most abundant nutrient is glutamine, which serves as a shuttle of carbon and nitrogen between organs. Glutamine is a major source of nitrogen for nonessential amino acids, nucleotides, and hexosamines [49]. Moreover, glucose and glutamine are versatile and in some cases can compensate each other to maintain TCA cycle function [48].

The TCA cycle (Fig. 1) begins with the condensation of acetyl-CoA with OAA to produce citrate, catalyzed by CS. Citrate can be then exported to the cytoplasm to be used as a precursor for lipid biosynthesis (via conversion by ATP-citrate lyase; ACLY) or stays in the mitochondria and is converted to isocitrate by ACO. Isocitrate is subsequently decarboxylated to α-ketoglutarate (α-KG) by IDH. α-KG is then converted to succinyl-CoA by α-KGDH complex or can exit mitochondria and serve as a precursor for amino acid biosynthesis. Succinyl-CoA is transformed to succinate in the reaction catalyzed by SUCLG or it can be utilized for porphyrin biosynthesis. Succinate is then oxidized to fumarate by SDH, which also represents complex II of the ETC. Fumarate is hydrated to malate by FH and, finally, malate is oxidized by MDH to restore OAA. The TCA cycle can be divided into two stages: (1) decarboxylating, in which citrate is converted to succinyl-CoA and releases two CO2 molecules and (2) reductive, which comprises the successive oxidations of succinate to OAA [46, 50]. A defect in any of the TCA cycle enzymes as well as depletion or abundance of its substrates leads to malfunction of the cycle and activation of adaptive mechanisms to assure cell survival. Many of these mechanisms/pathways are related to processes linked with tumorigenesis.

Fig. 1
figure 1

TCA cycle. Acetyl-CoA, the source of energy for TCA cycle, is formed from pyruvate through oxidative decarboxylation by PDH enzyme complex. Acetyl-CoA then enters and fuels TCA cycle as described in the text. α-KG alpha-ketoglutarate, α-KGDH alpha-ketoglutarate dehydrogenase, acetyl-CoA acetyl coenzyme A, ACLY ATP-citrate lyase, ACO aconitase, ADP adenosine diphosphate, ATP adenosine triphosphate, CO 2 carbon dioxide, CS citrate synthase, FAD flavin adenine dinucleotide, FADH 2 reduced FAD, FH fumarate hydratase, GDP guanosine diphosphate, GTP guanosine triphosphate, H 2 O water, IDH isocitrate dehydrogenase, MDH2 malate dehydrogenase 2, NAD + nicotinamide adenine dinucleotide (oxidized), NADH reduced form of NAD, PDH pyruvate dehydrogenase, SDH succinate dehydrogenase complex, succinyl-CoA succinyl coenzyme A, SUCLG succinyl-CoA synthetase

Mitochondrial Substrates in Tumorigenesis

Genetic defects––recessive mutations in genes of TCA cycle enzymes have been known to be associated with multisystem disorders and severe neurodegenerative diseases for many years [5153]. In recent years, dominant mutations associated with tumorigenesis were also described in TCA cycle enzymes, namely in SDH, FH, IDH, and MDH [7, 24, 54, 55]. These mutations usually lead to accumulation of TCA cycle substrates that, via the feedback loop or epigenetic changes, can promote tumorigenesis.

Pyruvate

Pyruvate is the metabolic intermediate and glycolytic end product that plays a crucial role in the metabolic switch between aerobic and anaerobic metabolism and is an important precursor for glucose, amino acid, and lipid synthesis. It can be utilized both in the cytosol and in mitochondria; under normoxic conditions, mitochondria mostly utilize pyruvate. Pyruvate transfer to mitochondria is mediated by the mitochondrial pyruvate carrier (MPC), a process that links the mitochondrial TCA cycle with the cytosolic glycolytic pathway [45]. Within the mitochondrial matrix, conversion of pyruvate to acetyl-CoA, NADH, and CO2 is catalyzed by PDH. In mitochondria, pyruvate is oxidized to fuel the TCA cycle and oxidative phosphorylation ATP production (OXPHOS) [45]. In the cytosol, pyruvate is reduced to l-lactate by LDH, and during this reaction, one molecule of ATP is produced and NAD+ is regenerated [56].

The PDH complex is the gatekeeper enzyme that links glycolysis to the TCA cycle and lipogenic pathways [5759]. Pseudohypoxia is one of the mechanisms involved in pyruvate deregulation. In cancer cells, metabolic changes to aerobic glycolysis may be due, in part, to inhibition of the PDH complex that is caused by increased expression of pyruvate dehydrogenase kinase (PDK) 1, 2, and 4 isoforms, leading to mitochondrial dysfunction. The activity of PDK is stimulated by ATP, NADH, and acetyl-CoA; inhibitory effects are exerted by ADP, NAD+, CoA-SH, and pyruvate [60]. Moreover, HIF-α activation also mediates the change of pyruvate distribution in the cell. HIF-1α was found to induce transcriptional upregulation of some of the proteins participating in pyruvate metabolism; for instance, PDK1 and LDH-A [61, 62], reviewed in [45]. PDK1 phosphorylates the E1 subunit of PDH complex that leads to its inactivation and consequently prevents the entry of pyruvate into mitochondria and, thus, its accumulation in the cytosol [59, 63]. In the metabolic environment of cancer cells, the allosteric modifications of PDK activity results in downstream mitochondrial dysfunctions, such as defects in the TCA cycle and/or the electron transport chain and, thus, are responsible for changes in the metabolic phenotype [60]. Besides the TCA cycle, the PDH complex is closely connected to the ETC. In the situation when PDH complex activity (and as a consequence the TCA cycle) is impaired, there is decrease in generation of the reducing equivalents NADH and FADH2, which donate electrons to the respiratory chain to complete the OXPHOS by generation of ATP. As a result of a defective PDH complex, TCA cycle, or respiratory chain, cells develop common features such as lactic acidosis, alteration in redox status, and deregulation of the ADP/ATP ratios and free CoA/acetyl-CoA [64].

The other player in disruption of pyruvate metabolism is pyruvate kinase isoenzyme M2 (PKM2), overexpressed in many cancers [56, 65]. Its knockdown was shown to decrease glycolysis and inhibit cell proliferation in cancer cells [66]. Upregulation of PKM2 is controlled by the pohosphoinositide-3-kinase (PI3K)/protein kinase B (AKT)/mammalian target of rapamycin (mTOR) signaling pathway and is transcriptionally mediated by HIF-1α and by c-Myc-heterogenous nuclear ribonucleoprotein-dependent gene splicing. PKM2 is a direct transcriptional target of HIF-1α and simultaneously promotes HIF-1α-mediated transactivation and reprogramming of glucose metabolism. Hydroxylation of PKM2 by PHD3 facilitates direct interaction with HIF-1α and promotion of transcriptional transactivation of HIF-1α target genes. Both PKM2 and PHD3 are targets of HIF-1α, meaning that the positive feedback loop between them sustains the expression of glycolytic genes and, thus, facilitates the Warburg effect. Moreover, PKM2 acts as a transcriptional co-activator not only for HIF-1α but also for Oct-4, β-catenin, and STAT3 (reviewed in [45]).

Succinate

Succinate, as mentioned above, is oxidized to fumarate by complex II of ETC, also known as SDH or succinate:ubiquinone oxidoreductase (SQR). ETC consists of four multimeric protein complexes that are anchored to the inner mitochondrial membrane. Together, they catalyze the oxidation reducing equivalents (NADH), using O2 as a terminal acceptor of electrons [67, 68]. Complex II is a unique structure that consists of four functionally different subunits broken up into two groups: the SDH portion consists of SDHA and SDHB whereas SDHC and SDHD make up the SQR component [69, 70]. Its function is to couple the oxidation of succinate to fumarate in the mitochondrial matrix (as part of the TCA cycle) with the reduction of ubiquitin (UQ) in the membrane (as part of the ETC) [7, 32]. Two SDH accessory factors, SDH complex assembly factors (SDHAFs) 1 and 2, were identified lately and they are important for SDH complex assembly and flavination of the SDH catalytic subunit, respectively [8, 71]. The SDH portion of complex II catalyzes succinate to fumarate oxidation and the simultaneous production of FADH2 while the SQR pathway mediates the transfer of electrons generated during oxidation of succinate to reduce ubiquinone to ubiquinol and, subsequently, to respiratory complex III. Electron transfer within the ETC is coupled at specific points to the extrusion of protons into the mitochondrial intermembrane space. This fuels the ATP synthase complex (complex V) for ATP generation. Coupling between ETC and ATP synthesis is called OXPHOS [68].

Impaired function of complex II was shown to be associated with tumorigenesis via stabilization and activation of HIF-α and increased ROS production. Recent works have described complex II as an important site of ROS production in the form of superoxide [7274] and have also linked it to ROS-mediated execution of apoptotic cell death [75]. Albayrak et al. [76] demonstrated that cell death induction by SDHC expression is associated with a transient inhibition of complex II and ROS generation. Moreover, cells deficient in SDHC are resistant to several proapoptotic cytostatic agents. It seems that complex II acts as a cell death sensor responding to acidification upon toxic stimuli by its disassembly and thereby further facilitating ROS-mediated apoptosis [68, 77]. The other oncogenic mechanism is mediated through SDH complex subunits dysfunction resulting from gene mutations. In SDHx-mutated PHEOs/PGLs, high succinate accumulation has been detected [78, 79] and the succinate:fumarate ratio has been found to be higher in SDHB-mutated and metastatic PHEOs/PGLs in contrast to SDHC/D-related tumors or tumors without metastases [79, 80]. Succinate has a structure similar to α-KG and exerts its ability to modulate the activity of α-KG-dependent dioxygenases––the enzyme family with diverse functions, including epigenetic regulation, oxygen sensing, collagen biosynthesis, fatty acid metabolism, and translation regulation [35, 81, 82]. The α-KG-dependent dioxygenase family also includes HIF prolyl hydroxylases (PHDs), which are needed for HIF-α hydroxylation and its further recognition by pVHL and eventual proteasome degradation [8385]. Succinate acts as a competitive inhibitor of PHDs and thus promotes the activation of the HIF-α signaling pathway and expression of HIF target genes (Fig. 2) ([8688]; reviewed in [32]). Additionally, several studies have demonstrated succinate’s ability to remodel the epigenome and alter gene expression. Succinate accumulation leads to DNA hypermethylation by inhibition of ten-eleven-translocation methylcytosine dioxygenase (TET) in PHEOs/PGLs and gastrointestinal stromal tumors [89, 90]. These findings emphasize the interconnection between the TCA cycle and epigenomic changes in cells. Moreover, succinate was shown to act as a signal for inflammation; during the inflammatory process, it accumulates in immune cells, leading to HIF-α activation [91]. These findings are very interesting, particularly because inflammation and tumorigenesis seem to be closely related.

Fig. 2
figure 2

Some of the metabolic pathways involved in HIF-α activation in PHEOs/PGLs. α-KG alpha-ketoglutarate, α-KGDH alpha-ketoglutarate dehydrogenase, acetyl-CoA acetyl coenzyme A, ACO aconitase, Akt RAC-alpha serine/threonine-protein kinase, CS citrate synthase, c-Myc Myc proto oncogene, eIF-4E eukaryotic translation initiation factor 4E, ERK mitogen-activated protein kinase 2, FH fumarate hydratase, HIF-α hypoxia-inducible factor α, IDH isocitrate dehydrogenase, MAX myc-associated factor X, mTORC1 mammalian target of rapamycin complex 1, mTORC2 mammalian target of rapamycin complex 2, MDH2 malate dehydrogenase 2, NF1 neurofibromin 1, PDH pyruvate dehydrogenase, PHD prolyl hydroxylase domain protein, PI3K phosphoinositide 3-kinase, pVHL von Hippel-Lindau protein, Raptor regulatory associated protein of mTOR, RAS rat sarcoma oncogene, RET rearranged during transfection proto-oncogene, Rheb RAS homolog enriched in brain, S6K S6 kinase, SDH succinate dehydrogenase complex, succinyl-CoA succinyl coenzyme A, SUCLG succinyl-CoA synthetase, TMEM127 transmembrane protein 127, TSC1/2 tuberous sclerosis complex 1/2, UQ ubiquitin

Fumarate

Fumarate is a TCA cycle metabolite that is converted to malate by FH. Inactivation of FH, e.g., due to its mutations, leads to fumarate accumulation. In high levels, fumarate can act as an oncometabolite through several mechanisms. Similarly to succinate, it acts as an α-KG-dependent dioxygenase inhibitor and consequently leads to HIF-α activation by PHD inhibition (Fig. 2) [35, 82, 86]. Recent findings suggest that there are also other HIF-independent mechanisms of oncogenesis. Fumarate is a moderately reactive α,β-unsaturated electrophilic metabolite that can covalently bind to cysteine residues of proteins in a process called succination and is a feature of FH-deficient tumors [92, 93]. Succination is a non-enzymatic and irreversible reaction [94]. Succinated proteins have been detected in a variety of animal and cellular models, and some of these proteins were identified to be associated with tumorigenesis [82, 93]. Furthermore, FH-mutated PHEOs/PGLs displayed the same epigenetic changes as SDH-mutated tumors involving the alterations in DNA methylation [21, 89].

Fumarate accumulation is probably also associated with mutations in another TCA cycle enzyme, MDH2. MDH2-mutated tumors have a transcriptional profile similar to SDH-mutated tumors. Like succinate and fumarate, malate also inhibits PHDs [95, 96] and, thus, HIF-α prolyl hydroxylation. Nevertheless, a high fumarate:succinate ratio was detected in MDH2-mutated tumors, suggesting fumarate accumulation, which is in contrast to SDHx-mutated tumors [24].

Citrate

Citrate occupies a central position in cellular metabolism as a substrate of the TCA cycle and as an important switch between metabolic pathways [45, 97]. It serves as a metabolic sensor and is involved in many biological processes, including inflammation, cancer, insulin secretion, and histone acetylation. After its synthesis in mitochondria, citrate becomes a substrate in the TCA cycle and, after oxidation, serves as a major source of cellular ATP production. Citrate inhibits and induces important strategic enzymes involved in glycolysis, the TCA cycle, gluconeogenesis, and fatty acids synthesis [98]. Under physiological conditions when citrate is sufficiently produced by TCA cycle, it represses the TCA cycle by inhibition of PDH and SDH enzymes and, when transported outside the mitochondria by the citrate carrier (CIC), acts as a glycolytic inhibitor (reviewed in [45, 56]). This means that it slows down or arrests both the TCA cycle and glycolysis and stimulates gluconeogenesis and lipid synthesis. In normal cells, together with ATP, citrate inhibits enzymatic activity of phosphofructokinase (PFK) 1 (a rate-limiting enzyme in glycolysis) and PFK2 [45, 56]. PFK2 produces fructose 2,6-bisphosphate, which is a powerful allosteric activator of PFK1 in cancer cells. Increased levels of PFK2 enable to overcome citrate and ATP inhibition of PFK1 when glucose uptake is high. Moreover, overexpression of the nuclear isoform of PFK2 (PFK3B) sustains a high rate of glycolysis and stimulates proliferation. Inhibition of PFK2 suppresses cell proliferation [99101]; reviewed in [45, 56]. Citrate also indirectly inhibits pyruvate kinase (PK) by inhibition of PFK1, leading to decreased levels of fructose 1,6-bisphosphate, a strong allosteric activator of PK [102]. Moreover, cytosolic citrate is cleaved by ACLY to OAA and acetyl-CoA, which are used for pyruvate, fatty acid, and sterol synthesis, respectively [98]. ACLY has been shown to be involved in metabolic modulation mediated by PI3K/AKT signaling pathway. Activation of the PI3K/AKT pathway induces glucose uptake and promotes glucose carbon flux into biosynthetic pathways but its central role in oncogenesis is to relay the reprogamming of citrate metabolism [103]. In PHEOs/PGLs arising from SDHx and VHL mutations, lower levels of citrate, isocitrate, and ACO have been observed compared to those harboring RET, NF1, or TMEM127 mutations, demonstrating decreased OXPHOS and presence of pseudohypoxic state in those tumors [79].

The citrate level in the cytosol depends on the mitochondrial export rate, which can be modulated by the level of CIC expression; this is regulated by several transcription factors, including Sp1 and NF-κB [104]. CIC was shown to be overexpressed in cancer cells with its inhibition having antitumor activity, albeit no toxicity on adult normal tissues [105]. Citrate was also suggested to be used in cancer treatment since it was shown to induce cancer cell death directly or facilitate toxic effects of conventional anticancer drugs; reviewed in [45].

Glutamine, α-KG, and D2HG

Glutamine plays a crucial role in the metabolism of cancer cells since it is involved in cell bioenergetics and is a source of nitrogen for the production of molecules essential for cell growth [106108]. Glutamine is the second most important carbon-based source of energy after glucose [45]. It serves a number of anabolic processes in cancer cells, such as protein and nucleotide synthesis and conversion to glutamate. Glutamate is used for amino acid synthesis or is transformed into α-KG that enters the TCA cycle [109]. Besides that, glutamine regulates the activity of mammalian target of rapamycin complex 1 (mTORC1). mTORC1 is a master regulator of cell growth, activates protein translation, and inhibits macroautophagy in response to abundance of amino acids and growth factor signaling [110, 111]. Glutamine conversion to glutamate is mediated by glutaminases, GLS and GLS2, which are under control of transcription factor c-Myc [98, 112]. The role of GLS in cancer is isoform-specific: GLS was proposed to work as an oncogene whereas GLS2 seems to exert antitumor activity [107, 108, 113115]. Oncogenic expression of GLS was demonstrated to be dependent on the c-Myc transcription factor mediated by suppression of miR-23a/b [116] and c-Myc has also been found to stimulate glutamine catabolism and cell proliferation via the upregulation of glutamine importer SLC1A. GLS expression can also be activated by ErbB2 signaling through activation of NF-κB [45, 117].

The product of GLS activity, glutamate, is an essential precursor for antioxidant glutathione synthesis and also serves as a donor of amino groups for non-essential amino acids. Glutamate is transformed to α-KG in a reaction catalyzed either by transaminases for production of non-essential amino acids or by glutamate dehydrogenase (GDH). Under conditions of limited glucose supply, GDH-mediated conversion is prevalent. α-KG is then utilized in the TCA cycle, either for succinate (via α-KGDH) or for citrate (via IDH) production. Both of them are important substrates—succinate for the TCA cycle and citrate for fatty acid synthesis. The reductive pathway of citrate production is favored in cells under (pseudo)hypoxic conditions to compensate the drop in glucose-derived citrate production [45]. Lately, it has been described how low levels of intracellular citrate and high levels of α-KG are associated with HIF-dependent reductive carboxylation of glutamine [118120]. According to Sun and Denko [121], hypoxia causes a decrease in glucose-derived citrate because of decreased PDH activity and also increases α-KG levels as a result of decreased αKGDH activity. These changes in substrate concentrations also drive the reverse reaction of IDH. This metabolic switch keeps maintenance of lipid synthesis that is required for the growth of cells in hypoxia.

D2HG is an oncometabolite that is synthesized in place of α-KG in IDH1/2-mutated cells [122]. D2HG is structurally very similar to α-KG and glutamate; this similarity results in the competitive inhibition of α-KG-dependent dioxygenases [123]. Unlike succinate and fumarate, the effect of D2HG on PHDs is not completely clear since there is evidence of both inhibition [123, 124] and activation of PHDs [125, 126]; reviewed in [127]. According to the work of Williams et al. [128], mutant IDH1 expression in gliomas was not sufficient to stabilize HIF-1α. But there is evidence that accumulation of D2HG is associated with epigenetic changes, disruption of defense mechanisms against ROS, and changes in redox homeostasis. Mechanisms of its involvement in tumorigenesis are not fully understood yet but are thought to be based on epigenetic modifications that result in differential expression of genes involved in cell proliferation, changes in HIF-α levels, as well as aberrant extracellular matrix structures induced by changes in collagen synthesis [39, 45, 127]. Further studies are needed to fully explore metabolic and physiological defects present in IDH-mutated tumors as well as their consequences and roles in tumorigenesis.

Current Views on Cancer Cell Metabolism Linked to Craving for Oxygen in the Pathogenesis of PHEO/PGL

Although in recent years scientists were able to reveal several crucial signaling pathways involved in tumor development, precise metabolic changes still remain hidden and further research is needed. It is currently known that cancer cells switch from OXPHOS-producing ATP to aerobic glycolysis-producing lactate and suppression of mitochondrial function. This is mostly mediated by changes in the fate of the end product of glycolysis––pyruvate. Under normal conditions in healthy cells, pyruvate is directed to mitochondria, metabolized by PDH to acetyl-CoA, and enters the TCA cycle to fuel OXPHOS. But in most cancer cells, PDH activity is suppressed, leading to reduced flow of pyruvate to the mitochondria and a decrease in OXPHOS. Under these conditions, pyruvate is converted to lactate by LDH [45]. This is the situation when the cell starts to “gasp” for the air even in the presence of oxygen. This feature, referred to as the “Warburg effect” or “aerobic glycolysis,” was first observed by Otto Warburg in 1924 and is considered a defect in mitochondrial respiration and/or ATP production [129131]. Although glycolysis is much less efficient in the production of ATP compared to OXPHOS (2 vs. 36 molecules of ATP), it makes cancer cells more resistant to hypoxia (oxygen deprivation) as a result of excessive growth or their high metabolic activity and poor oxygen supply. Thus, aerobic glycolysis is present especially in the early avascular phase of tumorigenesis, where it provides an advantage in hypoxic conditions [132]. This metabolic shift also provides the cell with resources to sustain proliferation. In the past, it was anticipated that these metabolic changes are caused by defects in mitochondria but it is now clear that they emerge from specific metabolic regulatory signaling. Should we, therefore, refer to some cancers as a metabolic disease? Certainly, yes, as suggested for kidney cancer by Linehan et al. [40, 133]. Mutations in known kidney cancer susceptibility genes lead to dysregulation of at least one metabolic pathway involved in the sensing of oxygen, iron, or nutrients, and particularly in the TCA cycle enzymes. Special attention should also be given to PHEOs/PGLs as several hereditary PHEOs (related to mutations in VHL, SDHx, FH) are closely linked to the development of kidney cancer and can occur concurrently. Thus, for PHEOs/PGLs, as outlined above, several metabolic hits in the TCA cycle occur that lead to pseudohypoxia. Succinate or fumarate accumulation occurs in SDHx- or FH- and MDH2-mutated PHEOs/PGLs, respectively. In SDHx- and VHL-mutated PHEOs/PGLs, lower levels of citrate, isocitrate, and ACO were detected, which is associated with epigenetic and metabolic changes involved in tumorigenesis. All these metabolic changes result in a metabolic shift from OXPHOS to aerobic glycolysis with further “gasping for air” by TCA metabolite-specific activation of the HIF signaling pathway. HIF-α stabilization and inhibition of α-KG-dependent dioxygenases was described in VHL- and SDHx-mutated tumors [28, 134139]. HIF enhances the glycolytic pathway by increasing expression of target genes involved in the glycolytic and anabolic processes (glucose transporters (GLUT) 1, 2, hexokinase (HK) 2, PKM2, LDH-A) [34, 140]. Although some studies have not found the GLUT1 overexpression of SDHx-related tumors [29, 135, 141], there was still increased expression of GLUT3 and HK2 as well as LDH-A [135, 141]. Functional analysis of PHEO/PGL tumor tissue from patients with somatic HIF2A and germline PHD1 and PHD2 mutations also showed increased activity of the HIF signaling pathway and its target gene expression, including GLUT1 and in PHD-mutated PHEOs/PGLs also LDH-A [12, 16]. Recently, Berkel et al. [142] reported high expression of HK2 in SDHx-related PHEOs/PGLs when compared to sporadic and RET-, NF1-, and MAX-related tumors. All PHEOs/PGLs in their study showed similar GLUT1 and GLUT3 expression.

HIF-α activation is becoming a significant detriment to PHEOs/PGLs but most likely, not solely, results in tumorigenesis acceleration, tumor growth, and metastasis development. Now, these tumors become truly hypoxic and are “gasping for air,” culminating in no escape from normal cell function collapse and ongoing tumorigenesis. Indeed, this is not the only oncogenic mechanism in PHEOs/PGLs but it seems to play an important role in their development.

Cells, in craving for oxygen, activate a number of adaptive responses, coordinated by various signaling pathways, most of them controlled by HIFs [143145]. HIFs are transcription factors comprised of ubiquitously expressed three α subunit isoforms (oxygen-regulated, HIF-1α, HIF-2α, HIF-3α) and a constitutively expressed β subunit [144146]. HIF-α regulates the expression of a large number of genes related to adaptation to hypoxia, such as those involved in angiogenesis, glycolysis, survival, differentiation, or proliferation. Under normoxic conditions, HIF-1α and HIF-2α are degraded via the ubiquitin-proteasome pathway, a process involving prolyl hydroxylation by PHDs (PHD1, PHD2, and PHD3) and the recognition of hydroxylated HIF-α subunits by pVHL. pVHL is a component of an E3 ubiquitin ligase complex that targets HIF-α for proteasomal degradation [145]. Under hypoxic conditions, HIF-α becomes stabilized, heterodimerizes with HIF-β, recruits co-activators, binds to the DNA at hypoxia-responsive elements (HREs), and activates transcription of the target genes (reviewed in [32]). There are also multiple other oxygen-independent oncogenic pathways involved in HIF-α regulation. Two main pathways that induce the pseudohypoxic state independently of oxygen tension are the PI3K and mTOR pathways. Their signaling leads to stabilization of HIF-1α, which causes a strong increase in glycolysis––the Warburg phenotype [45, 147, 148]. Full glucose oxidation in mitochondria is decreased, resulting in decreased ATP and citrate production. In turn, the feedback of these molecules on the main regulator enzyme for glycolysis, PFK1, is broken. This leads to acceleration of glycolysis and enhanced production of lactic acid [147]. HIF-α enhances glycolysis by tilting the balance between PDH and LDH in favor of LDH [63, 77, 149]. Occasionally, impairment of mitochondrial respiration can be observed, resulting in defective OXPHOS and contributing towards tumorigenesis [147]. Reduced activity of the respiratory chain results in reduced ROS generation and, thus, a decrease in apoptotic signals [77, 150]. Interestingly, HIF-α-driven pseudohypoxia might drive overproduction of TCA substrates, for instance D2HG [151], giving the bidirectional connection of TCA with metabolic pathways in cell.

SDHx mutations disrupt complex II activity, resulting in succinate accumulation and inhibition of PHD-catalyzed HIF-α hydroxylation required for HIF-α proteasomal degradation [8688, 152], thus, mimicking hypoxia. A similar scenario happens in FH- and MDH2-deficient tumors, since fumarate has the ability to competitively inhibit PHDs (Fig. 2). Additionally, both metabolites inhibit other members of α-KG-dependent dioxygenases, including histone lysine demethylases (KDMs) and TET enzymes [89, 90, 153, 154]. Competitive inhibition of chromatin-modifying α-KG-dependent dioxygenases results in marked impairment of epigenetic regulation of gene expression and DNA hypermethylation phenotype [89, 90]. Abnormal fumarate and succinate accumulation also exerts tumorigenic effects by other mechanisms: overproduction of ROS may participate in oncogenesis and tumor progression by irreversible DNA modifications and protein oxidation. Moreover, protein succination may result in the constitutive activation of the NRF2-mediated antioxidant defense pathway that has the ability to promote tumorigenesis by enhancing ROS elimination as well as by generating a reductive environment that can facilitate cell proliferation and survival. The functional impairment of multiple proteins leads to subsequent dysregulation of cellular metabolism and activation of oncogenic pathways [151, 154157].

In PHEOs/PGLs, besides the mutations in genes encoding TCA cycle enzymes, other mutations in genes that directly or indirectly influence the TCA cycle, HIF signaling, and other pathways involved in tumorigenesis were described. Mutations in VHL, PHD1/2, and HIF2A genes are associated with direct HIF-α pathway dysregulation (Fig. 2) [24, 32, 34]. In VHL mutations, oxygen sensing is impaired, resulting in HIF-α stabilization in different ranges of activation depending on the type of mutation. In VHL type 1 disease, mutations are usually nonsense or deletions and result in complete dysfunction of pVHL, whereas VHL type 2 disease mutations are commonly missense and cause mild changes in pVHL structure and compromise the activity of ubiquitin-ligase complex in variant extent [158160]. Interestingly, VHL mutations associated with mild changes in pVHL function confer a higher risk for PHEO/PGL development than those associated with complete dysfunction of pVHL, suggesting that besides pseudohypoxia there are also other mechanisms involved in PHEO/PGL pathogenesis (reviewed in [161]). As mentioned above, HIF-α prolyl hydroxylation is a crucial step for its further proteasomal degradation. Germline loss-of-function PHD1 and PHD2 mutations described in patients with PHEO/PGL and polycythemia lead to stabilization and activation of HIF-α by the decreased or lost ability to hydroxylate HIF-α prolyl residues [16, 17, 144, 145]. The role of HIF-α, especially HIF-2α, in PHEO/PGL development is supported by the detection of somatic and germline gain-of-function mutations in the HIF2A gene, causing HIF-α stabilization. HIF2A mutations are associated with a specific clinical phenotype: multiple PHEOs/PGLs, polycythemia, and, often, duodenal somatostatinomas [12, 13, 162, 163].

Indirect activation of HIF-α can be found in cluster 2 PHEOs/PGLs arising from NF1, RET, TMEM127, K-RAS, H-RAS, and MAX mutations (Fig. 2). HIF-α activation in these tumors is not expressed to the same extent as in cluster 1 tumors. However, these clusters are most probably interconnected via mTOR and PI3K signaling pathways [32]. The PI3K signaling pathway is linked to growth control, glucose metabolism, and can also activate HIF [164]. Activation of PI3K and mTOR pathways leads to HIF-α accumulation and to induction of its downstream genes transcription. In RET, NF1, and K-/H-RAS mutations, HIF up-regulation is driven through activation of the RAS/RAF/MEK/ERK, or mitogen-activated protein kinase (MAPK), signaling pathway, resulting in activation of downstream pathways (PI3K and mTORC) involved in oncogenic processes (reviewed in [32]). mTOR-dependent translation of HIF-α is induced by the loss of tuberous sclerosis 1 and 2 (TSC1, TSC2) or serine/threonine kinase LKB1 tumor-suppressor genes and via the activation of the AKT signaling pathway (Fig. 2) [165170]. Moreover, mutations in tumor suppressor MAX lead to deregulation of c-Myc signaling that enhances both the glycolytic pathway and OXPHOS. c-Myc is a very powerful gene transcription activator, labeled also as the “oncogene from hell” [171]. It regulates genes involved in the biogenesis of ribosomes and mitochondria, in the regulation of glucose and glutamine metabolism, and can also induce DNA damage, increase ROS production, and genome instability [151, 172176]. c-Myc was shown to regulate glycolysis in cells under normoxic conditions via direct activation of LDH-A and other glycolytic genes [148]. Metabolic reprogramming in cancer cells can be mediated via crosstalk between HIF-α and c-Myc. c-Myc transcriptional activity is also directly regulated by sirtuin 1 (SIRT1), either by c-Myc deacetylation or by binding c-Myc and promoting its association with MAX [151, 177, 178].

In K-RAS-mutated tumors, in addition to activation MAPK signaling pathway and subsequent HIF-α activation, the strong link between RAS, mitochondria, and the Bcl-2 family proteins was described. Increased K-RAS expression is associated with mitochondrial dysfunction, mitochondrial ROS production, and, thus, with tumorigenesis [179, 180].

The KIF1β tumor suppressor gene is a downstream target of PHD3 necessary for neuronal apoptosis and seems to be connected to both the HIF and MAPK signaling pathways [14, 15].

Conclusions and Future Therapeutic Perspectives

There is increasing evidence that pseudohypoxia—gasping for the air—plays an important role in PHEO/PGL development and is a result of interplay between several metabolic signaling pathways. The role of mitochondria in tumorigenesis is indubitable, since most metabolic pathways in the cell are somehow interconnected with this organelle and with the TCA cycle. Mitochondrial dysfunction is quite common in cancers and results in changes of mitochondrial energy metabolism, specifically in the metabolic switch from OXPHOS to glycolysis. Pseudohypoxia and the metabolic switch to aerobic glycolysis are associated with a higher aggressiveness of tumors, resistance to systemic and radiation therapies, along with worse prognosis for patients [181185].

This article shows how initial stages of “gasping for air” in PHEOs/PGLs are closely linked to metabolic and hypoxic (HIF signaling pathways directly) reprogramming of these tumors. In the past years, there is clear evidence of the direct role of TCA cycle enzymes and substrates in tumor development, including PHEOs/PGLs. Defects in SDH, FH, IDH, and MDH2 enzymes lead to accumulation of TCA metabolites, resulting in dysregulation of cell metabolism, generation of a pseudohypoxic state, alterations in epigenetic homeostasis, and other protumorigenic processes. The key role of TCA cycle enzymes in tumorigenesis designates them as very promising therapeutic targets. Based on actual understanding of the biochemical routes that redirect and reprogram cell metabolism in PHEOs/PGLs, two strategies in treatment could be used: one that directly focuses on inhibiting HIF pathway signaling and another aimed at fixing metabolic reprogramming in cancer cells.

Nowadays, several agents affecting HIF-1α signaling have been introduced, with varying results depending on a cancer’s HIF-α phenotype [43, 186188]. They include antiangiogenic agents, such as inhibitors of vascular endothelial growth factor (VEGF), mTOR, and heat shock protein 90 inhibitors, or agents that restore or activate PHD enzyme activity. Drugs selectively targeting HIF-2α signaling are under development; for example, two cembrane diterpenes were found to selectively inhibit HIF-2α and modulate its downstream effectors [189191]. In terms of inhibiting HIF signaling pathway, agents targeting both HIF-1α and HIF-2α subunits, such as PI3K inhibitors, dual PI3K/mTOR inhibitor, or JNK inhibitor are of great interest since they can activate different genes and, in many cancers, both are expressed in a certain balance [138, 192194].

Dysregulated TCA cycle is another very promising therapeutical target. Restoring the activity of malfunctional enzymes or providing the depleted substrates could reverse energy production back to OXPHOS and, thus, restore mitochondrial functions. Several promising agents are under development, e.g., drugs restoring the functionality of mitochondrial enzymes, small molecule inhibitors of certain proteins, or glycolysis inhibitors. For example, proteostasis regulators, such as histone deacetylase inhibitors (HDACis), have the potential to increase the stability and, in effect, the total amount of mitochondrial SDHB protein in SDHB-deficient cells [195]. Another compound––redox-silent vitamin E analog α-tocopheryl succinate (α-TOS)––with known antiproliferative effects, has been found to target the UbQ binding site of complex II [196] and shown to be effective in several experimental cancers in vitro and in vivo with no toxicity to normal cells [197200]. The other possibility of treatment is inhibition of glycolysis by small molecule inhibitors of GLUT1 [201, 202] or its reversion using dichloroacetate (DCA) [203, 204] or 3-bromopyruvate (3BP) [205207]. Activation of mitochondrial respiration can also be induced by downregulation of LDH by small interfering ribonucleic acid (siRNA), which prevents conversion of pyruvate to lactate and, as a result, allows for the entry of pyruvate into the mitochondria in order to fuel the TCA cycle and OXPHOS [208]. Discovery of IDH1/2 mutations led to the development of small-molecule inhibitors that reduced D2HG levels in cancer cells and inhibited tumorigenesis; IDH1/2 inhibitors were proven to be successful in a glioma xenograft model and produced cytostatic rather than cytotoxic effects [209211]. Since glutamine is highly utilized in cancer cells, several approaches aimed to influence different steps of glutamine metabolism are studied: suppressing cancer cell glutamine uptake, suppressing glutamine-dependent anaplerosis, inhibiting complex I, targeting glutamine-dependent mTOR activation, and enzymatic lowering of blood glutamine levels (reviewed in [111]). Therapies targeted against glutamine seem to be very promising, particularly in tumors displaying glutamine dependence. Another attractive target in PHEO/PGL and cancer treatment is citrate, levels of which are decreased in tumor cells and have the capability to induce apoptotic cell death or facilitate the toxic effects of conventional anticancer drugs [212214].

The other direction in PHEO/PGL research is development of novel diagnostic methods based on biomarkers associated with altered metabolic pathways and pseudohypoxia that will help choose targeted therapy and, thus, personalize anticancer treatment.