Edited by: Jane Mellor, University of Oxford, United Kingdom
Reviewed by: Cinzia Allegrucci, University of Nottingham, United Kingdom; Gokul Gopinath, Texas A&M University Baylor College of Dentistry, United States
This article was submitted to Epigenomics and Epigenetics, a section of the journal Frontiers in Genetics
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Cancer cells reprogram their metabolism to meet their demands for survival and proliferation. The metabolic plasticity of tumor cells help them adjust to changes in the availability and utilization of nutrients in the microenvironment. Recent studies revealed that many metabolites and metabolic enzymes have non-metabolic functions contributing to tumorigenesis. One major function is regulating epigenetic modifications to facilitate appropriate responses to environmental cues. Accumulating evidence showed that epigenetic modifications could in turn alter metabolism in tumors. Although a comprehensive understanding of the reciprocal connection between metabolic and epigenetic rewiring in cancer is lacking, some conceptual advances have been made. Understanding the link between metabolism and epigenetic modifications in cancer cells will shed lights on the development of more effective cancer therapies.
One hallmark of tumor cells is their rewired metabolism to meet the requirement for macromolecular biosynthesis, survival, and proliferation (
Epigenetics is defined as heritable changes in gene expression independent of mutations in genomic DNA. It originally includes histone post-translational modifications such as acetylation, methylation, ubiquitination, phosphorylation, SUMOylation and DNA modifications. With the development of proteomics and mass spectrometry technology, the repertoire of chromatin modifications is expanding with more epigenetic modifications identified such as acylation (crotonylation, succinylation, propionylation, β-hydroxybutyrylation),
Cellular metabolism and epigenetic modifications interact with one another and are regulated in a reciprocal manner. Most chromatin post-translational modifications, such as phosphorylation, acetylation, methylation, acylation, and O-linked
Most chromatin modifying enzymes use intermediary metabolites as cofactors or substrates and thus their activity is regulated by the availability of these metabolites. In the following sections, we will first describe the metabolism of four common intermediates (acetyl-CoA, SAM, α-ketoglutarate, NAD+). Then we will discuss the effect of metabolites from glycolysis, tricarboxylic acid (TCA) cycle, and fatty acids metabolism on chromatin modifications.
Histone acetylation is performed by lysine acetyltransferases (KATs) that transfer the acetyl group from acetyl-CoA to histones. Acetylation neutralizes the positive charges of lysine residues on histones, which eliminates the electrostatic interaction between histones and DNA, leading to a less compact chromatin structure permissive for gene transcription. Histone acetylation is sensitive to changes of global acetyl-CoA levels. As the acetyl group donor, acetyl-CoA is generated from three major sources: glucose, fatty acids, and acetate (
Metabolism of acetyl-CoA and histone acetylation. Glucose-derived pyruvate is metabolized to acetyl-CoA by PDC in the mitochondria. Mitochondrial acetyl-CoA needs to be converted to citrate or acetylcarnitine in order to be exported into cytoplasm and nucleus. In the nucleus, acetyl-CoA is regenerated from citrate and acetylcarnitine by ATP-citrate lyase (ACLY) and carnitine acetyltransferase (CAT), respectively, for histone acetylation. Nucleus acetyl-CoA can also be produced from glucose-derived pyruvate by nucleus PDC. Fatty acids (octanoate) can be oxidized to produce acetyl-CoA in the mitochondria but it is unknown how it is transported into the nucleus. ACSS2 synthesizes acetyl-CoA from acetate, which is derived from the media or deacetylation reactions. Glutamine can be used to synthesize citrate through reductive carboxylation in the mitochondria. Citrate can then be translocated into the nucleus to generate nucleus acetyl-CoA. OAA, oxaloacetate; ACLY, ATP-citrate lyase; PDC, pyruvate dehydrogenase complex; ACSS2, acetyl-CoA synthetase short chain family member 2; CAT, carnitine acetyltransferase; Ac, acetylation.
Fatty acids are also a bona fide source of carbon for histone acetylation, contributing up to 90% of certain histone acetylation markers, i.e., H3K9ac, in immortalized hepatocytes (
Cancer cells also synthesize acetyl-CoA from acetate by acetyl-CoA synthetase short chain family member 2 (ACSS2) (
For cancer cells that cannot undergo normal oxidative phosphorylation in the mitochondria, they use glutamine-dependent reductive carboxylation as the major pathway to generate citrate and acetyl-CoA (
DNA methylation and histone methylation are catalyzed by methyltransferases with SAM as the methyl donor. SAM is derived from combined activities of one-carbon metabolism and methionine metabolism through a vitamin-dependent metabolic cycle (
Metabolism of SAM. SAM is synthesized from methionine and ATP by methionine adenosyltransferase (MAT). In methylation reactions, SAM is sequentially converted to S-adenosylhomocysteine (SAH), homocysteine (Hcy) and methionine with 5-methyl-tetrahydrofolate (THF) or with betaine as the methyl donor. Serine-glycine metabolism provides one-carbon unit to the folate cycle. Serine biosynthesis is controlled by LKB-AMPK-mTOR pathway. In the methionine salvage pathway, SAM is converted to 5′-methylthioadenosine (MTA), which is salvaged back for SAM generation. MTA inhibits the activity of PRMT5. MAT, methionine adenosyltransferase; MTA, 5′-methylthioadenosine; THF, tetrafolate; LKB1, liver kinase B1; mTOR, mammalian target of rapamycin complex; SAH, S-adenosylhomocysteine; Hcy, homocysteine; MTAP, MTA phosphorylase; DMG, dimethylglycine; PRMT5, arginine methyltransferase 5.
S-adenosylmethionine can also be regenerated by methionine salvage pathway, where SAM is decarboxylated to form 5′-methylthioadenosine (MTA), which is then salvaged back to methionine and SAM (
Due to the tight connection between SAM availability and DNA and histone methylation, factors that perturb SAM levels or SAM/SAH ratio could determine DNA and histone methylation status (
DNA and histone methylation can be actively removed by demethylases. There are two major classes of demethylases: flavin adenine dinucleotide (FAD)-dependent LSD demethylases and α-KG-dependent JmjC family demethylases. The LSD family of histone demethylases (LSD1 and LSD2) use oxygen to remove methyl groups from mono- or dimethylated histones in a FAD-dependent manner. JmjC demethylases use oxygen and α-KG as substrates, producing succinate and CO2. JmjC demethylases include a diverse family of enzymes responsible for histone demethylation, DNA 5-methyl-cytosine hydroxylation, RNA
α-KG is either generated as an intermediary metabolite of the TCA cycle or produced by transamination of glutamate derived from glutamine (
Metabolism of α-KG. α-KG is synthesized from the TCA cycle or transamination of glutamate derived from glutamine. α-KG can be metabolized to succinate and fumarate in the TCA cycle. Succinate and fumarate are competitive inhibitors of α-KG-dependent demethylases. α-KG can also be converted to 2-HG by mutated isocitrate dehydrogenase 1 and 2 (IDH1 and IDH2). 2-HG is a competitive inhibitor of α-KG-dependent demethylases. α-KG, α-ketoglutarate; 2-HG, 2-hydroxyglutarate; Mut IDH1/2, mutated isocitrate dehydrogenase 1 and 2.
There are four classes of histone deacetylases (HDAC classes I, II, III, and IV) that remove acetyl moieties from histone lysine residues (
Metabolism of NAD+. Sirtuins consume NAD+ and produce nicotinamide (NAM). NAM is recycled to produce NAD+ by the NAD+ salvage pathway. In the NAD+ salvage pathway, NAM is converted to nicotinamide mononucleotide (NMN) by nicotinamide phosphoribosyltransferase (NAMPT), and NMN is then converted to NAD+ by nicotinamide mononucleotide adenylyltransferases (NMNATs). AMPK is required for NAMPT expression and thus controls intracellular NAD+. NAM, nicotinamide; NAMPT, nicotinamide phosphoribosyltransferase; NMNATs, nicotinamide mononucleotide adenylyltransferases; PRPP, phosphoribosyl pyrophosphate.
During glycolysis, NAD+ is converted to NADH, leading to reduced NAD+/NADH ratio and downregulated activity of sirtuins. Therefore, it is possible that increased aerobic glycolysis in cancer may reduce the activity of sirtuins by decreasing NAD+/NADH ratio, leading to histone hyperacetylation, decondensed chromatin structure and dysregulation of gene expression (
There are three NAD+-producing NMNATs (NMNAT1, NMNAT2, and NMNAT3) with distinct subcellular localizations. NMNAT1 exists in the nucleus and is responsible for nuclear production of NAD+; NMNAT2 and NMNAT3 are localized in the Golgi complex and mitochondria, respectively (
Glycolysis provides several intermediary metabolites that regulate chromatin modifications. In addition to providing acetyl-CoA for histone acetylation, glycolysis enhances histone acetylation by inhibiting the reverse process. The product of glycolysis, pyruvate acts as an inhibitor for histone deacetylase 1 and 3 (HDAC1/3) and thus promotes histone acetylation (
Glycolysis is required for pyruvate kinase-mediated histone H3T11 phosphorylation and glycolytic metabolites fructose 1, 6-biphosphate (FBP) and phosphoenolpyruvate (PEP) function as the cofactor and substrate, respectively (
Another extensively studied metabolism that regulates chromatin modifications is the tricarboxylic acid (TCA) cycle. As described earlier, the TCA intermediary metabolite α-KG serves as the substrate for JmjC demethylases. Other TCA intermediary metabolites especially for those structurally related metabolites including succinate and fumarate are two competitive inhibitors of α-KG-dependent demethylases (
The oncometabolite D-2-hydroxyglutarate (D2-HG) is another competitive inhibitor of α-KG-dependent enzymes. Due to its structural similarity to α-KG, D2-HG outcompetes α-KG for binding to histone demethylases, i.e., JHDM. D2-HG is typically maintained at low levels in normal cells but is significantly elevated in tumor cells, i.e., glioma and melanoma (
Fatty acid oxidation is an important source for acetyl-CoA and histone acetylation. As
Recent studies showed that fatty acid metabolism regulates histone acetylation by modulating histone deacetylation. For example, several long-chain free fatty acids including myristic, oleic and linoleic acids have been shown to bind to SIRT6 and induce up to a 35-fold increase of its activity toward H3K9ac (
Histone methylation is regulated by lipid metabolism (
Recent studies showed that metabolic enzymes also directly regulate chromatin modifications independent of their produced metabolites, which is one of non-metabolic functions of metabolic enzymes (
Some metabolic enzymes can phosphorylate non-histone proteins. An unbiased quantitative phosphoproteomic approach revealed that PKM2 can phosphorylate a total of 974 proteins, including mammalian target of rapamycin (mTOR) inhibitor AKT1 substrate 1 (AKT1S1) (
Certain glycolytic enzymes regulate histone gene expression or histone cleavage. Nuclear glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and lactate dehydrogenase (LDH) promote histone
The demethylation reaction is a redox reaction and it has been well established that histone demethylation is regulated by oxygen availability. Hypoxia has been reported to increase H3K4me3 in mammalian cells by inhibiting the activity of the oxygen-dependent H3K4 demethylase JARID1A (
To regulate the activity of epigenetic modifiers, the concentration of metabolites as substrates should be well above the enzymatic Km, which refers to the substrate concentration that produces half-maximal velocity and is used to measure the binding affinity of enzymes to their substrates. In fact, for most chromatin modifiers with the exception of protein kinases, their measured Km is within the range of physiological concentrations of metabolites. Therefore, their enzymatic activity could be regulated by the availability of substrates and cofactors derived from the metabolic pathways. However, accumulating evidence showed that perturbations in metabolites availability only influence certain types of chromatin modifications. For example, increasing the cellular SAM levels by methionine and folate amendment specifically increased H3K4me2/me3 but not H3K79me3 (
Different chromatin modifying enzymes have different kinetic parameters, such as Km. This different Km implied that a change in substrate concentrations would differentially influence the enzymatic activity and there is a hierarchy of sensitivity to nutritional limitations. The Km of H3K4 methyltransferase MLL1 for SAM is 10.4 μM whereas that of H3K27 methyltransferase EZH2 is 1.64 μM, which makes MLL1-catalyzed H3K4me3 more sensitive to changes in the intracellular SAM levels than EZH2-catalyzed H3K27me3 (
Instead of causing global chromatin changes, most nutritional alterations affect chromatin modifications on specific locus, which cannot be explained by the differential kinetic properties of modifying enzymes. For example, elevated acetyl-CoA levels have been shown to increase histone acetylation only at a subset of genes, i.e., growth-promoting genes (
NAD+ is also produced in the same way to determine the specificity. The NAD+-producing enzymes GAPDH and LDH have been reported to translocate into the nucleus and interact with transcription factors and chromatin modifiers (
The epigenetic landscape plays a crucial role in cellular adaptation to changes in nutrient availability and utilization. But the remaining question is how nutrient changes are transduced to alterations in epigenetic modifications? In addition to intermediary metabolites, such as acetyl-CoA and SAM, which function as an indicator to reflect the cells’ potential to generate energy (
Phosphoinositide 3-kinase (PI3K)/Akt pathway is a critical signaling cascade in response to growth factor stimuli and reflects in acetyl-CoA and histone acetylation changes (
The distinct metabolic pathways in cancer cells and their connection to epigenetic modifications. Cancer cells have increased glucose and glutamine uptake, leading to accelerated glycolysis and biomass accumulation. PI3K/AKT is activated by growth factors to regulate glycolysis and acetyl-CoA generation. AKT induces the expression of GLUT1 to increase glucose uptake and phosphorylates HK1 and PFK to enhance aerobic glycolysis. AKT phosphorylates ACLY to increase acetyl-CoA production. Increased lactate dehydrogenase (LDH) activity and decreased pyruvate dehydrogenase (PDC) activity result in increased lactate export, attenuated TCA cycle and diversion of glycolysis to pentose phosphate pathway. ACLY, ACSS2 and PDC contribute to nuclear acetyl-CoA production and subsequent histone acetylation. Glycolysis is increased in cancer cells with NAD+ converted to NADH, leading to reduced NAD+/NADH ratio and downregulated activity of sirtuins, which results in histone hyperacetylation and dysregulation of gene expression. Serine and one-carbon metabolism is also accelerated in cancer cells to produce SAM, which is controlled by LKB-AMPK-mTOR pathway. α-KG is primarily produced by transamination of glutamate derived from glutamine in cancer cells. Mutations of IDH1/2 lead to accumulation of 2-HG, which increases histone methylation by inhibiting α-KG dependent enzymes. The potential targets for anti-cancer therapy were labeled in red color. ACSS2, acetyl-CoA synthetase short chain family member 2; AMPK, AMP-activated protein kinase; F-6-P, fructose-6-phosphate; FBP, fructose-1,6-biphosphate; GA3P, glyceraldehyde-3-phosphate; G-6-P, glucose-6-phosphate; Gln, glutamine; Glu, glutamate; HK1/2, hexokinase 1/2; LDHA, lactate dehydrogenase; MCT1, monocarboxylate transporter 1; MCT4, monocarboxylate transporter 4; MDH1, malate dehydrogenase 1; Mut IDH1/2, mutated isocitrate dehydrogenase 1 and 2; MTA, 5′-methylthioadenosine; MTAP, MTA phosphorylase; NAM, nicotinamide; NAMPT, nicotinamide phosphoribosyltransferase; NMNATs, nicotinamide mononucleotide adenylyltransferases; PCK1, phosphoenolpyruvate carboxykinase 1; PDC, pyruvate dehydrogenase complex; PEP, phosphoenolpyruvate; PFK, phosphofructokinase; PI3K, phosphoinositide 3-kinase; PKM2, pyruvate kinase M2; PRMT5, arginine methyltransferase 5; PRPP, phosphoribosyl pyrophosphate; α-KG, α-ketoglutarate; 2-HG, 2-hydroxyglutarate; 3-P-G, 3-phosphoglycerate; ACLY, ATP-citrate lyase.
AMP-activated protein kinase (AMPK) is an AMP-sensitive protein kinase that functions as an energy sensor to regulate mitochondrial biogenesis in cells (
The mammalian target of rapamycin complex (mTORC) participates in signal transduction pathways that transduce growth factor signals and nutrients signals to transcription and translational control, thus determining cell growth and proliferation status (
A recent report showed that mTORC2 but not mTORC1 signaling pathway regulates histone acetylation H3K56ac (
It is noteworthy that there is a bidirectional relationship between epigenetic modifications and metabolism. On one hand, cell metabolites and metabolic enzymes modulate epigenetic modifications; on the other hand, epigenetic changes at metabolic genes regulate the transcription of genes involved in metabolism, which eventually affects cell metabolism (
The bidirectional regulation between metabolism and epigenetic modifications could lead to a feedback control of cell metabolism: intracellular metabolism perturbations change epigenetic modifications at metabolic genes, which influences the transcription of these genes and metabolic pathways. The feedback regulation of metabolism could enable cells to respond to changes in microenvironment in a prompt and accurate way. Glucose metabolism has been shown to stimulate pyruvate kinase 1 (Pyk1)-catalyzed H3T11 phosphorylation, which represses the transcription of
Another way to regulate cell metabolism is modifying metabolic enzymes, which may affect their activity, stability and/or subcellular localization.
Acetylation is an important way to control the activity of many metabolic enzymes. For example, a number of mitochondrial proteins have been reported to be inactivated by acetylation to suppress mitochondrial functions (
Acetylation also controls the stability of some metabolic enzymes. ACLY is deacetylated by SIRT2 to become unstable (
Acetylation promotes the translocation of several glycolytic enzymes to the nucleus where they function as transcriptional regulators. For example, PKM2 is acetylated by p300, which promotes its translocation into the nucleus and contributes to tumor cell proliferation and tumorigenesis (
Metabolic enzymes also undergo other modifications, i.e., phosphorylation, acylation, etc. Upon DNA damage, nuclear ACLY is phosphorylated, which enhances its ability to synthesize the nuclear acetyl-CoA pool and increases histone acetylation required for efficient double-strand break repair by homologous recombination (
In addition to modulate the activity of metabolic enzymes and the expression of metabolic genes, epigenetic modifications also serve as the storage for metabolites. The typical example is recycling the acetyl group from acetylated proteins in the form of acetate by class I and II HDACs. ACSS2 then synthesizes acetyl-CoA from acetate. Based on the potential acetylation sites, yeast histones can store up to 65-fold more acetyl groups than acetyl-CoA and mammalian proteins can store ∼100-fold more acetyl moieties than free acetyl-CoA (
Histone demethylation results in hydroxylation of the enzymatic substrate to generate formaldehyde, which is an endogenous protein and DNA cross-linking agent. Formaldehyde can be detoxified by converting to formate, which then functions as one-carbon unit to fuel nucleotide biosynthesis (
Cancer cells have distinct metabolic pathways and epigenetic landscapes with their normal counterparts, which contribute to tumorigenesis (
Although tremendous progress has been made in understanding the connection between cancer metabolism and epigenetics, there are several open and outstanding questions need to be addressed. Firstly, the list of metabolic enzymes present in the nucleus is continually expanding. Understanding their roles in the nucleus is critical to elucidate the connection between metabolism and epigenetic regulation. However, since many enzymes lack a canonical nuclear localization sequence (NLS), it remains unclear how they enter into the nucleus. Moreover, many nuclear metabolic enzymes function within a complex via interaction with other proteins. Thus, characterizing their interaction partners in the nucleus will help us better understand their non-metabolic functions. Secondly, numerous transcriptomic studies showed that the effect of metabolites on gene transcription was specific rather than global. For example,
XY and SL conceptualized the study and wrote, reviewed, and edited the manuscript. XY, RM, YW, YZ, and SL wrote the draft.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.