Keywords

1 Introduction

The theory put forward by Otto Warburg many years ago is that the faster growth rate of cancer cells than that of their normal counterparts is largely because cancer cells tend to drive glycolysis followed by lactate fermentation in the cytosol as energy source rather than oxidation of pyruvate fed into mitochondria, even in the presence of oxygen. This claim is now known as Warburg Effect (Wang et al. 2009; Warburg 1956). After more than eight decades, discovery and characterization of specific metabolic changes in cancer cells continue to progress rapidly and reprogramming energy metabolism is considered an emerging hallmark of cancers (Hanahan and Weinberg 2011). During cancer development, in order to meet the excessive demand of cellular growth, cancer cells may adapt a serial of strategic alterations regarding energy metabolism. For instance, glucose, as a ubiquitous energy source in most organisms, can supply carbon, oxygen, hydrogen for anabolic processes, and its uptake is frequently enhanced in cancer cells. Glutamine, a non-essential amino acid, serves as carbon source for energy production, conduces to biosynthetic reaction by supplying carbon and nitrogen, regulates redox homeostasis through synthesis of glutathione (Dang 2010). Increased glutaminolysis is emerged as a novel feature of the metabolic profile of cancer cells, along with increased aerobic glycolysis. Another energy resource, fatty acid, which is seldom appreciated before but under the spotlight now, is required for energy storage, membrane synthesis and generation of signaling molecules (Currie et al. 2013).

Today, mutations in oncogenes and tumor suppressors are known to be answer for the malignant transformation. Furthermore, Warburg Effect is generally accepted as the result of these mutations, rather than a cause. Several famous oncogenes and tumor suppressors are conversantly participated in modifying glycolysis, glutaminolysis, lipid metabolism and various biosynthetic pathways. For instance, PI3K pathway/hypoxia/c-Myc promotes glucose uptake (Cairns et al. 2011). AMP-activated protein kinase (AMPK) regulates glucose switch through mTOR pathway (Cairns et al. 2011). Oncogenic Kras enhances glucose import at gene expression and protein levels as well as regulates glutamine metabolism in tissue-specific PDAC model (Ying et al. 2012; Son et al. 2013). Both c-Myc and p53 can manipulate glutaminolysis through glutaminase1 or glutaminase2 (Gao et al. 2009; Hu et al. 2010; Suzuki et al. 2010). Hypoxia can mediate lipogenesis by IDH1 through reductive glutamine metabolism (Metallo et al. 2011).

In the meantime, nearly two decades ago, microRNAs (miRNAs), which are approximately 22 nucleotides long, were first discovered in Caenorhabditis elegans (Lee et al. 1993; Reinhart et al. 2000). They are encoded by miRNA genes which are transcribed by RNA polymerase II or III, then forming the primary miRNA in the nucleus, the primary miRNA is processed by Drosha and DGCR8, resulting in pre-miRNA, then it is exported from nucleus into cytoplasm by exportin 5 and reprocess by Dicer. Dicer is a RNAse III enzyme that processes the pre-miRNA to a 22 nucleotides miRNA. Presumably, miRNAs regulate a third of human genes through base pairing with regions within the 3′ untranslated region of the target mRNA (Cannell et al. 2008). There is abundant evidence on the multitudinous roles of miRNAs in biological processes, covering proliferation, differentiation, aging, apoptosis, metastasis, signal transduction and so on (Babashah and Soleimani 2011).

Evidence is emerging that miRNAs are involved in multiple cancer development process including cancer metabolism. Whether we can cut off the specific energy supply for cancer cell growth by restoring normal aerobic metabolism, and how miRNAs involved in this alteration, are among the fiercely investigated topics for cancer researchers. Here, we will focus on the special roles of miRNAs played on tumor metabolism, including glucose, glutamine and lipid metabolism.

2 MicroRNAs Regulate Glucose Metabolism

2.1 MiRNAs and Glucose Homeostasis

As summarized in Table 4.1, miRNAs are shown to regulate insulin secretion, biosynthesis and sensitivity to maintain the glucose homeostasis. Lin28/let-7 axis is demonstrated as the central regulator of mammalian glucose metabolism in mouse models (Zhu et al. 2011). Let-7 is the well-known tumor suppressor, whose biogenesis blocked by the RNA-binding protein Lin28. In mice, lin28 and let-7 play the opposite role in glucose metabolism through insulin-PI3K-mTOR pathway. Global knockdown of the let-7 family with an antimiR was sufficient to prevent and treat impaired glucose tolerance in mice with diet-induced obesity (Frost and Olson 2011). Let-7 and let-7-targets such as c-Myc and Kras all have significant effects on cancer metabolism, hence, the regulation of glucose homeostasis by lin28/let-7 axis may have profound implications in cancer metabolism (Zhu et al. 2010).

Table 4.1 MicroRNA regulations in glucose and glutamine metabolism
Full size table

The expression of miR-143 is up-regulated in the liver of genetic and dietary mouse model of obesity (Jordan et al. 2011). Highly expressed miR-143 inhibits insulin-stimulated AKT activation and impairs glucose metabolism by directly targeting oxysterol-binding-protein-related protein (ORP). In the regulation of glucose homeostasis, the pancreatic specific miRNA miR-375 (Kloosterman et al. 2007) has a profound effect. miR-375 as well as miR-124 and let-7b regulate insulin exocytosis in the pancreatic beta cell by suppressing myotrophin (MTPN) which helps secretary granules to exocytotic sites (Gauthier and Wollheim 2006). Recent study showed that miR-375 directly targets 3′-phosphoinositidedependent protein kinase-1(PDPK1), and significantly impairs the activity of PDPK1 (El Ouaamari et al. 2008), a master kinase related to human disease such as diabetes and cancer.

2.2 MiRNAs and Glycolysis

It is well known that the most fundamental metabolism alteration in tumor cells is increased levels of glucose uptake and glycolysis (also known as Warburg Effect). Glucose can’t pass across the hydrophobic cell membrane freely. They need a transport, glucose transporter proteins, also called solute carrier family 2 (GLUT). The level of glucose uptake can be affected by the activity of GLUT in cell membranes. GLUT3 is inhibited by miR-195-5p (Fei et al. 2012), which directly targets GLUT3 3′-untranslated region in human bladder cancer T24 cell lines. Via this molecular mechanism, miR-195-5p suppresses proliferation and promotes apoptosis of cancer cells in vitro. GLUT4, another transporter of GLUT family, can be up-regulated by miR-223 (Lu et al. 2010) and miR-133 (Horie et al. 2009) in rat cardiomyocytes.

MiRNAs regulate glycolysis by directly targeting key metabolic enzymes. Hexokiase 2 (HK2), the first enzyme of glycolysis that phosphorylates glucose to produce glucose 6-phosphate, was identified as a target of miR-143 (Fang et al. 2012). MiR-143 was shown to down-regulate HK2 protein level and inhibit glucose metabolism in lung cancer. This phenomenon was also confirmed in breast cancer. Further it was found that pro-inflammatory cytokines promote glycolysis mediated by miR-155, which up-regulates HK2 in two cascades of miR-155-SOCS1-STAT3-HK2 and miR-155-C/EBPβ-miR-143-HK2 (Jiang et al. 2012). Phosphogulcose isomerase (PGI) is a housekeeping cytosolic enzyme that brings the interconversion of glucose-6-phoshpate and fructose-6-phosphate, playing an important role in glycolysis and gluconeogenesis. A decrease of PGI expression leads to over-expression of miR-200 (Ahmad et al. 2011), which is associated with reversal of mesenchymal-epithelial transition (EMT) phenotype in human breast cancer cells (Davalos et al. 2012). MiR-15a/16-1 cluster, the first tumor suppressor miRNAs described in human malignancies (Calin et al. 2002), are reported to target the enzyme Aldo A which catalyzes a reversible aldol reaction in glycolysis (Calin et al. 2008). MiR-122 is the most abundantly expressed miRNA in the liver, and may be involved in lipid and cholesterol metabolism. While PTP1B is a newly identified direct target of miR-122, it is a negative regulator of insulin signaling cascade (Yang et al. 2012). And miR-122 is also predicted to target aldo A (Chen et al. 2012). Thus, these miRNAs are reported to have roles in cancer glycolysis.

2.3 MiRNAs and Glucose Stress

The change of miRNA expression is also a cause of abnormal glucose condition and subsequent cellular functions. In mesenchymal stem cells, hyperglycemic condition lead to a decrease of miR-32-5p, which promotes cell cycle by targeting tensin homologs deleted on chromosome 10 (PTEN) (Zhu et al. 2013). MiR-30a-5p over-expression induce beta cell dysfunction in vitro, which can be induced by high glucose at the concentration of 33.3 mmol/l (Kim et al. 2013). Another glucose induced miRNA, miR-29c, is known as a tumor suppressor which induces cell apoptosis, increases extracellular matrix protein accumulation and activates Rho kinase in mouse diabetic nephropathy model (Long et al. 2011). In glioma cells, a single miRNA was shown to adapt the cells to different glucose conditions. Glucose induced miR-451 promotes cell growth. While decreased miR-451 expression in low glucose slows proliferation, it enhances migration and survival (Godlewski et al. 2010). Deprivation of glucose leads to a change of miRNAs such as miR-466h-5p (Druz et al. 2012). Cultivation of mouse cell line with glucose-free medium causes the oxidative stress. The inhibition of histone deacetylases (HDACs) activity by this oxidative stress increases acetylation in miR-466h-5p promoter region, which lead to the activation of this miRNA.

3 MicroRNAs Modulate Lipid Metabolism

Alteration and its impact of lipid metabolism in cancer cells have long been ignored compared to the intensive interest in two other star nutrients, glucose and glutamine. This sounds a bit strange considering the pivotal roles of fatty acid in energy storage, cell proliferation and generation of membrane signal. However, last few years have witnessed a rapid development of the field and now a growing body of research is focused on lipid metabolism. Herein, we will summarize the role of several miRNAs that are involved in regulation of lipid homeostasis (Fig. 4.1).

Fig. 4.1
figure 1

MicroRNA-mediated regulation of lipid metabolism. Many miRNAs are involved in regulation of the expression of target coding genes essential for fat acid synthesis and degradation. Focus of this figure is on the vital roles of miR-122, miR-33a/b and miRNAs related with PPARs in lipid metabolism

Full size image

One of the most notable miRNAs involved in lipid metabolism is liver-specific miR-122, which accounts for approximately 70 % of total miRNAs in the liver. In addition to its known roles in regulating HCV/HBV (hepatitis C/B virus) replication and expression, hepatocyte differentiation and glucose metabolism (Fukuhara et al. 2012; Qiu et al. 2010; Hu et al. 2012), miR-122 is the first miRNA to be linked recently to fat and cholesterol metabolism control. Two studies demonstrated the fundament roles of miR-122 in regulating lipid metabolism by 2′-O-methoxyethyl (2′-MOE) phosphorothioate-modified antisense oligonucleotide (ASO) in vivo (Esau et al. 2006; Fabani and Gait 2007). More recently, two independent groups discovered the tumor suppressor role of miR-122 in mouse liver using genetic deletion technology (Hsu et al. 2012; Tsai et al. 2012). Hsu et al. utilized mice which is germline knockout (KO) or liver-specific knockout (LKO) of miR-122 to describe the characteristics of fatty acid metabolism in vivo. They found that in LKO mice, the hepatic lipid level, especially triglyceride (TG) accumulation, was increased by up-regulation of Agpat1 and Mogat1 which are responsible for TG biosynthesis. But the secretion of TG into serum is reduced, thus resulting in hepatic microsteatosis in early adult life of LKO mice. Moreover, many genes involved in development, cellular proliferation, death and cancer, such as, c-Myc, c-Jun, CCND1, Igf2, Epcam, Rhoa, are aberrantly expressed in addition to the variation of lipid metabolism genes. In KO mice, increased monocytes and neutrophils are recruited to livers, leading to inflammation, the major factor attributed to malignant transformation of hepatocellular carcinoma. Production of pro-tumorigenic cytokines including IL-6 and TNF-α are also increased in LKO and KO mice which possess higher HCC incidence along with aging (Hsu et al. 2012). Tsai et al. also described the similar tumor suppressor role of miR-122 using genetic deletion in mice. They demonstrated that serum HDL and VLDL were significantly reduced in the miR-122−/− mice, which, on the contrary, causes liver steatosis. In addition, miR-122 can control hepatic fibrogenesis through one target named Krüppel-like factor 6 (KLF6) that is expressed mainly in hepatic carcinoma, whereas its expression is much lower in the normal hepatocytes of WT livers (Tsai et al. 2012). As a whole, these studies represented endeavors to decipher and manipulate the vital roles of miR-122 in regulating liver homeostasis using miR-122 antisense oligonucleotides or using an loss-of-function model in vivo (Wen and Friedman 2012).

SREBP, abbreviation of sterol regulatory element binding transcription factor, includes two protein named SREBP1 and SREBP2 whose vital function is played in sterol biosynthesis and cholesterol homeostasis by regulating transcription of sterol-regulated genes. Known miRNAs that are related with SREBP include miR-33a/b. MiR-33a is encoded in intron 16 of SREBP2 gene on chromosome 22, whereas miR-33b is encoded in intron 17 of SREBP1 gene on chromosome 17, respectively (Najafi-Shoushtari et al. 2010). The two mature miRNAs only differ in two nucleotides, but can regulate plentiful overlapping target genes, such as ATP-binding cassette transporter sub-family A member 1 (ABCA1), which is a cholesterol transporter and can mediate cholesterol efflux from within the cell to lipid-free apolipoprotein A1 (APOA-I). When ABCA1 is decreased because of elevated miR-33, then high density lipoprotein (HDL) cholesterol levels formation is destroyed, so the incidence of cardiometabolic diseases increased (Najafi-Shoushtari et al. 2010; Horie et al. 2010; Rayner et al. 2010). In mouse macrophages, miR-33 also targets ATP-binding cassette, sub-family G (white), member 1 (ABCG1), reducing cholesterol efflux to nascent HDL (Rayner et al. 2010; Marquart et al. 2010). Meantime, miR-33a and miR-33b also contribute to the regulation of several proteins involved in fatty acid β-oxidation, modulating the expression of carnitine O-octanoyltransferase (CROT), carnitine palmitoyltransferase 1A (CPT1A), and hydroxyacyl-CoA dehydrogenase-3-ketoacyl-CoA thiolase-enoyl-CoA hydratase (trifunctional protein) β-subunit (HADHB). CROT, CPT1A, and HADHB regulate the transport and degradation of fatty acids in the mitochondria (Rayner et al. 2011; Davalos et al. 2011).

Nevertheless, Tae-Il Jeon and colleagues claimed that additional miRNAs may be more robustly activated by SREBPs than miR-33 to regulate intracellular cholesterol, due to the fact that intracellular cholesterol levels are tightly controlled (Jeon et al. 2013). They performed a genome-wide analysis searching for miRNAs that are differentially expressed in livers of mice through controlled dietary, normal diet supplemented with excess cholesterol versus another diet with a combination of lovastatin plus ezetimibe (LE), drugs inhibit both endogenous cholesterol synthesis and dietary absorption of cholesterol. Intriguing finding is that miR-96/182/183 cluster comes to light. They are expressed from a unique primary transcript and the promoter for this locus is a direct target regulated by SREBP2 in mice. In this paper, miR-96 and miR-182 negatively regulate the expression of INSIG-2 and FBXW7 by targeting their 3′UTR, respectively. INSIG-2 and FBXW7 are two proteins that weaken the maturation of SREBP2, especially nuclear SREBP2 which plays the role of a transcription factor. Further analyses revealed that, when mice are fed with diet with cholesterol, normal, LE ingredient, miR-182 presents the gradient increase tendency because of the different density of cholesterol. Inversely, FBXW7 will decrease and result in accumulation of nuclear SREBP2 in LE group. This phenomenon is also conserved in human cells. So it demonstrates the coordinate and reciprocal regulation of nuclear SREBP2 with miR-182 and FBXW7. Also, these miRNAs can modulate synthesis of lipid through regulate nuclear SREBP.

Peroxisome-proliferator-activated receptors (PPARs) are nuclear hormone receptors that exert a transcriptional activity regulating genes involved in cell proliferation, cell differentiation, energy homeostasis, diabetes, obesity, atherosclerosis, and other basic metabolically processes (Peyrou et al. 2012). Three major isoforms are known: PPAR-alpha, PPAR-beta/delta, PPAR-gamma. At least five different miRNAs have been reported to regulate PPARs expression in liver cells directly or indirectly. Such as miR-10b, miR-21/miR-122 can directly regulate PPARα/PPARβ(δ) respectively. MiR-27b/miR-132 modulate PPARα/PPARγ indirectly (Zheng et al. 2010; Kida et al. 2011; Gatfield et al. 2009; Mann et al. 2010). The performance of relationship between these miRNAs and PPARs is embodied in steatosis, cholesterol and lipid metabolism. Recently, Alessio Papi et al. found that, when mammospheres derived from breast cancer stem cell are exposed with breast tumor fibroblasts supernatant, enhanced autocrine tumor necrosis factor-a (TNFα) lead to functional interplay between peroxisome proliferator activated receptor-α and hypoxia inducible factor-1α (PPARα/HIF1α), two nuclear transcriptional factors. But high PPARγ expression will antagonize the PPARα/HIF1α interplay. Further, they demonstrated that siHIF1α and siPPARα can decrease miR-130b expression, and pre-miR-130b in turn facilitates PPARα expression and decrease PPARγ. Another miRNA, miR17-5p, down-regulates PPARα, whereas it increases PPARγ expression. Through this regulatory network, they suggest that the antagonist interplay between PPARα and PPARγ is mediated by miR130b and miR17-5p. Next, they pinpoint that apolipoprotein E (ApoE) is over-expressed in mammospheres and PPARα over-expression can induce ApoE expression at mRNA and protein levels. Hence, PPARα/HIF1α interplay can regulate lipid homeostasis through control of ApoE expression (Zissel et al. 2013).

4 MicroRNAs Regulate Glutamine Metabolism

Since most cancer cells rely on aerobic glycolysis for proliferation and metabolism, fewer metabolites produced from glycolysis enter the tricarboxylic acid (TCA) cycle for energy supply. Thus, other alternative metabolites such as glutamine will play such a vital role as feeding the TCA cycle and redox homeostasis, not only as a nitrogen donor (Table 4.1). The c-Myc transcription factor, which is well known to regulate cell proliferation and glucose metabolism, has been validated to stimulate glutamine catabolism by repressing miR-23a/b that target glutaminase 1 (GLS1), resulting in an increased uptake and catabolism of glutamine (Gao et al. 2009). GLS1 catalyzes the conversion of glutamine to glutamate, the latter is further catabolized to α-ketoglutarate, which serves as a TCA substrate. Moreover, glutamate can also be converted to proline in metabolic process. C-Myc not only increases GLS, but also induces proline biosynthesis from glutamine. In 2012, Liu et al. found that c-Myc robustly suppresses the expression of POX/PRODH protein, the first enzyme in proline catabolism, primarily through increasing miR-23b*, which is processed from the same transcript as miR-23b (Liu et al. 2012).

In recent years, energy metabolism has been reported to play a major role in somatic reprogramming. TDH-mediated threonine catabolism could stimulate reprogramming process in mouse embryonic fibroblast (Han et al. 2013). In Drosophila, bioinformatics screen reveals that miR-277 acts as a metabolic switch in modulating amino acid catabolism. Hence, the regulation of amino acid metabolism by miRNAs is emerging to exert significant impact in cancer biology and beyond (Stark et al. 2003).

5 Summary and Future Perspectives

As we have discussed in this chapter, a plethora of studies now focus on the diverse roles of miRNAs played on tumor metabolism including glucose, glutamine and lipid metabolism with an aim to understand and combat cancers. Given that cancer cells adapt specific strategies for energy metabolism, one might expect to explore the regulatory mechanisms of cancer specific metabolism by miRNAs for cancer therapy. Indeed, restoration of miR-122 constitutes a novel approach that could be beneficial for therapy of chronic liver diseases and HCCs that express low levels of miR-122. Moreover, high miR-122 expression abolishes hepatic insulin resistance, resulting in lower incidence of diabetes. Nevertheless, the actual roles and regulatory networks of miRNAs in the real biological process of development and diseases are far more complicated than previously thought. In cancer metabolism, substantial cross-talks exist among the functions of miRNAs, oncogenes, nutrients enzymes and metabolites, dissection of which will greatly enhance our understanding of the complex process underlying human malignancies as well as provide insight for cancer therapy.