Keywords

1 Introduction

The liver maintains blood glucose and lipids at physiological levels. As such, it is central to the metabolic flexibility of the body to utilize appropriate energy substrates at a given time of the day. The ability to switch between glucose and fatty acid utilization during physiological feeding/fasting is achieved by several hormonal signaling pathways, notably insulin and glucagon. De novo lipogenesis, occurring primarily in the liver, plays a key role in maintaining metabolic flexibility, as it converts unused glucose at the fed state to fatty acids for storage in adipose tissue that are released at the fasted state as a main energy source. Not surprisingly, dysregulation in liver lipogenic pathway often leads to hepatic and systemic metabolic dysfunction, as demonstrated in mouse genetic models . In this chapter, we will discuss regulatory mechanisms governing de novo lipogenesis in the liver with an emphasis on transcriptional regulation. The first layer of control (in addition to insulin) is mediated by pre-programmed, rhythmic mRNA expression of lipogenic genes regulated by the circadian clock transcription factors . A second layer is mediated by several nuclear hormone receptors that are downstream of the molecular clock and activated by dietary lipids, thereby integrating nutrient status with energy metabolism. We will also summarize metabolic phenotypes in mice with loss- or gain-of-function studies of lipogenic genes. These mouse genetic models raise the possibility of liver-derived lipid metabolites that serve as mediators for crosstalk between the liver and other peripheral tissues in a coordinated effort to maintain metabolic homeostasis. Two examples of such lipid metabolites will be discussed.

2 Metabolic Rhythm of Hepatic Lipogenesis

2.1 Physiology of Feeding/Fasting Responses

As the first organ flooded with dietary nutrients carried by the portal vein, the liver is key in controlling energy substrate homeostasis. A central mechanism in the process is hepatic de novo lipogenesis, whereby energy from dietary sources can be efficiently repackaged. Following a meal, elevation in glycemia stimulates insulin secretion by pancreatic beta cells. Energy homeostasis is maintained in part by insulin’s direct action on the liver, including inhibition of hepatic gluconeogenesis and glycogenolysis and stimulation of glucose storage in the form of glycogen. In addition, insulin induces de novo lipogenesis, allowing for the transfer of the energy contained in excess dietary carbohydrates toward exportable and energy dense lipids.

In short, the canonical signaling pathway of insulin commences with its binding to the insulin receptor (IR) , a cell surface receptor tyrosine kinase. Subsequent autophosphorylation and activation of IR follows and initiates a cascade of phosphorylation events through the insulin receptor substrate (IRS) family of proteins 1–6, which then act through the phosphatidylinositol-3-kinase (PI3K)/Protein kinase B (PKB or AKT) signaling pathway to regulate metabolism (Boucher et al. 2014). The action of insulin triggers many metabolic pathways downstream of AKT. For instance, AKT inhibits Tuberous Sclerosis Complex (TSC) 1/2, which leads to activation of Sterol regulatory element-binding transcription factor (SREBP) 1 and up-regulation of lipogenic gene expression through the mammalian target of rapamycin complex 1 (mTORC1) (Yecies et al. 2011).

Storage of glucose as glycogen is limited and the liver converts excess glucose to fatty acids through de novo lipogenesis. In the liver, glucose exists as glucose-6-phosphate that is processed to pyruvate through glycolysis. Pyruvate decarboxylation generates acetyl coenzyme A (acetyl-CoA), a substrate for acetyl-CoA carboxylase α and β (ACACA/B or Acc1/2) to synthesize malonyl-CoA, which enters the committed step for lipogenesis by fatty acid synthase (FASN). Additionally, malonyl-CoA accumulation inhibits hepatic carnitine palmitoyltransferase I (CPT1), a rate-limiting enzyme regulating fatty acids transport to the mitochondria for β oxidation. In addition, dietary fatty acids taken-up by the liver are re-esterified to form triglycerides in the liver and exported as very low-density lipoproteins (VLDLs) for peripheral use and storage. At the postprandial state, insulin levels drop, allowing the release of fatty acids from adipose tissue to be used as a primary energy substrate. The action of glucagon also increases hepatic gluconeogenesis to maintain blood glucose concentrations.

In addition to hormonal and substrate-driven metabolic regulation, transcriptomic studies have illustrated the presence of circadian control in metabolism. In the liver, ~15 % of genes transcripts oscillate rhythmically (Panda et al. 2002; Vollmers et al. 2009), many of which play a role in glucose and lipid metabolism. This anticipatory metabolic pre-programming is thought to promote metabolic efficiency and is coordinately controlled by several “molecular clocks ”. This will be discussed in detail below.

2.2 Circadian Regulation of Liver Metabolism

The observation of endogenous biological rhythms dates back to the 1700s, when French scientist Jean-Jacques d’Ortous De Mairan discovered that the leaf movements of a plant retained a periodicity of 24-h in complete darkness. This suggested that cycling biological phenomena could be not only a simple response to external stimuli, but also a consequence of autonomous internal clocks (Somers 1999). Much later in the 1950s, studies in Drosophila by Pittendrigh and in humans by Aschoff, suggested the existence of innate circadian rhythms (Aschoff 1960; Pittendrigh 1954). Throughout the years, the reliance of systemic metabolic homeostasis on the precision and coordination of the molecular processes within tissues and cells has become increasingly evident.

The master regulator of the mammalian circadian clock is expressed within the pacemaker neurons of the supra chiasmic nucleus (SCN) in the hypothalamus (Inouye and Kawamura 1979; Stephan and Zucker 1972). The rhythm is entrained by daylight through neuronal connection to photoreceptors in the retina. This establishes a close functional coupling of the light and dark cycle. The SCN in turn, installs a “standard time” for a number of basic physiological functions such as alertness, blood pressure, or body temperature. Peripheral organs, although synchronized under the influence of the brain central clock, possess their own circadian machinery to control organ-specific cellular processes. With regard to the liver, the brain clock directly influences feeding and fasting, which then entrains the hepatic clock to control lipid and glucose metabolism (Yamazaki et al. 2000; Yoo et al. 2004).

2.3 The Molecular Clock

At the core of the circadian clock machinery is a well-synchronized transcription-translation feedback loop generated by transcription factors aryl hydrocarbon receptor nuclear translocator-like (ARNTL or BMAL1) and circadian locomotor output cycles kaput (CLOCK), together with co-regulators Period (PER1, PER2, and PER3) and Cryptochrome (CRY1 and CRY2) (Asher and Schibler 2011; Bass and Takahashi 2010; Zhang and Kay 2010). The BMAL:CLOCK heterodimer binds to the E-box response element in the promoter of its target genes, which include the clock repressors, PER and CRY, and genes involved in the regulation glucose and lipid metabolism (Fig. 11.1). At the peak of BMAL:CLOCK activity, PER and CRY begin to accumulate, forming an inhibitory complex that binds and suppresses BMAL:CLOCK activity. The cycle is then completed by releasing the inhibition of BMAL:CLOCK with phosphorylation-dependent proteolytic degradation of PER through casein kinases (CK1ε and CK1δ) and CRY through AMP-activated protein kinase (AMPK) (Gallego and Virshup 2007).

Fig. 11.1
figure 1

The transcriptional network of the molecular clock and its downstream effectors. The circadian clock is a coordinated transcription-translation feedback loop generated by transcriptional activators, BMAL1 and CLOCK, and repressors PER and CRY. BMAL1 and CLOCK heterodimerize and bind to the E-box response element in the promoter of their target genes, including PER and CRY. As PER and CRY accumulate, they form a complex that inhibits the BMAL1:CLOCK dimer. The cycle renews with the proteolytic degradation of PER and CRY through phosphorylation by CK1ε/δ and AMPK. In mice, the peak DNA binding of BMAL1:CLOCK is in the light cycle, while the peak expression of target genes is in the dark cycle

Full size image

As mentioned previously, studies have demonstrated the existence of cell/tissue specific autonomous clock machinery. Interestingly, the liver, even after 16 days of a shifted light/dark cycle, does not completely adapt to the change (Yamazaki et al. 2000). Correspondingly, using a PER2:Luciferase protein as a reporter of circadian activity, it was revealed that peripheral tissues in isolation were able to self-sustain circadian oscillations for more than 20 days (Yoo et al. 2004). Furthermore, lesions in the SCN did not abolish peripheral circadian oscillations, but instead resulted in phase desynchrony among tissues (Yoo et al. 2004). This suggests that the brain clock is necessary for phase control of clocks in peripheral organs but not required for tissue-specific circadian oscillations. Instead, there appears to be tissue-specific cues that are more effective in synchronizing peripheral clocks. In the liver, the inversion of the feeding cycle rapidly alters the hepatic expression of metabolic genes, suggesting that the liver metabolic program is more rapidly entrained to rhythmic feeding and fasting, even in the presence of a functioning central pacemaker (Damiola et al. 2000; Stokkan et al. 2001), providing a link between the circadian clock machinery and metabolic regulation.

2.4 Clock Regulators and Metabolic Homeostasis

The important role of the circadian machinery in metabolism became evident through genetic deletion or mutations of individual core clock components. Specifically, mice with loss-of-function of Bmal1 or Clock developed metabolic disorders. Clock knockout mice are hyperphagic and obese and display a range of abnormal metabolic parameters such as hyperlipidemia, β cell dysfunction, hepatic steatosis, and hyperglycemia (Pan et al. 2010; Rudic et al. 2004; Turek et al. 2005). Additional studies have indicated a potential role of Clock in lipid uptake, absorption, synthesis, and degradation. Microsomal triglyceride transfer protein (MTP) is an essential component to the hepatic secretion of apolipoprotein B-containing lipoproteins, such as chylomicrons and VLDLs. Clock regulates Mtp and plasma triglycerides circadian changes via Shp, which binds to Hnf4α/Lrh-1 at the Mtp promoter. Clock loss of function and Shp knockdown mice lose diurnal expression of Mtp and have high plasma triglycerides at all times (Pan et al. 2010). Clock knockout mice also display altered rhythmic expression of enzymes involved in lipid biosynthesis and fatty acid degradation, including Acsl4 and Fabp1 (Kudo et al. 2007).

Similarly, Bmal1 knockout mice are arrhythmic in complete darkness, develop hyperlipidemia and hepatic steatosis, and are unable to maintain oscillations in plasma glucose and triglycerides (Rudic et al. 2004; Shimba et al. 2011). Bmal1 regulates de novo lipogenesis in the liver via insulin-mTORC1-Akt signaling (Zhang et al. 2014), and Bmal1 overexpression in mouse liver significantly elevated the mRNA level of Srebp-1c and of lipogenic enzymes such as Scd1, Fasn, and Acaca (Zhang et al. 2014). Furthermore, Bmal1 KO primary hepatocytes displayed impaired expression of lipogenic enzymes, Acaca, Fasn, Scd1, glycerol-3-phosphate acyltransferase (Gpat), and lipogenic transcription factors, Chrebp, Srebp1c, and Pgc1β (Zhang et al. 2014), which highlights the critical function of Bmal1 in hepatic lipogenesis. In addition, abolishment of the negative regulators, Per and Cry, has also resulted in altered lipid metabolism (Grimaldi et al. 2010; Rudic et al. 2004; Turek et al. 2005) with reduced plasma triglyceride level in mice lacking Per1/Per2 (Grimaldi et al. 2010). These studies collectively have suggested a critical role of the clock machinery as a novel regulator that couples circadian rhythms to metabolism.

2.5 Circadian Regulation of Liver Metabolism via Nuclear Receptors

While the molecular clock directly regulates gene expression, it also “outsources” metabolic regulation to several downstream transcriptional regulators. Notable among these are nuclear receptors , which are transcription factors that contain both DNA and ligand binding domains (Chawla et al. 2001). They can be activated by a range of metabolic ligands including steroid hormones, fatty acids, oxysterols, bile acids, and heme. Several studies have suggested that tissue-specific expression and rhythmicity of nuclear receptors link peripheral circadian clocks to tissue-specific metabolic outputs. Pertinently, 20 of the 41 nuclear receptors expressed in the mouse liver display a rhythmic expression pattern. Indeed several of these receptors have been implicated in hepatic lipid metabolism (Yang et al. 2006). Moreover, endogenous ligands for some of these nuclear receptors are known to oscillate in a circadian fashion, thus adding another layer of integration between the molecular clock and nuclear receptors in regulating metabolic homeostasis.

RORs and REV-ERBs . In addition to the core molecular clock described earlier, the nuclear receptors RAR-related orphan receptors (RORs) and REV-ERΒs act as auxiliary clock components that directly regulate clock gene expression. ROR and REV-ERB are clock output proteins that bind ROR response elements (ROREs) in target enhancer regions and act as constitutive transcriptional activators and repressors, respectively (Dumas et al. 1994; Forman et al. 1994; Retnakaran et al. 1994). Pertinently, Rors and Rev-erbs directly regulate Bmal1 gene expression, thus acting as accessory clock components that facilitate rhythmic Bmal1 expression. The expression and rhythmicity of Ror (α, β, and γ) and Rev-erb (α and β) isoforms is tissue-specific. Specifically, Rorγ and Rev-erbα/β display rhythmic expression in the liver (Yang et al. 2006). In circadian terminology, a standard of time is defined based on the period of a cue given by the environment (called a zeitgeber). Under standard light-dark cycles, the time of lights on defines zeitgeber time zero (ZT0) and the time of lights off defines zeitgeber time twelve (ZT12). For nocturnal animals, ZT0 and ZT12 represent the beginning of a physiological fasting and feeding phases, respectively. In the mouse liver, mRNA levels of Rev-erbs α and β peak during the light/fasting cycle at ZT4 and ZT8, respectively, while Rorγ expression peaks during the dark cycle at ZT16. Furthermore, crosstalk exists between Rev-Erb and Ror proteins and occurs in a context-dependent manner. For example, Rev-Erbα has an RORE in its promoter region, and thus it represses its own expression and is induced by Rors (Delerive et al. 2002; Raspe et al. 2002).

Several studies have established that Rev-erb α and β play crucial and direct roles in regulating hepatic lipogenesis (Fig. 11.2). Cistromic analyses have shown that Rev-erbα co-localizes with the histone deacetylase Hdac3 and its associated co-repressor NCoR at several loci encoding genes involved in lipogenesis (Feng et al. 2011). Consistently, hepatic deletion of either Hdac3 or Rev-Erbα in mice results in dramatic hepatosteatosis, thus consolidating their roles in the repression of lipogenesis. The cistromic analyses further revealed that Rev-Erbα and Hdac3 occupancy displayed a circadian pattern inversely related to histone acetylation and RNA polymerase II recruitment. Therefore, during the active/feeding period Hdac3 occupancy decreases allowing for increased expression of lipogenic genes. In fact, Rev-Erbα appears to be necessary for Hdac3 recruitment to metabolic genes. These data suggest that the circadian epigenomic remodeling controlled by Hdac3 and Rev-erbα is essential for homeostasis of the lipogenic process in the liver. Furthermore, similar cistromic analysis has shown that loci that are mutually bound by both Rev-erbs α and β show a significant enrichment in genes involved in lipid and lipoprotein metabolism (Bugge et al. 2012), suggesting functional redundancy of the β isoform in the diurnal regulation of lipogenesis.

Fig. 11.2
figure 2

Circadian partitioning of hepatic lipid synthesis and catabolism . The circadian clock orchestrates the transcriptional partitioning of hepatic lipid usage through several downstream regulators. The gene expression patterns of PPARα and PPARδ, REV-ERBα and REV-ERBβ, and SREBP1C all display circadian rhythmicity. Further, the activating ligands for these nuclear receptors as well as for FXR and LXR also display rhythmic accumulation throughout the fasting/feeding cycle. This transcriptional coordination of de novo lipogenesis by the molecular clock and its downstream transcriptional regulators coincides with post-prandial regulation of lipogenesis by insulin. Diagram illustrates liver metabolism in mice

Full size image

While the negative feedback loop generated by the core clock (namely BMAL1, CLOCK, CRY, and PER) is sufficient to form a circadian oscillator, the establishment of a secondary, or auxiliary, feedback loop via the ROR and REV-ERB proteins is thought to make the circadian oscillator robust and tunable (Stricker et al. 2008; Tigges et al. 2009). Thus, it has been proposed that endogenous ligands may tune molecular clock systems via the ROR and REV-ERB nuclear receptors in response to metabolic cues. As of now, heme is the only known endogenous ligand of REV-ERBα and β, and heme levels are known to oscillate in a circadian fashion (Kaasik and Lee 2004; Raghuram et al. 2007; Yin et al. 2007). Heme is a large heterocyclic prosthetic group that is involved in a range of biological functions including diatomic gas transfer, electron transfer, and chemical catalysis. How this and other ligands of the REV-ERB and ROR nuclear receptors may act to coordinate molecular clock function with dynamic metabolic needs in peripheral tissues has yet to be characterized in depth.

PPARs . Peroxisome Proliferator-activated Receptors (PPARs) play multi-faceted roles in the regulation of lipid metabolism in a variety of tissues and cell types. PPARs form heterodimers with an obligate interaction partner, Retinoid X Receptor (RXR). Endogenous ligands of the three PPAR isoforms (α, β/δ, and γ) include mono- and polyunsaturated fatty acids and eicosanoids. Several synthetic ligands have also been generated for these receptors. All three isoforms show rhythmic expression in the mouse liver, with expression of Pparγ and Pparα peaking during the light cycle at ZT8 and ZT12, respectively, and Pparβ/δ peaking during the dark cycle at ZT20 (Yang et al. 2006). Pparγ is most abundantly expressed in adipose tissue and plays an indispensable role in adipocyte differentiation (Grimaldi et al. 2010); however, its role in the liver is less well understood. PPARα is a master regulator of fatty acid catabolism (Kersten et al. 1999; Leone et al. 1999), while PPARδ promotes de novo lipogenesis (Liu et al. 2011).

PPARα mediates metabolic responses to fasting including fatty acid oxidation and ketogenesis. It regulates the expression mitochondrial enzymes involved in fatty acid import and oxidation, including medium-chain acyl-CoA dehydrogenase and carnitine palmitoyltransferase I, as well as extramitochondrial enzymes such as acyl-CoA oxidase, cytochrome P450 4A3, and Abcd2 and Abcd3 which mediate peroxisomal fatty acid oxidation (Fourcade et al. 2001; Leone et al. 1999). PPARα also regulates the expression of fatty acid binding proteins which chaperone intracellular fatty acids and transport PPARα ligands to the nucleus (Wolfrum et al. 2001), thus enacting a feed-forward process.

PPARδ is a positive regulator of hepatic lipogenesis. In contrast to its counterpart Rev-erbα, which suppresses lipogenesis during the light/fasting cycle in mice, Pparδ promotes lipogenesis during the dark/feeding cycle (Fig. 11.2). Pparδ directly regulates the expression of the rate-limiting enzyme of de novo lipogenesis, acetyl-coA carboxylase 1 (Acc1) (Liu et al. 2013). Thus, Rev-erbα and Pparδ coordinate the circadian rhythmicity of lipid synthesis in the liver. Furthermore, recent research has implicated Pparδ-dependent hepatic lipogenesis in interorgan communication via lipid signaling molecules, which will be discussed in detail later.

LXR and FXR . The Liver X Receptors (LXRα and LXRβ) promote de novo lipogensis at least in part through activating the expression of SREBP1C, a direct transcriptional activator of lipogenic enzymes (Repa et al. 2000). Like the PPARs, RXR is the obligate LXR binding partner. On the other hand, the Farnesoid X Receptor (FXR) represses de novo lipogenesis and promotes triglyceride clearance. Fxr indirectly represses the expression of several genes involved in lipid metabolism including Srebp1c through induction of Shp (Watanabe et al. 2004).

While Lxr and Fxr are constitutively expressed in mouse liver (Yang et al. 2006), studies have suggested that the molecular clock regulates the activity of these nuclear receptors through indirect mechanisms. First, the clock regulates Lxr and Fxr protein stability via the NAD-dependent protein deacetylase sirtuin-1 (Sirt1) (Kemper et al. 2009; Li et al. 2007; Nakahata et al. 2009). The Bmal1-Clock heterodimer drives the rhythmic accumulation of the redox substrate NAD+ partly by controlling the transcription of the enzyme nicotinamide phosphoribosyltransferase (Nampt), which is rate-limiting in the NAD+ salvage pathway (Ramsey et al. 2009). This poises the cyclic activation of Sirt1, which in turn deacetylates and destabilizes Lxr and Fxr during fasting.

LXR activity can also be regulated by the rhythmic accumulation of is putative ligands, cholesterol and oxysterols, throughout fasting/feeding cycles (Janowski et al. 1996). Furthermore, LXRs promote the conversion of cholesterol to bile acids through up-regulation of the CYP7A1 enzyme (Peet et al. 1998). Bile acids are known FXR ligands (Wang et al. 1999). In turn, FXR generates a negative feedback loop by repressing CYP7A1 expression through SHP (Goodwin et al. 2000). As such, the protein stability and ligand activation of LXRs and FXR are coordinately regulated by fasting and feeding.

3 Hepatic Lipogenesis and Peripheral Metabolism

In the first part of this chapter, we discussed critical regulatory nodes controlling metabolic rhythm of hepatic metabolism. The remainder of the chapter will focus on the impact of dysregulation in the hepatic lipogenic pathway on systemic metabolic homeostasis.

3.1 The Lipogenic Pathway

Lipogenesis involves several enzymatic steps, from the mitochondrial synthesis of citrate and its transport to the cytosol for generation of acetyl-CoA, to malonyl CoA synthesis followed by production of palmitate that serves as a building block for synthesis of complex lipids (Fig. 11.3). The steps involved in the generation of the shared intermediate acetyl-CoA directly link lipid and carbohydrate metabolism and dictate the capacity to synthesize fatty acids from carbohydrates . De novo lipogenesis is essential for embryonic development. Whole-body knockouts of ATP citrate lyase (Acly, catalyzes cytosolic formation of acetyl-CoA from citrate), ACACA, or FASN (the latter two are rate-limiting enzymes of de novo lipogenesis), are embryonic lethal (Abu-Elheiga et al. 2005; Beigneux et al. 2004; Chirala et al. 2003). In the case of liver-specific loss- or gain-of-function, mice are viable but appear to have both hepatic and systemic metabolic phenotypes. A summary of the mouse genetic models , focusing on liver specific modulation, is provided in Table 11.1. As discussed earlier, most enzymes involved in liver de novo lipogenesis display a 24-h oscillation in mRNA transcript (Hughes et al. 2009; Miller et al. 2007; Panda et al. 2002). The studies reported herein do not necessarily focus on time-dependent effects. They remain consistent with the notion that hepatic lipogenesis has strong regulatory effects on metabolism in peripheral organs and on whole body energy balance.

Fig. 11.3
figure 3

Hepatic de novo lipogenesis and the generation of lipid mediators. Glycolysis is a major pathway providing substrates for de novo lipogenesis in the fed state. Mitochondrial citrate is exported to cytosol and converted to acetyl-CoA. Malonyl CoA synthesis is followed by production of palmitate that serves as a building block for synthesis of complex lipids. Products of hepatic de novo lipogenesis, such as PC(16:0/18:1) and PC(18:0/18:1), have been implicated in mediating postprandial fat utilization in the liver and skeletal muscle, respectively. Gck glucokinase, Pfkfb 6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase, Pdha1 puruvate dehydrogenase, Acly ATP citrate lyase, Acaca acetyl-Coenzyme A carboxylase alpha, Fasn fatty acid synthase, Scd1 stearoyl-Coenzyme A desaturase 1, SFA saturated fatty acids, MUFA mono unsaturated fatty acids

Full size image
Table 11.1 Metabolic phenotypes of mouse models with liver specific modulation of de novo lipogenesis
Full size table

3.2 Metabolic Phenotypes Associated with Gain- or Loss-of Function Mouse Models

Gck : Carbohydrate metabolism is an important source of acetyl-CoA through glycolysis and pyruvate oxidation. Glucokinase (Gck) and 6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase (Pfkfb) are rate limiting enzymes for glycolysis which are also central in driving glucose flux for hepatic lipogenesis. Adenovirus mediated hepatic overexpression of Gck or Pfkfb increases glycolysis (Wu et al. 2005). However, only Gck overexpression causes lipid accumulation in liver and circulation. Interestingly, both hepatic Gck and Pfkfb dramatically increase fatty acid oxidation in skeletal muscle.

Pdha1 . The pyruvate dehydrogenase complex is at the crossroads of glucose and fatty acid metabolism, as it catalyzes mitochondrial pyruvate decarboxylation into acetyl-CoA. Liver specific knockout of the Pdha1 subunit blocks the incorporation of glucose carbon into fatty acids by hepatocytes (Choi et al. 2010). These mice displayed reduced fat (33 %) and lean mass (9 %), increased lipogenic capacity in epididymal adipose tissue, and improved peripheral insulin sensitivities during a hyperinsulinemic euglycemic clamp.

Acly . ATP citrate lyase (Acly) regulates the synthesis of acetyl-CoA from citrate. Liver-specific Acly knockdown via adenovirus-mediated RNA interference reduces hepatic contents of acetyl-CoA and malonyl-CoA, inhibits hepatic de novo lipogenesis, and allaviates hepatic steatosis in db/db mice (Wang et al. 2009). Interestingly, these mice display improved glucose tolerance (GTT) and increased muscle insulin sensitivity.

Acaca/b : Acetyl-CoA carboxylase catalyzes the synthesis of malonyl-CoA, which serves as the donor of 2-carbon units to fatty acid synthesis. The two isoforms of acetyl-CoA carboxylase, Acaca and Acacb, are encoded by separate genes and show distinct tissue distributions. Acaca is highly expressed in the liver, whereas Acacb is predominantly expressed in the skeletal muscle and heart, and to a lesser extent in the liver. In addition, Acacb is localized in the mitochondria while Acaca is cytosolic. Liver-specific Acaca knockout mice display a mild phenotype of decreased hepatic lipid content and serum non-esterified fatty acids (NEFA) under chow or fat free diet (Mao et al. 2006). It is not clear whether the up-regulation of Acacb observed in this model alleviates the phenotype by compensating for Acaca loss. Whole body Acacb knockout mice (Abu-Elheiga et al. 2001) have lowered liver fat content, an effect of increased β-oxidation as a result of the decrease in malonyl-CoA (a negative regulator of the mitochondrial carnitine palmitoyl-CoA shuttle system).

Fasn . Fatty acid synthase (Fasn) catalyzes the first committed step in fatty-acid biosynthesis. Liver Fasn knockout mice develop fatty liver and hypoglycemia under a fat-free lipogenic diet or prolonged fasting (Chakravarthy et al. 2005), a phenotype that resemble Pparα deficiency in the liver. Indeed, the hypoglycemia/steatohepatitis phenotype could be corrected by pharmacological activation of PPARα. It is suggested that that Fasn is required for the production of endogenous PPARα ligands under fat-free conditions, which will be discussed below.

Scd1 . Liver-specific knockout of stearoyl-CoA dehydrogynase (Scd1) , an enzyme that catalyzes a rate-limiting step in the synthesis of monounsaturated fatty acids (MUFA) from saturated fatty acids, protects mice from high carbohydrate diet-induced metabolic disorders (Cohen et al. 2002; Miyazaki et al. 2007) as a consequence of suppressed hepatic lipid accumulation and gluconeogenesis and reduced adipose tissue weight. In Scd1 knockout mice fed a high-sucrose, very low-fat diet, oleate feeding rescued the defective hepatic triglyceride secretion and hepatic lipogenesis. These results are consistent with the notion that hepatic SCD1 is required for carbohydrate-induced adiposity. In a follow-up study, the authors showed that lack of Scd1 in the liver lowered the MUFA levels of white adipose tissue (Flowers et al. 2012). Furthermore, liver and plasma triglycerides showed similar alterations in fatty acid composition, indicating that fatty acid content of plasma triglycerides is predictive of hepatic Scd1 activity. The data therefore support the existence of crosstalk between liver and adipose tissue, which also raises the possibility that blood-borne lipid metabolites can serve as signaling molecules for inter organ communication to achieve coordinated energy substrate utilization. In the following section, we discuss recent studies that identify bioactive lipids synthesized de novo or derived from dietary fats implicated in such communication.

3.3 De Novo Lipogenesis and Tissue Crosstalk

Despite ample genetic evidence demonstrating a clear role for hepatic de novo lipogenesis in metabolic homeostasis, the underlying mechanisms remain unclear. A simple explanation is that hepatic lipogenesis modulates energy substrate availability, which could exert metabolic consequences at the whole-body level. Recent studies also suggest that lipiogenic products may function as metabolic signaling molecules. A notable finding came from hepatic Fasn knockout mice. As discussed above, the hypoglycemia and fatty liver phenotype under a fat-free diet can be rescued with a synthetic PPARα agonist, suggesting that Fasn participates in the production of endogenous PPARα ligands (Chakravarthy et al. 2005). Mass spectrometry profiling of liver extracts from wild type or Fasn knockout mice was performed to screen for lipids bound by PPARα (Chakravarthy et al. 2009). Phosphatidylcholine (PC) (16:0/18:1) was identified as a putative PPARα ligand (Fig. 11.3). In fact, portal vein infusion of PC(16:0/18:1) increased fatty acid oxidation in a PPARα dependent manner. Of note, in mice Fasn and lipogenesis are most active in the dark (feeding) cycle, whereas Pparα is known to control fat catabolism in the light cycle. It is likely that postprandial production of PC(16:0/18:1) is in anticipation of the upcoming fasting state when Pparα-controlled β-oxidation and ketogenesis are critical for energy metabolism.

PPARδ is best known for its activity in the control of muscle oxidative metabolism in type I fibers (Evans et al. 2004). Previous work has shown that Pparδ knockout mice are glucose intolerant, whereas PPARδ activation by an agonist in db/db mice improves insulin sensitivity in insulin-responsive tissues (Lee et al. 2006). Molecular and functional analyses suggested that PPARδ activation reduces hepatic glucose production by increasing glycolysis and the pentose phosphate shunt to promote fatty acid synthesis in the liver. In fact, DNA array analysis showed that genes involved in fatty acid synthesis were up-regulated in the liver by PPARδ agonist treatment. As expected, PPARδ ligand treatment increased the expression of genes involved in fatty acid catabolism in muscle. These results indicated a coordinated regulation of glucose/lipid metabolism in the liver-muscle axis. Follow-up studies revealed that Pparδ regulates the diurnal expression of several lipogenic genes in the dark cycle, which correlates well with rhythmic alterations in serum lipid profiles and the activity of fatty acid uptake in muscle (Liu et al. 2013). Genetic evidence further suggests that hepatic Pparδ activity is sufficient to drive the observed change in diurnal muscle fatty acid utilization. Unbiased lipidomic profiling of serum or liver samples from mice with hepatic Pparδ activation/deletion or Acaca knockdown showed that PC(18:0/18:1), or 1-stearoyl-2-oleoyl-sn-glycero-3-phosphocholine (SOPC) was correlated with altered muscle fatty acid uptake phenotypes (Fig. 11.3). In vitro or in vivo delivery of this lipid enhanced muscle cell fatty acid uptake. Interestingly, this effect also requires intact PPARα signaling in the muscle, indicating that PC(16:0/18:1) and PC(18:0/18:1) share similar biological effects on PPARα activation. Therefore, functional identification of PC(16:0/18:1) and PC(18:0/18:1) suggests that these lipogenic products may provide signals for postprandial fat utilization in the liver-muscle axis.

4 Concluding Remarks

With the increasing prevalence of obesity and its associated metabolic pathologies becomes pandemic, many researchers have started to focus on identifying lipid metabolites that are involved in metabolic regulation or dysregulation. De novo lipogenesis, a process essential for embryonic development, is undoubtedly involved in both processes. As an important biosynthetic pathway in metabolic flexibility, hepatic de novo lipogenesis is regulated by several signaling pathways and a network of transcription factors, including the molecular clock and lipid-sensing nuclear receptors. These regulatory mechanisms integrate temporal, nutritional, and hormonal controls. Recent studies support the notion that the lipogenic pathway may generate bioactive lipids, which serve as intracellular or long-range metabolic mediators that coordinate fat utilization corresponding to feeding or fasting states. In addition to PC(16:0/18:1) and PC(18:0/18:1) discussed above, palmitoleate (C16:1n7) (Kuda et al. 2009) and branched fatty acid esters of hydroxy fatty acids (FAHFAs) (Yore et al. 2014) have been described to be associated with adipose tissue lipogenesis and to exert systemic insulin sensitizing effects. Therefore, understanding both the upstream signals and downstream effectors of the lipogenic pathway may provide new therapeutic opportunities for drug development to treat metabolic diseases.