Regulation of mammalian energy metabolism is an intricate process involving numerous hormones, transcription factors and signal transduction cascades. Much of the regulation occurs via secreted factors that relay information from one organ to another. One group of secreted factors that recently emerged as having a major impact on lipid and possibly glucose metabolism are the ANGPTLs (angiopoietin-like proteins). This includes ANGPTL3, ANGPTL4/FIAF (fasting-induced adipose factor), and ANGPTL6/AGF (angiopoietin-related growth factor). Although the receptors for these proteins have yet to be identified, it is nevertheless increasingly clear that these proteins have important effects on plasma triacylglycerol clearance, adipose tissue lipolysis, and adiposity. This review summarizes contemporary data on ANGPTLs with emphasis on the connection with energy metabolism.
- angiopoietin-like protein (ANGPTL)
- fasting-induced adipose factor
- glucose metabolism
- liver X receptor
- plasma triacylglycerol
As our modern society is facing a global epidemic of obesity, there is heightened interest in the regulation of energy metabolism. At the molecular level, much of the attention has focused on the role of various ligand-activated transcription factors as well as intracellular signal transduction cascades, which are amenable to pharmacological intervention. In the last few years it has become clear that much of the regulation also occurs via novel secreted factors, such as leptin and adiponectin, allowing the various organs to communicate with each other. One group of recently discovered factors for which there is now sound evidence linking them to energy metabolism are the ANGPTLs (angiopoietin-like proteins).
The angiopoietins and ANGPTLs encompass a group of about nine proteins that share a common modular structure consisting of an N-terminal signal sequence, a unique region of variable length, a coiled-coil domain, and a large fibrinogen/angiopoietin-like domain. The latter two well-conserved domains take up the majority of the protein and are probably involved in oligomerization, giving rise to several high molecular mass species. Although angiopoietins and ANGPTLs are indistinguishable structurally, the property that separates the two sets of proteins is the ability of angiopoietins to bind to the Tie2 receptor tyrosine kinase, affecting angiogenesis, blood-vessel maturation, and vascular-endothelium integrity [1,2]. Despite bearing a name that is suggestive of such a function, the evidence linking ANGPTLs with angiogenesis is less compelling. For example, studies have ascribed both a stimulatory and suppressive role for ANGPTL4 in angiogenesis [3,4]. At the same time, data abound connecting ANGPTLs, especially ANGPTL3, ANGPTL4 and ANGPTL6, with regulation of energy metabolism. Indeed, in the last few years it has become evident that ANGPTLs are important secretory proteins governing plasma lipid levels and adiposity. This review will concentrate on the connection between ANGPTLs and regulation of energy homoeostasis, addressing the specific role of ANGPTL3, ANGPTL4, and ANGPTL6.
ANGPTL3 was discovered while searching EST (expressed sequence tag) databases for signal sequences and amphipathic helices . The protein contains 460 and 455 amino acids in human and mouse respectively, and shares the characteristic modular structure of ANGPTLs, consisting of a signal sequence, a small unique region, a coiled-coil domain and a large fibrinogen/angiopoietin-like domain. In mouse, ANGPTL3 is expressed principally in liver, with much lower expression levels observed in kidney and lung. The first link between ANGPTL3 and lipid metabolism was established by Koishi et al. , who via positional cloning managed to map the genetic defect in KK/San mice to the Angptl3 gene. KK/San mice represent a mutant strain of KK obese mice with markedly lower plasma NEFA (non-esterified fatty acid) and TG (triacylglycerol) levels. Adenoviral-mediated overexpression and intravenous injection of ANGPTL3 were subsequently shown to increase plasma TGs, NEFAs, and cholesterol in both wild-type C57/B6 and KK/San mice. Elevation of fasting plasma TG levels by ANGPTL3 can be attributed to suppression of VLDL (very-low-density-lipoprotein) clearance via inhibition of the activity of LPL (lipoprotein lipase), a key enzyme in plasma TG hydrolysis, rather than stimulation of hepatic VLDL production [7,8]. Targeted deletion of the Angptl3 resulted in a similar phenotype as observed in the KK/San mice: markedly decreased plasma TG levels associated with elevated heparin-releasable LPL activity . It appears that region 17–165 of ANGPTL3 is sufficient to elicit the hypertriglyceridemic effect. This region is contained in the N-terminal fragment that is produced upon proteolysis of ANGPTL3. Mutating the putative cleavage sites reduces the potency of ANGPTL3 towards elevating plasma TG, suggesting that regulated proteolysis of ANGPTL3 may be an important mechanism to govern its activity and thus the activity of LPL .
ANGPTL3 causes elevation of plasma NEFA, yet inhibition of LPL is expected to lower rather than raise plasma NEFA levels. Instead, ANGPTL3 appears to stimulate adipose tissue lipolysis via direct interaction with adipocytes . The cellular mechanism behind the stimulatory effect of ANGPTL3 on adipose tissue lipolysis is unknown, but probably involves an extracellular receptor that has yet to be identified. It can thus be concluded that ANGPTL3 is a liver-derived factor that prevents plasma TG clearance and promotes adipose tissue lipolysis. Based on this information, it could be anticipated that Angptl3 knock-out mice have elevated adipose-tissue stores. However, this does not appear to be the case .
Expression of ANGPTL3 in liver is governed by the LXR (liver X receptor), a nuclear hormone receptor involved in the regulation of hepatic TG and cholesterol metabolism. Administration of the synthetic LXR agonist T0901317 markedly increases hepatic ANGPTL3 mRNA levels in mice and in human HepG2 hepatoma cells [12,13]. A functional LXR response element of the DR-4 type has been identified in the promoter of the human ANGPTL3 gene, establishing ANGPTL3 as a direct target gene of LXR. Treatment of mice with synthetic LXR agonists is known to cause pronounced hypertriglyceridaemia . This effect is probably mediated via up-regulation of ANGPTL3, since LXR agonists fail to elevate plasma TGs in mice expressing a defective form of ANGPTL3, despite up-regulation of numerous genes involved in lipogenesis . Although highly expressed in liver, ANGPTL3 does not appear to be under transcriptional control of PPARα (peroxisome-proliferator-activated receptor α) (F. Zandbergen and S. Kersten, unpublished work).
ANGPTL4 was discovered in parallel by numerous groups. Kim et al.  picked up ANGPTL4, which they named HFARP (hepatic fibrinogen/angiopoietin-related protein), by degenerative PCR while looking for additional members of the angiopoietin family. Kersten et al.  identified ANGPTL4 in a subtractive hybridization screen to isolate novel target genes of the nuclear receptor PPARα in liver, and called it FIAF (fasting-induced adipose factor). Finally, Yoon et al.  found ANGPTL4, which they named PGAR (PPARγ angiopoietin-related), in a similar screen looking for targets of the PPARγ agonist troglitazone in adipose tissue. In order to promote a uniform nomenclature for the angiopoietins and ANGPTLs, in the remainder of this paper the name ANGPTL4 will be used.
Contrary to ANGPTL3 and ANGPTL6, highest expression of ANGPTL4 is found in white adipose tissue, earning the classification adipo(cyto)kine, yet reasonable expression is also observed in liver, heart, skeletal muscle, and intestine [16,17]. The ANGPTL4 protein has a molecular mass of about 50 kDa, appears to be glycosylated, and similar to other members of the ANGPTL family is divided up into distinct regions. Furthermore, analogous to other proteins possessing a fibrinogen/angiopoietin-like domain, ANGPTL4 forms higher order oligomeric structures . Formation of these multimers is probably mediated via specific cysteine residues in its N-terminal domain creating disulphide bridges . In addition to multimerization, ANGPTL4 also undergoes proteolytic processing, giving rise to the full length form and a truncated form . Both forms of ANGPTL4 can be detected in human blood plasma, and their abundance displays marked inter-individual variation. In humans, interestingly, processing of ANGPTL4 appears to be tissue-dependent, with liver secreting truncated ANGPTL4 and adipose tissue secreting the full length form. Plasma levels of ANGPTL4 are reportedly lower in patients with type II diabetes compared to healthy subjects, but do not seem to be related to adiposity as expressed by body mass index [18,20].
ANGPTL4 is evolutionarily most closely related to ANGPTL3 and this resemblance is also observed at the functional level. Indeed, in analogy with ANGPTL3, ANGPTL4 has primarily been connected with regulation of lipid metabolism. By intravenous injection of recombinant ANGPTL4 protein, Yoshida et al.  were the first to show that ANGPTL4 potently raises plasma TGs. In vitro experiments using recombinant ANGPTL4 and purified LPL demonstrate that ANGPTL4 is able to inhibit LPL activity, providing a plausible explanation for its hypertriglyceridemic effect. Adenoviral-mediated overexpression of ANGPTL4 confirm these data, showing severely elevated plasma TG levels within several days after virus injection [19,20]. The elevated plasma TG levels under fasting conditions are due to decreased VLDL clearance . Further evidence for the inhibitory effect of ANGPTL4 on plasma TG clearance via LPL was gathered using various mouse models: transgenic mice overexpressing ANGPTL4 in heart or liver, and ANGPTL4 null mice [9,22]. Our studies with transgenic mice that overexpress ANGPTL4 in peripheral tissues show that moderate (three to five-fold) overexpression of ANGPTL4 causes a two to three-fold elevation of plasma TG levels, which was observed under both fed and fasted conditions. Additional experiments indicated that impaired clearance of TG-rich apoB-containing lipoproteins (represented by chylomicrons and VLDL), rather than increased production, is responsible for the observed hypertriglyceridaemia (S. Mandard, F. Zandbergen, E. van Straten, W. Wahli, F. Kuipers, M. Müller and S. Kersten, unpublished work). Interestingly, the ability of ANGPTL4 to elevate plasma TG seems to be dependent on oligomerization, which may stabilize the coiled-coil domain .
Overall, a wealth of data are now available showing a marked inhibitory effect of ANGPTL4 on clearance of TG-rich lipoprotein, which is mediated by inhibiting LPL. Impaired TG clearance leads to depletion of adipose-tissue stores, hypertriglyceridaemia and elevated hepatic TG storage (S. Mandard, F. Zandbergen, E. van Straten, W. Wahli, F. Kuipers, M. Müller and S. Kersten, unpublished work). Whether ANGPTL4 plays a similar role in TG metabolism in humans still has to be demonstrated. Nevertheless, it can be speculated that ANGPTL4 is a promising pharmacological target for the treatment of hypertriglyceridaemia.
In addition to lipid metabolism, recent data have also linked ANGPTL4 to regulation of plasma glucose metabolism. Adenoviral-mediated overexpression of ANGPTL4 in C57/B6 mice was found to be associated with a dramatic decrease in plasma glucose levels and improved glucose tolerance . In contrast, deletion of the ANGPTL4 gene or hepatic overexpression did not affect fed or fasted plasma glucose levels . In our experiments, peripheral ANGPTL4 overexpression did not alter plasma glucose levels, yet modestly impaired glucose tolerance, which became much more pronounced after feeding a high fat diet (S. Mandard, F. Zandbergen, E. van Straten, W. Wahli, F. Kuipers, M. Müller and S. Kersten, unpublished work). Thus the effect of ANGPTL4 on glucose homoeostasis remains controversial.
Blocking of LPL is expected to result in a decreased plasma NEFA level. However, injection of ANGPTL4 and peripheral ANGPTL4 overexpression were both associated with elevated plasma NEFA . Since plasma glycerol was increased as well, these data suggest that ANGPTL4 stimulates adipose-tissue lipolysis. One possible mediator of the effect of ANGPTL4 may be the recently identified ATGL (adipose TG lipase), which was found to be up-regulated in white adipose tissue of ANGPTL4 transgenic mice (S. Mandard, F. Zandbergen, E. van Straten, W. Wahli, F. Kuipers, M. Müller and S. Kersten, unpublished work). Additional studies are required to define better the intracellular events triggered by ANGPTL4 that lead to enhanced fat lipolysis.
The regulation of ANGPTL4 expression has been extensively studied in a variety of tissues. Expression of ANGPTL4 is highly up-regulated during hypoxia, an effect which is mediated by the transcription factor HIF-1α (hypoxia inducible factor 1α) [3,23,24]. This observation, and the structural analogy with angiopoietins, has led to the idea that ANGPTL4 may be involved in angiogenesis. Another stimulus that increases ANGPTL4 expression in several tissues is fasting [16,25]. The effect of fasting partially occurs via PPARα, a nuclear hormone receptor that is activated by fatty acids. PPARα is a very potent activator of ANGPTL4 transcription, governing ANGPTL4 expression in liver, heart, skeletal muscle and intestine. ANGPTL4 is also under transcriptional control of PPARβ/δ (in skin, adipose tissue and liver) and PPARγ (in adipose tissue). In fact, ANGPTL4 appears to be a general target of PPARs, responding to all three PPARs in numerous tissues. The genomic sequence responsive to PPAR stimulation was localized to intron 3, and differed little from the consensus DR-1 PPAR responsive element . Corroborating regulation by PPARα, plasma levels of truncated ANGPTL4 were found to be increased in patients after 4 weeks of treatment with fenofibrate, a potent synthetic PPARα agonist . Regulation of ANGPTL4 by all three PPARs in numerous tissues is hard to reconcile with an exclusive role in plasma TG clearance. Instead, these data suggest ANGPTL4 may be much more versatile, similar to what is observed for many other adipokines. Additional research will be necessary to substantiate this claim.
Further research will also be necessary to establish the functional differences between ANGPTL3 and ANGPTL4. So far the roles assigned to both proteins are completely overlapping (Figure 1), the only difference being their pattern and regulation of expression.
ANGPTL6, which is also called angiopoietin-like growth factor, was identified by screening EST databases for angiopoietin homologues. Human and mouse ANGPTL6 are comprised of 470 and 457 amino acids respectively, giving rise to a protein of about 50 kDa. In mouse, ANGPTL6 protein can be detected in liver, but profiling of mRNA expression levels among several tissues still has to be performed . Accordingly, it is unclear whether liver constitutes the major site of expression in mouse and human.
At the functional level, overexpression of ANGPTL6 in keratinocytes was shown to promote epidermal proliferation, remodelling, regeneration, and angiogenesis. Notwithstanding the effects of ANGPTL6 overexpression in skin, endogenous expression of ANGPTL6 in skin has yet to be demonstrated [26,27]. More recent data provide very compelling evidence that ANGPTL6 is a powerful modulator of energy metabolism and adiposity . Mice that survived deletion of the Angptl6 gene showed marked obesity, characterized by elevated fat mass and fat-cell size, hypercholesterolaemia, elevated plasma NEFA, hyperinsulinaemia and glucose intolerance. The increased fat mass could be ascribed to decreased energy expenditure, and was unrelated to food intake. Experiments with transgenic mice overexpressing ANGPTL6 in liver and with mice infected with ANGPTL6-expressing adenovirus provided strong support for a role of ANGPTL6 in stimulating energy expenditure, and consequently diminishing fat mass. Based on the published data, it is unclear whether dysregulation of plasma insulin, cholesterol, and NEFA can be attributed to direct ANGPTL6 action or whether the observed changes are secondary to its effect on adiposity. Plasma TG levels do not appear to be influenced by ANGPTL6 overexpression or deletion, suggesting that ANGPTL6, in contrast to ANGPTL3 and ANGPTL4, does not inhibit LPL activity. Future research will need to address the mechanisms by which ANGPTL6 increases energy expenditure, as well as the tissues targeted by the protein.
Concluding remarks and future research
Despite their relatively low profile, there is a large body of evidence showing that ANGPTLs have major effects on lipid metabolism in mice. ANGPTL3 and ANGPTL4 have been unambiguously established as potent inhibitors of plasma TG clearance, causing elevation of plasma TG levels. Accordingly, they represent interesting candidate genes for the study of the genetic basis of familial hyperlipidaemia, as well as promising pharmacological targets for the treatment of dyslipidaemia.
Proteins in Disease: A Focus Topic at BioScience2005, held at SECC Glasgow, U.K., 17–21 July 2005. Edited by B. Austen (St George's Hospital Medical School, London, U.K.), C. Connolly (Dundee, U.K.), B. Irvine (Belfast, U.K.), M. Sugden (Queen Mary, London, U.K.) and V. Zammit (Hannah Research Institute, Ayr, U.K.).
Abbreviations: ANGPTL, angiopoietin-like protein; EST, expressed sequence tag; LPL, lipoprotein lipase; LXR, liver X receptor; NEFA, non-esterified fatty acid; PPAR, peroxisome-proliferator-activated receptor; TG, triacylglycerol; VLDL, very-low-density lipoprotein
- © 2005 The Biochemical Society