**5. Role of PPARβ/δ in lipoprotein metabolism**

Treatment of the atherogenic dyslipidemia associated with type 2 diabetes mellitus and metabolic syndrome requires lowering triglycerides, increasing HDL-C and increasing the size of the LDL-C particle. Studies using the PPARβ/δ agonist GW501516 have demonstrated that this drug increased HDL-C (79%), and decreased triglycerides (56%), LDL-C (29%) and fasting insulin levels (48%) in obese rhesus monkeys, a model for human obesity and its associated metabolic disorders (Oliver, Jr. et al., 2001). A decrease in small dense LDL was also observed in treated animals (Oliver, Jr. et al., 2001). It has been suggested that the increase in HDL-C levels after PPARβ/δ treatment is caused by enhanced cholesterol efflux stimulated by a higher expression of the reverse cholesterol transporter ATP-binding cassette A1 (ABCA1) in several tissues, including human and mouse macrophages and intestinal cells and fibroblasts (Leibowitz et al., 2000; van, V et al., 2005). Apart from these beneficial effects of PPARβ/δ activation on HDL levels, treatment with this compound also increased HDL particle size in primates (Wallace et al., 2005), an effect which is thought to be protective against the progression of coronary artery disease in humans (Rosenson et al., 2002). In addition, PPARβ/δ activation reduces cholesterol absorption through a mechanism that may involve, at least in part, reduced intestinal expression of Niemann-Pick C1-like 1 (Npc1l1), the proposed target for the inhibitor of cholesterol absorption ezetimibe (van, V et al., 2005). However, additional studies are necessary to clearly demonstrate that the effects of these drugs are mediated through PPARβ/δ activation.

In obese and diabetic db/db mice, administration of a PPARβ/δ agonist modestly increased HDL particles, without affecting triglyceride levels (Leibowitz et al., 2000), whereas in a shorter treatment with GW501516 a reduction in plasma free fatty acids and triglyceride levels was observed in db/db mice, but not in mice exposed to a high fat diet (Tanaka et al., 2003).

In mice, deletion of PPARβ/δ led to enhanced LDL and triglyceride levels (Akiyama et al., 2004). It has been proposed that the increase in triglycerides observed in these PPARβ/δnull mice is caused by a combination of increased VLDL production and decreased plasma triglyceride clearance, as demonstrated by a decrease in postheparin LPL activity and increased hepatic expression of the LPL inhibitors Angptl3 and 4 (Akiyama et al., 2004). Recent findings obtained by our laboratory indicate that additional mechanisms can also contribute to the hypotriglyceridemic effect of PPARβ/δ (Barroso et al., 2011). Interestingly, the main factor influencing hepatic triglyceride secretion is fatty acid availability (Lewis, 1997). In liver, fatty acids are either incorporated into triglycerides or oxidized by

Peroxisome Proliferator-Activated

leading to AMPK activation.

(Akiyama et al., 2004).

Receptor β/δ (PPARβ/δ) as a Potential Therapeutic Target for Dyslipidemia 221

GW501516 prevents LPS-induced ERK1/2 phosphorylation in adipocytes (Rodriguez-Calvo et al., 2008). It is important to note that a previous study found that obesity leads to increased hepatic ERK1/2 activity and that caloric restriction blunts this increase and improves insulin sensitivity (Zheng et al., 2009). In our study, the improvement in glucose tolerance caused by GW501516 was also accompanied by the reduction in phospho-ERK1/2 levels. An additional mechanism could involve SIRT1, since it has recently been reported that pharmacological PPARβ/δ activation increases the expression of SIRT1 (Okazaki et al., 2010), a deacetylase which regulates AMPK activity (Ruderman et al., 2010) through LKB1 acetylation (Lan et al., 2008), and might be essential to the regulatory loop involving PPARα, PGC-1α and Lipin 1 (Sugden et al., 2010). However, our findings made this possibility unlikely given that the increase in SIRT1 levels induced by GW501516 did not modify the acetylation status of LKB1. Interestingly, we showed that GW501516 increased the AMP/ATP ratio in liver, indicating that, in line with a previous study in skeletal muscle cells (Kramer et al., 2007), the underlying mechanism responsible for the increase in AMPK phosphorylation induced by this drug could be a modification of the cellular energy status. Previous studies have suggested that the reduction in ATP levels caused by GW501516 can be the result of a specific inhibition of one or more complexes of the respiratory chain, an effect on the ATP synthase system, or to mitochondrial uncoupling (Kramer et al., 2007). These potential changes would reduce the yield of ATP synthesis by the mitochondria,

In agreement with the reported regulation of PGC-1α (Canto et al., 2009; Jeninga et al., 2010; Lee et al., 2006a) and Lipin 1 (Higashida et al., 2008) by AMPK, exposure to the HFD reduced both *Pgc-1*α and *Lipin 1* expression. The reduction in Lipin 1 was likely to be the result of the decrease of PGC-1α, since it has been reported that genetic reduction of hepatic PGC-1α decreases the expression of *Lipin 1* (Estall et al., 2009). In addition, it has been shown that physiological stimuli that increase mitochondrial fatty acid oxidation induce *Pgc-1*α gene expression, which in turn activates the expression of *Lipin 1* (Finck et al., 2006). Interestingly, it has been reported that upregulation of *Lipin 1* in liver increases PPARα activity by two mechanisms: transcriptional activation of the *Pparα* gene and direct coactivation of PPARα in cooperation with PGC-1α (Finck et al., 2006). Thus, Lipin 1 is considered to be an inducible "booster" that amplifies pathways downstream PGC-1α-PPARα, mainly mitochondrial fatty acid oxidation (Finck et al., 2006). In agreement with this, GW501516 treatment prevented the reduction in PGC-1α, increased the nuclear protein levels of Lipin 1 and amplified the PGC-1αPPARα pathway, as demonstrated by the increase in the transcriptional activation of *Ppar*α and the increase in PPARα transcriptional activity. These effects subsequently enhanced hepatic fatty acid oxidation, as shown by the increase in -hydroxybutyrate levels. The reduction in PGC-1α and Lipin 1 levels caused by the HFD and their restoration after GW501516 treatment observed in our study might also contribute to the changes of plasma triglyceride levels, since both proteins are involved in the control of hepatic triglyceride secretion and fatty acid oxidation (Zhang et al., 2004; Chen et al., 2008; Estall et al., 2009). Overall, these data implicated PGC-1α and Lipin 1 in the hypotriglyceridemic effect of PPARβ/δ and complemented the findings of a previous study reporting that elevated plasma triglyceride levels in PPARβ/δ-null mouse were related to a combination of increased VLDL production and decreased plasma triglyceride clearance

mitochondrial -oxidation. An increase in fatty acid oxidation in liver would thus reduce the availability of fatty acids and subsequent hepatic triglyceride secretion. However, it was unknown whether the hypotriglyceridemic effect observed following PPAR/ activation involved increased hepatic fatty acid oxidation and the mechanisms implicated. The ratelimiting step for mitochondrial -oxidation is the transport of fatty acid into mitochondria by liver carnitine palmitoyltransferase-1 (CPT1a). This fatty acid transporter is under the control of both PPARs and AMP-activated protein kinase (AMPK), which detects low ATP levels and in turn increases oxidative metabolism (Zhang et al., 2009) by reducing the levels of malonyl-CoA. Interestingly, PPAR/ activation can increase the activity of AMPK and the increase in fatty acid oxidation in human skeletal muscle cells following GW501516 treatment is dependent on both PPAR/ and AMPK (Kramer et al., 2007). It is worth noting that a recent discovered protein, lipin 1, plays an important role in hepatic fatty acid oxidation since it determines whether fatty acids are incorporated into triglycerides or undergo mitochondrial -oxidation. In addition, the expression and compartmentalization of lipin 1 controls the secretion of hepatic triglycerides (Bou et al., 2009). Thus, in the cytoplasm, lipin 1 promotes triglyceride accumulation and phospholipid synthesis by functioning as an Mg2+-dependent phosphatidate phosphatase (phosphatidic acid phosphatase-1, PAP-1). In contrast, in the nucleus lipin 1 acts as a transcriptional coactivator linked to fatty acid oxidation by regulating the induction of PGC-1α-PPARα-target genes (Finck et al., 2006). Lipin 1 induces PPARα gene expression and forms a complex with PPARα and PGC-1α leading to the induction of genes involved in fatty acid oxidation, including *Cpt1a* and *Mcad* (medium chain acyl-CoA dehydrogenase ) (Finck et al., 2006). When we examined the effects a high-fat diet (HFD) on hypertriglyceridemia and on the hepatic fatty acid oxidation pathway, we observed that exposure to HFD caused hypertriglyceridemia that was accompanied by reduced hepatic mRNA levels of PGC-1α and lipin 1, reduced hepatic phospho-AMPK levels and increased activity of extracellularsignal-regulated kinase 1/2 (ERK1/2) (Figure 2). Interestingly, drug treatment reduced hypertriglyceridemia, and restored hepatic phosphorylated levels of AMPK and ERK1/2. GW501516 treatment increased nuclear lipin 1 protein levels, leading to amplification of the PGC-1α-PPARα signaling system, as demonstrated by the increase in PPARα levels and PPARα-DNA binding activity and the increased expression of PPARα-target genes involved in fatty acid oxidation. These effects of GW501516 were accompanied by an increase in plasma -hydroxybutyrate levels, demonstrating enhanced hepatic fatty acid oxidation. The maintenance of AMPK phosphorylation following GW501516 treatment was accompanied by the recovery in the expression levels of *Lipin 1* and *Pgc-1*α and the increase in the mRNA levels of the *Vldl receptor* (Figure 2). Although we cannot rule out direct transcriptional activation of these genes by PPARβ/δ since it has been suggested that *Lipin 1*, the *Vldl receptor* (Sanderson et al., 2010) and *Pgc-1*α (Hondares et al., 2007) might be PPARβ/δ-target genes, most effects of GW501516 might be the result of the increase in AMPK phosphorylation (Kramer et al., 2007). In fact, it has been reported that this kinase upregulates the expression of *Lipin 1* (Higashida et al., 2008), the *Vldl receptor* (Zenimaru et al., 2008) and *Pgc-1*α (Lee et al., 2006b). The increase in AMPK phosphorylation following GW501516 treatment might involve several mechanisms. Since inhibitory crosstalk between ERK1/2 and AMPK has been reported (Du et al., 2008), the increase in phospho-AMPK levels could be the result of the inhibition by GW501516 of the phosphorylation of ERK1/2 induced by the HFD, which is in agreement with our previous study reporting that

mitochondrial -oxidation. An increase in fatty acid oxidation in liver would thus reduce the availability of fatty acids and subsequent hepatic triglyceride secretion. However, it was unknown whether the hypotriglyceridemic effect observed following PPAR/ activation involved increased hepatic fatty acid oxidation and the mechanisms implicated. The ratelimiting step for mitochondrial -oxidation is the transport of fatty acid into mitochondria by liver carnitine palmitoyltransferase-1 (CPT1a). This fatty acid transporter is under the control of both PPARs and AMP-activated protein kinase (AMPK), which detects low ATP levels and in turn increases oxidative metabolism (Zhang et al., 2009) by reducing the levels of malonyl-CoA. Interestingly, PPAR/ activation can increase the activity of AMPK and the increase in fatty acid oxidation in human skeletal muscle cells following GW501516 treatment is dependent on both PPAR/ and AMPK (Kramer et al., 2007). It is worth noting that a recent discovered protein, lipin 1, plays an important role in hepatic fatty acid oxidation since it determines whether fatty acids are incorporated into triglycerides or undergo mitochondrial -oxidation. In addition, the expression and compartmentalization of lipin 1 controls the secretion of hepatic triglycerides (Bou et al., 2009). Thus, in the cytoplasm, lipin 1 promotes triglyceride accumulation and phospholipid synthesis by functioning as an Mg2+-dependent phosphatidate phosphatase (phosphatidic acid phosphatase-1, PAP-1). In contrast, in the nucleus lipin 1 acts as a transcriptional coactivator linked to fatty acid oxidation by regulating the induction of PGC-1α-PPARα-target genes (Finck et al., 2006). Lipin 1 induces PPARα gene expression and forms a complex with PPARα and PGC-1α leading to the induction of genes involved in fatty acid oxidation, including *Cpt1a* and *Mcad* (medium chain acyl-CoA dehydrogenase ) (Finck et al., 2006). When we examined the effects a high-fat diet (HFD) on hypertriglyceridemia and on the hepatic fatty acid oxidation pathway, we observed that exposure to HFD caused hypertriglyceridemia that was accompanied by reduced hepatic mRNA levels of PGC-1α and lipin 1, reduced hepatic phospho-AMPK levels and increased activity of extracellularsignal-regulated kinase 1/2 (ERK1/2) (Figure 2). Interestingly, drug treatment reduced hypertriglyceridemia, and restored hepatic phosphorylated levels of AMPK and ERK1/2. GW501516 treatment increased nuclear lipin 1 protein levels, leading to amplification of the PGC-1α-PPARα signaling system, as demonstrated by the increase in PPARα levels and PPARα-DNA binding activity and the increased expression of PPARα-target genes involved in fatty acid oxidation. These effects of GW501516 were accompanied by an increase in plasma -hydroxybutyrate levels, demonstrating enhanced hepatic fatty acid oxidation. The maintenance of AMPK phosphorylation following GW501516 treatment was accompanied by the recovery in the expression levels of *Lipin 1* and *Pgc-1*α and the increase in the mRNA levels of the *Vldl receptor* (Figure 2). Although we cannot rule out direct transcriptional activation of these genes by PPARβ/δ since it has been suggested that *Lipin 1*, the *Vldl receptor* (Sanderson et al., 2010) and *Pgc-1*α (Hondares et al., 2007) might be PPARβ/δ-target genes, most effects of GW501516 might be the result of the increase in AMPK phosphorylation (Kramer et al., 2007). In fact, it has been reported that this kinase upregulates the expression of *Lipin 1* (Higashida et al., 2008), the *Vldl receptor* (Zenimaru et al., 2008) and *Pgc-1*α (Lee et al., 2006b). The increase in AMPK phosphorylation following GW501516 treatment might involve several mechanisms. Since inhibitory crosstalk between ERK1/2 and AMPK has been reported (Du et al., 2008), the increase in phospho-AMPK levels could be the result of the inhibition by GW501516 of the phosphorylation of ERK1/2 induced by the HFD, which is in agreement with our previous study reporting that GW501516 prevents LPS-induced ERK1/2 phosphorylation in adipocytes (Rodriguez-Calvo et al., 2008). It is important to note that a previous study found that obesity leads to increased hepatic ERK1/2 activity and that caloric restriction blunts this increase and improves insulin sensitivity (Zheng et al., 2009). In our study, the improvement in glucose tolerance caused by GW501516 was also accompanied by the reduction in phospho-ERK1/2 levels. An additional mechanism could involve SIRT1, since it has recently been reported that pharmacological PPARβ/δ activation increases the expression of SIRT1 (Okazaki et al., 2010), a deacetylase which regulates AMPK activity (Ruderman et al., 2010) through LKB1 acetylation (Lan et al., 2008), and might be essential to the regulatory loop involving PPARα, PGC-1α and Lipin 1 (Sugden et al., 2010). However, our findings made this possibility unlikely given that the increase in SIRT1 levels induced by GW501516 did not modify the acetylation status of LKB1. Interestingly, we showed that GW501516 increased the AMP/ATP ratio in liver, indicating that, in line with a previous study in skeletal muscle cells (Kramer et al., 2007), the underlying mechanism responsible for the increase in AMPK phosphorylation induced by this drug could be a modification of the cellular energy status. Previous studies have suggested that the reduction in ATP levels caused by GW501516 can be the result of a specific inhibition of one or more complexes of the respiratory chain, an effect on the ATP synthase system, or to mitochondrial uncoupling (Kramer et al., 2007). These potential changes would reduce the yield of ATP synthesis by the mitochondria, leading to AMPK activation.

In agreement with the reported regulation of PGC-1α (Canto et al., 2009; Jeninga et al., 2010; Lee et al., 2006a) and Lipin 1 (Higashida et al., 2008) by AMPK, exposure to the HFD reduced both *Pgc-1*α and *Lipin 1* expression. The reduction in Lipin 1 was likely to be the result of the decrease of PGC-1α, since it has been reported that genetic reduction of hepatic PGC-1α decreases the expression of *Lipin 1* (Estall et al., 2009). In addition, it has been shown that physiological stimuli that increase mitochondrial fatty acid oxidation induce *Pgc-1*α gene expression, which in turn activates the expression of *Lipin 1* (Finck et al., 2006). Interestingly, it has been reported that upregulation of *Lipin 1* in liver increases PPARα activity by two mechanisms: transcriptional activation of the *Pparα* gene and direct coactivation of PPARα in cooperation with PGC-1α (Finck et al., 2006). Thus, Lipin 1 is considered to be an inducible "booster" that amplifies pathways downstream PGC-1α-PPARα, mainly mitochondrial fatty acid oxidation (Finck et al., 2006). In agreement with this, GW501516 treatment prevented the reduction in PGC-1α, increased the nuclear protein levels of Lipin 1 and amplified the PGC-1αPPARα pathway, as demonstrated by the increase in the transcriptional activation of *Ppar*α and the increase in PPARα transcriptional activity. These effects subsequently enhanced hepatic fatty acid oxidation, as shown by the increase in -hydroxybutyrate levels. The reduction in PGC-1α and Lipin 1 levels caused by the HFD and their restoration after GW501516 treatment observed in our study might also contribute to the changes of plasma triglyceride levels, since both proteins are involved in the control of hepatic triglyceride secretion and fatty acid oxidation (Zhang et al., 2004; Chen et al., 2008; Estall et al., 2009). Overall, these data implicated PGC-1α and Lipin 1 in the hypotriglyceridemic effect of PPARβ/δ and complemented the findings of a previous study reporting that elevated plasma triglyceride levels in PPARβ/δ-null mouse were related to a combination of increased VLDL production and decreased plasma triglyceride clearance (Akiyama et al., 2004).

Peroxisome Proliferator-Activated

Receptor β/δ (PPARβ/δ) as a Potential Therapeutic Target for Dyslipidemia 223

The increase in fatty acid oxidation caused by GW501516 was apparently inconsistent with its lack of effects on hepatic triglyceride levels observed in our study. Several reasons may account for this. First, similar to the effects of GW501516, which restores Lipin 1 levels, hepatic *Lipin 1* overexpression leads to increased liver triglyceride content (Finck et al., 2006). This apparently conflicts with the effects of Lipin 1 on fatty acid oxidation, but it has been explained by hepatic triglyceride sequestration secondary to diminished triglyceride secretion, increased fatty acid uptake, or the PAP activity of Lipin 1 (Finck et al., 2006). Second, in our study we reported an additional possibility, the increase caused by GW501516 in the expression of the *Vldl receptor* in liver. The huge increase of this receptor observed in liver after GW501516 treatment might also reduce plasma triglyceride levels by increasing VLDL uptake by the liver. However, this can also lead to an increase in hepatic triglyceride content. Third, it has been reported that GW501516 improves hyperglycemia by increasing glucose flux through the pentose phosphate pathway and enhancing fatty acid synthesis in liver (Lee et al., 2006a). In that study, GW501516 increased liver triglyceride content but the authors reported that although this might raise concerns that long-term drug treatment might cause hepatic steatosis, they did not observe signs of fatty liver with treatment up to 6 months. In addition, long-term GW501516 treatment has been shown to reduce body weight and levels of circulating and liver triglycerides (Wang et al., 2004; Tanaka et al., 2003). In summary, our findings indicated that PPAR/ activation by GW501516 amplified the PPARα-PGC1-α pathway through the restoration of AMPK

In humans, there are conflicting reports as to whether PPARβ/δ polymorphisms are associated with changes in plasma lipoproteins. Thus, while some studies found an association between a PPARβ/δ polymorphism and plasma lipids (Skogsberg et al., 2003), this was not confirmed in other studies (Gouni-Berthold et al., 2005). These discrepancies could be caused by differences in gender or the influence of gene-environment interactions, since a recent study reported that the association between the PPARβ/δ -87T>C polymorphism and plasma HDL-cholesterol might be sex-specific, women showing a stronger association, and that this association was only observed in subjects consuming a low-fat diet (Robitaille et al., 2007). The authors of this study concluded that the presence of the PPARβ/δ -87T>C polymorphism, which may result in enhanced PPARβ/δ activity, is associated with lower risk of suffering metabolic syndrome and that this association depends on the amount of fat consumed. In summary, the findings available at present on the effects of PPARβ/δ activation on lipoprotein metabolism are so promising that

PPARβ/δ drugs are now in clinical trials for the treatment of human dyslipidemia.

resistance may also contribute to ameliorate the atherogenic dyslipidemia.

**6.1 PPARβ/δ, inflammation and insulin resistance in adipose tissue** 

As stated above insulin resistance plays a crucial role in the development of hypertrigliceridemia, resulting in a sequence of lipoprotein changes leading to atherogenic dyslipidemia. Thus, those drugs, such as the PPARβ/δ ligands, which improve insulin

The expansion of adipose tissue, mainly in the form of visceral obesity, may contribute to enhanced inflammation in this tissue and insulin resistance through several processes. First, macrophages can infiltrate in adipose tissue, which contributes to the overproduction of

**6. Role of PPARβ/δ in insulin resistance** 

activity, contributing to the hypotriglyceridemic effect of this drug.

Fig. 2. A schematic of the potential effects of GW501516 (dashed lines) on liver metabolism is shown. Drug treatment with the PPAR/ agonist GW501516 prevents the reduction in phospho-AMPK levels and the subsequent increase in phospho-ERK1/2 levels caused by the HFD. In addition, GW501516 prevents the reduction in PGC-1α and increases Lipin 1 protein levels in the nucleus leading to amplification of the PPARα-PGC-1α pathway, which subsequently induces hepatic fatty acid oxidation. This pathway is additionally increased by GW501516 through the enhanced synthesis of the hepatic PPARα endogenous ligand 16:0/18:1-PC. As a result of the increase in this pathway the availability of fatty acids to be secreted as triglycerides might be compromised. The increase in the hepatic levels of the Vldl receptor can also contribute to reduce plasma triglyceride levels.

The data reported in our study also demonstrated that PPARβ/δ activation by GW501516 can amplify the PPARα pathway by an additional mechanism. Previous studies had demonstrated that hepatic fatty acid synthase (FAS) was necessary for the normal activation of PPARα target genes but did not identify the ligand involved in this process (Chakravarthy et al., 2005). Recently, this endogenous PPARα ligand was identified as 16:0/18:1-phosphatidilcholine (PC) (Chakravarthy et al., 2009). The synthesis of this ligand requires FAS activity, which yields palmitate (16:0), whereas 16:0/18:1-PC is generated through the enzymatic activity of CEPT1 (Chakravarthy et al., 2009). Subsequent binding of 16:0/18:1-PC to PPARα in the nucleus turns on PPARα-dependent genes and affects hepatic lipid metabolism. Interestingly, activation of PPARβ/δ by GW501516 induces FAS expression in liver as a result of increased glycolysis and the pentose phosphate shunt (Lee et al., 2006a). Our findings confirmed that GW501516 also increased *Cept1* expression and the levels of 16:0/18:1-PC, contributing to further amplification of the PPARα pathway.

Fig. 2. A schematic of the potential effects of GW501516 (dashed lines) on liver metabolism is shown. Drug treatment with the PPAR/ agonist GW501516 prevents the reduction in phospho-AMPK levels and the subsequent increase in phospho-ERK1/2 levels caused by the HFD. In addition, GW501516 prevents the reduction in PGC-1α and increases Lipin 1 protein levels in the nucleus leading to amplification of the PPARα-PGC-1α pathway, which subsequently induces hepatic fatty acid oxidation. This pathway is additionally increased by GW501516 through the enhanced synthesis of the hepatic PPARα endogenous ligand 16:0/18:1-PC. As a result of the increase in this pathway the availability of fatty acids to be secreted as triglycerides might be compromised. The increase in the hepatic levels of the

The data reported in our study also demonstrated that PPARβ/δ activation by GW501516 can amplify the PPARα pathway by an additional mechanism. Previous studies had demonstrated that hepatic fatty acid synthase (FAS) was necessary for the normal activation of PPARα target genes but did not identify the ligand involved in this process (Chakravarthy et al., 2005). Recently, this endogenous PPARα ligand was identified as 16:0/18:1-phosphatidilcholine (PC) (Chakravarthy et al., 2009). The synthesis of this ligand requires FAS activity, which yields palmitate (16:0), whereas 16:0/18:1-PC is generated through the enzymatic activity of CEPT1 (Chakravarthy et al., 2009). Subsequent binding of 16:0/18:1-PC to PPARα in the nucleus turns on PPARα-dependent genes and affects hepatic lipid metabolism. Interestingly, activation of PPARβ/δ by GW501516 induces FAS expression in liver as a result of increased glycolysis and the pentose phosphate shunt (Lee et al., 2006a). Our findings confirmed that GW501516 also increased *Cept1* expression and the levels of 16:0/18:1-PC, contributing to further amplification of the PPARα pathway.

Vldl receptor can also contribute to reduce plasma triglyceride levels.

The increase in fatty acid oxidation caused by GW501516 was apparently inconsistent with its lack of effects on hepatic triglyceride levels observed in our study. Several reasons may account for this. First, similar to the effects of GW501516, which restores Lipin 1 levels, hepatic *Lipin 1* overexpression leads to increased liver triglyceride content (Finck et al., 2006). This apparently conflicts with the effects of Lipin 1 on fatty acid oxidation, but it has been explained by hepatic triglyceride sequestration secondary to diminished triglyceride secretion, increased fatty acid uptake, or the PAP activity of Lipin 1 (Finck et al., 2006). Second, in our study we reported an additional possibility, the increase caused by GW501516 in the expression of the *Vldl receptor* in liver. The huge increase of this receptor observed in liver after GW501516 treatment might also reduce plasma triglyceride levels by increasing VLDL uptake by the liver. However, this can also lead to an increase in hepatic triglyceride content. Third, it has been reported that GW501516 improves hyperglycemia by increasing glucose flux through the pentose phosphate pathway and enhancing fatty acid synthesis in liver (Lee et al., 2006a). In that study, GW501516 increased liver triglyceride content but the authors reported that although this might raise concerns that long-term drug treatment might cause hepatic steatosis, they did not observe signs of fatty liver with treatment up to 6 months. In addition, long-term GW501516 treatment has been shown to reduce body weight and levels of circulating and liver triglycerides (Wang et al., 2004; Tanaka et al., 2003). In summary, our findings indicated that PPAR/ activation by GW501516 amplified the PPARα-PGC1-α pathway through the restoration of AMPK activity, contributing to the hypotriglyceridemic effect of this drug.

In humans, there are conflicting reports as to whether PPARβ/δ polymorphisms are associated with changes in plasma lipoproteins. Thus, while some studies found an association between a PPARβ/δ polymorphism and plasma lipids (Skogsberg et al., 2003), this was not confirmed in other studies (Gouni-Berthold et al., 2005). These discrepancies could be caused by differences in gender or the influence of gene-environment interactions, since a recent study reported that the association between the PPARβ/δ -87T>C polymorphism and plasma HDL-cholesterol might be sex-specific, women showing a stronger association, and that this association was only observed in subjects consuming a low-fat diet (Robitaille et al., 2007). The authors of this study concluded that the presence of the PPARβ/δ -87T>C polymorphism, which may result in enhanced PPARβ/δ activity, is associated with lower risk of suffering metabolic syndrome and that this association depends on the amount of fat consumed. In summary, the findings available at present on the effects of PPARβ/δ activation on lipoprotein metabolism are so promising that PPARβ/δ drugs are now in clinical trials for the treatment of human dyslipidemia.
