Peroxisome Proliferator-activated Receptor β/δ Regulates Very Low Density Lipoprotein Production and Catabolism in Mice on a Western Diet

The results of recent studies using selective agonists for peroxisome proliferator-activated receptor β (PPARβ) suggest that this receptor may have a role in regulating levels of serum lipids in animal models of obesity and insulin resistance. To further examine this possibility, serum lipid profiles of mice lacking a functional PPARβ receptor were determined. PPARβ-null mice maintained on either normal chow or a 10-week high fat (HF) diet, a condition that has been shown to induce insulin resistance and obesity in mice, have elevated levels of serum triglycerides primarily associated with very low density lipoprotein (VLDL) with no difference in either total cholesterol or phospholipids. Consistent with this finding, PPARβ-null mice on a HF-diet were shown to have an increased rate of hepatic VLDL production as well as lowered lipoprotein lipase activity in serum compared with wild-type controls. The latter parallels an increase in the hepatic expression of the genes encoding angiopoietin-like proteins 3 and 4 in PPARβ-null mice on a HF diet, both proteins of which have recently been shown to inhibit lipoprotein lipase (LPL) activity in vivo. Consistent with elevated VLDL production, a marked increase in plasma VLDL apoB48, -E, -AI, and -AII, as well as a sharp depletion of the hepatic lipid stores was also found in PPARβ-null mice. In addition, PPARβ-null mice on a HF diet were shown to have increased adiposity, despite lower total body weight. Together, these results indicate a clear role for PPARβ in regulating levels of serum triglycerides in mice on a high fat Western diet by modulating both VLDL production and LPL-mediated catabolism of VLDL-triglycerides and also suggest a potential therapeutic role for PPARβ in the improvement of serum lipids in the setting of metabolic syndrome.

The peroxisome proliferator-activated receptors (PPARs) 1 are members of the nuclear hormone receptor superfamily.
Three major isoforms of PPARs (␣, ␤/␦, and ␥) have been identified (1-3), each of which forms an obligate heterodimer with retinoid X receptor ␣ (4,5). PPAR-retinoid X receptor ␣ heterodimers selectively bind to degenerate direct repeats of the hexameric nucleotide sequence, AGGTCA, separated by 1 bp (DR1) called peroxisome proliferator response elements on PPAR-responsive target genes (2). Peroxisome proliferator response elements mediate the response to PPARs and have been identified in the promoters of a number of PPAR target genes involved in lipid metabolism and adipocyte function (4, 6 -8).
PPAR␣ mediates the response in rodents to a diverse group of chemicals called peroxisome proliferators, which include hypolipidemic drugs, plasticizers, synthetic fatty acids, and pesticides (9). Peroxisome proliferators increase the number and size of peroxisomes and cause significant hepatomegaly in susceptible rodent species. PPAR␣ target genes encode proteins involved in peroxisomal and mitochondrial fatty acid ␤-oxidation, fatty acid transport, fatty acid synthesis, fatty acid-binding, and apolipoproteins, thus demonstrating a central role for this receptor in lipid metabolism and transport (9 -12). The fibrate class of hypolipidemic drugs has been shown to lower serum triglycerides, while elevating high density lipoprotein (HDL) cholesterol in humans and mice, an effect that requires a functional PPAR␣ in mice (13).
PPAR␥ has been implicated in adipocyte differentiation (8,14), insulin resistance (15)(16)(17), macrophage foam cell formation (18), and in the mediation of HF diet-induced adipocyte hypertrophy (19). Ligands for PPAR␥ include synthetic insulin-sensitizing thiazolidinedione compounds (TZDs), such as troglitazone, rosiglitazone, and pioglitazone (the latter two are currently in use for the treatment of type 2 diabetes in humans), and natural ligands, such as 9-and 13-hydroxyoctadecanoic acid, 15-hydroxyeicosatetraenoic acid, and linoleic acid (20 -23). In addition to their beneficial effects on insulin sensitivity, TZDs have been shown to improve serum lipid profiles (24 -27). Given that the major cause of mortality in diabetic patients is atherosclerosis, TZDs are being studied for their effectiveness in preventing atherosclerosis.
Compared with PPAR␣ and PPAR␥, relatively little is known about the function of PPAR␤ (27), which is ubiquitously expressed. Fatty acids and fibrate drugs such as bezafibrate (28) were originally identified as PPAR␤ ligands, although these compounds are not specific for this receptor (29). The recent development of animal models (30 -32) and discovery of selective agonists for PPAR␤ have allowed the identification of novel roles for this receptor. For example, PPAR␤ may be involved in keratinocyte proliferation in response to 12-O-tetradecanoylphorbol 13-acetate treatment of skin (33). PPAR␤ was also suggested to be involved in the regulation of embryo implantation in the mouse (34). Furthermore, PPAR␤ is regulated by adenomatous polyposis coli (APC) through ␤-catenin (35).
Apart from these functions, it was suggested that PPAR␤ plays an important role in improving serum lipid profiles in both rodent and monkey species. For instance, a PPAR␤ agonist, L-165041, was shown to raise plasma cholesterol concentrations in insulin-resistant db/db mice primarily associated with HDL particles, while having no effect on either glucose or triglycerides (36). L-165041 also reduced LPL activity in white adipose tissue (WAT). In addition, a potent and selective PPAR␤ agonist, GW501516, produced a dose-dependent rise in serum HDL-cholesterol, concomitant with a reduction in the levels of small-dense LDL, fasting triglycerides, and fasting insulin in obese rhesus monkeys (37). GW501516 was also shown to promote reverse cholesterol transport by increasing the expression of ABCA1 in macrophages, fibroblasts, and intestinal cells as well as inducing apoA1-specific cholesterol efflux (37). Although reasons for discrepancies in the effects of these PPAR␤ agonists are not entirely clear, their beneficial effects on serum lipids indicate a potential for use in the treatment of patients with the so-called "metabolic syndrome," marked by traits such as obesity, insulin resistance, hypercholesterolemia, and hypertriglyceridemia.
To further examine potential roles of PPAR␤ receptor in serum lipid regulation, levels of serum lipids were determined in mice lacking a functional PPAR␤ receptor (32). In this report, it is shown that PPAR␤-null mice on a HF diet are hypertriglyceridemic as a result of elevated levels of serum triglycerides primarily associated with very low density lipoprotein (VLDL). These results indicate that this phenotype may primarily be due to increased VLDL production and decreased catabolism of VLDL triglycerides by LPL in the absence of PPAR␤.

MATERIALS AND METHODS
Animals and Diet-Homozygous PPAR␤-null male mice and their control wild-type littermates on a C57Bl6/N background were housed in a pathogen-free facility under standard 12-h light/12-h dark cycle and were allowed water ad libitum. 8-to 12-week-old animals were maintained on either a control diet containing 5% fat (CD) (NIH-07 diet; Zeigler Bros, Inc., Gardeners, PA) or placed for 10 weeks on an AIN-93G-modified 35% high fat (HF) diet (F4048, Bio-Serv, Frenchtown, NJ).
Measurement of Plasma Lipids and Lipoproteins-Fasting plasma samples were collected at 0, 6, and 10 weeks after initiation of the HF diet. After a 4-h fast, blood was obtained from the retro-orbital plexus of mice anesthetized with methoxyflurane (Pitman-Moore, Mundelheim, IL), placed into pre-cooled tubes containing EDTA (final concentration 4 mM) and centrifuged at 2500 ϫ g for 20 min at 4°C. Plasma aliquots were stored at Ϫ70°C. Total cholesterol (TC) and triglycerides (TGs) (Sigma), as well as free cholesterol (FC) and phospholipid (PL) (Wako, Osaka, Japan) concentrations were measured in 12-l aliquots of plasma using commercial kits and the Hitachi 911 automated chemistry analyzer (Roche Applied Science). Plasma free fatty acids (FFA) were measured using the NEFA C kit (Wako Chemicals, Richmond VA). Fast protein liquid chromatography (FPLC) separation of serum lipoproteins from pooled plasma samples (100 l; n ϭ 5) was achieved by gel filtration using two Superose 6HR 10/30 columns (Amersham Biosciences) in series as previously described (38).
Hepatic TG and TC Assay-For liver TG and TC concentrations, total lipids were extracted as follows. Hepatic TG and TC were extracted from 0.  0.375 ml of chloroform:methanol (1:2) was then added. Following vortexing, the mixture was kept at room temperature for 5 min. 0.125 ml of chloroform was put in this tube and vortexed. Next, 0.125 ml of water was added followed by vortexing again. After centrifugation at 16,000 ϫ g for 15 min, the organic layer was separated into another tube. For complete extraction of liver lipid, an additional extraction was repeated using 100 l of chloroform. Lipids dissolved in organic solvents were dried down and re-suspended in the enzymatic kit buffers prior to TG and TC assays performed using kits (Sigma) as described above.
In Vivo TG Production and Lipoprotein Lipase and Hepatic Lipase Activities-Mice were injected with tyloxapol (500 g/g body weight; Sigma). Plasma VLDL/Chylomicron-TG clearance in mice is completely inhibited under these conditions. Blood samples were taken from each mouse at 0, 30, 90, and 150 min after injection. The accumulation of newly synthesized VLDL/Chylomicron-TG was measured in plasma aliquots. Post-heparin plasma lipoprotein lipase (LPL) and hepatic lipase (HL) activities were assayed in triplicate using 14 C-labeled triolein substrate in 5 M and 1 M NaCl as previously described (39).
Determination of Hepatic Glycogen Content-Hepatic glycogen was purified from whole liver according to an earlier report (40). Frozen tissue samples (50 -150 mg) were digested in a suitable volume of 30% KOH for 10 min at 100°C. Digestion was completed by the addition of one-fifth volume of 20% NaSO 4 . Macromolecules were precipitated by the addition of two volumes of absolute ethanol and overnight incubation at 20°C. Macromolecules were washed with 70% ethanol and dried, and glycogen was then hydrolyzed by digestion in 500 l of 4 N sulfuric acid for 10 min at 100°C. Samples were then neutralized with the addition of an equivalent volume of 4 N NaOH. The glucose resulting from glycogen hydrolysis was measured using an enzymatic kit (Sigma).
Northern Blot Analysis-Total RNA was isolated from the livers, epididymal white adipose, and muscle of PPAR␤-null and wild-type control mice by the acidic guanidine isocyanate/phenol/chloroform extraction method using the Ultraspec RNA isolation system (Biotecx Laboratories, Houston, TX). Total RNA was quantitated by optical densitometry at 260 nm and subjected to electrophoresis on a 1.0% agarose gel containing 7% formaldehyde in 20 mM MOPS, 8 mM sodium acetate, 1 mM EDTA buffer (pH 7.0) and transferred to Gene Screen Plus hybridization transfer membranes (PerkinElmer Life Sciences, Boston, MA). The cDNA probes used for Northern blotting included bifunctional enzyme, acyl-CoA oxidase, and thiolase (43), fatty acid translocase and apoCIII, glycerol-3-phosphate acyltransferase, stearoyl-CoA desaturase-1, sterol response element binding protein 1c, microsomal triglyceride transfer protein, hepatic lipase (HL) (44), fatty acid synthase (11), lipoprotein lipase (LPL), acetyl-CoA carboxylase (11), liver fatty acid-binding protein (L-FABP) (provided by Sam Sorof, Institute for Cancer Research, Fox Chase Cancer Center, Philadelphia, PA), apoAIV (provided by Kiyoto Motojima, Toho University, Tokyo, Japan), and L-carnitine palmitoyl transferase 1 (provided by Daniel Kelly, Washington University School of Medicine, St. Louis, MO). The probes for apoAV, angiopoietin-like protein 3 (Angptl4), angiopoietinlike protein 4 (Angptl3), and 36B4 cDNAs were amplified from a mouse liver cDNA library using gene-specific primers and cloned into pCR TOPO II (Invitrogen, Carlsbad, CA). The identity of the probes was confirmed by DNA sequencing. The intensities of hybridizing mRNAs were quantified by using ImageQuaNT and normalized to levels of ␤-actin mRNA.

Elevated Plasma TG and FFA Levels in PPAR␤-null Mice-
On a normal chow diet, PPAR␤-null mice were shown to exhibit elevated plasma TG and FFA but similar TC, FC, and PL levels compared with their wild-type controls (Table I). Following a HF dietary challenge for a period of 10 weeks, PL, TC, and FFA levels were dramatically increased but not significantly differ-
FIG. 4. Northern blot analysis of total RNA from liver of wild-type (؉/؉) and PPAR␤-null (؊/؊) mice on a HF diet. Total RNA (10 g) was denatured and electrophoresed in formaldehyde-containing 1% agarose gels, blotted onto nylon membranes, and probed with the indicated cDNA probes for angiopoietin-like protein 3 (Angptl3), Angptl4 and 36B4. Angptl3 and Angptl4 mRNA levels in PPAR␤-null mice were quantified relative to 36B4 expression and reported as a percentage of wild-type (ϩ/ϩ) levels. Data are mean Ϯ S.E., n ϭ 5 for both groups. *, p Ͻ 0.05 and ** p Ͻ 0.01 significantly different from respective wild-type (ϩ/ϩ) controls. ent between PPAR␤-null and control mice. However, TG levels remained significantly elevated in PPAR␤-null mice compared with their respective controls (Table I). The results of FPLC analysis indicates that the increase in plasma TG in PPAR␤null mice on a HF diet is primarily due to an increase in plasma VLDL levels, whereas TC FPLC profiles did not differ (Fig. 1). By Western blot analysis, the apolipoprotein content of the VLDL particles was also dramatically increased in PPAR␤-null mice; apoB48, apoE, apoA-I, and apoA-II were only detectable in PPAR␤-null VLDL (Fig. 1, inset).
The results of Western blot analysis show a significant increase in the hepatic expression of the LDLr and LRP (the two major receptors for apoB-containing lipoproteins) but not of the HDL receptor SRBI, in PPAR␤-null mice compared with controls (Fig. 2). This increase in LDLr and LRP paralleled that of their preferential ligand, apoE, in the plasma of PPAR␤-null mice versus controls.
PPAR␤ Deficiency Alters VLDL Catabolism-To elucidate possible mechanisms by which PPAR␤ deficiency results in increased plasma VLDL-TG levels, mice fed a HF diet for 10 weeks were injected with heparin, and their post-heparin LPL and HL plasma activities measured. Compared with controls, PPAR␤-null mice were shown to have a significant decrease in post heparin LPL activity, with no statistically significant change in HL activity (Fig. 3). The results of recent studies (45)(46)(47)(48) indicate that one of the factors that regulates LPL activity is angiopoietin-like proteins 3 (Ang-L3) and Ang-L4. Intravenous injection of recombinant Ang-L3 or Ang-L4 protein in mice inhibited LPL activity and concomitantly increased serum TGs (46,47). Angptl3 and Angptl4 mRNA expression were unchanged in PPAR␤-null mice on a normal chow diet (data not shown). On the other hand, consistent with hypertriglyceridemia and lowered LPL activity, the expression of the genes encoding Angptl3 and Angptl4 were elevated in the livers of PPAR␤-null mice on a HF diet (Fig. 4). Hepatic expression of the gene encoding apoCIII, another major inhibitor of LPL, was unchanged in PPAR␤-null mice (Fig. 5A).
The expression of genes encoding several key proteins involved in lipid catabolism was also examined. However, the levels of expression of the genes encoding LPL in white adipose tissue and HL in liver were not significantly different between genotypes (Fig. 5A). Furthermore, the expression in liver, muscle, and adipose (not shown) of genes encoding enzymes involved in the peroxisomal ␤-oxidation of TG fatty acids, a pathway induced by the HF diet, comprised of acyl-CoA oxidase, bifunctional enzyme (enoyl-CoA hydratase/hydroxyacyl-CoA dehydrogenase), thiolase, and straight-chain acyl-CoA oxidase that catalyzes the first, rate-limiting step of peroxisomal ␤-oxidation of very long chain fatty acids (49), were all similar among PPAR␤-null and control mice (Fig. 5B). Together, these data suggest that increased hepatic Ang-L3 and Ang-L4 expression may contribute to elevated VLDL-TGs by inhibiting LPL activity in PPAR␤-null mice on a HF diet.
PPAR␤ Deficiency Results in Increased Hepatic VLDL Secretion Rate-To further explore mechanisms that contribute to hypertriglyceridemia, VLDL-TG synthetic pathways in fasted PPAR␤-null and control mice previously fed a HF diet for 10 weeks were examined. Under these conditions, TG entering the plasma compartment is predominantly liver-derived and associated with VLDL. The hepatic VLDL-TG secretion rate was measured after injection of tyloxapol, which blocks the activity of plasma lipases. Consistent with increased plasma TG levels, the hepatic VLDL-TG production rate was significantly increased in PPAR␤-null mice on a HF diet compared with their control littermates (Fig. 6A). To determine whether the elevation in VLDL secretion rate could be explained, in part, by changes in the expression of genes encoding proteins involved in triglyceride synthesis, a series of Northern blot analyses using selected cDNA probes for genes associated with fatty acid synthesis was performed. Despite a dramatic increase in hepatic VLDL-TG secretion rate, the expression of genes encoding proteins involved in lipogenesis, including fatty acid synthase, acetyl-CoA carboxylase, nuclear lipogenic protein S14, sterol response element binding protein 1, steroyl-CoA desaturase 1, and glycerol-3-phosphate acyltransferase were not significantly different between PPAR␤-null mice and their controls (Fig. 6B). The expression levels of additional genes encoding liver proteins involved in, or associated with, TG/fatty acid metabolism, including apoAIV, apoAV, liver carnitine palmitoyl-transferase I, liver fatty acid-binding protein (L-FABP), and CD36/fatty acid translocase were also examined but not FIG. 5. A, Northern blot analysis of total RNA from liver and WAT of wild-type (ϩ/ϩ) and PPAR␤-null (Ϫ/Ϫ) mice on a HF diet. Total RNA (10 g) was denatured and electrophoresed in formaldehyde-containing 1% agarose gels, blotted onto nylon membranes, and probed with the indicated cDNA probes for hepatic lipase (HL), and apoCIII, lipoprotein lipase (LPL), and ␤-actin. B, Northern blot analysis of total RNA from liver and muscle of wild-type (ϩ/ϩ) and PPAR␤-null (Ϫ/Ϫ) mice on a HF diet. Total RNA (10 g) was denatured and electrophoresed in formaldehyde-containing 1% agarose gels, blotted onto nylon membranes, and probed with the indicated cDNA probes for acyl CoA oxidase (AOX), bifunctional enzyme (BIEN), thiolase (Thiol), and ␤-actin.
found to be significantly different among groups (Fig. 7). In addition, the mRNA encoding microsomal triglyceride transfer protein, which plays a critical role in the assembly of VLDL by transferring lipoprotein lipids to nascent apoB, was not found to be significantly different between wild-type and PPAR␤-null mice (Fig. 7).
PPAR␤ Deficiency Results in Hepatic Lipid Depletion in Mice on a HF Diet-On a normal chow diet, the levels of triglycerides, cholesterol, and glycogen in livers of PPAR␤-null and control mice were similar. In contrast, hepatic triglyceride and cholesterol levels were dramatically reduced in PPAR␤-null mice on a HF diet versus controls (Fig. 8). Hepatic glycogen was increased in the PPAR␤-null mice but the increase did not reach statistical significance (p Ͻ 0.09). The attenuation in hepatic lipid levels is consistent with the increased output of VLDL from the livers of PPAR␤-null mice on a HF diet.
PPAR␤-null Mice on an HF Diet Exhibit Lowered Body Weight and Increased Adiposity-PPAR␤-null mice on normal chow have lowered total body weight, consistent with a previous report (32) using the same mouse model (Fig. 9A). This difference in body weight between genotypes is maintained throughout a 10-week HF dietary challenge, and the rate of weight gain is similar between genotypes. This contrasts re-sults from an earlier study showing that PPAR␤-null mice exhibit an increased rate of weight gain on a HF diet (31). Possible reasons for this disparity between studies may include differences in genetic background of the mice and the design of the targeted allele of the PPAR␤ gene itself. PPAR␤-null mice either on normal chow or HF diet exhibit no overt signs of disease that might contribute to the lower body weight. To determine whether this difference in body weight could be attributed to food consumption, food intake levels were measured. Despite lowered body weight as compared with wild-type mice, PPAR␤-null mice had increased food intake (mean grams/d/g of body weight Ϯ S.E.: 0.38 Ϯ 0.04 versus 0.84 Ϯ 0.06, p Ͻ 0.001 for wild-type and PPAR␤-null mice, respectively). This suggests a possible role for PPAR␤ in the control of metabolic rate. Consistent with this notion, it was previously reported (31) that transgenic mice that overexpressed PPAR␤ specifically in adipose have dramatically reduced adiposity. Conversely, PPAR␤-null mice on HF diet were shown to have increased adiposity, exhibited by significant increases in the weight of white and brown adipose tissue as well as a slight degree of hepatomegaly (Fig. 9B). The reason for this hepatomegaly is not due to increased hepatic triglycerides and cholesterol, which actually decreased in the PPAR␤-null livers. It may be due to elevated glycogen contents. DISCUSSION A goal of this study was to further examine a potential role of PPAR␤ in regulating plasma lipids under normal and dietaryinduced conditions of obesity and insulin resistance using a gene knockout approach. To this end, the plasma lipoprotein profiles of mice lacking a functional PPAR␤ receptor (32) were measured on normal chow and following 10 weeks on a HF diet. Our results show that PPAR␤-null mice on either normal chow and HF diet have a pronounced hypertriglyceridemia that was almost exclusively associated with higher VLDL-TG. Interestingly, an elevation in plasma FFAs that was found in PPAR␤null mice on normal chow was normalized following the HF diet. On the other hand, TC and PL levels in serum were not affected by the absence of PPAR␤ under both dietary conditions. Recent reports (45)(46)(47)(48) have demonstrated that Angptl3 and Angptl4, which are synthesized and secreted by liver, inhibit LPL (but not HL) activity (which is the primary means of catabolism of TGs in VLDL into FFAs) distally in the plasma FIG. 6. A, rates of VLDL export in wild-type (ϩ/ϩ) and PPAR␤-null (Ϫ/Ϫ) mice fed a HF diet. Mice were injected with tyloxapol (500 g/g body weight; Sigma). Plasma VLDL/Chylomicron-TG clearance in mice is completely inhibited under these conditions. Blood samples were taken from each mouse at 0, 30, 90, and 150 min after injection. The accumulation of newly synthesized VLDL/Chylomicron-TG was measured in plasma aliquots. Total number of mice in each group given in parentheses. Values represent mean Ϯ S.D.; *, p Ͻ 0.05. B, Northern blot analysis of total RNA from liver of wild-type (ϩ/ϩ) and PPAR␤-null (Ϫ/Ϫ) mice fed a HF diet. Total liver RNA (10 g) was denatured and electrophoresed in formaldehyde-containing 1% agarose gels, blotted onto nylon membranes, and probed with the indicated cDNA probes for fatty acid synthase (FAS), acyl-CoA carboxylase (ACC), spot 14 (S14), sterol regulatory element-binding protein 1 (SREBP1), stearoyl-CoA desaturase-1 (SCD1), glycerol-3-phosphate acyltransferase (GPAT), and ␤-actin. compartment. Consistent with the fact that the hepatic expression of the genes encoding Angptl3 and Angptl4 were markedly increased in PPAR␤-null mice on HF diet, our results show that LPL (but not HL) activity is lowered. These data suggest that, in the absence of PPAR␤, the resulting increase in Ang-L3 and Ang-L4 expression causes an attenuation of LPL activity that leads to impaired triglyceride catabolism and thus hypertriglyceridemia, a possibility that remains to be examined. The fact that hepatic Angptl3 and Angptl4 gene expression was elevated in PPAR␤-null mice on the HF diet only (but unchanged on normal chow), suggest that the attenuation in LPL activity under this dietary condition results in decreased catabolism of TGs into FFAs, thus helping to normalize FFAs.
In addition to reduced LPL activity, the markedly increased rate of VLDL-TG secretion in the plasma compartment of PPAR␤-null mice on a HF diet compared with controls indicates a potential role for PPAR␤ in inhibiting hepatic VLDL secretion rate in response to a HF diet. Consistent with the elevated VLDL output from the livers of PPAR␤-null mice on a HF diet, hepatic triglyceride and cholesterol levels were dramatically reduced, indicating a depletion of hepatic lipid stores. This increase in VLDL production was also accompanied by elevated VLDL apoB48, apoE, apoAI, and apoAII. The increased levels of VLDL in PPAR␤-null mice may be associated with greater atherogenic risk, because aortic lesion size has been positively correlated with VLDL plasma retention time in mice (50,51) and proposed to be the same in humans (52,53). Interestingly, the expression of the LDLr and LRP were also increased in the livers of PPAR␤-null mice compared with controls. The increase of the two major hepatic receptors for apoB-containing lipoproteins may be a mechanism by which PPAR␤-null livers compensate for depleted lipid stores resulting from enhanced hepatic VLDL secretion, a possibility that remains to be examined. Similar compensatory changes in the expression of genes that increase hepatic cholesterol have been observed in mice overproducing liver derived HDL (54). Recently, it was reported that VLDL-TG hydrolyzed by LPL strongly activates the expression of a series of "lipid storage" genes in macrophages via PPAR␤, indicating that PPAR␤ is a VLDL sensor in these cells (55). It is conceivable that, in the absence of PPAR␤, extrahepatic cells such as macrophages, are unable to properly detect or uptake lipids found in VLDL and the increased hepatic production of VLDL represents a compensatory response. Despite the dramatic increase in VLDL-TG secretion rate, the expression levels of the major genes involved in fatty acid synthesis and in VLDL-TG secretion were not altered in the livers of PPAR␤-null versus control mice. The possibility that intestinally derived VLDL may account, at least in part, for increased VLDL-TG production in PPAR␤-null mice cannot be ruled out, even upon fasting conditions. Consistent with the observed hypertriglyceridemia in PPAR␤-null mice, the selective PPAR␤ agonist GW501516 was shown to lower serum TG (mostly VLDL-TG) in obese rhesus monkeys (37). On the other hand, treatment of obese and diabetic db/db mice with another PPAR␤ agonist (L-165041) did not affect the plasma TG levels but raised plasma HDL cholesterol levels (36). GW501516 also increased HDL cholesterol levels in obese monkeys. In contrast, our results indicate that genetic ablation of the PPAR␤ gene does not alter levels of serum cholesterol in mice. The differences between these studies may result from species differences, from variable specificity for PPAR␤ among agonists, and/or from indirect effects related to leptin receptor deficiency in db/db mice. In general, the physiological effects of receptor agonists may not necessarily be in direct opposition to the effects of gene deletion of the receptor. PPAR␤-null mice on either normal chow or HF diet exhibit lowered total body weight in the absence of overt signs of toxicity. The weights of individual tissues, however, including liver, white adipose, and brown adipose stores, were relatively higher in PPAR␤-null mice on a HF diet as compared with similarly treated wild-type mice. The diminution in total body weight cannot, however, be accounted for by dietary intake, which was actually found to be markedly elevated in PPAR␤null mice on a HF diet. Recent studies have reported a potential role for PPAR␤ in promoting fatty acid oxidation in muscle and adipose (31,56). Transgenic mice that overexpress a constitutively activated ligand-independent PPAR␤-VP16 fusion protein specifically in adipose were shown to have dramatically reduced adiposity due to increased fatty acid oxidation within this depot (31). These findings are consistent with increased adiposity in PPAR␤-null mice on a HF diet. It is noteworthy that hypotriglyceridemia was observed in this model of targeted activation of PPAR␤ (31), consistent with elevated triglycerides in our PPAR␤-null mouse model, further pointing to a role for this receptor in lipid regulation. In another study, the PPAR␤ agonist GW501516 administered to mice fed a high fat diet lowered the extent of diet-induced obesity and insulin resistance, an effect accompanied by enhanced fatty acid ␤-oxidation, proliferation of mitochondria, and a marked reduction of lipid droplets in skeletal muscles (56). Impairment of fatty acid oxidation within skeletal muscle of the PPAR␤-null mice used in this study, if present, might also be expected to contribute to elevated triglycerides and FFAs. In general, improper utilization of lipids as an energy source may also lead to stunted growth, despite increased dietary intake.
In summary, these results indicate that PPAR␤ deficiency contributes to VLDL hypertriglyceridemia by increasing the hepatic production of VLDL as well as decreasing LPL-mediated catabolism upon dietary conditions of obesity and insulin resistance. The latter effect is at least in part due to increased hepatic expression of Ang-L3 and Ang-L4 in PPAR␤-null mice. These combined studies suggest that PPAR␤ plays an important role in the metabolism of VLDL-TG, particularly in the setting of metabolic syndrome, as previously suggested (37). Further work is needed to better understand the exact coordinate function that VLDL and PPAR␤ may exert in the atherosclerotic process in vivo.