Loss of hepatic PPARα promotes inflammation and serum hyperlipidemia in diet-induced obesity

Abstract

Agonists for PPARα are used clinically to reduce triglycerides and improve high-density lipoprotein (HDL) cholesterol levels in patients with hyperlipidemia. Whether the mechanism of PPARα activation to lower serum lipids occurs in the liver or other tissues is unknown. To determine the function of hepatic PPARα on lipid profiles in diet-induced obese mice, we placed hepatocyte-specific peroxisome proliferator-activated receptor-α (PPARα) knockout (Ppara^HepKO) and wild-type (Ppara^fl/fl) mice on high-fat diet (HFD) or normal fat diet (NFD) for 12 wk. There was no significant difference in weight gain, percent body fat mass, or percent body lean mass between the groups of mice in response to HFD or NFD. Interestingly, the Ppara^HepKO mice on HFD had worsened hepatic inflammation and a significant shift in the proinflammatory M1 macrophage population. These changes were associated with higher hepatic fat mass and decreased hepatic lean mass in the Pparα^HepKO on HFD but not in NFD as measured by Oil Red O and noninvasive EchoMRI analysis (31.1 ± 2.8 vs. 20.2 ± 1.5, 66.6 ± 2.5 vs. 76.4 ± 1.5%, P < 0.05). We did find that this was related to significantly reduced peroxisomal gene function and lower plasma β-hydroxybutyrate in the Ppara^HepKO on HFD, indicative of reduced metabolism of fats in the liver. Together, these provoked higher plasma triglyceride and apolipoprotein B100 levels in the Ppara^HepKO mice compared with Ppara^fl/fl on HFD. These data indicate that hepatic PPARα functions to control inflammation and liver triglyceride accumulation that prevent hyperlipidemia.

INTRODUCTION

Because of the epidemic of obesity, hyperlipidemia has significantly increased over the past several decades and has led to the expansion of other related disorders, such as nonalcoholic fatty liver disease (NAFLD) and insulin-resistant type II diabetes (4, 69). NAFLD has been shown to worsen with hepatic inflammation (4, 56, 69), which also causes hyperlipidemia and glucose intolerance (4, 56). Hyperlipidemia is characterized by abnormally high levels of lipids (triglycerides and cholesterol) in the blood and can lead to complications like cardiovascular disease and stroke, reducing life expectancy (44). Current research for new therapeutic options is focused on targeting fat burning and inflammatory pathways (4, 56). Drugs such as statins and fibrates are commonly prescribed clinically for the treatment of hyperlipidemia. Fibrates activate the nuclear receptor peroxisome proliferator-activated receptor-α (PPARα), which is a transcription factor that regulates numerous genes involved in fatty acid metabolism. Fibrates are effective at reducing plasma triglyceride levels (61), possibly by lowering plasma apolipoprotein B100 (ApoB100) (35, 57). ApoB100 is essential for very-low-density lipoprotein (VLDL) excretion from the liver (8) that carries triglycerides and cholesterol out of the liver to the blood. In weight gain and obesity, plasma ApoB100 and triglyceride levels are increased (14), which indicates hypersecretion of VLDL and possibly the development of hyperlipidemia (8, 66). Reducing either NAFLD or fat mass lowers hyperlipidemia, but whether PPARα mediates these effects via the liver, adipose, muscle, or unknown tissue is yet to be determined.

PPARα is highly expressed in the liver where it plays an essential function in lipid metabolism, especially in response to fasting (1, 6, 21, 24, 25, 29, 62). However, it is also present in other tissues, such as brown adipose tissue (BAT) and white adipose tissue (WAT), where it mediates mitochondrial function and the browning of WAT for the burning of fat (50–52, 67). However, others have shown, using global PPARα knockout (PPARα ΚΟ) mice that PPARα is dispensable for cold-induced browning and BAT function, which was reported to occur by another isoform, PPARγ (11, 32). While the critical role of PPARα in the metabolic response in the liver, WAT, and BAT has been well studied, tissue-specific knockout of PPARα has not been well established.

Whole body PPARα knockout mice develop obesity in a sexually dimorphic fashion while additional studies on whole body knockout mice on Sv/129 or C57BL/6N genetic backgrounds observed no effect of global PPARα deficiency on the development of obesity (2, 9). Further studies by Kim et al. (30) demonstrated that mice deficient for PPARα on a mixed Sv/129/C57BL/6N background gained more weight in response to high-fat feeding as compared with wild-type control mice. The role of PPARα in the development of insulin resistance has been contentious. One study in global PPARα KO mice reported they are resistant to the development of high-fat-induced insulin resistance and hyperglycemia, while another study showed rodents with a PPARα deletion developed insulin resistance (18, 19). One possible explanation for the discrepancy in the observed phenotypes between these studies is the differences in methods used to determine insulin sensitivity with one study measuring in the fasted state and the other in the fed state (18, 19). Regardless, studies in tissue-specific PPARα KO mice are needed to reveal which tissue the nuclear receptor is responsible for improving hyperlipidemia. The effects may be tissue specific, or factors such as its role in insulin sensitivity may be compounding or from another set of tissues.

To elucidate the role of PPARα in tissues, floxed PPARα mice have been recently generated (43). We utilized hepatocyte-specific PPARα KO (Ppara^HepKO) mice to determine the capacity of hepatic PPARα in mitigating hyperlipidemia from diet-induced obesity. We examined the effect of hepatocyte PPARα on the plasma lipid and metabolite profile in response to high-fat (HFD) and normal fat (NFD) diets. We found that the Ppara^HepKO mice had significantly more inflammation, plasma triglycerides, and ApoB100 on HFD, which was associated with exacerbated hepatic steatosis in these mice. Our results demonstrate a principle role for PPARα in the protection against hepatic inflammation that leads to hyperlipidemia and fatty liver disease.

METHODS

Animals.

The experimental procedures and protocols of this study conform to the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committee of the University of Mississippi Medical Center. All mice had free access to food and water ad libitum. Animals were housed in a temperature-controlled environment with 12/12-h dark/light cycle. Ppara^HepKO and Ppara^fl/fl mice were obtained from Dr. Walter Wahli, Lee Kong Chian School of Medicine, Nanyang Technological University the Academia, Singapore as originally described (43). Albumin-Cre mice were used for Ppara^HepKO and obtained from Dr. Wahli. The mice were housed under standard temperatures between 24 and 25°C. The control NFD consisted of 17% fat diet (Teklad 22/5 rodent diet, cat. no. 860; Harland Laboratories, Indianapolis, IN). Studies were performed on 8-wk-old male mice initially housed under standard conditions with full access to standard mouse chow described above and water. After this time, mice were switched with full access to a 60% HFD (cat. no. D12492; Research Diets, New Brunswick, NJ) for 12 wk and allowed access to water as previously described (1, 24, 40, 62). Mice were fasted 8 h before euthanasia via isoflurane anesthesia and blood and tissues were extracted for further analysis.

Body and liver composition.

Body composition changes were assessed at 4-wk intervals throughout the study using magnetic resonance imaging (EchoMRI-900TM; Echo Medical System, Houston, TX). MRI measurements were performed in conscious mice placed in a thin-walled plastic cylinder with a cylindrical plastic insert added to limit movement of the mice. Mice were briefly submitted to a low-intensity electromagnetic field where fat mass, lean mass, free water, and total water were measured. Liver composition was measured in excised tissues in the same EchoMRI system at the end of 12 wk of HFD, immediately after euthanasia.

Liver triglyceride measurement.

Triglycerides were measured from 100 mg of liver tissue homogenized in 1 ml of 5% NP-40 in water. Homogenized tissues were then heated to 95°C for 5 min and then centrifuged (13,000 g) for 2 min. Tissue triglyceride levels were measured using a fluorometric assay kit according to the manufacturer’s guidelines (PicoProbe Trigylceride Fluorometric Assay Kit; BioVision, Milpitas, CA). Samples from individual mice were run in duplicate and averaged. The averages of individual mice were then used to obtain group averages.

Liver staining.

To determine the effects of treatment on hepatic lipid accumulation, livers were mounted and frozen in Tissue-Tek OCT and sectioned at 10 µm. Frozen sections were air dried and fixed in 10% neutral buffered formalin. Sections were briefly rinsed in tap water, followed by 60% isopropanol, and stained for 15 min in Oil Red O solution. Sections were further rinsed in 60% isopropanol and nuclei stained with hematoxylin followed by aqueous mounting and coverslipping. The degree of Oil Red O staining was determined at ×40 magnification using a color Axiocam 105 camera with Zen 2 software attached to a Zeiss microscope. Images were analyzed using ImageJ software, and to ensure the accuracy of measurement, six images of each animal were analyzed and averaged into a single measure. Measurements were obtained from three individual animals per group. Data are presented as the average mean ± SE of the percent Oil Red O staining for each group. Hemotoxylin-eosin (H&E) and Periodic Acid–Schiff staining were performed as previously described (23–25).

Fasting glucose and insulin.

After an 8-h fast, a blood sample was obtained via the orbital sinus under isoflurane anesthesia. Blood glucose was measured using an Accu-Chek Advantage glucometer (Roche, Mannheim, Germany). Fasting plasma insulin concentrations were determined by a LINCO ELISA kit).

Analysis of plasma lipids and metabolites.

After an 8-h fast, a blood sample was obtained via the orbital sinus under isoflurane anesthesia for plasma lipids and metabolites. Nuclear magnetic resonance (NMR) spectroscopy experiments were acquired using a 14.0-T Bruker magnet equipped with a Bruker AV-III console operating at 600.13 MHz. All spectra were acquired in 3-mm NMR tubes using a Bruker 5-mm QCI cryogenically cooled NMR probe. Plasma samples were prepared and analyzed according to the Bruker In Vitro Diagnostics research (IVDr) protocol. Sample preparation consisted of combining 50 μl plasma with 150 μl buffer supplied by Bruker Biospin specifically for the IVDr protocol. For 1D ¹H NMR, data were acquired using the 1D-NOE experiment that filters NMR signals associated with broad line widths arising from proteins that might be present in plasma samples that adversely affect spectral quality. Experiment conditions included: sample temperature of 310 K, 96 k data points, 20 ppm sweep width, a recycle delay of 4 s, a mixing time of 150 ms, and 32 scans. Lipoprotein subclass analysis was performed using regression analysis of the NMR data, which is done automatically as part of the IVDr platform as previously described (46).

Quantitative real-time PCR analysis.

Total RNA was harvested from Ppara^fl/fl and Ppara^HepKO mice by lysing livers using a Tissue Lyser LT (Qiagen) and then extraction by 5-Prime PerfectPure RNA Tissue Kit (Fisher Scientific). Total RNA was read on a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Wilmington, DE) and cDNA was synthesized using high capacity cDNA reverse transcription kit (Applied Biosystems). PCR amplification of the cDNA was performed by quantitative real-time PCR using TrueAmp SYBR Green qPCR SuperMix (Alkali Scientific) for gene-specific primers as previously described (23–26, 40, 62). The thermocycling protocol consisted of 5 min at 95°C, 40 cycles of 15 s at 95°C, and 30 sFF at 60°C, finished with a melting curve ranging from 60 to 95°C to allow distinction of specific products. Normalization was performed in separate reactions with primers to GAPDH.

Gel electrophoresis and Western blot analysis.

Mouse tissues were flash frozen in liquid nitrogen during harvesting and stored at −80°C. For gel electrophoresis, 50–100 mg of cut tissue was then resuspended in three volumes of CelLytic Buffer (Sigma C3228) plus 10% protease inhibitor cocktail (Sigma P2714–1BTL) and Halt phosphatase inhibitor cocktail (Fisher PI78420), and then incubated on ice for 30 min. The livers were lysed using a Qiagen Tissue Lyser LT and then centrifuged at 100,000 g at 4°C. Protein samples were resolved by SDS PAGE and electrophoretically transferred to Immobilon-FL membranes. Membranes were blocked at room temperature for 2 h in TBS (10 mM Tris·HCl (pH 7.4) and 150 mM NaCl) containing 3% BSA. Subsequently, the membranes were incubated overnight at 4°C with the following antibodies: PPARα (cat. no.sc-9000; Santa Cruz Biotechnology, Santa Cruz, CA) or heat shock protein 90 (HSP90) (cat. no. 13119; Santa Cruz). After three washes in TBS + 0.1% Tween 20, the membrane was incubated with an infrared anti-rabbit (IRDye 800, green) or anti-mouse (IRDye 680, red) secondary antibody labeled with IRDye infrared dye (LI-COR Biosciences) (1:10,000 dilution in TBS) for 2 h at 4°C. Immunoreactivity was visualized and quantified by infrared scanning in the Odyssey system (LI-COR Biosciences).

Statistical analysis.

Data were analyzed with Prism 7 (GraphPad Software, San Diego, CA) using ANOVA combined with Tukey’s posttest to compare pairs of group means or unpaired Student’s t tests. Results are expressed as means ± SE. Additionally, one-way ANOVA with a least significant difference post hoc test was used to compare mean values between multiple groups, and a Student’s two-tailed, and a two-way ANOVA was utilized in multiple comparisons, followed by the Bonferroni post hoc analysis to identify interactions. P values of 0.05 or smaller were considered statistically significant.

RESULTS

Body weights, fat, and lean mass of mice with a PPARα hepatocyte-specific knockout are comparable to floxed control mice on HFD or NFD.

The global PPARα null mice compared with their littermate controls have significant differences in body weight and in their metabolic phenotypes (43). Previously, Montagner et al. (43) showed that mice with a hepatocyte-specific PPARα KO had no change in body weight on a standard chow diet, but significantly lower body weight in response to a methionine-deficient and choline-deficient diet (MCD). Here, we used Ppara^fl/fl and Ppara^HepKO mice, confirmed by Western blot analysis in Fig. 1A, which showed the specific loss of PPARα in the liver (dotted blue line) but not in other organs in Ppara^HepKO mice. To determine how a diet high in fat may impact hyperlipidemia in mice with PPARα not expressed in hepatocytes, we placed the Ppara^fl/fl and Ppara^HepKO on HFD or NFD for 12 wk. We found that Ppara^fl/fl on HFD had significantly (P < 0.05) lower PPARα mRNA in the liver compared with NFD (Fig. 1B). The Ppara^HepKO had almost no detectable expression on NFD or HFD. The Ppara^fl/fl and Ppara^HepKO mice exhibited no difference with NFD between the groups and similar increases in body weight on HFD (Table 1). Also, body weight, body length, and heart and fat pad masses had no differences between the groups on NFD or HFD, but liver weights were significantly (P = 0.0353) higher for the Ppara^HepKO on HFD. There was no difference in percent body fat mass or lean mass on HFD or NFD (Fig. 1, C and D). These data show that body weights of mice with a hepatocyte-specific PPARα KO on NFD or HFD are not different and that the loss of PPARα in the liver does not significantly change body weight during diet-induced obesity. However, PPARα in the liver does regulate liver mass, which impacts hyperlipidemia and hepatic insulin resistance that typically occur in obesity.

Fig. 1.

Percent body fat and lean mass in peroxisome proliferator-activated receptor-α (PPARα) wild-type (Ppara^fl/fl) and PPARα hepatocyte-specific knockout (Ppara^HepKO) mice fed a high-fat (HFD) or normal fat diet (NFD). A: Western blot analysis for PPARα and heat shock protein 90 (HSP90) as loading control in various tissues from Ppara^fl/fl and Ppara^HepKO mice. H, heart, K, kidney, L, liver, P, pancreas, S, spleen. Dotted blue line, specific loss of PPARα in the liver. B: Ppara mRNA levels in NFD and HFD in Ppara^fl/fl and Ppara^HepKO mice. C: fat mass was directly measured at the end of the experimental protocol by noninvasive EchoMRI. D: lean mass was determined by noninvasive EchoMRI. Values are means ± SE; *P < 0.05 vs. NFD; #P < 0.05 vs. HFD Ppara^fl/fl; ^P < 0.05 vs. NFD Ppara^HepKO; n = 16 and13, Ppara^fl/fl NFD, HFD; n = 6 and 8, Ppara^HepKO NFD, HFD. n, number of mice. All circles, NFD; all squares, HFD.

Table 1.

Body and organ weights in Ppara^fl/fl and Ppara^HepKO mice

Parameter	Pparafl/fl NFD	Pparafl/fl HFD	PparaHepKO NFD	PparaHepKO HFD	Diet P value	Genotype P value	Interaction P value
n	16	13	6	8
BW, g	34.9 ± 1.4	55.2 ± 1.5	37.1 ± 1.5	51.1 ± 1.1	0.0115*	0.3629	0.6625
BL, cm	9.9 ± 0.09	10.5 ± 0.08	9.9 ± 0.13	10.5 ± 0.04	<0.0001*	0.3389	0.6402
HW, mg	159.3 ± 5.2	172.2 ± 10.8	176.2 ± 9.5	171.6 ± 4.3	0.4659	0.1258	0.6675
HW:BW, mg/g	4.6 ± 0.19	3.1 ± 0.15	4.7 ± 0.37	3.3 ± 0.08	<0.0001*	0.3215	0.8271
HW:BL, mg/cm	16.1 ± 0.55	16.5 ± 1	17.7 ± 1.1	16.2 ± 0.4	0.5029	0.4054	0.2561
LW, mg	1.3 ± 0.4	2.5 ± 0.25	1.3 ± 0.12	3 ± 0.4	<0.0001*	0.2465	0.2833
LW/BW, mg/g	0.038 ± 0.001	0.044 ± 0.003	0.035 ± 0.002	0.058 ± 0.002	0.0005*	0.1249	0.0353*
LW/BL, mg/cm	0.13 ± 0.004	0.23 ± 0.04	0.13 ± 0.01	0.28 ± 0.03	<0.0001*	0.1045	0.1139
Epidydmal FW, g	0.91 ± 0.13	1.8 ± 0.13	1.2 ± 0.2	1.57 ± 0.13	0.0167*	0.4358	0.6399
Visceral FW g	0.75 ± 0.1	2.9 ± 0.08	0.87 ± 0.16	2.8 ± 0.14	<0.0001*	0.3189	0.5906
Total fat, g	1.66 ± 0.24	4.77 ± 0.17	2.03 ± 0.37	4.37 ± 0.21	<0.0001*	0.3345	0.9053

Values are means ± SE: n = no. of mice. Ppara^fl/fl, peroxisome proliferator-activated receptor-α (PPARα) wild-type mice; Ppara^HepKO, knockout mice; NFD, normal fat diet; HFD, high-fat diet; BW, body weight; BL, body length; HW, heart weight; LW, liver weight; FW, fat weight.

^*P < 0.05.

Ppara^HepKO mice have HFD-induced increased plasma insulin and reduced hepatic glycogen levels.

PPARα global null mice have been shown to have altered glucose and insulin plasma levels (43, 62). The role of hepatic PPARα in HFD-induced insulin and glucose intolerance has yet to be elucidated. Both groups of mice on NFD had normal fasting insulin and glucose levels (Fig. 2, A and B). However, HFD significantly increased fasting plasma insulin (Fig. 2A) and glucose (Fig. 2B) levels in both groups as compared with NFD. The insulin and glucose were significantly (P < 0.05) higher in Ppara^HepKO as compared with Ppara^fl/fl mice in HFD groups. The proinsulin C-peptide serum levels indicate insulin turnover and secretion, where increased levels suggest less insulin clearance by the liver (10). The C-peptide plasma levels were not changed in NFD between Ppara^HepKO and Ppara^fl/fl mice but were significantly (P < 0.05) higher with HFD between the groups). Diet-induced obesity causes less insulin clearance and higher circulating C-peptide levels (20, 55), and the Ppara^HepKO mice (Fig. 2C e had significantly higher C-peptide plasma levels compared with Ppara^fl/fl. These data imply that hepatic PPARα regulates insulin clearance, which with reduced levels increases circulating insulin, leading to peripheral insulin resistance.

Fig. 2.

Altered glucose and insulin levels in peroxisome proliferator-activated receptor-α (PPARα) hepatocyte-specific knockout (Ppara^HepKO) mice fed a high-fat diet (HFD). A: fasting blood insulin levels. B: fasting blood glucose levels. C: serum C-peptide levels. D: hepatic Fgf21 and Gys2 mRNA levels in normal fat diet (NFD) and HFD in PPAR wild-type (Ppara^fl/fl) and Ppara^HepKO mice. E: representative Periodic-Acid-Schiff staining in livers from Ppara^fl/fl and Ppara^HepKO mice. Scale bar = 100 nm. Values are means + SE. *P < 0.05 vs. NFD Ppara^fl/fl, #P < 0.05 vs. HFD Ppara^fl/fl, ^P < 0.05 vs. NFD Ppara^HepKO; A: n = 6 and 8 mice, Ppar^fl/fl NFD, HFD; n = 6 and 8 mice, Ppara^HepKO NFD, HFD; B: n = 16 mice/group NFD, n = 13 mice/group HFD Ppara^fl/fl, n =6 mice/group NFD, n = 8 mice/group HFD, Ppara^HEPKO; C: n = 6 mice/group; D: n = 16 mice/group NFD, 13 mice/group HFD, Ppara^fl/fl, n = 6 mice/group NFD, n = 8 mice/group HDF, Ppara^HEPKO. All circles, NFD; all squares, HFD.

Two PPARα target genes in the liver that are known to mediate glucose levels in mice are Fgf21 (27, 36) and Gys2 (38). The Fgf21 produces the FGF21 hormone that is known to regulate lipid accumulation and glucose intolerance (5, 7). The Gys2 gene produces glycogen synthase 2 that regulates hepatic glycogen storage for glucose storage (23, 25). Both genes were significantly (P < 0.05) lower in HFD Ppara^fl/fl mice compared with NFD (Fig. 2D). In the Ppara^HepKO mice, Fgf21 levels were abolished in NFD and HFD compared with Ppara^fl/fl. The Gys2 mRNA levels were significantly (P < 0.05) lower in the Ppara^HepKO mice on HFD compared with Ppara^fl/fl. In comparison with the lower glycogen synthase 2 mRNA, Periodic acid-Schiff (PAS) staining, an indicator of glycogen levels (25), was also reduced in the livers of HFD for Ppara^HepKO and Ppara^fl/fl mice, with Ppara^HepKO mice lower compared with Ppara^fl/fl (Fig. 2E). Reduced PPARα expression in the obese lessens β-oxidation and fat burning leading to hepatic steatosis that corresponds to reduced glycogen levels. These data show that hepatic PPARα regulates insulin resistance and glycogen storage, which may alter circulating insulin levels leading to peripheral insulin resistance.

Ppara^HepKO mice have higher fat accumulation and reduced fat burning in the liver that is worsened on HFD.

The higher liver weights in the Ppara^HepKO mice indicate that the loss of hepatic PPARα impairs liver function. PPARα has a known role in increasing fatty acid β-oxidation in the liver, WAT, and BAT. Obese mice have lower liver PPARα expression and hepatic steatosis and hyperlipidemia (25, 40). Oil Red O staining showed higher fat accumulation in the Ppara^HepKO as compared with Ppara^fl/fl mice on NFD (Fig. 3A), and these mice have normal body weights and are not obese. The presence of lipid droplets in the Ppara^HepKO indicates the beginning stages of NAFLD, commonly referred to as steatosis. Diet-induced obesity by HFD in the Ppara^fl/fl and Ppara^HepKO mice showed significantly more steatosis in the liver in both groups compared with NFD. However, the Ppara^HepKO had more micro- and macrosteatosis on HFD and NFD compared with Ppara^fl/fl. These findings were confirmed biochemically by measurement of hepatic triglycerides that were significantly higher in the Ppara^HepKO in NFD and HFD as compared with Ppara^fl/fl mice (Fig. 3B). Hepatic fat mass was also measured by EchoMRI and was significantly (P < 0.05) increased in both groups in response to HFD (Fig. 3C), and similar findings showed that hepatic fat mass was higher in Ppara^HepKO compared with Ppara^fl/fl mice. Hepatic lean mass was decreased in both groups in response to HFD but was further suppressed in the Ppara^HepKO in comparison with Ppara^fl/fl mice (Fig. 3D). PPARα transcriptionally regulates several genes involved in metabolic processes, such as the G0/G1 switch gene 2 (G0s2) gene, which has been shown to mediate metabolism and the cell cycle (73). Previously, G0s2 was shown to be significantly reduced in global PPARα KO animals (73). Here, we found similar results that G0s2 was substantially lower in the Ppara^HepKO compared with Ppara^fl/fl mice (Fig. 3E). These indicate that hepatic fat accumulation may be higher in the Ppara^HepKO by reduced hepatic metabolic gene actions.

Fig. 3.

Peroxisome proliferator-activated receptor-α (PPARα) hepatocyte-specific knockout (Ppara^HepKO) mice have enhanced high-fat diet (HFD)-induced hepatic steatosis. A: Oil red O staining in livers from PPARα wild-type (Ppara^fl/fl) and Ppara^HepKO mice on normal fat diet (NFD) and HFD. B: biochemical measurement of hepatic triglyceride levels. C: hepatic fat mass as measured by noninvasive EchoMRI at the end of the study. D: hepatic lean mass as measured by noninvasive EchoMRI at the end of the study. Measurement of hepatic metabolic gene expression in NFD and HFD in Ppara^fl/fl and Ppara^HepKO mice for G0s2 (E), Ctp1a (F), and peroxisomal gene expression for Abcd3, Acox1, Crot, and Pex7 (G). Values are means ± SE; *P < 0.05 vs. NFD Ppara^fl/fl; #P < 0.05 vs. HFD Ppara^fl/fl; ^P < 0.05 vs. NFD Ppara^HepKO. A: n = 9 and 8 Ppara^fl/fl and  Ppara^HEPKO NFD and HFD mice. B: n = 6 mice.  C and D: n = 16 and 13 Ppara^fl/fl NFD and HFD mice and n = 6 and 8 Ppara^Hepko NFD and HFD mice. E–G, n = 16 and 8 Ppara^fl/fl NFD and HFD mice and n = 6 and 8 Ppara^HepKO NFD and HFD mice. All circles, NFD; all squares HFD.

PPARα mediation of fat-burning β-oxidation is targeted by several genes in the pathway. One major contributor is the carnitine palmitoyltransferase 1a (Cpt1a), which brings fatty acids into the mitochondria for the burning of fat. The Ppara^fl/fl mice had significantly (P < 0.05) lower Cpt1a mRNA expression on HFD compared with NFD (Fig. 3F). The Ppara^HepKO mice had significantly (P < 0.01) reduced Cpt1a mRNA expression on HFD compared with Ppara^fl/fl. Another contributor is peroxisomes that are critical in aiding the mitochondria for the oxidation of very-long-chain-fatty acids and preventing mitochondrial and cellular DNA damage by reactive oxygen species (ROS) generated by β-oxidation of fatty acids (16). Peroxisomal gene expressions for Abcd3, Acox1, Crot, and Pex7 were significantly (P < 0.05) reduced in HFD Ppara^fl/fl mice compared with NFD (Fig. 3G). Ppara^HepKO mice showed a significant (P < 0.05) reduction in peroxisomal gene expression in NFD and HFD compared with Ppara^fl/fl mice. These genes represent peroxisomal biogenesis, fatty acid import, fatty acid degradation, and fatty acid export. These data indicate that mitochondrial dysfunction (Cpt1a) is mirrored by reduced peroxisomal activity (peroxisomal genes), which supports the generally linked activity of these protective organelles against NAFLD.

The major metabolite that PPARα increases during β-oxidation is the ketone β-hydroxybutyrate, which is released from the liver into blood (29). Plasma lipids as well as other metabolites such as amino, carboxylic, and keto acids were analyzed using a novel NMR spectroscopy approach. The Ppara^HepKO mice on HFD had significantly (P < 0.05) reduced plasma levels of β-hydroxybutyrate (measured as 3-hydroxybutyric acid) compared with control (Table 2). Interestingly, there was no difference between the groups for plasma pyruvic acid, the simplest of the α-keto acids. Other metabolites measured were carboxylic and amino acids. There was no difference between Ppara^HepKO compared with Ppara^fl/fl for plasma levels of carboxylic acid groups: acetic acid, citric acid, formic acid, or lactic acid (Table 2). Plasma amino acids between the mice groups showed significantly (P < 0.05) higher levels of alanine and significantly (P < 0.05) lower levels of glutamic acid in the HFD Ppara^HepKO group (Table 2). No change was observed in the other amino acids measured. These results show that Ppara^HepKO mice have reduced fat burning via the β-oxidation pathway, have microsteatosis on NFD, and early-stage NAFLD, which is worsened on HFD.

Table 2.

Metabolite analysis in Ppara^fl/fl and Ppara^HepKO mice

Plasma metabolite	Pparafl/fl NFD	Pparafl/fl HFD	PparaHepKO NFD	PparaHepKO HFD	Diet P Value	Genotype P Value	Interaction P Value
n	6	12	6	8
Keto acids
3-Hydroxybutyric acid, mmol/l	0.228 ± 0.05	0.294 ± 0.01	0.171 ± 0.01	0.21 ± 0.02	0.2045	0.0366*	0.9781
Pyruvic acid, mmol/l	0.048 ± 0.006	0.055 ± 0.005	0.05 ± 0.005	0.046 ± 0.003	0.6424	0.9641	0.7987
Carboxylic acids
Acetic acid, mmol/l	0.12 ± 0.01	0.046 ± 0.004	0.11 ± 0.02	0.056 ± 0.005	0.001*	0.4856	0.9938
Citric acid, mmol/l	0.188 ± 0.04	0.096 ± 0.006	0.188 ± 0.04	0.088 ± 0.004	0.002*	0.8615	0.8615
Formic acid, mmol/l	0.026 ± 0.006	0.028 ± 0.004	0.035 ± 0.007	0.026 ± 0.001	0.1275	0.1529	0.5817
Lactic acid, mmol/l	5.55 ± 1.8	1.56 ± 0.12	4.42 ± 1.9	1.72 ± 0.16	0.012*	0.6062	0.4943
Amino acids
Alanine, nmol/l	0.193 ± 0.012	0.143 ± 0.005	0.156 ± 0.021	0.167 ± 0.005	0.1816	0.3625	0.0271*
Glutamine, nmol/l	0.03 ± 0.033	0.353 ± 0.0008	0.098 ± 0.069	0.306 ± 0.0008	0.0262*	0.0312*	0.0495*
Glycine, nmol/l	0.181 ± 0.005	0.105 ± 0.011	0.171 ± 0.014	0.125 ± 0.004	0.001*	0.374	0.1202
Histidine, nmol/l	0.031 ± 0.004	0.04 ± 0.001	0.04 ± 0.007	0.042 ± 0.005	0.5002	0.1356	0.8656
Isoleucine, nmol/l	0.035 ± 0.002	0.035 ± 0.003	0.035 ± 0.002	0.033 ± 0.002	0.3506	0.3506	0.1656
Leucine, nmol/l	0.046 ± 0.006	0.033 ± 0.002	0.038 ± 0.005	0.036 ± 0.002	0.0553	0.488	0.1555
Tyrosine, nmol/l	0.023 ± 0.008	0.026 ± 0.003	0.01 ± 0.007	0.031 ± 0.005	0.6644	0.5774	0.1608
Valine, nmol/l	0.088 ± 0.005	0.074 ± 0.003	0.068 ± 0.01	0.077 ± 0.004	0.6517	0.1395	0.0422*

Values are means ± SE: n = no. of mice. NFD, normal fat diet; HFD, high-fat diet; Ppara^fl/fl, peroxisome proliferator-activated receptor-α (PPARα) wild-type mice; Ppara^HepKO, knockout mice.

^*P < 0.05.

Ppara^HepKO mice on HFD exhibit inflammation and higher proinflammatory cytokines expression.

Hepatic inflammation induces liver fat accumulation (40), which most likely occurs from oxidative stress and suppression of PPARα (25). To ascertain if there were inflammation in the Ppara^HepKO and Ppara^fl/fl mice in response to diet, we performed H&E staining and determined visible inflammatory foci that assembled in the livers. The Ppara^fl/fl mice on NFD showed no significant visible inflammatory foci (Fig. 4A). The Ppara^fl/fl on HFD displayed a variable, low level of inflammatory foci, but no signficant inflammation was observed. The Ppara^HepKO on NFD had obvious inflammatory foci present, but no significant difference was obeserved from Ppara^fl/fl mice on NFD for the inflammation score. However, the Ppara^HepKO on HFD had focal-to-mild lobular inflammation in 100% of the mice tested (Fig. 4A, arrows). These data suggest that PPARα protects against HFD-induced inflammation in the liver.

Fig. 4.

The peroxisome proliferator-activated receptor-α (PPARα) hepatocyte-specific knockout (Ppara^HepKO) mice develop hepatic inflammation on high-fat diet (HFD). A: H&E staining in livers from PPAR wild-type (Ppara^fl/fl) and Ppara^HepKO mice on normal fat diet (NFD) and HFD and inflammation score. Yellow arrows, Ppara^HepKO had focal-to-mild lobular inflammation on 100% of the HFD mice tested. B and C: biochemical measurement of hepatic triglyceride levels. Measurement of hepatic immune gene expression in NFD and HFD in Ppara^fl/fl and Ppara^HepKO mice for Cd20 and Cd3 (B) and macrophage markers Adgre1, Tnfa, Nos2, and Arg1 (C). Values are means ± SE; *P < 0.05 vs. NFD Ppara^fl/fl; #P < 0.05 vs. HFD Ppara^fl/fl; ^P < 0.05 vs. NFD Ppara^HepKO. A: n = 13 and 8, Ppara^fl/fl NFD, HFD; n = 6 and 8, Ppara^HepKO NFD, HFD. B and C: n = 13 and 7, Ppara^fl/fl NFD, HFD; n = 6 and 8, Ppara^HepKO NFD, HFD. n, number of mice. All circles, NFD; all squares, HFD.

Next, to ascertain the type of immune response that was elicited in the Ppara^HepKO and Ppara^fl/fl mice in response to diet, we measured established markers of immune cells, Cd20 and Cd3, which are markers for B-lymphocytes and thymocytes, respectively. In both groups of mice, hepatic Cd20 expression signficantly (P < 0.05) increased (Fig. 4B). However, there was no change in Cd20 expression between Ppara^fl/fl and Ppara^HepKO on HFD. The Cd3 expression, however, was significantly higher only in the Ppara^HepKO on HFD. We have previously shown that hepatic steatosis can occur by inducing a shift in macrophage population from anti-inflammatory M2 to proinflammatory M1 macrophages (40). We found that the Adgre1 mRNA expression that encodes for the general macrophage marker F480 had no significant difference between Ppara^fl/fl and Ppara^HepKO on NFD or HFD (Fig. 4C). This finding was not surprising as we have reported no change in total macrophage numbers, but that the change occurred in the M2 and M1 populations (40). In Fig. 4, D and E, the Ppara^HepKO on HFD had significantly higher expression of proinflammatory M1 macrophage markers TNFα (Tnfa) and inducible nitric oxide synthases iNOS (Nos2), and in Fig. 4F reduced anti-inflammatory M2 macrophage markers arginase 1 (Arg1). These data imply that hepatic PPARα reduces HFD-induced inflammation by mediating the M2 to M1 macrophage population shift.

Ppara^HepKO mice exhibited worsened HFD-induced hyperlipidemia.

A clinically known function of the activation of PPARα is to reduce plasma triglycerides for patients with hyperlipidemia. However, whether this is mediated from PPARα expressed in liver, WAT, BAT, muscle, or another tissue is unknown. The Ppara^HepKO on HFD had significantly (P < 0.05) higher plasma triglycerides as compared with HFD Ppara^fl/fl mice (Fig. 5A). Measurement of the triglyceride distribution showed that the triglyceride increase in the Ppara^HepKO was mostly in the VLDL lipoprotein and not intermediate-density lipoprotein (IDL, LDL, or HDL (Fig. 5B). The VLDL is assembled in the liver with triglycerides, cholesterol, and apolipoproteins, especially ApoB100 (8). Plasma ApoB100 was significantly (P < 0.05) higher in the HFD Ppara^HepKO compared with HFD Ppara^fl/fl mice (Fig. 5C). ApoB100 and the microsomal triglyceride transfer protein are critical for excretion of the VLDL molecule from liver (8). Despite the alteration in plasma apolipoprotein levels, the hepatic apoB mRNA expression was not changed between the HFD groups (Fig. 5D). However, hepatic Mttp mRNA was significantly (P < 0.05) lower in the HFD Ppara^HepKO compared with HFD Ppara^fl/fl mice, which could be due to the higher circulating levels of ApoB100 found in Ppara^HepKO mice.

Fig. 5.

Mice with a loss of hepatic peroxisome proliferator-activated receptor-α (PPARα) on a high-fat diet (HFD) have higher triglycerides and apolipoprotein B100 levels. A: plasma triglycerides as measured by NMR spectroscopy. Ppara^fl/fl, PPAR wild-type; Ppara^HepKO, PPARα hepatocyte-specific knockout mice. B: plasma triglyceride distribution as measured by NMR spectroscopy. Plasma ApoB100 as measured by NMR spectroscopy. C: plasma ApoB100 as measured by NMR spectroscopy. mRNA levels in livers from Ppara^fl/fl and Ppara^HepKO mice on normal fat diet (NFD) and HFD: Apob and Mttp (D), Npc2 (E), and Ldlr (F). Values are means ± SE; *P < 0.05 vs. NFD Ppara^fl/fl; #P < 0.05 vs. HFD Ppara^fl/fl; ^P < 0.05 vs. NFD Ppara^HepKO; n = 6 and 12, Ppara^fl/fl NFD, HFD; n = 6 and 8, Ppara^HepKO NFD, HFD; A and C, n = 8 and 12; B, n = 16 and 8 Ppara^fl/fl NFD, HFD; D, E, and F, n = 6 and 8, Ppara^HepKO NFD, HFD. n, number of mice. All circles, NFD; all squares, HFD.

The Npc2 gene encodes for the Niemann-Pick C2 (NPC2) protein that mediates cholesterol egress from lysosomes (28) and regulates sterol transport between membranes (72). The Npc2 expression was significantly reduced in Ppara^fl/fl mice on HFD compared with NFD (Fig. 5E). However, Ppara^HepKO had no signficant change in Npc2 expression in NFD or HFD. The Ppara^fl/fl mice on HFD had significantly lower Npc2 compared with Ppara^HepKO HFD. The movement of cholesterol aids homeostatic responses and the cellular cholesterol pool, including LDL receptor expression and de novo cholesterol synthesis (41). The hepatic LDL receptor mRNA (Ldlr) expression was signficantly (P < 0.05) lower in Ppara^HepKO and Ppara^fl/fl mice on HFD compared with their NFD controls. However, comparison between the groups showed no significant difference.

Next, to determine how these signaling mechanisms impact plasma cholesterol levels, we used Bruker NMR spectroscopy with the IVDr) protocol. The total and LDL particle numbers were significantly increased with HFD in both groups (Fig. 6, A and C.). The VLDL particle number was significantly lower with HFD in both groups (Fig. 6B). Interestingly, the total, VLDL, and LDL particle numbers were not changed between the HFD or NFD between the two groups (Fig. 6, A–C). The lipoprotein subclass analysis was performed by Bruker NMR spectroscopy using regression analysis of the NMR data, which is done automatically as part of the IVDr platform as previously described (46). The distribution of cholesterol, free cholesterol, and phospholipid was not changed for VLDL, IDL, LDL, or HDL (Fig. 6D). These data were also supported with no change in total plasma cholesterol, LDL cholesterol, HDL cholesterol, or LDL:HDL ratio (Fig. 7, A–E). However, the L5PN subfraction of LDL was significantly increased in Ppara^HepKO as compared with Ppara^fl/fl mice (Fig. 7C). Apolipoproteins ApoA1 and ApoA2 are scavengers that help remove cholesterol from peripheral tissues. There was no significant difference between the HFD groups for plasma ApoA1 (P = 0.0977) or ApoA2 (P = 0.2605) (Fig. 8, A and B). This was also reflected in the ApoA1 and ApoA2 distribution profiles for each fraction (Fig. 8, C and D).

Fig. 6.

Plasma total, very-low-density lipoprotein (VLDL), and low-density lipoprotein (LDL) particle numbers and cholesterol distributions in PPAR wild-type (Ppara^fl/fl) and PPARα hepatocyte-specific knockout (Ppara^HepKO) mice are not different between the groups. Total particle number (A), plasma VLDL particle number (B), and plasma LDL particle number (C) as measured by NMR spectroscopy. D: distribution profiles of cholesterol, free cholesterol, and phospholipid in plasma as for VLDL, intermediate-density lipoproteins (IDL), LDL, and high-density lipoprotein (HDL). Values are means ± SE; *P < 0.05 vs. NFD Ppara^fl/fl; ^P < 0.05 vs. normal fat diet (NFD) Ppara^HepKO; n = 6 and 12, Ppara^fl/fl NFD, high-fat diet (HFD); n = 6 and 8, Ppara^HepKO NFD, HFD. n, number of mice. All circles, NFD; all squares, HFD.

Fig. 7.

Plasma cholesterol in peroxisome proliferator-activated receptor-α (PPARα) wild-type (Ppara^fl/fl) and PPARα hepatocyte-specific knockout (Ppara^HepKO) mice fed a high-fat diet (HFD). Plasma total cholesterol (A), low-density lipoprotein (LDL0 cholesterol (B), LDL cholesterol subfractions (C), high-density lipoprotein (HDL) cholesterol (D), and LDL:HDL ratio (E) as measured by NMR spectroscopy. Values are means ± SE; *P < 0.05 vs. normal fat diet (NFD) Ppara^fl/fl; #P < 0.05 (vs. Ppara^fl/fl mice); ^P < 0.05 (vs. NFD Ppara^HepKO); n = 6 and12, Ppara^fl/fl NFD, HFD; n = 6 and 8, Ppara^HepKO NFD, HFD). n, number of mice. All circles, NFD; all squares, HFD.

Fig. 8.

Plasma apolipoproteins A1 (ApoA1) and ApoA2 levels in peroxisome proliferator-activated receptor-α (PPARα) wild-type (Ppara^fl/fl) and PPARα hepatocyte-specific knockout (Ppara^HepKO) mice fed a high-fat diet (HFD). Plasma ApoA1 (A) and Apo-A2 (B) as measured by NMR spectroscopy. ApoA1 (C) and ApoA2 (D) distribution profiles for very-low-density lipoprotein (VLDL), intermediate-density lipoproteins (IDL), low-density lipoprotein (LDL), and high-density lipoprotein (HDL). Values are means ± SE; *P < 0.05 vs. normal fed diet (NFD) Ppara^fl/fl; ^P < 0.05 vs. NFD Ppara^HepKO; n = 6 and12, Ppara^fl/fl NFD, HFD; n = 6 and 8, Ppara^HepKO NFD, HFD. n, number of mice. All circles, NFD; all squares, HFD.

Hyperlipidemia is regulated by the production of cholesterol in the liver (41). To determine changes in cholesterol conversion or transport gene changes, we measured the major players Cyp7a1, Cyp27a1, Hmgcr1, Cd36, Abca1, and Abcg1 (41). The hepatic mRNA expression for Cyp7a1, Cyp27a1, Hmgcr1, Abca1, and Abcg1 was significantly (P < 0.05) reduced in Ppara^fl/fl mice on HFD compared with NFD, but scavenger receptor Cd36 was unchanged (Fig. 9). The Ppara^HepKO mice had reduced Cyp27a1 and increased Cd36 expression on NFD compared with HFD. The other genes for Ppara^HepKO compared with the Ppara^fl/fl had no significant differences.

Fig. 9.

Cholesterol conversion or transport gene expression in peroxisome proliferator-activated receptor-α (PPARα) wild-type (Ppara^fl/fl) and PPARα hepatocyte-specific knockout (Ppara^HepKO) mice. Hepatic Abca1, Abcg1, Cd36, Cyp27a1, Cyp7a1, and Hmgcr1 mRNA levels in NFD (all circles) and HFD (all squares) in Ppara^fl/fl and Ppara^HepKO mice. Values are means ± SE; *P < 0.05 vs. NFD Ppara^fl/fl; ^P < 0.05 vs. NFD Ppara^HepKO; n = 16 and 13, Ppara^fl/fl NFD, HFD; n = 6 and 8, Ppara^HepKO NFD, HFD. n, number of mice.

Overall, these data indicate that hepatic PPARα mediates immune cell foci inclusion in the liver during high-fat feeding and that this causes a shift in the macrophage population from M2 to M1. These events lead to hepatic steatosis, which also increases plasma triglyceride and ApoB100 levels during high-fat diet-induced hyperlipidemia.

DISCUSSION

To determine the specific purpose of hepatic PPARα, here we used mice lacking PPARα specifically in hepatocytes as previously described (43). The effect of the loss of hepatocyte PPARα on inflammation and lipid accumulation in response to fasting and MCD had been studied in these mice (43). However, the role of hepatic PPARα on lipid accumulation in the liver as well as regulation of plasma triglycerides, apolipoproteins, cholesterol, and other metabolites in response to high-fat feeding had not been previously examined. Here, we found no difference in body weight between the Ppara^fl/fl and Ppara^HepKO on a high-fat or normal-fat diet. However, there were significant differences observed with the MCD diet in the Ppara^HepKO but not the global Ppara KO (43). These differences are likely due to the type of calorie load on the liver.

Small, dense LDL is a type of LDL cholesterol that is an emerging risk factor for cardiovascular disease (31, 58, 64). It is smaller and heavier than typical LDL cholesterol and can increase the risk of developing atherosclerosis and other cardiovascular diseases (58, 64). The Ppara^HepKO mice showed significantly increased levels of LDL subfraction (L5) in response to high-fat feeding. Several studies have demonstrated that alterations in small, dense LDL may mediate the increased risk for atherosclerosis and cardiovascular disease in NAFLD (60, 63). These are the first data to demonstrate a relationship between hepatic PPARα and small, dense LDL particles in the plasma. Further studies are needed to fully elucidate the role of alterations in hepatic PPARα and small, dense LDL particles in the development of cardiovascular disease in NAFLD.

Diet-induced obesity may result in elevated levels of glucose and fatty acids resulting in enhanced $O_2^{-}$ production leading to increased formation of ROS. Oxidative stress magnifies the adverse effects of obesity by inducing inflammation in the liver leading to NAFLD and, on occasion, worsens to nonalcoholic steatohepatitis (4). The onset of NAFLD has been linked to cardiovascular events by increasing levels of the proinflammatory markers TNFα and C-reactive protein (CRP) (42). CRP is a liver-specific protein that has been extensively studied as a biomarker of inflammation in cardiovascular disease. Exacerbated ROS, due to obesity-induced NAFLD, increases CRP expression and levels in blood increasing risk of diabetes, hypertension, and cardiovascular disease (48). Both statin and fibrate drugs have been shown to decrease CRP, but if this occurs by their role in regulating hepatic inflammation is unknown. Here, we found that proinflammatory M1 macrophage markers TNFα and iNOS were significantly increased and that intralobular inflammatory foci were increased in the liver of the Ppara^HepKO on HFD compared with Ppara^fl/fl control.

Ketone body metabolism is central for fine-tuning metabolic roles that optimize organ and organism performance in varying nutrient states and protect from inflammation and injury in multiple organ systems (49). Ppara^HepKO mice on a high-fat diet showed significantly lower levels of plasma 3-hydroxybutyric acid, which is converted to β-hydroxybutyrate (βOHB). These are the major metabolites released from the liver during the burning of fat. Interestingly, there was no change in the α-keto acid, pyruvic acid. Ketogenesis is a series of reactions that lead to the formation of ketone bodies, including βOHB. PPARα regulates ketogenesis via a transcriptional network that includes AMP-activated protein kinase (AMPK), PPARγ coactivator 1α (PGC-1α), mammalian target of rapamycin, and FGF21 (17). Lower levels of βOHB production have been associated with NAFLD and NASH (39), which may be reduced from inflammation in the liver, possibly from lipid peroxidation that worsens NAFLD (4). Obesity increases oxidative stress (ROS) and simultaneously decreases expression and activity of key cytoprotective systems including heme oxygenase (HO) and bilirubin as well as PPARα, while increasing inflammatory cytokines and insulin resistance (13, 34, 47, 59, 68). Increasing HO-1 activity produces the antioxidant bilirubin that has been shown to reduce adiposity (53, 65), and recently found to function as a hormone by activating PPARα by direct binding (62). In obese mice, increasing HO-1production of bilirubin reduces oxidative stress and fatty liver by increasing PPARα and FGF21 levels (25, 54). Here, the observed increased hepatic steatosis was associated with lower levels of hepatic Ffg21 and Cpt1a, which are genes involved in fatty acid β-oxidation (22, 45, 69). FGF21 is regulated via PPARα and has been previously reported to reverse hepatic steatosis and glucose intolerance (3, 12, 25, 33). Others have shown that the Ppara^HepKO mice have significantly lower fasting plasma glucose levels on normal chow diet compared with littermate controls (6, 43). However, here, we found that the Ppara^HepKO mice in response to high-fat feeding exhibited significantly higher hyperglycemia and hyperinsulinemia compared with Ppara^fl/fl mice. This alteration in blood glucose regulation in response to high-fat feeding may be due in part to reduced hepatic Fgf21 in Ppara^HepKO mice, as FGF21 has been shown to function in glucose homeostasis (5, 70, 71).

Other factors such as branched-chain amino acids (BCAAs) are involved in the metabolism of glucose, and oxidation of BCAAs may increase fatty acid oxidation and play a role in obesity (37). Although the loss of hepatic PPARα did not have any significant effect on the levels of plasma BCAAs in response to high-fat diet feeding compared with control, it did have an impact on the levels of alanine and glutamic acid in the present study. The elevations in serum alanine could be reflective of the altered glucose metabolism in the Ppara^HepKO mice that would interfere with the glucose-alanine cycle (15).

In conclusion, to develop novel therapeutics for the treatment of NAFLD, a better understanding of the ability of the liver to respond to inflammation and how these induce alterations in apolipoprotein and triglyceride metabolism are needed. These may be mediated from the early onset of reduced PPARα, causing inflammation and hepatic steatosis, which also occurs in obesity. PPARα is the major isoform present in the liver where it helps to orchestrate the hepatic responses to fasting (6, 29), and here we show that it is also quintessential for high fat-induced hepatic inflammation. PPARα is also expressed in a variety of extrahepatic tissues, such as muscle, adipose, pancreas, kidney, and brain, which may also play influential roles in the clearance of apolipoproteins for triglycerides and cholesterol homeostasis. This may aid in the protection of the cardiovascular system from lipid peroxidation or oxidized cholesterol and plaque development that can lead to heart attack or stroke. Increasing HDL cholesterol by either exercise or fibrate treatment is beneficial for hyperlipidemia. However, we did not find a significant difference in HDL levels for the hepatic PPARα KO compared with control mice. The PPARα regulation of HDL levels may be mediated by other tissues, such as adipose or muscle. Further studies using tissue-specific PPARα KO in extrahepatic tissues such as adipose and muscle, which has not been performed, will reveal the actions of the nuclear receptor responses to apolipoprotein signaling.

Perspectives and Significance

The results of the present study establish an essential role for hepatic PPARα in protection against high-fat feeding that induced inflammation-induced hyperlipidemia, hepatic steatosis, and insulin resistance. Moreover, our data show a role for hepatic PPARα in the regulation of plasma triglycerides, apolipoproteins, amino acids, and keto acids in response to high-fat feeding. Our results support the development of drugs that target induction of hepatic PPARα as potential novel therapies for dietary obesity-induced inflammation that is known to cause hypertriglyceridemia, hepatic steatosis, and alterations in ketone and fatty acid metabolism.

GRANTS

This work was supported by National Institutes of Health Grant L32MD009154 (to T. D. Hinds, Jr.), National Heart, Lung and Blood Institute Grants K01HL-125445 (to T. D. Hinds, Jr.) and P01 HL05197-11 (to D. E. Stec), and the National Institute of General Medical Sciences Grant P20GM104357-02 (to D. E. Stec). Nuclear magnetic resonance instrumentation was supported in part by National Science Foundation Grant 0922862, National Institutes of Health Grant S10 RR025677, and Vanderbilt University matching funds. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

D.E.S., C.D.A., D.F.S., and T.D.H. conceived and designed research; D.E.S., D.M.G., J.A.H., S.H., Z.L.M., N.R.F., J.W.R., D.F.S., and T.D.H. performed experiments; D.E.S., D.M.G., J.A.H., S.H., Z.L.M., N.R.F., J.W.R., D.F.S., and T.D.H. analyzed data; D.E.S., D.M.G., J.A.H., Z.L.M., N.R.F., J.W.R., C.D.A., D.F.S., and T.D.H. interpreted results of experiments; D.E.S. and T.D.H. prepared figures; D.E.S. and T.D.H. drafted manuscript; D.E.S., D.M.G., J.A.H., S.H., Z.L.M., N.R.F., J.W.R., C.D.A., D.F.S., and T.D.H. edited and revised manuscript; D.E.S., D.M.G., J.A.H., S.H., Z.L.M., N.R.F., J.W.R., C.D.A., D.F.S., and T.D.H. approved final version of manuscript.