A Novel Peroxisome Proliferator-activated Receptor (PPAR)α Agonist and PPARγ Antagonist, Z-551, Ameliorates High-fat Diet-induced Obesity and Metabolic Disorders in Mice*

Background: Obesity is one of the principal causes of metabolic syndrome. Results: A novel PPARα agonist/γ antagonist, Z-551, ameliorates obesity, insulin resistance, and impairment of glucose and lipid metabolisms in mice. Conclusion: Z-551 might be clinically useful for preventing or treating obesity and obesity-related metabolic disorders. Significance: A novel combination of PPARα agonist/γ antagonist is effective to improve obesity and obesity-related metabolic disorders. A novel peroxisome proliferator-activated receptor (PPAR) modulator, Z-551, having both PPARα agonistic and PPARγ antagonistic activities, has been developed for the treatment of obesity and obesity-related metabolic disorders. We examined the effects of Z-551 on obesity and the metabolic disorders in wild-type mice on the high-fat diet (HFD). In mice on the HFD, Z-551 significantly suppressed body weight gain and ameliorated insulin resistance and abnormal glucose and lipid metabolisms. Z-551 inhibited visceral fat mass gain and adipocyte hypertrophy, and reduced molecules involved in fatty acid uptake and synthesis, macrophage infiltration, and inflammation in adipose tissue. Z-551 increased molecules involved in fatty acid combustion, while reduced molecules associated with gluconeogenesis in the liver. Furthermore, Z-551 significantly reduced fasting plasma levels of glucose, triglyceride, free fatty acid, insulin, and leptin. To elucidate the significance of the PPAR combination, we examined the effects of Z-551 in PPARα-deficient mice and those of a synthetic PPARγ antagonist in wild-type mice on the HFD. Both drugs showed similar, but weaker effects on body weight, insulin resistance and specific events provoked in adipose tissue compared with those of Z-551 as described above, except for lack of effects on fasting plasma triglyceride and free fatty acid levels. These findings suggest that Z-551 ameliorates HFD-induced obesity, insulin resistance, and impairment of glucose and lipid metabolisms by PPARα agonistic and PPARγ antagonistic activities, and therefore, might be clinically useful for preventing or treating obesity and obesity-related metabolic disorders such as insulin resistance, type 2 diabetes, and dyslipidemia.

Obesity is one of the principal causes of metabolic syndrome, and has currently become a serious social problem (1,2). Excess energy due to overeating or a lack of physical activity causes increased accumulation of subcutaneous and visceral fat, resulting in adipocyte hypertrophy, macrophage infiltration into the adipose tissue, and abnormal adipokine productions (3). Free fatty acids (FFA) 4 and adipokines including monocyte chemoattractant protein-1 (MCP-1) and TNF-␣ overproduced by hypertrophic adipocytes infiltrated with macrophages especially in visceral fat are known to provoke and induce systemic chronic inflammation, insulin resistance, and impairment of glucose and lipid metabolisms by inhibiting insulin-signaling pathway in target organs such as the liver and skeletal muscle (4 -6). Thus, visceral fat obesity has been elucidated to be closely related to the onset of obesity-related metabolic disorders such as insulin resistance, type 2 diabetes, dyslipidemia, and fatty liver diseases. Accordingly, drugs that can counteract obesity by inhibiting visceral fat accumulation, may provide an effective treatment of insulin resistance and obesity-related metabolic disorders. Drugs that may increase the expenditure of excess energy stored as triglyceride (TG) are also desirable.
PPARs are ligand-activated transcription factors and belong to a nuclear receptor superfamily, which consists of three subtypes (␣, ␥, and ␦) in mammals (7,8). PPAR␣ and PPAR␥ have been noticed as therapeutic targets for dyslipidemia and type 2 diabetes, respectively. PPAR␣ is mainly expressed in the liver, and plays wide-ranging roles in lipid metabolism, energy homeo-stasis, and inflammation (9,10). Fibrates, PPAR␣ agonists, and drugs for dyslipidemia promote peroxisomal and mitochondrial ␤-oxidation, and inhibit fatty acid synthesis in the liver due to activation of PPAR␣, resulting in amelioration of insulin resistance owing to decreased ectopic fat accumulation in peripheral tissues (11)(12)(13)(14)(15). On the other hand, PPAR␥ has two isoforms, PPAR␥1 and PPAR␥2 formed by selective splicing. PPAR␥1 is expressed at low levels in many tissues, whereas PPAR␥2 is specifically expressed in the adipose tissue and functions as a master regulator in adipocyte differentiation and glucose metabolism (16). Thiazolidinediones (TZDs), PPAR␥ agonists, and drugs for type 2 diabetes promote loss of hypertrophic adipocytes by apoptosis and differentiation of preadipocytes into normal adipocytes, thereby preventing adipocyte hypertrophy and ameliorating insulin resistance (17,18), whereas these drugs cause weight gain. In contrast, we have reported that heterozygous PPAR␥-deficient (PPAR␥ ϩ/Ϫ ) mice are resistant to high-fat diet (HFD)-induced obesity, adipocyte hypertrophy, and insulin resistance (18 -20). Consistent with these observations, the Pro12Ala polymorphism in the human PPAR␥2 gene resulting in partial reduction of PPAR␥ transcriptional activity has been shown to be associated with reduced weight gain and improved insulin sensitivity (21,22). Inhibition of PPAR␥ activity by the retinoid X receptor (RXR) antagonist (23) or PPAR␥ antagonists (24,25) has been reported to ameliorate HFD-induced obesity and insulin resistance by inhibiting fat accumulation in adipocytes. These findings suggest that inhibiting the onset and progression of obesity is essential for prevention and treatment of obesity-related metabolic disorders. These results indicate that PPAR␥ antagonists could be potential drugs to prevent and treat obesity and obesity-related metabolic disorders and may be even more superior to PPAR␥ agonists in terms of suppression of fat formation and improvement of glucose metabolism (26). Taken together, it is desirable to utilize a PPAR modulator having both PPAR␣ agonistic and PPAR␥ antagonistic activities as an effective treatment of diet-induced obesity, insulin resistance, and impairment of glucose and lipid metabolisms by suppressing visceral fat accumulation and increasing energy expenditure in our concept of drug discovery.
Based on these backgrounds, to develop a drug with a novel mechanism of action for the prevention or treatment of obesity and obesity-related metabolic disorders, we have designed a novel PPAR modulator, Z-551, which has both PPAR␣ agonistic and PPAR␥ antagonistic activities.
In the present study, we examined the preventive and therapeutic effects of Z-551 on obesity and obesity-related metabolic disorders in two different animal models, wild-type (WT) mice and wild-type diet-induced obese (WT DIO) mice on the HFD. To elucidate the contribution of PPAR␥ antagonistic activities of Z-551 and the significance of the PPAR combination, we examined the effects of Z-551 in PPAR␣-deficient mice and those of a selective PPAR␥ antagonist, CZD-2, in WT mice on HFD. Finally, we will discuss the potential of Z-551 as a drug for preventing or treating obesity and obesity-related metabolic disorders such as insulin resistance, type 2 diabetes, dyslipidemia, and fatty liver diseases. Cell Culture-CV-1 cells were purchased from American Type Culture Collection (ATCC, Manassas, VA). CV-1 cells were cultured in DMEM (Invitrogen, Carlsbad, CA) supplemented with 10% charcoal-treated FBS (Dainippon Pharmaceutical Co., Ltd., Osaka, Japan) and 1% Antibiotic-Antimycotic solution (Invitrogen) at 37°C in 5% CO 2 .
Animals-Male C57BL/6J mice (WT mice) were purchased from Charles River Japan (Yokohama, Japan). 129S4/SvJae-PparatmiGonz/J mice (PPAR␣-deficient mice) were purchased from Jackson Laboratories (Bar Harbor, ME) and the male PPAR␣-deficient mice were obtained by crossbreeding. Mice were group-housed on a 12-h light/dark cycle in an animal room maintained at 23 Ϯ 3°C, and fed ad libitum either a normal diet (ND) (CE-2, CLEA Japan, Tokyo, Japan) or a HFD (32.0% safflower, 33.1% casein, 0.5% DL-methionine, 17.6% sucrose, 1.4% vitamin mixture, 9.8% mineral mixture, and 5.6% cellulose powder) (27). Cumulative food intake was measured every 2 weeks over an 8-week period in an individual group and the food intake was expressed as an amount of food consumed (g/day/mouse). All the experiments in the present study were conducted on male littermates. The animal care and experimental procedures were approved by the Animal Care Committee of the University of Tokyo.
Drug Administration-1) To examine the preventive effects of Z-551 or CZD-2 in C57BL/6J mice (WT mice) and PPAR␣deficient mice on the HFD (Figs. 2, 3, 6, 7, and 8), 4-week-old male mice were acclimatized to the HFD for 1 week followed by administration of Z-551 or CZD-2 for 9 to 16 weeks. 2) To examine the therapeutic effects of Z-551 in wild-type diet-induced obese (WT DIO) C57BL6/J mice (Figs. 4 and 5), 4-weekold male mice were given the HFD for 10 to 14 weeks and then administered Z-551 for 8 to 16 weeks. Z-551 and CZD-2 were given as a 0.1% (w/w) food admixture. In all experiments, based on the two factors of body weight and fasting plasma glucose level, the mice were assigned to a control group and a group of drug-treatment before administration of the drugs.
Glucose Homeostasis-The oral glucose tolerance test (OGTT) and insulin tolerance test (ITT) were performed by the methods described previously (28) with slight modifications regarding the amounts of glucose and insulin administered, and fasting periods as described in the figure legends. The pyruvate challenge test was conducted as described previously (29).
Assays for Blood and Tissue Samples-Plasma glucose and insulin levels were determined using a GluTest ProR (Sanwa Kagaku Research, Nagoya, Japan), and an Insulin ELISA Kit (Shibayagi, Gunma, Japan), respectively. Plasma TG and FFA levels were measured using a Triglyceride E-Test Wako and a NEFA C-Test Wako (Wako Pure Chemical Industries, Osaka, Japan), respectively. Plasma leptin levels were determined using a Leptin ELISA Kit (Morinaga Institute of Biological Science, Yokohama, Japan). Tissue TG in the liver and skeletal muscle was extracted with a hexane/isopropanol mixture (3:2, v/v) from the tissue homogenate, and dissolved in dioxane after evaporation at room temperature, followed by measurement using a Triglyceride E-Test Wako. The thiobarbituric acid reactive substance (TBARS) levels in the liver were determined using a TBARS Assay Kit (ZeptoMetrix Corporation, Buffalo, NY) according to the manufacturer's instructions.
Rectal Temperature-Rectal temperature was measured by inserting a digital thermistor thermometer (KN-91, Natsume Seisakusho, Tokyo, Japan) into the mouse rectum.
Oxygen Consumption-Oxygen consumption was measured every 1 h for 24 h in WT mice on the HFD using an O 2 /CO 2 metabolism measurement device (Oxymax system; model 7540, Columbus Instruments, Columbus, OH). After Z-551 treatment for 3 weeks, each mouse was placed in a sealed chamber with an air flow rate of 600 ml/min at room temperature. The amount of oxygen consumed was normalized by body weight of each mouse and expressed as liter/kg/h.
Histological Analysis-The isolated epididymal white adipose tissue (WAT) and liver were fixed in 10% formaldehyde/ PBS, and maintained at 4°C until use. The fixed tissues were dehydrated and processed for paraffin embedding, and 10-m sections were cut and mounted on silanized slides, followed by staining with hematoxylin and eosin (HE) (18).
Measurement of Adipocyte Size-The adipocyte size was measured by the method described previously (18) with slight modifications. In brief, the epididymal WAT was routinely processed for paraffin embedding, and 10-m sections were cut and mounted on silanized slides, followed by staining with HE. The individual adipocyte area was manually traced and the adi-pocyte size was analyzed with Image J software (NIH, Bethesda, MD). The white adipocyte size was measured in 200 cells/ mouse in each group.
Real-time Quantitative RT-PCR-Real-time quantitative RT-PCR was performed by the method described previously (30,31). Total RNA was prepared from the epididymal WAT and liver using TRIzol (Invitrogen) according to the manufacturer's instructions. After purification, total RNA was converted into first-strand cDNA. The first-strand cDNA obtained was subjected to real-time quantitative PCR with TaqMan Gene Expression Master Mix (Applied Biosystems, Foster City, CA) and TaqMan Gene Expression Assays (Applied Biosystems) containing a set of TaqMan probe and PCR primers for each gene using an ABI PRISM 7900 Sequence Detection System (Applied Biosystems). The TaqMan probes were used for mouse acyl-CoA oxidase 1 (ACOX1 encoded by Acox1), . The primers and the probe for mouse cyclophilin A (Ppia) were as follows: the forward primer was 5Ј-ggtcctggcatcttgtccat-3Ј, the reverse primer was 5Јcagtcttggcagtgcagataaaa-3Ј, and the probe was 5Ј-ctggaccaaacacaaacggttccca-3Ј (31). The relative amount of each transcript was normalized to the amount of Rplp0 or Ppia transcripts as internal standards.
Akt Phosphorylation-Mice were anesthetized after 6-h fasting, and insulin (0.5 unit/kg of body weight) (Humulin R, Eli Lilly, Indianapolis, IN) or saline (vehicle) was injected into the inferior vena cava, followed by isolating the liver and skeletal muscle 5 min later. Western blot analysis was carried out as described previously (32). The Akt phosphorylation was analyzed using rabbit anti-mouse phospho-Akt (Ser-473) antibody (Cell Signaling Technology Inc., Beverly, MA) and rabbit anti-mouse Akt antibody (Santa Cruz Biotechnology, Santa Cruz, CA). The insulin-stimulated Akt phosphorylation was expressed as a ratio of phosphorylated Akt to Akt (pAkt/Akt).
Peripheral Quantitative Computed Tomography (pQCT)-The right femur was removed from each mouse. The isolated femur was fixed in 80% alcohol, and maintained at 4°C until use. pQCT was performed using an XCT Research SA Plus (Stratec Medizintechnik GmbH, Pforzheim, Germany). Total and trabecular bone mineral densities in the proximal femur of mice were measured by the method described previously (33).
Statistical Analysis-All the results were expressed as means Ϯ S.E. Difference between two groups was assessed using Student's t test. Data involving more than two groups were assessed by analysis of variance statistical (ANOVA) and Dunnett's test. p Ͻ 0.05 was considered statistically significant.

Z-551 Is a Novel PPAR Modulator Having Both PPAR␣ Agonistic and PPAR␥ Antagonistic
Activities-Agonistic and antagonistic activities of Z-551 (structure shown in Fig. 1A) for each PPAR were quantitatively determined by the luciferase reporter assay. Z-551 indicated partial agonistic activities for mouse and human PPAR␣ with half-maximal effective concentrations (EC 50 values) of 539 nM and 1.22 M, respectively (Fig. 1C). The maximal agonistic activities for both species were ϳ60% of that of PPAR␣ full agonist, Wy-14643. In addition, Z-551 exhibited antagonistic activities for mouse and human PPAR␥ with halfmaximal inhibitory concentrations (IC 50 values) of 956 nM and 481 nM, respectively (Fig. 1D), obtained with PPAR␥ full agonist, rosiglitazone. These results indicate no significant differ-ences in the PPAR␣ agonistic or PPAR␥ antagonistic activity of Z-551 between mouse and human. In contrast, Z-551 indicated neither agonistic nor antagonistic activity for mouse and human PPAR␦. These results indicate that Z-551 is a novel dual PPAR modulator, which has both PPAR␣ partial agonistic and PPAR␥ antagonistic activities.

Z-551 Has Preventive Effects on HFD-induced Obesity and Obesity-related Metabolic Disorders in WT Mice on the HFD-
To elucidate preventive effects of Z-551 on HFD-induced obesity and obesity-related metabolic disorders, we examined the effects using WT mice on the HFD. Z-551 significantly suppressed an increase in body weight without any effect on food intake, and weight of the visceral fat tissue and epididymal WAT, compared with the mice on the HFD alone (control) (Fig.  2, A-C). In the epididymal WAT, histological analysis showed that HFD-induced hypertrophy of adipocytes was suppressed by administration of Z-551 (Fig. 2D). Gene expression analysis in the epididymal WAT demonstrated that Z-551 significantly inhibited gene expressions involved in fatty acid uptake (Lpl and Cd36), fatty acid synthesis (Fasn), energy metabolism (Lep), macrophage recruitment (Ccl2), and cell surface markers of macrophage (Emr1) (Fig. 2E). Furthermore, Z-551 significantly reduced plasma leptin levels as well as mRNA levels (Fig. 2F). These findings indicate that Z-551 has anti-obesity effects and inhibits adipocyte hypertrophy and macrophage infiltration in adipose tissue. In the liver, the histological analysis showed that Z-551 suppressed the formation of large lipid droplets, compared with the control (Fig. 2G). Z-551 also significantly increased liver weight (Fig. 2H). The gene expression analysis in the liver demonstrated that Z-551 significantly increased gene expressions involved in fatty acid uptake (Lpl and Cd36), fatty acid oxidation (Acox1), and energy expenditure (Ucp2) (Fig. 2I). These findings indicate that Z-551 enhances fatty acid combustion including fatty acid uptake, fatty acid oxidation, and energy expenditure in the liver. Consistent with these results, plasma TG and FFA levels were significantly reduced by administration of Z-551 (Fig. 2, J and K). To determine whether the basal metabolism is altered by administration of Z-551, we measured the rectal temperature and oxygen consumption. Z-551 signif- icantly raised the rectal temperature and oxygen consumption during the active period in the dark phase in WT mice on the HFD (Fig. 2, L and M). Z-551 did not change the oxygen consumption per mice in the dark phase despite showing significant decrease of body weight compared with the control, indicating that Z-551 significantly increased oxygen consumption per body weight (Fig. 2M). These findings suggest that oxygen consumption is one of causes of anti-obesity effects by Z-551 administration. Furthermore, Z-551 significantly increased the expression levels of brown adipose tissue (BAT) markers of thermogenesis such as Ppargc1a and Ucp1, suggesting that Z-551 activated the thermogenesis in BAT of the mice (Fig. 2N). On the other hand, Z-551 had no effect on the cumulative fecal weight and lipids excreted in feces for 2 weeks after starting Z-551 administration (data not shown). The OGTT and ITT were performed to elucidate the effects of Z-551 on insulin resistance in WT mice on the HFD. The OGTT indicated that Z-551 significantly decreased plasma glucose and insulin levels after glucose administration compared with the control, and attenuated the glucose intolerance (Fig. 2, O and P). The ITT indicated that Z-551 significantly decreased plasma glucose levels after insulin injection, and improved decreased insulin sensitivity induced by the HFD (Fig. 2Q).
The Western blot analysis showed that Z-551 significantly increased insulin-stimulated Akt (Ser-473) phosphorylation in the liver, skeletal muscle (Fig. 3, A and B), and epididymal WAT (data not shown), suggesting that Z-551 enhances insulin sensitivity. Z-551 significantly reduced increased fasting plasma glucose and insulin levels, which were shown at time zero in the OGTT (Fig. 2, O and P). These results indicate that Z-551 significantly improved hyperglycemia and hyperinsulinemia in WT mice on the HFD. Taken together, these findings suggest that Z-551 has preventive effects on HFD-induced obesity and obesity-related metabolic disorders due to ameliorating HFD-induced obesity, insulin resistance, and abnormal glucose and lipid metabolisms.

Z-551 Has Therapeutic Effects on HFD-induced Obesity and Obesity-related Metabolic Disorders in WT DIO
Mice-To elucidate the therapeutic effects of Z-551 on HFD-induced obesity and obesity-related metabolic disorders, we examined the effects using WT DIO mice. The results demonstrate that the therapeutic effects of Z-551 were essentially comparable to the preventive effects in WT mice on the HFD as described above (Fig. 2). Z-551 significantly suppressed increased body weight without changing food intake (data not shown), increased epididymal WAT weight and adipocyte hypertrophy induced by the HFD compared with the control (Fig. 4, A-C, Fig. 5, A-D). In the epididymal WAT, Z-551 significantly inhibited gene expressions involved in fatty acid uptake (Lpl and Cd36) and energy metabolism (Lep) (Fig. 5E). We examined the effects of Z-551 on macrophage infiltration and inflammation in the adipose tissue in detail. Z-551 significantly inhibited the expressions of common macrophage markers (Emr1, Cd68, and Itgam), classically activated (M1) macrophage markers (Ccl2 and Tlr4), and M1 macrophage/inflammation markers (Tnf and Il1b), whereas Z-551 had no effect on the expression of an alternatively activated (M2) macrophage marker (Mrc1) (Figs. 4D and 5E). Z-551 also significantly reduced plasma leptin levels as well as mRNA levels (Fig. 4E). These findings indicate that Z-551 not only has anti-obesity effects and inhibits adipocyte hypertrophy and macrophage infiltration in the adipose tissue similar to the preventive effects of Z-551 as described above, but also suppresses inflammation in adipose tissue. In the liver, similarly, Z-551 suppressed the formation of large lipid droplets in the liver (Fig. 4F), whereas Z-551 significantly increased liver weight compared with the control (Fig.  4G). The gene expression analysis in the liver also showed similar profiles involved in fatty acid uptake (Lpl and Cd36), fatty acid oxidation (Acox1), and energy expenditure (Ucp2) (Fig.  4H). We examined whether Z-551 had effects on TG content, oxidative stress, and gluconeogenesis in the liver. Z-551 significantly decreased concentrations of hepatic TG, and TBARS which is an indicator of lipid peroxidation (Fig. 4, I and J). Z-551 significantly inhibited gene expression of Pck1 involved in gluconeogenesis in the liver and decreased plasma glucose levels in the pyruvate challenge test, which evaluates gluconeogenesis H, mRNA expression levels in the liver. I and J, TG content (I) and TBARS levels (J) in the liver. K, plasma glucose in the pyruvate challenge test Week 9 after Z-551 administration. In the pyruvate challenge test, sodium pyruvate (1.5 g/kg BW) was intraperitoneally injected after 6-h fasting and blood samples were obtained at the indicated times. L and N, plasma TG (L) and FFA (N) Week 13 after Z-551 administration. M, TG content in skeletal muscle. O, rectal temperature. Rectal temperature was measured during the dark phase Week 15 after Z-551 administration. P and Q, plasma glucose Week 7 after Z-551 administration. In the OGTT, glucose (1.5 g/kg BW) was orally administered after 6-h fasting. In the ITT, insulin (0.5 unit/kg BW) was intraperitoneally injected. Blood samples were obtained at the indicated times. All results are expressed as mean Ϯ S.E. (n ϭ 6). *, p Ͻ 0.05; **, p Ͻ 0.01; n.s., not significant. (Fig. 4, H and K). These results indicate that Z-551 enhances fatty acid combustion including fatty acid uptake and fatty acid oxidation, and inhibits oxidative stress and gluconeogenesis in the liver. Furthermore, Z-551 significantly reduced plasma TG levels and ectopic TG accumulation in the skeletal muscle, and tended to reduce plasma FFA compared with the control (Fig. 4,  L-N). Z-551 significantly raised the rectal temperature during the active period in the dark phase in WT DIO mice (Fig. 4O). The OGTT and ITT indicated that Z-551 had effects (Fig. 4, P-R) similar to the preventive action in WT mice on the HFD as described above (Fig. 2) by ameliorating the glucose intolerance and improving the reduced insulin sensitivity induced by the HFD. Likewise, Z-551 significantly improved hyperglycemia and hyperinsulinemia in WT DIO mice as shown by fasting plasma glucose and insulin levels (Fig. 4, P and Q). Taken together, these findings suggest that Z-551 has therapeutic effects on HFD-induced obesity and obesity-related metabolic disorders by ameliorating HFD-induced obesity, insulin resistance, and abnormal glucose and lipid metabolisms as well as the preventive effects.
Both PPAR␣ Agonistic and PPAR␥ Antagonistic Activities of Z-551 Are Required for Ameliorating HFD-induced Obesity and Obesity-related Metabolic Disorders-To elucidate the contribution of PPAR␥ antagonistic activities of Z-551 and the signif-icance of the PPAR combination, we examined the preventive effects of Z-551 using PPAR␣-deficient mice and those of a selective PPAR␥ antagonist, CZD-2 using WT mice on the HFD.
(i) Effects of Z-551 in PPAR␣-deficient Mice on the HFD-Z-551 significantly increased the expression of PPAR␣ target genes in the liver of WT mice (Fig. 2), whereas no effect of Z-551 was found in the liver of PPAR␣-deficient mice (data not shown). Z-551 was found to have similar but weaker preventive effects in PPAR␣-deficient mice on the HFD (Fig. 6) compared with those in WT mice on the HFD as described above (Fig. 2), though no effects on fasting plasma TG and FFA levels and liver weight were observed. Z-551 significantly suppressed increases in body weight and epididymal WAT weight, though to lower degrees than those in WT mice on the HFD (Fig. 2) without changing food intake (data not shown), whereas Z-551 tended to inhibit an increase in body weight compared with the control (Fig. 6, A-C). Z-551 also suppressed adipocyte hypertrophy in the epididymal WAT induced by the HFD (Fig. 6D). In the epididymal WAT, gene expressions involved in fatty acid uptake (Lpl and Cd36), fatty acid synthesis (Srebf1 and Fasn), energy metabolism (Lep), macrophage recruitment (Ccl2), and cell surface markers of macrophage (Cd68) were inhib- ited by administration of Z-551 (Fig. 6E). Z-551 also significantly reduced plasma leptin to a much lower degree (Fig.  6F). The OGTT and ITT indicated that Z-551 had similar effects (Fig. 6, G-I) to those in WT mice on the HFD (Fig. 2, O-Q) by ameliorating the glucose intolerance and improving the reduced insulin sensitivity induced by the HFD. Additionally, Z-551 significantly reduced fasting plasma glucose and insulin levels to lower degrees (Fig. 6, G and H). In contrast, Z-551 did not reduce fasting plasma TG and FFA levels (Fig. 6, J and K), suggesting that PPAR␣ activation by Z-551 administration is necessary for the reduction of fasting plasma TG and FFA levels. Z-551 had no effect on liver weight in PPAR␣-deficient mice compared with the control (Fig. 6L), whereas Z-551 significantly increased liver weight in WT mice on the HFD (Fig. 2H). These results indicate that PPAR␣ activation by Z-551 was likely to cause the liver hypertrophy in mice.
(ii) Effects of PPAR␥ Antagonist, CZD-2, in WT Mice on the HFD-CZD-2 is a selective PPAR␥ antagonist with IC 50 of 844 nM for mouse PPAR␥ and its antagonistic activity is equivalent to that of Z-551 (Fig. 1). In WT mice on the HFD, CZD-2 exhibited very similar effects with different intensities (Fig. 7, A-E, I, and J) compared with those of Z-551 in PPAR␣-deficient mice on the HFD as described above (Fig. 6) except some effects. The OGTT indicated that CZD-2 decreased plasma insulin, without any difference in plasma glucose levels compared with the con-  JUNE 5, 2015 • VOLUME 290 • NUMBER 23 trol (Fig. 7, F and G). The ITT indicated that CZD-2 had a significant decrease in plasma glucose levels after insulin injection (Fig. 7H), suggesting that a selective PPAR␥ antagonist significantly improves insulin sensitivity.

PPAR␣ Agonist/␥ Antagonist, Obesity, and Metabolic Disorders
Taken together, these findings indicate that not only PPAR␣ agonistic activity but also PPAR␥ antagonistic activity of Z-551 is required for ameliorating HFD-induced obesity, insulin resistance, and abnormal glucose and lipid metabolisms.

Z-551 Has No Effect on Bone Loss in WT Mice on the HFD-
pQCT, MRI, and MRS were used to elucidate the effect of Z-551 on bone loss in the femur of WT mice on the HFD. In this experiment, both Z-551 (0.01%, 0.03%, and 0.1%) and rosiglitazone (0.01% and 0.03%) markedly improved hyperglycemia and insulin sensitivity in the mice (data not shown). Rosiglitazone exhibited medulla ossium flava (fatty or yellow marrow), and significantly decreased total and trabecular bone mineral den- sities, leading to the trabecular bone loss in the proximal femur of mice (Fig. 8, A-D) as previously reported (35). On the other hand, Z-551 had no effect on morphology, color, bone mineral density, and MRI of the proximal femur in the mice compared with the control (Fig. 8, A-D). MRS analysis confirms that Z-551 had a closely similar spectrum to that of the control obtained from the proximal femur of mice, whereas rosiglitazone showed new signals of 1H-MRS, indicating the presence of lipids (Fig. 8E). These results suggest that Z-551 has no effect on bone loss, whereas rosiglitazone increases the adiposity of bone marrow, leading to an increased risk of bone fracture.

Discussion
In the present study, we investigated the preventive and therapeutic effects of Z-551 using WT mice and WT DIO mice on the HFD to elucidate the mechanisms of Z-551 actions, and thereby examining its potential of being a clinically useful drug for diet-induced obesity and obesity-related metabolic disorders such as insulin resistance, type 2 diabetes, dyslipidemia, and fatty liver diseases.
We demonstrated that Z-551 had anti-obesity effects in both WT mice and WT DIO mice, and unique features of suppressing body weight gain without any reduction of food intake. The gene expression analysis suggests that the anti-obesity effects of Z-551 are caused by the suppression of increased WAT weight due to the inhibition of fatty acid uptake, and fatty acid synthesis in the adipose tissue, and the activation of fatty acid uptake, fatty acid oxidation, and energy expenditure in the liver. Obesity reportedly causes the onset of leptin resistance with hyperleptinemia and a decline in basal metabolism including thermogenesis (36 -38). Leptin is an important adipokine that plays a key role in the regulation of energy intake and expenditure. The present study demonstrates that the anti-obesity effects of Z-551 lead to a significant decrease in plasma leptin levels and the significant elevation in rectal temperature and oxygen consumption during dark phase, suggesting increased energy expenditure. In addition, Z-551 significantly increased the expression levels of BAT markers of thermogenesis. These data raised the possibility that an enhanced BAT thermogenesis could be an underlying cause of the increase in energy expenditure induced by Z-551. Since there are reports to the effect that PPAR␣ agonists increase fatty acid combustion in the liver and promote lipid degradation in the adipose tissue (11,15), and PPAR␥ antagonists show anti-obesity effects (24,25), we consider that both PPAR␣ agonistic and PPAR␥ antagonistic activities of Z-551 contribute to the anti-obesity effects. In fact, the effects, of Z-551 in PPAR␣-deficient mice and CZD-2 in WT mice suggest that the anti-obesity effects of Z-551 are not only due to the PPAR␣ activation but also due to the PPAR␥ inhibition.
The effects of Z-551 on amelioration of insulin resistance were examined. In WT and WT DIO mice, the OGTT and ITT indicated that Z-551 improved hyperglycemia and hyperinsulinemia induced by the HFD, resulting in amelioration of insulin resistance, whereas the HFD worsened insulin resistance, compared with mice given the ND (data not shown). The observations that treatment with Z-551 markedly reduced plasma insulin levels and blunted the increase in this hormone induced by oral glucose intake in the OGTT (Figs. 2P and 4Q) may be a consequence of the decreased insulin requirement caused by the improvement in glucose homeostasis and insulin sensitivity induced by Z-551 treatment, since Z-551 does not affect ␤-cell insulin secretion (data not shown). Z-551 significantly enhanced insulin-stimulated Akt phosphorylation in the liver and skeletal muscle, which is a principal regulatory factor for insulin signaling and glucose metabolism, suggesting that Z-551 improves insulin sensitivity.
It is known that the hypertrophy of visceral adipocytes is closely related to the formation of chronic low-grade inflammation in the periphery and the induction of insulin resistance mediated by abnormal production of pro-inflammatory adipokines (39 -41). There are reports to the effect that two types of macrophages, M1 and M2, are found in the adipose tissue suffering from macrophage infiltration, and that increases in M1 macrophages and the M1/M2 polarization of macrophages in the adipose tissue are closely involved in the onset and progression of insulin resistance (42)(43)(44)(45). Furthermore, Murakami et al. reported that the PPAR␣ activation reduces macrophagederived inflammation in cultured adipocytes by suppressing the expression of Tnf through the inhibition of the nuclear factor-B (NF-B) signaling pathway (46). Z-551 inhibited the expressions of the M1 macrophage markers, Ccl2, Tnf, and Tlr4 in the adipose tissue of WT DIO mice, without any effect on the expression of the M2 macrophage marker, Mrc1; implicating that Z-551 suppresses the increase in M1 macrophage infiltration into the adipose tissue and induction of inflammation. The gene expression analysis in PPAR␣-deficient mice administered Z-551 and WT mice administered CZD-2 demonstrates that both Z-551 and CZD-2 tended to inhibit gene expressions of macrophage markers. This observation suggests that both drugs probably inhibited HFD-induced macrophage infiltration into the adipose tissue. Taken together, these findings suggest that Z-551 ameliorates obesity-induced insulin resistance in part by the inhibition of macrophage infiltration and inflammation through not only PPAR␣ activation, but also PPAR␥ inhibition.
We investigated plasma glucose-lowering effect of Z-551. Z-551 significantly reduced fasting plasma glucose levels in WT and WT DIO mice, whereas a selective PPAR␥ antagonist, CZD-2, did not reduce the levels in WT mice. On the other hand, Z-551 reduced plasma insulin levels in WT, WT DIO, or PPAR␣-deficient mice, while CZD-2 reduced the level in WT mice. These results suggest that PPAR␣ activation is mainly associated with the glucose-lowering action of Z-551, although not only PPAR␣ activation, but also PPAR␥ inhibition are involved in the improvement of insulin sensitivity. No suppressive effect of Z-551 on food intake indicates involvement of the modulations of glucose and lipid metabolisms including fat accumulation, gluconeogenesis, and glucose utilization in the liver or skeletal muscle in the glucose-lowering action of Z-551. In fact, Z-551 decreased the concentration of hepatic TG content and improved insulin sensitivity. This observation was probably attributable to the inhibition of gluconeogenesis in the liver and the increase in glucose utilization in the skeletal muscle. However, the precise mechanisms for the glucose-lowering effect of Z-551 remain to be elucidated.
Z-551 reduced plasma TG and FFA levels in both WT and WT DIO mice, whereas Z-551 and CZD-2 had no effect on those levels in PPAR␣-deficient and WT mice, indicating that the reductions of plasma TG and FFA levels are caused by PPAR␣-dependent action of Z-551, which enhances fatty acid combustion in the liver via PPAR␣ activation. The increase in plasma FFA is known to provoke insulin resistance by inhibiting the insulin-signaling pathway in the liver and skeletal muscle as well as TNF-␣, and also inducing inflammation in the adipose tissue infiltrated with macrophages (18,47). Based on these findings, we have thought that the decrease in plasma FFA levels by Z-551 is partially responsible for the improved insulin resistance.
Excessive fat accumulation and increased oxidative stress due to reactive oxygen species in the liver are known to contribute to the onset of insulin resistance and fatty liver (48,49). Z-551 has the following effects: to significantly enhance the expressions of PPAR␣ target genes involved in fatty acid uptake and oxidation such as Lpl and Acox1 in the liver in WT and WT DIO mice, and to reduce the concentrations of hepatic TG and TBARS in WT DIO mice. According to these results, Z-551 will be clinically useful for the prevention and treatment of fatty liver diseases including nonalcoholic fatty liver diseases (NAFLD) and nonalcoholic steatohepatitis (NASH).
Several reports on thiazolidinediones (TZDs), PPAR␥ agonists, demonstrate that these compounds inhibited ectopic fat accumulation and inflammation in the peripheral tissues and increased insulin sensitivity by converting hypertrophic adipocytes (large adipocytes) to normal adipocytes (small adipocytes) (17,18). TZDs are, however, known to develop various adverse effects including increased body weight, fluid retention, and cardiac failure (50,51). On the other hand, the present study demonstrates that Z-551 and CZD-2 having PPAR␥ antagonistic activities inhibited body weight gain. Z-551 had no effect on hematocrit level, which is a marker of fluid retention, whereas rosiglitazone, a TZD, significantly decreased hematocrit level (data not shown). In recent years, epidemiological studies have discussed the problem of bone fragility in diabetic patients administered TZDs and a close relationship between PPAR␥ activation and bone fracture risk (52,53). Although the effects of PPAR␥ inhibition on bone fracture have not been sufficiently elucidated, drugs inhibiting PPAR␥ are considered to be able to avoid an increased risk of bone fracture. In fact, Z-551 had no effect on bone mineral density, morphology, and adiposity, whereas rosiglitazone significantly decreased bone mineral density and increased adiposity, leading to an increased risk of bone fracture in the femur of WT mice on the HFD (Fig.  8). Thus, the adverse effects of PPAR␥ agonists such as TZDs may be avoidable with Z-551.
Analysis of liver weight in WT, WT DIO, and PPAR␣-deficient mice administered Z-551 shows that the liver hypertrophy caused by Z-551 was likely to be mediated by PPAR␣ activation. However, Z-551 did not significantly increase plasma ALT and AST levels of the liver toxicity markers in WT mice on the HFD (data not shown). We have thought that the risk of liver hypertrophy in humans is low, because PPAR␣ agonists such as fenofibrate used in humans provoke no liver hypertrophy. Chronic hepatic PPAR␣ activation is reported to cause an increased risk of hepatic cancer in rodents, but recent studies have confirmed that there are differences between rodents and humans in the transcription control mechanism of cancer-related genes and gene expression levels of PPAR␣; thus, the risk of hepatic cancer induced by PPAR␣ activation in humans can be little (54,55). Based on these findings, Z-551 may have little risk of liver hypertrophy and hepatic cancer in humans. Further studies on the adverse effects of Z-551 are warranted.
The present study demonstrates that a novel and promising PPAR modulator, Z-551, having both PPAR␣ agonistic and PPAR␥ antagonistic activities, ameliorates HFD-induced obesity, insulin resistance, and impairment of glucose and lipid metabolisms in mice as expected in our concept of drug discovery (Fig. 9). Furthermore, Z-551 causes no adverse effect on body weight gain, and might reduce risks of fluid retention and bone fracture, which are induced by PPAR␥ agonists including TZDs. In conclusion, we consider that Z-551 will be clinically useful for the prevention or treatment of diet-induced obesity and obesity-related metabolic disorders such as insulin resistance, type 2 diabetes, dyslipidemia, and fatty liver diseases.