|Home | About | Journals | Submit | Contact Us | Français|
Peroxisome proliferator–activated receptor (PPAR) β/δ–null mice exhibit exacerbated hepatotoxicity in response to administration of carbon tetrachloride (CCl4). To determine whether ligand activation of the receptor protects against chemical toxicity in the liver, wild-type and PPARβ/δ-null mice were administered CCl4 with or without coadministration of the highly specific PPARβ/δ ligand GW0742. Biomarkers of liver toxicity, including serum alanine aminotransferase (ALT) and hepatic tumor necrosis factor (TNF) α mRNA, were significantly higher in CCl4-treated PPARβ/δ-null mice compared to wild-type mice. Hepatic expression of TNF-like weak inducer of apoptosis receptor (TWEAKr) and S100 calcium–binding protein A6 (S100A6/calcyclin), genes involved in nuclear factor kappa B signaling, was higher in the CCl4-treated PPARβ/δ-null mice compared to wild-type mice. GW0742 treatment resulted in reduced serum ALT concentration and lower expression of CCl4-induced TNF-α, S100A6, monocyte chemoattractant protein-1 (MCP1), and TWEAKr in wild-type mice, and these effects were not observed in PPARβ/δ-null mice. Expression of TNF-α was higher in PPARβ/δ-null primary hepatocytes in response to interleukin-1β treatment compared to wild-type hepatocytes, but GW0742 did not significantly modulate TNF-α expression in hepatocytes from either genotype. While PPARβ/δ-null hepatic stellate exhibited higher rates of proliferation compared to wild-type cells, GW0742 did not affect α-smooth muscle actin expression in these cells. Combined, these findings demonstrate that ligand activation of PPARβ/δ protects against chemically induced hepatotoxicity by downregulating expression of proinflammatory genes. Hepatocytes and hepatic stellate cells do not appear to directly mediate the inhibitory effects of ligand activation of PPARβ/δ in liver, suggesting the involvement of paracrine and autocrine events mediated by hepatic cells.
Peroxisome proliferator–activated receptors (PPARs) are members of the nuclear hormone receptor superfamily. In response to ligand activation, PPARs heterodimerize with the retinoid X receptor α leading to the disassociation of corepressors, recruitment of coactivators, binding to peroxisome proliferator response elements in target genes, and an increase in transcription (Berger and Moller, 2002; Willson et al., 2000). In addition to this classic receptor-mediated transcriptional upregulation of target genes, PPARs can also modulate gene expression by physically interacting with other transcription factors. For example, PPARs can interact with nuclear factor kappa B (NF-κB) (Delerive et al., 1999; Planavila et al., 2005; Westergaard et al., 2003) and the c-Jun molecule of the AP-1 complex (Delerive et al., 1999). There are three PPAR subtypes, PPARα (NR1C1), PPARβ/δ (NR1C2), and PPARγ (NR1C3) (Shearer and Hoekstra, 2003). All three PPARs can be activated by fatty acids and fatty acid derivatives (Forman et al., 1997; Kliewer et al., 1997). Synthetic agonists for the three receptors have also been developed. PPARα is activated by the fibrate class of lipid-lowering drugs, and the insulin-sensitizing thiozolidinediones are agonists for PPARγ (reviewed in Peraza et al., 2006). PPARβ/δ is activated by GW0742 and GW501516 with high affinity (reviewed in Peraza et al., 2006).
In addition to ligand selectivity, the expression pattern of PPARs provides another level of regulation. For example, PPARα is highly expressed in the liver (Auboeuf et al., 1997; Braissant et al., 1996), and PPARγ is expressed significantly in adipose tissues (Braissant et al., 1996; Fajas et al., 1997). In contrast, PPARβ/δ is ubiquitously expressed in most tissues (Escher et al., 2001) with notably higher expression in intestine, keratinocytes, and liver (Girroir et al., 2008b). Consistent with their expression pattern, PPARα has an essential role in regulating fatty acid oxidation in the liver, and PPARγ regulates terminal differentiation of adipose and lipid storage (reviewed in Peraza et al., 2006). Given the ubiquitous expression pattern of PPARβ/δ, it is not surprising that a variety of biological roles for this receptor have been elucidated. Ligand activation of PPARβ/δ by GW501516 can significantly improve glucose homeostasis (Lee et al., 2006), increase serum high density lipoprotein cholesterol (Leibowitz et al., 2000; Oliver et al., 2001), and increase skeletal muscle fatty acid catabolism (Tanaka et al., 2003). PPARβ/δ also mediates terminal differentiation (Aung et al., 2006; Burdick et al., 2007; Di Loreto et al., 2007; Hollingshead et al., 2008; Kim et al., 2006; Man et al., 2007; Marin et al., 2006; Matsuura et al., 1999; Nadra et al., 2006; Saluja et al., 2001; Schmuth et al., 2004; Tan et al., 2001; Varnat et al., 2006; Vosper et al., 2003; Werling et al., 2001; Westergaard et al., 2001) and inhibits cell proliferation in a variety of model systems (Ali et al., 2005; Burdick et al., 2007; Di Loreto et al., 2007; Fukumoto et al., 2005; Girroir et al., 2008a; Hollingshead et al., 2007a, 2008; Kim et al., 2004a, 2005, 2006; Lim et al., 2008; Man et al., 2007; Marin et al., 2006; Martinasso et al., 2006; Matthiessen et al., 2005; Michalik et al., 2001; Müller-Brüsselbach et al., 2007; Otsuyama et al., 2007; Peters et al., 2000; Planavila et al., 2005; Sertznig et al., 2008; Teunissen et al., 2007; Westergaard et al., 2001; Yang et al., 2008). Lastly, PPARβ/δ has anti-inflammatory activities including inhibition of cytokine production, modulation of cell adhesion molecules, and inhibiting NF-κB signaling (Arsenijevic et al., 2006; Barish et al., 2008; Bassaganya-Riera et al., 2004; Ding et al., 2006; Dyroy et al., 2007; Fan et al., 2008; Graham et al., 2005; Hollingshead et al., 2007b; Jakobsen et al., 2006; Kang et al., 2008; Kilgore and Billin, 2008; Kim et al., 2006, 2008; Nagasawa et al., 2006; Odegaard et al., 2008; Peters et al., 2000; Planavila et al., 2005; Polak et al., 2005; Ravaux et al., 2007; Riserus et al., 2008; Rival et al., 2002; Rodriguez-Calvo et al., 2008; Schmuth et al., 2004; Shan et al., 2008; Sheng et al., 2008; Takata et al., 2008; Welch et al., 2003; Woo et al., 2006). Given that there is strong evidence from multiple independent laboratories indicating that PPARβ/δ mediates terminal differentiation, inhibits cell proliferation, and has anti-inflammatory activities, it is not surprising that there are also reports demonstrating that PPARβ/δ attenuates carcinogenesis (Harman et al., 2004; Hollingshead et al., 2008; Kim et al., 2004a; Marin et al., 2006; Reed et al., 2004). However, others report that ligand activation of PPARβ/δ promotes tumorigenesis (Gupta et al., 2000; Wang et al., 2004, 2006; Xu et al. 2006a,b). Such conflicting data may suggest context- and/or tissue-specific roles for PPARβ/δ.
The role of PPARβ/δ in liver function has not been examined extensively. Previous work shows that PPARβ/δ can ameliorate carbon tetrachloride (CCl4)–induced liver toxicity (Shan et al., 2008). In response to administration of CCl4, liver toxicity is exacerbated in the absence of PPARβ/δ expression as shown by significantly elevated serum alanine aminotransferase (ALT) and hepatic tumor necrosis factor (TNF) α in PPARβ/δ-null mice as compared to similarly treated wild-type mice (Shan et al., 2008). Enhanced expression of proinflammatory genes that are regulated by the NF-κB signaling pathway including TNF-like weak inducer of apoptosis receptor (TWEAKr), CD14, and ICAM1 was also noted in PPARβ/δ-null mice following CCl4 administration (Shan et al., 2008). These changes were consistent with increased NF-κB DNA–binding activity in CCl4-treated PPARβ/δ-null mouse liver as compared to the wild type (Shan et al., 2008) and suggest that PPARβ/δ may protect against liver toxicity from CCl4 by modulating excessive NF-κB activity. This idea is supported by other studies showing that PPARs can interfere with NF-κB signaling (Delerive et al., 1999; Planavila et al., 2005; Westergaard et al., 2003). Combined, these findings suggest that PPARβ/δ protects against liver toxicity in the absence of exogenous ligands and support the hypothesis that ligand activation of PPARβ/δ with high-affinity agonists might provide further protection against liver toxicity. This hypothesis was examined in the present study. Additionally, whether PPARβ/δ modulates cell-specific activities that might influence CCl4-induced toxicity was examined using primary hepatocyte and primary hepatic stellate cell cultures.
Animal experiments were approved by the Institutional Animal Care and Use Committee at The Pennsylvania State University, which conforms to the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health. Male wild-type or PPARβ/δ-null mice on a C57BL/6 genetic background, aged 6–8 weeks, were housed in a temperature-controlled environment (25°C) with a 12-h light/dark cycle. The animals were fed a standard rodent chow with ad libitum access to drinking water. A cohort of wild-type and PPARβ/δ-null mice were administered CCl4 (0.5 ml/kg body weight) by ip injection twice per week for 17 days. Another cohort of mice from both genotypes were administered the highly specific PPARβ/δ agonist GW0742 (5 mg/kg) by oral gavage 5 days a week for 17 days in addition to the CCl4 treatment. Control mice were administered the same volume of corn oil. Mice were euthanized at end of the second week, and the liver was removed and snap frozen in liquid nitrogen.
Liver tissue was fixed in 10% neutral buffered formalin and embedded in paraffin. Sections were prepared and stained with hematoxylin and eosin and examined under a light microscope. The entire representative liver section from the same lobe from five mice per group was examined and scored by a pathologist. For each section, the percentage of tissue that showed coagulative centrilobular necrosis was determined using light microscopy. Similarly, the amount of inflammatory cell infiltrate was estimated based on the distribution and severity of the lesion. Based on this analysis, a scoring system was developed, whereby a score of 1 was assigned when the least amount of necrosis, the least amount of inflammatory cell infiltrate, and few mitotic figures were observed; a score of 2 was assigned when mild to moderate necrosis, mild to moderate inflammatory cell infiltrate, and a mild to moderate number of mitotic figures were observed; a score of 3 was assigned when moderate necrosis, moderate inflammatory cell infiltrate, and a moderate number of mitotic figures were observed; and a score of 4 was assigned when severe necrosis, severe inflammatory cell infiltrate, and numerous mitotic figures were observed.
Serum was obtained from mice, and the concentration of L-alanine:2-oxoglutarate aminotransferase (ALT; EC220.127.116.11) was measured using INFINITY125 ALT Reagent (Thermo Electron, Melbourne, Australia) according to the manufacturer's recommended procedure.
The mRNAs encoding TNF-α, TWEAKr, S100 calcium–binding protein A6 (S100A6), monocyte chemoattractant protein-1 (MCP1), and 18s rRNA were quantified using real-time PCR analysis. The cDNA was generated from total RNA using 2.5 μg total RNA with MultiScribe Reverse Transcriptase kit (Applied Biosystems, Foster City, CA). Primers were designed for real-time PCR using the Primer Express software (Applied Biosystems). The sequence and Genbank accession numbers for the forward and reverse primers used to quantify mRNAs were TNF-α (NM_013693) forward: 5′-TGGAGTCATTGCTCTGTGAAGGGA-3′ and reverse: 5′-AGTCCTTGATGGTGGTGCATGAGA-3′, TWEAKr (NM_013749) forward: 5′-GGCGCTGGTTTCTAGTTTCCT-3′ and reverse: 5′-GGGTGCTCCTCACTGGATCA-3′, S100A6 (NM_011313) forward: 5′-GTACTCTGGCAAGGAAGGTGACA-3′ and reverse: 5′-CAGCATCCTGCAGCTTGGA-3′, MCP1 (NM_011333) forward: 5′-TCTCTCTTCCTCCACCACCATG-3′ and reverse: 5′-GCGTTAACTGCATCTGGCTGA-3′, and 18s RNA (X006686) forward: 5′-TCGATGCTCTTAGCTGAGTGTCCC-3′ and reverse: 5′-TATTCCTAGCTGCGGTATCCAGGC-3′. Real-time PCRs were carried out using SYBR green PCR master mix (Finnzymes, Espoo, Finland) in the iCycler and detected using the MyiQ real-time PCR detection system (Bio-Rad Laboratories, Hercules, CA). The following conditions were used for PCR: 95°C for 15 s, 94°C for 10 s, 60°C for 30 s, and 72°C for 30 s and repeated for 45 cycles. The PCR included a no-template control reaction to control for contamination and/or genomic amplification. All reactions had > 90% efficiency. Melting curves were performed for all quantitative real-time PCRs (qPCRs). Relative expression levels of mRNA were normalized to 18s rRNA.
Liver cytosol fractions were isolated from mice as previously described (Kim et al., 2005). Cytosolic proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride membranes as previously described (Kim et al., 2005). Membranes were incubated with primary antibody against TWEAKr (Cell Signaling Technology, Inc., Beverly, MA) or actin (Rockland, Inc., Gilbertsville, PA), followed by incubation with biotinylated secondary antibodies and then 125I-labeled streptavidin as previously described (Kim et al., 2005). Hybridization signals were normalized to actin or lactate dehydrogenase and are presented as the fold change from wild-type control.
Primary hepatocytes from wild-type and PPARβ/δ-null mice were isolated as described previously (Stoehr and Isom, 2003). Two experiments were performed with the primary hepatocytes. For the first experiment, wild-type and PPARβ/δ-null primary hepatocytes were cultured with or without TNF-α (2500 U/ml) for either 24 or 48 h. Bromodeoxyuridine (BrdU) was added to the medium for the final 24 h of culture. The percentage of BrdU-labeled nuclei was quantified at the end of the experiment. For the second experiment, wild-type and PPARβ/δ-null primary hepatocytes were incubated with either dimethyl sulfoxide or GW0742 (0.1, 0.5, or 1.0μM) for 1 h. Interleukin (IL) 1β (PeproTech, Inc., Rocky Hill, NJ) was then added to the medium (10 ng/ml) for an additional 4 h of culture. Cells were lysed with TRIZOL reagent (Invitrogen, Carlsbad, CA), and RNA was isolated as described above. Total RNA was used to quantify mRNA-encoding TNF-α using qPCR as described above.
Hepatic stellate cells from wild-type and PPARβ/δ-null mice were isolated by a modified method previously described (Liu et al., 2003). Mice were anesthetized with nebutal. In situ liver perfusion and digestion was performed with pronase E and collagenase B, and the resulting liver cell suspension was purified by density gradient centrifugation using 13% Nycodenz. Hepatic stellate cells were plated on uncoated 60-mm dish at a density of 5 × 106 cells per dish and maintained in Dulbecco’s modified Eagle’s medium with F12 (HyClone, Logan, UT) supplemented with 50 μg/ml gentamycin (Sigma, St Louis, MO), 50 ng/ml amphrotericin B (Sigma), and 20% fetal bovine serum. Cell viability was assessed by vitamin A autofluorescence microscopy. Culture medium was changed every 2–3 days. Cells were passaged up to nine times, and cell growth was monitored visually during the culture period based on the relative number of cell passages. Protein samples were obtained from wild-type and PPARβ/δ-null hepatic stellate cells at different passages and examined for α-smooth muscle actin (α-SMA) expression as a marker of stellate cell activation using Western blot analysis as described above.
Differences between treatments were determined using ANOVA followed by Dunnett or Bonferroni post hoc tests (Prism 4.0, GraphPad Software, Inc., San Diego, CA). Significant differences were determined when p≤0.05.
To test the hypothesis that ligand activation of PPARβ/δ protects against chemically induced liver damage, wild-type and PPARβ/δ-null mice were treated with CCl4 with and without GW0742 for 17 days. Histological analysis revealed a significant increase in centrilobular coagulative necrosis in PPARβ/δ-null mouse liver in response to CCl4 as compared to wild type (Fig. 1A and B). Similarly, the average level of necrosis and inflammatory cell infiltrate was higher in CCl4-treated PPARβ/δ-null mouse liver as compared to wild type (Fig. 1C). Administration of GW0742 did not significantly decrease the level of centrilobular necrosis (Fig. 1A and B) or necrosis and inflammatory cell infiltrate (Fig. 1C), but there was a trend for lower severity of these indices in the wild-type mice cotreated with GW0742 as compared to wild type treated solely with CCl4, and this trend was not found in similarly treated PPARβ/δ-null mice (Fig. 1A–C). Consistent with the histopathology, serum ALT and hepatic expression of mRNA-encoding TNF-α were significantly higher in CCl4-treated PPARβ/δ-null mice as compared to the similarly treated wild-type mice (Fig. 2A and B). Ligand activation of PPARβ/δ caused a significant decrease in serum ALT and hepatic expression of mRNA-encoding TNF-α in CCl4-treated wild-type mice, and this effect was not found in PPARβ/δ-null mice (Fig. 2A and B). Thus, markers of CCl4-induced hepatotoxicity are significantly reduced by the presence and activation of PPARβ/δ.
Previous work showed that proinflammatory genes involved in the NF-κB signaling pathway are overexpressed in response to CCl4 in PPARβ/δ-null mouse liver as compared to wild type, including TWEAKr and S100A6 (Shan et al., 2008). Hepatic expression of mRNA-encoding TWEAKr and S100A6 was notably higher in CCl4-treated PPARβ/δ-null mice as compared to wild type (Fig. 3A and C). In response to ligand activation of PPARβ/δ in wild-type mice, a reduction of CCl4-induced S100A6 and MCP1 expression was observed, and this effect did not occur in similarly treated PPARβ/δ-null mice (Fig. 3A and B). In response to ligand activation of PPARβ/δ, no statistically significant decrease in CCl4-induced TWEAKr mRNA expression was observed (Fig. 3C). However, Western blot analysis of TWEAKr expression demonstrated that ligand activation of PPARβ/δ caused a marked decrease in CCl4-induced TWEAKr expression in wild-type mice that was not observed in the absence of PPARβ/δ expression (Fig. 3D).
To begin to determine the specific liver cell types in which PPARβ/δ modulates CCl4-induced liver toxicity, primary hepatocytes were examined from wild-type and PPARβ/δ-null mice. TNF-α is a major cytokine that mediates CCl4-induced liver toxicity and can cause a mitogenic response in cultured primary hepatocytes (Stoehr and Isom, 2003). A modest but statistically insignificant increase in relative BrdU labeling was observed in hepatocytes from PPARβ/δ-null mice following 24-h exposure to TNF-α, and this was not found in TNF-α–treated wild-type hepatocytes (Fig. 4A). After 48 h of treatment with TNF-α, an increase in relative BrdU labeling was observed in wild-type hepatocytes (Fig. 4A). However, relative BrdU labeling was markedly increased in PPARβ/δ-null hepatocytes after 48-h exposure to TNF-α as compared to similarly treated hepatocytes from wild-type mice (Fig. 4A). Hepatic stellate cells have critical roles in the initiation and progression of liver fibrogenesis, which can occur with CCl4 toxicity (Friedman, 2008). Thus, it is possible that PPARβ/δ modulates essential signaling pathways in hepatic stellate cells that attenuate CCl4-induced hepatotoxicity. Morphological examination revealed no difference in the shape or size of hepatic stellate cells between wild-type and PPARβ/δ-null mice (Fig. 4B). However, hepatic stellate cells from PPARβ/δ-null mice proliferated much faster than wild-type cells and required passaging much sooner as compared to hepatic stellate cells from wild-type mice (data not shown). Examination of α-SMA expression in passaged hepatic stellate cells indicated that PPARβ/δ-null cells were comparatively more activated as there was an increase in α-SMA expression as compared to hepatic stellate cells from wild-type mice (Fig. 4C).
To determine the effect of ligand activation of PPARβ/δ in primary hepatocytes, cells were cultured in the presence of IL-1β with and without pretreatment (4 h) with GW0742. IL-1β can directly upregulate TNF-α expression in human cell lines (Bethea et al., 1992) and in mouse primary hepatocytes (data not shown). Treating primary hepatocytes with GW0742 at concentrations known to specifically activate PPARβ/δ in liver cells (Hollingshead et al., 2007a) did not influence the IL-1β–induced increase in TNF-α mRNA expression (Fig. 5A). However, the level of IL-1β–induced TNF-α expression in primary hepatocytes was significantly higher in the absence of PPARβ/δ expression as compared to wild-type cells (Fig. 5A). Similarly, α-SMA expression was notably higher in PPARβ/δ-null stellate cells, but treating hepatic stellate cells with GW0742 did not consistently influence α-SMA expression (Fig. 5B and data not shown). No consistent changes in the expression of transforming growth factor (TGF) β, collagen type I α1 (COL1α1), platelet-derived growth factor β (PDGFβ), and its receptor (PDGFrβ) were observed between either the genotype or in response to GW0742 (data not shown).
PPARβ/δ-null mice exhibit exacerbated hepatotoxicity in response to administration of CCl4 (Shan et al., 2008), suggesting that ligand activation of the receptor may protect against CCl4-induced liver toxicity. Results from the present study support this hypothesis as markers of CCl4-induced liver toxicity are significantly reduced in response to the highly specific PPARβ/δ ligand, GW0742. Since reduced levels of CCl4-induced serum ALT concentrations and inhibited expression of hepatic TWEAKr, MCP1, and S100A6 mRNAs were not changed in similarly treated PPARβ/δ-null mice, this demonstrates that the protective effect of GW0742 requires a functional receptor. TWEAK is a potent activator of the NF-κB signaling pathway in multiple cell types, including skeletal muscle (Dogra et al., 2006), central nervous system (Polavarapu et al., 2005), and macrophages (Kim et al., 2004b). TWEAK binds to its cognate receptor, TWEAKr or Fn14 (fibroblast growth factor–inducible 14), which in turn is bound by tumor necrosis factor receptor-associated factors 1, 2, 3, and 5, culminating in NF-κB activation (Brown et al., 2003). Similarly, S100A6 is a transcriptional target of NF-κB, and an NF-κB–binding sequence has been localized in the promoter of S100A6 (Joo et al., 2003). Previous work demonstrated that PPARβ/δ protects against CCl4-induced liver toxicity in the absence of exogenous ligands by interfering with excessive NF-κB signaling (Shan et al., 2008). This is consistent with other findings showing that PPARs modulate NF-κB transcriptional activity (Delerive et al., 1999; Planavila et al., 2005; Westergaard et al., 2003). Thus, the protective effect observed in the absence of exogenous ligands may be the result of endogenous ligand activation of PPARβ/δ and subsequent interaction with NF-κB. Indeed, ligand activation of PPARβ/δ can attenuate NF-κB function in cardiomyocytes (Ding et al., 2006), endothelial cells (Rival et al., 2002), and skeletal muscle cells (Woo et al., 2006). Results from the present study suggest that modulation of NF-κB signaling by ligand activation of PPARβ/δ is one of the mechanisms by which PPARβ/δ attenuates CCl4 hepatotoxicity.
In contrast to the present findings, enhanced CCl4-induced liver toxicity is reported to occur in rats when administered in combination with another PPARβ/δ ligand, L165041 (Hellemans et al., 2003). Hellemans et al. examined hepatotoxicity after a single CCl4 treatment and after 5 weeks of CCl4 treatment and reported evidence of exacerbated hepatotoxicity by coadministration of L165041 in both models. In contrast, GW0742 did not protect or exacerbate hepatotoxicity following a single CCl4 treatment in the mouse model as detected by serum ALT levels (data not shown) while GW0742 protected against hepatotoxicity following 17 days of coadministration. The reason for the observed differences in the Hellemans study and the present study cannot be determined from the present work. Since the dosing paradigms of both CCl4 and the PPARβ/δ ligand examined by both groups are very similar, it remains possible that this difference in response is due in part to species differences (rat vs. mouse) or the ligand used (L165041 vs. GW0742). However, it is also worth noting that there is a large body of evidence that is consistent with the idea that ligand activation of PPARβ/δ can protect against chemically induced liver toxicity. For example, ligand activation of PPARβ/δ by GW501516 can significantly inhibit experimentally induced steatohepatitis in mice including reduced hepatic expression of TGF-β1, IL-6, IL-1β, MCP1, TNF-α, and NF-κB (Nagasawa et al., 2006). These findings are highly consistent with results from the present studies and previous work (Shan et al., 2008) where PPARβ/δ was also noted to inhibit TNF-α expression and NF-κB activity. Additionally, these effects are also consistent with a large body of evidence from multiple laboratories demonstrating anti-inflammatory activities of PPARβ/δ (Arsenijevic et al., 2006; Barish et al., 2008; Bassaganya-Riera et al., 2004; Ding et al., 2006; Dyroy et al., 2007; Fan et al., 2008; Graham et al., 2005; Hollingshead et al., 2007b; Jakobsen et al., 2006; Kang et al., 2008; Kilgore and Billin, 2008; Kim et al., 2006, 2008; Nagasawa et al., 2006; Odegaard et al., 2008; Peters et al., 2000; Planavila et al., 2005; Polak et al., 2005; Ravaux et al., 2007; Riserus et al., 2008; Rival et al., 2002; Rodriguez-Calvo et al., 2008; Schmuth et al., 2004; Shan et al., 2008; Sheng et al., 2008; Takata et al., 2008; Welch et al., 2003; Woo et al., 2006). In particular, it is of interest to note that GW501516 can suppress IL-6–mediated acute phase reactions by inhibiting STAT3 in human and rat hepatocytes (Kino et al., 2007). Collectively, while there is a report suggesting that ligand activation of PPARβ/δ can enhance CCl4-induced liver toxicity in rats (Hellemans et al., 2003), this is highly inconsistent with a larger body of literature and may be due to differences in species and/or ligands.
To begin to determine the specific cell types where PPARβ/δ functions to protect against liver toxicity, primary hepatocytes and hepatic stellate cell cultures were examined. Consistent with previous work showing enhanced CCl4-induced liver toxicity in the absence of PPARβ/δ expression, enhanced BrdU labeling in response to TNF-α was observed in primary hepatocytes from PPARβ/δ-null mice as compared to wild type. Similarly, TNF-α mRNA was significantly higher in IL-1β–treated PPARβ/δ-null hepatocytes as compared to the wild type. However, GW0742 did not modulate TNF-α expression in either wild-type or PPARβ/δ-null hepatocytes. In addition, expression of mRNA-encoding TWEAKr or S100A6 was not influenced by GW0742 treatments in either genotypes (data not shown). These findings suggest that hepatocytes may not be the primary cell where PPARβ/δ modulates liver toxicity. Other cell types in the liver, including Kupffer cells, endothelial cells, and hepatic stellate cells, are also known to have major roles in producing proinflammatory molecules that impact neighboring hepatocytes (Li and Friedman, 1999). Indeed, cell proliferation was considerably accelerated in PPARβ/δ-null hepatic stellate cell as compared to the wild type, and constitutive expression of α-SMA was markedly higher in passaged hepatic stellate cells from PPARβ/δ-null mice, suggesting that PPARβ/δ can modulate stellate cell activation. However, markers of hepatic stellate cell activation and proliferation including α-SMA, COL1α1, TGFβ1, PDGFβ, and PDGFrβ were not consistently altered by GW0742 (data not shown), and TWEAKr and S100A6 were not expressed in a PPARβ/δ-dependent manner in hepatic stellate cells either (data not shown). Since ligand activation of PPARβ/δ in hepatocytes did not alter IL-1β–induced expression of TNF-α, this suggests that PPARβ/δ does not appear to directly modulate this pathway in hepatocytes. Similarly, since ligand activation of PPARβ/δ in hepatic stellate cells did not consistently modulate stellate cell activity, this suggests that PPARβ/δ may not directly interact in this cell type. However, there are a number of other signaling pathways that should be examined in more detail in these cell types before concluding that PPARβ/δ does not protect against liver toxicity by modulation of molecular events in these cell types. Since Kupffer cells are the liver macrophages that regulate the severity of liver toxicity by releasing the majority of inflammatory cytokines (Roberts et al., 2007), the idea that ligand activation of PPARβ/δ may regulate Kupffer cell activity should be examined further. This is particularly interesting given the recent observations that PPARβ/δ inhibits inflammation and inflammatory signaling in macrophages (Kang et al., 2008). Similarly, PPARγ ligands can suppress inflammation by interrupting signaling pathways in Kupffer cells (Enomoto et al., 2003; Planaguma et al., 2005; Uchimura et al., 2001).
In summary, these results demonstrate that ligand activation of PPARβ/δ can inhibit CCl4-induced liver toxicity through a mechanism that involves PPARβ/δ-dependent downregulation of proinflammatory signaling. This effect is likely due in part to interactions of PPARβ/δ with NF-κB. While hepatocytes and hepatic stellate cells do not appear to be the direct cell types that mediate the protective effect, it remains possible that paracrine and/or autocrine signaling pathways involving more than one hepatic cell type underlies this protective effect. The possible role of PPARβ/δ in Kupffer cell activity should be explored further, especially due to the recent demonstration of significant PPARβ/δ-dependent anti-inflammatory activities in Kupffer cells and macrophages (Kang et al., 2008; Odegaard et al., 2008). Lastly, characterization of the mechanism underlying the regulation of NF-κB signaling pathway by PPARβ/δ may provide new molecular targets for treating a variety of inflammation-dependent diseases, including atherosclerosis, diabetes, and cancer.
National Institutes of Health (CA124533 to J.M.P., ES04869 to G.H.P., CA023931 to H.C.I.); Tobacco Settlement funds from the Pennsylvania State Department of Health.
The authors gratefully acknowledge Andrew Billin and Timothy Willson for providing GW0742 for these studies and Roberta Horner for technical assistance with these studies.