Cytochrome P450 4F isoforms catalyze the hydroxylation of eicosanoids such as leukotriene B4, prostaglandins, and lipoxins as well as hydroxyeicosatetraenoic acids. In the present study, we report the molecular cloning of two novel mouse CYP4F isoforms, CYP4F15 and CYP4F16. Sequence comparison showed that CYP4F15 has 93.5% homology to CYP4F4 and CYP4F16 has 90.8% homology to CYP4F5, therefore they are the orthologs for rat CYP4F4 and CYP4F5, respectively. Both isoforms are expressed in liver and also in extrahepatic tissues but the patterns of expression are slightly different. To elucidate further the regulation and regulatory mechanism of the two isoforms, renal and hepatic CYP4F15 and CYP4F16 expression were analyzed using wild-type (SV/129) mice and peroxisome proliferator-activated receptor (PPAR) α null mice with or without challenge by bacterial endotoxin (LPS) or clofibrate. Renal expression of CYP4F15 was induced by LPS and clofibrate in (+/+) mice, and these effects were absent in the (−/−) mice. Renal expression of CYP4F16 was not affected by LPS or clofibrate in (+/+) or (−/−) mice. In contrast, hepatic expression of CYP4F15 and CYP4F16 was significantly reduced by LPS-treatment in (+/+) mice. A lesser reduction was also seen in the (−/−) mice, suggesting that PPARα is partially responsible for this down-regulation. Clofibrate treatment caused the reduction of hepatic CYP4F16 expression and this effect was not dependent on PPARα. Clofibrate treatment had no effect on hepatic CYP4F15 expression. Together, our data indicate that CYP4Fs are regulated in an isoform-specific, tissue-specific, and species-specific manner.
The cytochromes P450 (P450) 4F comprise a growing P450 subfamily that to date includes four human isoforms (Kikuta et al., 1993, 1994; Bylund et al., 1999; Cui et al., 2000) and four rat isoforms (Chen and Hardwick, 1993; Kawashima and Strobel, 1995). The heterologously expressed enzymes catalyze the hydroxylation of several eicosanoids such as leukotriene B4, prostaglandin A1 and E1, as well as lipoxins and hydroxyeicosatetraenoic acids (Kawashima et al., 1997; Jin et al., 1998; Kikuta et al., 1998, 1999; Bylund et al., 1999). These eicosanoids are important mediators involved in the inflammatory cascade, which includes attraction of leukocytes, regulation of blood vessel tone, increase of vascular permeability, etc. It is a reasonable speculation that CYP4Fs may modulate the extent of inflammation through regulation of the level of these inflammatory mediators. Several lines of evidence support this hypothesis. Recombinant CYP4F3 catalyzes the conversion of LTB4 to 20-hydroxy LTB4with a K m of 0.64 μM, which matches the K m of LTB4hydroxylation in vivo by human neutrophils (Powell, 1984). This suggests that CYP4F3 might be solely responsible for inactivation of LTB4 in neutrophils. One study showed that leukocytes from patients with inflammatory bowel disease have a higherV max value for LTB4 than those from healthy volunteers (Ikehata et al., 1993).
Little is known about the regulation of CYP4Fs. Kawashima et al. (1997)showed that CYP4F4, 4F5, and 4F6 were down-regulated in rat liver when the animals were treated with clofibrate, a ligand for peroxisome proliferator-activated receptor α (Kawashima et al., 1997). In contrast, this drug is a strong inducer for CYP4As in both liver and kidney. The mechanism of CYP4A induction is a well documented one in which the ligand-activated PPARα heterodimerizes with the retinoid X receptor, binds to the peroxisome proliferator responsive elements (PPREs) in 5′ flanking region of a CYP4A gene and activates gene transcription (Muerhoff et al., 1992). PPARα activation also down-regulates the expression of several genes such as acute-phase-responsive genes and Apo-AI (Corton et al., 1998; Vu-Dac et al., 1998). Whether PPARα is involved in rat CYP4F down-regulation by clofibrate is unknown at the present time.
Evidence accumulated to date also suggests a potential role for PPARα in the inflammatory process. Besides its pharmacological ligands such as fibrate class compounds and Wy14,643, some endogenous compounds like fatty acids and eicosanoids are found to be ligands for PPARα (Issemann et al., 1993; Forman et al., 1995; Kliewer et al., 1995; Krey et al., 1997). Most of these natural ligands belong to the lipid mediator family that modulates the inflammatory response. Devchand et al. (1996) demonstrated that LTB4 may control a generalized inflammatory response by activating PPARα. PPARα activation results in an increase in expression of enzymes involved in the ω- or β-oxidation pathways, and hence, stimulates the catabolism of lipid mediators, thus serving as a feedback control for inflammation. The study of CYP4F catalytic activities showed that CYP4Fs are capable of eliminating LTB4 by ω-oxidation, and should fall into the category of the genes that could be regulated by PPARα activation.
Inflammation is recognized as a pathophysiological condition that can affect the expression of CYP 450s (Morgan, 1997). Endotoxin-treatment is a standard inflammation model used in studying P450 regulation. It has been shown that expression of CYP2C11, 2E1, and 3A2 are repressed in rat liver, whereas the expressions of CYP4As are induced by LPS treatment (Sewer et al., 1996). Similar results were obtained in mice, in that expression of CYP2A5, 2C29, and 3A11 were shown to be down-regulated. Surprisingly, the expression of CYP4A10 and 4A14 were also reduced in mice liver but induced in mice kidney (Barclay et al., 1999). The induction of renal CYP4As was absent in PPARα null mice, whereas the repression of the hepatic P450s was partially attenuated.
Based on our hypothesis that CYP4F might affect the inflammation cascade by modulating LTB4 levels, inflammation itself might be expected to change CYP4F expression. Thus, we set out to clone members of the CYP4F subfamily in the mouse, and to define their expression in LPS-treated mice. Additionally, clofibrate, a typical CYP4A inducer, was used to determine whether mouse CYP4Fs are regulated in a similar manner as 4As. Furthermore, the regulation was compared in wild-type and PPARα null mice (Lee et al., 1995) to assess the role of PPARα in mouse CYP4F regulation.
Materials and Methods
Construction of Mouse cDNA Library.
Poly(A)+ RNA was isolated from male mouse (129/SV-+Tyr c-ch Ter/+) brain and kidney by standard methods (Chomczynski and Sacchi, 1987). cDNA libraries were prepared with 5.6 μg of mouse brain mRNA and 6.2 μg of mouse kidney mRNA using the ZAP cDNA synthesis kit (Stratagene, La Jolla, CA). The mouse brain library contained 480,000 independent clones, and the mouse kidney library had 960,000 independent clones (per package reaction). The size of inserts ranged from 0.5 to 6.0 kb in both libraries.
cDNA Cloning and Sequencing.
The CYP4F4 cDNA containing the full-length coding sequence was radiolabeled with [α-32P]dCTP (NEN, Boston, MA) using the random primer labeling method. Plaque hybridization was carried out using 32P-labeled cDNA at 65°C in 50 mM Tris buffer (pH 7.5) containing 1 M NaCl, 10 mM EDTA, 0.1% sodiumN-lauroylsarcosinate, 0.2% polyvinylpyrrolidone, 0.2% Ficoll, 0.2% bovine serum albumin, and 100 μg/ml salmon sperm DNA. The filters were washed twice at 65°C for 30 min in 3× standard saline citrate (SSC) and 0.1% SDS. The positive plaques were purified through two more rounds of the screening procedures described above, and all the positive clones were in vivo-excised using ExAssist Helper Phage (Stratagene). The phagemid DNAs of the positive clones were subjected to Southern blot analysis, after digesting them withEcoRI and XhoI, using the32P-labeled cDNA of CYP4F4 as a probe. The sequence analysis was performed with deletion clones prepared using a ExoIII/Mung Bean Nuclease Deletion kit (Stratagene). The sequence reactions were carried out using BigDye Primer Cycle Sequence Ready Reaction kit or BigDye Terminator Cycle Sequencing kit (Applied Biosystems, Foster City, CA). Sequencing was performed in both directions for every chosen clone.
All animal treatment procedures were approved by the Institutional Animal Care and Use Committee of Emory University. Groups of four female 9- to 11-week-old wild-type (+/+) or (−/−) mice (F5 SV/129) homozygous for a disruption in the ligand-binding domain of the PPARα gene were used. Animals were housed in group cages containing three to five animals per cage and provided food and water. Animals were given a single i.p. injection of either 1 mg/kg of chromatographically purified Escherichia coli LPS, serotype 0127:B8 (Sigma Chemical Co., St. Louis, MO) dissolved in sterile 0.9% saline, 40 mg/kg clofibrate (Sigma Chemical Co.) dissolved in corn oil (Wesson), or 0.9% saline in volume of 0.1 ml. Animals were sacrificed 24 h after injection by CO2 asphyxiation and decapitation before collection of organs.
Northern Blot Analysis.
For study of the tissue distribution of mouse cytochrome P450 4F expression study, a poly(A)+ RNA blot was purchased from Clontech (Palo Alto, CA) and used as instructed by the manufacturer. For experiments to determine the role of PPARα in CYP4F regulation, total RNA was fractionated by electrophoresis on a 1.3% agarose gel in the presence of 5% formaldehyde and transferred to a charged nylon membrane (Schleicher & Schuel, Keene, NH). Hybridization was carried out at 65°C in RapidHyb buffer (Amersham, Arlington Heights, IL) overnight. The blot was washed twice in 2× SSC at 65°C for 30 min, and then in 0.1× SSC, 0.1% SDS at 65°C for 30 min. To control for RNA loading and transfer variation, Northern blots were normalized to the content of glyceraldehyde 3-phosphate dehydrogenase mRNA with a cDNA probe. The amount of the probe bound to the filter was quantitated with a PhosphorImager (Packard, Meriden, CT).
The cDNA probe for mouse CYP4F15 was amplified by polymerase chain reaction using 4F15 up (5′-CTC ACT GTT CCT GAG AGA AG-3′) as forward primer, m4Flow (5′-ATG TCC TCA TCT GAC AGC TC-3′) as reverse primer, and pBluescript KS plasmid harboring full-length CYP4F15 cDNA as template. The polymerase chain reaction product was electrophoresed on 1% agarose gel and purified with Geneclean (Bio101, La Jolla, CA). The cDNA probe for CYP4F16 was generated by the same method, except CYP4F16-specific upstream primer 4F16 up (5′-TAG AGG ACC AGC AGC TCC TG-3′) and the plasmid harboring full-length CYP4F16 cDNA were used. The probes were labeled with random labeling kit (Stratagene) and [α-32P]dCTP (NEN).
In Vitro Translation.
The TNT T3 coupled reticulocyte lysate system (Promega, Madison, WI) was chosen for in vitro translation. Two micrograms of pBluescript SK (+/−) vector, which harbored the full-length CYP4F15 or CYP4F16 cDNA, was used as template in in vitro transcription and translation from the T3 promoter. [35S]methionine (1200 Ci/mmol) (Amersham) was incorporated into the newly synthesized protein. The empty pBluescript SK (+/−) vector served as control. The reactions were incubated at 30°C for 2 h. The translation products were analyzed on a 10% polyacrylamide-SDS gel and exposed to Kodak XR X-ray film overnight.
Slot Blot Assay.
Plasmids harboring the full-length mouse cytochrome P450 4F cDNAs were loaded onto charged nylon membranes in a slot blot manifold by gentle vacuum in presence of 10× SSC solution. The DNAs were denatured by floating the blots on 0.5 N NaOH, 0.15 M NaCl for 2 min, and neutralized with 0.6 M Tris-HCl (pH 7.5), 0.15 M NaCl. The blots were soaked in 2× SSC and UV cross-linked. The prehybridization, hybridization, and wash conditions were same as for Northern blotting described above.
Sequence Analysis and Accession Numbers.
The nucleotide sequences from library screening were analyzed by BLAST and FASTA at National Center for Biotechnology Information (www.ncbi.nlm.nih.gov). The homology between mouse and rat isoforms was analyzed by MEGALIGN of DNA Star and GCG (Genetics Computer Group, Inc., Madison, WI) programs. The GenBank accession numbers of CYP4F15 and CYP4F16 are AF233645 andAF233646, respectively.
For each treatment group, the mRNA content of CYP4F15 and CYP4F16 was expressed as a percentage of the control group mean ± S.E.M. Two-way ANOVA analysis of variance was used to analyze the significant differences between the group mean values.
Cloning and Sequence Analysis of CYP4F15 and CYP4F16.
To identify mouse cytochrome P450 4F subfamily members, 550,000 plaques from a brain cDNA library and 390,000 plaques from a kidney cDNA library were screened with 32P-labeled CYP4F4 cDNA as probe. Thirty-four positive clones from kidney and 23 positives from brain were isolated after the three rounds of screening. They were analyzed by Southern blot hybridization with the same probe after restriction enzyme digestion. Thirteen clones from brain and nine from kidney were selected for sequencing from both the 5′ and 3′ ends. The data showed the existence of five novel cDNAs. One full-length clone #33 from brain and one full-length clone K19 from kidney are reported here.
As shown in Fig. 1, clone #33 contains 2158 nucleotides, including 154-bp 5′ untranslated region and 424-bp 3′ untranslated region. Two open reading frames were identified at position 140 to 1744 and 155 to 1744. By comparison with other 4F family members, the second reading frame was judged to be the one more likely to be used. The presence of two in-frame termination codons upstream of the initiation methionine further proved that the full-length N-terminal coding sequence is present. Clone K19 covers 2192 nucleotides. An open reading frame at position 29 to 1603 was followed by 589-bp 3′ untranslated sequences. Both sequences have intact poly(A) tails preceded by polyadenylation signals (ATTAAA).
The amino acid sequences of both clones are shown in Fig.2A. Clone #33 encodes 529 amino acid residues with a calculated molecular mass of 60,628 Da. Clone K19 encodes 524 amino acid residues with predicted molecular mass of 60,229 Da. In vitro translation of PBluescript KS plasmids harboring the full-length clones #33 and K19 in the T3 coupled reticulocyte lysate system produced two proteins with molecular masses of 60,000 Da, which are in close agreement with calculated molecular masses (Fig.2B). The two proteins have a 72.5% amino acid sequence homology to each other, and contain the conserved CYP4 family consensus sequence and heme-binding region with the invariant cysteine highlighted in Fig.2A. These two novel mouse cytochrome P450s are named CYP4F15 and CYP4F16 according to the standardized nomenclature. Database searches and alignments reveal that the encoded CYP4F15 and CYP4F16 proteins have 66 to 80% sequence similarities to four known human CYP4Fs, rat CYP4F1, and rat CYP4F6. Furthermore, CYP4F15 has 93.5% identity to rat CYP4F4 and CYP4F16 has 90.8% identity to rat CYP4F5, suggesting that CYP4F15 and CYP4F16 are the orthologs for CYP4F4 and CYP4F5, respectively (Table 1).
Expression of CYP4F15 and CYP4F16 in Mouse Tissues.
Because we are dealing with the isoforms from a protein subfamily with relatively high homology, we first prepared isoform-specific probes and tested them in a slot blot analysis (Fig. 3). Under stringent conditions (0.1× SSC, 0.1% SDS at 65°C), probes for CYP4F15 and CYP4F16 predominantly hybridized to CYP4F15 and CYP4F16 cDNA, respectively, and with a 30-fold lower cross-hybridization signal to other mouse 4F isoforms as calculated by PhosphoImager analysis. The same stringent conditions were used for the following Northern blots.
Northern blotting analysis was performed on poly(A)+ RNA from eight mouse tissues, using CYP4F15- and CYP4F16-specific probes (Fig.4). The data indicated that CYP4F15 and CYP4F16 are expressed as a major transcript of 2.8 kb in multiple tissues and an additional 2.4-kb transcript only seen in kidney. As expected, both CYP4F15 and CYP4F16 are highly expressed in liver. CYP4F15 is also expressed in brain, kidney, and lung, although with much lower expression levels. Interestingly, brain has a much stronger signal compared with that of kidney, which has not been seen for other cytochrome P450s with the exception of aromatase (CYP19). No signal was detected in heart, skeleton muscle, spleen, or testis. CYP4F16 has a somewhat broader tissue distribution than that of CYP4F15, with transcripts detected in liver, kidney, and heart as well as brain, spleen, and lung.
Regulation of Renal CYP4F15 and CYP4F16 Expression by LPS and Clofibrate in PPARα (+/+) and (−/−) Mice.
Our previous data showed that rat 4F isoforms were down-regulated by clofibrate and LPS (X. Cui, A. M. Robida, E. T. Morgan, and H. W. Strobel, unpublished data) in liver and remained unchanged in kidney. To test whether the mouse isoforms are regulated in the same manner, wild-type mice were given a single injection of LPS or clofibrate. Expressions of CYP4F15 and CYP4F16 at the mRNA level were analyzed by Northern blot. The availability of PPARα knockout mice allowed us to investigate further other putative mechanisms of regulation and to compare 4F regulation with that of the CYP4A isoforms. The PPARα null mice were given the same treatments, and the expression levels were compared with the respective wild-type control.
In kidney, wild-type expression of CYP4F15 gene was low as shown in Fig. 4 and knockout of PPARα did not change the constitutive expression level (Fig. 5, A and B). Unlike its ortholog CYP4F4, CYP4F15 expression in wild-type mice was significantly induced up to 3-fold (p < 0.005) and 6.5-fold (p < 0.005) in kidney by treatments with LPS and clofibrate, respectively. However, administration of LPS and clofibrate to PPARα (−/−) mice resulted in no change in renal CYP4F15 mRNA level compared with PPARα (−/−) or wild-type untreated group. Absence of a regulatory effect of LPS and clofibrate on CYP4F15 expression in PPARα (−/−) mice suggested that PPARα is a major determinant in renal CYP4F15 regulation. The expression of CYP4F16 is different from that of CYP4F15. Clofibrate treatment caused no significant change of CYP4F16 expression in either PPARα (+/+) or (−/−) mice. LPS treatment led to a slight increase of CYP4F16 expression, but due to the variation within the group, the increase is not significant statistically (p = 0.1).
Regulation of Hepatic CYP4F15 and CYP4F16 Expression by LPS and Clofibrate in PPARα (+/+) and (−/−) Mice.
In liver, ablation of PPARα did not change the constitutive expression level of either CYP4F15 or CYP4F16 (Fig. 6). In contrast to the induction of CYP4F15 expression by LPS administration in PPARα (+/+) kidney, LPS treatment caused a striking decrease of CYP4F15 expression to 11% (p < 0.005) of the untreated wild-type level in liver. CYP4F16 mRNA was reduced to the comparable level (9%,p < 0.005) in the same group. It is noteworthy that hepatic CYP4F15 and CYP4F16 mRNA levels for PPARα null mice were also reduced, although to a lesser extent (52% for CYP4F15,p < 0.05; 78% for CYP4F16, p < 0.05). This result seems consistent with the notion that PPARα is not solely responsible for the reduction of hepatic CYP4F15 and CYP4F16 by LPS. There was no change of CYP4F15 level observed when wild-type mice were treated with clofibrate. However, the same clofibrate treatment led to a decreased mRNA level in null mice. For CYP4F16, clofibrate caused a slight repression in both wild-type and null mice in a PPARα-independent manner.
To study the regulation of CYP4F gene expression by LPS and clofibrate as well as the involvement of PPARα in wild-type and PPARα knockout mice, mouse CYP4F isoforms must be first identified. Five new mouse CYP4F cDNAs have been identified from mouse brain and kidney; two of them are used in our gene regulation study, namely, CYP4F15 and CYP4F16. Both clones contain full-length open reading frames with the conserved family 4 consensus sequence and heme-binding region. Two in-frame initiation methionines are present in CYP4F15 cDNA, separated by four amino acid residues. However, the second initiation methionine of CYP4F15 showed higher homology with other CYP4Fs with similar length and hydrophobicity of the N-terminal inner membrane domain. Therefore, we think that the second ATG is the real initiation codon used for protein expression. CYP4F15 encodes a protein with 529 amino acid residues with a predicted molecular weight of 60,628, whereas CYP4F16 encodes 524 amino acid residues with an estimated molecular weight of 60,229. The in vitro translated proteins showed molecular masses of 60 kDa, values that are consistent with the calculated estimates. CYP4F15 and CYP4F16 are determined to be the orthologs of rat CYP4F4 and CYP4F5 because their amino acid sequence homologies are higher than 90%. Therefore, we used CYP4F15 and CYP4F16 in the subsequent gene regulation study to compare with our regulation study of rat CYP4Fs by LPS.
Both CYP4F15 and CYP4F16 mRNA are highly expressed in liver, with some expression in extrahepatic tissues. CYP4F16 showed a broader tissue distribution than that of CYP4F15. Interestingly, CYP4F15 has a higher expression level in brain than in kidney. It is also possible that CYP4F15 might be concentrated in certain regions of the brain. Whether that higher expression is related to certain brain physiological functions is unknown at the present time.
A major finding of this work is that CYP4F15 mRNA expression is highly inducible by clofibrate treatment in kidney, and this induction seems totally dependent on PPARα. To our knowledge this is the first time that a CYP4F isoform has been found to show CYP4A-like induction by clofibrate. The requirement of functional PPARα to activate CYP4A gene transcription in response to clofibrate is well characterized and is due to a functional PPRE in the CYP4A promoter (Muerhoff et al., 1992). It is very likely that CYP4F15 induction by clofibrate is regulated through the same mechanism, i.e., clofibrate acts as a ligand for PPARα and activated CYP4F15 gene transcription due to the presence of PPRE. Interestingly, the induction of CYP4F15 expression can be mimicked by LPS treatment in kidney; and this is also PPARα-dependent. The same phenomenon was observed for CYP4A10 and CYP4A14 expression in clofibrate- and LPS-treated mouse kidney (Barclay et al., 1999). Apparently, PPARα is activated in mouse kidney during the response to LPS. Consistent with this observation, another PPARα-responsive gene, acetyl-CoA oxidase, is also induced by LPS treatment (Barclay et al., 1999). It is possible that one or more endogenous compounds are generated during the LPS-induced inflammatory condition and function as agonists for PPARα. The candidate endogenous molecules that could activate PPARα are the lipid inflammatory mediators such as arachidonic acid, prostaglandins, leukotrienes, and hydroxyeicosatetraenoic acids, which have been shown to be potent activators of PPARs (Kliewer et al., 1995; Yu et al., 1995; Devchand et al., 1996).
In contrast to the CYP4F15 mRNA induction in kidney, LPS and clofibrate treatments had no significant effects on CYP4F16 mRNA expression in either wild-type or PPARα knockout mouse kidneys. There is no doubt that PPARα is activated in kidney by these treatments; therefore, this unresponsiveness is likely due to the differences between CYP4F15 and CYP4F16 promoter sequences. Previous studies have shown that fibrate administration can cause both positive and negative regulation in related genes. For instance, fibrates lower high-density lipoprotein concentration in rats by decreasing hepatic apolipoprotein A-I expression, but it will increase the apolipoprotein A-I expression in humans (Vu-Dac et al., 1998). The molecular mechanism underlying this phenomenon seems to be the sequence differences in promoter regions. On one hand, the positive regulation sequence PPRE is present in human Apo A-I promoter, whereas the functional negative element RevRE is present in rat Apo A-I promoter (Vu-Dac et al., 1998). Whether the isoform-specific regulation of CYP4Fs in mouse kidney reflects the same mechanism needs further characterization of the regulatory elements in their respective promoters and of the trans-activating factors that control gene transcriptions.
The regulation of CYP4F15 and CYP4F16 mRNA in liver by clofibrate and LPS revealed a more complicated picture than that seen in kidney. In contrast to kidney, hepatic CYP4F15 and CYP4F16 mRNA expression is almost abolished in wild-type mice treated with LPS compared with control. Similar results were observed for CYP4A10 and CYP4A14 (Barclay et al., 1999). This tissue specificity may due to the activation of Kupffer cells, which may lead to high local cytokine concentrations around the hepatocytes (Fox et al., 1996). Studies have shown that hepatic expression of CYP2E1, CYP2C11, and CYP3A2 are repressed by LPS (Sewer et al., 1996). We also observed that expression of CYP4F4 and CYP4F5 are reduced in LPS-treated rat liver (X. Cui, A. M. Robida, E. T. Morgan, and H. W. Strobel, unpublished data). In PPARα null mice, down-regulations of CYP4F15 by LPS were less than those of LPS-treated controls with 50 to 80% of control expression levels remaining after treatments. These results suggest that PPARα is involved in this down-regulation, but it is not solely responsible for the reduction of mRNA expression. Despite their different responses to LPS in kidney, mRNA of CYP4F15 and CYP4F16 showed similar down-regulation by LPS in liver. In this circumstances, perhaps PPARα does not function in a manner that requires direct binding to DNA elements in CYP4F15 and 4F16 5′ flanking regions. It may be that PPARα exerts its function by interacting with protein factors in other signaling pathways, and contributes to the CYP4F15 and CYP4F16 down-regulation indirectly. Recent research has revealed cross talk between PPARs and other signaling pathways. PPARs can interfere with other nuclear receptors such as retinoic acid receptor and glucocorticoid receptor by competing for the limited amount of the partners or coactivators (Kamei et al., 1996; Dowell et al., 1997). PPARs also interact with NF-κB, signal transducer and activator of transcription, and activator protein-1 pathways in a DNA-binding independent manner (Jiang et al., 1998; Ricote et al., 1998; Staels et al., 1998). Many studies favor the notion that PPARs have anti-inflammation properties. It has been shown that PPARα activation inhibits IL-1-induced IL-6, prostaglandins, and cyclooxygenase-2 production in smooth muscle cell by interfering with NF-κB, activator protein-1, and signal transducer and activator of transcription pathways (Staels et al., 1998). PPARγ agonists can inhibit the production of IL-1, IL-6, and TNFα in monocytes (Jiang et al., 1998). More controversial data have been also presented. In liver, peroxisome proliferators can increase the TNFα production (Bojes et al., 1997) and activate NF-κB activity (Li et al., 1996; Espandiari et al., 1998; Rusyn et al., 1998). Our data favor the latter findings in that PPARα mediates decreased CYP4F15 and CYP4F16 mRNA expression.
Clofibrate did not cause a change in CYP4F15 expression in wild-type mouse liver. However, the decrease of CYP4F15 expression by clofibrate treatment in PPARα knockout mouse livers suggested that the induction effect of clofibrate via PPARα is obscured by the negative effect on CYP4F15 expression mediated through another pathway(s). Vu-Dac et al. (1998) showed that fibrates could dramatically increase the level of liver Rev-erbα, a member of an orphan nuclear receptor family that functions as negative regulator for gene transcription. It is clear that the decrease of apolipoprotein A-I expression by fibrate treatment is due to the activation of Rev-erbα, which binds to RevRE on the promoter region to suppress gene expression. It remains to be determined whether Rev-erbα functions in clofibrate-mediated down-regulation of CYP4Fs in both rat and mouse. Another possible explanation for clofibrate-initiated down-regulation is through the activation of Kupffer cells. This hypothesis is supported by the finding that peroxisome proliferators can activate Kupffer cells, leading to rapid NF-κB activation and subsequent induction of TNFα production (Rusyn et al., 1998). That CYP4F16 expression is reduced in both wild-type and knockout mice implies that PPARα is not involved in the regulation of CYP4F16 mRNA by clofibrate. This is consistent with our findings in kidney. The reduction of expression of CYP4F16 mRNA further illustrates the existence of a PPAR-independent negative regulation pathway(s) by peroxisome proliferators.
In summary, we have shown that PPARα is involved in mouse 4F regulation both in liver and kidney under inflammatory conditions. CYP4F15 is a novel 4F isoform, which like 4As can be up-regulated through PPARα activation. CYP4F16 behaves similarly to the other rat 4Fs. Together, these results indicate that expression of CYP4Fs is regulated in a species-, tissue-, and isoform-specific manner.
We thank Laura Bankey for excellent computer expertise.
- Received June 29, 2000.
- Accepted August 28, 2000.
Send reprint requests to: Dr. Henry W. Strobel, Department of Biochemistry and Molecular Biology, University of Texas Medical School at Houston, P.O. Box 20708, Houston, TX 77030. E-mail:
This work is supported by Grants MH58297 and GM46897 (to E.T.M.) from the National Institute of Mental Health, Department of Health and Human Services, and U.S. Army Grant DAMD-17-98-1-8002 “Disaster Relief and Emergency Medical Service” project. The data presented here form part of the dissertation of Xiaoming Cui submitted to the faculty of The University of Texas Health Science Center at Houston, Graduate School of Biomedical Sciences in partial fulfillment of the requirements for the Doctor of Philosophy degree.
- CYP and P450
- cytochrome P450
- leukotriene B4
- peroxisome proliferator activated-receptor α
- peroxisome proliferator responsive element
- standard saline citrate
- base pair
- nuclear factor-κB
- tumor necrosis factor
- U.S. Government