Polyunsaturated Fatty Acid Suppression of Hepatic Fatty Acid Synthase and S14 Gene Expression Does Not Require Peroxisome Proliferator-activated Receptor α*

Dietary polyunsaturated fatty acids (PUFA) induce hepatic peroxisomal and microsomal fatty acid oxidation and suppress lipogenic gene expression. The peroxisome proliferator-activated receptor α (PPARα) has been implicated as a mediator of fatty acid effects on gene transcription. This report uses the PPARα-deficient mouse to examine the role of PPARα in the PUFA regulation of mRNAs encoding hepatic lipogenic (fatty acid synthase (FAS) and the S14 protein (S14)), microsomal (cytochrome P450 4A2 (CYP4A2)), and peroxisomal (acyl-CoA oxidase (AOX)) enzymes. PUFA ingestion induced mRNAAOX (2.3-fold) and mRNACYP4A2(8-fold) and suppressed mRNAFAS and mRNAS14 by ≥80% in wild type mice. In PPARα-deficient mice, PUFA did not induce mRNAAOX or mRNACYP4A2, indicating a requirement for PPARα in the PUFA-mediated induction of these enzymes. However, PUFA still suppressed mRNAFAS and mRNAS14 in the PPARα-deficient mice. Studies in rats provided additional support for the differential regulation of lipogenic and peroxisomal enzymes by PUFA. These studies provide evidence for two distinct pathways for PUFA control of hepatic lipid metabolism. One requires PPARα and is involved in regulating peroxisomal and microsomal enzymes. The other pathway does not require PPARα and is involved in the PUFA-mediated suppression of lipogenic gene expression.

Peroxisome proliferators encompass a wide variety of compounds including hypolipidemic drugs (WY14,643, gemfibrozil, and clofibrate), plasticizers (di-(2-ethyhexyl)phthalate), steroids (dehyrdoepiandrosterone and dehyrdoepiandrosteronesulfate), and dietary fatty acids (25)(26)(27). Collectively, peroxisome proliferator-induced changes in gene expression are mediated by activating a nuclear receptor, the peroxisome proliferator-activated receptor (PPAR). PPARs are members of the steroid/thyroid superfamily of nuclear receptors (20 -28). Activation of PPAR-mediated transcription is achieved through PPAR-RXR heterodimers which bind DNA motifs called peroxisome proliferator response elements (PPRE) located in promoters of target genes (20 -33). PPARs also have inhibitory effects on gene transcription. For example, apolipoprotein CIII and transferrin gene expression is inhibited by PPAR-RXR competition for an HNF-4 binding site within the promoters of these genes (34,35). PPAR␣ has also been shown to interfere with thyroid hormone action by sequestering RXR␣, a factor required for thyroid hormone receptor binding to DNA (36 -38).
The fact that PPARs are activated by fatty acids in conjunction with the known effects of PUFA on peroxisomal and microsomal fatty acid oxidation suggest that PUFA regulation of these pathways might utilize a common transcriptional mediator, i.e. PPAR. While several PPAR subtypes (␣, ␤ (also known as ␦, Nuc1, FAAR), ␥1, and ␥2) have been identified in rodents, PPAR␣ is the predominant form in rodent liver (22-24, 28, 31, 39). Recent gene targeting studies clearly demonstrate that PPAR␣ is required for the pleiotropic response to peroxisome proliferators including an increase in hepatic mRNAs encoding peroxisomal and microsomal enzymes (40).
This report examines the role of PPAR␣ in PUFA-regulation of mRNAs encoding hepatic lipogenic, microsomal and peroxisomal enzymes. In this work, we assessed PUFA regulation of the S14 gene and fatty acid synthase, models for lipogenic gene expression, and acyl-CoA oxidase (AOX) and cytochrome P450A2 (CYP4A2), enzymes involved in peroxisomal and microsomal fatty acid oxidation, respectively. The recently developed PPAR␣-null mouse (40) was used to determine whether PPAR␣ mediates PUFA regulation of hepatic AOX, CYP4A2, S14, and FAS gene expression. This work shows that while PPAR␣ is required for the PUFA-mediated induction of both AOX and CYP4A2 gene expression, it is not required for the PUFA-mediated inhibition of either S14 or FAS gene expression. These and other studies indicate that PUFA regulation of hepatic gene transcription involves at least two distinct pathways, a PPAR␣-dependent and a PPAR␣-independent pathway.
The cDNA for CYP4A2 was cloned by differential display screening (43) of rat liver RNA. First strand cDNA was prepared (Delta RNA Fingerprinting Kit, K1801-1, CLONTECH) from rat liver RNA of rats fed olive oil and fish oil for 5 days. PCR was performed using a Perkin-Elmer DNA thermal cycler 9600 as follows: one cycle of 94°C, 10 s; 40°C, 5 min; 68°C, 5 min; two cycles of 94°C, 5 s; 40°C, 5 min, 68°C, 5 min; 25 cycles of 94°C 5 s; 60°C, 1 min; and 68°C, 2 min for 30 cycles of 95°C/min; 60°C/min; 72°C/min. 35 S-Labeled PCR products were separated by denaturing polyacrylamide gel electrophoresis and visualized by autoradiography. A band present in the fish oil derived cDNA and diminished in the olive oil derived cDNA was excised from the gel, reamplified by PCR, and cloned into pGEM-T (Promega). The insert was sequenced by dideoxynucleotide sequencing (44). Computer assisted analysis (GCG, University of Wisconsin) of the cloned sequence with the GenBank TM data base indicated that the cloned sequence was identical to a region of the rat cytochrome P450 4A2 sequence (accession no. 57719) at nucleotides 12624 -12728 for the rat gene. This sequence was in the noncoding 3Ј-untranslated region. To obtain the sequence unique to CYP4A2 and containing the translated sequence, two PCR primers (5Ј-primer, 5Ј-atatatgaattc 7370 CACATGGGACCACCTGG; 3Јprimer, 5Ј-atatatatcgat 12650 GCTGGGAAGGTG TCTGGAGT (underlined sequence represents EcoRI and ClaI cloning sites, respectively; the superscripted numbers represent the position of the nucleotide in the rat CYP4A2 gene)) were used to generate a cDNA from liver RNA of fish oil-fed rats by reverse transcriptase-PCR. The identity of the clone (pCYP4A2-2) was verified by DNA sequencing and used to measure CYP4A2 mRNA. mRNA levels were measured by dot and Northern blot analyses (17) and the level of hybridization was quantified using a Molecular Dynamics PhosphorImager (Sunnyvale, CA) or by videodensitometry using an Agfa-2 scanner linked to a Macintosh computer with NIH Image software.
Plasmid Construction and Primary Hepatocytes-The construction of the reporter gene with the rat AOX PPRE fused upstream from the thymidine kinase promoter (TKCAT223) was described previously (38).
Primary hepatocytes were obtained from rat liver by the collagenase perfusion method and transfected with specific DNAs in the presence of Lipofectin (16). Hepatocytes were treated with triiodothyronine along with specific fatty acids or peroxisome proliferators (WY14,643 or gemfibrozil dissolved in Me 2 SO) (16,38). After 48 h of treatment, hepatocytes were analyzed for protein and CAT activity (38). CAT activity is defined as CAT units ϭ counts/min of 14 C-butylated chloramphenicol/ h/100 g of protein.
Statistical Analysis-All data are presented as the mean Ϯ S.E. Statistical comparisons were made by a single-factor factorial analysis of variance using Microsoft Excel version 7.

The Effects of Olive and Fish Oil on Hepatic Gene Expression in Wild-type (ϩ/ϩ) and PPAR␣ Null (Ϫ/Ϫ) Mice-PPAR␣
is the predominant PPAR subtype in rodent liver and has a central role in regulating the transcription of genes encoding hepatic peroxisomal and microsomal enzymes (24,28,40). To determine whether PPAR␣ mediates PUFA regulation of hepatic gene expression wild type (ϩ/ϩ) and PPAR␣ null (Ϫ/Ϫ), mice were fed an olive oil or fish oil diet for 5 days. Northern analyses show that feeding (ϩ/ϩ) mice fish oil for 5 days resulted in a ϳ2-fold (p Ͻ 0.003) and ϳ9-fold (p Ͻ 0.001) increase in hepatic mRNA AOX and mRNA CYP4A2 , respectively (Fig. 1, A  and B). In contrast, fish oil did not significantly induce mRNA AOX and mRNA CYP4A2 in the PPAR␣ null (Ϫ/Ϫ) mice. These results indicate that PPAR␣ is required for the PUFAmediated induction of AOX and CYP4A2 mRNAs. While hepatic ␤-actin mRNA was elevated in the (Ϫ/Ϫ) mice when compared with the (ϩ/ϩ) mice, it was not affected by dietary manipulation.
Analysis of mRNAs encoding S14 and FAS shows both mRNAs were suppressed (Ն70%) in both the wild type (ϩ/ϩ) and PPAR␣ null (Ϫ/Ϫ) mice following fish oil feeding. Since FIG. 1. The effects of olive and fish oil feeding on levels of mRNA AOX , mRNA CYP4A2 , mRNA S14 , and mRNA FAS from both wild-type (؉/؉) and PPAR␣ null (؊/؊) mice. Eight mice of each genotype were meal-fed with diets supplemented with 10% olive oil for 10 days. Four of each genotype were switched to a diet supplemented with 10% fish oil for 5 days. Total RNA was prepared from mouse livers and measured for mRNAs encoding AOX, CYP4A2, S14, FAS, and ␤-actin by Northern analysis. Ethidium bromide staining of the RNA on a typical agarose gel indicates that the RNA was evenly loaded. The results of the hybridization are displayed (A) and quantified (B). A, The AOX, CYP4A2, S14, and FAS and ␤-actin blots were exposed to x-ray film for 36, 18, 24, 24, and 18 h, respectively. B, the level of hybridization was quantified by a PhosphorImager and normalized against the level of expression of the corresponding mRNAs in olive oil-fed mouse liver; n ϭ 4. The statistical comparison (analysis of variance, ANOVA) was for olive oil versus fish oil-fed mice. N.S., not significant.
PUFA rapidly inhibits the transcription of both the S14 and the FAS gene (16,17), these observations indicate that PPAR␣ is not required for the PUFA-mediated suppression of transcription of these genes.
The Effects of Olive Oil and Fish Oil on Hepatic AOX and S14 Gene Expression-Previous studies have shown that S14 and FAS are regulated by PUFA and peroxisomal proliferators in rat liver or primary hepatocytes (17,45). The studies described below compare the PUFA and peroxisome proliferator regulation of S14 and AOX in rat liver and primary hepatocytes. These in vivo and in vitro (rat primary hepatocytes) studies were performed to gain additional support for the idea that PUFA regulation of lipogenic gene expression is a result of a different pathway than PUFA regulation of AOX gene expression.
Time Course of Fish Oil and Gemfibrozil Effects on Hepatic S14 and AOX Gene Expression in Vivo-The rapidity of fish oil action on hepatic mRNA AOX and mRNA S14 was examined in rats fed fish oil for 1 to 5 days (Fig. 3A). Rats were meal-fed a high carbohydrate diet supplemented with10% triolein oil for 10 days. Subsequently, half of the rats were maintained on this diet and half were switched to a high carbohydrate diet supplemented with 10% fish oil. Fish oil feeding induced a rapid suppression of mRNA S14 ; a 60% suppression was observed within 1 day of switching the diet from triolein to fish oil. Similar effects on mRNAs encoding FAS and L-pyruvate kinase Total hepatic RNA was prepared and examined by dot-blot analysis for the effect of feeding on mRNA AOX and mRNA S14 levels. The results were quantified and normalized against the level of hepatic mRNA expressed in chow-fed rats, i.e. 1 unit. Analysis of variance for both mRNA AOX and mRNA S14 levels for menhaden oil-, 20:5-, and 22:6-fed versus olive oil-fed, p Ͻ 0.002. These results are representative of two separate studies; n ϭ 4.
FIG. 3. Time course of fish oil and gemfibrozil effects on rat hepatic S14 and AOX gene expression. A, rats were meal-fed diets supplemented with 10% (w/w) triolein oil for 10 days. Half of the rats were maintained on triolein oil, while the other half were switched to the fish oil diet. Both triolein-and fish oil-fed rats were killed after 1, 2, and 5 days. Total liver RNA was prepared and examined by dot-blot analysis for mRNA AOX and mRNA S14 levels; n ϭ 4 -10 per time point. These results are representative of two separate studies. The results were quantified and normalized against the level of hepatic mRNA expressed in chow-fed rats, i.e. 1 unit. Analysis of variance: a, 1-, 2-, and 5-day fish oil feeding on mRNA S14 versus triolein oil fed mRNA S14 , p Ͻ 0.001; b, 2-and 5-day fish oil feeding on mRNA AOX versus triolein oil fed mRNA AOX , p Ͻ 0.001. B, after meal-training rats to a high carbohydrate fat-free diet, half of the rats were maintained on the high carbohydrate, fat-free diet, while the other half was switched to a diet supplemented with 0.2% (w/w) gemfibrozil for 2, 4, and 8 days. Dot-blot analysis was used to measure the mRNA AOX and mRNA S14 levels. The results were quantified and normalized against the level of hepatic mRNA expressed in chow-fed rats, i.e. 1 unit; n ϭ 3-4 per time point. These results are representative of two separate studies. Analysis of variance: a, 2, 4, and 8 days, gemfibrozil versus control diet fed, mRNA AOX , p Ͻ 0.004; b, 4-day gemfibrozil fed versus control diet fed, mRNA S14 , p ϭ 0.002. have were reported previously (17). In contrast, mRNA AOX remained unaffected after 1 day on the fish oil diet, yet was induced 2-and 3.5-fold after 2 and 5 days, respectively. Such studies indicate that changes in S14 mRNA precede changes in AOX mRNA following initiation of fish oil feeding, but they do not argue against PPAR␣ as a common mediator for the PUFAregulation of AOX and S14.
In an effort to separate the induction of AOX from the suppression of S14, the peroxisome proliferator, gemfibrozil was fed to rats at 0.2% (w/w) for up to 8 days (Fig. 3B). mRNA AOX was induced ϳ4-fold after 4 days on gemfibrozil, a level comparable to the level of mRNA AOX after 5 days on fish oil. In contrast, gemfibrozil did not significantly suppress mRNA S14 (Fig. 3B) or mRNA FAS (not shown). Only a modest 22% inhibition of mRNA S14 was seen after 4 days of gemfibrozil feeding. These results show that mRNAs encoding both S14 and AOX are affected by PUFA within 2 days of initiating fish oil feeding. However, the absence of a significant inhibition of mRNA S14 following 8 days of gemfibrozil feeding argues against PPAR␣ as a common mediator for PUFA regulation of both AOX and S14 gene expression.
Effect of Fatty Acids on AOX and S14 Gene Expression in Primary Hepatocytes-Primary hepatocytes provide a method to assess the direct effects of PUFA on hepatic gene expression (16,17). To examine the effects of fatty acids on S14 and AOX mRNAs, primary rat hepatocytes were treated with albumin alone or albumin plus various fatty acids (Fig. 4). Treatment of primary hepatocytes with 18:1, 18:2, 18:3 (both n-3 and n-6), and 20:4 did not induce mRNA AOX . Only 20:5 treatment induced mRNA AOX (ϳ2-fold). This finding is consistent with the effects of highly unsaturated fatty acids on AOX gene expression in vivo (Figs. 1 and 2). Oleic acid (18:1) did not affect S14 gene expression when compared with controls. However, 18:2 (n-6) resulted in ϳ50% decline and 18:3 (both n-3 and n-6), 20:4 (n-6) and 20:5 (n-3) treatment resulted in Ͼ80% suppression of hepatocyte mRNA S14 levels. These findings demonstrate that a broader spectrum of fatty acids affect S14 gene expression than AOX gene expression in primary rat hepatocytes.
Fatty Acid Effects on Reporter Gene Activity in Primary Hepatocytes-To determine if fatty acids activate PPAR␣ in liver, primary hepatocytes were transfected with a reporter gene containing the AOX PPRE fused to the thymidine kinase promoter, i.e. TKCAT223 (38). Primary hepatocytes were cotransfected with pSG5 (empty vector) or pSG5-mPPAR␣, a PPAR␣ expression vector (Fig. 5). In the absence of co-transfected PPAR␣, TKCAT223 CAT activity was expressed at low levels (Ͻ150 CAT units) and this activity was marginally affected by fatty acid or peroxisome proliferator (WY14,643, gemfibrozil) treatment. Co-transfection with pSG5-PPAR␣ led to at least a 10-fold stimulation of the TKCAT223 activity. Treatment of PPAR␣-transfected hepatocytes with 18:1 and 20:4 had no effect on CAT activity, while 20:5 treatment induced CAT activity by ϳ2-fold. This pattern of fatty acid regulation of PPAR␣ is consistent with the effects of these fatty acids on mRNA AOX (Fig. 4). By comparison, both WY14,643 and gemfibrozil induced CAT activity by 4-fold. These studies show that long chain unsaturated fatty acids such as 18:1, 18:2, 18:3 (n-3 and n-6), and 20:4 do not activate PPAR␣ in primary hepatocytes. Only the highly unsaturated fatty acid, 20:5, n-3 activates PPAR␣, albeit to a level less than WY14,643 or gemfibrozil. This pattern of control contrasts with the known effects of fatty acids on mRNA S14 (16,17) (Fig. 4) and S14CAT activity 3 (17). Total RNA was prepared from the hepatocytes and analyzed by dot blot analysis for mRNA coding for AOX and S14. The results were quantified and normalized against the level of AOX or S14 mRNA expressed in hepatocytes receiving no fatty acid treatment (albumin control). Analysis of variance: a, 18:1 (n-9)-, 18:2 (n-6)-, 18:3 (n-3)-, 18:3 (n-6)-, or 20:4 (n-6)-treated mRNA AOX versus albumin-treated mRNA AOX , p Ն 0.07; b, 20:5 (n-3)-treated mRNA AOX versus albumin-treated mRNA AOX , p ϭ 0.006; c, 18:1-treated mRNA S14 versus albumin-treated mRNA S14 , p ϭ 0.01; d, 18:2 (n-6)-, 18:3 (n-3)-, 18:3 (n-6)-, 20:4 (n-6)-, or 20:5 (n-3)treated mRNA S14 versus albumin-treated mRNA S14 , p Յ 0.0004. These results are representative of two separate studies; n ϭ 4. peroxisomal and microsomal fatty acid oxidation (24,40). Since several peroxisomal, microsomal and lipogenic enzymes are regulated by PUFA at the pretranslational level, we tested the hypothesis that dietary PUFA regulate hepatic fatty acid oxidation and de novo lipogenesis through a common mediator, i.e. PPAR␣. Interestingly, all studies reporting on fatty acid regulation of PPAR␣ have been carried out by over expressing receptors in established cell lines. No studies have directly examined the role PPAR␣ may have in fatty acid-regulated hepatic gene transcription. The PPAR␣-null mouse allows such an analysis. Coupling this genetic approach with other studies has allowed us to show for the first time that 1) PPAR␣ is required for PUFA-mediated induction of hepatic mRNA AOX and mRNA CYP4A2 (Fig. 1); 2) PPAR␣ is not required for PUFAmediated suppression of mRNA S14 or mRNA FAS (Fig. 1); 3) while 18:2 (n-6), 18:3 (n-6 and n-3), 20:4 (n-6) and 20:5 (n-3) suppress mRNA S14 and mRNA FAS , only 20:5 (n-3) induces mRNA AOX in primary hepatocytes (Fig. 4); 4) while gemfibrozil induces hepatic mRNA AOX , it has little or no effect on mRNA S14 or mRNA FAS (Fig. 3). Taken together, these studies indicate that PUFA control of peroxisome/microsomal fatty acid oxidation and de novo lipogenesis in rat liver does not involved PPAR␣ as a common mediator. The differential effect of specific fatty acids, i.e. 18:2, 18:3 (n-3 and n-6), 20:4 (n-6), versus gemfibrozil underscores the lack of coordinate regulation of these pathways in rat liver. Such studies indicate that PUFA regulates at least two pathways in liver, one involves PPAR␣ and controls expression of genes encoding proteins involved in peroxisomal and microsomal fatty acid oxidation. The other mechanism is PPAR␣-independent and is involved in the PUFA-mediated suppression of lipogenic gene expression.
PUFA suppress hepatic mRNA S14 and mRNA FAS levels by inhibiting gene transcription (15)(16)(17). From the data reported above, this inhibitory mechanism does not require PPAR␣. Although the mechanism of PUFA induction of hepatic mRNA AOX and mRNA CYP4A2 has not been established, the following studies implicate transcription as the principal mode of PUFA regulation of AOX and CYP4A: 1) peroxisomal proliferators rapidly induce transcription of genes encoding AOX, the bifunctional enzyme, thiolase and CYP4A subtypes 1-3 (26,40); 2) PPAR␣ is required for the induction of these genes (40); 3) PPAR␣ binds PPREs as PPAR/RXR heterodimers in the promoters of these genes and stimulates transcription of cislinked reporter genes (20 -33, 38); 4) fatty acids activate PPAR␣ and stimulate transcription of cis-linked reporter genes (20 -33) (Fig. 5); and 5) PPAR␣ is required for the PUFA induction of hepatic mRNA AOX and mRNA CYP4A2 (Fig. 1).
Previous efforts to examine the involvement of PPAR␣ in PUFA regulation of lipogenic gene expression showed that the cis-regulatory targets for PUFA and PPAR in the S14 promoter (38, 45) did not converge. Analysis of stearoyl-CoA desaturase 1 gene expression indicated that peroxisome proliferators/ PPAR induced by PUFA-suppressed transcription (46). Such studies argued against PPAR as a mediator of PUFA effects on lipogenic gene transcription. However, the over expression of receptors does not necessarily reflect physiologically relevant processes. The use of the PPAR␣-null mouse allows us to directly evaluate the role PPAR␣ plays in PUFA regulation of hepatic gene expression. In contrast to (ϩ/ϩ) mice, hepatic mRNA AOX and mRNA CYP4A2 was not significantly induced in PPAR␣ (Ϫ/Ϫ) mice by the PUFA diet indicating a requirement for PPAR␣ in the PUFA-mediated induction of these enzymes. The fact that hepatic mRNA S14 and mRNA FAS was suppressed in both (ϩ/ϩ) and (Ϫ/Ϫ) mice provides strong evidence against a requirement for PPAR␣ for PUFA-mediated suppression of S14 and FAS gene transcription. While these studies confirm our earlier suggestion that PPAR did not mediate PUFA suppression of S14 gene transcription, they provide new information on the requirement for PPAR␣ in the PUFA-induction of AOX and CYP4A2 and the lack of involvement of PPAR␣ in PUFA-mediated suppression of FAS gene transcription or Lpyruvate kinase gene expression. 2 While other PPAR subtypes (PPAR␥ and PPAR␦) (24,39) are expressed in liver, Northern analyses suggests PPAR␥ and ␦ are minor subtypes in rodent liver. However, their role in PUFA control of hepatic gene expression cannot be excluded.
An important outcome of these studies is the finding that of all the PUFA tested, only 20:5 (n-3) activates PPAR␣ in liver. Several groups have reported on fatty acid activation of PPARs in established cells lines like CV-1 and HeLa (20 -24). Recently three groups reported that specific fatty acids, i.e. 18:2 (n-6), 18:3 (n-3 and n-6), and 20:4 (n-6) are ligands for PPAR␣ (47)(48)(49). These same ligands do not activate PPAR␣ or induce mRNA AOX in primary hepatocytes (Figs. 4 and 5). Feeding animals soybean or corn oil, oils containing 18:2 and 18:3 fatty acids, does not induce peroxisomal enzymes (4). This apparent conflict can be reconciled by the fact that primary hepatocytes have a high capacity for fatty acid oxidation, triglyceride synthesis and very low density lipoprotein secretion (7). We speculate that these pathways prevent intracellular fatty acids from accumulating to levels that activate PPAR␣. Interestingly, 20:5 (n-3) was the only PUFA tested here that activated PPAR␣. 20:5 (n-3) is reported to be poorly oxidized in mitochondria and poorly incorporated into complex lipids, such as triglycerides (7). Thus, 20:5 (n-3) might accumulate in the cell and mimic a state of fatty acid overload in the liver. Fatty acid overload resulting from high fat feeding (Ͼ50% calories as fat), uncompensated diabetes and liver disease have all been reported to increase peroxisomal ␤-oxidation (26,50,51). Alternatively, 20:5 (n-3) might be metabolized to an active ligand. Recent studies have suggested that the leukotriene, LTB4, is a ligand for PPAR␣ (52). Indeed, LTB4 is derived from 20:4 (n-6) by the action of 5-lipoxygenase and LTA4 hydrolase. If this pathway were operative, we would expect 20:4 (n-6) treatment of hepatocytes to activate PPAR␣ and induce mRNA AOX . The lack of a 20:4 (n-6) effect on mRNA AOX and PPAR␣ along with the low LTA4 hydrolase activity associated with liver cells (53) suggest that LTB4 is not the operative ligand for 20:5 (n-3) activation of PPAR␣. However, the in vitro model used in this work may lack factors present in the in vivo system.