Saturated fatty acids, but not unsaturated fatty acids, induce the expression of cyclooxygenase-2 mediated through Toll-like receptor 4*

Results from our previous studies demonstrated that activation of Toll-like receptor 4 (Tlr4), the lipopolysaccharide (LPS) receptor, is sufficient to induce nuclear factor kappaB activation and expression of inducible cyclooxygenase (COX-2) in macrophages. Saturated fatty acids (SFAs) acylated in lipid A moiety of LPS are essential for biological activities of LPS. Thus, we determined whether these fatty acids modulate LPS-induced signaling pathways and COX-2 expression in monocyte/macrophage cells (RAW 264.7). Results show that SFAs, but not unsaturated fatty acids (UFAs), induce nuclear factor kappaB activation and expression of COX-2 and other inflammatory markers. This induction is inhibited by a dominant-negative Tlr4. UFAs inhibit COX-2 expression induced by SFAs, constitutively active Tlr4, or LPS. However, UFAs fail to inhibit COX-2 expression induced by activation of signaling components downstream of Tlr4. Together, these results suggest that both SFA-induced COX-2 expression and its inhibition by UFAs are mediated through a common signaling pathway derived from Tlr4. These results represent a novel mechanism by which fatty acids modulate signaling pathways and target gene expression. Furthermore, these results suggest a possibility that propensity of monocyte/macrophage activation is modulated through Tlr4 by different types of free fatty acids, which in turn can be altered by kinds of dietary fat consumed.


SUMMARY
Results from our previous studies demonstrated that activation of toll-like receptor 4 (Tlr4), the lipopolysaccharide (LPS) receptor, is sufficient to induce NFκB activation and expression of inducible cyclooxygenase (COX-2) in macrophages. Together, these results suggest that both SFA-induced COX-2 expression and its inhibition by UFAs are mediated through a common signaling pathway derived from Tlr4. These results represent a novel mechanism by which fatty acids modulate signaling pathways and target gene expression. Furthermore, these results suggest a possibility that propensity of monocyte/macrophage activation is modulated through Tlr4 by different plasma free fatty acids, which in turn can be altered by kinds of dietary fat consumed.

INTRODUCTION
Cyclooxygenase [COX; prostaglandin endoperoxide (PGH2) synthase] catalyzes the conversion of arachidonic acid to prostaglandin endoperoxide. This is the ratelimiting step in prostaglandin (PG) and thromboxane biosynthesis. Two isoforms of COX have been cloned from various animal cells: constitutively expressed COX-1 (1)(2)(3)(4)(5) and mitogen-inducible COX-2 (6)(7)(8)(9)(10)(11). Numerous studies have demonstrated that the levels of PGs in various tumors, or the tumor's biosynthetic capacity of PGs, are greater when compared with normal tissues (12)(13)(14)(15)(16). Recently, it has been shown that the inducible form of COX is overexpressed in sites of inflammation and in many types of tumor tissues (17)(18)(19)(20). Overexpression of COX-2 in tumor tissues occurs in both tumor cells and stromal cells including macrophages (21). What causes the overexpression of COX-2 in such pathological states is not clearly understood. COX-2 belongs to a family of immediate early response genes which do not require precedent protein synthesis for their expression (22). Therefore, elucidating the signaling pathways leading to the expression of COX-2 is a key to understanding why COX-2 is overexpressed in such pathological states, and can provide critical information for identifying potential targets of modulation by pharmacological and dietary factors.
COX-2 expression is induced by various mitogenic stimuli in different cell types (6,9,11,23). The cis-acting NFκB element is present in the 5′-flanking regions of COX-2 genes of different species (24,25). Results from our previous studies demonstrated that the activation of NFκB is required to induce maximal expression of COX-2 in the lipopolysaccharide (LPS)-stimulated macrophage cell line (26,27). Pro-inflammatory 4 cytokines, such as TNFα and IL-1, also activate NFκB and induce COX-2 expression in many cell types (28,29).
The recent finding that murine Tlr4 is the LPS receptor (30) provided a new impetus in elucidating LPS-induced signaling pathways and target gene expression.
Results from our previous studies indicated that murine Tlr4 confers LPS responsiveness, and that activation of Tlr4 is sufficient to induce NFκB activation and expression of COX-2 in macrophages (27). The lipid A moiety possesses most of the biological activities of LPS (31). Lipid A of Escherichia coli and Salmonella typhimurium is a β,1-6 linked disaccharide of glucosamine, acylated with R-3-hydroxylaurate or myristate and phosphorylated at positions 1 and 4′. The 3-hydroxyl groups of these saturated fatty acids are further 3-O-acylated by lauric acid, myristic acid or palmitic acid (31). These acyl-linked saturated fatty acids are subject to hydrolysis by acyloxyacyl hydrolase; the deacylated LPS loses its endotoxic properties (32,33). This implies that fatty acids acylated in Lipid A may play an important role in LPS-mediated signaling pathways. In light of the finding that murine Tlr4 is the LPS receptor (30), it is important to determine whether these fatty acids modulate Tlr4-mediated signaling pathways and the expression of target gene products. If they do, this will represent a new paradigm for the mechanism by which gene expression is regulated by fatty acids.
Activation of monocytes/macrophages is an important initial step in the cascades of events leading to many inflammatory diseases including endotoxemia (34). If activation of macrophages is modulated by types of fatty acids through Tlr4, it can be inferred that risk for such diseases may also be modified by different types of fatty acids. 5
Lipopolysaccharide (LPS) was purchased from DIFCO (Detroit, Michigan). Bovine serum albumin (BSA, fatty acid free and low endotoxin, Cat.No. A8806) and human recombinant TNFα were purchased from Sigma. Polyclonal antibodies for COX-2 were prepared and characterized as described previously (21,23). Antibodies for iNOS and IL-1α were purchased from Santa Cruz Biotech (Santa Cruz, CA). Donkey anti-rabbit immunoglobulin G (IgG) antibodies conjugated to horseradish peroxidase were Preparation of fatty acids-albumin complexes-All fatty acids were solubilized in ethanol. They were combined with fatty acid free and low endotoxin BSA at a molar ratio of 10:1 (fatty acid: albumin) in serum poor medium (0.25% FBS). Fatty acidsalbumin complex solution was freshly prepared prior to each experiment.

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and
immunoblotting-These were performed as previously described (26,35). Briefly, solubilized proteins were subjected to 8% SDS-PAGE for COX-2, iNOS, IL-1α and GAPDH immunoblot analyses. Following electrophoresis, the gel was transferred to a PVDF membrane and the membrane was blocked to prevent non-specific binding of antibodies in TBS-T [20 mM Tris HCl, 137 mM NaCl, 0.05% (v/v) Tween 20, pH 7.6] containing 5% non-fat dried milk (Carnation). Immunoblotting was performed using respective polyclonal antibodies followed by incubation with anti-rabbit IgG coupled to horseradish peroxidase. The membrane was exposed on an X-ray film (Kodak) using ECL western blot detection reagents (Amersham). Transient transfection and luciferase assay-These were performed as described in our previous studies (27,35). Briefly, RAW 264.7 cells were plated in 6-well plates (5 8 x 10 5 cells/well) and transfected with luciferase reporter plasmids and HSP70-βgalactosidase plasmid as an internal control using SuperFect transfect reagent (Quiagen)

Plasmids-The
according to the manufacturer's instruction. Luciferase and β-galactosidase enzyme activities were determined using the Luciferase Assay System and β-galactosidase Enzyme System (Promega, Madison, WI) according to the manufacturer's instruction.
Luciferase activity was normalized by β-galactosidase activity.

Induction of COX-2 expression by saturated fatty acids is mediated through the activation of NFκ κ κ κBIn our previous studies it was demonstrated that activation of
NFκB is sufficient and required to induce maximal expression of COX-2 in LPS-9 stimulated RAW 264.7 cells (27). Therefore, we determined whether saturated fatty acid-induced COX-2 expression is mediated through the activation of NFκB in RAW 264.7 cells. Lauric acid activated NFκB in a dose-dependent manner ( Fig. 2A). The expression of COX-2 induced by lauric acid was inhibited by co-transfection of a dominant-negative mutant of IκBα plasmid (Fig. 2B). In addition, lauric acid-induced COX-2 expression was significantly reduced in the COX-2 promoter reporter gene containing mutated NFκB site as compared with the one containing wild-type NFκB site  induced NFκB activation and COX-2 expression (Fig. 4A and B). These results suggest that the upstream target in the signaling pathways through which saturated fatty acids mediate NFκB activation and COX-2 expression is Tlr4 or its associated molecules.

Saturated fatty acid-induced COX-2 expression is inhibited by a dominant
However, these results do not allow us to conclude whether saturated fatty acids directly interact with Tlr4.

Unsaturated fatty acids inhibit saturated fatty acid-induced COX-2 expression,
and this inhibition is mediated through suppression of NFκ κ κ κB   Unlike saturated fatty acids, unsaturated fatty acids were unable to induce COX-2 expression (Fig. 1D).
Furthermore, they inhibited saturated fatty acid-induced NFκB activation (Fig. 5A) and 11 COX-2 expression (Fig. 5B). These results indicate that inhibition of saturated fatty acidinduced COX-2 expression by unsaturated fatty acids is mediated through suppression of NFκB signaling pathway. Together, these results suggest that both the induction of COX-2 by saturated fatty acids and its inhibition by unsaturated fatty acids are mediated through NFκB signaling pathway.

DISCUSSION
Most long-chain fatty acids are esterified in cellular lipids in mammalian cells.
Therefore, the concentrations of unesterified fatty acids are believed to be low. However, fatty acids are rapidly released by the action of various phospholipase A 2 and monoacylglycerol and diacylglycerol lipases in response to various extracellular stimuli.
In plasma the average concentration of free fatty acid in postabsorptive state is < 0.7 mM, 13 and this concentration may be much higher in absorptive phase after ingestion of a fatty meal (45). Therefore, blood cells such as monocytes are constantly exposed to relatively high concentrations of free fatty acids. Fatty acids are known to regulate the expression of many genes involved in lipid metabolism (45) and modulate activity of signaling molecules such as phospholipase C and protein kinase C (46, 47). The mechanism by which fatty acids can regulate gene expression is still not well understood. However, some conceptual framework has been proposed for the possible mechanism of actions. 5′-flanking region. Deletion of these sequences did not affect the promoter activity of COX-2 reporter gene (Fig. 3) suggesting that the PPRE-like sequences do not appear to be required for saturated fatty acid-induced COX-2 expression. However, the possibility that the saturated fatty acids in part stimulate or inhibit other PPAR-responsive gene products which in turn cause the induction of COX-2 expression cannot be ruled out.
It was shown that unsaturated fatty acids induce COX-2 expression in mammary epithelial cells (43). Whether this induction is mediated through PPARs has not been determined. However, to our surprise, saturated fatty acids, but not unsaturated fatty acids, induce COX-2 in RAW 264.7 cells (Fig. 1). Greater potency of lauric acid and palmitic acid in inducing COX-2 expression among saturated fatty acids tested (Fig. 1C) coincides with the abundance of these fatty acids in lipid A molecule (31). Lauric acid, myristic and palmitic acid are known to be major fatty acids acylated in lipid A molecule (31). The fact that deacylation of these fatty acids from LPS results in loss of endotoxic activity (32,33) implies an important role of these fatty acids in LPS-mediated signal transmission. NFκB is one of the major downstream signaling pathways derived from activation of LPS receptor, Tlr4 in RAW 264.7 cells (27). The results demonstrating that induction of COX-2 by lauric acid is mediated through activation of NFκB (Fig. 2) and that this activation is inhibited by a dominant-negative mutant of Tlr4 (Fig. 4A), suggest that the most upstream signaling components affected by saturated fatty acids include Tlr4 or molecules associated with Tlr4. Whether saturated fatty acids can directly interact with Tlr4, or they interact with molecules associated with either extracellular or intracellular domains of Tlr4, remains to be determined. 15 The results presented in Fig. 4 and Fig. 6 suggest that activation of NFκB and COX-2 expression induced by saturated fatty acids and inhibition of this induction by unsaturated fatty acids are mediated through a common signaling pathway derived from Tlr4. The possibility that saturated fatty acids may act as a physiologically relevant endogenous ligand for Tlr4, and that unsaturated fatty acids interfere with saturated fatty acids in interacting with Tlr4 or molecules associated with Tlr4 remains to be determined.
While the detail mechanism by which saturated and unsaturated fatty acids interact with Tlr4 or its associated molecules is not known, the results presented in this report represent a novel mechanism by which fatty acids modulate signaling pathways and the expression of target genes. Furthermore, the results strongly imply that cellular

ACKNOWLEDGMENT:
We thank Dr. Walter A. Deutsch for reading the manuscript, and Wei Fan for technical assistance.