INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

A GROUP OF 18-C FATTY ACID isomers referred to as conjugated linoleic acids (CLA) has been shown to have many biological effects, including modulation of the immune system (9, 26, 37, 38). We previously showed that dietary CLA lowered prostaglandin (PG)E₂ and histamine release in response to antigen challenge (42) in a guinea pig model for type I (immediate) hypersensitivity, a standard model for studying asthma and allergies (12, 41). The mechanism by which CLA lower PGE₂ levels may be, at least in part, inhibition of cyclooxygenase (COX), the enzyme responsible for the initial steps in converting arachidonic acid into prostanoids [PG and thromboxanes (TX)] (2). CLA isomers are similar to linoleic acid (the precursor of arachidonic acid), except the double bonds are in a conjugated formation (i.e., a 1,3-diene, not methylene interrupted). This structural difference (a conjugated diene) in similar 20-C fatty acids has been shown to inhibit the enzymatic activity of COX in vitro as measured by PG production (28).

For many years, prostanoids have been implicated in the pathogenesis of type I hypersensitivity disorders such as asthma. We hypothesized that feeding dietary CLA would decrease prostanoids and leukotrienes (LT) released in response to antigen challenge. A liquid chromatography/tandem mass spectrometry (LC/MS/MS) method was used to quantify the prostanoids released from sensitized guinea pig tissues exposed to antigen in tissue baths. Stable products of each of the five possible pathways downstream of COX were analyzed. LT were quantified using a combination of LC/MS/MS and enzyme immunoassay. Previous investigations of type I hypersensitivity indicated that increased abundance of COX-2 (the inducible isoform) was associated with increased prostanoid release (2). Therefore, we evaluated the effects of the antigen sensitization and challenge on COX-2 protein in lung tissues studied.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Diets and sensitization. Female Hartley guinea pigs (Harlan Sprague-Dawley, Madison, WI) weighing 200-350 g were housed in a humidity- and temperature-controlled room with 12:12-h light-dark cycle in accordance with University of Wisconsin-Madison Research Animal Resources Center. Diets consisted of a standard guinea pig chow¹ (#7006, Harlan-Teklad, Madison, WI) supplemented with 0.25% of safflower oil (control) or CLA² (Natural Lipids, Hovdebygda, Norway). Guinea pigs were given free access to food and water for 14 days before and during active sensitization to chicken egg ovalbumin (OVA; A5503, Sigma, St. Louis, MO) for a total of 32 days on the diet. Body weights and feed consumption were recorded weekly, but did not differ between diet groups, and therefore, they were not reported. Guinea pigs were sensitized to OVA with an initial intraperitoneal injection of 50 µg OVA in PBS with 1 mg aluminum hydroxide followed 14 days later by a subcutaneous injection (flank) of 200 µg OVA in PBS emulsified with incomplete Freund's adjuvant (1:1 vol:vol) (42). Guinea pigs were killed with an intraperitoneal injection of pentobarbital sodium (100 mg/kg) 4 days after the subcutaneous injection.

Tissue baths. Lungs, trachea, and bladder were removed, placed in physiological saline solution (PSS; 37°C, gassed with 95% O₂-5% CO₂), trimmed of excess tissue, and cut several times with scissors (but kept intact) to increase cellular exposure to PSS and antigen. The PSS was a bicarbonate buffer solution containing (in mmol/l): 118 NaCl, 1.0 NaH₂PO₄, 4.7 KCl, 2.5 CaCl₂, 0.5 MgCl₂, 11 glucose, and 25 NaHCO₃, pH = 7.4. One lung was frozen in liquid nitrogen immediately following removal for subsequent fatty acid analysis or COX-2 Western blot analysis. Tissues were incubated in PSS for 1 h. PSS was discarded and replaced with fresh PSS every 15 min. After equilibration, tissues were placed in 10 ml fresh PSS for 1 h. PSS was collected for analysis of basal mediator release. Tissues were then incubated in 10 ml fresh PSS containing 0.01 g/l OVA for 1 h. PSS was collected for analysis for antigen-induced mediator release (challenge). PSS samples from basal and challenge incubations were stored on ice after collection (<4 h) until analysis by enzyme immunoassay or processing through C-18 Bond Elut solid-phase extraction columns (Varian, Walnut Creek, CA). Column extraction was performed according to previously published methods (33). Column-extracted samples were then stored at 80°C until analysis.

Mediator analysis. Lipid-derived mediators were quantified by HPLC coupled with negative ion electrospray tandem mass spectrometry, as similarly described previously (25). Briefly, a reverse-phase HPLC separation was used to resolve all compounds of interest and to remove interfering compounds. On-line detection was achieved using electrospray tandem mass spectrometry. Consistently accurate and precise recovery of target compounds from a blank matrix was demonstrated over a range of 1 to 100 ng in 10-ml aqueous samples for all prostanoids except PGD₂, which was detectable as low as 5 ng/ml. Linearity of response to standards was demonstrated over a range of 0.10 to 10 ng/ml in the injected solution. This method was successfully used to profile five prostanoids (TXB₂, 6-keto-PGF₁, PGF₂, PGE₂, and PGD₂) and LTB₄ in this study. Because of low recovery from solid-phase extraction columns, a separate assay (enzyme immunoassay) was performed to analyze LTC₄/D₄/E₄ (RPN 224, Amersham, Arlington Heights, IL) in tissue bath samples. LTD₄ and E₄ are synthesized from LTC₄ by sequential enzymatic cleavage steps. This assay does not distinguish LTC₄, LTD₄, and LTE₄ from each other (% cross-reactivity for LTC₄ and LTD₄ is 100%, and for LTE₄ it is 30%; all other eicosanoids and fatty acids do not cross-react).

Microsomal isolation. Microsomes were isolated from unchallenged (frozen immediately after removal) and challenged lungs (frozen immediately following tissue bath experiment described above) of three guinea pigs from each treatment. Lungs were thawed on ice, minced with scissors, and homogenized in 3 vol sucrose buffer (250 mM sucrose, 10 mM Tris · HCl, and 1 mM EDTA, pH = 7.8). Homogenates were centrifuged at 12,000 g for 20 min at 4°C. Supernatants were then centrifuged at 110,000 g for 60 min at 4°C. The pellet was washed with 50 mM Tris · HCl and resuspended in 2 ml ice-cold 50 mM Tris · HCl with 0.1% Triton X-100, frozen in liquid nitrogen, and stored at 80°C until Western blot analysis.

Western blot analysis. COX-2 expression was measured by Western blot analysis. Protein content of microsomal isolations was determined by Bradford assay (6) using Protein Assay Dye Reagent (Bio-Rad Laboratories, Hercules, CA) and bovine serum albumin for a standard curve. Thirty micrograms of protein from each sample were steamed for 5 min and then separated by SDS-PAGE with a Versatile Mini-Protean II Electrophoresis Cell (Bio-Rad Laboratories) using a 4% stacking and a 10% separating acrylamide gel. COX-2 enzyme (ovine, Cayman Chemical, Ann Arbor, MI) was also electrophoresed as a positive control. Protein was electrophoretically transferred to Immobilon-P Transfer Membranes (Millipore, Bedford, MA) using a Mini Trans-Blot Transfer Cell (Bio-Rad Laboratories), and membranes were then soaked in blocking buffer (100 mM Tris · HCl, pH = 7.5; 0.9% NaCl, 0.1% Tween 20, 5% nonfat dry milk) at 4°C overnight. Membranes were incubated with goat polyclonal anti-COX-2 antibody (1:500 dilution in blocking buffer) specific for COX-2 (Cox-2 N-20:sc-1746, Santa Cruz Biotechnology, Santa Cruz, CA) at 37°C for 1.5 h, rinsed with TST buffer (100 mM Tris · HCl, pH = 7.5; 0.9% NaCl, 0.1% Tween 20) five times for a total of 1 h, incubated with anti-goat horse radish peroxidase (HRP)-conjugated antibody (1:500 dilution in blocking buffer, Santa Cruz Biotechnology) at 37°C for 45 min, rinsed as before with TST buffer, incubated with Renaissance Western Blot Chemiluminescence Reagent (NEN Life Science Products, Boston, MA) for 1 min, and exposed to X-ray film for 30 s and 1 min.

Fatty acid analysis. Fatty acid composition of the diet and CLA isomer composition of the diet and guinea pig tissues were determined by gas chromatography (GC) using previously published methods (8).

Statistical analysis. Prostanoid and LT data were analyzed as a 2 × 2 factorial design with random effects using the mixed procedure to test for diet effect on mediator release at basal and challenge levels (SAS, Cary, NC). For each mediator and tissue, we fit the following model
where Y denotes the mediator response, µ indicates the overall mean, day is a random effect that represents day-to-day variability, time(day) represents the effect of time-to-time variability nested within day, animal(time) represents animal-to-animal variability nested within time, and  denotes the within-animal variability. A Tukey-Kramer correction was made for the multiple comparisons. Fatty acid contents of lungs, tracheas, and bladders from control vs. CLA-fed guinea pigs were compared by Student's t-test followed by a Bonferroni adjustment for multiple comparisons. A 95% confidence interval was used to test significance.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Tissue fatty acid analysis. Concentrations of c9, t11/t9, c11, t10, and c12 CLA isomers in both the neutral lipid and phospholipid fractions of all tissues analyzed from guinea pigs fed CLA were significantly increased (Tables 1-3) except in the bladder phospholipid fraction. Although the amounts of c9, t11/t9, c11 isomers and t10, c12 isomers were similar in the diet (41% compared with 43%), in the tissues, the range of c9, t11/t9, c11 isomers was 52-63% of the total CLA isomers, and the range of t10, c12 isomer was 35-44% of the total CLA isomers. Other investigators (22, 38) reported this same observation. In the tracheas, 14:0 was significantly increased in the phospholipid fraction of the CLA group, and 18:1 (n-9) was significantly decreased in the neutral lipid fraction of the CLA group. All other fatty acid levels in all tissues studied (not including CLA isomers) were not significantly affected by the dietary treatment.

Prostanoid and LT release. Antigen-induced release of all prostanoids from tracheas, four of five from lungs, and one of five (other 4 of 5 decreased, but not statistically significantly) from bladder was significantly reduced (P  0.05) in those tissues from CLA-fed guinea pigs compared with control-fed guinea pigs. However, there was no dietary effect on prostanoid release at basal levels in tracheas, lungs, and bladders (Figs. 1-3). The interaction of diet and antigen challenge was significant for all prostanoids released from lungs and tracheas except for PGE₂ in both tissues and PGD₂ in lungs. The diet by challenge interaction was not significant for any of the prostanoids released from bladders.

View larger version (27K): [in this window] [in a new window]   Fig. 1.   Prostanoid release from lungs of guinea pigs fed a control diet or a conjugated linoleic acid (CLA)-supplemented diet. Tissues were incubated in fresh physiological saline solution (PSS) for 1 h to determine basal levels of mediator release followed by incubation for 1 h in PSS + ovalbumin (0.01 g/l) to determine mediator release in response to antigen challenge. Prostanoids were analyzed using liquid chromatography/tandem mass spectrometry. Values are least-squared means, error bars represent standard error of the least-squared means, and n = 8 per diet treatment. Bars within a prostanoid group with different letters are significantly different at P  0.05. A Tukey-Kramer correction for multiple comparisons was made. An ANOVA table from a 2 × 2 factorial analysis is also shown. TXB2, thromboxane B2; PG, prostaglandin.

View larger version (29K): [in this window] [in a new window]   Fig. 2.   Prostanoid release from tracheas of guinea pigs fed a control diet or a CLA-supplemented diet. See Fig. 1 for description of terms and methods.

View larger version (26K): [in this window] [in a new window]   Fig. 3.   Prostanoid release from bladders of guinea pigs fed a control diet or a CLA-supplemented diet. See Fig. 1 for description of terms and methods.

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Cysteinyl leukotriene (LTC4/D4/E4) release from lungs, tracheas, and bladders of guinea pigs fed a control diet or a CLA-supplemented diet. Leukotrienes were analyzed by enzyme immunoassay. See Fig. 1 for description of terms and methods.

Western blot analysis of COX-2 protein levels. COX-2 protein levels of guinea pig lungs before (immediately after death) and after antigen challenge were analyzed by immunoblotting to determine if the decreased release of prostanoids in response to antigen challenge was due to decreased translation of COX-2. There were no apparent differences between control- and CLA-fed groups before or following antigen challenge (Fig. 5).

View larger version (55K): [in this window] [in a new window]   Fig. 5.   Western blot analysis of cyclooxygenase-2 protein levels before (lungs frozen immediately after death) and after antigen challenge (lungs frozen immediately after antigen challenge in tissue bath). Western blot image shown is representative of 6 lungs from each dietary treatment.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The most significant finding in this study was decreased release of various prostanoids and cysteinyl LTs in response to antigen challenge by tissues of guinea pigs fed CLA. A similar decrease in PGE₂ release from antigen-sensitized and -challenged guinea pig tracheas was shown in an earlier study (42). Dietary CLA has also been shown to decrease prostanoid levels in serum (37, 38), bone (22), spleen (38), and cultured keratinocytes (24). However, all of those studies only reported the effect of CLA on constitutive or chemically induced PGE₂ production and did not evaluate antigen-induced release. To our knowledge, the study reported here is the first to investigate the effect of dietary CLA on antigen-stimulated release of the entire range of prostanoids and LTs. In addition to the significance of decreased release of eicosanoids in response to antigen sensitization and challenge, it is also important to note that there was no effect of dietary CLA on basal prostanoid or LT release, indicating that the protective role played by constitutive release of some of these mediators is not compromised by dietary CLA.

On the basis of Western blot analysis results (no apparent differences associated with dietary treatment), feeding CLA before and during antigen sensitization did not alter existing COX-2 protein levels at the time of antigen challenge. Therefore, decreased prostanoid release in response to antigen challenge does not appear to be due to decreased COX-2 enzyme protein levels.

Although it has been proposed that the biological effects of CLA may be due to changes in composition of phospholipid fatty acids (4, 26), the effects of CLA on prostanoid and LT release in this study were not likely due to changes in fatty acid precursors, because phospholipid concentrations of linoleic acid and arachidonic acid were not affected by CLA. Although other studies showed a similar lack of effect of CLA feeding on linoleic acid and arachidonic acid levels in a broad range of tissues (18, 24, 38, 40), some studies show that CLA decreases linoleic acid and/or arachidonic acid in the liver (4, 22, 24). This could depend, in part, on the amount of linoleic acid in the comparison group (24) and whether or not phospholipid and neutral lipids were analyzed separately.

A possible mechanism by which CLA decreases prostanoid and LT release is downregulation of enzyme activity of either COX and lipoxygenase or phospholipase A2, the enzyme responsible for cleaving precursor fatty acids from the phospholipid before cyclooxygenation or lipoxygenation. As discussed earlier, evidence already exists supporting this hypothesis for COX. Elongated and desaturated 20-C counterparts to CLA (c8, t12, c14 eicosatrienoate and c5, c8, t12, c14 eicosatetraenoate) competitively inhibited COX enzyme activity very strongly (28). These 20-C metabolites of CLA have been recovered from livers of rats fed CLA (36). A more recent report also demonstrates inhibition of COX by various isomers of CLA (7). If COX inhibition is the method by which CLA is acting, we hypothesize that one or more enzymes in the lipoxygenase pathway are also being inhibited because rather than potentiation of the LT release [which is typically seen when COX alone is inhibited (19)], we showed that these mediators are also downregulated in CLA-fed animals.

The purpose of this study was to determine 1) if the entire profile of products from the COX and lipoxygenase pathways was decreased in CLA-fed animals, and if the decrease was during basal and/or antigen-challenged conditions; 2) if the decrease in these products was due to a decrease in abundance of COX-2; and 3) if the decrease in the products was due to a decrease in substrate availability. Although the decrease does not seem to be due to the latter two mechanisms, and published data support the possible role of CLA inhibition of the enzymes involved, additional questions remain to be answered. For example, it is unclear whether or not CLA decreases prostanoid and LT release through a histamine-related mechanism. We previously showed that tracheas from CLA-fed animals have decreased histamine release in response to antigen challenge (42). Because histamine induces release of the mediators measured (32), CLA could be acting through inhibition of histamine release rather than directly affecting prostanoid and LT synthesis. However, because of the methods required for prostanoid and LT analysis, we were unable to analyze histamine levels in this study. Alternatively, a decrease in prostanoids early in the feeding protocol could have resulted in decreased sensitization of the guinea pigs as PGE₂ increases formation of IgG₁ (the reaginic immunoglobulin in guinea pigs) and IgE (34). However, this study was not designed to investigate the effects of CLA on antigen sensitization of the guinea pigs.

Results from this study have implications in several disease states. For example, in the airways, prostanoids contribute to airway hypersecretion and inflammation (14, 19) and cysteinyl LTs are potent bronchoconstrictors (11, 21). Drug therapies for airway disorders have been targeted at limiting production of lipid mediators produced by the COX (3) and lipoxygenase (3, 10) pathways. The ability of CLA to downregulate both pathways is of particular importance in aspirin-sensitive airway disorders. In these patients, inhibition of the COX pathway with aspirin can induce or worsen asthma, possibly because of a shift in substrate from the COX pathway to the lipoxygenase pathway (30). Because CLA decreases cysteinyl LTs and has no effect on LTB₄, the downregulation of the COX-mediated pathway by CLA would not likely lead to increased LT synthesis as aspirin does in these patients.

Results from this study also have implications for bladder disorders. Use of the sensitized guinea pig model to study allergy-related conditions affecting bladder has been established (35). The capacity of dietary CLA to decrease lipid-mediator release, especially cysteinyl LTs, from bladders in response to antigen challenge could prove relevant for noninfectious inflammatory bladder conditions such as interstitial cystitis (5).

Although not specifically investigated in this study, these results may also be relevant to a broader range of diseases that are influenced by mediators that are decreased by CLA. Increased PG synthesis is associated with atherosclerosis (13), and inhibition of COX has been shown to improve endothelial dysfunction associated with atherosclerosis (16). In fact, there is evidence from atherosclerosis models to suggest that CLA may be beneficial (20). COX-2 is upregulated in systemic lupus erythematosus-associated nephritis, and prostanoids contribute to the associated renal dysfunction (39). CLA has also been shown to have a beneficial effect in a lupus model, increasing the survival time of mice after they developed proteinuria (43). COX-2-specific inhibitors are being tested as chemopreventors, especially in colon cancer (15), and COX-2 has been reported to have a role in the pathogenesis of bladder cancer (27, 29). CLA has been shown in numerous models to be anticarcinogenic (17, 31). COX-2 is also upregulated in arthritis, and COX-2 inhibition reverses edema in arthritic paws of rats (1). COX-2 inhibitors have also been implicated in the prevention and treatment of Alzheimer's disease (23).

Perspectives