Nitrolinoleic acid: An endogenous peroxisome proliferator-activated receptor γ ligand

Materials and Methods

Materials. LNO₂ and [¹³C]LNO₂ were synthesized and purified by using LA (NuCheckPrep, Elysian, MN) and [¹³C]LA (Spectra Stables Isotopes, Columbia, MD) subjected to nitroselenylation as described (9). LNO₂ concentrations were quantified spectroscopically and by chemiluminescent nitrogen analysis (Antek Instruments, Houston) by using caffeine as a standard (9). Anti-CD36 Ab was kindly provided by F. Debeer (University of Kentucky Medical Center, Lexington, KY); anti-PPARγ and anti-β-actin Abs were from Santa Cruz Biotechnology; and anti-aP2 Ab was from Chemicon. Horseradish peroxidase-linked goat anti-rabbit IgG and Coomassie blue were from Pierce. [³H]rosiglitazone was from American Radiolabeled Chemicals (St. Louis). 2-deoxy-d-[³H]glucose was from Sigma. ScintiSafe Plus 50% was from Fisher Scientific. Rosiglitazone, ciglitazone, 15-deoxy-Δ^12,14-PGJ₂, conjugated LA (CLA1 and CLA2), and GW9662 were from Cayman Chemical (Ann Arbor, MI). The 1-palmitoyl-2-hydroxy-sn-glycero-3-phosphate [16:0 lysophosphatidic acid (LPA)], 1-O-9-(Z)-octadecenyl-2-hydroxy-sn-glycero-3-phosphate (18:1 LPA), 1-O-hexadecyl-2-azelaoyl-sn-glycero-3-phosphocholine (azPC), and 1-palmitoy-2-azelaoyl-sn-glycero-3-phosphocholine (azPC ester) were from Avanti Polar Lipids.

Cell Transient Transfection Assay. CV-1 cells from American Type Culture Collection were grown to ≈85% confluence in DMEM/F12 supplemented with 10% FBS and 1% penicillin–streptomycin. Then, cells were transiently cotransfected with a plasmid containing the luciferase gene under regulation by four Gal4 DNA-binding elements (UAS_G × 4 TK-luciferase, a gift from Ronald M. Evans, The Salk Institute, La Jolla, CA), in concert with plasmids containing the ligand-binding domain for the different nuclear receptors fused to the Gal4 DNA-binding domain. For assessing full-length PPAR receptors, CV-1 cells were transiently cotransfected with a plasmid containing the luciferase gene under the control of three tandem PPAR response elements (PPRE) (PPRE × 3 TK-luciferase) and hPPARγ, hPPARα, or hPPARδ expression plasmids, respectively. In all cases, GFP expression plasmid was cotransfected as the control for the transfection efficiency. After the transfection (24 h), cells were cultured for another 24 h in OPTIMEM (Invitrogen). Then, cells were treated with different compounds as indicated in the figure legends for 24 h in OPTIMEM. Reporter luciferase assay kits from Promega were used to measure the luciferase activity, according to the manufacturer's instructions, with a luminometer (Victor II, Perkin–Elmer). Luciferase activity was normalized by GFP units. Each condition was performed with n ≥ 3 for each experiment. All experiments were repeated at least three times.

PPARγ Competition-Binding Assay. Human PPARγ1 cDNA was inserted into pGEX (Amersham Pharmacia Biosciences) containing the gene encoding GST. GST-PPARγ protein induction and receptor binding was assessed as described in ref. 17.

The 3T3-L1 Differentiation and Oil Red O Staining. The 3T3-L1 preadipocytes were propagated and maintained in DMEM containing 10% FBS. To induce differentiation, 2-d postconfluent preadipocytes (designated day 0) were cultured in DMEM containing 10% FBS plus 3 μM LNO₂ for 14 d. The medium was changed every 2 d. Rosiglitazone (3 μM) and LA (10 μM) were used as the positive and negative controls, respectively. The differentiated adipocytes were stained by Oil red O as described in ref. 18.

The 2-Deoxy-d-[³H]glucose Uptake Assay in Differentiated 3T3-L1 Adipocytes. The 2-Deoxy-d-[³H]glucose uptake assay was performed as described in ref. 19. 3T3-L1 preadipocytes were grown in 24-well tissue culture plates; 2-d postconfluent preadipocytes were treated by 10 μg/ml insulin, 1 μM dexamethasone, and 0.5 mM 3-isobutyl-1-methylxanthine (all from Sigma) in DMEM containing 10% FBS for 2 d, and then cells were kept in 10 μg/ml insulin also in DMEM containing 10% FBS for 6 d (changed medium every 3 d). After induction of adipogenesis (8 d), test compounds in DMEM containing 10% FBS were added for an additional 2 d (changed medium every day). The PPARγ-specific antagonist GW9662 was included 1 h before other additions. After two rinses with serum-free DMEM, cells were incubated for 3 h in serum-free DMEM and rinsed at room temperature three times with freshly prepared KRPH buffer (5 mM phosphate buffer/20 mM Hepes/1 mM MgSO₄/1 mM CaCl₂/136 mM NaCl/4.7 mM KCl, pH 7.4). The buffer was replaced with 1 μCi/ml 2-deoxy-d-[³H]glucose in KRPH buffer for 10 min at room temperature. The treated cells were rinsed carefully three times with cold PBS, lysed overnight in 0.8 M NaOH (0.4 ml per well), and neutralized with 13.3 μl of 12 M HCl. Lysate (360 μl) was added to 4 ml of ScintiSafe Plus 50% in a scintillation vial, and the vials were mixed and counted.

RNA and Protein Preparation and Analysis. RNA and protein expression levels were analyzed by quantitative real-time PCR and Western blot analysis as described in ref. 20.

LNO₂ Decay. LNO₂ decay was induced by incubating 3 μM LNO₂ in medium plus serum at 37°C. At different times, samples were removed and analyzed for bioactivity or for LNO₂ content by Bligh and Dyer extraction in the presence of 1 μM [¹³C]LNO₂ as an internal standard. Nondecayed LNO₂ was quantified by means of triple quadruple mass spectrometric (MS) analysis (Applied Biosystems/MDS Sciex, Thornhill, Ontario, Canada) as described in ref. 9.

Results

To characterize LNO₂ as a potential ligand for a lipid-binding nuclear receptor [e.g., PPARα, PPARγ, PPARδ, androgen receptor, glucocorticoid receptor, mineralocorticoid receptor, progesterone receptor, and retinoic X receptor α (RXRα)], CV-1 reporter cells were cotransfected with plasmids containing the ligand-binding domain for these nuclear receptors fused to the Gal4 DNA-binding domain and the luciferase gene under regulation of four Gal4 DNA-binding elements. LNO₂ (1 μM) induced significant activation of PPARγ (620%), PPARα (325%), and PPARδ (221%), with no impact on androgen, glucocorticoid, mineralocorticoid, progesterone, or RXRα receptor activation (Fig. 1A). To further explore PPAR activation by LNO₂, CV-1 cells were transiently cotransfected with a plasmid containing the luciferase gene under three PPRE in concert with PPARγ, PPARα, or PPARδ expression plasmids. Dose-dependent activation by LNO₂ was observed for all PPARs (Fig. 1B), with PPARγ showing the greatest response at clinically relevant LNO₂ concentrations.

PPARγ activation by LNO₂ rivaled that induced by ciglitazone and rosiglitazone and exceeded that of 15-deoxy-Δ^12,14-PGJ₂ (21, 22), which only occurred at concentrations three orders of magnitude greater than found clinically (Fig. 1C and Inset). Several other reported endogenous PPARγ activators added in equimolar concentration with LNO₂ [LA, conjugated LA (CLA1, CLA2), LPA (16:0, 18:1), azPC, and azPC ester; 1 and 3 μM], displayed no significant activation of PPARγ reporter gene expression when compared with vehicle control (Fig. 1C) (23–25). LNO₂-mediated PPARγ activation was inhibited by the PPARγ-specific antagonist GW9662 and enhanced ≈180% by coaddition of the RXRα agonist 9-cis-retinoic acid, which facilitates PPRE promoter activation via heterodimerization of activated RXRα with PPARγ (Fig. 2A).

LNO₂ slowly undergoes decay reactions in aqueous solution, displaying a 30- to 60-min half-life (Fig. 2B), yielding NO and an array of oxidation products (data not shown). The activation of PPARγ paralleled the presence and concentration of the LNO₂ parent molecule, as determined by concomitant receptor activation analysis and MS quantitation of residual LNO₂ in reaction mixtures undergoing different degrees of aqueous decomposition before addition to CV-1 cells (Fig. 2B). Affirmation that activation of PPARγ is LNO₂-specific, rather than a consequence of LNO₂ decay products, was established by treatment of PPARγ-transfected CV-1 cells with NO donors and oxidized LA derivatives. Additionally, the fatty acid oxidation products 9-oxo-10(E), 12(Z)-octadecadienoic acid (9-oxoODE) and 13-oxo-9(E),11(Z)-octadecadienoic acid (13-oxoODE) are reported endogenous PPARγ stimuli (26, 27). The 13-oxoODE had no effect on PPARγ-dependent reporter gene expression (Fig. 2C). Only high and nonphysiological concentrations of 9-oxoODE, and the concerted addition of high concentrations of 13-oxoODE and S-nitrosoglutathione or spermine-NONOate [(Z)-1-{N-[3-Aminopropyl]-N-[4-(3-aminopropylammonio)butyl]-amino}-diazen-1-ium-1,2-diolate], resulted in modest PPARγ activation (Fig. 2C). The NO donors added individually did not activate PPARγ, even at high concentrations, eliminating the possibility that PPARγ-mediated signaling by LNO₂ is due to NO derivatives released during LNO₂ decay (Fig. 2C).

Competititive PPARγ binding analysis quantified the displacement of [³H]rosiglitazone by unlabeled rosiglitazone, LNO₂, and LA. The K_i for rosiglitazone was 53 nM, consistent with reported values (Fig. 2D, 40–50 nM; ref. 28). LNO₂ displayed an estimated K_i of 133 nM and LA an estimated K_i of >1,000 nM (previously reported as 1.7–17 μM; ref. 28). This reveals that the K_i for LNO₂ displacement of [³H]rosiglitazone from PPARγ is comparable to this highly avid ligand.

The PPARγ agonist actions of LNO₂ also were examined in a biological context, by using cell models noted for well established PPARγ-dependent functions. The scavenger receptor CD36 is expressed in diverse cell types, including platelets, adipocytes, and macrophages. In macrophages, CD36 is a receptor for oxidized low-density lipoprotein, with expression positively regulated by PPARγ (29). Treatment of mouse RAW264.7 macrophages with LNO₂ induced greater CD36 receptor protein expression than an equivalent rosiglitazone concentration, a response partially inhibitable by the PPARγ-specific antagonist GW9662 (Fig. 3A). Moreover, quantitative real-time PCR revealed that LNO₂-dependent PPARγ transactivation induced a dose-dependent increase in CD36 mRNA expression in these macrophages (data not shown).

PPARγ plays an essential role in the differentiation of adipocytes (30, 31). In support of this precept, selective disruption of PPARγ results in impaired development of adipose tissue (18, 32). To define whether LNO₂ induces PPARγ-activated adipogenesis, 3T3-L1 preadipocytes were treated with LNO₂, rosiglitazone, or LA for 2 wk. Adipocyte differentiation was assessed both morphologically and by means of Oil red O staining, which reveals the accumulation of intracellular lipids. Vehicle and LA did not affect differentiation, whereas LNO₂ induced >30% of 3T3-L1 preadipocyte differentiation (Fig. 3B). Rosiglitazone treatment affirmed a positive PPARγ-dependent response. LNO₂- and rosiglitazone-induced preadipocyte differentiation also resulted in expression of specific adipocyte markers (PPARγ2 and aP2), an event not detected for LA (Fig. 3C). PPARγ ligands play a central role in glucose metabolism, with thiazolidinediones widely used as insulin-sensitizing drugs. Addition of LNO₂ (1–10 μM) to differentiated adipocytes induced a dose-dependent increase in glucose uptake (Fig. 3D Left). The impact of LNO₂ on glucose uptake was greater than that observed for equimolar 15-deoxy-PGJ₂, equivalent to rosiglitazone and similarly inhibited by the PPARγ-specific antagonist GW9662 (Fig. 3D Right). In aggregate, these observations establish that LNO₂ induces well characterized PPARγ-dependent signaling actions toward macrophage CD36 expression and adipogenesis.

Discussion

The identification of bona fide high affinity endogenous PPARγ ligands has been a provocative issue that, when resolved, will advance our understanding of PPARγ modulation and reveal new means for intervention in diverse metabolic disorders and disease processes. This will also shed light on the broader contributions of this nuclear hormone receptor family to both the maintenance of tissue homeostasis and the regulation of cell and organ dysfunction. Of relevance, the generation of PPAR-activating intermediates from complex lipids often requires hydrolysis by phospholipase A₂, with the identity of phospholipid sn-2 position fatty acid derivatives that serve as PPAR agonists still forthcoming (33, 34). Thus, the fact that ≈80% of LNO₂ present in a variety of tissue compartments is esterified encourages the notion that this pool of high affinity PPAR ligand activity will be mobilized upon the phospholipase A₂ activation that occurs during inflammation to regulate the secondary formation of cell signaling molecules.

Comparison of PPARγ-dependent gene expression induced by LNO₂ with that of thiazolidinediones and putative fatty acid- and phospholipid-derived PPARγ ligands affirmed the robust activity of LNO₂ as a PPARγ ligand. Although a number of endogenous lipophilic species are proposed as PPARγ ligands, their intrinsically low binding affinities and in vivo concentrations do not support a capability to serve as physiologically relevant signaling mediators. Presently reported endogenous PPARγ agonists include free fatty acids, components of oxidized plasma lipoproteins (9- and 13-oxoODE, azPC), conjugated LA derivatives (CLA1 and CLA2), products of phospholipase hydrolysis of complex lipids (LPA), platelet-activating factor, and eicosanoid derivatives such as the dehydration product of PGD₂, 15-deoxy-Δ^12,14-PGJ₂ (14, 22, 23). Herein, we observed minimal or no activation of PPARγ reporter gene expression by 1–3 μM concentrations of these putative ligands, in contrast to the dose-dependent PPAR activation by LNO₂ that, for PPARγ, was significant at clinically relevant concentrations as low as 100 nM. A dilemma exists in that some putative endogenous PPARγ agonists have only been generated by aggressive in vitro oxidizing conditions (e.g., Cu-mediated low-density lipoprotein oxidation) and have not been clinically quantified or detected (for instance, azPC and conjugated LA). Other lipid derivatives proposed as PPAR ligands are present in <100 nM tissue concentrations, orders of magnitude below their binding affinities (1–15 μM), and are not expected to result in significant receptor occupancy and activation in vivo. This latter category includes free fatty acids, eicosanoids, 9- and 13-oxoODE, platelet-activating factor, and 15-deoxy-Δ^12,14-PGJ₂ (14). For example, although LPA is a notable PPARγ ligand of relevance to vascular, inflammatory, and cell-proliferative diseases, the plasma concentration of LPA is well below 100 nM (35, 36) and its binding affinity for PPARγ has not been established. Also, the in vivo downstream vascular signaling actions of LPA are inconsistent with its proposed PPARγ ligand activity (24, 37–39). Future studies using mice genetically deficient for PPARγ and NO synthase isoforms should assist in defining the mechanisms of formation and endogenous PPAR ligand activity of nitrated fatty acids.

In summary, we show that LNO₂ is a high-affinity ligand for PPARs, especially PPARγ, that activates both reporter constructs and cells at physiological concentrations. Fatty acid nitration products, generated by NO-dependent reactions, thus are expected to display broad cell signaling capabilities as endogenous nuclear receptor-dependent paracrine signaling molecules with a potency that rivals thiazolidinediones (Fig. 1C). Present data reveal that LNO₂ mediates cell differentiation in adipocytes, CD36 expression in macrophages, and inflammatory-related signaling events in endothelium (e.g., inhibition of vascular cell adhesion molecule-1 expression and function; data not shown) with an important contribution from PPARγ-dependent mechanisms (Fig. 3). Nitroalkene derivatives of fatty acids thus represent a unique class of receptor-dependent cell differentiation, metabolic, and anti-inflammatory signaling molecules that serve to converge NO and oxygenated lipid redox signaling pathways.