Arginase I Induction by Modified Lipoproteins in Macrophages: A Peroxisome Proliferator-Activated Receptor-γ/δ-Mediated Effect that Links Lipid Metabolism and Immunity

OxLDL Induces Arginase I in Bone-Marrow-Derived Macrophages

The metabolic fate of l-arginine in macrophages between NOS and arginase has been demonstrated to be regulated by the activation stage of the cell (16). In the context of atherosclerosis, the internalization of modified lipoproteins inhibits the inducible form of NOS (17), whereas in endothelial cells, arginase is up-regulated in response to oxLDL (14). To analyze the regulation of arginase isoforms in in vitro lipid-loaded cells, we measured arginase activity in macrophages treated with increasing concentrations of oxLDL and acetylated LDL (acLDL) (Fig. 1A). Both modified lipoproteins were able to increase arginase activity in a dose-dependent manner, with 30 μm and 24 h treatment being the optimal conditions (Fig. 1B). The response was specific for modified LDL because native LDL (nLDL) did not increase arginase levels significantly. Resting bone marrow-derived macrophages (BMDM) constitutively express arginase I and II (Fig. 1D). Modified lipoproteins selectively induced arginase I mRNA (Fig. 1E) and protein (Fig. 1D).

It is well documented that uptake and internalization of oxidized lipoproteins activate PPARγ expression and activity in macrophages (18). Thus, to better understand the regulation of arginase I by modified lipoproteins, we treated lipid-loaded cells with the PPAR antagonist GW9662 (19). Our results show that pretreatment of macrophages with 1 μm GW9662 potently reduced the induction of arginase I activity by oxLDL and acLDL (Fig. 1C) and also blocked the induction of ArgI protein (Fig. 1D) and mRNA (Fig. 1E). Moreover, the induction of the enzyme by oxLDL was also inhibited by 1 μm of each, GW5393 (20) and T0070907 (21), two structurally different PPAR antagonists (supplemental Fig. 1, published as supplemental data on The Endocrine Society’s Journals Online web site at http://mend.endojournals.org).

Although the PPAR antagonists used here are known to be specific for PPARγ, at the concentrations used in this work (1 or 2 μm), they are inhibiting all PPAR isoforms, according to the references cited above. The mode of action of GW9662 and T0070907 is to covalently modify a cysteine residue in the ligand-binding domain of PPAR (19, 21), whereas GW5393 was designed as a PPAR ligand unable to recruit the coactivator cAMP response element binding protein-binding protein (CBP) to the receptor (20). Because murine macrophages constitutively express more than one PPAR isoform, we used these molecules as pan-antagonists to be sure that the induction of arginase I by modified LDL were dependent on PPAR activation.

Collectively, these results strongly suggest that the induction of arginase I in foam cells is mediated through a PPAR-dependent mechanism. Indeed, under the same experimental conditions, modified lipoproteins triggered the expression of PPARγ and others known PPAR target genes such as CD36, AP2 (22), or ADRP (adipose differentiation-related protein) (Fig. 1F). These genes have been found to be induced by oxLDL and up-regulated in atherosclerotic lesions (23, 24).

It is important to note that in Fig. 1, A and B, the levels of arginase-specific activity were calculated as a function of protein concentration. We paid special attention to this aspect because several reports have demonstrated that oxLDL induces macrophage proliferation (25). In our experimental conditions, we did not observe changes in protein concentration that could significantly affect enzyme-specific activity.

Arginase I Is Selectively Induced by PPARγ and -δ Agonists

To identify the PPAR isoforms responsible for arginase I expression, we next treated primary macrophages with different PPAR agonists. The results are presented in Fig. 2. Interestingly, the natural activators of PPAR, such as the eicosanoid 15dPGJ2 and the oxidized derivative of linoleic acid, (13-HODE), dose-dependently induced arginase I (Fig. 2A). We also used the thiazolidinedione rosiglitazone (Fig. 2B), a potent PPARγ synthetic activator and a widely used insulin sensitizer (26). This drug also induced arginase I and showed maximal effects at concentrations above 1 μm. Finally, GW7845 and GW1929, two non-thiazolidinedione, potent synthetic PPARγ agonists, also significantly increased arginase activity and expression.

We also treated cells with PPARδ and PPARα agonists under the same experimental conditions. Remarkably, the specific PPARδ agonist GW0742 (27) induced arginase activity very efficiently, showing significant induction at 25–100 nm concentrations (Fig. 2C). In contrast, treatment with low micromolar concentrations of two different PPARα ligands (GW7647 and WY14643) did not result in significant changes in arginase activity (data not shown). These results are consistent with the minor expression of PPARα in mouse macrophages, in agreement with previous reports (28). Both RNA expression levels measured by real-time quantitative PCR (Fig. 2D) and protein expression (Fig. 2E) confirmed the induction of arginase I by PPARγ and -δ ligands, whereas the two PPARα activators did not increase protein or mRNA levels. Together, these results demonstrate that arginase I expression and activity are specifically modulated by PPARγ and -δ activators in primary macrophages.

The Induction of Arginase by PPARγ and -δ Ligands Is Increased by RXR Agonists

Because PPAR/RXR heterodimers can be activated through both the PPAR and the RXR arms of the complex, we analyzed arginase activity using the RXR ligand 9-cis-retinoic acid (9-cRA), in the presence or absence of different PPAR agonists. The data of Fig. 3 demonstrate an additive effect on arginase activity by cotreatment with PPARγ and -δ activators (Fig. 3, A and B). On the other hand, cis-RA alone also induced arginase activity at concentrations from 50 nm (Fig. 3D). As a control, we verified that PPAR ligands increased the expression of the known target genes CD36, AP2, and ADRP (24, 29, 30) by real-time PCR (Fig. 3E).

Arginase I Induction Is Inhibited by PPAR Antagonists and Reduced in Cells from PPARγ and PPARδ Knockout Mice

To confirm that the induction of arginase I expression by PPAR agonists was receptor dependent, we used two different experimental approaches. First, arginase activity and protein levels were significantly inhibited by two structurally unrelated PPAR antagonists, GW9662 and GW5393 (Fig. 4, A and B). Second, we analyzed macrophages with conditional disruption of PPARγ and macrophages from PPARδ-deficient mice. Macrophages from C57BL/6 wild-type (WT) mice showed increased mRNA expression of arginase I when treated for 12 h with 100 nm PPARγ agonist GW7845 or with 100 nm GW0742 (Fig. 4C). However, arginase I induction was completely abrogated in cells from either PPARγ0 or PPARδ-deficient mice. Moreover, because arginase I is induced by IL-4 in murine macrophages, we used this Th2-derived cytokine as a positive control in all our experiments.

It has also been documented that IL-4 induces PPARγ expression in macrophages (31). Therefore we checked whether arginase induction by IL-4 could be inhibited by PPAR antagonists. Pretreatment with either GW9662 or GW5393 resulted in more than 50% inhibition of enzyme activity (supplemental Fig. 1). As expected, expression of arginase I mRNA by IL-4 was highly increased by cotreatment with GW7845. Furthermore, the synergistic effect of GW7845 on IL-4-induced arginase I expression was completely suppressed in PPARγ-deficient macrophages.

Together, these results demonstrate that arginase I is specifically induced by activation of PPARγ and -δ isoforms in macrophages.

PPAR Ligands Promote L. major Growth in Macrophages in an Arginase-Dependent Manner

The mouse model of L. major infection between the susceptible BALB/c and resistant C57BL/6 strains is one of the best characterized examples in which macrophage arginase I, induced in the context of a predominant Th2 response in susceptible mice, is triggering the growth of parasites inside macrophages (32). To determine the functional impact of PPAR-dependent arginase regulation for immune responses, we treated Leishmania-infected BMDM with PPAR ligands. The results presented in Fig. 5 clearly demonstrate that both GW7845 and GW0742, when added in vitro to infected macrophages, significantly increased the growth of intracellular Leishmania amastigotes (Fig. 5, A and B, and supplemental Fig. 2). Moreover, Leishmania growth was prevented by pretreatment with 2 μm PPAR antagonist GW5393 or by adding N^ω-hydroxy-nor-l-arginine (nor-NOHA), the specific arginase inhibitor, together with either the PPARδ agonist GW0742 (Fig. 5, A and B) or PPARγ ligand GW7845 (supplemental Fig. 2). We also show the potent induction of parasitic growth achieved in the presence of IL-4 as a positive control for arginase I induction. Intracellular amastigote growth was strictly correlated with the levels of macrophage arginase I activity. In cells cotreated with PPAR/RXR ligands (supplemental Fig. 2), the amount of intracellular parasites was similar to those infected in the presence of IL-4, and both effects were reverted by Nor-NOHA.

Together, these results establish a role for PPARγ and PPARδ in arginase I expression and alternative macrophage activation.

Reagents and Chemicals

Lipoproteins, oxLDL, acLDL, and nLDL, were from Biomedical Technologies, Inc. (Stoughton, MA) and used according to the manufacturer’s instructions. The RXR agonist 9-cRA; PPARα agonists WY14643 and GW7647, PPARγ agonist GW1929, the PPAR antagonist GW9662 (2-chloro-5-nitrobenzanilide), and the natural PPAR ligands 15dPGJ2 and 13-HODE were purchased from Sigma-Aldrich (Madrid, Spain), whereas the PPARδ agonist GW0742, PPARγ agonist GW7845, and the PPAR antagonist GW5393 (1,3,4-oxadiazole) were provided by Tim Willson and Jon Collins (GlaxoSmithKline, Research Triangle Park, NC). The PPAR antagonist T0070907 (2-chloro-5-nitro-N-4-pyridinyl-benzamide) and rosiglitazone were purchased from Cayman Chemical (Ann Arbor, MI). PPAR and RXR ligands were dissolved in dimethylsulfoxide before use. Recombinant murine IL-4 was obtained from PeproTech, EC Ltd. (London, UK) and used in complete culture media at a concentration of 2.5 ng/ml.

Cell Culture

BMDM were obtained by flushing the femurs from 6-wk-old BALB/c mice (Harlam Iberica, Barcelona, Spain) and cultured in hydrophobic Teflon bags (Cell Genix, Freiburg, Germany) in DMEM containing 10% inactivated fetal calf serum, 2 mm l-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin (all from Sigma, Spain), and 15% of L929 supernatant as a source of macrophage colony-stimulating factor.

Measurement of Arginase Activity

Arginase activity was measured in macrophages lysates as previously described (43). One unit of enzyme activity is defined as the amount of enzyme that catalyzes the formation of 1 μmol urea/min.

RNA and Protein Analysis

RNA was isolated using TRIzol reagent (Sigma-Aldrich, Spain). Reverse transcription was made by using the high-capacity cDNA archive kit (Applied Biosystems, Foster City, CA) according to the manufacturer’s protocol. TaqMan or SYBR Green real-time quantitative assays were performed using an Applied Biosystems 7500 sequence detector system. The relative expression of arginase I (Arg1) and arginase II (Arg2) genes were analyzing by using TaqMan Gene Expression Assays, and the relative expression in each experiment was compared by the 2^−ΔΔCt calculation method as described (44). The culture conditions used in this study did not alter the expression of 18S rRNA, thus validating its use as endogenous control. Primers and probes for arginase I and II and 18S rRNA were obtained from Applied Biosystems.

The rest of the genes were analyzed using SYBR Green gene expression assays. In these experiments, each sample was run in duplicate and was normalized to 36B4, as previously described (45). Primers and probes are published as supplemental data on The Endocrine Society’s Journals Online web site at http://mend.endojournals.org (supplemental Table 1).

Protein expression was assayed by Western blot using macrophage lysates with the same protein content (assayed by the Bradford method; Bio-Rad, Spain). Briefly, Proteins were separated by 10% SDS-PAGE, transferred to a polyvinylidene difluoride membranes, and then incubated with anti-arginase I (1:1000; BD Transduction Laboratories, Lexington, KY), anti-arginase II (1:5000; Santa Cruz Biotechnology, Santa Cruz, CA), or anti-β-Actin (1:3000; Sigma) which served as loading control. Goat antirabbit IgG-horseradish peroxidase secondary antibody (1:10000; Santa Cruz Biotechnology) was used against arginase II and β-actin, and goat antimouse IgG (1:3000; Pierce, Rockford, IL) were used against arginase I. Detection was made with and ECL-Plus western blotting detection system (Amersham Biosciences, Uppsala, Sweden).

Leishmania Infection

L. major promastigotes (MHOM/IL/80/Friedlin) were cultured as previously described (31). Cells were cultured onto round slides inside 24-well plates. PPAR antagonists were added to macrophage cultures 24 h before infection, and PPAR agonists were added 2 h before the infection. Then, stationary-phase promastigotes were added to BMDM at a parasite to cell ratio of 3:1. After 24 h infection, slides were mounted and stained with Diff-Quik (Panreac, Barcelona, Spain). Results are presented as the number of intracellular amastigotes per 100 cells.

Arginase I Expression in Macrophages from PPARγ- and PPARδ-Deficient Mice

Primary peritoneal macrophages were elicited from thioglycollate-injected mice. Cells were derived from floxed PPARγ-MxCre^−/+ mice and PPARδ^−/− mice (C57BL/6 background). The PPARγ fl/fl mice were provided by Ron Evans, and the PPARδ^−/− mice were provided by Frank Gonzalez. Mice were maintained on standard chow diet in pathogen-free conditions. The cells were flushed and cultured in DMEM with 10% fetal bovine serum (Omega Scientific, Tarzana, CA) and 1% penicillin/streptomycin (Cellgro, Lawrence, KS). BMDM were derived by flushing cells from the femur and tibia of floxed PPARγ-MxCre^−/+ and PPARδ^−/− mice. Cells were then differentiated using DMEM, 20% fetal bovine serum, 1% penicillin/streptomycin, and 30% L929-conditioned media. Cells were activated by ligand for 12 h before harvesting total RNA using TRIzol reagent (Invitrogen, Carlsbad, CA). Alternatively, cells were pretreated with 2.5 ng/ml IL-4 for 12 h before activating with ligand for 24 h. Total RNA was harvested after 24 h using TRIzol reagent (Invitrogen). Expression of arginase I was analyzed by quantitative real-time PCR (SYBR Green) using an Applied Biosystems 7900 sequence detector.

Statistical Analysis

All data were processed by using GraphPad Prism 4.0 software and are the result of at least three independent experiments performed in triplicate. Errors bars were calculated as sd of the data, and when necessary, we have analyzed significance by the Student’s t test (paired t test).

NURSA Molecule Pages:

• Ligands: 15-Deoxy-Δ12 | 9-cis-Retinoic acid | GW 1929 | GW 7647 | GW 9662 | GW0742X | Pirinixic acid | Rosiglitazone;

• Nuclear Receptors: PPARα | PPARγ | PPARδ.