The THP-1 cell line is widely used as a model for the investigation of macrophage biology. These are suspension cells which could be induced to differentiate in the presence of the protein kinase C activator, phorbol ester (PMA), after treatment with which they became adherent to the culture substratum. This process was accompanied by an up-regulation in the PPARδ mRNA (Fig. 1A) beginning within 2 h of the PMA stimulus and reaching a peak at 24 h. This elevation in the mRNA was accompanied by an increase in protein expression (Fig. 1B).

In order to investigate the effect of PMA on PPARδ signalling, we employed a transient transfection approach. COS-1 cells were transfected with a luciferase reporter plasmid containing 4 copies of the liver fatty acid binding protein PPRE in front of a minimal TK promoter and a plasmid directing the expression of full length PPARδ . The ability of increasing concentrations of PMA to activate PPARδ was tested in a dose response (Fig. 1C) and the results indicated that phorbol ester is a very effective activator of the receptor, achieving a 4–5 fold maximal induction compared to the vehicle-only control at approximately 1 nM PMA. This appeared to be due entirely to activation of PPARδ as no increase in PPARδ protein was observed upon treatment of the COS-1 cells with PMA (data not shown). Transient transfection experiments (Fig. 1D) using a chimeric receptor in which the Gal4 DNA-binding domain is fused in frame to the ligand binding domain of PPARδ demonstrated activation by PMA. Thus, activation of the UAS reporter construct therefore reflects PMA signalling mediated by the ligand-binding domain of PPARδ . This is in agreement with the results of Tan et al. 29

As PMA appears to modulate both expression and activity of PPARδ, we investigated the dynamics of the THP-1 response to phorbol ester and the PPARδ agonist, compound F. (Fig. 2). PMA induced THP-1 cells to differentiate in a dose-dependent manner, with an EC₅₀ of approximately 3 nM. Co-treatment of the cells with 100 nM compound F resulted in a reduced EC₅₀ of 1 nM PMA (Fig. 2A). If THP-1 cells were treated with low levels of PMA (>1 nM), very few were seen to differentiate, unless simultaneously treated with compound F, when the number of cells undergoing differentiation and adhering to the culture substratum was seen to increase significantly (p < 0.0001) in a compound F dose-dependent manner (Fig. 2B), with maximal activity at approximately 10 nM. PMA-induced differentiation was dependent on continued PMA signalling: when cells were treated with a single dose of PMA (1 nM), and the PMA was subsequently withdrawn 48 h later, the cells began to de-differentiate and detach from the substratum. After 8 days, less than 20% of the initial total remained adherent. When the cells were exposed to 20 nM compound F treatment following the PMA withdrawal, the differentiated phenotype was retained for significantly longer (p < 0.0001), with some 50% of cells still attached (Fig. 2C). These results indicate that PPARδ activity is required for both the initiation and maintenance of adhesion. However, despite the fact that compound F treated THP-1 cells showed acute sensitivity to very low doses of PMA, compound F alone was not sufficient to induce differentiation indicating that other factors in addition to expression and ligand activation of PPARδ are involved in the induction of macrophage differentiation.

Cells were treated with increasing doses of PMA for up to 48 h with medium containing 0.5% FCS, in the presence (broken line) or absence (solid line) of compound F. Cells were lyzed and cDNA prepared for QPCR analysis. Several gene targets were analyzed for their known role in macrophage differentiation. CD68 (Fig. 3A) is a classical marker of macrophage differentiation. In the absence of compound F there was a slight increase in CD68 mRNA with increasing concentrations of PMA. In the presence of 20 nM compound F, there was a profound increase in message in response to PMA treatment. Similar effects were observed with IL8 (Fig. 3B), a potent macrophage chemoattractant, known to be important in the development of the inflammatory phenotype. Adipophilin (Fig. 3C) and adipocyte fatty acid binding protein (AFABP) (Fig. 3D) both showed a similar synergistic response following treatment with PMA and compound F. Both of these are crucial to lipid trafficking and storage during the differentiation of macrophages into a foam cells and we and others have shown previously that they are targets of PPARδ 2223. Levels of IL1β were not significantly affected (Fig. 3E).

MMP9 (Fig. 3F) is a matrix metalloproteinase important in inflammation. MMP9 facilitates the entry of macrophages into peripheral tissues by allowing them to penetrate the basement membrane underlying the endothelium. Osteopontin (Fig. 3G) is another macrophage differentiation marker. Both of these gene targets showed an up-regulation in message in response to PMA, and this effect was still present in the cells treated with the compound F. In this case, the effect of the PPARδ agonist was to blunt the PMA response.

Stable cell lines over-expressing PPARδ were generated as previously described 22. The effect of this on the PMA-induced differentiation response is shown in Fig. 4A. The PPARδ over-expressing cells differentiated at much lower concentrations of PMA when compared to the parental THP-1 cells. This effect was even greater than that observed with compound F. Over-expression of PPARδ was also able to limit the de-differentiation in the absence of a prolonged PMA stimulus in a manner similar to compound F treated parental cells (p < 0.0001, Fig. 4B).

Stable cell lines were also generated that expressed PPARδ anti-sense mRNA. These cell lines also appeared to be normal with respect to their routine culture characteristics. Fig. 5A shows western blotting of parental, SENSE and ANTISENSE cell lines, both in the presence and absence of PMA. PPARδ expression is undetectable unless the cells are treated with PMA, while untreated SENSE cells show a level of expression similar to that observed with the PMA-treated parental cells. ANTISENSE cells do not express PPARδ even when stimulated with PMA. Figs. 5B and 5C show the response of the cell lines to treatment with 8 nM PMA compared to the wild type. Cells were plated at a density of 5 × 10⁵/ml and left for 48 h, during which time all three cell lines underwent a doubling in number. At 48 h, the cells were treated with a single dose of PMA. This was left on the cells for a further 48 h, after which the medium was removed and replaced with medium without PMA. The cells were left for a further 96 h. Following PMA treatment, both adherent (Fig. 5B) and non-adherent (Fig. 5C) cell number was determined by SYBR staining at 48 h intervals. Fig. 5C shows that no ANTISENSE cells ever became adherent following PMA treatment, indicating that PPARδ is required for the differentiation of THP-1 cells. When the non-adherent cells are considered (Fig. 5B) it can be seen that, for the duration of the PMA treatment, ANTISENSE cells did not proliferate. Once the PMA was removed, the cells rapidly resumed proliferation, indicating that PMA-induced growth arrest of THP-1 cells is PPARδ-independent.

RPMI 1640, DMEM, fetal bovine serum (FBS) and SuperScript II were purchased from Gibco Life Technologies, UK. Phorbol-12-myristate-13-acetate (PMA) and G418 were obtained from Sigma-Aldrich, UK and the PPARδ ligand 3-propyl-4-(3-(3-trifluoromethyl-7-propyl-6-benz- 35-isoxazoloxy)-propylthio)phenylacetic acid (compound F) were a gift from GlaxoSmithKline, UK.

The human THP-1 cell line was obtained from ATCC and grown in RPMI 1640 supplemented with 10% heat-inactivated FBS and 20 μM β-mercaptoethanol. Cells were maintained at 37°C and 5% CO₂. Experiments were performed in 6-well plates where differentiation was induced by resuspending the cells in RPMI containing either 0.5% FBS and PMA as described in the figure legends. Experiments were performed in both the presence, or absence of compound F (either 20 or 100 nM final concentration). All drugs were added with DMSO as a vehicle and were replaced at intervals of 48 hours unless otherwise stated.

At the end of each experiment, adherent THP-1 cells were fixed with a solution of 4% paraformaldehyde, 1% Triton-X for 30 minutes at room temperature, following which the DNA was stained with a 1: 10,000 SYBR green solution (Molecular Probes). Plates were incubated for 24 h at 4°C in the dark. Cells were washed in PBS and the relative numbers determined from a standard curve by fluorimetry using a LabSystems FluoroSkan Ascent FL microplate reader.

PPARδ over-expressing cell lines (SENSE) and THP-1 anti-sense cells (ANTISENSE) were prepared as described previously 22. Briefly, THP-1 cells were transfected with an expression vector containing the entire human PPARδ coding sequence in either sense or anti-sense orientation respectively. Transfection was by a modified DEAE-dextran procedure 36 and pools of stable cells were selected by maintaining cells in medium containing 1 mg/ml G418, 10% FBS and 10% THP-1 conditioned medium, with vigorous washing to remove dead cells. This procedure was repeated until cell killing stopped and robust growth was observed. Six pools from independent transfections were obtained, each pool displayed a similar phenotype.

COS-1 cells were transfected in 6-well plates using DEAE-dextran as described previously 37. The reporter construct used was pFABPLuc, which consists of four copies of the human peroxisome proliferator response element from the human liver fatty acid binding protein (FABP) gene in front of the herpes simplex virus thymidine kinase promoter, cloned immediately upstream of the cDNA encoding firefly luciferase in pGLBAS (Promega). The PPARδ expression vector (pCLDNhPPARδ) contained the coding sequence for human PPARδ under the control of the enhancer/promoter of the human cytomegalovirus. The Gal4/LBD plasmid consisted of an in-frame fusion of the Gal4 DNA-binding domain and the ligand-binding domain of PPARδ . The reporter plasmid without the PPRE (TKLuc) and the empty expression vector (pCLDN) were used as controls. pSV β-galactosidase (Promega) was co-transfected in all cases in order to control for transfection efficiency. Following transfection and recovery, the cells were incubated for 18 h with DMEM containing 10% FBS and 1% streptomycin/penicillin, supplemented with vehicle (DMSO), compound F or PMA. Cells were analysed for β-galactosidase and luciferase activities using assay kits as described by the manufacturer (Promega). Data are presented as the ratio of relative light units obtained with the luciferase assay to the absorbance obtained at 415 nm in the β-galactosidase assays (relative luciferase).

THP-1 wild type, SENSE and ANTISENSE cells were treated with PMA or vehicle for 48 h following which they were lyzed in SDS-PAGE loading buffer and analysed using standard Western blotting procedures. The primary antibody was either PPARδ antiserum (a gift from Dr David Bell), used at a 1:2000 dilution, or an antibody raised against the AB domain (residues 1–100) of PPARδ, used at a dilution of 1:500. The AFAR antiserum (obtained from Prof. John Hayes) was used at a 1:3000 dilution. The secondary detection reagent, a peroxidase conjugated mouse anti-rabbit IgG antiserum (Sigma) was used at a dilution of 1:3000. Bands were visualised using enhanced chemiluminesence (ECL+) as described by the manufacturer (Amersham Pharmacia Biotech).

RNA was extracted from differentiated and undifferentiated THP-1 cells using the Qiagen RNeasy kit. cDNA was synthesised from this RNA using SuperScript II. QPCR analysis was performed on triplicate experiments using the standard Applied Biosystems procedures, with primers and probes described in Table 1. 18S RNA was quantified using an assay from Applied Biosystems.

All graphs and statistics were prepared using Graphpad Prism for the Macintosh v3.0 (Graphpad Inc., San Diego, CA). p values were calculated using two-way ANOVA, unless otherwise stated.