NF-κB p65, p50, and p105 subunits antibodies and Ik-β and anti-rabbit secondary antibody were obtained from Cell Signaling Technology (MA, USA). The β-actin, anti-ubiquitin, anti-biotin, PSMD2, PSMD3, PSMD11, 20S Proteasome β1, 20S Proteasome α1, polyclonal anti-mouse secondary antibodies, and Protein A/G plus agarose beads were from Santa Cruz Biotechnology (CA, USA). Anti-PSMD1 and anti-TBP antibodies were from Sigma-Aldrich (Dorset, UK).

Human aortic endothelial cells were obtained from Cascade Biologics – Invitrogen cell culture (Carlsbad, CA, USA) and were grown in EC culture media MV plus growth supplements from Promo Cell (Heidelberg, DE, USA) supplemented with 100 units/ml penicillin and 100 μg/ml streptomycin. For experiments, 80% confluent ECs were incubated in EC basal media from Promo Cell (2% fetal calf serum, 0.4% EC growth supplements, 90 μg/ml heparin and with 100 U/ml penicillin, and 100 μg/ml streptomycin) for 24 h before treatments. THP-1 monocytes were purchased from the LGC Promochem (Middlesex, UK) and were maintained in RPMI 1640 media supplemented with 10% fetal bovine serum, 1% l-glut, and 1% P/S. The cells were passaged twice weekly in complete media at a ratio of 1:5. All cell lines were maintained in an incubator at 37°C and 5% CO₂. For cell counts and viability determination, the cells were diluted 1:1 with trypan blue and counted using a hemocytometer. 15-Deoxy-Δ^12,14-prostaglandin J₂, biotin-15-deoxy-Δ^12,14-prostaglandin J₂, prostaglandin D₂, and biotin-prostaglandin D₂ were purchased from Cayman Chemical (Ann Arbor, MI, USA). To change the solvent for experiments, prostaglandins were dried down and re-suspended in ethanol at a concentration of 1 mg/ml. For experiments, ethanol was used as vehicle control.

Endothelial cells were treated with prostaglandins for 4 h at 37°C. After incubation, the EC were solubilized in lysis buffer (1% Non-idet P-40, 0.1% SDS, 150 mM NaCl, 50 mM Tris–HCl, pH 7.2) and complete protease inhibitor (Roche Diagnostics, Germany). Immunoprecipitation of human plaques proteins was carried out by solubilizing the tissues in T-PER Tissue Protein Extraction Reagent (Thermo Scientific, IL, USA) and complete protease inhibitor (Roche Diagnostics, Germany) and subsequent homogenization in a tissue lyser. The samples were centrifuged (12,000 × g for 10 min at 4°C) and 300 μg of total cell extract was diluted to 500 μl with lysis buffer. The sample was incubated in 2 μg of antibody at 4°C for 2 h and 25 μl of Protein A/G plus agarose beads added to the sample prior to incubation overnight at 4°C. The immunoprecipitates were washed four times with lysis buffer and proteins were eluted with SDS sample buffer (63 mM Tris–HCl, 10% glycerol, 2% SDS, 0.0025% bromophenol blue, pH 6.8) and analyzed by SDS-PAGE, followed by Western blotting. Antibodies to PSMD2, PSMD3, PSMD11, 20S Proteasome α1, and 20S Proteasome β1 were obtained from Santa Cruz Biotechnology (CA, USA) and used for immunoprecipitation experiments.

Protein samples of interest were run on 4–20% polyacrylamide gels and subjected to electrophoresis. The gels were electrophoresed in Tris–HEPES running buffer (0.1M Tris, 0.1M HEPES, 3 mM SDS, pH 8) using a Mini-PROTEAN 3 Electrophoresis System (Biorad) at 80 V for 15 min and 120 V for the duration of the run. SDS-PAGE separated proteins were transferred to nitrocellulose membrane in transfer buffer (25 mM Tris, 192 mM glycine, 20% v/v MeOH). Transfer was carried out for 90 min at 80 V on ice. The membranes were stained with the reversible stain Ponceau-S solution to check transfer efficiency and destained in distilled water. The membranes ware then blocked for 2 h in 5% BSA in TBS-T (20 mm Tris, pH 7.6, 150 mm NaCl, 0.1% Tween 20) at room temperature on an orbital shaker. The membranes were subsequently incubated with the appropriate dilution of primary antibody overnight. Following incubation, the membranes were washed three times in TBS-T and incubated with the appropriate dilution of secondary antibody for 1.5 h. After three washes in TBS-T for 15 min, the membranes were developed using Supersignal enhanced Chemiluminesence reagents (Thermo Scientific, IL, USA) and exposed to photographic X-ray film.

The 358 proteins covalently modified by 15d-PGJ₂, in EC (27), were subjected to pathway mapping analysis and were distributed into categories according to their cellular component, molecular function, and biological process using Gene Ontology (GO), a web-based platform available for interpreting set of GO hierarchies and the PANTHER classification system.¹ The identified proteins were subjected to pathway mapping analysis using the KEGG pathway database (DAVID Bioinformatics Resources 6.7). To filter the results, a threshold of two proteins per pathway and a significance level of p < 0.01 were applied.

The cell viability of EC treated with 15d-PGJ₂ or MG132 was evaluated by MMT assay. The following dose–response was applied for 15d-PGJ₂: 0, 1, 5, 10, and 25 μM. The cell viability after treatment with MG132 was assayed at 100 nM (the dose used for experiments). After 18 h of treatment, 5 mg/ml of dimethylthiazol-diphenyltetrazolium bromide (MTT) (Sigma-Aldrich, Ireland) solution was added to EC and incubated at 37°C, 5%CO₂ for 3.5 h. Following cell lysis with DMSO, absorbance was spectrophotometrically read at 570 nm. For experiments, 15d-PGJ₂ doses which did not significantly reduce cell viability were chosen (97.2 ± 4% viability with 1 μM 15d-PGJ₂, 95.4 ± 4% with 5 μM, and 90.5 ± 6% with 10 μM), while EC viability was significantly decreased at 72.7 ± 8% with 25 μM 15d-PGJ₂. EC viability treated with 100 nM MG132 was 93.5 ± 4%.

Lysates of EC treated with 15d-PGJ₂ or MG132 for 18 h were used to measure the chymotrypsin-like activity of the proteasome using the peptide substrate SUC-LLVY-AMC and 50 μg of proteins per assay. A commercially available kit (Cayman Chemical, Ann Arbor, MI, USA) was used to perform the assay according to the manufacturers’ guidelines. AMC hydrolysis was measure in a Spectramax M2 (Molecular Devices, CA, USA) plate fluorescence reader with 360-nm excitation and 460-nm emission wavelength. The proteasome inhibitor epigallocatechin gallate (EGCG) was used as control to demonstrate specificity of the substrate in this assay. The experiments were repeated in triplicate.

A high-binding polystyrene microplate (R&D, MN, USA) was coated overnight at 4°C with 40 μg/ml of purified 19S Proteasome (Boston Biochem, MA, USA) in PBS. After incubation, the plate was washed four times with 0.05% Tween 20 in PBS and 100 μl solution of biotinylated-15d-PGJ₂ or biotinylated-15d-PGD₂ in PBS was added to the wells and incubated for 2 h at RT with gentle shaking. A binding competition assay was performed to determine the binding specificity of biotinylated-15d-PGJ₂ to the proteasome proteins. Specifically, the 19S-coated plate was pre-incubated with 10- and 20-fold molar excess of 15d-PGJ₂ for 2 h at RT and subsequently with biotinylated-15d-PGJ₂ for a further 2 h at RT. After incubation, the plate was washed four times with 0.05% Tween 20 in PBS. After washing, a blocking solution of 4% BSA in PBS was subsequently added for 2 h at RT with gentle shaking. Subsequently, a solution of 1:750 (v:v) streptavidin–horseradish peroxidase conjugate (GE Healthcare, USA) was incubated for 1 h at RT. After washing four times with 0.05% Tween 20 in PBS, 100 μl of substrate reagent solution (R&D, MN, USA) was incubated for 30 min at RT with gentle shaking. The reaction was stopped with 25 μl of 2M sulfuric acid, and the absorbance was read at 450 nm with correction at 540 nm in a Spectramax M2 (Molecular Devices, CA, USA) plate reader.

Endothelial cells were grown on cover slips for fluorescence microscopy and treated with 15d-PGJ₂ or MG132 for 18 h and subsequently stimulated with 0.2 ng/ml TNF-α for 30 min. The cells were fixed with formaldehyde (3%) for 15 min followed by blocking in 5%BSA in PBS for 1 h. EC were then incubated with the appropriate dilution of primary antibody followed by incubation with fluorescent secondary antibody (Alexa Fluor 488, Invitrogen). Subsequently, the cells were stained with Alexa Fluor 568-phalloidin to stain actin microfilaments and with DAPI to visualize nucleic acid. The stained cells were analyzed by using a Zeiss AxioImager M1 fluorescent microscope, and the images were captured using an Olympus digital camera (Optronics, Goleta, CA, USA). Pictures were processed using AxioVision software.

For static adhesion assay, 4 × 10⁴ EC were seeded in a 96-well plate for 24 h and then treated with 15d-PGJ₂ or MG132, for 18 h. Subsequently, EC were stimulated with TNF-α (0.2 ng/ml) for 6 h. After treatments, the wells were washed three times with medium, and 1 × 10⁵ fluorescein-labeled THP-1 monocytes were added to each well and incubated for 30 min at 37°C. After incubation, the wells were washed three times with fresh medium and adherent monocytes were measured in a Spectramax M2 (Molecular Devices, CA, USA) plate fluorescence reader with 485-nm excitation and 530-nm emission wavelength. For photomicrographs, fluorescence-labeled adherent monocytes in the 96-well plate were analyzed using a Zeiss AxioImager M1 fluorescent microscope. Adhesion assay experiments were repeated in triplicate, and the average value was expressed as a percentage of vehicle control.

THP-1 monocytes were applied to 24-well transwell migration plates (Invitrogen, USA), consisting of upper and lower chambers, separated by membranes punctuated with pores of 5 μm in diameter. The 1 × 10⁵ monocytes were seeded in a final volume of 150-μl basal medium into the upper chamber. Conditioned medium was prepared from confluent EC treated with 15d-PGJ₂ or MG132 for 18 h; the cells were washed, fresh medium was added, and then the cells were treated with TNF- α (0.2 ng/ml for 6 h). After incubation, conditioned medium was cleared by centrifugation (14,000 × g for 10 min at 4°C). Conditioned medium was added to the lower chamber in a final volume of 500 μl as a chemoattractant. Plates were incubated for 2 h at 37°C. Membranes were washed with PBS and were placed into 4% formaldehyde for 15 min to fix cells adherent to the underside of the membrane. Membranes were washed with PBS, and the upper chambers immersed in DAPI solution for 3 min, for nucleic acid staining. Following two PBS washes, the adherent cells were visualized using a Zeiss AxioImager M1 fluorescent microscope. The number of cells in five random 20× fields was counted, and the average value was expressed as a percentage of control.

Following 15d-PGJ₂ or MG132 treatments (18 h) and activation with TNF- α (0.2 ng/ml for 6 h), the EC were harvested to be stained for flow cytometric analysis. To block non-specific binding, EC were incubated in 2% BSA in PBS (blocking buffer) for 15 min before adding the antibodies. FITC-VCAM-1, PE-ICAM-1, and APC-E-selectin antibodies (BD, Rankling Lakes, NJ, USA) were used to label EC in 2% BSA in PBS. The antibodies were incubated for 30 min at room temperature. Following two washes with blocking buffer, the cells were fixed in 2% paraformaldehyde. Forward and side scatter gates were established to exclude non-viable cells and cell debris from the analysis. The mean fluorescence intensity of 2 × 10⁵ cells was analyzed in each sample. Auto-fluorescence signals generated by unlabeled cells were used as negative controls in each experiment. Flow cytometric analysis was performed on an Accuri C6 instrument and analyzed with CFlow^® Software (Accuri, Ann Arbor, MI, USA). The data were expressed as median fluorescence intensity of three independent experiments.

The data were analyzed by parametric unpaired Student’s t-test and analysis of variance (ANOVA). The Student’s t-test was used to compare the mean of two data sets. ANOVA was used to examine any overall differences between treatments groups. Results were expressed as mean ± SEM. Experimental points were performed in triplicate with a minimum of three independent experiments. A value of p < 0.05 was considered statistically significant.

In our previous work (27), conjugation targets of 15d-PGJ₂ were identified by a chemical proteomic approach where human EC were exposed to 5 μM 15d-PGJ₂ that had been biotinylated at the carboxyl group, leaving the reactive cyclopentenone ring unmodified. A total of 358 proteins complexing with 15d-PGJ₂ were identified by affinity purification and subsequent liquid chromatography–tandem mass spectrometry (LC–MS/MS) (27). The MS data are available in PRIDE proteomics identification database² under accession numbers 27957–27962. Functional annotations were available for 355 of the 358 proteins identified. The analysis showed the proteasome complex to be disproportionally represented (fold enrichment = 8.30, p = 2.20E−08) with 13 proteins in this pathway identified in our dataset. The identified proteasome proteins, together with the corresponding MS identification details, are detailed in Table 1. Validation of several 15d-PGJ₂ targeted proteasome proteins (Rpn1, Rpn2, Rpn3, and Rpn6) was performed by immunoprecipitation from EC treated with 5 μM biotin-15d-PGJ₂ and subsequent anti-biotin Western blot. As shown in Figures 1B,C, only the 19S proteasome proteins were modified by biotin-15d-PGJ₂, while proteins in the 20S proteasome (α1 and β1) were not modified. These results show an extraordinary level of specificity for the interaction of the proteasome with 15d-PGJ₂.

To further examine the covalent modification of 19S proteasome proteins by 15d-PGJ₂, an immunoplate was coated with purified 19S particle of the 26S proteasome and incubated in vitro with biotin-15d-PGJ₂. Biotin labeling of 19S proteasome was observed at nanomolar concentration of biotin-15d-PGJ₂ (Figure 2A) demonstrating direct interaction of 15d-PGJ₂ and proteasome proteins at low levels of the prostaglandin. The specificity of the interaction was confirmed by pre-incubating the coated plate with 10- and 20-fold molar excess of unlabeled 15d-PGJ₂, which prevented the binding of biotinylated prostaglandin, and by incubating 10 μM biotin-15d-PGJ₂ with an irrelevant protein, BSA. In addition, biotin-PGD₂ showed no binding, confirming the specificity of the interaction between 15d-PGJ₂ and the 19S. In our previous work, we demonstrated that covalent modification by 15d-PGJ₂ profoundly altered the biological function of several proteins (14). Here, we explored the effect of 15d-PGJ₂ covalent modification of proteasome components on proteasome catalytic activity and assayed in lysates of EC pretreated with the prostaglandin. The effects were compared to that of MG132, a well-known synthetic proteasome inhibitor which acts by covalently binding to the beta subunits of the proteasome (28). The proteolytic activity of the proteasome was assayed using a fluorogenic substrate (SUC-LLVY-AMC) (Figure 2B). The treatment with 15d-PGJ₂ inhibited the proteasome activity by 16 ± 2% at 1 μM, 36 ± 4% at 5 μM, and 44 ± 7% at 10 μM compared to the vehicle control. For comparison, 100-nM MG132 decreased the proteasome activity by 58 ± 1%. The proteasome system degrades proteins that have been targeted for proteolysis by the addition of ubiquitin molecules. As shown in Figure 2C, anti-ubiquitin Western blot analysis of EC treated with 10-μM 15d-PGJ₂ or with 100-nM MG132 resulted in accumulation of proteins conjugated to ubiquitin, consistent with a decreased proteasomal activity.

The proteasome system plays a critical role in the regulation of the inflammatory process in part by modulating the activation of the NF-κB pathway. In particular, the proteasome regulates two crucial steps of NF-κB activation, IκB-α degradation, and p105 processing, leading to the formation of active p50/p65 dimers. If proteasome activity is inhibited by 15d-PGJ₂, then degradation of both IkB-α and p105 would be prevented, and consequently, NF-κB pathway activation would be suppressed. As shown in Figures 3A,B, treatment of EC with increasing concentrations of 15d-PGJ₂ (1–10 μM) or with the proteasome inhibitor MG132 prevented the proteolysis of IκB-α and p105 induced by TNF-α. It should be noted however that 15d-PGJ₂ inhibits a number of other components of the NF-κB signaling pathway resulting in the accumulation of IκB-α, including IKK and the p50 subunit, the latter by direct covalent modification (19).

In addition, NF-κB is normally held in an inactive state in the cytoplasm through interactions with IκB-α. TNF-α induces IκB-α degradation, freeing NF-κB to translocate to the nucleus and initiate transcription of proinflammatory genes. The nuclear localization of active NF-κB was detected by immunofluorescence experiments using a fluorescein isothiocyanate (FITC)-conjugated p65 antibody (Figure 3C). While in resting EC, p65 showed strong staining in an exclusively cytoplasmic localization, p65 was predominantly localized to the nucleus in the presence of TNF-α. In contrast, in the presence of 15d-PGJ₂ and MG132, nuclear translocation of p65 induced by TNF-α was reduced. These findings were confirmed by Western blotting of nuclear fractions obtained from TNF-α-activated EC, which showed that 15d-PGJ₂ inhibited the nuclear localization of p65 and p50 induced by TNF-α, as did the synthetic proteasome inhibitor, MG132 (Figures 3D,E). In summary, in TNF-α-activated EC, treatment with 15d-PGJ₂ resulted in the inhibition of the NF-κB pathway activation, by modulating the processing the NF-κB inhibitors, similarly to the effects of the proteasome inhibitor MG132.

Inhibition of the nuclear translocation of p65 and p50 by 15d-PGJ₂ would be expected to suppress the expression of NF-κB target genes induced in response to TNF-α, including adhesion molecules and chemokines. The effect of 15d-PGJ₂ on TNF-α-induced expression of adhesion molecules was determined in EC activated by TNF-α for 6 h (Figure 4A). TNF-α markedly increased mRNA expression of VCAM-1, ICAM-1, and E-selectin compared to vehicle control (p < 0.001). Pretreatment with either 15d-PGJ₂ or MG132 significantly suppressed the expression of mRNA for VCAM-1, ICAM-1, and E-selectin. These results were confirmed by flow cytometry (Figure 4B), which showed a dose-dependent decrease of EC surface protein expression of VCAM-1, ICAM-1, and E-selectin.

The induction of ICAM-1, VCAM-1, and E-selectin, among others, facilitates the rolling and firm adhesion of monocytes to the EC surface of the vessel wall, a key process in vascular inflammation. Having confirmed the ability of 15d-PGJ₂ to reduce adhesion molecule expression in EC treated with TFN-α, further experiments were undertaken to examine the effect of 15d-PGJ₂ on monocyte adhesion to TFN-α-treated EC. As shown in Figures 5A,B, TNF-α caused a significant increase in the adhesion of human monocytes compare to vehicle (~19-fold, p < 0.001). Pretreatment with 15d-PGJ₂ inhibited the adhesion of monocytes on TNF-α-activated EC (by 11 ± 6, 89 ± 2, and 87 ± 4% with 1, 5, and 10 μM 15d-PGJ₂, respectively). Similarly, MG132 decreased adhesion of monocytes by 90 1%, consistent with a role for the proteasome in regulating the pathways leading to monocyte adhesion.

Monocytes may be recruited to the vascular endothelium in early atherosclerosis through chemotactic factors such as MCP-1, MCP-4, and IL-8 that are released in response to activation of NF-κB pathway. As shown in Figure 6A, conditioned medium from 15d-PGJ₂-treated EC that had been exposed to TNF-α promoted monocyte chemotaxis and this was suppressed by prior incubation of the EC with 15d-PGJ₂ in a concentration-dependent manner. The proteasomal inhibitor MG132 showed similar effects, inhibiting moncoyte chemotaxis by the conditioned medium by 78 ± 2%. Furthermore, as shown in Figure 6B, both 15d-PGJ₂ and MG132 significantly decreased EC generation of MCP-1, MCP-4, and IL-8 in the conditioned media assayed by ELISA, explaining, at least in part, their inhibition of monocyte chemotaxis.