The hinge-helix 1 region of peroxisome proliferator-activated receptor gamma1 (PPARgamma1) mediates interaction with extracellular signal-regulated kinase 5 and PPARgamma1 transcriptional activation: involvement in flow-induced PPARgamma activation in endothelial cells

Abstract

Peroxisome proliferator-activated receptors (PPAR) are ligand-activated transcription factors that form a subfamily of the nuclear receptor gene family. Since both flow and PPARgamma have atheroprotective effects and extracellular signal-regulated kinase 5 (ERK5) kinase activity is significantly increased by flow, we investigated whether ERK5 kinase regulates PPARgamma activity. We found that activation of ERK5 induced PPARgamma1 activation in endothelial cells (ECs). However, we could not detect PPARgamma phosphorylation by incubation with activated ERK5 in vitro, in contrast to ERK1/2 and JNK, suggesting a role for ERK5 as a scaffold. Endogenous PPARgamma1 was coimmunoprecipitated with endogenous ERK5 in ECs. By mammalian two-hybrid analysis, we found that PPARgamma1 associated with ERK5a at the hinge-helix 1 region of PPARgamma1. Expressing a hinge-helix 1 region PPARgamma1 fragment disrupted the ERK5a-PPARgamma1 interaction, suggesting a critical role for hinge-helix 1 region of PPARgamma in the ERK5-PPARgamma interaction. Flow increased ERK5 and PPARgamma1 activation, and the hinge-helix 1 region of the PPARgamma1 fragment and dominant negative MEK5beta significantly reduced flow-induced PPARgamma activation. The dominant negative MEK5beta also prevented flow-mediated inhibition of tumor necrosis factor alpha-mediated NF-kappaB activation and adhesion molecule expression, including vascular cellular adhesion molecule 1 and E-selectin, indicating a physiological role for ERK5 and PPARgamma activation in flow-mediated antiinflammatory effects. We also found that ERK5 kinase activation was required, likely by inducing a conformational change in the NH(2)-terminal region of ERK5 that prevented association of ERK5 and PPARgamma1. Furthermore, association of ERK5a and PPARgamma1 disrupted the interaction of SMRT and PPARgamma1, thereby inducing PPARgamma activation. These data suggest that ERK5 mediates flow- and ligand-induced PPARgamma activation via the interaction of ERK5 with the hinge-helix 1 region of PPARgamma.

FIG. 1.

MEK5-ERK5 activation increases PPARγ1-mediated transactivation of the (PPRE)₃-tk-luciferase reporter construct in HUVECs, which is independent of PPARγ1 S82 phosphorylation. (A and B) MEK5-ERK5 activation induced PPARγ1 transcriptional activity, but DN-ERK5 did not inhibit CA-MEK5α-mediated PPARγ activity. PPARγ1 transcriptional activity was measured by transfection of full-length PPARγ1 and the (PPRE)₃-tk-luciferase reporter construct in HUVECs. PPARγ-mediated transactivation was determined with the transfection of wild-type ERK5 (ERK5a, lane2), empty vector (lane 3), or ERK5 mutants (DN-ERK5 [lane 4] or ERK5b [lane5]) with vehicle (A) or 10 μM ciglitazone (B). Results are the mean ± SD of three to six independent experiments. Luciferase activity of the (PPRE)₃-tk-luc construct with CA-MEK5α and ERK5a in the absence of transfected PPARγ at ciglitazone concentrations of 0 and 10 μM were 0.8 ± 0.2 and 1.1 ± 0.2 (relative PPARγ luciferase activity), respectively. Luciferase activity of the TK promoter alone with CA-MEK5α and ERK5a was below 0.1 relative PPARγ luciferase activity. (C) Activation of MEKK1 inhibited PPARγ activation. CA-MEKK1, as indicated, was transfected in Cos7 cells, and pcDNA3.1 vector was used to provide equal amounts of transfected DNA. Results are the mean ± SD of three independent experiments. (D) ERK5 did not phosphorylate PPARγ1 in an in vitro kinase assay. CHO cells were transfected with vector or CA-MEK5α, and ERK5 was immunoprecipitated with ERK5 antibody. An immune complex kinase assay was then performed with GST or GST-PPARγ1 mutants (GST-PPARγ1-AF-1, GST-PPARγ1-DBD, and GST-PPARγ1-LBD). (E) CA-MEK5α- and/or ciglitazone-induced full-length PPARγ1 wild type or mutants (PPARγ1S82A or PPARγ1S82D) mediated transactivation of the (PPRE)₃-tk-luciferase reporter construct in HUVECs. Results are the mean ± SD of three independent experiments.

FIG. 2.

Endogenous ERK5 associates with endogenous PPARγ at the hinge-helix 1 region of PPARγ1, and the hinge-helix 1 region of the PPARγ1 fragment inhibited the ERK5-PPARγ interaction and CA-MEK5α-mediated PPARγ transcriptional activity. (A) HUVECs were stimulated with 10% serum for 30 min, whole-cell extract was immunoprecipitated with anti-PPARγ antibody or an equal amount of rabbit IgG, and Western blot analysis was performed with anti-ERK5 antibody (top). No difference in the amount of ERK5 (middle) or PPARγ (bottom) was observed in samples by Western blot analysis with anti-ERK5 (middle) or anti-PPARγ (bottom) antibody. (B and C) Association of activated ERK5a with PPARγ1 hinge-helix 1 was tested in a mammalian two-hybrid assay. The activation domain VP16 was fused to wild-type ERK5a and the PPARγ1 deletion mutants. Luciferase activity was normalized relative to the mean luciferase activity of the empty VP16 transfection (white bar; set as 1-fold). Constructs fused to the Gal4 binding domain were cotransfected with the Gal4-responsive luciferase reporter pG5-luc with or without cotransfection of CA-MEK5α in Cos7 cells for 40 h. The total transfected DNA amount was normalized with empty VP16 vector. Results are the mean ± SD of three independent experiments (B). (C) Association of activated ERK5a with PPARγ1 hinge-helix 1 was tested with PPARγ1 hinge-helix 1 truncated deletion mutant (PPARγ1 Δaa202-231) or small fragment of PPARγ1 (aa195-227) with or without CA-MEK5α or VP16-ERK5a, in a mammalian two-hybrid assay. The total transfected DNA amount was normalized with empty VP16 vector. (D) The VP16-PPARγ1(aa 195-227) fragment inhibited coimmunoprecipitation of ERK5 with PPARγ (top). No difference in the amount of PPARγ (middle) or ERK5 (bottom) was observed in samples by Western blot analysis with anti-PPARγ (middle) or anti-ERK5 (bottom) antibody. (E) Cells were transfected with plasmids expressing VP16 or the VP16-PPARγ1(aa 195-227) fragment and 1 μg of (PPRE)₃-tk-luciferase, 0.5 μg of pSG5-PPARγ, and vector to provide equal amounts of transfected DNA, as described for Fig. 1. After 16 h of stimulation with or without ciglitazone, luciferase PPARγ1 transcriptional activity was assayed as described in the legend for Fig. 1. The total transfected DNA amount was normalized with empty VP16 vector. Results are the mean ± SD of three to six independent experiments. (F) Cotransfection of plasmid expressing the VP16-PPARγ1(aa 195-227) fragment did not inhibit CA-MEK5α-induced PPARγ1 activation in HUVECs. The total transfected DNA amount was normalized with empty VP16 vector.

FIG. 3.

Flow-induced PPARγ1 transcriptional activation by the ERK5-PPARγ interaction and ERK5 activation. (A to C) Effect of short-term flow on PPARγ1 activity. At 24 h after transfection, growth-arrested HUVECs were stimulated by ciglitazone (5 μM), and then after 3 h of ciglitazone stimulation HUVECs were exposed for 20 min to flow (12 dynes/cm²) or no flow with or without plasmid expressing the VP16-PPARγ1(aa 195-227) fragment (B) or DN-MEK5β (C), as indicated. After 16 h of ciglitazone stimulation, luciferase PPARγ1 transcriptional activity was assayed as described in the legend for Fig. 1. (D and E) Effects of long-term flow on PPARγ1 and ERK5 activities. (D) Transfection medium contained 2 μg of PPRE reporter plasmid, 1 μg of pSG5-PPARγ, and vector to provide equal amounts of transfected DNA with or without plasmid expressing the VP16-PPARγ1(aa 195-227) fragment. At 48 h after transfection, growth-arrested HUVECs were exposed for 9 h to flow (5 dynes/cm²) or static condition with or without stimulation by ciglitazone (5 μM), as indicated. Luciferase PPARγ1 transcriptional activity was assayed as described in the legend for Fig. 1 after 9 h of ciglitazone or vehicle stimulation. (E) Gal4-ERK5a transcriptional activity was detected as described for Fig. 7a. Results are the mean ± SD of three independent experiments. *, P < 0.05; **, P < 0.01.

FIG. 4.

Flow-inhibited TNF-α-mediated NF-κB activation and VCAM-1 expression by ERK5 and PPARγ activation. (A) HUVECs were transfected with pFR-Luc plasmid and pNF-kBLuc-plasmid. To control transfection efficiency, pRL-TK was transfected as a luciferase control reporter vector. After 24 h of transfection, HUVECs were treated with the following protocol: cells were maintained under static conditions for 20 min followed by vehicle (lanes 1 and 2) or TNF-α stimulation (20 ng/ml; lanes 3 and 4) under the same static conditions with (lanes 2 and 4) or without (lanes 1 and 3) DN-MEK5β transfection, or cells were subjected to flow (shear stress of 5 dynes/cm²) for 20 min followed by TNF-α stimulation (lanes 5 and 6) with (lane 6) or without (lane 5) DN-MEK5β transfection under continuous flow. After 6 h of TNF-α stimulation, luciferase NF-κB transcriptional activity was assayed using the dual-luciferase reporter assay system, and luciferase luminescence was counted in a Luminometer and then normalized to cotransfected luciferase activity as described in Materials and Methods. Results are the mean ± SD of three independent experiments. **, P < 0.01. (B) Effect of long-term flow on TNF-α-mediated VCAM-1 and E-selectin expression. After 24 h of transfection, BLMECs were treated in the following protocol: cells were maintained under static conditions for 60 min followed by vehicle (lanes 1 and 2) or TNF-α (20 ng/ml) stimulation (lanes 3 and 4) under the same static conditions with (lanes 2 and 4) or without (lanes 1 and 3)DN-MEK5β transfection, or cells were subjected to flow (shear stress of 5 dynes/cm²) for 60 min followed by TNF-α stimulation with (lane 6) or without (lane 5) DN-MEK5β transfection under continuous flow. (B, left) After 4 h of TNF-α stimulation, VCAM-1 (upper) and E-selectin (lower) mRNA levels were determined by relative quantitative RT-PCR. 18S rRNA was used as an internal control. (Right) After 6 h of TNF-α stimulation, VCAM-1, hemagglutinin-tagged MEK5β, and β-actin expression were determined by Western blot analysis. (C) Effect of ERK5 and PPARγ activation on TNF-α-mediated VCAM-1 expression. At 24 h after transfection, growth-arrested BLMECs were stimulated with TNF-α (20 ng/ml) with or without plasmid expressing CA-MEK5α or DN-PPARγ1, as indicated. (C, left) After 4 h of TNF-α stimulation, VCAM-1 (upper) and E-selectin (lower) mRNA levels were determined as described for panel B. (Right) VCAM-1, MEK5, PPARγ, and β-actin expression levels were determined by Western blot analysis.

FIG. 5.

Activated ERK5 disrupts the association of corepressor SMRT with PPARγ1. (A) Scheme of the hinge-helix 1 region of PPARγ1. (B) The hinge-helix 1 region of PPARγ1 is critical for CA-MEK5α and ciglitazone-induced PPARγ1-mediated transactivation of the (PPRE)₃-tk-luciferase reporter construct. (C) DN-SMRT increased PPARγ1-mediated transactivation of the (PPRE)₃-tk-luciferase reporter construct. (D and F) Interaction of ERK5a with the hinge-helix 1 region of PPARγ1 disrupts SMRT/PPARγ1. Cos7 cells were transfected with plasmids expressing Gal4, VP16, Gal4-SMRT, ERK5a, or CA-MEK5α with wild-type VP16-PPARγ1-LBD (D) or VP16-PPARγ1-LBD Δaa202-231 (F), as indicated, in a mammalian two-hybrid assay. (E) ERK5 activation by cotransfection of CA-MEK5α and ERK5a inhibited coimmunoprecipitation of SMRT with PPARγ (top). No difference in the amount of PPARγ and SMRT was observed in samples by Western blot analysis with anti-PPARγ or anti-SMRT antibody (middle and bottom).

FIG. 6.

ERK5a binding site of PPARγ1. (A and B) Requirement of ERK5a kinase activity and the COOH-terminal region of ERK5 for the ERK5a-PPARγ1 interaction. (A) Cos7 cells were transfected with plasmids expressing wild-type Gal4-PPARγ1-LBD(aa 202-475), VP16, and CA-MEK5α with VP16-ERK5a, VP16-DN-ERK5, or VP16-ERK5(aa 1-418), as indicated, in a mammalian two-hybrid assay. (B) The middle region of ERK5 (aa 419 to 577) in COOH-terminal region is critical for ERK5a-PPARγ1 interaction. Cos7 cells were transfected with plasmids expressing Gal4-PPARγ1-LBD(aa 202-475), VP16, or CA-MEK5α with several COOH-terminal or NH₂-terminal deletion mutants of VP16-ERK5a, as indicated, in a mammalian two-hybrid assay.

FIG. 7.

Transcriptional activation domains of ERK5a. (A) Cos7 cells were transfected with Gal4-dependent (Gal4-luc) reporter constructs with dominant negative forms of Gal4-DN-ERK5 or ERK5b (upper) and Gal4-ERK5 COOH-terminal deletion mutants (lower). Luciferase activity was measured in unstimulated cells. (B) Cos7 cells were transfected with Gal4-dependent (Gal4-luc) reporter constructs with several COOH-terminal fragments of Gal4-ERK5, as indicated. Luciferase activity was measured in unstimulated cells.

FIG. 8.

Both transcriptional domains and the PPARγ1 binding site in the COOH-terminal region of ERK5 are critical to fully activate PPARγ1. (A and B) Transfection medium contained 1 μg of (PPRE)₃-tk-luciferase, 0.5 μg of pSG5-PPARγ1, and vector to provide equal amounts of transfected DNA. pcDNA3.1-CA-MEK5α (A) and plasmids expressing deletion mutants of the COOH-terminal tail of ERK5a (A) or several VP16-fused truncated mutant fragments of the COOH-terminal region of ERK5a (B) were transfected in HUVECs as indicated, and the pcDNA3.1 vector was used to provide equal amounts of transfected DNA. After 24 h of transfection, growth-arrested HUVECs were stimulated with or without ciglitazone (5 μM). Luciferase PPARγ1 transcriptional activity was assayed as described in the legend for Fig. 1. Results are the mean ± SD of three independent experiments. Expression of full-length and truncated ERK5 was demonstrated by Western blotting with anti-ERK5 antibody (A, right panel). We also performed immunostaining with anti-VP16 antibody and confirmed expression of these constructs in HUVECs (data not shown).

FIG. 9.

Model of the ERK5a-PPARγ1 interaction activating PPARγ1 activity. The position of H12 is regulated by a ligand. In the ligand binding receptor, H12 folds back to form part of the coactivator binding surface. By contrast, H12 inhibits corepressor binding to PPARγ and other nuclear receptors (29). The corepressor interaction surface requires H3, H4, and H5, thereby overlapping the coactivator interaction surface (14). In the present study we found a critical role for the hinge-helix 1 domain in regulating PPARγ1 transcriptional activity. The inactive NH₂-terminal kinase domain of ERK5a partially inhibits the association of PPARγ1 on COOH-terminal ERK5 and also inhibits its transcriptional activity. Following activation, the inhibitory effect of NH₂-terminal ERK5 decreases, and the middle region of ERK5a fully interacts with the hinge-helix 1 region of PPARγ1. The association of ERK5a with the hinge-helix 1 region of PPARγ1 releases corepressor SMRT and induces full activation of PPARγ1.