Relaxin Activates Peroxisome Proliferator-activated Receptor γ (PPARγ) through a Pathway Involving PPARγ Coactivator 1α (PGC1α)^*

Cells and Transfections

The HEK-RXFP1 cells are HEK-293T cells stably expressing RXFP1 described previously (12). For reporter assays, the HEK-RXFP1 cells were transfected using Fugene6 (Roche) as described previously (12). The plasmids included ACO-PPRE-luc (24) provided by Brian Seed (Harvard University), PGC1α-promoter luciferase reporter (25) and plasmids encoding PKI and inactive mutant (PKImut) (26) from Addgene, dominant-negative CREB (KCREB, Clontech, Mountain View, CA), cAMP response element reporter pGL4.29 (CRE-luc), and Renilla luciferase plasmid (pTK-RL) (all from Promega). For cotransfection of siRNA with reporter plasmids, Dharmafect Duo reagent (Invitrogen) was used according to the manufacturer's instructions. The siRNA used were Silencer Select human PGC1α (s21395), PKA (s11065), p38-MAPK (s3585), or nontargeting negative control (Invitrogen).

The LX2 cells are a human hepatic stellate cell line (27), and were provided by Dr. S. L. Friedman (Mount Sinai School of Medicine, New York). The LX2 cells were transfected using a Nucleofector 4D device (Lonza) using reagent S.E. and program EW-113. The ACO-PPRE-luc and pTK-RL plasmids described above were found to be unsuitable for the LX2 cells. Therefore, a new PPRE-luciferase reporter (pGL4–3xPPRE-luc) was constructed. Briefly, a DNA fragment containing three tandem copies of the PPAR response element (28) was inserted between the KpnI and HindIII sites of the pGL4.26 vector (Promega) containing a minimal promoter upstream of the firefly luciferase gene. The Renilla luciferase plasmid used was an SV40-driven construct (pGL4.72, Promega).

Cells were treated with purified porcine relaxin (kindly provided by Dr. O. David Sherwood, University of Illinois at Urbana-Champaign) as indicated in the figure legends. Relaxin was used at a concentration of 1 nm, which is in the range sufficient for activation of canonical cAMP-related pathways through RXFP1 (2). Forskolin, the p38 MAPK inhibitor PD-169316, the PKA inhibitor H89, the PI3K inhibitor LY294002, the ERK1/2 inhibitor PD98059, the nitric-oxide synthase inhibitor L-NAME, and pertussis toxin were purchased from EMD Biochemicals. The cyclic nucleotides Sp-6-Bnz-cAMPS and Sp-8-pCPT-2′-O-Me-cAMPS were from Biolog. The nitric oxide donors SNAP (S-nitroso-N-acetyl-d,l-penicillamine) and GSNO (S-nitrosoglutathione) were from Sigma.

Reporter Assays

Cells were cotransfected with firefly and Renilla luciferase reporter plasmids as described above. After 24 h, cells were treated as described for the individual figures, then assays were performed 48 h after transfection. Activation of reporter constructs was monitored using the Dual-Glo Luciferase assay (Promega) and measured in a Spectramax M5 plate reader. The data are expressed as the firefly luciferase activity normalized to the Renilla luciferase in the same sample.

Quantitative Real-time PCR and Conventional RT-PCR

Total cellular RNA was extracted using the Purelink RNA Mini kit (Invitrogen), and treated with RNase-free DNase to remove genomic DNA. The RNA concentration was measured using the Ribogreen assay (Invitrogen). A total of 2 μg of RNA was reverse transcribed to cDNA using the High Capacity cDNA Reverse Transcription kit (Applied Biosystems, Carlsbad, CA) in a total volume of 20 μl. The cDNA was then subjected to TaqMan real-time PCR. The reaction mixtures were comprised of 2 μl of cDNA (diluted 1:15), 10 μl of Taqman universal PCR master mix, 2.5 μl Taqman primer/probe mix in a final volume of 20 μl per reaction. The human gene expression assays (Applied Biosystems) used were PGC1α (Hs00173304_m1) and TATA binding protein (TBP) (4333769F). Gene expression was normalized to the level of TBP within each sample using the relative _ΔΔCT method. Gene expression is shown as relative expression to control. The data shown are representative of three independent experiments. For conventional RT-PCR analysis, total RNA was extracted and cDNA prepared as above from HEK-RXFP1, THP1, HepG2, and LX2 cells. Expression of RXFP1 was determined using intron-spanning primers for human RXFP1 (sense 5′-TCTTGGTATTAATTTGGCCGC-3′, antisense 5′-CATTAACTCAGGTGGCATCTCC-3′, or GAPDH (sense 5′-ACCACAGTCCATGCCATCAC-3′, antisense 5′-TCCACCACCCTGTTGCTGTA-3′) as a loading control.

cAMP and PKA Assays

The production of cAMP production was measured by treating the cells with the indicated concentrations of relaxin in the presence of the phosphodiesterase inhibitors IBMX (0.5 mm) and Ro20–1724 (0.1 mm) for 30 min at room temperature. The level of cAMP was measured using the cAMP-Glo assay (Promega). The level of PKA activation was determined using the Pep-Tag assay (Promega) according to the manufacturer's instructions.

Western Blots

Cells were treated as described for the individual experiments, using mammalian protein extraction reagent (M-PER, Pierce, Thermo Scientific, Rockford, IL) containing a protease & phosphatase inhibitor mixture (Pierce). The lysates were sonicated briefly to rupture the cell and nuclear membranes. The protein concentration was estimated by the BCA assay (Pierce), and lysates were heated in SDS loading buffer at 95 °C for 5 min. The proteins were resolved on 10% polyacrylamide gels and transferred to PVDF membranes. The membranes were blocked with Odyssey blocking buffer (Li-Cor, Lincoln, NE), then incubated with diluted primary antibody overnight at 4 °C. The primary antibodies used and their dilutions were total p38 MAPK (9228, 1:1000), phospho-p38 MAPK (Thr-180/Tyr-182) (4511, 1:1000), CREB (9104, 1:1000), phospho-CREB (Ser-133) (9198, 1:1000), or PKA (4282, 1:1000), all from Cell Signaling Technology, Danvers, MA), PGC1α (101707, 1:500, Cayman Chemical, Ann Arbor, MI), and GAPDH (MAB374, Millipore, Temecula CA). After a 1-hour incubation with fluorescently tagged secondary a-antibodies (Li-Cor, Lincoln, NE), the membranes were scanned on an Odyssey infrared scanner (Li-Cor).

Statistical Analysis

Curve-fitting and statistical analysis was performed using Prism5 software (GraphPad, La Jolla, CA). Differences were analyzed using one-way or two-way analysis of variance as appropriate, with the Bonferonni post-test. Data are expressed as mean ± S.E. of at least three independent determinations as indicated in the figure legends.

Relaxin Activates PKA to Activate PPAR Activity

One major signaling pathway triggered by relaxin is via induction of cAMP production and activation of cAMP-activated protein kinase (PKA). To examine the role of cAMP/PKA in activation of PPARγ, HEK-293 cells stably expressing RXFP1 (HEK-RXFP1 cells) were used. We previously showed that relaxin increased cAMP production and activation of a PPAR response element (PPRE) via activation of PPARγ in these cells, and that these responses were dependent on RXFP1 (12). Relaxin also increased PKA activity in these cells, which was blocked by treatment with the PKA inhibitor H89 (Fig. 1A). Using a PPRE reporter, treatment of HEK-RXFP1 cells with either forskolin, a direct adenylyl cyclase activator, or relaxin, increased transcriptional activity through a PPAR response element (PPRE) reporter, and in both cases this activity was inhibited by the PKA inhibitor H89 (Fig. 1B). These data confirm that PKA is activated in response to relaxin, and that the cAMP/PKA pathway is involved in activation of PPAR transcriptional activity in these cells.

Because relaxin activates multiple signaling pathways in cells expressing RXFP1, a panel of activators and inhibitors of various pathways was used to screen for potential mechanisms for relaxin-induced PPARγ activation (Fig. 1C). Both forskolin and the PKA-specific activator Sp-6-Bnz-cAMPS caused increased PPRE transcriptional activity, but the exchange protein activated by cAMP (Epac) activator Sp-8-CPT-cAMPS had no effect. The relaxin-stimulated PPRE activation was completely inhibited by H89, and partially inhibited by pertussis toxin and the PI3K inhibitor LY294002. The ERK1/2 inhibitor PD98059 had no influence on the relaxin effect, but the p38 MAPK inhibitor PD169316 partially blocked the effect. The nitric oxide donors (SNAP and GSNO) did not stimulate transcriptional activity, and the nitric-oxide synthase inhibitor L-NAME did not inhibit the ability of relaxin to activate PPARγ, suggesting that the nitric oxide is not a mediator of the relaxin effects.

p38 MAPK Is a Downstream Mediator of PKA-induced PPAR Activity

Because the p38 MAPK inhibitor reduced the relaxin effect (Fig. 1C), this pathway was studied in more detail. The p38 MAPK inhibitor caused a partial decrease in the activation of the PPRE reporter by relaxin (Fig. 2A), but had no significant effect on the baseline levels. To determine whether relaxin causes phosphorylation and activation of p38 MAPK, HEK-RXFP1 were treated with relaxin for various periods of time, and phosphorylation of p38 MAPK was monitored (Fig. 2B). Relaxin caused a rapid and transient increase in p38MAPK phosphorylation that was maximal between 5 and 15 min, with a return to baseline levels with increased time. Furthermore, the relaxin-induced p38 MAPK activation was completely blocked by treatment with H89 (Fig. 2C), suggesting that p38 MAPK is activated downstream of PKA.

CREB Mediates Relaxin-induced PPAR Activity

A common downstream target of both PKA and p38MAPK activation is cAMP response element-binding protein (CREB). Relaxin treatment of HEK-RXFP1 cells stimulated CREB transcriptional activity of HEK-RXFP1 cells transfected with a CREB luciferase reporter (Fig. 3A). Furthermore, overexpression of the dominant-negative CREB (KCREB) resulted in a complete loss of relaxin-induced PPRE activation (Fig. 3B). Treatment with relaxin resulted in phosphorylation of CREB at Ser-133 (Fig. 3C). The phosphorylation of CREB in response to relaxin was completely blocked by a PKA inhibitor, and partially blocked by the p38 MAPK inhibitor. The partial reduction by the p38 MAPK inhibitor of CREB phosphorylation (36.0 ± 2.1%) compares favorably with the decrease in relaxin-stimulated PPRE activity (47.4 ± 2.7%, Fig. 2A). Together, these data suggest that relaxin causes activation of CREB downstream of PKA and p38 MAPK.

PGC1α Is the Mechanism for Activation of PPARγ Activity by Relaxin

We have shown earlier that relaxin activation of PPARγ in HEK-RXFP1 cells occurs by a ligand-independent mechanism (12). The PPARγ coactivator PGC1α is capable of activating PPARγ in the absence of PPARγ ligand, and furthermore is regulated by both cAMP/PKA and p38 MAPK in a number of cells (29, 30). Therefore, the effect of relaxin on the expression level of PGC1α was tested. Relaxin increased the RNA expression level of PGC1α in HEK-RXFP1 cells (Fig. 4A). The increased RNA expression was evident by 2 h of treatment, and started decreasing by 4 h. This increase in expression level was mirrored in cells expressing a reporter containing the PGC1α promoter upstream of luciferase (Fig. 4B). We also examined the PGC1α protein level, and found that relaxin caused an increase in PGC1α protein by 4.5 h of treatment (Fig. 4C). The increased expression of PGC1α by relaxin was diminished by the PKA inhibitor H89 (Fig. 4D). Furthermore, coexpression of the dominant-negative KCREB construct decreased the relaxin-stimulated increase in PGC1α protein levels (Fig. 4E). The involvement of PGC1α in the ability of relaxin to stimulate PPARγ activation was examined using siRNA-mediated knockdown of PGC1α (Fig. 4, F and G). The PGC1α-siRNA reduced the stimulation of PPARγ activity by relaxin, while control siRNA had no effect, suggesting that PGC1α mediates the effect of relaxin on PPARγ. Overall, these results suggest that relaxin increases PGC1α expression through PKA and CREB activation.

Relaxin Regulates PGC1α in Cells Endogenously Expressing RXFP1

The signaling pathways activated by relaxin in HEK-RXFP1 cells were tested in cells naturally expressing RXFP1. In the liver, the myofibroblastic hepatic stellate cells are the predominant RXFP1-expressing cells (31, 32). A human hepatic stellate cell line (LX2) that recapitulates many of the properties of activated hepatic stellate cells (27) was used. Relaxin induced an increase in cAMP levels (Fig. 5A) and activated PKA (Fig. 5B), confirming that LX2 cells are responsive to relaxin. Expression of RXFP1 was detected in LX2 cells at a modest level compared with HEK-RXFP1 cells or THP1 monocytes previously shown to express high levels of RXFP1 (Fig. 5C). Relaxin treatment resulted in an increase in transcription through the PPRE reporter, consistent with activation of PPARγ (Fig. 5D). Similar to the results using HEK-RXFP1 cells, relaxin stimulated phosphorylation of p38 MAPK (Fig. 5E). Furthermore, relaxin stimulated phosphorylation of CREB that was reduced by both H89 and the p38 MAPK inhibitor (Fig. 5F). The degree of reduction of relaxin-stimulated CREB phosphorylation with by the p38 MAPK inhibitor (38.4 ± 4.8%) compares favorably to the response seen in HEK-RXFP1 cells. Finally, the expression level of PGC1α was increased in response to relaxin (Fig. 5G). Taken together, these results confirm that the signaling pathways leading to PGC1α expression and PPARγ activation are intact in LX2 cells endogenously expressing RXFP1.