Figure 1.
Expression and localization of NR fusion proteins.
(A) Localization of THR, RXR and PPARG2 proteins in the nucleus. Hela cells transfected with the respective EYFP fused NR expression vector were stained with Hoechst 33342. Left images show the specific localization of each EYFP fusion protein; middle images show the nuclei revealed by Hoechst staining; right images show merge images between EYFP and DNA staining. (B) Protein identity confirmed by Western blot. HEK293T cells transfected with the respective EYFP or Luc fused NR expression vectors were lysed in RIPA buffer and analyzed by Western blot. Anti-GFP antibodies were used to detect THR-EYFP and PPARG2-EYFP, and anti-Renilla Luciferase antibody was used to visualize RXR-Luc.
Figure 2.
Functional validation of NR fusion proteins.
(A) Electrophoretic mobility shift assay (EMSA) experiments with NR fusion proteins encoding RXR-Luc, THR-EYFP and PPARG2-EYFP. Nuclear extracts were prepared from HEK293T cells co-transfected with RXR-Luc and THR-EYFP or RXR-Luc and PPARG2-EYFP encoding plasmids. Several biotinylated RE oligonucleotides were mixed with nuclear extracts containing overexpressed NR fusion proteins: A DR4 probe was used to test the DNA binding of RXR-Luc+THR-EYFP (lane 4), and a DR1 probe was used for RXR-Luc+PPARG2-EYFP (lane 7). In the absence of nuclear extracts no signal was obtained (lanes 3 and 6). As a positive control for EMSA, a biotin-labeled 60 bp duplex bearing the Epstein-Barr Nuclear antigen 1 (EBNA-1) binding sequence was incubated with an extract containing the EBNA protein (lane 1). For each condition, the specificity of the gel shift experiment was determined by the addition of a 200-fold excess of the corresponding unlabeled double stranded consensus sequence (lanes 2, 5 and 8). (B) Histograms represent transcriptional activity of NR fusion proteins RXR-Luc, PPARG2-EYFP (left) or RXR-Luc and THR-EYFP (right) monitored with gene reporters containing DR1 responsive elements or thyroid responsive element palindomic (TREPAL) in their promoters: Hela cells were transiently transfected with RXR-Luc and PPARG2-EYFP or RXR-Luc and THR-EYFP expression plasmids together with a PPAR DR1 (left) or a TREPAL firefly luciferase gene reporter construct (right), respectively. A pCMV-β-galactosidase plasmid was used for normalization. 6 hours after transfection some cells were stimulated 24 h with 10−6 M rosiglitazone (ROSI) or 10−6 M of triodothyronine (T3) and 10−6 M of 9cis retinoic acid (9cis RA). After cells lysis, transcriptional activity was measured by monitoring firefly luciferase activities normalized by β-galactosidase activities. (C) Histograms represent transcriptional fold activation by ligand of transfected tagged constructs, RXR-Luc and PPARG2-EYFP (left) or RXR-Luc and THR-EYFP (right) compared to non tagged plasmids. Values shown are mean ± SD. (n = 2–3). Statistical differences were analyzed by one-way ANOVA followed by Bonferroni’s post hoc test: *P<0.05; **P<0.01; ***P<0.001.
Figure 3.
Titration BRET experiments in living cells.
BRET between RXR-Luc and PPARG2-EYFP, RXR-Luc and THR-EYFP. Regression curves are represented with the BRET value as a function of the fluorescence/luminescence ratio (EYFP/Luc). HEK293T cells were transfected with a fixed amount of RXR-Luc donor plasmid together with increasing amount of acceptor plasmids (PPARG2-EYFP above or THR-EYFP and EYFP lower) and BRET was measured in living cells after stimulation with control DMSO or ligands either on resuspended cells 5–10 minutes in PBS (left panel), or on adherent cells during an overnight (O/N) incubation in culture medium (right panel). Values represent BRET measures (each in triplicate) integrated over a 20 min reading (A), BRET titration curves between RXR-Luc and PPARG2-EYFP from control cells (open square) and from cells stimulated 5 min with ligands 9cis RA+TROG (filled triangle) in PBS resuspended cells (left panel) or from adherent cells stimulated O/N (right panel) (B), BRET titration curves between RXR-Luc and THR-EYFP from control cells (open circles) and from cells stimulated 5 min with 9cisRA+T3 (filled circles) in PBS (left panel) or from adherent cells stimulated O/N (right panel) (right panel). A negative control saturation experiment is also shown between a RXR-Luc donor and an unfused EYFP acceptor protein (open triangle). Apparent affinity (apparent Kd) between RXR-Luc and PPARG2-EYFP or RXR-Luc and THR-EYFP represents EYFP/Luc ratio corresponding to the BRETmax/2 value (BRET50). Shown are cumulative data from three independent experiments in triplicate.
Figure 4.
In vitro BRET shift between NRs in cleared cell lysates in the presence of a DNA RE.
(A and B) For these in vitro titration BRET experiments, a fixed amount of PLB cell lysate expressing RXR-Luc protein (80 ku luciferase) is mixed with increasing amount of a PLB cell lysate expressing PPARG2-EYFP or THR-EYFP (0, 10, 20, 30, 50, 75, 100 or 150 ku of fluorescence) and regression curves are represented as the BRET value (recorded over a 20 min period) as a function of the fluorescence/luminescence ratio in the absence (control TE: Tris 10 mM PH 7,5; EDTA 1 mM) or presence of 100 nM of dsDNA RE (diluted in TE). (A) In vitro BRET saturation between RXR-Luc and PPARG2-EYFP with TE (control TE, filled squares) or 100 nM of dsDNA DR1 (filled triangles). (B) In vitro BRET saturation between RXR-Luc and THR-EYFP with TE (control TE, filled squares) or 100 nM of dsDNA DR4 (filled triangles). (C) One hour BRET kinetic monitoring in vitro interaction between 80 ku of donor RXR-Luc and 40 ku or 100 ku fluorescence of PPARG2-EYFP in the absence (control TE) or presence of 100 nM of dsDNA RE DR1. (D) One hour BRET kinetic monitoring in vitro interaction between 80 ku of donor RXR-Luc and 10 ku, 40 ku or 100 ku fluo of THR-EYFP in the absence (control TE) or presence of 100 nM of dsDNA RE DR4.
Figure 5.
DNA dose responses of in vitro BRET shift.
Dose response experiments showing in vitro BRET shift induced by different concentrations of a dsDNA RE. Values represent BRET measures (each in triplicate) integrated over a 10 min reading for DR1 and DR4 DNA RE and over a 20 min recording for NF DNA RE. (A), BRET values between RXR-Luc and PPARG2-EYFP as a function of the log[dsDNA] of DR1 (0,1 to 300 nM) (B), BRET values between RXR-Luc and PPARG2-EYFP as a function of the log[dsDNA] of NF (1 to 300 nM) (C) BRET values between RXR-Luc and THR-EYFP as a function of the log[dsDNA] of DR4 (0.1 to 100 nM) (D) BRET values between RXR-Luc and THR-EYFP as a function of the log[dsDNA] of NF. In each graph, the black arrow indicated the BRET value corresponding to a concentration of 100 nM of DNA RE. Shown are data from two to three independent experiments in triplicate and values represent mean ± SD. For each dose response, the R2 of the fitting slope and EC50 (corresponding to the half maximum effective concentration of a DNA RE) are shown.
Figure 6.
Analysis of DNA binding mutants.
(A), BRET shift between 80 ku of RXR-Luc and 40 ku of PPARG2-EYFP-WT or mutants PPARG2-EYFP-C142R and PPARG2-EYFP-F375A in presence or absence of 100 nM DR1 dsDNA RE. Values represent BRET measures (each in triplicate) integrated over a 10 min reading. Values shown are means ± SD of three independent experiments (n = 3). Statistical differences of PPARG2WT DR1 relative to TE control (***P<0.001), PPARG2-C142R DR1 (###P<0.001) and PPARG2-F375A DR1 (¶¶¶P<0.001), as well as PPARG2WT TE control relative to PPARG2-F375A TE (ΔP<0.05) were analyzed by one-way ANOVA followed by Dunnett’s multiple comparison post hoc test. (B), One hour BRET kinetic monitoring interaction between 80 ku of donor RXR-Luc and 40 ku or 100 ku fluo of PPARG2-EYFPC142R in the absence (control TE) or presence of 100 nM of dsDNA RE DR1. (C), One hour BRET kinetic monitoring interaction between 80 ku of donor RXR-Luc and 40 ku or 100 ku fluo of PPARG2-EYFPF375A in the absence (TE) or presence of 100 nM of dsDNA RE DR1. For comparison with PPARG2WT in graphs 6B and 6C, a BRET kinetic recording interaction between 80 ku of donor RXR-Luc and 100 ku fluo of PPARG2-EYFPWT in the presence of 100 nM of dsDNA RE DR1 is shown (open circle).
Figure 7.
Stuctural requirements of the DNA targets.
In vitro BRET shift recorded between RXR-Luc and EYFP-RXR, or RXR-Luc and PPARG2-EYFP, or RXR-Luc and THR-EYFP in the presence of 100 nM of different responsive elements (described in Table 1). Values represent BRET measures (each in triplicate) integrated over a 10 min reading. (A), BRET shift obtained between 80 ku of RXR-Luc and 40 ku of PPARG2-EYFP with control (TE), single strand (ss) consensus DR1 and 13 different double strand DNA RE. (B), BRET shift obtained between 80 ku of RXR-Luc and 40 ku of THR-EYFP with control (TE), single strand (ss) consensus DR4 and 13 different responsive elements. (C), BRET shift between 80 ku luciferase of donor RXR-Luc and 40 ku fluo of acceptor EYFP-RXR in absence (control TE) or in presence of 100 nM ds DNA DR1 or DR4 RE. Values shown are means ± SD (n = 3). Statistical differences relative to control (TE) were analyzed by one-way ANOVA followed by Dunnett’s multiple comparison post hoc test: ***P<0.001 and *P<0.05.
Table 1.
Different responsive elements used for in vitro BRET shift assays.
Figure 8.
Effect of polydIdC and MgCl2 on in vitro BRET shift assay.
(A) Effect of polydIdC addition on RXR/PPARG2 and RXR/THR BRET assay. A BRET shift assay in PLB was performed in presence or absence of polydIdC (50 ng/µl) and effect on basal BRET and specific DNA RE induced BRET shift was measured for RXR-Luc and PPARG2-EYFP (**P<0.01; ***P<0.001), or RXR-Luc and THR-EYFP (#P<0.05; ###P<0.001), Statistical differences relative to control (TE) were analyzed by one-way ANOVA followed by Dunnett’s multiple comparison post hoc test. (B, C and D) Effect of MgCl2 addition on BRET shift RXR/PPARG2, RXR/THR and RXR/RXR BRET assay. BRET shift experiments were carried out in Gel Shift Buffer (GSB) supplemented with 0, 2 or 5 mM of MgCl2. Calculation of the DR1/DR4 BRET shift ratio for RXR/PPARG2 (B), the DR4/DR1 BRET ratio for RXR/THR (C) and the DR1/DR4 BRET ratio for RXR/RXR (D), illustrated the higher specificity of BRET shift observed by increasing MgCl2 concentration. Histograms represent mean BRET value ± SD of at least 3 experiments in triplicate. Values represent BRET measures integrated over a 20 min reading.
Figure 9.
Subcellular localization of the interaction between RXR and PPARG2.
BRET imaging was recorded in (A), HEK293T cells co-transfected with RXR-Luc and DsRed (as a transfection reporter) or RXR-Luc and PPARG2-EYFP or (B), HEK293T cells co-transfected with RXR-Luc and PPARG2-EYFP-F375A or PPARG2-EYFP-C142R. The pictures show expression of PPARG2-EYFP or mutants (GFP), RXR-Luc (Em480), PPARG2-EYFP or mutants excited by energy transfer (Em535) and BRET signal generated by the two tagged proteins (535/480). Please note the high and clustered BRET signals obtained between RXR-Luc and PPARG2-EYFP. The pixel-by-pixel 535 nm/480 nm ratios were calculated by dividing the absolute light intensities per pixel of images obtained at 535 nm over 480 nm. These numerical ratios (comprised between 0 and 1.5) were translated and visualized with a continuous 256 pseudo-color look-up table (LUT) as displayed in the figures (C) Titration curves were obtained by expressing for each cell the mean BRET intensity (Em535/Em480) as a function of the mean fluorescence/luminescence ratio (EYFP/Luc). Histograms represent the mean BRET intensity (D) and standard deviation (clusterization index) (E). A high standard deviation indicates a clusterization of the signal. Statistical differences were analyzed by one-way ANOVA followed by Bonferroni post hoc test: *P<0.05 (significant against RXR-Luc); #P<0.05 (significant against PPARG2-EYFP-F375A); ¶P<0.05 (significant against PPARG2-EYFP-C142R). The number of cells assayed for each condition was the following: 11 cells for RXR-Luc alone (control); 23 cells for RXR-Luc and PPARG2-EYFP; 15 cells for RXR-Luc and PPARG2-EYFP-F375A; and 21 cells for RXR-Luc and PPARG2-EYFP-C142R.