beta-arrestin 2 oligomerization controls the Mdm2-dependent inhibition of p53

Abstract

beta-arrestins (beta-arrs), two ubiquitous proteins involved in serpentine heptahelical receptor regulation and signaling, form constitutive homo- and heterooligomers stabilized by inositol 1,2,3,4,5,6-hexakisphosphate (IP6). Monomeric beta-arrs are believed to interact with receptors after agonist activation, and therefore, beta-arr oligomers have been proposed to represent a resting biologically inactive state. In contrast to this, we report here that the interaction with and subsequent titration out of the nucleus of the protooncogene Mdm2 specifically require beta-arr2 oligomers together with the previously characterized nucleocytoplasmic shuttling of beta-arr2. Mutation of the IP6-binding sites impair oligomerization, reduce interaction with Mdm2, and inhibit p53-dependent antiproliferative effects of beta-arr2, whereas the competence for receptor regulation and signaling is maintained. These observations suggest that the intracellular concentration of beta-arr2 oligomers might control cell survival and proliferation.

Fig. 1.

β-arr2 IP6-binding site mutants show impaired oligomerization. (a) β-arr2 constructs used in the study. Boxes indicate N- and C-terminal globular domains of β-arr2 (N-ter arrestin and C-ter arrestin, respectively). Positions of alanine substitutions to generate the mutants of the N-terminal and C-terminal IP6-binding sites (ΔIP6-N and ΔIP6-C, respectively) and of the NES of β-arr2 (ΔNES) are shown. (b) BRET saturation curves obtained by measuring BRET ratios in HEK293T cells expressing fixed quantities of BRET donor (β-arr2-Rluc) and increasing amounts of BRET acceptors (YFP-tagged β-arr2 constructs and control YFP). Relative amounts of BRET acceptor are expressed as the ratio between the fluorescence of the acceptor over the luciferase activity of the donor. YFP° corresponds to background fluorescence in cells expressing the BRET donor alone. Error bars indicate SD of mean specific BRET-ratio values from 21 to 42 individual transfections grouped as a function of the amount of BRET acceptor.

Fig. 2.

Inhibition of β-arr2 oligomerization impairs cytosolic delocalization of Mdm2. (a) HeLa cells expressing β-arr2-YFP, β-arr2ΔIP6-N-YFP, or β-arr2ΔIP6-C-YFP constructs and Mdm2. Anti-Mdm2 monoclonal antibody was revealed by using a secondary Cy3-labeled donkey anti-mouse IgG antibody. Similar results were obtained in H1299, COS, and HEK293T cells (data not shown). (b) Determination of predominant nuclear, cytosolic, or diffuse Mdm2 localization in the indicated number of cells used in immunofluorescence experiments. Bars indicate the percentage of cells in each category. Similar results were obtained in H1299 cells (data not shown). (c) BRET saturation experiments (as described in Fig. 1b) in HEK293T cells expressing Rluc-Mdm2 as BRET donor and the indicated acceptors (individual transfections, n = 33–66). (d) Interaction between Mdm2 and β-arr2 assessed by coimmunoprecipitation experiments. COS7 cells were cotransfected with (1 μg) of β-arr2, β-arr2ΔIP6-N, or β-arr2ΔIP6-C Flag-tagged constructs or control empty vector and Rluc-Mdm2 cDNA. After immunoprecipitation with anti-Flag antibodies, immunoblots were probed with anti-RLuc antibodies; 10% of input is shown. (e) FLIM in unstimulated HEK293 live cells expressing EGFP-Mdm2 (the FRET donor), alone or in association with the indicated FRET acceptors: mCherry, β-arr2-mCherry, β-arr2 ΔIP6-N-mCherry, and β-arr2 ΔIP6-C-mCherry. Average values (±SEM, n = 10) of fluorescence lifetime (τ), shown in the figure with a pseudocolor scale (from 1.5 to 3 ns) were 2.386 + 0.004, 2.259 ± 0.014, 2.386 ± 0.005, and 2.405 ± 0.012, respectively. Only β-arr2-mCherry significantly decreased the fluorescence lifetime of GFP-Mdm2 (P < 0.05).

Fig. 3.

Visualization of β-arr2–Mdm2 interaction and of β-arr2 oligomers in the nucleus. (a) Nuclear and cytosolic FRET between EGFP-Mdm2 (FRET donor) and mCherry-tagged β-arr2 constructs (FRET acceptors) expressed in live HEK293 cells. Fluorescence lifetime (τ) of FRET donor in both nuclear (τ_nucl) and cytosolic (τ_cyto) areas of cells displaying diffuse distribution of both Mdm2 and WT β-arr2 (gray columns in Fig. 2b) is indicated. Values measured in the nuclear area in the presence of mCherry, β-arr2 ΔIP6-N-mCherry and β-arr2 ΔIP6-C-mCherry were, respectively, 2.386 ± 0.106, 2.378 ± 0.182, and 2.406 ± 0.138 (mean ± SD); they were significantly higher (P < 0.001) than that measured with β-arr2-mCherry. (b) FLIM in unstimulated HEK293 live cells expressing EGFP-β-arr2 (FRET donor), alone or in association with β-arr2-mCherry (FRET acceptor), under basal condition or in the presence of 20 nM LMB. Fluorescence lifetime is shown by using a pseudocolor scale. (c) Comparison of fluorescence lifetime values after LMB treatment in cells expressing the FRET donor alone or both FRET donor and acceptor. **, significant difference between τ values were observed in both nucleus and cytosol (P < 0.001).

Fig. 4.

β-arr2-dependent enhancement of p53 activity requires oligomerization and nuclear shuttling of β-arr2. (a) H1299 cells growing in six-well plates were cotransfected with the p53-cis reporter-Luc (200 ng; Stratagene) and/or cDNA coding for p53 with or without Mdm2 cDNA (0.5 μg each) and the indicated β-arr2 YFP-tagged constructs (0.5 μg). Total transfected DNA was kept constant by using an empty vector. p53 promoter-dependent transcription (the luciferase signal) and β-arr-associated fluorescence (for normalization) were measured; (**, P < 0.001, β-arr2 vs. vector in cells expressing p53 alone). (b) H1299 cells were transfected with equivalent amounts of cDNAs coding for p53, the indicated β-arr2 YFP-tagged constructs or empty vector. At 48 h after transfection, nuclei were stained with propidium iodide (50 μg/ml), and cell cycle phases were analyzed by FACS. Data indicate the fold increase of cells expressing YFP constructs in G₂–M phase, comparatively with cells transfected with vector DNA; *, P < 0.05 WT β-arr2 vs. vector. Cell distribution in the various phases of cell cycle is shown in SI Table 1. (c) MEF β-arr2KO were transfected with p53-Luc reporter and reconstituted with the indicated β-arr2 YFP-tagged constructs (1 μg per 35-mm dish). p53 promoter-dependent transcription and β-arr-associated fluorescence were measured as in a; (**, P < 0,001). (d) Immunoblot analysis of WT and mutant β-arr2 in reconstituted MEF β-arr2KO, comparatively with WT MEF and human peripheral blood mononuclear cells (PBMC), as described in SI Fig. 6.