MDMX Recruits UbcH5c to Facilitate MDM2 E3 Ligase Activity and Subsequent p53 Degradation In Vivo

Abstract

MDM2 regulates p53 degradation by functioning as an E3 ubiquitin ligase. The role of MDMX, an MDM2 homolog that lacks E3 ligase activity, in the regulation of p53 degradation remains incompletely understood and sometime controversial. This confusion is due at least in part to studies of p53 degradation mainly carried out in in vitro settings, as elimination of either MDM2 or MDMX from mice results in p53-dependent embryonic lethality, thus obfuscating in vivo studies of the individual roles of MDM2 and MDMX in p53 degradation. To overcome this problem, we generated mice expressing an inducible p53 allele under various MDM2 and MDMX deletion and mutation statuses and studied in vivo p53 degradation. Degradation of p53 in vivo was largely prevented in mice and mouse embryonic fibroblast retaining MDM2 but lacking MDMX. Although MDM2 and MDMX interacted with p53 in the absence of each other, they bound p53 more efficiently as a heterodimer. MDMX, but not MDM2, interacted with ubiquitin-conjugating enzyme UbcH5c, an interaction that was essential for MDMX to enable MDM2 E3 ligase activity for p53 degradation. Grafting the C-terminal residues of MDMX to the C-terminus of MDM2 allowed MDM2 to interact with UbcH5c and enhanced MDM2-mediated p53 degradation in the absence of MDMX. Together, these data indicate that MDMX plays an essential role for p53 degradation in vivo by recruiting UbcH5c to facilitate MDM2 E3 ligase function. SIGNIFICANCE: This study provides the first in vivo evidence of MDMX facilitating MDM2-mediated p53 degradation, clarifying its role in the regulation of this critical tumor suppressor.

Figure 1.. Deletion of MDMX in mice causes p53 accumulation.

Early passages of (A) Mdm2^+/+;Mdmx^+/+;p53^ER/- and (B) Mdm2^+/+;Mdmx^−/−;p53^ER/- mouse embryonic stem (MEF) cells were treated with 4-hydroxytamoxifen (4-OHT) for the indicated times. The levels of MDM2, MDMX, p53 and actin were analyzed by western blot. (C) The amounts of p53 remained at each time point in A-B were quantified by densitometry, normalized to actin, and plotted. (D) Spleen and (E) thymus tissues were isolated from 8 weeks old Mdm2^+/+;Mdmx^+/+;p53^ER/- and Mdm2^+/+;Mdmx^−/−;p53^ER/- mice after treatment with 4-OHT for 6 hours. Tissue lysates were analyzed by western blot. Relative amounts of p53 were indicated. (F-K) MEFs of varies MDM2 and MDMX statuses were treated with 4-OHT for the indicated times and cell lysates were analyzed by western blot. Relative amounts of p53 were indicated.

Figure 2.. Deletion of MDMX impedes p53 polyubiquitination and proteasomal degradation.

(A) MEFs of indicated genotypes were treated with 4-OHT for 2 hours and then with 10 μM MG132 for 5 hours before harvesting. The cell lysates were isolated and immunoprecipitated with an anti-p53 antibody and analyzed by western blot using an anti-ubiquitin antibody. (B) MEFs of indicated genotypes were treated with 4-OHT for 2 hours and then with MG132 for additional 4 hours. Cell lysates were isolated and the levels of p53 and actin were analyzed by western blot. Relative amounts of p53 were quantified and shown as a bar graph below. MEFs of Mdm2^+/+;Mdmx^+/+;p53^ER/-(C), Mdm2^−/−;Mdmx^+/+;p53^ER/- (D), and Mdm2^+/+;Mdmx^−/−;p53^ER/- (E) genotypes were treated with 4-OHT for 2 hours followed by treatment with 100 μg/mL cycloheximide and harvested at the indicated time points. MDM2, MDMX, p53 and actin were analyzed by western blot. The relative levels of p53 (F), MDM2 (G), and MDMX (H) was quantified, normalized to actin, and plotted.

Figure 3.. MDM2-MDMX heterodimerization facilitates MDM2-p53 binding.

(A) MEFs of indicated genotypes were treated with 4-OHT for 2 hours. The cells were then harvested, lysed, and immunoprecipitated (IP) for MDM2 and MDMX, and blotted with the indicated proteins. Input represents 5% of total cell lysate utilized for IP. (B) MEFs of indicated genotypes were treated and blotted as in A. (C) MEFs of indicated genotypes were treated and blotted as in A.

Figure 4.. MDM2-MDMX heterodimerization facilitates p53 cytoplasmic localization.

Cell lysates were isolated from (A) Mdm2^+/+;Mdmx^+/+;p53^ER/-, (B) Mdm2^487/487;Mdmx^+/+;p53^ER/-, (C) Mdm2^462/462;Mdmx^+/+;p53^ER/-, and (D) Mdm2^+/+;Mdmx^−/−;p53^ER/- MEFs. The lysates were either untreated (Total) or separated into cytoplasmic (Cyto) and nuclear (Nucl) fractions and analyzed for MDM2, MDMX, and p53 by western blot. Tubulin and PARP were used as controls for cytoplasmic fraction and nuclear fraction, respectively. The relative amounts of p53 in each fraction were shown. (E-H) MEFs of indicated genotypes were stained with p53 antibody and counter-stained with DAPI, and cell images were taken by microscope. Representative images were shown.

Figure 5.. MDMX interacts with UbcH5c.

(A) MEFs of indicated genotypes were treated with 4-OHT for 2 hours, cell lysates were harvested and immunoprecipitated (IP) for MDM2 or MDMX, and blotted for indicated proteins. Input represents 5% of total cell lysate utilized for IP. (B) MEFs of indicated genotypes were treated as in A and immunoprecipitated for MDM2 and MDMX and blotted for indicated proteins. (C) SJSA and MCF-7 cells were immunoprecipitated (IP) for MDM2 or MDMX, and immunoblotted with indicated antibodies. Input represents 5% of total cell lysate utilized for IP. (D) MEFs of indicated genotypes were treated with siRNA against Mdmx (MDMX) or non-specific sequences (NS) for 24 hour followed by treatment with 4-OHT for 2 hours. Cell lysates were harvested and immunoprecipitated for MDM2 and blotted for indicated proteins. (E) MDMX expressed in bacterial was analyzed by SDS-PAGE and Commassie Blue staining. (F) MDMX-UbcH5c complexes were resolved by SDS-PAGE and blotted for MDMX, GST, and UbcH5c.

Figure 6.. MDMX C-terminus is essential for UbcH5c binding and p53 degradation.

(A) U2OS cells were transfected with empty vector DNA, MDMX or MDMX^ΔC7 plasmid DNA. IP was performed with an anti-MDMX antibody and blotted for the indicated proteins. (B) U2OS cells were transfected with vector, MDM2 or MDM2^XC7 plasmid DNA. IP was performed with an anti-MDM2 antibody and blotted for with the indicated proteins. (C) 2KO (Mdm2^−/−;p53^−/−) MEFs were co-transfected with p53 and MDM2 or MDM2^XC7 plasmid DNA. The levels of MDM2, p53 and actin were analyzed by western blot. (D) 2KO MEFs were co-transfected with vector, MDMX^ΔC7 or MDM2^XC7 plasmid DNA along p53 DNA. The cell lysates were isolated and immunoprecipitated with an anti-p53 antibody and analyzed by western blot using an anti-ubiquitin antibody. (E) 2KO MEFs were treated with siRNA against MDMX or non-specific sequences (NS) for 24 hours followed by transfection with indicated plasmid DNA for another 24 hours. Cell lysates were blotted for the indicated proteins. (F) 2KO MEFs were treated with siRNA against UbcH5c or non-specific sequences (NS) for 24 hours followed by transfection with indicated plasmid DNA for another 24 hours. Cell lysates were blotted for the indicated proteins.

Figure 7.. A hypothetic model for MDMX facilitating MDM2 E3 ligase function.

Singular MDM2 or MDMX weakly interact with p53 (shown by dashed molecules of p53) through their N-terminus. When MDM2 and MDMX form a heterodimer via their respective C-terminal RING domains, the binding affinity of the MDM2-MDMX heterodimer with p53 is increased (shown by solid molecules of p53). At the same time, MDMX brings UbcH5c to MDM2 proximity, activates MDM2 RING E3 ligase activity to transfer ubiquitin to p53.