MDMX acidic domain inhibits p53 DNA binding in vivo and regulates tumorigenesis

Abstract

The MDM2 homolog MDMX oncoprotein is indispensable for inhibition of p53 during normal embryonic development and malignant transformation, yet how MDMX harnesses p53 functions is unclear. In addition to a canonical N-terminal p53-binding domain, recent work suggests the central acidic domain of MDMX regulates p53 interaction through intramolecular mimicry and engages in second-site interaction with the p53 core domain in vitro. To test the physiological relevance of these interactions, we generated an MDMX knockin mouse having substitutions in a conserved WW motif necessary for these functions (W201S/W202G). Notably, MDMX^SG cells have normal p53 level but increased p53 DNA binding and target gene expression, and rapidly senesce. In vivo, MDMX^SG inhibits early-phase disease in Eµ-Myc transgenic mice but accelerates the onset of lethal lymphoma and shortens overall survival. Therefore, MDMX is an important regulator of p53 DNA binding, which complements the role of MDM2 in regulating p53 level. Furthermore, the results suggest that the WW motif has dual functions that regulate p53 and inhibit Myc-driven lymphomas independent of p53.

Keywords: CK1α; DNA binding; MDMX; Myc; p53.

Fig. 1.

MDMX acidic domain regulates MDMX and p53 through intra- and intermolecular interactions. (A) A model of p53 inhibition by MDMX. The central acidic domain (AD) of MDMX engages in intramolecular interactions with the p53 binding domain (p53BD) and RING domains in the absence of p53. CK1α binding disrupts intramolecular interactions of AD with p53BD, increasing the apparent affinity of the MDMX N terminus for p53. Following formation of the initial MDMX–p53 complex, the AD also establishes interaction with the p53 core domain, stabilizing the complex and blocking p53 binding to DNA. (B) Partial sequence of the MDMX AD showing the conserved WW motif important for autoinhibition and the substitution mutations introduced into the mouse germline.

Fig. 2.

Generation of MDMX^SG mutant mice and embryo fibroblasts. (A) Diagram of the knockin targeting construct for introducing the W201S/W202G substitutions into mouse MDMX. (B) Growth curve of primary MEFs. MEFs of the indicated genotypes (from littermates) were cultured under identical conditions, and cell numbers were determined at the indicated intervals. Three cell lines were tested for each genotype. (C) MEFs were plated at low density (500 cells per 10-cm plate) and cultured for 14 d. Colonies were stained with crystal violet. The result is representative of three cell lines for each genotype. (D) The rates of cell proliferation in the indicated MEFs (n = 3) were determined by pulse labeling with [³H]thymidine and measuring the incorporation into DNA. (E) Primary MEFs from wild-type and MDMX^SG/SG littermate embryos were cultured following the 3T3 protocol. The cell density at each passage is plotted. The result is representative of three cell lines for each genotype. Error bars indicate standard deviation.

Fig. 3.

MDMX^SG/SG MEFs undergo spontaneous growth arrest and express hallmarks of chronic DNA damage. (A) Low-passage MEFs (P6) of the indicated genotypes were treated with 1 nM actinomycin D for 18 h or 10 Gy gamma irradiation for 4 h, and analyzed for expression of p53 pathway markers. The result is representative of three cell lines for each genotype. (B) MDMX^SG/SG and WT control MEFs cultured for 10 passages were stained for SA-β-gal activity as a marker for senescence. The result is representative of three cell lines for each genotype. (C) MEFs cultured for 10 passages were analyzed for markers of the DNA damage response.

Fig. 4.

Increased p53 activity in MDMX^SG/SG mouse tissues. (A) The thymus of mice 4 h after irradiation was analyzed for apoptosis using TUNEL staining. (B) Quantitation of apoptotic cells in the thymus 4 h after 6 Gy irradiation (n = 3; P < 0.01). Error bars indicate standard deviation. (C) The tissues of mice 4 h after 6 Gy irradiation were analyzed for p53 pathway markers by Western blot. The result is representative of >3 mice for each genotype.

Fig. 5.

Increased basal p53 DNA binding and transcriptional output in MDMX^SG/SG mice. The thymocytes and splenocytes of mice 4 h after treatment with 6 Gy IR (n = 3) or untreated controls (n = 3) were analyzed for p53 target gene mRNA expression (A, C, and E) by RT-PCR and p53 binding (B, D, and F) to the corresponding promoters by ChIP. *P < 0.05. Sample pairs without a P value indicates no statistical difference.

Fig. 6.

Biochemical defects of MDMX^SG. (A) MDMX–p53 binding in the thymus was determined by IP-Western blot before and after irradiation. The result is representative of three animals for each genotype. WCE, whole cell extract. (B) MDMX–CK1α binding in mouse tissues was determined by IP-Western blot. (C) Mouse tissues were analyzed for MDMX S289 phosphorylation by IP-Western blot using a phospho-specific pS289-MDMX antibody. (D) FLAG-tagged MDMX was coexpressed with p53 and CK1α in H1299 cells. The MDMX–p53 complex was purified using FLAG antibody beads. The DNA binding activity of the complex was determined by pull down using a biotinylated oligonucleotide with a p53 binding site. FLAG-p53 and FLAG-MDM2 were used as positive controls. (E) Detection of MDMX domain interaction with p53 using a fragment release assay. Beads loaded with GST-p53 were used to capture protease-cleavable MDMXc3 expressed in H1299 cells. The complex was cleaved with PreScission, and the partition of MDMX fragments on the beads and in the supernatant was determined by Western blot. SQ, region with ATM phosphorylation sites. (F) Result of the fragment release assay using MDMX, the MDMX–CK1α complex, and SG mutant. The ratios of each MDMX fragment on the beads and in the supernatant indicate the binding affinity to p53 after cleavage of MDMXc3. Specificity control: The MDMX AD fragment does not bind to GST-p53-1–82 without the core domain.

Fig. 7.

MDMX^SG inhibits Eµ-Myc lymphoma onset but shortens survival. (A) MDMX^SG/SG mice were crossed with the Eµ-Myc transgenic mice to generate the three cohorts Eµ-Myc;MDMX^+/+ (median survival 184 days), Eµ-Myc;MDMX^SG/+ (median survival 121 days), and Eµ-Myc;MDMX^SG/SG (median survival 84 days). Littermates were followed for lymphoma development. (B) Lymphoma tissues (enlarged spleens) of the indicated genotypes were analyzed for expression of p53 pathway markers. The status of p53 in each tumor sample was determined by RT-PCR amplification of p53 mRNA and sequencing of the entire p53 ORF. A tumor with a heavy background band that obscured the p53 blot is marked with an asterisk. (C) The peripheral blood B lymphocytes (CD19⁺) of 30-d-old littermates were analyzed by forward scatter. The fraction of large cells was determined. A histogram of one mouse for each genotype is shown. (D) Average fraction of large peripheral B lymphocytes in groups of five mice for each genotype (**P < 0.01). Error bars indicate standard deviation.

Fig. 8.

p53-independent growth inhibition by MDMX is inactivated by the SG mutation. (A) H1299 cells (p53-null) infected with lentivirus expressing GFP or RFP were used for secondary infection with lentivirus expressing LacZ or MDMX. The RFP-labeled MDMX-expressing cells and GFP-labeled LacZ control were mixed and cocultured for eight passages. The RFP/GFP ratio was determined before and after coculture by FACS. (B) Same as A except MDMX^SG was expressed. (C and D) Same as A and B, except RFP/GFP labels were switched. (E) Western blot confirming similar expression levels of MDMX and MDMX^SG in the fluorescence-labeled cells.