DNAJB1 stabilizes MDM2 and contributes to cancer cell proliferation in a p53-dependent manner

Both MDM2 and MDMX regulate p53, but these proteins play different roles in this process. To clarify the difference, we performed a yeast 2 hybrid (Y2H) screen using the MDM2 acidic domain as bait. DNAJB1 was found to specifically bind to MDM2, but not MDMX, in vitro and in vivo. Further investigation revealed that DNAJB1 stabilizes MDM2 at the post-translational level. The C-terminus of DNAJB1 is essential for its interaction with MDM2 and for MDM2 accumulation. MDM2 was degraded faster by a ubiquitin-mediated pathway when DNAJB1 was depleted. DNAJB1 inhibited the MDM2-mediated ubiquitination and degradation of p53 and contributed to p53 activation in cancer cells. Depletion of DNAJB1 in cancer cells inhibited activity of the p53 pathway, enhanced the activity of the Rb/E2F pathway, and promoted cancer cell growth in vitro and in vivo. This function was p53 dependent, and either human papillomavirus (HPV) E6 protein or siRNA against p53 was able to block the contribution caused by DNAJB1 depletion. In this study, we discovered a new MDM2 interacting protein, DNAJB1, and provided evidence to support its p53-dependent tumor suppressor function.     

Expansion of protein interaction maps by phage peptide display using MDM2 as a prototypical conformationally flexible target protein

Expanding on the possible protein interaction partners in a biochemical pathway is one key molecular goal in the post-genomic era. Phage peptide display is a versatile in vitro tool for mapping novel protein-protein interfaces and the advantage of this technique in expanding protein interaction maps is that in vitro manipulation of the bait protein conformational integrity can be controlled carefully. Phage peptide display was used to expand on the possible types of binding proteins for the conformationally responsive protein MDM2. Peptides enriched differ depending upon whether MDM2 is ligand-free, zinc-bound, or RNA-bound, suggesting that MDM2 conformational changes alter the type of peptide ligands enriched. Classes of putative/established MDM2-binding proteins identified by this technique included ubiquitin-modifying enzymes (F-box proteins, UB-ligases, UBC-E1) and apoptotic modifiers (HSP90, GAS1, APAF1, p53). Of the many putative MDM2 proteins that could be examined, the impact of HSP90 on MDM2 activity was studied, since HSP90 has been linked with p53 protein unfolding in human cancers. Zinc ions were required to reconstitute a stable MDM2-HSP90 protein complex. Zinc binding converted MDM2 from a monomer to an oligomer, and activated MDM2 binding to its internal RING finger domain, providing evidence for a conformational change in MDM2 protein when it binds zinc. Reconstitution of an HSP90-MDM2 protein complex in vitro stimulated the unfolding of the p53 tetramer. A p53 DNA-binding inhibitor purified from human cells that is capable of unfolding p53 at ambient temperature in vitro contains co-purifying pools of HSP90 and MDM2. These data highlight the utility of phage peptide display as a powerful in vitro method to identify regulatory proteins that bind to a conformationally flexible protein like MDM2.     

Structure of Human MDM4 N-Terminal Domain Bound to a Single-Domain Antibody

The N-terminal domain of MDM4 binds to the N-terminal transactivation domain of the tumor suppressor p53 and is an important negative regulator of its transactivation activity. As such, inhibition of the binding of MDM4 to p53 is a target for anticancer therapy. The protein has not been crystallized satisfactorily for structural studies without the addition of an N-terminal p53 peptide. We selected a single-domain antibody (VH9) that bound to the human domain with a dissociation constant of 44 nM. We solved the structure of the complex at 2.0-Å resolution. The asymmetric unit contained eight molecules of VH9 and four molecules of MDM4. A molecule of VH9 was located in each transactivation domain binding site, and the four non-MDM4-bound VH9 domains provided additional crystal contacts. There are differences between the structures of human MDM4 domain bound to VH9 and those of human and zebra fish MDM4 bound to a p53 peptide. Molecular dynamics simulations showed that the binding pocket in the three MDM4 structures converged to a common conformation after removal of the ligands, indicating that the differences are due to induced fit. The largest conformational changes were for the MDM4 molecules bound to p53. The simulated and observed structures should aid rational drug design. The use of single-domain antibodies to aid crystallization by creating a molecular scaffold may have a wider range of applications.     

Tetra-substituted imidazoles as a new class of inhibitors of the p53–MDM2 interaction

Capitalizing on crystal structure information obtained from a previous effort in the search for non peptide inhibitors of the p53–MDM2 interaction, we have discovered another new class of compounds able to disrupt this protein–protein interaction, an important target in oncology drug research. The new inhibitors, based on a tetra-substituted imidazole scaffold, have been optimized to low nanomolar potency in a biochemical assay following a structure-guided approach. An appropriate strategy has allowed us to translate the high biochemical potency in significant anti-proliferative activity on a p53-dependent MDM2 amplified cell line.     

Human Glioblastoma Multiforme: p53 Reactivation by a Novel MDM2 Inhibitor

Cancer development and chemo-resistance are often due to impaired functioning of the p53 tumor suppressor through genetic mutation or sequestration by other proteins. In glioblastoma multiforme (GBM), p53 availability is frequently reduced because it binds to the Murine Double Minute-2 (MDM2) oncoprotein, which accumulates at high concentrations in tumor cells. The use of MDM2 inhibitors that interfere with the binding of p53 and MDM2 has become a valid approach to inhibit cell growth in a number of cancers; however little is known about the efficacy of these inhibitors in GBM. We report that a new small-molecule inhibitor of MDM2 with a spirooxoindolepyrrolidine core structure, named ISA27, effectively reactivated p53 function and inhibited human GBM cell growth in vitro by inducing cell cycle arrest and apoptosis. In immunoincompetent BALB/c nude mice bearing a human GBM xenograft, the administration of ISA27 in vivo activated p53, inhibited cell proliferation and induced apoptosis in tumor tissue. Significantly, ISA27 was non-toxic in an in vitro normal human cell model and an in vivo mouse model. ISA27 administration in combination with temozolomide (TMZ) produced a synergistic inhibitory effect on GBM cell viability in vitro, suggesting the possibility of lowering the dose of TMZ used in the treatment of GBM. In conclusion, our data show that ISA27 releases the powerful antitumor capacities of p53 in GBM cells. The use of this MDM2 inhibitor could become a novel therapy for the treatment of GBM patients.     

MDM2 is a negative regulator of p21WAF1/CIP1, independent of p53

The MDM2 oncogene has both p53-dependent and p53-independent activities. We have previously reported that antisense MDM2 inhibitors have significant anti-tumor activity in multiple human cancer models with various p53 statuses (Zhang, Z., Li, M., Wang, H., Agrawal, S., and Zhang, R. (2003) Proc. Natl. Acad. Sci. U. S. A. 100, 11636-11641). We have also provided evidence that MDM2 has a direct role in the regulation of p21, a cyclin-dependent kinase inhibitor. Here we provide evidence supporting functional interaction between MDM2 and p21 in vitro and in vivo. The inhibition of MDM2 with anti-MDM2 antisense oligonucleotide or Short Interference RNA targeting MDM2 significantly elevated p21 protein levels in PC3 cells (p53 null). In contrast, overexpression of MDM2 diminished the p21 level in the same cells by shortening the p21 half-life, an effect reversed by MDM2 antisense inhibition. MDM2 facilitates p21 degradation independent of ubiquitination and the E3 ligase function of MDM2. Instead, MDM2 promotes p21 degradation by facilitating binding of p21 with the proteasomal C8 subunit. The physical interaction between p21 and MDM2 was demonstrated both in vitro and in vivo with the binding region in amino acids 180-298 of the MDM2 protein. In summary, we provide evidence supporting a physical interaction between MDM2 and p21. We also demonstrate that, by reducing p21 protein stability via proteasome-mediated degradation, MDM2 functions as a negative regulator of p21, an effect independent of both p53 and ubiquitination.     

The central valine concept provides an entry in a new class of non peptide inhibitors of the p53-MDM2 interaction

Disrupting the interaction between the p53 tumor suppressor and its regulator MDM2 is a promising therapeutic strategy in anticancer drug research. In our search for non peptide inhibitors of this protein-protein interaction, we have devised a ligand design concept exploiting the central position of Val 93 in the p53 binding pocket of MDM2. The design of molecules based on this concept has allowed us to rapidly identify compounds having a 3-imidazolyl indole core structure as the first representatives of a new class of potent inhibitors of the p53-MDM2 interaction.     

Identification of a Catalytic Active but Non-Aggregating MDM2 RING Domain Variant

As a key regulator of the tumour suppressor protein p53, MDM2 is involved in various types of cancer and has thus been an attractive drug target. So far, small molecule design has primarily focussed on the N-terminal p53-binding domain although on-target toxicity effects have been reported. Targeting the catalytic RING domain of MDM2 resembles an alternative approach to drug MDM2 with the idea to prevent MDM2-mediated ubiquitination of p53 while retaining MDM2′s ability to bind p53. The design of RING inhibitors has been limited by the extensive aggregation tendency of the RING domain, making it challenging to undertake co-crystallization attempts with potential inhibitors. Here we compare the purification profiles of the MDM2 RING domain from several species and show that the MDM2 RING domain of other species than human is much less prone to aggregate although the overall structure of the RING domain is conserved. Through sequence comparison and mutagenesis analyses, we identify a single point mutation, G443T, which greatly enhances the dimeric fraction of human MDM2 RING domain during purification. Neither does the mutation alter the structure of the RING domain, nor does it affect E2(UbcH5B)–Ub binding and activity. Hence, MDM2-G443T facilitates studies involving binding partners that would be hampered by the low solubility of the wild-type RING domain. Furthermore, it will be valuable for the development of MDM2 RING inhibitors.     

Mutual dependence of MDM2 and MDMX in their functional inactivation of p53

MDMX, an MDM2-related protein, has emerged as yet another essential negative regulator of p53 tumor suppressor, since loss of MDMX expression results in p53-dependent embryonic lethality in mice. However, it remains unknown why neither homologue can compensate for the loss of the other. In addition, results of biochemical studies have suggested that MDMX inhibits MDM2-mediated p53 degradation, thus contradicting its role as defined in gene knockout experiments. Using cells deficient in either MDM2 or MDMX, we demonstrated that these two p53 inhibitors are in fact functionally dependent on each other. In the absence of MDMX, MDM2 is largely ineffective in down-regulating p53 because of its extremely short half-life. MDMX renders MDM2 protein sufficiently stable to function at its full potential for p53 degradation. On the other hand, MDMX, which is a cytoplasmic protein, depends on MDM2 to redistribute into the nucleus and be able to inactivate p53. We also showed that MDMX, when exceedingly overexpressed, inhibits MDM2-mediated p53 degradation by competing with MDM2 for p53 binding. Our findings therefore provide a molecular basis for the nonoverlapping activities of these two p53 inhibitors previously revealed in genetic studies.     

Anti-tumor efficacy of a novel antisense anti-MDM2 mixed-backbone oligonucleotide in human colon cancer models: p53-dependent and p53-independent mechanisms

BACKGROUND: The MDM2 oncogene is amplified or overexpressed in many human cancers and MDM2 levels are associated with poor prognosis. MDM2 not only serves as a negative regulator of p53 but also has p53-independent activities. This study investigates the functions of the MDM2 oncogene in colon cancer growth and the potential value of MDM2 as a drug target for cancer therapy, by inhibiting MDM2 expression with an antisense anti-human-MDM2 oligonucleotide. MATERIALS AND METHODS: The selected antisense mixed-backbone oligonucleotide was evaluated for its in vitro and in vivo antitumor activity in human colon cancer models: LS174T cell line containing wild-type p53 and DLD-1 cell line containing mutant p53. The levels of MDM2, p53 and p21 proteins were quantified by Western blot analysis. RESULTS: In vitro antitumor activity was found in both cell lines, resulting from specific inhibition of MDM2 expression. In vivo antitumor activity of the oligonucleotide occurred in a dose-dependent manner in both models and synergistically or additive therapeutic effects of MDM2 inhibition and the cancer chemotherapeutic agents 10-hydroxycamptothecin and 5-fluorouracil were also observed. CONCLUSIONS: These results suggest that MDM2 have a role in tumor growth through both p53-dependent and p53- independent mechanisms. We speculate that MDM2 inhibitors have a broad spectrum of antitumor activities in human cancers regardless of p53 status. This study should provide a basis for future development of anti-MDM2 antisense oligonucleotides as cancer therapeutic agents used alone or in combination with conventional chemotherapeutics.     

Design of a novel MDM2 binding peptide based on the p53 family

p53 is a major tumor suppressor protein, that binds to, and is negatively regulated by MDM2. In tumors over expressing MDM2, p53 function can be rescued through the disruption of the MDM2-p53 interactions by small molecules and peptides. It is known that MDM2 also binds p73 but not p63, the two homologues of p53. We dissect the structural and energetic reasons underlying this discrimination and have identified a peptide that is intrinsically less helical than p53 and yet has a higher affinity for MDM2. The increased disorder has been introduced by localizing a cationic residue in between two anionic residues, imparting a degree of frustration to the system. In addition, the introduction of a bulkier hydrophobic group towards the centre of the peptide enables the peptide to adapt a bound conformation that on the one hand is most strained, and yet enables the peptide to straddle the largest surface of MDM2, amongst all the peptides. Computations also reveal that this peptide is a dual inhibitor, binding to MDMX. The computed affinity of the new peptide has been validated against MDM2 using fluorescence-based thermal shift assays.     

Mimicking a p53‐MDM2 interaction based on a stable immunoglobulin‐like domain scaffold

Antibodies recognize protein targets with great affinity and specificity. However, posttranslational modifications and the presence of intrinsic disulfide‐bonds pose difficulties for their industrial use. The immunoglobulin fold is one of the most ubiquitous folds in nature and it is found in many proteins besides antibodies. An example of a protein family with an immunoglobulin‐like fold is the Cysteine Protease Inhibitors (ICP) family I42 of the MEROPs database for protease and protease inhibitors. Members of this protein family are thermostable and do not present internal disulfide bonds. Crystal structures of several ICPs indicate that they resemble the Ig‐like domain of the human T cell co‐receptor CD8α As ICPs present 2 flexible recognition loops that vary accordingly to their targeted protease, we hypothesize that members of this protein family would be ideal to design peptide aptamers that mimic protein‐protein interactions. Herein, we use an ICP variant from Entamoeba histolytica (EhICP1) to mimic the interaction between p53 and MDM2. We found that a 13 amino‐acid peptide derived from p53 can be introduced in 2 variable loops (DE, FG) but not the third (BC). Chimeric EhICP1‐p53 form a stable complex with MDM2 at a micromolar range. Crystal structure of the EhICP1‐p53(FG)‐loop variant in complex with MDM2 reveals a swapping subdomain between 2 chimeric molecules, however, the p53 peptide interacts with MDM2 as in previous crystal structures. The structural details of the EhICP1‐p53(FG) interaction with MDM2 resemble the interaction between an antibody and MDM2.     

Small-molecule antagonists of p53-MDM2 binding: research tools and potential therapeutics

p53 regulates a key pathway which protects normal tissues from tumor development that may result from diverse forms of stress. In the absence of stress, growth suppressive and proapoptotic activity of p53 is inhibited by MDM2 which binds p53 and negatively regulates its activity and stability. MDM2 antagonists could activate p53 and may offer a novel therapeutic approach to cancer. Recently, we identified the first potent and selective low molecular weight inhibitors of MDM2-p53 binding, the Nutlins. These molecules activate the p53 pathway and suppress tumor growth in vitro and in vivo. They represent valuable new tools for studying the p53 pathway and its defects in cancer. Nutlins induce p53-dependent apoptosis in human cancer cells but appear cytostatic to proliferating normal cells. Their potent activity against osteosarcoma xenografts suggests that MDM2 antagonists may have clinical utility in the treatment of tumors with wild-type p53.     

Structures of low molecular weight inhibitors bound to MDMX and MDM2 reveal new approaches for p53-MDMX/MDM2 antagonist drug discovery

Intensive anticancer drug discovery efforts have been made to develop small molecule inhibitors of the p53-MDM2 and p53-MDMX interactions. We present here the structures of the most potent inhibitors bound to MDM2 and MDMX that are based on the new imidazo-indole scaffold. In addition, the structure of the recently reported spiro-oxindole inhibitor bound to MDM2 is described. The structures indicate how the substituents of a small molecule that bind to the three subpockets of the MDM2/X-p53 interaction should be optimized for effective binding to MDM2 and/or MDMX. While the spiro-oxindole inhibitor triggers significant ligand-induced changes in MDM2, the imidazo-indoles share similar binding modes for MDMX and MDM2, but cause only minimal induced-fit changes in the structures of both proteins. Our study includes the first structure of the complex between MDMX and a small molecule and should aid in developing efficient scaffolds for binding to MDMX and/or MDM2.