Activation of the pathway by small-molecule- induced and MDMX dimerization

Bradford Graves1, Thelma Thompson, Mingxuan Xia, Cheryl Janson, Christine Lukacs, Dayanand Deo, Paola Di Lello, David Fry, Colin Garvie, Kuo-Sen Huang, Lin Gao, Christian Tovar, Allen Lovey, Jutta Wanner, and Lyubomir T. Vassilev1

Roche Research Center, Hoffmann-La Roche Inc., Nutley, NJ 07110

Edited by Alan R. Fersht, Medical Research Council Laboratory of Molecular Biology, Cambridge, United Kingdom, and approved May 30, 2012 (received for review March 3, 2012)

Activation of p53 tumor suppressor by antagonizing its negative marginal in many tumor cell lines expressing normal levels of regulator murine double minute (MDM)2 has been considered an MDM2, suggesting that cancer uses other mechanisms to at- attractive strategy for cancer therapy and several classes of p53- tenuate or disable p53 signaling (20), such as the overexpression MDM2 binding inhibitors have been developed. However, these of the other negative p53 regulator, MDMX. High levels of compounds do not inhibit the p53-MDMX interaction, and their MDMX can make MDM2 antagonists, which have effectiveness can be compromised in tumors overexpressing shown very low activity against p53-MDMX binding, ineffective – MDMX. Here, we identify small molecules that potently block in killing cancer cells (21 23). Thus, simultaneous inhibition of p53 binding with both MDM2 and MDMX by inhibitor-driven MDM2 and MDMX is needed to release the full activity of stabilized p53 (15, 17). Therefore, recent efforts have been fo- homo- and/or heterodimerization of MDM2 and MDMX . fi Structural studies revealed that the inhibitors bind into and oc- cused on identi cation of dual MDM2/MDMX antagonists. Because of distinct structural differences between MDM2 and clude the p53 pockets of MDM2 and MDMX by inducing the for- – mation of dimeric protein complexes kept together by a dimeric MDMX in their p53-binding pockets (24 26), small molecules optimized for MDM2 have shown very low affinity for MDMX small-molecule core. This mode of action effectively stabilized p53 (27). For example, the first potent and selective small-molecule and activated p53 signaling in cancer cells, leading to cell cycle MDM2 antagonist, nutlin-3a, has ∼400-fold lower potency arrest and . Dual MDM2/MDMX antagonists restored against MDMX than MDM2 (28). This trend has been followed p53 apoptotic activity in the presence of high levels of MDMX by other MDM2 inhibitors (19). Efforts to identify MDMX- MEDICAL SCIENCES and may offer a more effective therapeutic modality for MDMX- specific inhibitors have recently yielded a class of small molecules overexpressing cancers. with in vitro binding activity in the high nanomolar range but relatively poor cellular potency and uncertain mechanism of he tumor suppressor p53 is a powerful growth-suppressive cellular activity (29). Nearly equipotent MDM2/MDMX peptide Tand proapoptotic protein tightly controlled by its negative inhibitors have been identified and characterized structurally but regulators: murine double minute (MDM)2 and MDMX (1, 2). their activity has been detected only in cell-free systems (30). These proteins bind p53 with their structurally similar N-termi- Recently, a cell-penetrating “stapled” peptide with good MDMX nal domains and effectively inhibit p53 transcriptional activity (1, binding affinity has been identified and evaluated in cancer cells 3). They both possess a RING (really interesting new ) (31). Although cellular potency against p53-MDMX interaction domain in their C termini, but it is only functional in MDM2, has been found adequate, this peptide was unable to disrupt which serves as a specific E3 ligase and main regulator of p53 effectively p53-MDM2 binding, and it has been combined with stability (4, 5). Despite its RING domain, MDMX does not have the MDM2 antagonist, nutlin-3, to assess the antitumor potential an intrinsic ligase activity and does not affect directly p53 sta- of this emerging therapeutic modality. bility (6). However, MDMX can enhance ligase activity of Here, we identify a class of small molecules that can potently MDM2 toward p53 by forming MDM2/MDMX heterodimers (7, inhibit p53 interactions with both MDM2 and MDMX by in- 8). It has been reported that the MDM2/MDMX complex is duced protein dimerization and effectively restore p53 activity in responsible for polyubiquitination of p53, whereas MDM2 alone MDMX-overexpressing cancer cells. We show that antagonizing primarily induces monoubiquitination (9). Targeted disruption both negative p53 regulators significantly improves the apoptotic of MDM2/MDMX heterocomplexes is embryonic-lethal in mice, outcome in cancer cells overproducing MDMX. suggesting that complex formation is essential for p53 regulation in vivo (10). On the other hand, MDM2 can also ubiquitinate Results MDMX and is, therefore, responsible for its stability as well (11, Identification of Indolyl Hydantoins as MDM2/MDMX Antagonists. A 12). MDM2 is a transcriptional target of p53, and both proteins diverse library of small molecules was screened for suppression form an autoregulatory feedback loop by which they mutually of p53-MDMX binding (Table S1). The hits were then tested for control their cellular levels (13). activity against the p53-MDM2 interaction. One series of indolyl The functional relationship between MDM2 and MDMX is hydantoin compounds emerged as potent, dual MDM2/MDMX still being refined at the molecular level, but it is well established that these two negative regulators play a critical role in con- trolling p53 tumor-suppressor function in normal cells (2, 14). Author contributions: B.G., T.T., P.D.L., D.F., C.G., K.-S.H., L.G., C.T., J.W., and L.T.V. de- This is why they are frequently overproduced through gene signed research; B.G., T.T., M.X., C.J., C.L., D.D., P.D.L., C.G., L.G., C.T., and J.W. performed amplification and/or overexpression in tumors that retain wild- research; A.L. and J.W. contributed new reagents/analytic tools; B.G., T.T., M.X., C.L., D.D., type p53 (14). Therefore, antagonizing the binding of MDM2 P.D.L., D.F., C.G., K.-S.H., L.G., C.T., and L.T.V. analyzed data; and B.G. and L.T.V. wrote and MDMX to p53 is expected to restore p53 function and may the paper. offer a strategy for cancer therapy (15). Recently identified Conflict of interest statement: The authors are employees of Hoffmann-La Roche Inc. small-molecule inhibitors of the p53-MDM2 interaction have This article is a PNAS Direct Submission. fi validated this approach, and the rst pharmacological MDM2 Data deposition: The atomic coordinates and structure factors have been deposited in the antagonists are now undergoing clinical evaluation (16, 17). , www.pdb.org (PDB ID codes 3U15 and 3VBG). MDM2 inhibitors have shown effective p53 activation followed 1To whom correspondence may be addressed. E-mail: [email protected] or by cell cycle arrest, induction of apoptosis, and tumor regression [email protected]. in cancer cells with mdm2 gene amplification (18, 19). However, This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. their apoptotic activity has been found to be moderate to 1073/pnas.1203789109/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1203789109 PNAS Early Edition | 1of6 Downloaded by guest on October 1, 2021 antagonists. For example, RO-2443 (Fig. 1A) showed a re- MDMX N-terminal fragment (Fig. 1B) in the presence and ab- markably similar inhibitory activity against both MDM2 (IC50 = sence of the small molecule indicated that: (i) the compound was 33 nM) and MDMX (IC50 = 41 nM) binding to p53. For its size, binding to the p53 pocket; and (ii) there was a substantial up- RO-2443 appeared highly potent (ligand efficiency, defined as field shift for Y63, which is consistent with shielding by an aro- binding energy per heavy atom, is 0.36) and likely to bind into, at matic group. A similar shift had been observed upon binding of most, two of the three subpockets on the surface of MDMX or a p53 peptide to MDMX attributable to shielding by Phe19.In MDM2. These pockets were defined by the original structure of addition, there was an overall resonance broadening in the a p53 peptide bound to MDM2 (32) which showed that there presence of the compound, which was manifested in the HSQC 19 23 were three key residues from the p53 peptide, Phe , Trp , and spectrum as a decrease in cross-peak intensities, suggesting for- 26 Leu . Throughout this report, these binding pockets on the mation of a higher-molecular-mass species. To determine the surface of MDM2 and MDMX are referred to as the Phe, Trp, effect of RO-2443 on the state of MDMX in solution, size-ex- and Leu pockets. clusion chromatography with static light scattering (SEC-SLS) To investigate the mechanism of action of RO-2443, we first 1 15 was performed (Fig. 1C). In the absence of RO-2443, the protein used NMR spectroscopy. The [ H- N]HSQC spectrum of the gave a SLS-calculated mass of 12.8 kDa, which agrees with the theoretical mass of a monomer (12.3 kDa). Addition of RO-2443 resulted in the protein eluting at an earlier elution volume, in- dicating that the shape and/or mass of the protein had changed. A B The SLS-calculated mass of the complex was 24.1 kDa, which F suggests that MDMX forms a dimer when bound to RO-2443. RO-2443 O F G54 100 N Kinetic analysis of the binding of RO-2443 to MDMX (Fig. 1D) G54 N O showed that the binding was in line with a two-molecule binding 80 H model. Isothermal calorimetry (ITC) confirmed a 1:1 ratio be- Cl N 60 H Y63 tween compound and protein but could not distinguish between IC50 40 Y63 1:1, 2:2, or higher-order species (Fig. 1E). It revealed a binding

% of control % of p53-MDM2: 33 nM 20 MDM2 constant (Kd = 78 nM) that is in good agreement with the MDMX p53-MDMX: 41 nM 0 binding assay. The ITC also shows that the binding is completely 1.E-04 1.E-02 1.E+00 dominated by the entropy component, consistent with binding Concentration (uM)

N (ppm) interactions involving primarily hydrophobic effects. Thus, C 15 15N-MDMX a consistent picture emerged that RO-2443 induces some sort of 250 MDMX(15-106)MDMX 15N-MDMX/RO-2443 MDMX(15-106)MDMX + RO-2443 + RO5457707 dimer formation of MDMX and MDM2. 200 MDMX(15-106) + RO5472443 1 H (ppm) Crystal Structures Reveal Tight MDMX Dimer Formation. Crystals of 150 D MDMX bound to RO-2443 were grown that diffracted to rela- 100 tively high resolution (1.8 Å) and molecular replacement with

Absorbtion (mAU) the structure of MDMX bound to a p53 peptide (33) was suc- 50 4000 cessful (Table S2). The structure is comprised of four monomers 0 1-molecular model in the asymmetric unit arranged as a pair of dimers (Fig. 2A). 1.2 1.4 1.6 1.8 2 2.2 2.4 FI Δ These dimers show at their core two molecules of the inhibitor, Elution volume (ml) 2000 2-molecular model each of which has binding interactions to both protein mono- Time (min) mers. For each inhibitor molecule, the indolyl-hydantoin moiety E 0102030400 0.02 occupies the Phe pocket of one protein monomer, whereas the 0 50 100 150 fl 0.00 RO-2443 (uM) di- uoro-phenyl group reaches into the Trp pocket of the other. From a different viewpoint, Fig. 2B shows an overlay of the in- -0.02 5.0 DG hibitor dimer structure with that of the p53 peptide (33). Clearly, -0.04 DH µcal/sec -TDS among the key interactions that the compounds form is an ex- -0.06 tensive aromatic stacking interaction between the two indolyl- -0.08 0.0 hydantoin groups with interplanar distances ranging from 3.3 to 0.2 Kd = 0.078 μM 3.7 Å. The stacking interaction extends to include a tyrosine 0.0 ΔH = -1.13 kcal/mole -0.2 residue (Y63) from each of the protein monomers, resulting in

kcal/mole -TΔS = -8.6 kcal/mole -0.4 a four-level π-sandwich. The small molecule–protein interactions -5.0 ΔG = -9.69 kcal/mole -0.6 were even slightly shorter, ranging from 3.1 to 3.5 Å. Another -0.8 positive interaction between the small molecules is the σ-hole of -1.0 the chlorine attached to the indole ring pointing at the di-fluoro- -1.2 kcal/mole of injectant -10.0 phenyl ring of the other molecule. There are multiple contacts at 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Molar Ratio 4 Å between the chlorine atom and the phenyl ring, which are within the range described by Bissantz et al. (34). Fig. 1. RO-2443 inhibits the interaction of p53 with MDM2 and MDMX. (A) Modeling studies indicated that MDM2 should be able to form Chemical structure and in vitro activity of RO-2443. (B)[1H-15N]HSQC NMR the exact same homodimer and that MDMX and MDM2 could spectra of hzMDMX (His6-15-106, L45V, V95L) in the absence (black contours) form heterodimers. Indeed, crystal structures of MDM2 bound and presence (red contours) of RO-2443. The decrease in cross-peak in- to RO-2443 and other analogs confirm the expected MDM2 tensities observed upon binding of RO-2443 is evidenced by fewer contours homodimer structure (Fig. S1A). Using both structures, a model in the red spectrum. (C) Binding of RO-2443 to MDMX in solution changes its of the potential heterodimer can be assembled showing no se- mobility consistent with protein dimerization. SLS detection of N-terminal rious conflicts (Fig. S1B). Fig. 2C provides a conceptually clearer zebrafish MDMX fragment eluting from a size-exclusion column shows the shift of the protein peak to a position consistent with that of a dimer. (D) picture of the nature of the MDMX/RO-2443 dimer. As sug- Kinetic analysis of the change in Trp fluorescence intensity over a wide range gested by the model (also see Fig. S2A), the p53 binding pockets of RO-2443 concentrations. Curves based on a binding model with the on MDMX (or MDM2) are nearly completely occluded. The monomer or a dimer show that the data are consistent with the dimer model most exposed part of the inhibitor is the methylene bridge

(Kdhigh =0.4μM and Kdlow = 57.2 μM). (E) Isothermal calorimetry shows the between the hydantoin and phenyl groups, which provides impact of binding RO-2443 to hMDMX and that the energy of the in- a means for extending the inhibitor to reach into the Leu pocket teraction is dominated by the entropy component. (Fig. S2B). Partly for this reason, but also to impact the

2of6 | www.pnas.org/cgi/doi/10.1073/pnas.1203789109 Graves et al. Downloaded by guest on October 1, 2021 A Nutlin-3a: IC50 RO-5963: IC50 A MDM2: 18.7 nM MDM2: 17.3 nM OH OH 120 MDMX: 9300 nM 120 MDMX: 24.7 nM O NH 100 100 O F 80 80 N HN 60 60 Cl O F MDM2 N 40 MDMX 40 H % of control MDM2 20 20 MDMX 0 0 1.E- 1.E- 1.E- 1.E- 1.E+ 1.E+ 1.E+1.E- 1.E- 1.E- 1.E- 1.E+ 1.E+ 1.E+ 04 03 02 01Log00 concentration01 02 04 03 02 (µM)01 00 01 02 B C 40 p21 MDM2 B 30 MDMX Control 5963- 2.5 5963- 5 5963-10 5963-20 Nutlin-10 20 MIC-1 p53 BAX p21 10 MDM2

MDMX (fold change) mRNA 0 Actin 1.251234 2.5 5 10 Concentration (µM) D Control 5963-20 Control 5963-10 Nutlin 5963-10 Nutlin 5963-20 Control 5963-20 Nutlin 5963-10 C p53 High Kd MDMX Step 1: XMDM + XMDM MDM2 RO-2443 Actin IP: p53 IP: MDMX

MDMX Fig. 3. RO-5963 stabilizes p53 and activates the p53 pathway in cancer cells. MDMX Low Kd (A) Chemical structure and in vitro inhibitory activity of RO-5963 and the + Step 2: MDMX MDM2 antagonist, nutlin-3a. (B) RO-5963 stabilizes p53 and elevates protein MDMX levels of p53 targets, p21 and MDM2. Log-phase MCF7 cells were incubated with RO-5953 for 24 h, and cell lysates were analyzed by Western blotting. (C) Dose-dependent induction of p53 target in MCF7 cells 24 h post MEDICAL SCIENCES Fig. 2. RO-2443 binds to the p53 pocket of MDMX and induces protein RO-5963 addition. (D) RO-5963 inhibits p53-MDM2 and p53-MDMX binding dimerization. (A) Crystal structure of RO-2443 bound to MDMX. Close-up in cancer cells. MCF7 cells were incubated with 10 μM nutlin-3a and 10 or 20 view of the p53 binding regions of two MDMX molecules forming a dimer. μM RO-5963 for 4 h, and the levels of p53, MDM2, and MDMX were de- One MDMX molecule is shown as a surface rendition (carbon, white; oxygen, termined in protein complexes immunoprecipitated with anti-MDMX or red; nitrogen, blue; and sulfur, yellow). The other MDMX molecule is shown anti-p53 antibodies by Western blotting. as a stick figure with the same color scheme, but the carbon atoms are colored cyan. The two RO-2443 molecules are shown as stick figures. The molecule with cyan-colored carbon atoms is binding with the indole- transcription targets, p21 and MDM2, in a dose-dependent hydantoin moiety in the Phe pocket of MDMX, shown as a stick figure and manner (Fig. 3B). The increase in p21 and MDM2 protein levels the di-fluoro-phenyl group in the Trp pocket of MDMX shown as a surface. was attributable to induction of their transcription as revealed by The molecule with green-colored carbon atoms binds in the reverse mode. the dose-dependent increase of their mRNA and of two other ∼ (B) View of the two inhibitor molecules ( 90° rotation) showing how they p53 transcriptional targets, macrophage inhibitory cytokine-1 relate to the binding of a p53 peptide with carbon atoms colored magenta (MIC-1) and bcl-2 associated X protein (BAX), but not MDMX, (32). The indolyl group of one inhibitor (green) overlays with Phe19 of the peptide, whereas the di-fluoro-phenyl group of the other inhibitor (cyan) which is not under p53 control (Fig. 3C). Stabilization and ac- overlays with Trp23. It is worth noting that the chlorine atom of the 6-chloro- tivation of p53 was induced by disrupting its interaction with tryptophan of the peptide is nearly coincident with the parafluoro atom of MDM2 and MDMX in MCF7 cells as demonstrated by immu- the inhibitor. (C) Dimer model for binding of RO-2443 to MDMX. This ren- noprecipitation of either p53 or MDMX proteins from cells ex- dition of the possible steps in the formation of the dimer indicates that posed to RO-5963 followed by Western analysis (Fig. 3D). At 20 monomeric interactions likely form first, followed by a pairing of monomers μM, RO-5963 was equivalent to 10 μM nutlin in inhibiting p53- to form the dimer. This is based on the lack of any detection of dimer MDM2 binding and also effectively blocked p53-MDMX binding interactions by the small molecules alone. A binding model in which the at both 10 and 20 μM concentration. As expected, the MDM2- formation of the monomer is followed by the addition of a second small specific inhibitor, nutlin-3a, showed no effect on the p53-MDMX molecule and subsequent addition of a vacant copy of MDMX cannot be interaction. A notable increase in MDMX and MDM2 proteins ruled out. pulled down by immunoprecipitated MDMX (Fig. 3D, Right) was consistent with predicted RO-5963-induced formation of physicochemical properties of the compounds, this site was tar- MDMX/MDMX and MDMX/MDM2 dimers in MCF7 cells. geted for modification, resulting in analogs such as RO-5963. RO-5963 Activates p53 Pathway in Cancer Cells Expressing Wild-Type RO-5963 Inhibited p53 Binding to MDM2 and MDMX in Cancer Cells. p53. Inhibition of p53-MDM2 binding should induce p53 sig- RO-2443 showed potent MDM2/MDMX inhibitory activity naling only if the cells express wild-type but not mutant p53, in vitro, but poor water solubility did not allow for a meaningful which generally loses its transcriptional activity. Indeed, the EC50 ∼ assessment of its cellular activity. Further chemical optimization of RO-5963 was 10-fold lower in cancer cells expressing wild- of RO-2443 yielded RO-5963 (Fig. 3A), a close analog with type p53 compared with mutant p53 cells (Fig. 4A) and only slightly increased potency but substantially improved solubility. activated the p53 transcriptional targets p21 and MDM2 in p53 In the same binding assay, RO-5963 showed p53-MDM2 in- wild-type but not p53 mutant cancer cells (Fig. S3). RO-5963 hibitory activity (IC50, ∼17 nM) similar to that of nutlin-3a (IC50, affected the viability of HCT116 cells but not their nutlin-re- ∼19 nM) (Fig. 3A). Its p53-MDMX inhibitory activity (IC50, ∼24 sistant clone HCT116R1, which has lost ability to induce p53 nM) was nearly equivalent to MDM2 activity but ∼400-fold response (Fig. S4). Treatment of four cancer cell lines with RO- Ser15 better than the MDMX potency of nutlin-3a (IC50, ∼9 μM). RO- 5963 did not increase the levels of p53 phosphorylation, 5963 penetrated MDMX-overexpressing breast cancer cells suggesting that p53 activation is not caused by genotoxic stress (MCF7), stabilized p53, and elevated protein levels of its induced by the compound (Fig. 4B).

Graves et al. PNAS Early Edition | 3of6 Downloaded by guest on October 1, 2021 A B Dox effectively activate p53 and elevate p21 and MDM2 levels, sug- Mut-p53 gesting that it penetrates well cultured cells and can be used in 100 diverse cellular context (Fig. 4D). H460 80 MCF-7 RKO HCT116 HCT116 RKO WT-p53 - + - + - + - + - - 5963 60 p53 RO-5963 Effectively Activates Main Functions of the p53 Pathway in Cancer Cells. One of the main functions of activated p53 is in- 40 MCF-7 p-p53Ser15 HCT-116 duction of cell cycle arrest. As previously demonstrated by the RKO % of control p21 20 SW480 specific MDM2 antagonist, nutlin-3 (20), RO-5963 potently MDA-MB435 MDM2 0 arrested cell cycle progression in exponentially growing cancer Actin 0110 cells in G1 and G2 phase, effectively depleting the S phase Concentration (µM) compartment (Fig. S5A). Induction of apoptosis is another major C p53 function that is frequently altered in cancers expressing wild- - 1.25 2.5 5 10 - 1.25 2.5 5 10 - 1.25 2.5 5 10 µM RO-5963 - - - - - + + + + + + + + + + Nutlin type p53 (20). Similar to MDM2, MDMX overexpression has p53 been shown to effectively disable this function by inhibiting p53 p21 transcriptional activity (6, 14). Therefore, MDMX could be MDM2 a barrier to p53 apoptotic activity even in the presence of MDM2 MDMX – Actin antagonists (21 23). The SJSA1 osteosarcoma line, which expresses very high levels of MDM2 protein, is presumed to be G401 H460 D free of other defects in the p53 pathway and has shown a robust apoptotic response to the specific MDM2 antagonists (20). However, its engineered clone, SJSA-X, is nearly completely RKO LnCaP U2OS A498 22RV1 HCT116 H460 LOX MCF7 A549 G401 p53 resistant to nutlin because of the high levels of exogenously p21 expressed MDMX from a CMV promoter (23). Therefore, the MDM2 SJSA-X clone offers an excellent mechanistic model for assess- Actin ing the cellular activity of p53-MDMX inhibitors. The high levels RO-5963 - + - + - + - + - + - + - + - + - + - + - + of both MDM2 and MDMX proteins in SJSA-X cells represent a fairly high hurdle to dual inhibitors. When RO-5963 was tested Fig. 4. RO-5963 activates p53 signaling in diverse cellular context by a nongenotoxic mechanism. (A) Antitumor activity of RO-5963 depends on for apoptotic activity on the parental cell line, SJSA-V, it showed the p53 status. Viability of three wild-type p53 (MCF7, HCT116, RKO) and slightly lower but still strong Annexin V signal compared with two mutant p53 (SW480, MDA-MB-435) cancer cell lines was determined by nutlin (Fig. S5B). As expected, similar apoptotic activity was the CellTiter-Glo assay after 5 d of incubation with RO-5963 and expressed as measured also in the SJSA-X clone in which nutlin was practi- percentage of controls ± SD. (B) RO-5963 does not induce genotoxic re- cally inactive. The apoptotic activity of RO-5963 was dose-de- sponse in cancer cells. Cells were incubated with 10 μM RO-5963 or 1 μM pendent and was enhanced when both molecules were combined. doxorubicin for 24 h, and the levels of total and Ser15-phosphorylated p53 Western blotting of SJSA-X cell lysates revealed that 20 μM RO- were determined by Western blotting. (C) Binding of RO-5963 prevents 5963 induced p53 accumulation comparable to 10 μM nutlin-3a MDM2-mediated degradation of MDMX. G401 and H460 cells were in- because p53 stabilization is caused by disruption of the p53- cubated with the indicated concentrations of RO-5963 with or without MDM2 interaction (Fig. 5C). However, p53 transcriptional ac- 10 μM nutlin-3a, and relative levels of p53, p21, MDM2, and MDMX were tivity, indicated by p21 and MDM2 levels, was higher in the determined by Western blotting. Blots representative of three independent presence of RO-5963, presumably because of blocking not only experiments are shown. (D) RO-5963 activates p53 signaling in multiple p53-MDM2 but also p53-MDMX binding and liberating the el- cancer cell lines with wild-type p53. Exponentially growing cancer cell were μ evated p53 protein from both inhibitors. Protein levels of p21 incubated with 10 M RO-5963 for 24 h, and the relative levels of p53, p21, and MDM2 were increased further by combining RO-5963 with and MDM2 were determined by Western blotting. nutlin, reflecting the increased transcriptional activity of p53 (Fig. S5C) that explains enhanced apoptotic activity of the combination (Fig. S5B). These results indicate that RO-5963 can By disrupting p53-MDM2 binding and the autoregulatory effectively inhibit both p53-MDM2 and p53-MDMX binding and feedback loop, small-molecule antagonists (e.g., nutlins) can ele- can induce apoptosis in a cancer cell model overexpressing high vate MDM2 protein levels and, thus, facilitate MDMX ubiquiti- levels of both negative p53 regulators. nation and degradation. As a result, MDMX protein levels were found reduced in many cancer cell lines in the presence of nutlin-3 RO-5963 Overcomes the Resistance of MDMX-Overexpressing Cancer (28). If RO-5963 induces homo- or heterodimerization of MDM2 Cells to MDM2 Antagonists. Cancer cells overexpressing MDMX and MDMX inside living cells, that might interfere with MDM2 have substantially reduced apoptotic response to small-molecule ligase activity and/or its ability to ubiquitinate MDMX. To in- MDM2 antagonists (e.g., nutlins) (21–23). This resistance may be vestigate this possibility, we incubated G401 cancer cells with in- partly attributable to the inability of nutlin to inhibit p53-MDMX creasing concentrations of RO-5963 in the presence or absence of binding, leading to incomplete restoration of p53 activity. MCF7 μ 10 M nutlin-3a. As expected, nutlin substantially reduced breast cancer cell line represents such a MDMX-dependent MDMX protein levels (Fig. 4C). RO-5963 dose-dependently in- model. These cells have high levels of MDMX protein and are creased p53, MDM2, and p21 levels but only slightly reduced fairly insensitive to nutlin-induced apoptosis. Therefore, we first MDMX. However, it protected MDMX from nutlin-induced tested the apoptotic activity of RO-5963 in MCF7 and two other degradation by MDM2 in both G401 and H460 cells (Fig. 4C). breast cancer cell lines with low (ZR75-1) and intermediate These results suggest that RO-5963-induced dimerization of (ZR75-30) MDMX levels (35). As an MDM2-only inhibitor (20, MDM2 and MDMX is the likely cause for protection of MDMX 28), nutlin-3a had relatively low apoptotic activity in MCF7 cells from MDM2-mediated degradation. However, despite increased and practically no activity in ZR75-30 cells (Fig. 5A). ZR75-30 MDMX levels in cells exposed to RO-5963, compared with nutlin, cells express moderate levels of MDMX but very low levels of MDMX was prevented from binding to and inhibiting p53 in the MDM2, suggesting that MDMX plays an important role in p53 presence of the dual inhibitor. regulation in these cells. RO-5963 (20 μM) showed much higher Next, we tested RO-5963 in a wider panel of 11 solid tumor apoptotic activity than nutlin in both MCF7 and ZR75-30 cell cell lines expressing wild-type p53 and representing diverse tu- lines. These results suggest that RO-5963 is capable of releasing mor types: breast (MCF7), prostate (LNCaP, 22Rv1), colon p53 from MDMX inhibition and restoring its activity. Although (HCT116, RKO), lung (H460, A549), kidney (A498), osteosar- thisledtoefficient apoptosis and cell death in the MCF7 line (Fig. coma (U2OS), melanoma (LOX). The dual inhibitor was able to 5 A and B), ZR75-30 cells had an enhanced but incomplete

4of6 | www.pnas.org/cgi/doi/10.1073/pnas.1203789109 Graves et al. Downloaded by guest on October 1, 2021 interaction is critical for stabilization of p53 and that simulta- A B neous inhibition of p53-MDMX binding can enhance p53 apo- 80 Series1DMSO ptotic activity in cancer cells with MDMX overexpression. 70 Series2Nutlin (10 μM) However, inhibition of both MDM2 and MDMX does not sub- Series3RO-5963 (10 μM) 60 Series4RO-5963 (20 μM) stantially enhance p53-dependent apoptotic response in cancer 50 Control cells with normal or low levels of MDMX that is likely affected 40 by other factors than MDMX. 30 Discussion 20 fi

Annexin V+ cells (%) The role of MDMX in the ne regulation of p53 is still emerging, 10 but it is an established fact that MDMX overexpression can 0 ZR-75- 1 ZR-75-30 MCF- 7 block p53 function and render cancer cells resistant to MDM2 RO-5963 ZR75-1 ZR75-30 MCF7 antagonists (6, 15). Even normal levels of MDMX could partially silence activated p53 because all known inhibitors of the p53- C MDM2 interaction are unable to liberate p53 from the remain- ing MDMX, suggesting that dual MDM2/MDMX inhibitors may MDM2 substantially improve the outcome of this p53 activation strategy. However, recent efforts to develop inhibitors of p53 binding with MDMX both MDM2 and MDMX have been hindered by the structural Actin differences in the p53 pockets of the proteins (26). Using high- throughput screening and a diverse small-molecule library, we D identified a class of indolyl-hydantoin compounds that are 80 Series1DMSO roughly equipotent. The series demonstrated clear SAR and the 70 Series2Nutlin (10 μM) better analogs showed remarkable potency given their small size 60 Series3RO-5963 (10 μM) and the ability to occupy only two of the three surface subpockets 50 Series4RO-5963 (20 μM) in the p53 binding region. The presence of the 6-chloro-indole 40 group provided the expectation that the binding mode could be 30 predicted because this same group is used in some of the known 20 MDM2 antagonists (36). MEDICAL SCIENCES Annexin V+ cells (%) 10 The predicted binding mode of the dual MDM2/MDMX antagonists placed the chloro-indole moiety in the Trp pocket 0 with the remainder of the compound extending into the Phe 9 S -5 0 16 4T O l pocket. This was, in general, consistent with the available in- RKO 7 2 -435 A54 H460 T1 1 AGS G401 U me B SW48 HC LS Sk -M formation. First, the fact that the compounds were equipotent A D M against MDM2 and MDMX made it less likely that they bound in the Leu pocket, where the two protein structures diverge the Fig. 5. Apoptotic activity of RO-5963 and MDMX status. (A) Breast cancer most. Secondly, the large up-field shift in the NMR spectrum for cells respond to RO-5963 depending on the levels of MDM2 and MDMX. Y63 meant that this residue was likely interacting with an aro- Three cell lines with variable ratios of MDMX/MDM2 proteins were exposed matic group from the inhibitor, the di-fluoro-phenyl group, in to nutlin-3a (10 μM) and RO-5963 (10 or 20 μM) for 48 h, and the percentage this model. However, the actual structure turned these pre- ± of apoptotic cells ( SD) was determined by the Annexin V assay. (B) Cyto- dictions completely upside down (Fig. 2A). The indole and toxicity of RO-5963 on MCF7 cells. Phase-contrast images were taken 48 h μ hydantoin moieties remain essentially coplanar and occupy an after addition of 20 M RO5963. (C) Relative protein levels of MDM2 and extended Phe pocket. In this orientation, it is the indolyl- MDMX in a panel of solid tumor cell lines. Cells lines at subconfluent stage of growth were analyzed for protein levels by Western blotting. (D) Apoptotic hydantoin that is providing the shielding for the Y63 residue. It also forms a stacking interaction with the phenolic ring of Y63 response to RO-5963 and nutlin in a panel of cancer cell lines with wild-type fl and mutant p53. Log-phase cells were incubated with nutlin-3a and RO-5963 (Fig. S1A). Most surprisingly, the di- uoro-phenyl group of one for 48 h and analyzed for apoptosis as in A. Two mutant p53 cell lines, MDA- molecule bound to one protein monomer reaches over to bind MB-435 and SW480, were included as controls. into the Trp pocket of the other protein monomer (Fig. 2B). Although unusual, the structure nicely explained the preference for fluorine in the paraposition for binding to MDMX. It also apoptotic response. This is likely attributable to other abnormal- provided an explanation for the preference for chlorine at the 6- ities in cancer signaling that can attenuate p53 apoptotic pathways position of the indole. as seen in many epithelial cancer cell lines with normal MDM2 Although MDM2 and MDMX naturally form homodimers and relatively low MDMX levels (20, 28). Interestingly, the dual and heterodimers, these are driven by the C-terminal RING antagonist was less active than nutlin in the ZR75-1 cell line domains (6, 37). Thus, the dimerization of the N-terminal p53- binding domains in this fashion has not been observed pre- expressing barely detectable MDMX levels. This may result from fi the fact that nutlin is slightly more effective than RO-5963 at viously. It represents a signi cant advantage in that small mol- ecules are able to achieve potent binding and inhibitory effects inhibiting p53-MDM2 binding in cells. and to do it in a way that avoids addressing the Leu subpocket, Next, we tested the apoptotic activity of RO-5963 and nutlin where the two protein structures are the most different. As against a randomly selected panel of nine cancer cell lines a result, these compounds are equipotent against MDM2 and expressing wild-type p53 and two with mutant p53. As expected, MDMX. Nevertheless, it is possible to modify these compounds no apoptotic activity was detected in the mutant p53 lines, and to enable them to address the Leu pocket and RO-5963 (Fig. 3A) variable levels of apoptosis were measured in the wild-type lines represents one such approach. (Fig. 5D). RO-5963 showed better apoptotic activity than nutlin RO-5963 penetrated cancer cells and inhibited the binding of in 4/9 (H460, RKO, LS174T, AGS) and similar or slightly lower p53 to MDM2 and MDMX. Although nearly equipotent in vitro, in five of nine wild-type p53 cell lines. Three out of the four lines RO-5963 was more effective in disrupting the p53-MDMX in- with enhanced apoptosis had relatively high levels of MDMX. teraction than p53-MDM2 interaction in cultured cancer cells. Overall, MDMX protein levels correlated with the enhanced One possible reason for this is the increasing cellular levels of apoptotic response to the dual MDM2/MDMX antagonists (Fig. MDM2 in the presence of MDM2 inhibitors. Whereas MDMX 5 C and D). These data suggest that inhibition of p53-MDM2 levels are generally stable, MDM2 protein levels rise as a result

Graves et al. PNAS Early Edition | 5of6 Downloaded by guest on October 1, 2021 of inducing its expression by the accumulating p53. Increasing using RNA interference to address the role of MDMX in the MDM2 protein may change the balance between homo- and response to nutlin (23, 28, 35). heterodimers induced by RO-5963 and make MDM2 inhibitory Nearly one-fifth of breast, colon, and lung cancers overexpress activity a limiting factor in the course of exposure to the dual MDMX because of gene amplification (38), and that could be inhibitor. This is a likely reason why addition of nutlin further the only abnormality in their p53 signaling as proposed for enhanced the apoptotic outcome of RO-5963 treatment espe- MDM2-amplified tumors (20). Our data suggest that patients cially in the MDM2 amplified SJSA-X line (Fig. S5B). with this signature may greatly benefit from treatment with dual Forced or natural overexpression of MDMX can significantly MDM2/MDMX antagonists. The identification of the indolyl impede the activity of MDM2 antagonists (19, 21–23). These hydantoin class of dual inhibitors is the first step toward de- agents have been optimized for binding to the p53 pocket of velopment of MDM2/MDMX-targeted cancer therapy. Further MDM2 and are practically inactive against the p53-MDMX in- optimization of the potency and pharmacological properties of teraction (18, 19, 28). Unobstructed binding of MDMX to p53 this chemical class should allow the extension of our observations could silence its transcriptional activity and may alter other to in vivo cancer models and eventually the clinic. properties, leading to inadequate activation of p53 functions. Materials and Methods Therefore, the effectiveness of MDM2 antagonists in tumors with high levels of MDMX will be diminished. Our experimental Materials and detailed methods are available in SI Materials and Methods. results with the dual MDM2/MDMX inhibitor are in full Cell proliferation/viability was evaluated by the methyl-thiazolyl-tetra- zolium (MTT) and CellTiter-Glo (Promega) assays. Cell cycle analysis, Annexin agreement with this prediction. RO-5963 restored p53 tran- V assays, and Western blotting were performed as described previously (20). scriptional activity and overcame the apoptotic resistance of MDM2-p53 and MDMX-p53 binding was assessed by time-resolved (TR)-FRET MDMX-overexpressing cell line, SJSA-X, to nutlin-3 (Fig. 5C). binding and fluorescence quenching assays. NMR spectra were collected The same is true in the “naturally” MDMX-overexpressing using 15N-labeled humanized-zebrafish (hz)MDMX. For cocrystallization with breast cancer cell lines MCF7 and ZR75-30 (Fig. 5). Both cell RO-2443, a 1.3:1 stoichiometric excess of compound was added to HDMX lines exhibit higher MDMX to MDM2 ratios, suggesting that protein. The protein construct (14-111,C17S) was as described in ref. 33. Crys- MDMX may play a role in their resistance to nutlin-induced tallization conditions, data collection, molecular replacement, and refinement apoptosis. The relatively incomplete apoptotic response in were performed as detailed in SI Materials and Methods. Statistics for the ZR75-30 line is likely attributable to factors other than MDMX refined model are in Table S2. attenuating p53-dependent apoptosis signaling. In MCF7 cells, used routinely as a cellular model of MDMX-overexpressing ACKNOWLEDGMENTS. We thank Ann Petersen and Honju Li for chemical synthesis; Charles Belunis for protein purification; Sonal Sojitra, John Koss, breast cancer, RO-5963 showed massive cell death with nearly all and Stephen Wasserman for X-ray data collection; Geoffrey Wahl for the gift cells losing their viability within 72 h of treatment (Fig. 5B). of SJSA-X cells; and Nader Fotouhi for his suggestions and critical reading of These results are in agreement with previously published data the manuscript.

1. Harris SL, Levine AJ (2005) The p53 pathway: Positive and negative feedback loops. 21. Koblish HK, et al. (2006) Benzodiazepinedione inhibitors of the Hdm2:p53 complex Oncogene 24:2899–2908. suppress human tumor cell proliferation in vitro and sensitize tumors to doxorubicin 2. Wade M, Wang YV, Wahl GM (2010) The p53 orchestra: Mdm2 and Mdmx set the in vivo. Mol Cancer Ther 5:160–169. tone. Trends Cell Biol 20:299–309. 22. Patton JT, et al. (2006) Levels of HdmX expression dictate the sensitivity of normal and 3. Toledo F, Wahl GM (2006) Regulating the p53 pathway: In vitro hypotheses, in vivo transformed cells to Nutlin-3. Cancer Res 66:3169–3176. – veritas. Nat Rev Cancer 6:909 923. 23. Wade M, Wong ET, Tang M, Stommel JM, Wahl GM (2006) Hdmx modulates the 4. Marine JC, Lozano G (2010) Mdm2-mediated ubiquitylation: p53 and beyond. Cell outcome of p53 activation in human tumor cells. J Biol Chem 281:33036–33044. – Death Differ 17:93 102. 24. Popowicz GM, et al. (2007) Molecular basis for the inhibition of p53 by Mdmx. Cell 5. Lenos K, Jochemsen AG (2011) Functions of MDMX in the modulation of the p53- Cycle 6:2386–2392. response. J Biomed Biotechnol 2011:876173. 25. Popowicz GM, Czarna A, Holak TA (2008) Structure of the human Mdmx protein 6. Marine JC, Jochemsen AG (2005) Mdmx as an essential regulator of p53 activity. Bi- bound to the p53 tumor suppressor transactivation domain. Cell Cycle 7:2441–2443. ochem Biophys Res Commun 331:750–760. 26. Riedinger C, McDonnell JM (2009) Inhibitors of MDM2 and MDMX: A structural 7. de Graaf P, et al. (2003) Hdmx protein stability is regulated by the ubiquitin ligase – activity of Mdm2. J Biol Chem 278:38315–38324. perspective. Future Med Chem 1:1075 1094. 8. Kawai H, Lopez-Pajares V, Kim MM, Wiederschain D, Yuan ZM (2007) RING domain- 27. Vassilev LT (2005) p53 Activation by small molecules: Application in oncology. J Med – mediated interaction is a requirement for MDM2’s E3 ligase activity. Cancer Res 67: Chem 48:4491 4499. 6026–6030. 28. Xia M, et al. (2008) Elevated MDM2 boosts the apoptotic activity of p53-MDM2 9. Wang X, Wang J, Jiang X (2011) MdmX protein is essential for Mdm2 protein-medi- binding inhibitors by facilitating MDMX degradation. Cell Cycle 7:1604–1612. ated p53 polyubiquitination. J Biol Chem 286:23725–23734. 29. Reed D, et al. (2010) Identification and characterization of the first small molecule 10. Huang L, et al. (2011) The p53 inhibitors MDM2/MDMX complex is required for inhibitor of MDMX. J Biol Chem 285:10786–10796. control of p53 activity in vivo. Proc Natl Acad Sci USA 108:12001–12006. 30. Pazgier M, et al. (2009) Structural basis for high-affinity peptide inhibition of p53 11. Pan Y, Chen J (2003) MDM2 promotes ubiquitination and degradation of MDMX. Mol interactions with MDM2 and MDMX. Proc Natl Acad Sci USA 106:4665–4670. Cell Biol 23:5113–5121. 31. Bernal F, et al. (2010) A stapled p53 helix overcomes HDMX-mediated suppression of 12. Kawai H, et al. (2003) DNA damage-induced MDMX degradation is mediated by p53. Cancer Cell 18:411–422. MDM2. J Biol Chem 278:45946–45953. 32. Kussie PH, et al. (1996) Structure of the MDM2 oncoprotein bound to the p53 tumor 13. Momand J, Zambetti GP, Olson DC, George D, Levine AJ (1992) The mdm-2 oncogene suppressor transactivation domain. Science 274:948–953. product forms a complex with the p53 protein and inhibits p53-mediated trans- 33. Kallen J, et al. (2009) Crystal Structures of Human MdmX (HdmX) in Complex with p53 – activation. Cell 69:1237 1245. Peptide Analogues Reveal Surprising Conformational Changes. J Biol Chem 284: 14. Marine JC, et al. (2006) Keeping p53 in check: Essential and synergistic functions of 8812–8821. – Mdm2 and Mdm4. Cell Death Differ 13:927 934. 34. Bissantz C, Kuhn B, Stahl M (2010) A medicinal chemist’s guide to molecular inter- 15. Wade M, Wahl GM (2009) Targeting Mdm2 and Mdmx in cancer therapy: Better living actions. J Med Chem 53:5061–5084. through medicinal chemistry? Mol Cancer Res 7:1–11. 35. Lam S, et al. (2010) Role of Mdm4 in drug sensitivity of breast cancer cells. Oncogene 16. Vassilev LT (2007) MDM2 inhibitors for cancer therapy. Trends Mol Med 13:23–31. 29:2415–2426. 17. Brown CJ, Lain S, Verma CS, Fersht AR, Lane DP (2009) Awakening guardian angels: 36. Yu S, et al. (2009) Potent and orally active small-molecule inhibitors of the MDM2-p53 Drugging the p53 pathway. Nat Rev Cancer 9:862–873. – 18. Vassilev LT, et al. (2004) In vivo activation of the p53 pathway by small-molecule interaction. J Med Chem 52:7970 7973. antagonists of MDM2. Science 303:844–848. 37. Linke K, et al. (2008) Structure of the MDM2/MDMX RING domain heterodimer re- 19. Shangary S, et al. (2008) Temporal activation of p53 by a specific MDM2 inhibitor is veals dimerization is required for their ubiquitylation in trans. Cell Death Differ 15: selectively toxic to tumors and leads to complete tumor growth inhibition. Proc Natl 841–848. Acad Sci USA 105:3933–3938. 38. Danovi D, et al. (2004) Amplification of Mdmx (or Mdm4) directly contributes to tu- 20. Tovar C, et al. (2006) Small-molecule MDM2 antagonists reveal aberrant p53 signaling mor formation by inhibiting p53 tumor suppressor activity. Mol Cell Biol 24: in cancer: Implications for therapy. Proc Natl Acad Sci USA 103:1888–1893. 5835–5843.

6of6 | www.pnas.org/cgi/doi/10.1073/pnas.1203789109 Graves et al. Downloaded by guest on October 1, 2021