Investigation of the inhibitory mechanism of apomorphine against MDM2–p53 interaction

Article abstract of DOI:10.1016/j.bmcl.2017.03.082

Mirror-image screening using D-proteins is a powerful approach to provide mirror-image structures of chiral natural products for drug screening. During the course of our screening study for novel MDM2–p53 interaction inhibitors, we identified that NPD6878 (R-(?)-apomorphine) inhibited both the native L-MDM2–L-p53 interaction and the mirror-image D-MDM2–D-p53 interaction at equipotent doses. In addition, both enantiomers of apomorphine showed potent inhibitory activity against the native MDM2–p53 interaction. In this study, we investigated the inhibitory mechanism of both enantiomers of apomorphine against the MDM2–p53 interaction. Achiral oxoapomorphine, which was converted from chiral apomorphines under aerobic conditions, served as the reactive species to form a covalent bond at Cys77 of MDM2, leading to the inhibitory effect against the binding to p53.

Full text of DOI:10.1016/j.bmcl.2017.03.082

Bioorganic & Medicinal Chemistry Letters 27 (2017) 2571–2574
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Investigation of the inhibitory mechanism of apomorphine against
MDM2–p53 interaction
Hiroyuki Ishiba a, Taro Noguchi a, Keitou Shu a, Hiroaki Ohno a, Kaori Honda b, Yasumitsu Kondoh b,
Hiroyuki Osada b, Nobutaka Fujii a, Shinya Oishi a,
⇑
a Graduate School of Pharmaceutical Sciences, Kyoto University, Sakyo-ku, Kyoto 606-8501, Japan
b Chemical Biology Research Group, RIKEN Center for Sustainable Resource Science, Wako, Saitama 351-0198, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Mirror-image screening using D-proteins is a powerful approach to provide mirror-image structures of
Received 6 February 2017
Revised 23 March 2017
Accepted 27 March 2017
Available online 29 March 2017
chiral natural products for drug screening. During the course of our screening study for novel MDM2–
p53 interaction inhibitors, we identified that NPD6878 (R-(À)-apomorphine) inhibited both the native
L-MDM2–L-p53 interaction and the mirror-image D-MDM2–D-p53 interaction at equipotent doses. In
addition, both enantiomers of apomorphine showed potent inhibitory activity against the native
MDM2–p53 interaction. In this study, we investigated the inhibitory mechanism of both enantiomers
of apomorphine against the MDM2–p53 interaction. Achiral oxoapomorphine, which was converted from
chiral apomorphines under aerobic conditions, served as the reactive species to form a covalent bond at
Cys77 of MDM2, leading to the inhibitory effect against the binding to p53.
Keywords:
MDM2
p53
Mirror-image screening
Apomorphine
Oxoapomorphine
Michael acceptor
Ó 2017 Elsevier Ltd. All rights reserved.
Mirror-image screening is a powerful approach for identifying
potential pharmaceutical candidates from homochiral
resources.1–3 This unique concept was first demonstrated by the
identification of
-peptide ligands for the c-Src SH3 domain.4 By
taking advantage of the synthetic mirror-image protein ( -SH3)
of the c-Src SH3 domain, several -SH3 selective binding peptides
were isolated from a phage display library. The mirror-image pep-
tide ( -peptide) of the phage-displayed peptide reproduced the
oligonucleotides are resistant against degradation by naturally
occurring enzymes such as peptidases and nucleases, respectively,
they have significant advantages over native biomolecules for use
in clinical applications.
Recently, we established a novel approach for screening a vir-
tual mirror-image library of chiral natural product derivatives.16,17
In the first case study, we explored potential inhibitors against
MDM2–p53 interactions18 (Fig. 1). Initially, we synthesized a fluo-
D
D
D
D
binding affinity to the native c-Src SH3 domain. Currently, mir-
ror-image phage display experiments have been applied to a num-
ber of target molecules including HIV-1 gp41, amyloid Ab1–42, and
an oncoprotein MDM2.5–10 Mirror-image nucleotide ligands
(named Spiegelmers) have also been characterized via similar
rescent
thetic
D-protein of MDM2 (D-MDM2). Using the resulting syn-
D
-MDM2 protein, more than 22,000 natural products and
derivatives were screened by a chemical array technology.19,20
The secondary analysis of the screening hits by a fluorescence
polarization assay revealed that an
NP843 possessed selective inhibitory activity against the
-MDM2– -p53 interaction. The subsequent synthetic study of
a-tocopherol derivative
selection and amplification processes.11 A number of
L
-RNA Spie-
gelmers were identified by screening using mirror-image biomole-
cules such as -adenosine, -arginine and
-GnRH.12–14 Similarly,
identifications of -DNA ligands for vasopressin and GnRH were
D
D
L
D
D
the mirror-image structure of NP843 followed by biological evalu-
ations led to the identification of ent-NP843 as a novel MDM2–p53
inhibitor. In this process, two syntheses of mirror-image molecules
L
also reported, which inhibited the respective receptor activations
by native ligands.14,15 Because these mirror-image peptides and
including D-MDM2 protein and ent-NP843 facilitated the screening
without synthesis of a large number of mirror-image natural
products.
Abbreviations: MDM2, mouse double minute
polarization; SPR, surface plasmon resonance.
2 homolog; FP, fluorescence
During the course of the screening study, we revealed that
NPD6878 (apomorphine) was an MDM2–p53 inhibitor candidate
with high potency.16 R-(À)-apomorphine inhibited both the
⇑
Corresponding author.
E-mail address: soishi@pharm.kyoto-u.ac.jp (S. Oishi).
http://dx.doi.org/10.1016/j.bmcl.2017.03.082
0960-894X/Ó 2017 Elsevier Ltd. All rights reserved.

2572
H. Ishiba et al. / Bioorganic & Medicinal Chemistry Letters 27 (2017) 2571–2574
Fig. 2. Structures of apomorphine enantiomers.
Fig. 1. Concept of mirror-image screening.
Table 1
Inhibitory activities of R-(À)-apomorphine and S-(+)-apomorphine reagents.
Compounds
IC50
(l
M)a
L
-MDM2–
L-p53
D-MDM2–D-p53
R-(À)-apomorphine (1a)b
S-(+)-apomorphine (1b)b
0.215
0.175
0.195
–
c
a
IC50 values were determined by the fluorescence polarization (FP) assay using
0.5 nM FAM labeled P4 and 10 nM MDM2.
Fig. 3. Time course of R-(À)-apomorphine oxidation under aerobic conditions (in
50% PBS/CH3CN). Closed circle, R-(À)-apomorphine (1a); open circle, oxoapomor-
phine (2).
b
DMSO solutions of the reagents were employed for the assays.
c
Not determined.
D-MDM2–D-p53 and L-MDM2–L-p53 interactions (Table 1). Addi-
tionally, the enantiomer, S-(+)-apomorphine, also showed equipo-
tent inhibitory activity against the MDM2–p53 interaction (Fig. 2).
We focused on how both apomorphine enantiomers with a unique
scaffold worked as MDM2–p53 inhibitors. In this study, we inves-
tigated the inhibitory mechanism of apomorphine against the
MDM2–p53 interaction.
In the initial stages of the study, we focused our attention on the
color change of a solution of apomorphine reagent during the
assay. Although the blue and/or green color did not show signifi-
cant interference of the assay systems by fluorescence polarization,
we suspected that the oxidation of R-(À)-apomorphine might
occur during the assay to provide achiral oxoapomorphine.21
Indeed, when R-(À)-apomorphine was incubated in a PBS-CH3CN
(1:1) mixture under air at room temperature, this compound was
gradually oxidized into oxoapomorphine (Fig. 3). More than 80%
of R-(À)-apomorphine was transformed into dark blue oxoapomor-
phine after 24 h, suggesting that the inhibitory activity against the
MDM2–p53 interaction was derived from planar oxoapomorphine
3.
Next, we assessed the biological activities of oxoapomorphine,
which was synthesized via oxidation of commercially available
R-(À)-apomorphine (Scheme 1).22 To avoid interference from the
color of oxoapomorphine, we evaluated the inhibitory activity
against the MDM2–p53 interaction by surface plasmon resonance
(SPR) analysis, in which binding of MDM2 to immobilized p53 pep-
tide can be detected. The synthetic oxoapomorphine showed
potent inhibitory activity against the MDM2–p53 interaction,
while no inhibitory activity was observed with a freshly prepared
apomorphine solution (Fig. 4). After incubation in assay buffer
for 24 h, the apomorphine sample also reproduced the inhibitory
activity, suggesting that the oxidized form was the reactive species
responsible for bioactivity.23 Interestingly, the dose-response curve
Scheme 1. Synthesis of oxoapomorphine.
of inhibition did not fit to an ideal sigmoidal shape. Previously,
Smith et al. reported that the o-quinone substructure of oxoapo-
morphine could be susceptible to Michael addition of amino and
sulfhydryl groups on serum proteins.24 There are a number of
nucleophilic groups in MDM2, including a single Cys residue at
position 77 and nine Lys residues, and one or some of these func-
tional group(s) might contribute to the direct interaction with
oxoapomorphine. This postulate is also supported by the report
by Bista et al., where covalent bond formation via Cys77 of
MDM2 with Michael acceptors attenuated the interaction between
MDM2–p53.25
To elucidate the interactive residue in MDM2 with oxoapomor-
phine, we attempted to evaluate the inhibitory activities against
interaction between several mutant MDM2 proteins and p53. We

H. Ishiba et al. / Bioorganic & Medicinal Chemistry Letters 27 (2017) 2571–2574
2573
Table 2
Inhibitory activity of apomorphine and oxoapomorphine against mutant MDM2
proteins.
Compounds
IC50 (l
M)a
MDM2
MDM2/Cys(Me)77
MDM2/Ser77
P4
0.0524
>30
1.70
0.0553
>30
>30
0.0631
>30
>30
R-(À)-apomorphineb
oxoapomorphine
a
IC50 values were determined by SPR analysis using 100 nM MDM2 protein.
Freshly prepared solutions of apomorphine were employed for the assay.
b
Table 3
Inhibitory activity of p-benzoquinone and o-catechol against the MDM2–p53
interaction.
Fig. 4. Dose-dependent inhibition of apomorphine derivatives against the MDM2–
p53 interaction. The inhibitory activity against MDM2–p53 interaction was
evaluated by SPR analysis. Closed circle, R-(À)-apomorphine (freshly prepared
solution); open circle, R-(À)-apomorphine incubated in PBS-DMSO (8:2) for 24 h;
open triangle, oxoapomorphine.
designed and synthesized a Ser mutant (MDM2/Ser77) and a Cys
(Me) mutant [MDM2/Cys(Me)77] of MDM2 protein. Ser77 is signif-
icantly less nucleophilic compared with the intact cysteine. The
methyl group of S-methylcysteine [Cys(Me)] impedes the nucle-
ophilic attack by a sulfhydryl group in cysteine. The mutant
MDM2 proteins had affinity to the p53 peptide that was very sim-
ilar to that of native MDM2 (Fig. 5). Using these proteins, we car-
ried out the competitive inhibition experiments by SPR analysis
(Table 2, Supplementary Figs. S1 and S2). While the inhibitory
activity of the P4 peptide (p53-like peptide: H-LTFEHYWAQLTS-
NH2) for native MDM2 was reproducible for both MDM2/Ser77
and MDM2/Cys(Me)77 with equipotent doses, no inhibition of this
interaction by oxoapomorphine was observed for these mutants.
This suggests that Cys77 of MDM2 is involved in the inhibitory
activity of oxoapomorphine. According to the crystal structure of
MDM2,26 Cys77 is located on the opposite side to the binding site
of p53, indicating that the covalent bond formation with oxoapo-
morphine might lead to a drastic conformational change to
MDM2 that interferes with the interaction with p53.
Compounds
IC50 (l
M)a
MDM2
MDM2/Ser77
>100
À
p-Benzoquinone (5)
o-Catechol (6)
13.9
>100
b
a
IC50 values were determined by SPR analysis using 280 nM MDM2 protein.
Not determined.
b
The studies described above suggest that the inhibitory activity
against the MDM2–p53 interaction require both a Michael accep-
tor, such as a quinone skeleton, and a cysteine residue on MDM2.
We assessed the inhibitory activities of a pair of compounds with
simple structures: p-benzoquinone (5) and o-catechol (6), which
corresponded to a pair of oxoapomorphine and apomorphine,
respectively. p-Benzoquinone showed moderate inhibitory activity
against the MDM2–p53 interaction, whereas o-catechol showed no
inhibition. Interestingly, oxoapomorphine was more potent com-
pared with p-benzoquinone, indicating that the overall structure
of oxoapomorphine may also contribute to the high binding affin-
ity to MDM2 before covalent bond formation (Table 3).27 Of note,
the bioactivity of p-benzoquinone was attenuated for the MDM2/
Ser77 mutant, suggesting that covalent bond formation at Cys77
is also relevant to the MDM2–p53 inhibitory activity.
In conclusion, we demonstrated the inhibitory mechanism of
apomorphine, which is a hit compound from mirror-image screen-
ing for MDM2–p53 inhibitors. The chiral apomorphine was con-
verted into achiral oxoapomorphine under aerobic conditions by
air oxidation, which potently inhibited the MDM2–p53 interaction
via covalent bond formation with Cys77 of MDM2 (Scheme 2).24 In
the mirror-image screening of a natural product library, chiral
To demonstrate the formation of the oxoapomorphine adduct,
we performed MALDI-TOF MS analysis of MDM2 proteins after
incubation with oxoapomorphine under assay conditions
(280 nM MDM2 and 10 lM oxoapomorphine). When native
MDM2 was employed, the mass of the oxoapomorphine adduct
was detected. On the other hand, the adduct was not detected from
a mixture of MDM2/Cys(Me)77 and oxoapomorphine (Supplemen-
tary Figs. S3 and S4). These results support the observation that a
covalent bond between oxoapomorphine and Cys77 of MDM2
was formed. We also investigated crystallization conditions for
X-ray crystal structure analysis of the MDM2-oxoapomorphine
complex; however, crystallization failed because of likely instabil-
ity and the very low solubility of the MDM2-oxoapomorphine
complex.
Fig. 5. SPR analysis of the binding affinity of mutant MDM2 proteins to the p53 peptide (55.7 RU).

2574
H. Ishiba et al. / Bioorganic & Medicinal Chemistry Letters 27 (2017) 2571–2574
HO
HO
O
O
References
[O]
1. Wiesehan K, Willbold D. ChemBioChem. 2003;4:811.
2. Funke SA, Willbold D. Mol Biosyst. 2009;5:783.
3. Zhao L, Lu W. Curr Opin Chem Biol. 2014;22:56.
4. Schumacher TNM, Mayr LM, Minor Jr DL, et al. Science. 1996;271:1854.
5. Eckert DM, Malashkevich VN, Hong LH, Carr PA, Kim PS. Cell. 1999;99:103.
6. Welch BD, Van Demark AP, Heroux A, et al. Proc Natl Acad Sci USA.
2007;104:16828.
H
N
Me
Me
N
R-(–)-apomorphine(1a)
oxoapomorphine(3)
7. Wiesehan K, Buder K, Linke RP, et al. ChemBioChem. 2003;4:748.
8. van Groen T, Wiesehan K, Funke SA, et al. ChemMedChem. 2008;3:1848.
9. Liu M, Pazgier M, Li C, et al. Angew Chem Int Ed. 2010;49:3649.
10. Liu M, Li C, Pazgier M, et al. Proc Natl Acad Sci USA. 2010;107:14321.
11. For a review, see: Vater A, Klussmann S. Drug Discov Today. 2015;20:147.
12. Klussmann S, Nolte A, Bald R, et al. Nat Biotechnol. 1996;14:1112.
13. Nolte A, Klussmann S, Bald R, et al. Nat Biotechnol. 1996;14:1116.
14. Leva S, Lichte A, Burmeister J, et al. Chem Biol. 2002;9:351.
15. Williams KP, Liu XH, Schumacher TN, et al. Proc Natl Acad Sci USA.
1997;94:11285.
16. Noguchi T, Oishi S, Honda K, et al. Chem Commun. 2016;52:7653.
17. Noguchi T, Ishiba H, Honda K, et al. Bioconjug Chem. 2017. http://dx.doi.org/
10.1021/acs.bioconjchem.6b00692.
18. MDM2 is a negative regulator of tumor-suppressor protein p53 and is involved
in tumor development and survival.28,29 Many inhibitors that block the
interaction between MDM2–p53 have been identified30,31 using various types
of high throughput screening technologies.32
SH
HO
HO
S
MDM2
Me
Cys77
MDM2
N
4
Scheme 2. Plausible mechanism of inhibitory activity of apomorphine against the
MDM2–p53 interaction.24
compounds may be identified as hit(s) for both L- and D-proteins.
19. Kanoh N, Kumashiro S, Simizu S, et al. Angew Chem Int Ed. 2003;42:5584.
20. Noguchi T, Oishi S, Honda K, et al. Bioorg Med Chem Lett. 2013;23:3802.
21. Cheng HY, Strope E, Adams RN. Anal Chem. 1979;51:2243.
The majority of stable compounds have nonselective binding affin-
ity for both protein enantiomers, whereas a fraction of hits, like
apomorphine, may have unstable structures that convert into achi-
ral compounds during the screening process.
22. Garrido JMPJ, Delerue-Matos C, Borges F, et al.
2002;2:1713.
J Chem Soc Perkin Trans.
23. It was reported that o-quinone intermediate 2a would be involved in the
irreversible reaction with glutathoine, see ref 21. Although intermediate 2a is a
chiral molecule, the possible contribution of 2a to the inhibition against
MDM2–p53 interaction cannot be ruled out. Indeed, the incubated
apomorphine sample appparently exhibited more potent inhibitory activity,
indicating that alternative inhibitory compound(s) with higher potency might
be included in the sample.
24. Smith RV, Velagapudi RB, McLean AM, Wilcox RE. J Med Chem. 1985;28:613.
25. Bista M, Smithson D, Pecak A, et al. PLoS One. 2012;7:e37518.
26. Kussie PH, Gorina S, Marechal NP. Science. 1996;274:948 (PDB entry 1YCR).
27. The possible contribution of the hydrophobic property of oxoapomorphine to
the initial association process onto the hydrophobic surface and/or pocket of
MDM2 cannot be ruled out.
Acknowledgements
This work was supported by Grants-in-Aid for Scientific
Research from JSPS, Japan (24659046, 15H04653, 26-7738); Tar-
geted Protein Research Program; and Platform for Drug Discovery,
Informatics, and Structural Life Science from MEXT, Japan; the
Tokyo Biochemical Research Foundation; and Mochida Memorial
Foundation for Medical and Pharmaceutical Research. T.N. is grate-
ful for Research Fellowships for Young Scientists from JSPS, Japan.
We are grateful to Dr. Toru Nakatsu (Kyoto University) for his gen-
erous support of this research.
28. Toledo F, Wahl GM. Nat Rev Cancer. 2006;6:909.
29. Wade M, Wang YV, Wahl GM. Trends Cell Biol. 2010;11:694.
30. Wade M, Li YC, Wahl GM. Nat Rev Cancer. 2013;13:83.
31. Vassilev LT, Vu BT, Graves B, et al. Science. 2004;303:844.
32. Tsuganezawa K, Nakagawa Y, Kato M, et al. J Biomol Screen. 2013;18:191.
A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2017.03.082.