Identification of a new p53/MDM2 inhibitor motif inspired by studies of chlorofusin

Article abstract of DOI:10.1016/j.bmcl.2015.06.014

Previous studies on the natural product chlorofusin have shown that the full peptide and azaphilone structure are required for inhibition of the interaction between MDM2 and p53. In the current work, we utilized the cyclic peptide as a template and introd

Full text of DOI:10.1016/j.bmcl.2015.06.014

Bioorganic & Medicinal Chemistry Letters 25 (2015) 4878–4880
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Identification of a new p53/MDM2 inhibitor motif inspired by studies
of chlorofusin
Marco M. D. Cominetti a, Sarah A. Goffin a, Ewan Raffel a, Kerrie D. Turner a, Jordann C. Ramoutar a,
Maria A. O’Connell a, Lesley A. Howell a, , Mark Searcey a,b,
⇑
⇑
a School of Pharmacy, University of East Anglia, Norwich Research Park, Norwich NR4 7TJ, UK
b School of Chemistry, University of East Anglia, Norwich Research Park, Norwich NR4 7TJ, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Previous studies on the natural product chlorofusin have shown that the full peptide and azaphilone
structure are required for inhibition of the interaction between MDM2 and p53. In the current work,
we utilized the cyclic peptide as a template and introduced an azidonorvaline amino acid in place of
the ornithine/azaphilone of the natural product and carried out click chemistry with the resulting pep-
tide. From this small library the first ever non-azaphilone containing chlorofusin analog with
MDM2/p53 activity was identified. Further studies then suggested that the simple structure of the
Fmoc-norvaline amino acid that had undergone a click reaction was also able to inhibit MDM2/p53 inter-
action. This is an example where studies of a natural product have led to the serendipitous identification
of a new small molecule inhibitor of a protein–protein interaction.
Received 29 April 2015
Revised 29 May 2015
Accepted 2 June 2015
Available online 14 June 2015
Keywords:
p53
MDM2
Chlorofusin
Click chemistry
Protein–protein interactions
Ó 2015 Elsevier Ltd. All rights reserved.
Chlorofusin (Fig. 1) is a natural product that was first identified
in a screen for inhibitors of the MDM2/p53 interaction.1,2 It is a
cyclic peptide in which one of the amino acids, an ornithine, is con-
densed with a densely functionalized azaphilone. Interest in both
its biological activity, as an inhibitor of a protein–protein interac-
tion, and its structure initiated a series of studies of the natural
compound. The cyclic peptide was synthesized by both solid3
and solution phase4–6 methods, its stereochemistry and its lack
of inherent biological activity were established. The total synthesis
initially proved elusive as the first approach gave an erroneous
structure to the azaphilone portion,7 although this was closely fol-
lowed by the first total synthesis from the Boger laboratory.8,9
Interestingly, this approach allowed access to all the diastereomers
of the azaphilone structure and showed that the presence of the
azaphilone was required for MDM2/p53 inhibition but the overall
stereochemistry of the molecule had little effect.10 This has
recently been confirmed through a further total synthesis.11 We
have also shown that the introduction of substituents on the
ornithine in order to generate simpler analogs does not lead to
active MDM2/p53 inhibitors.12 Yao and co-workers have recently
described a synthesis of chlorofusin analogs that incorporate a
triazole moiety in the linker between the peptide and the aza-
philone and these are the first chlorofusin analogs to show inhibi-
tory activity in the MDM2/p53 assay.13 This has prompted us to
disclose our own studies of a click chemistry-based approach to
chlorofusin analogs that gave us the first simple non-azaphilone
peptide analogs of chlorofusin with anti-MDM2 activity.
However, on further probing, we found that the simple ‘clicked’
amino acid that we incorporated into the chlorofusin peptide also
had activity and was a relatively potent inhibitor of MDM2/p53
interactions. In order to facilitate these studies, we have also devel-
oped a fluorescence polarization (FP) based assay for screening
purposes that is simpler than the previously described12 ELISA
based assay (Fig. 1).
Incorporation of an azide moiety into our solid phase protocol
would give us a compound that could undergo click chemistry14
either while still on the resin or following cleavage. For this, we
required an Fmoc protected azide-containing amino acid. We uti-
lized a route from Boc-glutamic acid
an azide analog of ornithine in a modest yield, but through a scal-
able process (Scheme 1). Boc-glutamic acid -t-butyl ester 1 was
a-t-butyl ester 1 to generate
a
treated with carbonyldiimidazole and then reduced to alcohol 2
with sodium borohydride in a one pot, two step procedure, produc-
ing the expected compound in excellent yield. Mesylation and sub-
sequent treatment of the crude with sodium azide15 generated the
required side chain on the amino acid (4). The product does not
⇑
Corresponding authors.
E-mail addresses: l.howell@uea.ac.uk (L.A. Howell), m.searcey@uea.ac.uk (M.
Searcey).
http://dx.doi.org/10.1016/j.bmcl.2015.06.014
0960-894X/Ó 2015 Elsevier Ltd. All rights reserved.

M. M. D. Cominetti et al. / Bioorg. Med. Chem. Lett. 25 (2015) 4878–4880
4879
(CH2)6CH3
O
It seemed reasonable to explore the binding of the ornithine
amino acid analog to confirm that the peptide structure of chloro-
fusin was contributing to the binding of the click chemistry prod-
uct. The 4-(4-methylphenyl)triazole structure is significantly
different from the azaphilone of the natural product and it was a
possibility that the side chain itself was responsible for the biolog-
ical activity. With this in mind, the ornithine azide with Fmoc-pro-
tection 5 was treated with 4-ethynyltoluene and, as a further novel
analog, 1-ethynyl-4-trifluoromethylbenzene under the conditions
of the click chemistry reaction (Scheme 3). The yields were excel-
lent and the products could be isolated without chromatographic
purification. Nonetheless, to ensure ideal purity, they were addi-
tionally subjected to preparative HPLC. When these compounds
were tested in the FP assay, we were surprised to discover that
they also inhibited the MDM2/p53 interaction, with the 4-methyl-
phenyl compound having improved activity in comparison with
the full peptide structure (Table 1).
O
Cl
O
OH
O
O
O
O
HN
N
H
HN
N
O
O
OH
HN
HO
NH
Yao13 generated
analogs with a
trizole linker
through click
chemistry
O
NH
O
O
O
O
NH
O
HN
O
N
H
NH2
NH2
Chlorofusin
Figure 1. Chlorofusin.
The presence of halides is often seen as preferential for MDM2
binding ligands and so we also generated a number of analogs con-
taining the Cl, Br and I phenyl systems. In order to probe the poten-
tial size of the binding site, we made a methoxy analog as the
methylphenyl-compound was more active than the trifluo-
romethylphenyl. When screened in the FP assay, all of these com-
pounds except for the chlorophenyl-compound were inhibitors of
the p53 peptide/MDM2 interaction. The most potent compound
with the highest binding affinity was the iodo-derivative although
the binding affinity of the bromo compound was very similar.
Binding of the 4-methoxy compound was also enhanced compared
with the methyl, suggesting that the larger group is filling a bind-
ing pocket on the MDM2 protein more efficiently than the smaller
CH3. All of these compounds compare well with the p53 peptide,
with higher affinity. When the well-known anti-MDM2 compound
nutlin 3a was compared in the same assay, its IC50 value was about
5 fold lower than that of the iodo compound, suggesting that it is
maintaining a slightly higher potency than these new compounds
in this assay.
require purification and was treated with a saturated solution of
hydrochloric acid in ethyl acetate at room temperature to remove
both the Boc and tBu protecting groups. Evaporation of the solvent
and reaction of the residue with Fmoc-Cl in the presence of sodium
carbonate gave the target amino acid 5.
Incorporation of 5 in to the solid phase synthesis3 proceeded
smoothly (Scheme 2) and the azide was formed on resin, as
demonstrated by cleavage of a small sample and analysis by mass
spectrometry. A convenient route to derivatizing this compound
would be to carry it out on resin, provided the triazole-forming
reaction proceeded under these conditions. We had previously
found that acylation of the ornithine on resin was a convenient
approach to generate analogs.12 To our satisfaction, treatment of
the resin bound peptide (6) with a selection of alkynes with differ-
ent functional groups led to the successful generation of click ana-
logs. These could be removed from the resin using a standard
cleavage cocktail (95% TFA, 2.5% H2O, 2.5% TIPS) and purified by
preparative HPLC.
Removal of the Fmoc group from the compound 18 to generate
24 leads to a complete loss of activity, suggesting that a hydropho-
bic group is required in this position of the molecule. Screening of
the t-Bu ester version of the iodo compound, which we suspected
would lead to cellular uptake of these compounds, also led to loss
of activity for binding.
A preliminary screen in a proliferation assay was also under-
taken. Compounds 19–23 showed no activity against either SJSA-
1 (osteosarcoma, amplification of MDM2) or A375 cell lines (mel-
anoma, wild type p53, normal MDM2). However, both compounds
17 and 18 showed some activity, with 17 being active in the SJSA-1
The compounds were screened by a convenient FP assay,
adapted from several that have been previously published.16 The
high affinity peptide LTFEHYWAQLTS was synthesised on Rink
amide resin to generate the amide at the C-terminus and was
tagged with FAM at the N-terminus. Human MDM2 protein (17-
125) was used in the polarization assay and the wildtype p53 pep-
tide (residues 15–27) was used as a positive control and had an
IC50 of 14.45
lM and Ki of 1.82
lM. Screening the click chlorofusin
analogs 7–16 at 100
lM concentration revealed that only the 4-
methylphenyl (8) inhibited the MDM2/peptide interaction. A full
inhibition curve for compound 8 revealed the methylphenyl com-
cell line (33.1
both cell lines (31.2
l
M) and inactive in A375 whereas 18 was active in
M, 49.3 M). Cell-based assays such as these
pound had an IC50 of 66.01 lM and Ki of 8.309 lM. This is the first
l
l
time that simple analogs of chlorofusin have been shown to have
activity against MDM2 and compound 8 has a similar affinity to
the wildtype p53 peptide.
are not direct indicators of whether the compounds are acting
through the inhibition of MDM2/p53 and the fact that 18 is active
in both cell lines is indicative of a different mechanism or mecha-
nisms of action. Cell based assays that offer a more definitive
demonstration of intracellular activity have recently been
described17 and we will explore the utility of these approaches.
In conclusion, we have developed new analogs of chlorofusin
via click chemistry that have the ability to prevent the binding of
the p53 peptide to the MDM2 protein. Chlorofusin could be consid-
ered to be the product of a natural ‘fragment-based’ approach to
small molecule development—with the condensation of an aza-
philone and a cyclic peptide leading to an active molecule. In a sim-
ilar vein, we have used the cyclic peptide as a template to generate
new molecules to bind to MDM2. Further investigation of those
molecules has led to the identification of a new small molecule
motif, based around the click product of an Fmoc-protected
ornithine, which can also inhibit the p53/MDM2 interaction. It
O
O
2.4 eq NaN3,
dry DMF,
OtBu
1) 1.24 eq CDI,
THF, RT, 10 min
HOOC
OtBu
RO
67oC, 5h
2) 1.52 eq NaBH4,
H2O, 0oC-RT, 1h
HN
HN
Boc
Boc
1
1.5 eq Mesyl chloride,
1.5 eq Et3N, dry DMF
0oC, 45 min
R = H 2 91%
R = Ms 3 85%
1) HCl(sat), EtOAc,
O
O
0oC, 40 min - RT, 7 h
N3
OtBu
N3
OH
2) 1 eq Fmoc-Cl,
3.5 eq Na2CO3,
H2O-Dioxane, 0oC-RT 4 h
HN
4 100%
HN
Boc
Fmoc
5 71%
Scheme 1. Synthesis of the azide analog of ornithine.

4880
M. M. D. Cominetti et al. / Bioorg. Med. Chem. Lett. 25 (2015) 4878–4880
1. Fmoc-Asp-ODmab
HOBt, HBTU, DIPEA
2. 40% Piperidine, DMF
3. 20% Piperidine, DMF
4. 20% Piperidine, DMF
1. Fmoc-A.A.-OH
HOBt, HBTU, DIPEA
2. 40% Piperidine, DMF
3. 20% Piperidine, DMF
4. 20% Piperidine, DMF
O
O
ODmab
O
H
N
1. 2% N2H2 in DMF,
2. 5% DIPEA in DMF
3. DIC, HOBt,
1. 40% Piperidine, DMF
2. 20% Piperidine, DMF
3. 20% Piperidine, DMF
N
H
NHFmoc
NH2
HN
O
HN
O
OtBu
O
NH2
N3
DIPEA, DMF
NH
HN
MBHA Rink amide
tBuO
O
O
NH
O
O
HN
O
FmocNH
HN
ODm ab
O
TrtNH
O
NH
Cl
O
N
O
O
O
10 eq Alkyne
10 eq CuSO4.5H2O
10 eq sodium
R
N
N
HN
O
HN
O
H
H
HN
HN
O
O
OtBu
OtBu
N3
N
ascorbate, DMF
F
N
NH
NH
HN
HN
S
tBuO
O
tBuO
O
R =
O
O
O
O
NH
NH
O
NH
O
NH
Br
O
O
HN
HN
N
H
TrtNH
N
7
8
9
10
11
12
H
O
O
TrtNH
O
O
O
O
O
NH
NH
O
O
O
O
6
N
N
O
13
14
15
16
Scheme 2. Solid phase synthesis of analogs of chlorofusin peptide.
Studies are continuing to identify the binding site and the struc-
ture–activity relationship for this new class of compound.
0.25 eq CuSO4.5H2O
2.5 eq sodium
ascorbate, DMF,
CO2H
N
N
N
CO2H
NHFmoc
N3
Acknowledgments
NHFmoc
RT, 4 h
We thank the UEA for funding for MMDC and SAG and the
EPSRC Mass Spectrometry Service, Swansea for HRMS.
R
R
17 R = CH3 80%
18 R = CF3 95%
19 R = Cl 95%
Supplementary data
20 R = Br 90%
21 R = I 81%
22 R = OCH3 77%
23 R = NHBoc 53%
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2015.06.
014. These data include MOL files and InChiKeys of the most
important compounds described in this article.
CO2H
CO2H
NH2
N
N
N
N
N
N
NHFmoc
References and notes
1. Duncan, S. J.; Grüschow, S.; Williams, D. H.; McNicholas, C.; Purewal, R.; Hajek,
M.; Gerlitz, M.; Martin, S.; Wrigley, S. K.; Moore, M. J. Am. Chem. Soc. 2001, 123,
554.
2. Clark, R. C.; Lee, S. Y.; Searcey, M.; Boger, D. L. Nat. Prod. Rep. 2009, 26, 465.
3. Malkinson, J. P.; Zloh, M.; Kadom, M.; Errington, R.; Smith, P. J.; Searcey, M. Org.
Lett. 2003, 5, 5051.
1% piperidine,
1% DBU, CH2Cl2,
RT, 10 min
24 55%
18
F3C
F3C
Scheme 3. Synthesis of compounds 19–26.
4. Desai, P.; Pfeiffer, S. S.; Boger, D. L. Org. Lett. 2003, 5, 5047.
5. Mori, T.; Miyagi, M.; Suzuki, K.; Shibasaki, M.; Saikawa, Y.; Nakata, M.
Heterocycles 2007, 72, 275.
Table 1
Binding of compounds 17, 18 and 20–23 to MDM2 using the fluorescence polarization
assay
6. Lee, S. Y.; Boger, D. L. Tetrahedron 2009, 65, 3281.
7. Qian, W. J.; Wei, W. G.; Zhang, Y.-X.; Yao, Z.-J. J. Am. Chem. Soc. 2007, 129, 6400.
8. Sang, Y. L.; Clark, R. C.; Boger, D. L. J. Am. Chem. Soc. 2007, 129, 9860.
9. Clark, R. C.; Sang, Y. L.; Boger, D. L. J. Am. Chem. Soc. 2008, 130, 12355.
10. Clark, R. C.; Lee, S. Y.; Hwang, I.; Searcey, M.; Boger, D. L. Tetrahedron Lett. 2009,
50, 3151.
11. Qiu, H. B.; Qian, W. J.; Yu, S. M.; Yao, Z. J. Tetrahedron 2015, 71, 370.
12. Woon, E. C. Y.; Arcieri, M.; Wilderspin, A. F.; Malkinson, J. P.; Searcey, M. J. Org.
Chem. 2007, 72, 5146.
13. Qiu, H. B.; Chen, X. Y.; Li, Q.; Qian, W. J.; Yu, S. M.; Tang, G. L.; Yao, Z. J.
Tetrahedron Lett. 2014, 55, 6055.
14. Kolb, H. C.; Finn, M. G.; Sharpless, K. B. Angew. Chem., Int. Ed. 2004, 2001, 40.
15. Howell, L. A.; Bowater, R. A.; O’Connell, M. A.; Reszka, A. P.; Neidle, S.; Searcey,
M. ChemMedChem 2012, 7, 792.
a
Compounds
IC50
(
lM)
Kib
(lM)
17
18
20
21
22
23
Nutlin 3A
24.26
49.80
5.51
2.90
5.66
3.05
6.96
0.69
0.36
0.71
7.22
57.36
0.61
a
Concentration of substrate required to decrease polarization fluorescence by
50%.
16. Knight, S. M. G.; Umezawa, N.; Lee, H. S.; Gellman, S. H.; Kay, B. K. Anal.
Biochem. 2002, 300, 230.
b
Apparent inhibition constant.
17. Li, Y.-C.; Rodewald, Luo W.; Hoppmann, C.; Wong, Ee T.; Lebreton, S.; Safar, P.;
Patek, M.; Wang, L.; Wertman, Kenneth F.; Wahl, Geoffrey M. Cell Rep. 2014, 9,
946.
would be interesting to see if the azaphilone click analogs recently
described13 also have activity as the Fmoc-amino acid structure.