New anticancer agents mimicking protein recognition motifs

Article abstract of DOI:10.1021/jm400947b

The novel tetrasubstituted pyrrole derivatives 8g, 8h, and 8i showed selective cytotoxicity against M14 melanoma cells at low micromolar concentration. Structure-activity relationships (SARs) indicated the presence of three aromatic substituents on the pyrrole core as necessary for biological activity. Computational studies strongly suggest that the peculiar 3D orientation of these substituents is able to reproduce the hydrophobic side chains in LxxLL-like protein recognition motifs. Biological results showed altered p53 expression and nuclear translocation in cells sensitive to the compounds, suggesting p53 involvement in their anticancer mechanism of action. Unfortunately, because of poor solubility of the active analogues, it was not possible to perform further investigation by NMR techniques. Pharmacophore models were generated and used to perform 3D searches in molecular databases. Results indicated that two compounds share the same pharmacological profile and the same pharmacophoric features with our new derivatives, and one of them inhibited MDM2-MDM4 heterodimer formation.

Full text of DOI:10.1021/jm400947b

Subscriber access provided by the Library Service | University of Stellenbosch
Article
New Anticancer Agents Mimicking Protein Recognition Motifs
Marco Persico, Anna Ramunno, Vita Maglio, Silvia Franceschelli, Chiara Esposito,
Alfonso Carotenuto, Diego Brancaccio, Valeria De Pasquale, Luigi Michele Pavone,
MICHELA VARRA, Nausicaa Orteca, Ettore Novellino, and Caterina Fattorusso
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/jm400947b • Publication Date (Web): 23 Jul 2013
Downloaded from http://pubs.acs.org on August 1, 2013
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
New Anticancer Agents Mimicking Protein
Recognition Motifs
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
#
,§
±,
±
±,
Marco Persico, Anna Ramunno, * Vita Maglio, Silvia Franceschelli, * Chiara Esposito,
±
#,§
#
€
€
Alfonso Carotenuto, Diego Brancaccio, Valeria De Pasquale, Luigi Michele Pavone,
#
#
#
#,
Michela Varra, Nausicaa Orteca, Ettore Novellino, Caterina Fattorusso *
#
Dipartimento di Farmacia, Università di Napoli “Federico II”, Via D. Montesano 49, 80131
Napoli, Italy
±
Dipartimento di Farmacia, Università di Salerno, Via Giovanni Paolo II 132, 84084 Fisciano
(SA), Italy
€
Dipartimento di Biochimica e Biotecnologie Mediche, Università di Napoli “Federico II”,
Via S. Pansini 5, 80131 Napoli, Italy
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Abstract
The novel tetrasubstituted pyrrole derivatives 8g, 8h, and 8i showed selective cytotoxicity
against M14 melanoma cells at low micromolar concentration. Structureꢀactivity
relationships (SARs) indicated the presence of three aromatic substituents on the pyrrole core
as necessary for biological activity. Computational studies strongly suggest that the peculiar
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3ꢀD orientation of these substituents is able to reproduce the hydrophobic side chains in
LxxLLꢀlike protein recognition motifs. Biological results showed altered p53 expression and
nuclear translocation in cells sensitive to the compounds, suggesting p53 involvement in their
anticancer mechanism of action. Unfortunately, due to poor solubility of the active analogues,
it was not possible to perform further investigation by NMR techniques. Pharmacophore
models were generated and used to perform 3ꢀD searches in molecular databases, it resulted
that two compounds share the same pharmacological profile and the same pharmacophoric
features with our new derivatives, and one of them inhibited MDM2ꢀMDM4 heterodimer
formation.
2
ACS Paragon Plus Environment

Page 3 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Introduction
The identification of small molecules able to modulate specific intraꢀ and intermolecular
proteinꢀprotein interactions is necessary to clarify the dynamics of many cellular processes,
and promote the development of novel therapeutic approaches especially in the case of multiꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
factorial diseases such as cancer. Cancer cells are indeed characterized by the modification
of one or more molecular pathways involved in the regulation of cell life cycle. Thus, in this
field many proteinꢀprotein interactions are promising druggable targets, such as, those
involving transcription factors and their regulators (e.g., p53, MDM2), microtubules, proꢀ
2,3
apoptotic and apoptotic proteins (e.g., Bclꢀxl, Bcl2, Bax, Bid, XIAP).
The binding domains implicated in proteinꢀprotein interactions can be small binding clefts or
4
large surfaces, and generally include smaller regions, called hot spots, which represent
energetic focal points on protein surfaces that are major contributors to the binding energy of
the complex. These highly conserved binding subꢀregions have a functional and structural
adaptivity, and typically are able to accommodate small organic compounds based on
5
different scaffolds. It is noteworthy that the same (or homologous) hot spot can adapt the
same (or homologous) residues in different structural contexts. For example, the same
MDM2ꢀbinding domain can be recognized by p53, as well as, by its homologous proteins,
p63 and p73, and potentially by many other proteins containing similar binding motifs, whose
6
biological relevance is, in some cases, still not clarified. On the other hand, recognition
7
motifs, such as, polyproline and LxxLLꢀlike motifs, can bind to different protein domains.
Consequently, a small molecular scaffold able to mimic protein recognition motifs is
potentially able to alter different proteinꢀprotein interactions of a given molecular pathway.
Understanding the structural basis for peptidomimetic inhibition holds the key to develop
novel, highly effective, and selective drug candidates. As an example, we previously reported
that the particular spatial orientation of the three aromatic rings of paclitaxel is able to mimic
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the hydrophobic side chains of the interacting residues of SH3 domains, thus accounting for
8
its activity towards multiple protein targets involved in cellular apoptosis.
On these bases, in the search for new anticancer leads, we synthesized a small set of
tetrasubstituted pyrrole derivatives with the aim to develop stable mimics of LxxLLꢀlike
motifs, which have been reported as one of the most representative proteinꢀrecognition motifs
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
in cell cycle regulation. Two series of compounds were synthesized, namely a) 4ꢀacetylꢀ (8aꢀ
d, Table 1) and b) 4ꢀbenzoyl derivatives (8eꢀn, Table1), and their antiproliferative activity
was evaluated against different cancer cell lines (MCF7, Huh7, M14, Jurkat), as well as, on
mouse monocyte macrophages (Raw 264ꢀ7) cell line, by means of the MTT assay. Three of
them, 8gꢀi, showed a selective cytotoxicity against M14 melanoma cells at low micromolar
concentration. Then, active derivatives 8gꢀi were further investigated for their mechanisms of
action by means of biological, NMR, and molecular modeling studies.
Results and Discussion
Chemistry. We used a synthetic approach that enabled us to easily introduce: 1) three
lipophilic groups with different size and substituents in 1,2,4ꢀpositions of the pyrrole core; 2)
one hydrogen bond donor (NH) and two hydrogen bonds acceptors (C=O) in 2 and 4
positions. Synthetic route of the title compounds 8aꢀn is outlined in Scheme 1.
4
ACS Paragon Plus Environment

Page 5 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
Scheme 1
Boc
N
H
H N
2
COOCH3
OH
N
COOCH3
OH
COOCH3
a
Ts
b
HCl
Ts
CH2
1
2
3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
c,d
O
O
O
R1
H C
R1
H C
R1
COOH
N
COOCH
3
X
COOCH
3
3
N
3
N
H
f
e
4
4
4
a, R = CH , X = CH
1 3 3
b, R = C H , X = CH
3
R2
R2
CH Br
1
6
5
2
c, R = CH , X = C H
7
a-e
6a-e
1
3
6
5
R2
6
6
6
6
6
a, 7a R = CH , R = H
5a, R = H
1
3
2
2
b, 7b R = Phe, R = H
g, R NH2
1
2
5b, R = CH
3
2
3
c, 7c R = Phe, R = CH
1
2
3
5c, R = NO
2
2
d, 7d R = Phe, R = NO
1
2
2
5d, R = CF
2
3
e, 7e R = Phe, R = CF
1
2
3
O
R1
H C
CONHR
3
3
N
R2
8
a-n
a
Reagents and conditions: a) tosyl chloride, TEA, DCM, 0 °C to rt, overnight; b) (Boc)
2
O,
DMAP, AcCN, rt; c) Cs CO , dry AcCN, pentaneꢀ2,4ꢀdione for 4a, and 1ꢀphenylꢀ1,3ꢀ
2
3
butanedione for 4b and 4c; d) THF, DCM, rt; e) 50% w/v NaOH, TBAH, DCM, 0 °C to rt,
overnight; f) 2M NaOH, MeOH, reflux, 2 h; g) HOBT, HBTU, NMM, DMF, rt.
The 4ꢀacetylꢀ2ꢀmethoxycarbonylꢀ5ꢀmethylpyrrole 4a and its 4ꢀbenzoyl analogue 4b were
synthesized according to a slightly modified procedure previously reported by Ferreira et
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
9
,10
al.
In detail, reaction of (±)ꢀserine methyl ester hydrochloride 1 with pꢀtoluensulfonyl
11
chloride afforded the derivative 2, which in turn was reacted with tertꢀbutyl pyrocarbonate,
(
Boc) O, and a catalytic amount of 4ꢀdimethylaminopyridine (DMAP) to give the
2
dehydroalanine derivative 3.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Then, the intermediate 3 was reacted with pentaneꢀ2,4ꢀdione, using one equivalent of Cs CO3
2
as base, and the crude mixture was treated with 10% THF in dichloromethane at room
temperature to yield the pyrrole derivative 4a. A similar procedure was performed with 1ꢀ
phenylꢀ1,3ꢀbutanedione instead of pentaneꢀ2,4ꢀdione to give a mixture of 4ꢀbenzoylꢀ5ꢀ
methylꢀ1Hꢀpyrroleꢀ2ꢀcarboxylic acid methyl ester (4b) and its regioisomer 4ꢀacetylꢀ5ꢀ
phenylꢀ1Hꢀpyrroleꢀ2ꢀcarboxylic acid methyl ester (4c), not previously described, in an
approximately 4:1 ratio. The structure of 4b and 4c was confirmed by the analysis of the
ROESY spectra. In particular, a ROE signal between pyrrole NH and C5ꢀmethyl was
observed only in the spectrum of compound 4b (Figure 1S1). For compound 4c, a diagnostic
ROE signal was observed between pyrrole NH and H1’/H5’ of phenyl ring (Figure 2S1). The
intraꢀmolecular distances associated to these two diagnostic ROE signals were evaluated by
integration of the ROESY spectra to be 3.20 Å and 3.30 Å, respectively. These values are in
good accordance with distances present in experimentally determined structures containing
the 2ꢀmethylꢀ1Hꢀpyrrole and the 2ꢀphenylꢀ1Hꢀpyrrole fragments, i.e., 3.10 and 3.45 Å,
respectively (Cambridge Structural Database (CSD) codes: POPPAM and IDOCUZ). To
1
13
notice that our H and C NMR data (see Experimental Section) of the regioisomer 4b do not
9
1
13
match those reported by Ferreira et al. for the same compound. Actually, H and C NMR
spectra of the isomer 4c are comparable to those reported by Ferreira et al. Hence, we argue
1
that there was a misassignment in that case. Indeed, our H NMR spectrum of 4b matches
1
2
that described by Toplak et al. who employed a different synthetic approach.
6
ACS Paragon Plus Environment

Page 7 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Nꢀbenzylation of pyrrole derivatives 4a,b with commercially available benzyl bromides 5aꢀd
afforded the esters 6aꢀe, which in turn were hydrolyzed with 2 M NaOH at room temperature.
The corresponding acids 7aꢀe were precipitated by adding 2 M HCl, and were subjected to a
coupling reaction with an appropriate amine using 1ꢀhydroxybenzotriazole (HOBT) and Oꢀ
benzotriazolꢀ1ꢀylꢀN,N,N′,N′ꢀtetramethyluronium hexafluorophosphate (HBTU) as coupling
reagents in the presence of diisopropylethylamine (DIPEA) in DMF to provide the final
compounds 8aꢀn (Table 1) in good yields.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1 Physical and chemical data for compounds 8aꢀn.
O
R1
H C
N
CONHR
3
3
R2
8a-n
a
b
c
Compd
R1
R2
H
R3
Yield mp (°C)
Formula
Anal.
8
a
CH
3
3
3
85
94
92
91
82
89
133ꢀ135
C
C
C
22
21
24
H
H
H
22
26
26
N
N
N
2
2
2
O
O
O
2
2
2
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
8b
CH
CH
CH
H
H
H
H
H
184ꢀ186
94ꢀ96
8
c
8
d
3
130ꢀ132
116ꢀ118
156ꢀ158
22 2 2
C H21IN O
I
8
e
C
C
27
26
H
H
24
28
N
N
O
2 2
8
f
O
2 2
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
8g
8h
8i
H
H
87
90
84
72
87
85
127ꢀ129
118ꢀ120
141ꢀ143
129ꢀ130
123ꢀ124
126ꢀ127
C
27
H
23IN
2
O
2
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
I
C
27
H
H
H
23ClN
25ClN
22ClN
2
2
3
O
O
O
2
2
Cl
Cl
Cl
Cl
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
CH
C
C
3
28
27
8j
NO
2
4
8k
8l
CF
3
C
28
H
22ClF
3
N
2
O
2
H
C
27
H
H
H
23
28
N
3
O
O
O
4
NO
2
MeO
8
m
H
H
91
142ꢀ144
125ꢀ126
C
C
29
32
N
N
2
4
2
C,H,N
C,H,N
OMe
8
n
77
26
2
a
c
b
Yields refer to isolated and purified materials. Recrystallization solvent: EtOAc/nꢀhexane.
All the compounds were analyzed within ± 0.4% of the theoretical values.
Cytostatic effect of 8gꢀi derivatives. To evaluate the inhibitory effect of 8aꢀn derivatives we
used concentrations ranging from 100 nM to 100 ꢁM in the culture media of different human
tumor cell lines (M14, Huh7, MCF7, Jurkat), and on a mouse monocyte macrophages (Raw
2
64ꢀ7) cell line. Cells were treated for 24 and 48 h, and the cell viability was measured by the
MTT assay. Compounds 8gꢀi showed antiꢀproliferative activity on a single cell line, M14,
and results are summarized in Table 2. The IC50 concentrations of compounds were used in
all subsequent experiments.
8
ACS Paragon Plus Environment

Page 9 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 2. Antiꢀproliferative activity of compounds 8gꢀi. IC50 measured against cancer (M14,
MCF7, Huh7, and Jurkat) and normal (Raw 264ꢀ7) cell lines after 48 h treatment.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
IC50 (ꢀM)
a
b
c
d
e
Compd
M14
MCF7 Huh7 Jurkat Raw 264ꢀ7
8
8
g
5.0 ± 0.1 >100
7.5 ± 0.1 >100
>100
>100
>100
>100
>100
>100
>100
>100
>100
h
8i
2.5 ± 0.1 >100
a
b
c
Human melanoma cell line. Human breast cancer cell line. Human hepatocarcinoma cell
d
e
line. Immortalized line of T lymphocyte cells. Cell line murine macrophage.
The selective cytotoxicity at tested concentrations of 8gꢀi toward M14 cell line, suggests a
mechanism of action based on the interaction with a molecular pathway which is altered in
this cancer cells.
To characterize the inhibitory effect, we determined growth rates of M14 cells in the absence
1
3
or presence of active compounds 8gꢀi by the trypan blue dye exclusion method. Since we
did not observe a significant increase in cell death (trypan blue positive) in treated cells
compared with control cells we can conclude that proliferation of exponentially growing cells
was inhibited by treatment with an inhibitory effect seen after 24 h suggesting that these
compounds have a cytostatic effect.
Two sets of M14 cells were treated with 8gꢀi for 24 h and the medium was then removed and
cells were reꢀsuspended in a compoundꢀfree medium. The cells were incubated for further 24
and 48 h and analyzed by MTT assay to assess cell viability.
Obtained data suggest that incubation of the cells in compoundꢀfree medium for 48 h was
sufficient to restore normal cells proliferation (data not shown). The reversible effect on cell
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
proliferation indicates a specificity of action of the compounds and lack of toxicity once
removed from the cells after 24 h of incubation.
Effect on cell cycle progression and apoptosis analyses. Based on the results obtained, we
characterized the cytostatic effect by performing distribution analysis by flow cytometry
during the different cell cycle phases.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
As shown in Figure 1 (panel A), all three tested compounds 8gꢀi induced accumulation of
cells in S phase, and a corresponding decrease in G0/G1 and G2/Mꢀphase already after a 24 h
treatment. Moreover, cell cycle deꢀregulation leading to apoptosis occurred after 48 h of
treatment (Figure 1, panel B).
Figure 1. Cell cycle (panel A) and apoptosis profile (panel B) in M14 cell line in presence of
8gꢀi derivatives after 24 and 48 h treatment.
10
ACS Paragon Plus Environment

Page 11 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The Sꢀphase arrest was confirmed by nucleusꢀDAPI staining where we observed differences
in the size of the nucleus of treated cells probably due to an increase in DNA content and
inability of the cells to divide (Figure 2).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. M14 cells were treated for 24 h with 8gꢀi derivatives and nuclear areas were
measured using ImageJ software after DAPI staining. Results are expressed as mean ± SEM
of at least twenty independent measurements; *** denotes P<0.001 vs control. Statistical
significance was measured using one way ANOVA analysis, the Bonferroni post test was
used to compare differences between the groups. All measured values were significant
according to statistical test (P<0.05).
Cyclin A expression and involvement of p53. To understand the role of pyrrole derivatives
8gꢀi on cell cycle arrest, we assessed their effect on cyclin A, a molecular marker associated
14
with G1/S phase by Western blotting (Figure 3).
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 3. Western blotting (left) and histogram profile (right) of cyclin A expression after 24
and 48 h of 8gꢀi derivatives treatment at IC50 concentration. Cyclin A upꢀregulation is
required for induction and maintenance of cell cycle arrest.
As reported in Figure 3, treatment of M14 cells with the active compounds 8gꢀi led to an
increase in the expression of cyclin A known to promote Sꢀphase entry in mammals. Our data
suggest that the tested compounds induce a block in the S phase of the cell cycle and are
consistent with our results obtained by flow cytometry analysis. Interestingly, the increase is
present at 24 hours for all compounds and at 48 hours for compounds 8g and 8i, but not for
8h, suggesting the presence of a different mechanism that needs to be further elucidated.
Since 8gꢀi compounds reduce cell viability, induce cell cycle arrest and affect cyclin A
expression, we investigated the expression of p53 by Western blotting in M14 and Huh7 cells
15
after 24 and 48 hrs of treatment (Figure 4). Huh7 cells were used as a control since they
also contain a missense mutation in p53 gene but they are not sensitive to 8gꢀi.
12
ACS Paragon Plus Environment

Page 13 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. p53 expression in M14 (A) and Huh7 (B) cell line after 24 and 48 h of 8gꢀi
derivatives treatment.
These compounds enhance p53 expression in M14 cells at 24 hrs (an effect that disappeared
after 48 h of incubation), but not in Huh7 cells (Figure 4). Thus, the variation of p53
expression observed in M14 cell line is due to a specific effect of these compounds. The
increase of p53 in M14 cells, sensitive to the compounds, might suggest that, although
mutated, it could still have a role in the regulation of cell proliferation.
Furthermore, we examined whether treatment with 8gꢀi modified the nuclear translocation of
p53. Accumulation of p53 in the nuclear fraction increased in cells treated for 24 h compared
with control cells, an effect that disappeared after 48 h of incubation in accordance with the
p53 expression results (Figure 5).
Figure 5. Nuclear p53 expression in M14 cells after 24 and 48 h of 8gꢀi treatment. Same
amounts of nuclear lysate were loaded on 12% SDSꢀPAGE, blotted onto nitroꢀcellulose
membrane, and p53 antibody was used to visualize the p53 nuclear fraction.
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
We focused our attention on p53 protein since the critical role of this transcription factor in
cell cycle control is well documented. In recent years the biological relevance of a specific
p53ꢀstatus (wildꢀtype or mutated) has been evaluated in more detail for many cell lines. For
example, it has been reported that some p53 mutants expressed in cell lines have a residual or
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
16
a temperatureꢀsensitive transcriptional activity ; additionally, some mutant p53 proteins
“
gain of function” and controversial p53 status in many cell lines have been recently
17
described.
To the best of our knowledge, only one mutation has been reported for the p53 protein in
1
8
19
M14 (G266E) and Huh7 cell line (Y220C). Furthermore, a recent study on a set of
spiro(oxindoleꢀ3,3′ꢀthiazolidine)ꢀbased derivatives tested on M14 melanoma cell line found a
20
potential inhibition of p53ꢀMDM2 interaction on the basis of NMR studies. Finally,
increasing evidences support a role for p53 and its major negative regulators,
MDM2/MDM4, in melanogenesis without an exhaustive analysis of single melanoma cell
21
line.
NMR analysis. The ability of compound 8i to block p53ꢀMDM2 interaction was investigated
1
5
by NMR analysis. Holak et al. have recently described a twoꢀdimensional NꢀHSQC based
22,23
NMR assay to determine the effect of antagonists on proteinꢀprotein interactions.
The
method, named AIDA (for antagonist induced dissociation assay), provides information on
whether an antagonist of a proteinꢀprotein interaction is strong enough to dissociate the
complex, and whether its mode of action is modulated by denaturation, precipitation, or
release of a protein in its functional folded state. AIDA requires the use of a large protein
fragment (larger than 30 kDa) to bind to a small reporter protein (less than 20 kDa). In
appropriate conditions (flexible residues), 1D proton NMR spectra may suffice for
monitoring the states of proteins in complexes upon treatment with ligands. Because of the
14
ACS Paragon Plus Environment

Page 15 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
highly flexible nature of the Nꢀterminal domain of p53, p53ꢀMDM2 complex is suitable for
N
ε
1
D proton NMR application. In particular, the H side chains of W23 and W53 produce
sharp lines in the free p53 1D proton spectrum. On forming the complex with MDM2, W23
signal disappears, since W23, together with the p53 residues 17ꢀ26, comprises the primary
binding site for MDM2. Upon binding, these residues participate in wellꢀdefined structures of
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
large p53ꢀMDM2 complexes, whereas W53 is still not structured when p53 is bound to
22ꢀ24
MDM2.
2
The observed 1/T transverse relaxation rate of the bound W23 in the complexes
increases significantly, and the broadening of NMR resonances results in the disappearance
of this signal from the spectra.
Figure 6 shows the NMR spectrum of the tryptophan residues of the p53ꢀMDM2 complex
N
ε
(
0.1 mM); only W53 H side chains signal can be detected. The addition of 8i from a stock
DMSO solution (0.1 mM formal final concentration) to the p53ꢀMDM2 complex caused
1
sample precipitation. The full 1D H NMR spectrum of the supernatant solution did not
change compared to that acquired before 8i addition, indicating that only 8i precipitated after
its addition. In particular, the W23 peak was still not observable (Figure 6b) indicating that
the p53ꢀMDM2 was not dissociated. Nutlinꢀ3 was also used as positive control (Figure 6c).
22,23
Its addition (0.1 mM) caused a complete p53 release.
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 6. Oneꢀdimensional proton spectrum of the side chains of tryptophans (W) of p53ꢀ
MDM2 complex (a), after addition of 8i (b), and nutlinꢀ3 (c).
16
ACS Paragon Plus Environment

Page 17 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Due to the observed precipitation of 8i, the solubility of the compound was measured in the
25
same buffer solution used for NMR experiments adopting Lipinski et al. method. The
compound was dissolved in DMSO at a concentration of 4,6 ꢀg/ꢀL to obtain a 10 mM stock
solution which was added to the phosphate buffer solution; precipitation was quantified by
measuring UV absorbance in the 500ꢀ900 nm range (Figure 3SI). It resulted that the
solubility of 8i is ≤3.1 ꢀg/mL (≤6.8 ꢀM).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
6
The high affinity of the MDM2ꢀp53 complex (K
D
~ 0.7 ꢂM) along with the high
concentration of the two proteins required for the NMR analysis (100 ꢀM), allows the
22,23
detection of only very soluble inhibitors.
Thus, in the experimental conditions needed for
NMR analysis, 8i would be not able to displace p53 even if it was a high affinity ligand of
MDM2.
Conformational analysis and SAR Studies. A comprehensive conformational analysis was
performed on the newly synthesized derivatives 8aꢀn (Table 1). The conformational space
was sampled by using a protocol including molecular dynamics (Simulated Annealing) and
mechanic calculations followed by quantum mechanical (PM6) full geometry optimization
(see Experimental Section). Obtained conformers were ranked by their conformational
energy values and grouped into families. For clarity of presentation, herein only results
obtained for compound 8i are reported (results obtained for 8e, 8g, 8h, 8j, 8k, and 8l are
available in the Supporting Information).
Firstly, low energy conformers, i.e. within 5 kcal/mol from the global minimum energy value
(ꢁEGM ≤5), were classified into different conformational families, named TCC, CCC, TCT,
TTC, and CCT, according to the synperiplanar or antiperiplanar conformation of the torsion
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
angles τ1, τ2, and τ3, and occurrence rates together with the range of ꢁEGM values were
calculated for each family (Table 3, Tables 1SIꢀ6SI).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 3. ꢁEGM values (kcal/mol), torsional angle values (degrees) and occurrence rates (%)
of conformational families of 8i considering PM6 conformers within 5 kcal/mol from the
global minimum.
e
O
τ
3
H
N
1
f
g
h
c
b
Cl
τ
τ
2
N
d
H C
Oa
3
H C
3
Occurrence
Rate
a
ꢁEGM
Torsional Angles
(°)
Family
(
kcal/mol)
(
%)
b
c
d
τ1
0.00ꢀ1.54 ~180
1.57ꢀ2.95 ~0.00 ~0.00 ~0.00
τ2
τ3
TCC
CCC
TCT
TTC
CCT
~0.00 ~0.00
35.0
19.0
25.0
15.0
6.00
2.08ꢀ3.81 ~180
2.46ꢀ3.85 ~180
~0.00
~180
~180
~0.00
~180
3.65ꢀ4.63 ~0.00 ~0.00
a
b
The values reported refer to the lowest and the highest energy conformers of the family. τ1
c
d
torsional angle is calculated considering the amide bond atoms. τ2: a, b, c, and d atoms. τ3:
e, f, g, and h atoms.
Since the presence of three aromatic substituents on the pyrrole ring is necessary for
biological activity, they were considered as pharmacophoric moieties, and their relative
distances were calculated in all low energy conformers (Tables 4, 7SI, and 8SI; Figure 7).
18
ACS Paragon Plus Environment

Page 19 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 4. Intraꢀmolecular distances (Å), torsional angle values (degrees) and occurrence rates
(%) of conformational subꢀfamilies of 8i considering PM6 conformers within 5 kcal/mol
from the energy global minimum.
O
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
H
Y
τ5
c
b
N
k
m
Z
Cl
l
Nd
τ4
H C
O
3
i
j
X
H C
3
Torsional
Occurrence
Rate
Subꢀ
a
Family
Distances (Å)
Angles
family
(°)
(%)
d1
XꢀY)
5.7
8.8
5.6
8.4
4.8
7.9
6.1
8.8
4.6
6.3
5.7
7.9
d2
(XꢀZ)
9.0
9.0
9.0
8.9
8.5
8.7
8.1
7.0
9.2
9.2
8.3
6.9
d3
b
c
τ4
τ5
(
(YꢀZ)
9.1
TCC
CCC
I
II
I
~90
~90
~85
~85
~80
~105
~90
~105
~85
~90
~85
~105
~ꢀ100
~90
20.0
15.0
10.0
6.00
1.00
2.00
12.5
12.5
14.0
1.00
5.00
1.00
8.0
8.6
~100
~ꢀ95
~60
II
III
IV
I
4.6
5.6
8.4
~ꢀ70
~ꢀ90
~90
TCT
TTC
CCT
11.2
11.1
10.7
8.9
II
I
~ꢀ100
~90
II
I
10.6
9.4
~100
~ꢀ80
II
a
Conformational enantiomers present the same absolute values of torsional angles, with τ4
and τ5 having the opposite sign. τ4: c, d, i, and j atoms. τ5: b, k, l, and m atoms.
b
c
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 7. Molecular models of 8i conformational families, classified as in Table 4. The
lowest energy conformer of each family is displayed as sticks. Conformers belonging to other
subꢀfamilies are superimposed by pyrrole ring atoms and displayed as lines. Heteroatoms are
colored by atom type: O, red; N, blue; S, yellow; Cl, light green. Hydrogens are omitted for
clarity with the exception of that bond to the amide nitrogen.
It resulted that each conformational family is characterized by more than one set of intraꢀ
molecular distances, according to the values of torsion angles τ4 and τ5, giving rise to
different subꢀfamilies.
Active compounds 8gꢀ8i showed a similar conformational behavior; moreover, since each set
of intraꢀmolecular distances is univocally associated to one conformational subꢀfamily, they
also share the same orientation of the aromatic rings X, Y, and Z (Tables 4, 7SI, and 8SI). By
consequence, the different potency showed by compounds 8gꢀ8i (activity rank order
8i>8g>8h, Table 2) is due to specific molecular interactions with the target established by: i)
20
ACS Paragon Plus Environment

Page 21 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
the paraꢀsubstituent of the X ring (R
2
; 8i vs. 8g and 8h, Table 1) and ii) the metaꢀsubstituent
of the Y ring (R ; 8i and 8h vs. 8g, Table 1). Indeed, structureꢀactivity relationships (SARs)
3
indicate strict requirements for the substitution of the aromatic rings X and Y. The presence
of a metaꢀhalogen substituent on the Y ring (R ) resulted crucial for biological activity (8e vs.
3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
8g and 8h, Tables 1ꢀ2), and its replacement with a nitro group led to an inactive compound
(8l). On the other hand, the paraꢀsubstitution of the X ring caused either an improvement or a
loss of biological activity, depending on the nature of the R substituent (8i, 8j, and 8k vs. 8h,
2
Tables 1ꢀ2). To further investigate these SARs, the conformational features of active
analogues 8gꢀ8i were compared to that of inactive analogues 8e (Table 1SI) and 8jꢀl (Table
4SIꢀ6SI). It resulted that for all compounds the three most populated conformational families
are TCC, CCC, and TCT, with the following rank order: TCC>TCT>CCC. Significant
differences in the conformational behavior of active vs. inactive compounds can only be
found in the rate of TTC conformers, although not in all cases. In detail, the unꢀsubstituted
analogue 8e shows a decrease of TTC conformers (1%) and an increase of TCC conformers
(43%) with respect to active compounds 8gꢀ8i, whereas the replacement of the chlorine atom
at R (8i) with a nitro group (8l) does not alter the conformational preference of the resulting
3
structure (Table 4 vs. Table 6SI). By consequence, it is likely that the inactivity of 8l is due to
unfavorable steric interactions with the target caused by the greater size of the nitro group
with respect to the chlorine atom of 8i. Also compound 8k, which presents a trifluoroꢀmethyl
substituent at R instead of the methyl of 8i (Table 1), showed a conformational behavior
2
comparable to 8i, with only a slight increase of TCC and a slight decrease of CCC
conformational families (Table 4 vs. Table 5SI). Since the trifluoroꢀmethyl of 8k and the
methyl group of 8i share similar steric parameters, then the strong electronegativity of the
CF substituent is responsible for the activity decay of 8k. Accordingly, also the inactive
3
compound 8j is characterised by the presence of an electronꢀwithdrawing group at R (Table
2
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1), however, in the case of 8j, an increase of TCT conformational family (35%, Table 4SI),
and the absence of TTC conformers are observed too.
Finally, inactive analogues 8m and 8n, both characterized by the presence of a bulky
substituent at the ortho position on ring Y (Table 1), presented different electronic, steric, and
conformational features compared to active 8gꢀ8i (data not shown).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Structural comparison with LxxLLꢀlike motifs. In order to evaluate the peptidomimetic
ability of our active analogues, a structural analysis was performed on the Xꢀray structures of
LxxLLꢀlike motifs involved in proteinꢀprotein interactions, such as: i) p53 in complex with
HDM2 protein (PDB ID: 1YCR) and ii) nuclear receptor coꢀactivator 5 in complex with the
estrogen receptor beta (PDB ID: 2J7X). Motif residues responsible for proteinꢀprotein
interactions were identified and the distances between the centroids of the interacting side
chains were calculated (Table 5).
Table 5. Distances between the centroids of the side chains of interacting residues in LxxLLꢀ
like motifs.
PDB
Motifs
Distances (Å)
ID
i,i+3 i,i+4 i,i+7 i+3,i+4 i+3,i+7 i+4,i+7
345
9
349
a
LxxLL
2J7X
7.9
7.8
5.0
5.3
ꢀ
8.6
1
26
b
FxxLWxxL
1YCR
8.2
10.2
9.7
4.3
a
b
Nuclear receptor coactivator 5. p53.
Intraꢀmolecular distances reported in Table 5 were crossed with those of 8gꢀi (Tables 4, 7SI,
and 8SI). Conformational subꢀfamilies matching the specified similarity criteria were selected
22
ACS Paragon Plus Environment

Page 23 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(see Experimental Section for details). Low energy PM6 conformers belonging to the
345
349
selected conformational subꢀfamilies were superimposed on the
LxxLL
and/or
19
26
FxxLWxxL motif by fitting the centroids of the aromatic rings X, Y, and Z on the
centroids of the side chains of the corresponding residues. It emerged that some
conformational subꢀfamilies of 8gꢀi are able to reproduce the orientation of i, i+3, i+4
residues of LxxLLꢀlike motifs, like those present in p53 (i.e. CCC IꢀII, TCT I, TTC I, CCT I)
and in nuclear receptor coꢀactivator 5 (i.e. TCC I, CCC IꢀII) (Figure 8 and Figure 4SI).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 8. Superimposition of TCC I (A), CCC I (B), CCC II (C), TCT I (D), TTC I (E), and
CCT I (F) lowest energy conformers of 8i (carbons in green) on the i, i+3, i+4 residues both
345
349
of
LxxLL
(A, B, C) motif present in 2J7X structure (carbons in pink) and
1
9
26
FxxLWxxL (D, E, F) motif present in 1YCR structure (carbons in pink); overlaying
amino acid position is indicated. The motif backbone and interacting amino acids are
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
displayed as stick. Heteroatoms are colored by atom type: O, red; N, blue; Cl, light green.
Hydrogens are omitted for clarity.
On the other hand, conformational subꢀfamilies TCC II, CCC IV, TCT II, TTC II, and CCT II
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
19
26
are able to mimic the i, i+3, i+7 residues of the FxxLWxxL motif present in 1YCR
(Figure 9 AꢀE).
Figure 9. Superimposition of TCC II (A), CCC IV (B), TCT II (C), TTC II (D), and CCT II
(E) lowest energy conformers of 8i (carbons in green) on the i, i+3, i+7 residues of
1
9
26
FxxLWxxL motif in 1YCR structure (carbons in pink) and of CCC III (F) conformer of 8i
24
ACS Paragon Plus Environment

Page 25 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
9
26
(carbons in green) on the i, i+4, i+7 residues of FxxLWxxL motif in 1YCR structure
(
carbons in pink); overlaying amino acid position is indicated. The motif backbone and
interacting amino acids are displayed as stick. Heteroatoms are colored as follows: O, red; N,
blue; Cl, light green. Hydrogens are omitted for the sake of clarity.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Thus, just by rotating rings X and Y, i.e., by varying τ4 and τ5 (Table 4, Figure 7), 8gꢀi are
able to mimic the i, i+3, and i+4, as well as, the i, i+3, and i+7 residues of LxxLLꢀlike motifs.
Nevertheless, only CCC III conformers are able to reproduce the orientation of the
hydrophobic side chains of i, i+4, i+7 residues (Figure 9 F), which represent the key
19
26
interacting residues of the p53 FxxLWxxL binding motifs in its interaction with MDM2.
It is noteworthy that, apart from the superimposition of the hydrophobic moieties, the overall
shape of the molecule and the positioning of the heteroatoms (including the direction of some
carbonyl functions) of the active analogues 8gꢀi reproduce those of the motif interacting
residues.
1
9
26
Interestingly, due to the pseudoꢀsymmetry of the FxxLWxxL motif, some 8gꢀi
conformational enantiomers are able to reproduce this functional motif by fitting specular
residues (Figure 5SI).
Pharmacophore generation and 3Dꢀdatabase searching. To further investigate the
molecular mechanism of action of the new derivatives, pharmacophore models were
generated and used to perform 3Dꢀsearches in molecular databases. A pharmacophore model
was produced for each conformational subꢀfamily of 8gꢀi, i.e., for each set of pharmacophoric
distances (Table 9SI), (Catalyst Software, see Experimental Section for details). The three
phenyl rings and the halogen atom on ring Y were selected as pharmacophore fragments. The
generated pharmacophore models were then employed to perform as many 3Dꢀdatabase
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
searches in molecular databases (see Experimental Section). The resulting 1529 hits were
analyzed and filtered on the basis of the compatibility of their reported biological activity
27
with the activity profile of our active analogues (PubChem BioAssay database;
www.ncbi.nlm.nih.gov/pcassay). Firstly, only compounds whose biological activity on M14
cells has been explored were taken into consideration, this reduced the number of hits to 189.
Among them, only 19 were reported to be active in the micromolar range (<10 ꢀM) on M14
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
cells.
Two
of
them
(Chart
1),
NSC671136
(CID
383092,
pubchem.ncbi.nlm.nih.gov/summary/summary.cgi?cid=383092) and NSC659611 (CID:
6712209, pubchem.ncbi.nlm.nih.gov/summary/summary.cgi?cid=6712209), presented the
same pharmacological profile of our derivatives, being active on M14 (AID=25; Source:
DTP/NCI), and not on MCF7 cells (AID=83; Source: DTP/NCINCI).
Chart 1. Selected 3Dꢀdatabase search hits.
Cl
O
S
Cl
N
HN
NH
O
HN
O
N
O
O
O
N
O
S
Br
S
HN
NSC671136
NSC659611
Compounds NSC659611 and NSC671136 fit the pharmacophore models of CCC II, TCT II,
TCC I, and TTC II conformers of 8gꢀi (Figures 10 and 6SI).
26
ACS Paragon Plus Environment

Page 27 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 10. A) NSC659611 low energy conformer resulting from 3Dꢀdatabase
pharmacophore search, matching distances are indicated. B) Superimposition of CCC II
lowest energy conformer of 8i (carbons: orange) on NSC659611 (carbons: white). C)
NSC671136 low energy conformer resulting from 3Dꢀdatabase pharmacophore search,
matching distances are indicated. D) Superimposition of TCT II lowest energy conformer of
8i (carbons: cyan) on NSC671136 (carbons: white). Heteroatoms are colored by atom type:
O, red; N, blue; S, yellow; Cl, light green.
While
for
NSC659611
only
growth
inhibition
assays
are
reported
(http://pubchem.ncbi.nlm.nih.gov/assay/assay.cgi?cid=6712209), several possible molecular
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
targets are indicated for NSC671136, since its biological activity was explored at the
molecular
level
using
cellꢀbased
assays
(pubchem.ncbi.nlm.nih.gov/assay/assay.cgi?cid=383092). Importantly, all reported possible
protein targets of NSC671136 are involved in proteinꢀprotein interactions regulating DNA
transcription. Moreover, the ability of NSC671136 to inhibit MDM2ꢀMDM4 heterodimer
formation has also been proven (AID=489028 and 485346; Source: Burnham Center for
Chemical Genomics). Noteworthy, the MDM2–MDM4 complex is more efficient in targeting
p53 for ubiquitylation and degradation than MDM2 homodimer, and inhibition of the
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
21
MDM4ꢀp53 interaction restored p53 function in melanoma cells. This fascinating
hypothesis on the mechanism of action of our derivatives needs deeper investigation to be
corroborated. The synthetic feasibility to increase the chemical diversity around the pyrrolic
core of our lead compounds, will allow the introduction of solubilising groups to provide
analogues more amenable to NMRꢀbased studies, in order to shed light on the putative
involvement of p53ꢀMDM2ꢀMDM4 axis in their mechanism of action.
Conclusions
According to the results obtained through biological investigations, our derivatives 8gꢀi
interfere with DNA replication, being selectively active against M14 melanoma cells. Our
lead 8i induce a block in the S phase of the cell cycle and an alteration of p53 expression
level and nuclear translocation in sensitive cells, but not in other cells containing different
p53 mutations. SAR and molecular modeling studies strongly support the hypothesis that the
ability of our derivatives to mimic LxxLLꢀlike protein recognition motifs is responsible for
their antiꢀproliferative activity. Moreover, the results of our 3Dꢀdatabase pharmacophore
search showed that 8gꢀi share pharmacophoric features and cytotoxicity profile with a
28
ACS Paragon Plus Environment

Page 29 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
compound reported to inhibit MDM2ꢀMDM4 complex formation, supporting their
hypothesized peptidoꢀmimetic ability.
Since proteinꢀprotein interactions follow a precise structural code, then understanding the
structural basis of this “language” holds the key to develop novel, highly effective, and
selective drug candidates. Although further evidences are needed, the selective cytotoxicity
toward M14 cell line suggests a mechanism of action based on the interaction with proteins
that are involved in a specific cell signaling pathway, representing a potential benefit in the
search for new leads in targeted cancer therapies.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The results of the present study will drive the rational synthesis of new analogues,
introducing substituents aimed to optimize, both, interaction with the hypothesized targets
and solubility needed to perform NMR analysis. This approach will allow us to explore the
pharmacological potentiality of our leads as mimetic of protein functional motifs.
Experimental section
Chemistry. Melting points were taken on a Gallenkamp melting point apparatus and are
1
13
uncorrected. H and C NMR spectra were recorded at 300 and 75 MHz respectively, using a
Bruker Avance 300 MHz spectrometer. Splitting patterns are described as singlet (s), doublet
(d), triplet (t), quartet (q), quintuplet (qt), and broad (br). Mass spectra were obtained by
electrospray ionization (ESIMS) using a Thermo Finningan LCQ Deca XP Max ionꢀtrap
mass spectrometer equipped with Xcalibur software. Chromatographic separations were
performed on silica gel (Kieselgel 40, 0.040ꢀ0.063 mm, Merck). Reactions and products
mixtures were routinely monitored by thinꢀlayer chromatography (TLC) on Merck 0.2 mm
precoated silica (60 F254) aluminum sheets, with visualization by irradiation with a UV
lamp. All target compounds meet the criteria of >95% purity, as confirmed by combustion
analysis performed on a PerkinꢀElmer 240 °C elemental analyzer. Yields refer to purified
29
ACS Paragon Plus Environment