Synthesis, in vitro, and in cell studies of a new series of [Indoline-3,2′-thiazolidine]-based p53 modulators

Article abstract of DOI:10.1021/jm400311n

Analogues of the previously described spiro[imidazo[1,5-c]thiazole-3, 3′-indoline]-2′,5,7(6H,7aH)-trione p53 modulators were prepared to explore new structural requirements at the thiazolidine domain for the antiproliferative activity and p53 modulation. In cell, antiproliferative activity was evaluated against two human tumor cell lines. Derivative 5-bromo-3′-(cyclohexane carbonyl)-1-methyl-2-oxospiro[indoline-3,2′- thiazolidine] (4n) emerged as the most potent compound of this series, inhibiting in vitro 30% of p53-MDM2 interaction at 5 μM and the cell growth of different human tumor cells at nanomolar concentrations. Docking studies confirmed the interactions of 4n with the well-known Trp23 and Phe19 clefts, explaining the reasons for its binding affinity for MDM2. 4n at 50 nM is capable of inducing the accumulation of p53 protein, inducing significant apoptotic cell death without affecting the cell cycle progression. Comparative studies using nutlin in the same cellular system confirm the potential of 4n as a tool for increasing understanding of the process involved in the nontranscriptional proapoptotic activities of p53.

Full text of DOI:10.1021/jm400311n

Subscriber access provided by Georgetown University | Lauinger and Blommer Libraries
Article
Synthesis, in vitro, and in cell studies of a new series
of [indoline-3,2'-thiazolidine]-based p53 modulators.
Alessia Bertamino, Maria Soprano, Simona Musella, Maria Rosaria Rusciano, Marina Sala,
Ermelinda Vernieri, Veronica Di Sarno, Antonio Limatola, Alfonso Carotenuto, Sandro Cosconati,
Paolo Grieco, Ettore Novellino, Maddalena Illario, Pietro Campiglia, and Isabel Gomez-Monterrey
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/jm400311n • Publication Date (Web): 12 Jun 2013
Downloaded from http://pubs.acs.org on June 13, 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 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Synthesis, in vitro, and in cell studies of a new series
of [indolineꢀ3,2'ꢀthiazolidine]ꢀbased p53 modulators.
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
a
b
a
b
a
Alessia Bertamino, Maria Soprano, Simona Musella, Maria Rosaria Rusciano, Marina Sala,
a
a
c
c
Ermelinda Vernieri, Veronica Di Sarno, Antonio Limatola, Alfonso Carotenuto, Sandro
d
c
c
b
a
Cosconati, Paolo Grieco, Ettore Novellino, Maddalena Illario, Pietro Campiglia, Isabel
c*
Gomez-Monterrey.
a
Depart. of Pharmacy, University of Salerno, 84084 Fisciano, Salerno, Italy
b
Depart. of Cellular and Molecular Biology, University of Naples Federico II, 80131 Naples,
Italy
c
Depart. of Pharmacy, University of Naples Federico II, 80131 Naples, Italy
d
DiSTABiF, Second University of Naples , 81100 Caserta, Italy
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 53
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
Analogs of the previously described spiro[imidazo[1,5ꢀc]thiazoleꢀ3,3’ꢀindoline]ꢀ2’,5,7(6H,7aH)ꢀ
trione p53 modulators, were prepared to explore new structural requirements at the thiazolidine
domain for the antiproliferative activity and p53 modulation. In cell, antiꢀproliferative activity
was evaluated against two human tumor cell lines. Derivative 5ꢀbromoꢀ3'ꢀ(cyclohexane
carbonyl)ꢀ1ꢀmethylꢀ2ꢀoxospiro[indolineꢀ3,2'ꢀthiazolidine] (4n) emerged as the most potent
compound of this series, inhibiting in vitro a 30% of p53ꢀMDM2 interaction at 5ꢀM and the cell
growth of different human tumor cells at nanomolar concentrations. Docking studies confirmed
the interactions of 4n with the well known Trp23 and Phe19 clefts explaining the reasons for its
binding affinity for MDM2. 4n at 50 nM is capable to induce the accumulation of p53 protein,
inducing significant apoptotic cell death without affecting the cell cycle progression.
Comparative studies using nutlin in the same cellular system confirms the potential of 4n as a
tool for increasing understanding of the process involved in the nonꢀtranscriptional proꢀapoptotic
activities of p53.
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
INTRODUCTION
p53 is the tumor suppressor protein which has been the most intensively studied for nearly 30
1
years. p53 functions mainly as a transcription factor by binding to specific DNA sequences and
2
by transactivating or repressing a large group of target genes. Through these downstream
targets, p53 pathway coordinates cell cycle arrest, DNA repair, apoptosis and senescence to
preserve genomic stability and prevent tumor formation in response to cellular insults, including
1
ꢀ4
DNA damage, hypoxia, and a deficiency of growth factors or nutrients. Recently, novel
functions of p53 in metabolism regulation and tumor motility, invasion and metastasis have
ACS Paragon Plus Environment
2

Page 3 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
5,6
been unveiled. p53 levels are tightly controlled in unstressed mammalian cells through a
continuous cycle of ubiquitylation and degradation by the 26S proteasome, regulated primarily
7
through the interaction of p53 with the ringꢀfinger ubiquitin E3 ligase MDM2. This protein
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
controls the p53 levels through a direct binding interaction that neutralizes p53 transactivation
activity, exports nuclear p53, and targets it for degradation via the ubiquitylationꢀproteasomal
8
ꢀ10
pathway.
Loss of p53 activity, either by deletion, mutation, or MDM2 overexpression, is the
1
1,12
most common event in the development and progression of cancer,
while the restoration of
1
,13
endogenous p53 function results in tumor regression in vivo. In this context, the rescue of the
impaired p53 activity and reꢀsensitization to apoptosis in cancer cells by disrupting the MDM2ꢀ
1
4
p53 interaction offers an opportunity for anticancer therapeutics.
The tractability of the
MDM2ꢀp53 interaction as a drug target has been demonstrated and a number of classes of potent
15
16
small molecule inhibitors have been developed. Vassilev et al. identified the first group of
molecules that target the MDM2ꢀp53 interaction (Figure 1). These imidazoline derivatives,
designated Nutlins, specifically bind and dissociate MDM2 from p53, leading to extensive p53
activation and induction of a fullꢀblown p53 response, which can trigger tumour shrinkage in
experimental animals. The derivative Nutlinꢀ3 is currently undergoing Phase I clinical evaluation
17
against advanced solid tumors and hematological malignancies. The benzodiazepines and the
18
spiroꢀoxindole–based compounds are other classes of small molecules that have been found to
target the p53ꢀMDM2 interaction. These results led to the preclinical development of
TDP665759 and MIꢀ319, which disrupt the binding of MDM2 to p53 in vitro and suppress the
growth of tumor cells both in vitro and in vivo. Both compounds limit tumor growth without
causing major toxicity in the surrounding tissue. Although these products induce high levels of
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 53
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
p53, these are not sufficient to activate the apoptotic cascade. Other smallꢀmolecule inhibitors
1
9
20
include chromenotriazolo pyrimidines (1), and oxoindoles (2).
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 1. Structure of different inhibitors of p53ꢀMDM2 interaction.
Most of the inhibitors we have described above share a common design of a rigid heterocyclic
scaffold highly functionalized with appropriate aryl/alkyl groups, which are supposed to mimic
the critical p53 residues that bind MDM2. Different structural studies reveal that the p53ꢀMDM2
protein interaction is mediated by the 15ꢀresidue α−helical transactivation domain of p53, which
21
inserts into a hydrophobic cleft on the surface of MDM2. Three residues within this domain,
Phe19, Trp23, and Leu26, are essential for MDM2ꢀbinding, making this site an attractive target
ACS Paragon Plus Environment
4

Page 5 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
to design small molecules able to mimic the contacts and the orientations of these key amino
2
2
acids. Previously, we have reported a series of spiro(oxindoleꢀ3,3’ꢀthiazolidine)ꢀbased
2
3
derivatives potentially able to mimic at least two critical p53 residues that bind MDM2.
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
Compound 3 inhibited cell growth of different human tumor cells at submicromolar
concentrations (IC in the range of 0.4ꢀ0.9 ꢀM). This derivative induced apoptotic cell death
5
0
after 24 hrs of treatment at cytotoxic concentrations, but did not alter the normal course of cell
cycle. However, 3 induced a timeꢀdependent increment of p53 expression, indicating that the
activity profiles of the compound might be regulated by this protein. More concretely, NMR
studies performed on compound 3 demonstrated the ability of this compound to block p53ꢀ
MDM2 interaction. In this paper, we describe the design, synthesis and SAR studies of a more
flexible series of spiro(oxindoleꢀ3,3’ꢀthiazolidine) derivatives, leading to antiproliferative
compounds with significantly improved potency and cellular activity over the parent compounds.
RESULTS AND DISCUSSION
Design
Compound 3 was designed starting from the known spiroꢀoxindole derivative MIꢀ319 (Figure
23
1
). Its spiro(oxindoleꢀ3,3’ꢀthiazolidine) scaffold allows interesting and efficient structural
modifications. One of them, opening the imidazole ring, is described in this paper. In fact, we
considered it of interest to manipulate the imidazo nucleus of compound 3 with the aim of
altering its conformational properties. The opening of this ring would allow us to access more
flexible structures (series 4 and 5, Figure 2) with a potential third point of diversification through
the 4’ꢀcarboxyl group of the thiazolidine moiety. This ring could allow the aromatic and/or alkyl
side chains to assume more appropriate orientations to interact with the binding site.
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 53
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
Accordingly, two small libraries of compounds were considered for synthesis. Both series retain
the ester group at position Cꢀ4’ while the oxoindole moiety carries either a weak releasing (CH3)
or a withdrawing (Br) electron groups (R). Series 4 contains a substituted phenyl, benzyl, or a
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
cyclohexyl side chains on the Nꢀ3’ (R ), while these groups were positioned on Nꢀ1 in the series
2
5 (R1).
Figure 2. Design of new series 4 and 5. Dashed lines define hypothetical interaction subsites.
Chemistry
The new (2’R,4'R)ꢀethyl 3'ꢀsubstitutedꢀ2ꢀoxospiro[indolineꢀ3,2'ꢀthiazolidine]ꢀ4'ꢀcarboxylate
derivatives (4aꢀf, 4lꢀn) were prepared applying the synthetic route shown in Scheme 1. Starting
spirooxoindolethiazolidine skeletons (13ꢀ18) were constructed by condensation between the
isatin derivatives (7ꢀ12), and cysteine ethyl ester in EtOH. These derivatives were obtained with
80–90% yields, as (2’R)/(2’S) epimeric mixtures ranging from 60/40 to 40/60 ratios as we
ACS Paragon Plus Environment
6

Page 7 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
23,24
previously described.
The 3’-acyl derivatives were obtained by reaction of compounds 13ꢀ18
with the corresponding 2ꢀ(4ꢀchlorophenyl)acetyl, 4ꢀchlorobenzoyl, 4ꢀmethylbenzoyl, benzoyl or
cyclohexanecarbonyl chlorides in THF using TEA as base.
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
a
Scheme 1.
Synthesis of 2ꢀOxospiro[indolineꢀ3,2'ꢀthiazolidine]ꢀ4'ꢀcarboxylate Derivatives
(
series 4 and 5)
a
Reagents and conditions: (i) CysꢀOEt, NaHCO in EtOH; (ii) R ꢀCOCl, TEA, THF, 2 h, room
temp. See Table 1 for the correspondence between number of final products and substituent R,
R , and R .
3
2
1
2
In these conditions, all final products (4aꢀ4f and 4hꢀ4q) were obtained as single diastereomers
in 32ꢀ58% overall yields. These stereoselectivity in the acylation reactions of thiazolidine
24
25
derivatives has been previously observed by us and other authors and can be explained by the
fact that thiazolidines undergo facile ring opening and closure reactions. This favors the
formation of thermodynamically more stable diastereoisomers. Moreover, the reaction of
intermediates 13ꢀ15 (R = H) with 4ꢀCl benzoyl chloride gave also the 1ꢀsubstituted derivatives
1
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 53
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
5
aꢀc with 8ꢀ12% yields. In the synthesis of the 3ꢀcyclohexylcarboxy derivatives 4nꢀ4q, two
1
isomers were obtained in a 11/1ꢀ6/1 ratio (estimated by H NMR), which differ for the cis/trans
configuration at the N3’ꢀCOC H amide bond. The major isomer (cis) in these mixtures was
6
11
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
identified on the basis of the ROE observed in the ROESY spectrum of 4n between the Hꢀ1” of
the cyclohexanecarbonyl group and the Hꢀ4’ of the thiazolidine (Figure 3 and Figure S1 in
Supporting Information). Cis/trans isomerization about amide bond was evidenced by 2D
2
6
NMR. In fact, interconversion was demonstrated by an exchange crossꢀpeak between the Hꢀ4’
hydrogen signals of the two isomers (Figure S1).
Figure 3. ROE interaction observed between Hꢀ4’ and Hꢀ1” in the ROESY spectrum of
compound 4n.
Changes in the condition of reactions that involve an increase of the reaction time (2h to 12 or
24h) do not significantly modify the above described results, while increase of temperature
resulted in a general decrease of reaction yields. However, with the use of microwave (Scheme
2
), the reaction of intermediate 15 with 4ꢀchlorobenzoyl chloride led to a mixture of two
diastereoisomers (2’S, 4’R) 4f and (2’R,4’R) 4g and the 1ꢀsubstituted derivative 5c in 33, 15, and
% yields respectively. In the same condition and using cyclohexane carbonyl chloride as
9
acylation agent we observed the formation of only a diastereoisomer (2’S, 4’R) 4q (42%), the 1ꢀ
ACS Paragon Plus Environment
8

Page 9 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
cyclohexacarbonyl derivative 5d (25%), and a minor product with structure of 2ꢀoxoꢀ5'Hꢀ
spiro[indolineꢀ3,2'ꢀthiazole]ꢀ4'ꢀethyloxycarbonyl 6d with a 15% yield.
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
Scheme 2. Reaction of Ethyl 5ꢀbromoꢀ2ꢀoxospiro[indolineꢀ3,2'ꢀthiazolidine]ꢀ4'ꢀcarboxylate with
4ꢀchlorobenzoyl and cyclohexanecarbonyl chlorides.
a
MW. 10 minutes, 100 °C, 2 bar
The assignment of the relative configuration at the Cꢀ2’ asymmetric center as R was
determined on the basis of a ROE enhancement between Hꢀ4’ and Hꢀ4 observed in the 2D
ROESY spectra of 4g. Absolute configurations were defined hypothesizing configuration
2
3
retention at Cꢀ4’ according to structural determination achieved with the previous series. In
contrast, this ROE enhancement was not observed in any of the other derivatives, which were
assigned to the S stereochemistry at Cꢀ2’.
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 53
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 1ꢀ(4ꢀchlorobenzoyl)ꢀ2ꢀoxospiro[indolineꢀ3,2’ꢀthiazolidine] derivatives (5aꢀ5c) were
prepared using an alternative synthetic route as showed in Scheme 3. Treatment of isatins 7ꢀ9
with 4ꢀchlorobenzoyl chloride in DCM and TEA gave the corresponding intermediates 19ꢀ21
which were condensed with cysteine ethyl ester in EtOH. In this condition, compounds 5aꢀ5c
were obtained as single (2’S,4’R) diastereoisomer in 60ꢀ63% yields according to their 1D and 2D
NMR spectra. In addition, working with the 5ꢀsubstituted intermediates 20 and 21, we also
observed the formation of thiazoline derivatives 6b and 6c with 11 and 8% respectively.
Moreover, the reaction of 7ꢀ9 with cyclohexanecarbonyl chloride did not lead to the desired 1ꢀ
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
(cyclohexylcarbonyl) derivative, recovering the starting materials. Poor reactivity of cyclohexane
carbonyl chloride could be explained, if compared to 4ꢀchlorobenzoyl chloride, by considering a
higher steric hindrance of the first. Modifications of reaction conditions including changes of
solvent, increase of temperature or use of microwaves led to the same negative results.
a
Scheme 3. Synthesis of (2’S,4'R) Ethyl 2ꢀoxospiro[indolineꢀ3,2'ꢀthiazolidine]ꢀ4'ꢀcarboxylate
Derivatives (5) and Ethyl 2ꢀoxoꢀ5'Hꢀspiro[indolineꢀ3,2'ꢀthiazole]ꢀ4'ꢀcarboxylate Derivatives (6)
a
Reagents and conditions: (i) 4ꢀClꢀC H COCl, TEA, THF, 2 h, room temp. (ii) CysꢀOEt,
6
4
NaHCO in EtOH.
3
ACS Paragon Plus Environment
10

Page 11 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
BIOLOGICAL EFFECTS
Antiproliferative activity
The spirooxoindolethiazolidine derivatives were examined for antiproliferative activity against
two tumor cell lines: the human breast adenocarcinoma MCFꢀ7, and human colon carcinoma
HTꢀ29 at 24 hrs. The obtained IC values are summarized in Table 1.
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
5
0
Table 1. Antiproliferative Activity of Spiro[indolineꢀ3,2'ꢀthiazolidine] (4 and 5) and
Spiro[indolineꢀ3,2'ꢀthiazole] (6) Derivatives.
O
O
O
OEt
OEt
OH
S
S
S
N
R2
N
R2
N
R2
R
Br
Br
O
O
O
O
O
O
N
N
N
H
R1
4a-f,4h-4q
4
g
4r
O
O
OEt
R
OEt
S
S
NH
R
N
O
N
O
N
R1
6
R1
5
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 53
a
IC (ꢀM±SD)
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
50
b
b
Comp.
R
H
R1
H
R2
MCFꢀ7
HT29
3
1.21± 0.6 1.60 ± 0.4
> 5 1.00±0.2
4a
H
H
CH
CH
CH
2
2
2
C
C
C
6
H
H
H
4
(4ꢀCl)
4b
CH
Br
H
3
H
6
6
4
4
(4ꢀCl) 4.81±1.0 0.78±0.2
(4ꢀCl) 2.90±0.8 0.66±0.1
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
4c
H
4d
H
C
C
C
C
C
C
C
6
6
6
6
6
6
6
H
H
H
H
H
H
H
4
4
4
4
4
4
4
(4ꢀCl)
(4ꢀCl)
(4ꢀCl)
(4ꢀCl)
(4ꢀCl)
(4ꢀCl)
(4ꢀCl)
2.15 ±0.7 3.69±0.9
2.12±0.7 1.09±0.6
0.90±0.2 0.11±0.09
3.00 ±0.2 2.00±0.8
4.52±1.1 0.18±0.09
1.23±0.4 0.12±0.07
0.52±0.3 0.08±0.01
4e
CH
Br
Br
H
3
H
4f
H
4g
H
4h
CH
CH
CH
CH
CH
CH
CH
CH
CH
H
3
3
3
3
3
3
3
3
3
4i
CH
Br
Br
Br
Br
Br
CH
H
3
4j
4k
CH C H (4ꢀCl) 0.27±0.1 0.36±0.09
2 6 4
4l
C
6
H
5
0.31±0.1 0.21±0.2
4m
C
6
H
4
(4ꢀCH
3
) 0.06±0.05 0.09±0.05
4n
Cyclohexyl 0.04±0.01 0.07±0.01
4
o
p
q
3
Cyclohexyl
1.20±0.6 1.10±0.6
2.30±0.8 1.90±0.8
0.22±0.1 0.56±0.1
2.01±0.9 1.90±0.7
1.01±0.6 1.03±0.8
3.46±0.9 0.23±0.1
0.15±0.1 0.02±0.01
2.08±0.8 1.40±0.8
2.78±0.9 0.21±0.1
0.86±0.4 0.20±0.1
1.63±0.6 0.85±0.4
4
Cyclohexyl
4
Br
Br
H
Cyclohexyl
4r
CH
3
Cyclohexyl
5a
COC
6
6
6
H
4
H
4
H
4
(4ꢀCl)
(4ꢀCl)
(4ꢀCl)
H
H
H
H
5b
CH
3
COC
5c
5d
6b
6c
6d
Br COC
Br
Cyclohexyl
CH
COC
6
H
4
4
(4ꢀCl)
(4ꢀCl)
3
Br COC
Br
H
6
Cyclohexyl
ACS Paragon Plus Environment
12

Page 13 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
b
Data represent mean values (SD) of three independent determinations. Human breast
c
adenocarcinoma cell line. Human colon carcinoma cell line.
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
Compound 3, considered our hit compound, showed a similar micromolar antiproliferative
activity against the two tumoral cell lines used in the assay. The opening derivatives 4a, 4b and
4c showed an elevated activity with IC50 values in the ꢁM range against MCFꢀ7 and
submicromolar against HT29 cell lines. Exchange of the 4ꢀchlorobenzyl group for 4ꢀ
chlorophenyl gave compounds 4dꢀ4f, more active against MCFꢀ7 cells while 4f was 6ꢀfold more
active against the HT29 cell line. The activity data for compounds 4aꢀ4c, 4dꢀ4f, and 4hꢀ4j (IC50
from 0.08 to > 4.0 ꢁM) indicated that the nature of the substituents on the oxoindole moiety
markedly affects the antiꢀproliferative activity profile of these compounds. Contrary to what was
2
3
observed in the precedent series (compound 3), the presence of an electronꢀwithdrawing group,
such as the bromide group, at position Cꢀ5 of the indole system caused an increase of the activity
of the corresponding analogues 4c, 4f, and 4j in both cell lines, but, particularly on HT29 cells.
We observed that the configuration at the 2’ carbon has a notable influence on the cytotoxic
activity on this cell line. In fact, (2’R, 4’R) 4g was 18ꢀfold less potent than its diastereoisomer
(2’S, 4’R) 4f on HT29 cells, and 3ꢀfold less potent on the MCFꢀ7 cell line. The introduction of a
methyl group at Nꢀ1 improved the activity of compounds 4hꢀj compared to the nonꢀmethylated
analogues (4dꢀf), specially on colon cell line. The most potent compound of this subseries, 4j,
gave IC values of 520 and 80 nM in MCFꢀ7 and HT29 cell lines, respectively. Modifications of
5
0
4j at the Nꢀ3’position produced different effects: the introduction of 4ꢀClꢀbenzyl group increased
2ꢀfold) the antiproliferative activity of analogue 4k on MCFꢀ7 cells and reduced (4ꢀfold) its
(
activity on HT29 cells. A similar behavior was observed with the compound 4l which contains a
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 53
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
phenyl group at Nꢀ3’ position. The introduction of a 4ꢀCH ꢀphenyl or a cyclohexyl group led to
3
compounds 4m and 4n, which showed an antiꢀproliferative activity in the nanomolar range (IC50
<
100 nM, for both cell lines). In particular, 4n was 14ꢀfold more potent than its analogue 4j on
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
MCFꢀ7 cells. Further modifications of this compound involving substitution or loss of the Br
atom at the C5 (compounds 4o and 4p), lack of the CH group at N1 (compound 4q), as well as
3
ethyl ester hydrolysis to carboxylic acid (4r), all resulted in a loss of activity in the resulting
compounds. These derivatives were less potent than 4n against MCFꢀ7 ( from 5ꢀ to 60ꢀfold) and
HT29 (from 8ꢀ to 28ꢀfold) cell lines.
Furthermore, switching the 4ꢀClꢀbenzoyl or cyclohexyl carbonyl groups from position N3’ to
position N1 led to contrasting results. 4ꢀchlorobenzoyl derivative 5c showed cytotoxic activity in
the nanomolar range on both cell lines (IC = 150 and 20 nM) and was ~ 6 fold more potent than
5
0
its regioisomer 4f. Derivative 5a was also 2ꢀ3 fold more potent than its regioisomer 4d, while
compound 5b containing a CH group at Cꢀ5 position showed a slight decrease of activity on
3
MCFꢀ7 cells compared to its analogue 4e. In contrast, compound 5b was 4ꢀfold more potent than
4e against HT29 cells. Surprisingly, the same change of the position for cyclohexyl carbonyl
group (4q versus 5d) led to a considerable decrease in activity. In fact, 5d was 10ꢀfold less
potent than its regioisomer 4q on MCFꢀ7 cells. Finally, the presence of a more planar thiazoline
ring in the structure produced different effects: derivative 6b retained the cytotoxic activity of its
thiazolidine analogue 5b, while compound 6c was 6ꢀ and 10ꢀfold less active than 5c on both cell
lines. In contrast, compound 6d showed a slight increase of activity (~1.5 fold) on both cell lines
compared to 5d.
The antiꢀproliferative activity of the most potent compound against both cell line, 4n, was also
analyzed against a panel of human tumor cell lines, including PC3 (prostate), U937 (leukaemia),
ACS Paragon Plus Environment
14

Page 15 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Calu (lung), HEPG2 (liver), and C643 (anaplastic thyroid) human cell lines (Table 2).
Doxorubicin and Nutlinꢀ3 were used as reference cytotoxic agents. Also data on the MCFꢀ7 cell
line are reported in Table 2. In all tested cells lines, compound 4n showed marked cytotoxic
potency with IC s in the range 0.07ꢀ0.55 ꢀM, while Nutlinꢀ3 was much less effective in our
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
5
0
2
7
panel in accordance with some data found in the literature. 4n was 18ꢀfold more potent than
doxorubicin on the Calu cell line and was equipotent to doxorubicin in PC3 and U937 cell lines.
Table 2. Antiproliferative Activity of 4n on Multiple Human Tumor Cell Lines and One Normal
Cell Line.
IC (ꢀM±SD)
5
0
Origin tumour
Breast
Cell line
MCFꢀ7
PC3
4n
Nutlinꢀ3
Dox
2
7a
0.04±0.01
0.41±0.21
0.07±0.01
0.10±0.06
0.14±0.06
0.55±0.08
2.9±0.31
30.3±2.9
0.02±0.01
0.75±0.10
0.12±0.03
1.81±0.33
0.08±0.01
0.07±0.01
2
7b
Prostate
Leukemia
Lung
U937
15.6±1.9
27.2±5.3
Calu
2
7c
Liver
HEPG2
C643
10.2±5.1
Anaplastic Thyroid
Origin normal
23±11.2
1.40±3.6
Human Gingival
Fibroblast
HGF
1.60±0.15
0.50±0.15
a
Data represent mean values (SD) of three independent determinations at 24 hrs.
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 53
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
Table 2 also shows that 4n inhibited the cellular growth of a human gingival fibroblast (HGF)
normal cell line at low micromolar concentration. This cytotoxicity was similar to that showed
by Nutlinꢀ3 on the same line and was 4 times less than that caused by doxorubicin. These data
seem to indicate that 4n has a good profile of cell selectivity.
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
In vitro modulation of p53ꢀMDM2 interaction
In order to test the ability of 4n to inhibit p53ꢀMDM2 interaction we performed an in vitro
TM
binding assay using ImmunoSet
p53/MDM2 complex ELISA (Figure S4). Nutlinꢀ3 and
compound 3 were also evaluated as references. In this assay and for all compounds, the
minimum effective concentration was determined at 5 ꢀM. At this concentration, the percentage
inhibition was 19% for Nutlinꢀ3 and 25% for 3, while compound 4n has proved more effective
inhibiting 30% of p53ꢀMDM2 interaction.
Molecular Modeling studies
To better rationalize the reasons behind the activity of our indolineꢀ3,2'ꢀthiazolidines molecular
docking studies were undertaken on 4n which is the most potent antiproliferative agent in this
series. To date, several Xꢀray structures are reported of the complex between MDM2 and
different small molecules and, among these, we decided to choose the structure having PDB code
2
8
1
LBL in which MDM2 was solved at a very high resolution (1.60 Å) in complex with a spiroꢀ
oxindole derivative that is structurally related with the derivatives described herein. Therefore,
2
9
this structure was used to dock, through the Autodock4.2 (AD4) software, compound 4n as
well as the coꢀcrystal ligand. Docking of the latter compound confirmed the good performances
of this software which was able to predict the experimental binding pose with a rootꢀmean square
ACS Paragon Plus Environment
16

Page 17 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
deviation (rmsd) of 1.03 Å. In regards to 4n, the same calculations converged towards a single
solution in which the lowest energy (ꢂGAD4) binding conformation was also belonging to the
most populated cluster (f ). In particular, in the binding pose (ꢂGAD4= ꢀ8.32 kcal/mol,
occ
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
focc=86/100) predicted for 4n (Figure 4) the ligand is inserted in the MDM2 binding site so that
the 3ꢀcyclohexylcarboxy substituent is buried in the soꢀcalled Trp23 pocket making favorable
van der Waals contacts with Ile61, Val75, Phe86, Phe91 and Ile99 sideꢀchains. On the other
hand, the MDM2 Phe19 subpocket is occupied by the ligand ethyl ester chain that is able to
make direct contacts with Ile61 and Met62 sideꢀchains. The presence of this interaction could
partially explain why hydrolysis of the ethyl ester chain to the corresponding carboxylic acid
results in the loss of activity, on the other hand, reduction of membrane penetration cannot be
ruled out. Interestingly, the Leu26 cleft does not seem to be completely filled by the 5ꢀbromoꢀ2ꢀ
oxoindoline nucleus that, instead, forms additional hydrophobic interactions with Leu54 through
its N1ꢀmethyl substituent. This latter hydrophobic interaction could explain why 4dꢀf are
generally less active than their methylated analogues 4hꢀj. In this position, edgeꢀface π–π
interaction with the Phe55 residue are also established in a shallow and rather liphophilic portion
of the MDM2 protein in which the 5ꢀbromine atom fills the crevice between Phe55, Met62, and
Gln59. Indeed, the interaction with Phe55 residue has already been detected through Xꢀray
3
0
studies for other structurally unrelated ligands in a very recent work by Olson’s group
Interestingly, in the binding pose suggested by AD4 the ligand N3’ꢀCOC H amide bond adopts
6
11
a cis configuration as already suggested by ROESY experiments.
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 53
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 4. Predicted binding mode for compound 4n in the MDM2 binding site. The ligand is
represented as pink stick while the protein as blue sticks and ribbons and transparent white
surface.
Modulation of p53ꢀMDM2 interaction in cell
To confirm in cell the ability of 4n to inhibit the MDM2ꢀp53 interaction, the expression levels
of these two proteins were evaluated by western blot after 24, 48 and 72 hrs of 4n treatment and
after 24 hrs with Nutlinꢀ3. Compound concentrations close to their IC on the MCFꢀ7 cell line
50
were used in this assay (i.e., 50 nM for 4n and 3 ꢁM for Nutlinꢀ3). In these conditions, 4n
induced the accumulation of p53 and MDM2 proteins, (Figure 5A and 5B). To determine
whether 4n was able to prevent MDM2ꢀp53 interaction, p53 expression levels were measured
after MDM2 immunoprecipitation (Figure 5C and 5D). Cells were treated with 4n for 24, 48, and
72 hrs or Nutlinꢀ3 for 24 hrs. MDM2 was immunoprecipitated and the samples underwent
western blot to visualize p53, and evaluate its association with MDM2. Treatment with 4n
ACS Paragon Plus Environment
18

Page 19 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
reduced the MDM2ꢀp53 interaction as evidenced by a significant decrease in the p53 levels
bound to MDM2 at 24 and 48 hrs. Nutlinꢀ3 seems to be inactive in this immunoprecipitation
assay. We can hypothesize that Nutlin acts more quickly than 4n; its unresponsiveness to the test
may be due to a recombination of p53 and MDM2 within 24 hrs. Interestingly, this
recombination is also observed for 4n after 72 hrs.
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 5. (A) MCFꢀ7 cells were treated with 50 nM 4n for 24, 48 and 72 hrs or 3 ꢀM Nutlinꢀ3
for 24 hrs. Total cell lysates were analyzed by western blotting for p53 and MDM2 with specific
antibodies. (C) A similar experiment was performed for p53 after MDM2 immunoprecipitation.
(B and D) Immunoblots were quantified by ImageQuant densitometric analysis. Protein
expression levels were measured in arbitrary densitometric units and data show the means ±
SEM calculated from relative protein expression levels determined in three separate experiments.
Cellꢀcycle progression
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 53
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
To investigate the effect in the reduction of tumor cell survival mediated by 4n, MCFꢀ7 cell
cycle was analyzed after 24, 48 and 72 hrs of treatment. Again, Nutlinꢀ3 was used for
comparison. Nutlinꢀ3 (3 ꢀM) induced cell cycle arrest, as shown by G0/G1 phase block, and in
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
1
6
accordance with literature data. Compound 4n (50 nM) did not significantly affect cell cycle.
Indeed, 4nꢀtreated cells showed the same distribution in G1, S, G2/M phases of untreated cell at
24, 48 or 72 hrs (Figure 6A, 6B, and 6C). When the same experiment was performed with
increasing concentration of 4n (500 nM, 1 ꢀM, 5 ꢀM), no difference in cell cycle effect was
found compared to initial concentration (Figure 6 D).
Figure 6. Effects of 4n on cell cycle progression analyzed for DNA content by FACS in breast
cancer MCFꢀ7 cells untreated or treated with 50 nM 4n or 3 ꢀM Nutlinꢀ3 at 24 (A), 48 (B) and
72 hrs (C) and (D) MCFꢀ7 cells untreated or treated with 50, 500, 1000 and 5000 nM of 4n at
ACS Paragon Plus Environment
20

Page 21 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
4 hrs. The distribution and percentage of cells in G1, S and G2/M phase of the cell cycle are
indicated. Data points are means SEM. Significance assumed at *p<0.05, **0.01<p<0.05,
***p<0.01
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
Apoptotic cell death
Cell death was evaluated with preliminary, qualitative assessment of apoptosis by subG1
analysis. The treatment with 4n (50 nM) showed a strong increase of the subG1 peak, that rose in
a time dependent manner (Figure 7A). Nutlinꢀ3 (3 ꢀM) also induced a significant increase of cell
death. However, the subG1 peak may consist of apoptotic and necrotic cells: to discriminate
between the two possibilities, an annexin V binding assay was performed. 4n and Nutlinꢀ3
induced a significant increase of cell fraction in early apoptosis (Figure 7B). In particular, the
apoptotic cell percentage increased from 4% of untreated cells to 10% and 12% of cells after 48
and 72 hrs of 4nꢀtreatment. Nutlinꢀ3 determined a lower increase of early apoptosis compared to
4
n.
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 53
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. (A) Representative subG1 populations calculated from FACS histograms. MCFꢀ7 cells
were incubated with 50 nM 4n or with 3 ꢁM Nutlinꢀ3 for 24, 48 and 72 hrs. Data are expressed
as the percentage of subG1 cells. (B) A similar experiment was performed and the cells were
analysed by FACS for the occurrence of apoptosis (FITCꢀannexin binding). Values represent the
mean ± SEM. Significance assumed at *p<0.05, **0.01<p<0.05 ***p<0.01
2,3
Although p53 plays a pivotal role in regulating cell cycle and apoptosis, treatment of MCFꢀ7
cells with 4n or Nutlinꢀ3 has a different impact on both mechanisms. Indeed, whereas Nutlinꢀ3
determined a cell cycle arrest, 4n induced apoptotic cell death. To investigate the cell death
mechanism induced by 4n, the p53ꢀdependent apoptotic pathway was analyzed by Western blot.
The analysis of the expression levels of p53 and p53 transcriptional targets, such as p21 and p27,
showed a progressive increase of these proteins during the treatment with 4n in a time dependent
ACS Paragon Plus Environment
22

Page 23 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
manner (Figure 8). However, these increases do not seem to be sufficient to induce cell cycle
arrest. On the contrary, Nutlinꢀ3 showed a greater accumulation of p21 than 4n, at 24, 48 and 72
hrs, inducing cell cycle arrest (Figure 8A and 8B). Cells committed to die via p53ꢀdependent
apoptosis typically follow the mitochondrial pathway, although p53 can also modulate cell death
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
3
1,32
through death receptors.
p53 has been reported to trigger apoptosis by modulation of gene
3
1ꢀ33
transcription of Bcl2 family members and by physical interaction with these proteins.
Western blot results indicated that 4n and Nutlin regulate Bclꢀ2 members, reducing the antiꢀ
apoptotic BclꢀxL protein and increasing the proꢀapoptotic BclꢀxS protein, although the two drugs
affect BclxL/S expression at different times.
.
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 53
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
Figure 9. (A) MCFꢀ7 cells were treated with 50 nM 4n or 3 ꢀM nutlin for indicated time
interval. Total cell lysates were analyzed by western blotting for phosphotyrosine p53, p21, p27,
BclꢀxL/xS, cytocrome c and caspaseꢀ3 with specific antibodies. (B) Immunoblots were
quantified by ImageQuant densitometric analysis. Protein expression levels were measured in
arbitrary densitometric units and data show the means ± SEM calculated from relative protein
expression levels determined in three separate experiments. *p<0.05 vs protein expression levels
in untreated control cells.
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
The induction of 4nꢀmediated apoptosis was confirmed by the release of cytochrome c from
mitochondria to the cytosol, that increased at various time intervals becoming particularly
evident at 72 hrs of drug treatment. Also Nutlinꢀ3 affects cytochrome c release at 48 h even if at
3
4,35
lower levels compared to 4n.
One of the final effectors of the apoptotic process is caspaseꢀ3,
that is activated both by extrinsic and intrinsic or mitochondrial pathways. The cleavage of
caspaseꢀ3 is a clear indicator of apoptosis, hence we analyzed the p53ꢀfragment with a specific
antibody. Western blot analysis revealed that both 4n and Nutlinꢀ3 induced an accumulation of
the cleavage product of caspaseꢀ3. All these finding suggest that, in the MCFꢀ7 breast cancer cell
line, cell cycle arrest is the main molecular mechanism mediated by Nutlinꢀ3. On the contrary,
50nM 4n did not block cell cycle progression but rather it induced apoptosis.
CONCLUSIONS
Here we report the synthesis, and biological evaluation of a series of new 2ꢀ
oxospiro[indolineꢀ3,2'ꢀthiazolidine] derivatives, designed as inhibitors of p53ꢀMDM2 proteinꢀ
protein interaction, which exhibit activity against different tumor cell lines. In particular,
compound 4n showed high efficacy in MCFꢀ7 (breast), HT29 (colon), calu (lung), and U937
(leukemia) human cell lines with IC values of 40, 70, 100, and 70 nM, respectively, with >10ꢀ
5
0
fold selectivity against the HGF normal cell line. The well known p53ꢀMDM2 inhibitor, Nutlinꢀ
ACS Paragon Plus Environment
24

Page 25 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
3
, was considerably less efficient in all tested cell lines and also in the in vitro p53ꢀMDM2
binding inhibition assay. Docking studies allow us to predict binding mode for 4n in the MDM2
binding site, in which our compound sets favorable van der Waals contacts with Ile61, Val75,
Phe86, Phe91 and Ile99 sideꢀchains, direct contacts with Ile61 and Met62 sideꢀchains, and
additional hydrophobic interactions with Leu54 and edgeꢀface π–π interaction with the Phe55
residue. In the MCFꢀ7 cell line, 4n induces a timeꢀdependent increment in p53 expression,
indicating that its activity profile, as in the case of nutlins, might be regulated by this protein.
However, preliminary studies on the induction of apoptosis and cell cycle progression, showed a
different behavior for these compounds. Nutlinꢀ3 markedly blocked the G0/G1 phase, causing a
delay of cell cycle progression in responsive cells, while treatment with 4n did not alter the
normal course of cell cycle. Moreover, 4n induces apoptotic cell death through the intrinsic
apoptotic pathway. According to these findings, 4n represents a new lead compound for the
design of agents able to reactivate p53ꢀmediated apoptosis in cancer cells and provides a new
tool for studying p53 signaling in the natural cellular context of tumors that retain wildꢀtype p53.
Further experiments aimed to identify more potent spirothiazolidinꢀbased derivatives and to
better understand the apoptotic mechanisms induced by this compound are underway.
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
EXPERIMENTAL SECTION
General. Reagents, starting materials, and solvents were purchased from commercial suppliers
and used as received. Analytical TLC was performed on plates coated with a 0.25 mm layer of
silica gel 60 F254 Merck and preparative TLC on 20 cm x 20 cm glass plates coated with a 0.5
mm layer of silica gel PF254 Merck. Silica gel 60 (300ꢀ400 mesh, Merck) was used for flash
chromatography. Melting points were determined by a Kofler apparatus and are uncorrected.
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 53
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
13
Optical rotations were measured on an Atago Polax 2ꢀL polarimeter. H NMR and C NMR
spectra were recorded with a Varianꢀ400 spectrometer, operating at 400 and 100 MHz,
respectively. Chemical shifts are reported in δ values (ppm) relative to internal Me Si, and J
4
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
36
values are reported in hertz (Hz). ROESY experiment was recorded at 25 °C in the phaseꢀ
3
7
sensitive mode using the method from States. Data block sizes were 2048 addresses in t2 and
5
12 equidistant t1 values. Before Fourier transformation, the time domain data matrices were
multiplied by shifted sin2 functions in both dimensions. A mixing time of 500 ms was used.
ESIMS experiments were performed on an Applied Biosystem API 2000 tripleꢀquadrupole
spectrometer. Starting spiro(oxoindolethiazolidine) ethyl ester derivatives (13ꢀ18) were
synthesized as described in references 23 and 24. As example, here we described the synthesis
of (3RS,4’R)ꢀethyl 5ꢀbromoꢀ1ꢀmethylꢀ2ꢀoxospiro[indolineꢀ3,2’ꢀthiazolidine]ꢀ4’ꢀcarboxylate
(
18). Combustion microanalyses were performed on a Carlo Erba CNH 1106 analyzer, and all
reported values are within 0.4% of calculated values (Table 1S in Supporting Information).
These elemental analyses confirmed >95% purity.
Synthesis of (3RS,4′R) ethyl 5ꢀbromoꢀ1ꢀmethylꢀꢀ2ꢀoxospiro[indolineꢀ3,2′ꢀthiazolidine]ꢀ4′ꢀ
carboxylate (18).
NaHCO (1.0 g, 12 mmol) and 5ꢀbromoꢀ1ꢀmethyl isatin (12, 2.4 g, 10 mmol) were added to a
3
solution of LꢀCysꢀOEt (2.3 g, 12 mmol) in ethanol (100 mL) and the suspension was stirring at
room temperature for 12 h. Then, the suspension was filtered and the filtrate was concentrated.
Spiro(oxoindolethiazolidine) ethyl ester residue was dissolved in DCM and washed with water
(
3 x 50 mL). The combined organic layer was dried over anhydrous sodium sulfate, filtered, and
concentrated. A 5:1 diastereoisomeric mixture of the title’s compound was obtained, as oil with
1
7
3% yied. H NMR (400 MHz, CDCl ) δ 1.36 (3H, m, CH ), 3.10 and 3.12 (3H, s, CH ), 3.31
3
3
3
ACS Paragon Plus Environment
26

Page 27 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
and 3.43 (1H, m, Hꢀ5’a), 3.71 and 3.93 (1H, m, Hꢀ5’b), 4.25 (4H, m, CH ), 4.46 and 4.66 (2H,
2
m, Hꢀ4’), 6.69 and 6.75 (1H, d, J = 8.0 Hz, Hꢀ7), 7.48 (1H, d, Hꢀ6), 7.60 and 7.71 (1H, s, Hꢀ
4).The compound was used in the next reaction without further purification
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
General Procedure for the Synthesis of the (2’R o 2’S, 4'R)ꢀethyl 3'ꢀsubstitutedꢀ2ꢀ
oxospiro[indolineꢀ3,2'ꢀthiazolidine]ꢀ4'ꢀcarboxylate Derivatives (4aꢀq). To a solution of
(
2’R,4’R)ꢀ and (2’S,4’R)ꢀethyl 2ꢀoxospiro[indolineꢀ3,2’ꢀthiazolidine]ꢀ 4’ꢀcarboxylate derivatives
13ꢀ18, 200 mg, 5 mmol) in dry THF (50 mL) was added a solution of corresponding 4ꢀ
(
chlorobenzoyl or 4ꢀmethylbenzoyl or 4ꢀchlorophenylacethyl or cyclohexane carbonyl chlorides
5.5 mmol) in THF (10 mL) and TEA (10 mmol). The reaction mixture was stirred at room
(
temperature for 2 h and water was then added. The organic solution was washed with water (3x
100 mL), dried over Na SO , and evaporated in vacuo. Flash chromatography on silica gel, using
2
4
ethyl acetate/nꢀhexane as eluent, overall yielded the correspondent final derivatives as oil.
2’S,4'R)ꢀethyl 3'ꢀ(2ꢀ(4ꢀchlorophenylacetyl)ꢀ2ꢀoxospiro[indolineꢀ3,2'ꢀthiazolidine]ꢀ4'ꢀ
carboxylate (4a)
Overall yield 43%. [α] ꢀ7.1 (c 0.1, MeOH). H NMR (400 MHz,CDCl ) δ 1.40 (t, 3H, CH );
(
25
1
D
3
3
3.44 (d, 1H, J = 12.0 Hz, Hꢀ5’a); 3.63 (s, 2H, CH ); 3.95 (dd, 1H, J = 6.0 and 11.6 Hz, Hꢀ5’b);
2
4
.39 (q, 2H, CH ); 5.08 (d, 1H, J = 6.0, Hꢀ4’); 6.76 ( d, 1H, J = 8.4 Hz, Hꢀ7); 7.04 (t, 1H, Hꢀ6);
2
7
.15 (d, 2H, J = 8.0 Hz, aryl); 7.24 (t, 1H, J = 7.9 Hz, Hꢀ5); 7.28 (d, 2H, aryl); 7.43 (s, 1H, NH);
1
3
7.47 (d, 1H, Hꢀ4). C NMR (100 MHz,CDCl ) δ 14.5 (CH ), 34.6 (Cꢀ5’), 41.4 (CH ), 62.5
3
3
2
(
CH ), 64.3 (Cꢀ4’), 73.9 (Cꢀ2’), 109.3, 115.8, 123.5, 128.0, 128.9, 130.2, 134.4, 136.2, and 143.1
2
(
aryl), 170.1, 173.9, and 178.0 (C=O). ESIMS m/z calcd for C H ClN O S, 430.08; found
21 19 2 4
4
30.16.
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 53
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
(
2’S,4'R)ꢀethyl
5ꢀmethylꢀ3'ꢀ(2ꢀ(4ꢀchlorophenyl)acetyl)ꢀ2ꢀoxospiro[indolineꢀ3,2'ꢀ
thiazolidine]ꢀ4'ꢀcarboxylate (4b)
2
5
1
Overall yield 39%.[α]
ꢀ13.6 (c 0.25, MeOH). H NMR (400 MHz,CDCl ) δ 1.42 (t, 3H,
D
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
CH ); 2.04 (s, 3H, CH ); 3.37 (d, 1H, J = 12.4 Hz, Hꢀ5’a); 3.61 (s, 2H, CH ); 3.98 (dd, 1H, J =
3
3
2
6.0 and 11.6 Hz, Hꢀ5’b); 4.43 (q, 2H, CH ); 5.04 (d, 1H, J = 6.0, Hꢀ4’); 6.70 ( d, 1H, J = 8.0 Hz,
2
Hꢀ7); 7.08 (d, 1H, Hꢀ6); 7.19 (d, 2H, J = 8.0 Hz, aryl); 7.31 (d, 2H, aryl); 7.40 (s, 1H, Hꢀ4); 7.79
1
3
(
s, 1H, NH). C NMR (100 MHz,CDCl ) δ 14.2 (CH ), 21.7 (CH ), 34.8 (Cꢀ5’), 41.6 (CH ),
3 3 3 2
6
2.3 (CH ), 64.7 (Cꢀ4’), 74.2 (Cꢀ2’), 109.8, 123.2, 128.5, 129.6, 130.7, 133.0, 137.4, and 142.8
2
(
aryl), 170.6, 173.4, and 178.3 (C=O). ESIMS m/z calcd for C H ClN O S, 444.09; found
22 21 2 4
444.15.
(2’S,4'R)ꢀethyl
5ꢀbromoꢀ3'ꢀ(2ꢀ(4ꢀchlorophenyl)acetyl)ꢀ2ꢀoxospiro[indolineꢀ3,2'ꢀ
thiazolidine]ꢀ4'ꢀcarboxylate (4c)
2
5
1
Overall yield 37%. [α]
ꢀ29.4° (c 0.4, MeOH). H NMR (400 MHz,CDCl ) δ 1.43 (t, 3H,
D
3
CH ); 3.42 (d, 1H, J = 12.6 Hz, Hꢀ5’a); 3.65 (s, 2H, CH ); 4.01 (dd, 1H, J = 6.0 and 12.0 Hz, Hꢀ
3
2
5’b); 4.45 (q, 2H, CH ); 5.00 (d, 1H, J = 6.0, Hꢀ4’); 6.73 (d, 1H, J = 7.6 Hz, Hꢀ7); 7.23 (d, 2H, J
2
1
3
=
8.0 Hz, aryl); 7.31 (d, 2H, aryl); 7.48 (d, 1H, Hꢀ6); 7.95 (s, 1H, Hꢀ4); 8.04 (s, 1H, NH).
C
NMR (100 MHz,CDCl ) δ 14.3 (CH ), 21.9 (CH ), 34.6 (Cꢀ5’), 41.5 (CH ), 62.4 (CH ), 64.9 (Cꢀ
3
3
3
2
2
4’), 74.4 (Cꢀ2’), 109.5, 116.2, 123.4, 128.2, 129.0, 131.1, 132.5, 133.3, 137.2, and 143.2 (aryl),
70.2, 173.5, and 178.9 (C=O). ESIMS m/z calcd for C H BrClN O S, 507.99; found 508.04.
1
2
1
18
2
4
(
2’S, 4'R)ꢀethyl 3'ꢀ(4ꢀchlorobenzoyl)ꢀ2ꢀoxospiro[indolineꢀ3,2'ꢀthiazolidine]ꢀ4'ꢀcarboxylate
(4d)
Overall yield 49%. [α]
25
1
ꢀ16.3 (c 0.25, MeOH). H NMR (400 MHz,CDCl ) δ 1.36 (t, 3H,
D
3
CH ); 3.40 (d, 1H, J = 11.6 Hz, Hꢀ5’a); 3.95 (dd, 1H, J = 6.0 and 11.6 Hz, Hꢀ5’b); 4.38 (q, 2H,
3
ACS Paragon Plus Environment
28

Page 29 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
CH ); 4.87 (d, 1H, J = 6.0 Hz, Hꢀ4’); 6.77 (d, 1H, J = 8.0 Hz, Hꢀ7); 7.05 (t, 1H, J = 7.6 Hz, Hꢀ6);
2
7.20 ( t, 1H, Hꢀ5); 7.33 (d, 2H, aryl, J = 7.6 Hz); 7.44 (d, 2H, aryl); 7.73 (d, 1H, Hꢀ4); 8.54 (s,
1
3
1
H, NH). C NMR (100 MHz,CDCl ) δ 14.5 (CH ), 34.5 (Cꢀ5’), 62.8 (CH ), 66.8 (Cꢀ4’), 72.4
3
3
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
(
Cꢀ2’), 110.7, 123.2, 125.5, 126,3, 128.4, 129.0, 130.3, 134.3, 137.1, and 141.2 (aryl), 168.4,
170.6, and 177.1 (C=O). ESIMS m/z calcd for C H ClN O S, 416.06; found, 416.14.
2
0
17
2
4
(
2’S,4'R)ꢀethyl
carboxylate (4e)
Overall yield 58%.[α] ꢀ5.7 (c 0.1, MeOH). H NMR (400 MHz,CDCl ) δ 1.38 (t, 3H, CH );
3'ꢀ(4ꢀchlorobenzoyl)ꢀ5ꢀmethylꢀ2ꢀoxospiro[indolineꢀ3,2'ꢀthiazolidine]ꢀ4'ꢀ
2
5
1
D
3
3
2
.34 (s, 3H, CH ); 3.34 (d, 1H, J = 11.6 Hz, Hꢀ5’a); 3.98 (dd, 1H, J = 6.0 and 11.6 Hz, Hꢀ5’b);
3
4.38 (q, 2H, CH ); 4.85 (d, 1H, J = 6.0 Hz, Hꢀ4’); 6.72 (d, 1H, J = 8.0 Hz, Hꢀ7); 7.03 (d, 1H, Hꢀ
2
1
3
6
); 7.33 (d, 2H, J = 8.0 Hz, aryl); 7.44 (d, 2H, aryl); 7.54 (s, 1H, Hꢀ4); 8.06 (s, 1H, NH). C
NMR (100 MHz,CDCl ) δ: 14.5 (CH ), 21.5 (CH ), 34.5 (Cꢀ5’), 62.7 (CH ), 66.7 (Cꢀ4’), 72.5
3
3
3
2
(Cꢀ2’), 110.2, 126.1, 128.4, 129.0, 130.8, 133.6, 136.2, and 138.4 (aryl), 167.6, 170.1, and 176.2
C=O). ESIMS m/z calcd for. per C H ClN O S, 430.08; found, 430.08.
(
2
1
19
2
4
(2’S,4'R)ꢀethyl
5ꢀbromoꢀ3'ꢀ(4ꢀchlorobenzoyl)ꢀ2ꢀoxospiro[indolineꢀ3,2'ꢀthiazolidine]ꢀ4'ꢀ
carboxylate (4f)
25
D
1
Overall yield 46%. [α]
ꢀ9.8 (c 0.24, MeOH). H NMR (400 MHz,CDCl ) δ 1.40 (t, 3H,
3
CH ); 3.35 (d, 1H, J = 12.0 Hz, Hꢀ5’a); 3.89 (dd, 1H, J = 6.0 and 11.8 Hz, Hꢀ5’b); 4.42 (q, 2H,
3
CH ); 4.84 (d, 1H, J = 6.0 Hz, Hꢀ4’); 7.36ꢀ7.50 (m, 5H, Hꢀ7, aryl); 7.78ꢀ7.81 (m, 2H, Hꢀ6, Hꢀ4);
2
1
3
8
.04 (s, 1H, NH). C NMR (100 MHz,CDCl ) δ 14.5 (CH ), 35.3 (Cꢀ5’), 63.1 (CH ), 66.7 (Cꢀ
3 3 2
4’), 72.3 (Cꢀ2’), 116.8, 118.8, 128.4, 128.6, 128.7, 129.2, 131.2, 139.3, and 140.9 (aryl), 168.7,
70.1, and 173.6 (C=O). ESIMS m/z calcd for. per C H Br ClN O S, 493.97; found, 494.09.
1
2
0
16
2
4
ACS Paragon Plus Environment
29