Synthesis, characterization, and optimization for in vivo delivery of a nonselective isopeptidase inhibitor as new antineoplastic agent

Article abstract of DOI:10.1021/jm501336h

Bis-arylidenecycloalkanones structurally related to the nonselective isopeptidase inhibitor G5 were synthesized and tested for cytotoxic activity against glioblastoma cells. Cytotoxicities correlate well with Hammett σ constants for substituted arylidene groups, confirming the proposed inhibition mechanism. A new inhibitor (2c) based on the 4-hydroxycyclohexanone scaffold, which favors apoptosis over necrosis, was selected for further development. 2c inhibited representative deubiquitinases with micromolar IC₅₀, and its proapoptotic activity was studied on several cancer cell lines. Inhibitor 2c was conjugated to PEG via dicarbamate and diester linkers. While the dicarbamate was inactive, the diester (2cPE) behaves like a prodrug and is converted into the active species 2c by secreted esterase activities. Finally, 2cPE was also tested in vivo on A549 lung carcinoma xenografts generated in mice. Intravenous treatment with 2cPE led to a significant reduction in primary tumor growth, without appreciable toxicity to mice.

Full text of DOI:10.1021/jm501336h

Subscriber access provided by UNIV OF PITTSBURGH
Article
Synthesis, Characterization and Optimization for in Vivo Delivery of
a Non-Selective Isopeptidase Inhibitor as New Anti-Neoplastic Agent
Ulma Cersosimo, Andrea Sgorbissa, Carmen Foti, Sara Drioli, Rosario Angelica, Andrea Tomasella, Raffaella
Picco, Marta Stefania Semrau, Paola Storici, Fabio Benedetti, Federico Berti, and Claudio Brancolini
J. Med. Chem., Just Accepted Manuscript • Publication Date (Web): 01 Feb 2015
Downloaded from http://pubs.acs.org on February 4, 2015
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 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Synthesis, Characterization and Optimization for in
Vivo Delivery of a NonꢀSelective Isopeptidase
Inhibitor as New AntiꢀNeoplastic Agent
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
†
#
‡#
‡
†
‡
Ulma Cersosimo, Andrea Sgorbissa, Carmen Foti, Sara Drioli, Rosario Angelica, Andrea
‡
‡
§
§
†
Tomasella, Raffaella Picco, Marta Stefania Semrau, Paola Storici, Fabio Benedetti,
*
†
*‡
Federico Berti, and Claudio Brancolini
†
Dipartimento di Scienze Chimiche e Farmaceutiche, Università degli Studi di Trieste. Via
Giorgieri 1 – 34127 Trieste ITALY.
‡
Dipartimento di Scienze Mediche e Biologiche Università degli Studi di Udine. P.le Kolbe 4 ꢀ
3100 Udine ITALY.
3
§
Structural biology laboratory, Elettra ꢀ Sincrotrone Trieste S.C.p.A., Area Science Park,
Basovizza, 34149 Trieste, Italy
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 70
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. Bis arylidenecycloalkanones structurally related to the Non Selective Isopeptidase
Inhibitor G5 were synthesized and tested for cytotoxic activity against glioblastoma cells.
Cytotoxicities correlate well with Hammett σ constants for substituted arylidene groups
confirming the proposed inhibition mechanism. A new inhibitor (2c) based on the 4ꢀ
hydroxycyclohexanone scaffold, that favors apoptosis over necrosis, was selected for further
development. 2c inhibited representative deubiquitinases with micromolar IC50 and its proꢀ
apoptotic activity was studied on several cancer cell lines. Inhibitor 2c was conjugated to PEG
via dicarbammate and diester linkers. While the dicarbammate was inactive, the diester (2cPE)
behaves like a proꢀdrug and is converted into the active species 2c by secreted esterase activities.
Finally, 2cPE was also tested in vivo, on A549 lung carcinoma xenografts generated in mice.
Intravenous treatment with 2cPE led to a significant reduction in primary tumor growth, without
appreciable toxicity to mice.
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
ACS Paragon Plus Environment
2

Page 3 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
Protein modification by the addition of the 8ꢀkDa Ubiquitin or Ubꢀlike proteins is a well
known and widespread postꢀtranslation modification not limited to influencing protein
destruction, but also their subcellular localization or the assembling into multiꢀproteins
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
ꢀ3
complexes. Ub and Ubl peptides are generally ligated to proteins by the sequential action of
three enzymes: an ubiquitinꢀactivating enzyme (E1), an Ubꢀcarrier enzyme (E2) and an Ubꢀ
protein ligase (E3). E3 enzymes show specificity in substrate choice and represent critical
3
,4
players in several signaling pathways.
Ub and Ublꢀlinkages are reversible and their cleavage entails on the involvement of a large
family of enzymes, known as isopeptidases. Although ispopetidases can be viewed as E3ꢀligase
antagonists, additional functions, such as maturation of Ub or Ubl peptides, are under their
supervision. The isopeptidase family includes deubiquitinating enzymes (DUBs), which in
principle should be specifically devoted to the rupture of Ubꢀlinkages, and other proteolytic
5
,6
enzymes, which target additional Ublꢀproteins.
7
The discovery of bortezomib and its approval for the treatment of relapsed multiple myeloma
and mantle cell lymphoma has open the field to new inhibitors targeting, more specifically,
critical enzymes of the ubiquitinꢀproteasome system (UPS) and showing less adverse side
8
,9
effects. In this scenario E3 ligases and DUBs/isopeptidases have gained increasing attention as
targets for drug development. These enzymes represent good candidates to influence the function
of proteins controlling the transformed phenotype or to induce cellular stresses, which can kill
9
ꢀ12
cancer cells.
In the last decade several reports have described the discovery, synthesis and characterization
13ꢀ23
of isopeptidase inhibitors.
Based on the specific target selectivity, they can be divided in two
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 70
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
classes: selective inhibitors, acting on a specific enzyme or on a limited number of enzymes, and
nonꢀselective isopeptidase inhibitors (NꢀSIIs), which in principle can affect the activity of
several isopeptidases. The two classes offer different and complementary advantages for the
development of new anticancer treatments. The first class guarantees advantages in terms of
selectivity when growth/survival of a specific tumor depends on a specific isopeptidase. On the
other hand, the second class, affecting more enzymes and multiple pathways, may offer
advantages in terms of effectiveness on different tumors. Hence, further studies to improve the
antiꢀneoplastic activities of inhibitors of both classes are of primary importance.
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
A subclass of NꢀSIIs brings together molecules characterized by the presence of a crossꢀ
1
3
conjugated α,βꢀunsaturated dienone with two sterically accessible electrophilic βꢀcarbons that
can act as Michael acceptors to target nucleophiles, like the catalytic cysteine of several
1
3ꢀ17
isopeptidases.
Recently, a molecule of this family, RA190, has been engineered to exert
2
3
specificity against proteasome ubiquitin receptor RPN13 (Figure 1).
Optimization of these inhibitors in terms of in vivo efficacy is a fundamental step towards their
possible use in the clinic. In this work, starting from G5, a NꢀSII previously identified by us
1
4
(Figure 1), we performed structureꢀactivity studies that lend support to the proposed
mechanism of inhibition and allowed us to develop a G5ꢀderivative optimized for in vivo antiꢀ
neoplastic activity.
ACS Paragon Plus Environment
4

Page 5 of 70
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 1. Crossꢀconjugated dienone NꢀSIIs RA190 and G5
RESULTS AND DISCUSSION
Structure-activity studies. The effect on tumor cell survival of two series of dienones was
investigated in glioblastoma U87MG cells, which in response to G5 enter both apoptosis and
24
necrosis. The first series (Table 1) is based on the bis(arylidene)ꢀtetrahydrothiapyranꢀ4ꢀoneꢀ
1,1ꢀdioxide scaffold of G5 with variations in the substituents on the aromatic rings. In the second
series (Table 2), the sulfone group of the G5 scaffold is replaced by other groups, while the 4ꢀ
nitro substituents on the aromatic rings are fixed.
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 70
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 1. Cytotoxicity of bis(arylidene)ꢀtetrahydrothiapyranꢀ4ꢀoneꢀ1,1ꢀdioxides against U87MG
glioblastoma 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
a,b
b
Compound
R1
R1’
H
R2
R2’
IC [ꢀM]
s.d.
50
1
a
H
H
H
H
H
H
H
H
H
H
5.11
4.74
6.73
3.57
2.94
2.21
1.35
1.14
0.76
2.17
2.09
1.39
0.77
1.67
0.21
0.15
0.72
0.57
0.23
0.15
0.18
0.07
0.27
0.42
0.34
0.21
1
b
CH
3
CH
OH
OCH
OPh
F
3
1
c
OH
OCH
OPh
F
H
1
d
3
3
H
1e
H
1
f
H
1
g
CN
OCH
H
CN
H
1h
3
OCH
3
NO
NO
H
2
NO
NO
H
2
1
i
H
2
2
1
j
NO
NO
NO
NO
2
2
2
2
OH
OCH
1k
3
H
H
1
l
CH
3
H
H
G5
NO
2
H
H
a
b
From cell viabilities measured with a resazurin assay, 48 hours after treatments.
Experiments were carried out in triplicate and are presented as mean values and standard
deviations (s.d.).
ACS Paragon Plus Environment
6

Page 7 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 2. Cytotoxicity of nitrobenzylidene dienones against U87MG glioblastoma cells.
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
b
Compound
X
IC [ꢀM]
s.d.
5
0
2
a
CH
2
27.5
2.0
2
b
CH(OCH
2
CH
2
O)
15.33
4.62
2.47
7.15
50.1
11.6
1.79
0.77
0.71
1.96
0.25
2.90
14.3
2.01
0.81
0.21
2
c
CHOH
2d
CHCOOC
2
H
5
2
e
NH
O
2
f
2g
S
2
h
SO
G5
SO
2
a
b
From cell viabilities measured with a resazurin assay, 48 hours after treatments.
Experiments were carried out in triplicate and are presented as mean values and standard
deviations (s.d.).
With the exception of the fluoro derivative 1f, symmetrical compounds 1a-i and G5 were
obtained by the acid catalyzed Knoevenagel condensation of tetrahydrothiapyranꢀ4ꢀoneꢀ1,1ꢀ
2
5,26
dioxide (Scheme 1, X = SO ) with aromatic aldehydes.
An alternative approach was adopted
2
for the synthesis of 1f (scheme 1), consisting in the condensation between the corresponding
aldehyde and tetrahydrothiapyranꢀ4ꢀone followed by mCPBA oxidation of the resulting sulfide
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 70
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
3
. The oxidation of the pꢀfluoro derivative 3 to the desired sulfone 1f was accompanied by the
formation of the BaeyerꢀVilliger byꢀproduct 4 (Scheme 1). The two products, in ratio 3 : 1, were
easily separated by crystallization from acetone/hexane.
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
a
Scheme 1. Synthesis of symmetrical derivatives of series 1 and 2
a
Reagents and conditions: (a) ArCHO, EtOH, aq. HCl, 15ꢀ78 %; (b) ArCHO, EtOH, Ba(OH)2,
52% (for 2b); (c) mCPBA, CH Cl , 41ꢀ63%.
2
2
Unsymmetrical compounds 1j-l were obtained as shown in Scheme 2. Tetrahydrothiapyranꢀ4ꢀ
one was initially converted into the monoꢀalkylidene derivative 5 by an aldol reaction with 4ꢀ
nitrobenzaldehyde followed by dehydration of the aldol. Acidꢀcatalyzed Knoevenagel reaction of
5
6
with the appropriate aromatic aldehyde was then followed by mCPBA oxidation of the sulfides
j-l.
a
Scheme 2. Synthesis of unsymmetrical derivatives of series 1
ACS Paragon Plus Environment
8

Page 9 of 70
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
a
Reagents and conditions: (a) (i) ArCHO, NaOH, MgSO , H O, (ii) HCl, EtOH 56%; (b)
4
2
ArCHO, EtOH, aq. HCl, 44ꢀ95%; (c) mCPBA, CH Cl , 21ꢀ90%.
2
2
All compounds of series 2 (Table 2) are symmetrical and, with the exception of 2h, were
obtained by the Knoevenagel condensation of the appropriate ketone and 4ꢀnitrobenzaldehyde
(
Scheme 1). For compound 2b, containing an acidꢀlabile acetal group, the condensation was
27
carried out in basic conditions. Sulfoxide 2h (Table 2) was obtained by mCPBA oxidation of
the corresponding sulfide 2g. Even under controlled conditions (1 eq. mCPBA, 0 °C) the reaction
gave a 1 : 1 mixture of sulfoxide 2h and sulfone G5, but the desired product 2d was readily
purified by flash chromatography on silica gel.
1
13
Hꢀ and CꢀNMR spectroscopy showed that all compounds of series 1 and 2 were single
stereoisomers with the aryl groups trans with respect to the carbonyl.
Data in Table 1 indicate that all the sulfones of series 1 are active against U87MG
glioblastoma cells with IC values in the range 0.8ꢀ7ꢀM, the dinitroderivatives G5 and 1i being
50
the most active compounds of the series. Variations of hydrophobicity (logP values) and solvent
accessible area are small in this series, reflecting the limited structural changes, and no general
correlation is observed between these parameters and citotoxicity. However, a clear correlation
can be observed between cytotoxicity and the electronic effect of the substituents on the aromatic
ring. A plot of the logarithm of the observed IC against the sum of the Hammett σ constants for
5
0
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 70
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
28
the substituents on the aromatic rings (Figure 2) reveals that cytotoxicity depends on the
electronꢀwithdrawing ability of the substituents, suggesting that the cytotoxic activity is directly
related to the electrophilicity of the enone βꢀcarbon. This is consistent with the observation that
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
alcohol 7, obtained by NaBH reduction of G5, is devoid of any cytotoxic activity and strongly
4
suggests that cytotoxicity may indeed result from the alkylation of cysteine residues present in
13,29
the catalytic site of isopeptidases by the dienone moiety.
observed also for cytotoxic
arylidenecyclohexanones, which have been postulated to selectively react as Michael
acceptors with cellular thiols.
Similar correlations have been
3
0
Nꢀacylꢀbis(arylidene)ꢀ4ꢀpiperidones,
and
3
1,32
29
ꢀ5.0
ꢀ5.5
1c
1a
1b
1d
1j 1k
1
e
1g
1f
1l
ꢀ6.0
G5
1h
1i
ꢀ6.5
ꢀ1
0
1
2
ΣσRi
Figure 2. Linear correlation between U87MG glioblastoma cells cytotoxicity (LogIC ) and
5
0
2
Hammett σ constants for compounds of series 1 (slope: ꢀ0.39 ± 0.04; r : 0.8808). ΣσRi is the
sum of the σ and/or σ for all the substituents on both rings.
p
m
ACS Paragon Plus Environment
10

Page 11 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Finally, the contribution of the X group was investigated in a series of diꢀnitro derivatives 2
(
Table 2). While all compounds in this series are cytotoxic, the presence of a polar, electron
withdrawing group in the sixꢀmembered ring appears in general to be beneficial to achieve strong
activity.
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
Assessment of the dienones-induced apoptotic and necrotic responses. In cells resistant to
apoptosis, such as mouse fibroblasts defective for Bax and Bak (Bax/Bak DKO) or glioblastoma
2
4,33
U87MG cells, G5 can also activate a caspaseꢀindependent death.
Although caspaseꢀ
independent death can engage different mechanisms, previous studies have excluded the
involvement of autophagy (Foti et al., 2009, Autophagy 5:4, 1ꢀ3) and have demonstrated that G5
3
3, 34
triggers a necrotic cell death.
Hence, we decided to study, on a small set of dienones and
using G5 as a reference, whether variations in the structure affect the proꢀnecrotic and proꢀ
apoptotic activities (Figure 3). To this purpose, we selected derivatives, 1b, 2c and 2h and also
used T98G glioblastoma cells, which preferentially die by apoptosis in response to G5 (Figure
2
4, 34
3).
As illustrated in Figure 3 (A, B) the sulfoxide 2h behaves similarly to G5, while compounds
b and 2c show a higher propensity to trigger apoptosis compared to G5 and 2h. In fact, unlike
1
the latter compounds, 1b and 2c are more effective in killing T98G than U87MG cells, as proved
by trypan blue staining and caspase activities. Cell death in T98G cells, induced by 1b and 2c, is
characterized by a robust caspase engagement. On the other side in U87MG cells, during cell
death induced by G5 and 2h, caspases are much less involved, thus testifying to the existence of
two different types of cell death. The differential induction of necrosis and apoptosis in the two
cell lines was confirmed, for 2c, by using the caspase inhibitor zVAD−fmk (Figure 3C), which
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 70
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
was effective in counteracting cell death only in T98G cells; similar results were obtained with
the other compounds (Supplementary Figure S1).
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. Apoptotic and necrotic responses induced by different dienones in U87MG (A) and
T98G cells (B). Cells were treated with the indicated concentrations of the different compounds
for 24 hours. Upper: cell death as measured by percent of trypan blueꢀpositive cells; lower:
caspase activity measured with rhodamine 110 bisꢀ(NꢀZꢀLꢀaspartylꢀLꢀglutamylꢀLꢀvalylꢀaspartic
ACS Paragon Plus Environment
12

Page 13 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
acid amide) fluorogenic substrate (relative fluorescence units) (black, G5; grey, 2h; light grey,
1b; white, 2c). (C): effect of caspase inhibitor zVAD−fmk on cell death induced by 2c (percent
of trypan blueꢀpositive cells).
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
In summary, the simultaneous presence of the sulfone/sulfoxide and nitro groups in 2h and G5
appears to promote a necrotic response; replacement of either group, as in 2c and 1b, lowers the
necrotic activity of the NꢀSII and favors the induction of apoptosis.
The ability of the compound 2c to kill preferentially through apoptosis suggests that it could
represent a better choice for in vivo applications, with respect to G5. Moreover, compound 2c,
possessing a reactive OH group, appears to be a good compromise between cytotoxicity and the
possibility to introduce modifications on the core to improve the inhibitor’s drugꢀlike properties.
For these reasons 2c was selected for further development.
Deubiquitinase inhibition by G5 and 2c. We evaluated the inhibitory profile of the two
compounds against purified isopeptidases using ubiquitinꢀAMC as a substrate. Previous studies
demonstrated that dienones are broad inhibitors, capable of inhibiting deubiquitinases but also
35
deSUMOylase activities. Thus, we investigated G5 and 2c activities against ubiquitin carboxyꢀ
terminal hydrolases UCHL1, UCHL5 and the ubiquitinꢀspecific protease USP2.
UCHL5/UCH37 is a proteasomeꢀassociated UCH, which, in conjunction with POH1/PSMD14
3
and USP14 provides to the proteasome the deubiquitinase activity. Unlike UCHL1 and UCHL3,
UCHL5 can also process polyꢀubiquitin chains and previous studies identified UCHL5 as a
16
target of dienoneꢀbased NꢀSIIs .
Both G5 and 2c weakly inhibit UCHL1 (Figure 4A) but are much less effective than other
3
6
smallꢀmolecule inhibitors. Conversely, both compounds exhibit a more pronounced inhibitory
activity against UCHL5. G5 is considerably more potent than 2c, as shown by the IC values of
5
0
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 70
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
3.58 µM and 42.99 µM, respectively. Thus, UCHL5 inhibition may play an important role in
eliciting the accumulation of polyꢀubiquitinated proteins in cells treated with these compounds.
When USP2 activity was analyzed, again G5 showed some inhibitory activity (IC 32.56 µM)
5
0
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
whereas 2c was almost inactive. This IC value is in agreement with previous studies where a
5
0
1
4, 35
compound related to G5 was used.
Finally, we also tested G5 and 2c inhibitory potency against the isopeptidase USP18, which
3
7
processes the Ubꢀlike ISG15 protein and is a key regulator of the interferon response. Both
compounds inhibit USP18 activity, and 2c appears to be slightly more active than G5, with IC50
21.20 µM and 33,23 µM, respectively. The ability of 2c to inhibit USP18 was verified also in
3
8
cells by coꢀexpressing USP18 and a fusion between ISG15 and GFP. When ISG15ꢀGFP was
coꢀexpressed with USP18, accumulation of free GFP is predominant (Figure 4B). By contrast, in
the presence of increasing concentrations of G5 or 2c, accumulation of the uncleaved ISG15ꢀ
GFP chimera can be appreciated. In the same assay, bortezomib failed in influencing USP18ꢀ
dependent cleavage of ISG15ꢀGFP.
Figure 4. (A) Inhibition of ubiquitinꢀspecific hydrolases UCHL1, UCHL5, USP2 and USP18
by G5 and 2c. (B) Immunoblot analysis showing inhibition of USP18 in A549 cells coꢀ
expressing USP18 and the ISG15ꢀGFP chimera; p53 levels were used to monitor the inhibition
ACS Paragon Plus Environment
14

Page 15 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
of the UPS. Inhibitor concentrations were: 2c (2.5, 5, 10 µM); G5 (1, 2.5, 5 µM); bortezomib
(25, 50, 100 nM). Cellular lysates were generated after 24h of treatment. Under these
concentrations comparable % of death in A549 lung cancer cells can be observed 48h later. (C)
Quantitative densitometric analysis of immunoblots. Data are presented as the mean of two
experiments.
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
Analysis of 2c pro-apoptotic activity. Next, to evaluate whether 2c shows a strong and broadꢀ
spectrum of antiꢀproliferative activities, we calculated its IC against several other cancer cell
5
0
lines. Data in Table 3 indicate that 2c antiꢀproliferative activity is indeed wide and mainly in the
low µM range, as already observed for U87MG and T98G cell lines.
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 70
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 3. Antiꢀproliferative activity of 2c and PEGylated derivative 2cPC (11) in different cell
lines
a
Cell line
IC (ꢀM)
5
0
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
2
c
11
6.9
5.7
22
U87MG
A375
glioblastoma
melanoma
4.62
6.1
16
A549
lung cancer
HT29
colorectal
adenocarcinoma
7.6
>100
Hep3B
hepatocellular
carcinoma
8.5
Mia PaCaꢀ2 pancreatic cancer
35.2
a
From cell viabilities measured with a resazurin assay, 48 hours after treatments.
We then evaluated in more detail the effect of 2c on the lung cancer cell line A549 that was
later chosen for in vivo experimentation (vide infra). Initially, we compared the apoptotic
response elicited by 2c, G5 and bortezomib to identify the concentration ranges of the three
inhibitors that induce comparable extents of cell death (Figure 5A). We then used these
concentrations to compare the effects of the three inhibitors on the activity of the UPS and on the
accumulation of the BH3ꢀonly protein Noxa, a sensor of endoplasmic reticulum stress and key
14
element of the death pathway elicited by UPS inhibitors. Immunoblot analysis (Figure 5B)
demonstrated that at the lower concentrations bortezomib is more effective in eliciting
accumulation of polyꢀubiquitinated proteins and in stabilizing p53 with respect to 2c or G5.
Interestingly, when cell death is induced with similar intensity accumulation of Noxa is
comparable in bortezomib and 2c treatments, thus suggesting that ER stress is similarly evoked
ACS Paragon Plus Environment
16

Page 17 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
by the two compounds. By contrast, high doses of G5 less efficiently promote Noxa upꢀ
regulation, consistent with the possible engagement of a necrotic response (Figure 5B).
Bortezomib
G5
2c
A
B
ꢁM
8
0
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
6
4
0
0
116
ꢀ
ꢀ
9
7
6
4
6
5
ꢀ
ꢀ
ꢀ
36
2
9
ꢀ
2
0
0
21ꢀ
p53
Noxa
Actin
ꢁM
2c
Bortezomib
G5
Figure 5. (A) Inhibitor concentrations eliciting comparable death percentage in A549 lung
cancer cells, after 48 hours of treatment. (B) Immunoblot analysis of extracts from A549 cells
treated as in A) and monitored for UPS inhibition, (polyꢀubiquitin and p53 accumulation) and
induction of ERꢀstress, (Noxa induction) after 24 hours of treatment.
Synthesis and activity of a 2c derivative optimized for in vivo studies. 2c, like many
anticancer compounds, is poorly soluble in aqueous solutions. A common strategy for
improving drugs solubility and bioavailability consists in the conjugation of the compound with
polyethylene glycol (PEG). Despite the intrinsic limitation of a low drug/carrier mass ratio when
used with low molecular weight drugs, polyethyleneglycol offers some advantages as a water
3
9ꢀ41
soluble carrier for anticancer compounds.
conjugating 2c to monoꢀmethoxy PEG (5000 Da) via the OH group and suitable linkers (Scheme
).
We thus synthesized two soluble derivatives by
3
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 70
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
a
Scheme 3. PEGylation of 2c.
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
a
Reagents and conditions: (a) N,N’ꢀdisuccinimidyl carbonate (75%); (b) succinic anhydride,
DMAP (82%); (c) MPEGꢀOH, EDC, HOBT, TEA (70%).
For the synthesis of the first conjugate (9), with a dicarbamate linker (2cPC, 2cꢀPEGꢀ
Carbamate), 2c was converted into the mixed carbonate 8 and then coupled with monoꢀ
PEGylated 1,3ꢀdiaminopropane (41 % overall yield). When tested on the panel of cancer cell
lines, however, this conjugate was inactive, incapable of triggering cell death and reducing cell
proliferation (data not shown). We thus synthesized a second derivative (11), in 57 % overall
yield, by conjugating 2c with PEG through a succinate linker (Scheme 3). We reasoned that this
conjugate (2cPE, 2cꢀPEGꢀEster), containing a more reactive diester linker, could act as a proꢀ
4
2
drug, releasing the active species 2c upon the action of cellular hydrolases and, indeed, we
found that 11 was active on three of the four cell lines tested (Table 3).
In melanoma A375 and glioblastoma U87MG cells the antiꢀproliferative activity of 2cPE (11)
is identical to that of the parent molecule 2c (Table 3). By contrast, in A549 and, even more so,
in HT29, respectively lung and colon cancer derived cell lines, the PEGylated molecule is less
effective in suppressing proliferation. To verify this observation, we compared the response of
ACS Paragon Plus Environment
18

Page 19 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
A549 cells to 2cPE (11) and 2c, by scoring cell death with trypanꢀblue assay. Dose dependent
studies (Figure 6A) and time course analysis (Figure 6C) confirm that 2cPE is less effective than
2c in triggering cell death in A549 cells. An observation further corroborated by the analysis 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
polyꢀubiquitin accumulation, p53 stabilization and Noxa induction (Supplementary Figure S2).
Similar results were obtained in HT29 cells (Supplementary Figure S3).
On the opposite, in A375 cells, which show similar sensitivity to 2c and 2cPE in resazurin
assay (Table 3), cell death (Figure 6B), accumulation of polyꢀubiquitinated proteins, p53
stabilization and Noxa upꢀregulation (Supplementary Figure S2) confirm that 2cPE behaves like
the active compound 2c. In summary, in A375 cells 2c and the PEGylated derivative 2cPE are
undistinguishable as UPS inhibitors and inducers of cell death, while A549 and HT29 cells show
some resistance to the PEGylated derivative 2cPE (11).
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 70
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. Cell death, after 48 hours, as a function of concentration in A549 (A) and A375 (B)
cells treated with 2.5ꢀ10 ꢀM 2c (black bars) and 2cPE (11) (grey bars). (C) Cell death as a
function of time in A549 cells treated with 10 ꢀM 2c (black bars) and 11 (grey bars) for 24ꢀ72
hours. (D) Response to 10 ꢀM 2c and 2cPE (11) of A549 cells grown in normal and A375
conditioned medium. (E) Response to 10 ꢀM 2c and 2cPE (11) of A375 cells grown in normal
and A549 conditioned medium. (F) Immunoblot analysis showing polyꢀubiquitin accumulation,
p53 stabilization and Noxa induction in A549 cells treated with 10 ꢀM 2c (second lane) and
2cPE (11) in normal and A375 conditioned medium (third and fourth lane, respectively). DMSO
is the negative control throughout.
Activation of the 2cPE pro-drug by secreted esterases. A different pattern of expression of
enzymes capable to hydrolytically release the active species 2c from the PEG conjugate might
explain the differential responsiveness of the four cell lines to 2cPE (Table 3). This hypothesis is
ACS Paragon Plus Environment
20

Page 21 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
supported by the observation that unresponsive A549 cells become fully responsive to 2cPE
when they are grown in the presence of A375 conditioned medium (Figure 6D), while A375 cells
retain their responsiveness to 2cPE also in the presence of A549 conditioned medium (Figure
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
6
E). Similar results were obtained with HT29 cells (Supplementary Figure S3). Accumulation of
polyꢀubiquitin chains, stabilization of p53 and Noxa induction (Figure 6F) confirm that, in the
presence of conditioned medium from A375 cells, the activities of 2c and 2cPE are
indistinguishable, thus indicating that the prodrug is readily converted into the active molecule
by this medium.
These results indicate that the response to 2cPE (Table 3) of different cell lines depends on
their ability to secrete an esterase activity capable of releasing 2c from the proꢀdrug. To gain
insight into the enzymes that might be responsible for such processing we applied a
bioinformatics analysis.
Initially, the different esterases encoded by the human genome were extracted using GEO
4
3
categories and HomoloGene. In this manner we generated a list of 173 putative esterases. Next,
to understand which secreted esterase exhibited an expression profile compatible with the
responsiveness to 2cPE, we interrogated gene expression profiles available for A375, A549 and
4
4
HT29 cells with our list. The expression of the candidate esterase is expected to be high in
A375, reduced in A549 cells and even more reduced in HT29 cells. Seven esterases satisfied this
expression profile, as illustrated in figure 7A. Among these potential candidates, we have
focused our attention on the Phospholipase A2 group 7 (encoded by the PLA2G7 gene). This
enzyme, also known as PAFꢀAH (platelet activating factor acetylhydrolase), has been
4
5
extensively characterized, and the crystal structures of both the ligandꢀfree and of the paraoxon
46
covalentlyꢀinhibited enzyme are reported. PLA2G7, which hydrolyses the ester bond at the snꢀ2
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 70
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
position of phospholipid substrates shows a canonical α/β hydrolase fold, and it has an interface
binding area that makes it capable of binding to the hydrophobic portions of lipoproteins, or to
the internal hydrophobic region of cell membranes. The interface is located at the end of a long
hydrophobic channel that leads to the catalytic serine and to the oxyanion hole. Polyethylene
glycol is known to activate the phospholipase activity of the enzyme, probably by stabilizing 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
47
interface complex between the lipase and the hydrophobic inner regions of cell membranes. To
verify that PLA2G7 can accept 2cPE as a substrate, we have built a model of the tetrahedral
covalent intermediate for the hydrolysis of 2c conjugated to an ethylene glycol dodecamer
through the same succinate linker present in 2cPE. The crystal structure of the paraoxon
covalently inhibited enzyme was used as a starting point and stepwise mutated and optimized to
build the model. The unique conformation corresponding to a stable and productive intermediate
4
8,49
according to the stereoelectronic theory of Deslongchamps
is reported in Figure 7B. In this
conformation, the first five units of the PEG chain lie inside the hydrophobic channel of the
enzyme, and 2c is partially buried inside the catalytic site, which appears to be broad and flexible
enough to host this rather large structure. The two nitro groups have a key role in stabilizing this
intermediate: one of them is found at hydrogen bond distance with the hydroxy groups of serines
319 and 349, while the aromatic ring establishes πꢀstacking interactions with phenylalanine 322
and further hydrophobic contacts with histidine 351 (omitted for clarity in the figure). The other
nitro group is in contact with the side chain of arginine 218. Interactions of arginine and serine
with the nitro groups of a ligand, such as those found in this model, are well known in antibodies
5
0
and receptor complexes.
ACS Paragon Plus Environment
22

Page 23 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
700
A
***
6
00
00
5
400
***
3
2
1
00
***
***
***
00
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
**
00
0
ENPP2 AOAH ABHD8 BCHE PNLIP LPL PLA2
RP1
G7
B
Ser349
Ser319
Ser273
Arg218
A549 cells expressing:
C
6
0
0
PURO
PLA2G7
5
40
***
30
20
10
0
DMSO
2c
2cPE
Figure 7. (A) Expression profiles in A375, A549 and HT29 cells of seven candidate esterases for
the hydrolysis of 2cPE. (B) Optimized model of the rate determining tetrahedral intermediate for
hydrolysis of 2cPE catalyzed by phospholipase A (PLA2G7). The model was built from the
crystal structure of PLA2G7 covalently inhibited by paraoxon (PDB: 3D5E). (C) Response to 10
ꢁ
M 2c and 2cPE of A549 cells engineered to express PLA2G7 or the control gene PURO.
Trypan blue analysis was performed 48 hours after treatment.
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 70
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 model confirms that 2cPE could be a substrate for phospholipase A2 secreted by A375 cells
and thus supports the hypothesis that lack of this enzyme, or of other esterases that can activate
the proꢀdrug, might explain the different low response of A549 and HT29 cell lines (Table 3). To
prove the involvement of PLA2G7 in proꢀdrug maturation, we generated by retroviral infection,
A549 cells expressing PLA2G7 isolated from A375 cells. As control we used A549 cells
expressing only the resistance gene PURO. In the presence of PLA2G7, A549 cells dramatically
increase responsiveness to the proꢀdrug in terms of cell death (Fig. 7C). A further evidence was
obtained by incubating 2cPE in the presence of a commercially available preparation of
PLA2G7; enzyme was able to fully hydrolyse 2cPE in phosphate buffer at room temperature,
yielding 2c and PEG5000, as stated by mass spectrometry measures (see supporting information
S31). Conversely, no hydrolysis was observed in the absence of the enzyme upon the same
incubation times.
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
2cPE inhibits tumor growth in mice. With convincing evidence in hand that the proꢀdrug
2cPE (11) is converted into the active form 2c by secreted hydrolases and PLA2G7 in particular,
we set to analyzing whether the PEG conjugate displays antiꢀtumor activity in vivo. To this end
we selected A549 lung cancer cells, even if these cells showed a low response to the proꢀdrug in
culture. However, we expected that, in animals, the proꢀdrug should be efficiently processed by
mouse secreted 2cPE activator PLA2G7 and, possibly, other secreted esterases.
Initially toxicity tests indicated that 2cPE is well tolerated with transient minimal effect on the
animal’s body weight up to 800 mg/kg. All treated groups showed slight (2%) body weight loss
over the course of the first week, independently from the doses. Subsequently, the body weight
of all animals steadily increased until the end of the study with final 6 – 8% gain in body weight
ACS Paragon Plus Environment
24

Page 25 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(
Supplementary Figure S4). No gross toxicity was observed throughout the treatment and 2cPE
did not induce any behavioral change or grossly visible pathological changes.
A549 lung carcinoma xenografts were generated in immuneꢀcompromised mice and when the
tumors reached the size of 0.1 mm, 2cPE (170 mg/kg) was administered three times every 4
days. After one week from last injection (day 15) the percentage of tumor volume inhibition
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
(TGI) was 72% and 78%, respectively for subcutaneous and intravenous treatments (Figure 8A).
At the end of the experiment, the reduction of primary tumor growth was particularly marked
(TGI 79%) in the i.v. treated group. No significant adverse effects were observed correlated to
drug administration.
To further characterize the in vivo antiꢀtumor activity of 2cPE we evaluated the cellular
responses in normal mouse tissues (liver, kidney and spleen) and in the tumor. Immunoblot
analysis of tissue extracts confirms also in vivo the upꢀregulation of p53 and Noxa following
2
cPE treatment (Figure 8B). This upꢀregulation was much more evident in the tumor compared
to normal tissue. To exclude speciesꢀspecific enrichment of the antibody against p53, we used, in
addition to DOꢀ1 antibody, the 421 antibody, which recognizes with similar affinity the human
and murine proteins.
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 70
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
PBS
A0.9
2cPE s.c.
0
0
0
.8
.7
.6
2cPE i.v.
0
0
.5
.4
0
0
.3
.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
*
**
0
.1
0
*
**
*
*
*
*
*
*
*
*
*
0
4
8
12
16 20
24 28
32 36 40
44 48
Days from first treatment
Injections
Liver Spleen Kidney Tumor
B
ꢀ
+
ꢀ
+
ꢀ
+
ꢀ
+
2cPE
p53 (DOꢀ1)
p53 (421)
Noxa
Actin
Figure 8. (A) Variations of A549 lung carcinoma xenografts volume in mice following
subcutaneous (s.c.) and intravenous (i.v.) treatments with 2cPE (170 mg/kg). (B) Immunoblot
analysis showing upꢀregulation of p53 and Noxa proteins in tissues of mice treated with 2cPE.
3
Mice in which the tumor mass reached the size of at least 0.5 cm were treated for the first time
with 170mg/kg of 2cPE and 48 hours later they were sacrificed for the generation of tissue
extracts. p53 was detected with DOꢀ1 and 421 antibodies, showing no significant difference.
ACS Paragon Plus Environment
26

Page 27 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
CONCLUSIONS
Several studies have established that compounds characterized by the presence of the 1,5ꢀ
diarylꢀ3ꢀoxoꢀ1,4ꢀpentadienyl pharmacophore exhibit cytotoxic activities in vitro against multiple
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
29
cancer cell lines. While different cellular thiols may in principle react with these Michael
acceptors, isopeptidases, a heterogeneous family of cysteineꢀproteases, are important targets for
their antiꢀneoplastic activity. These inhibitors, also called NꢀSIIs, can trigger both necrosis and
24,33,34
apoptosis, depending on the cellular context.
In this work we have shown that the
cytotoxicity of a family of dienones structurally related to sulfone G5 is directly correlated with
the electrophilicity of the dienone system. This provides indirect evidence that the biological
activity of these compounds is due to their ability to act as mechanismꢀbased inhibitors of
cysteine proteases. We have also demonstrated that by varying the substituents on the aromatic
rings and the structure of the cyclic scaffold it is possible to limit necrosis and to favor cell death
by apoptosis. This has led to the identification of compound 2c, based on a 4ꢀ
hydroxycyclohexanone scaffold, as a good candidate for further development.
Analysis of the inhibitor activities has evidenced that G5 is in general a more potent inhibitor
of DUBs such as UCHL1, UCHL5 and USP2 compared to 2c. By contrast, 2c functions, to some
extent, as a better inhibitor for the deISGylase USP18. In this scenario the dichotomy
apoptosis/necrosis might reflect different patterns of cysteineꢀprotease inhibition. Certainly, in
view of the existence of additional cellular targets, alternative hypothesis are possible and further
studies will be necessary to clarify this point.
It has been suggested that similar NꢀSIIs, although characterized by a broad spectrum of
1
5ꢀ17,35
activity, can nevertheless exhibit preference for certain targets.
In this view, the lower IC50
for proteasomeꢀassociated deubiquitinase UCHL5 compared to other USPs, displayed by both
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 70
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
13,14,16
G5 and 2c, can explain their action as UPS inhibitors,
even if additional mechanisms must
23
also be taken into account.
In this study we have also designed and synthesized the first NꢀSII proꢀdrug (2cPE), by
conjugating compound 2c with PEG, through a cleavable diester linker. This strategy has
overcome the solubility issue and has provided effective antiꢀtumoral activity in vivo, after
administration of the proꢀdrug by intravenous injections. From the diverse response in vitro of
different cell lines to the proꢀdrug, and by applying bioinformatics analysis and structural
studies, we have identified in PLA2G7/PAFꢀAH a key esterase involved in proꢀdrug maturation.
This esterase has been proposed as an activating factor for anticancer proꢀdrugs and delivery
systems due to its overexpression in breast, stomach, colorectal, pancreatic, prostate and liver
cancers. Anticancer lipids have been included in liposomes to be hydrolysed by phospholipase
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
51
A2, and designed phospholipids containing PEG350 or PEG2000 chains linked to
5
2
phosphatidylethanolamine by a carbamate linkage have proven to be substrates for the enzyme.
PLA2G7 is primarily produced by macrophages and circulates in plasma in active form as a
53
complex with LDL and HDL. This pattern of expression explains why A549 cells, which are
unresponsive to 2cPE in vitro, become fully responsive when evaluated in vivo.
The antineoplastic activity of 2cPE in vivo was accompanied by the upꢀregulation of p53 and
of the proꢀapoptotic protein Noxa. This upꢀregulation was more marked in tumor cells compared
to liver, spleen and kidney. Importantly, although proꢀapoptotic genes were upꢀregulated also in
normal tissues at some extents, the animals did not show any effect of toxicity. Both the reduced
upꢀregulation of the apoptotic program, in normal cells and their higher intrinsic resistance to
stress, can explain the absence of evident toxicity.
ACS Paragon Plus Environment
28

Page 29 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
In summary, isopeptidase inhibitors represent interesting tools in antiꢀneoplastic therapy.
Understanding mechanisms of actions, improving the delivery, for instance by the generation of
proꢀdrugs through a cleavable PEGylation as described here for the first time, and further
medicinal chemistry approaches to improve activity are efforts that need pursuing to provide
additional antiꢀcancer therapeutics.
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
All biologically evaluated compounds had a purity > 95% as determined by HPLC analyses
(see supporting information).
Chemistry: general protocols. Melting points are uncorrected and are given in Celsius
1
13
degrees. H NMR and C NMR spectra were recorded in CDCl , unless otherwise stated, on Jeol
3
EX400 (400 MHz) and Varian X500 (500 MHz) spectrometers; chemical shifts are given in ppm
relative to tetramethylsilane. IR spectra were obtained as Nujol mulls with a ThermoꢀNicolet
AVATAR 320 FTꢀinstrument. Electron impact mass spectra (MS) were obtained on a Varian
Saturn 2200 spectrometer equipped with a direct insertion probe. Electrospray mass spectra
(ESIꢀMS) were obtained with a Bruker Daltonics Esquire 4000 spectrometer. Flash
chromatography was performed on silica gel 60 (Merck, 230−400 mesh). 4HꢀThiopyranꢀ4ꢀoneꢀ
25
26
27
27
30
1
,1ꢀdioxide, and compounds 1a, 2a, 2b, 2e were synthesized according to the literature.
Poly(ethylene glycol) methyl ether (SigmaꢀAldrich) and all the PEGylated derivatives were coꢀ
evaporated twice with dry dichloromethane and dried under vacuum immediately prior to use.
(3Z)-3-[(4-nitrophenyl)methylene]-4H-tetrahydrothiapyran-4-one (5). A suspension of
tetrahydrothiapyranꢀ4ꢀone (23.2 g, 0.2 mol) and 4ꢀnitrobenzaldehyde (15.1 g, 0.1 mol) in 100
mL water containing MgSO ·7H O (14.8 g, 0.06 mol) and NaOH (5 g, 0.125 mol) was stirred at
4
2
ACS Paragon Plus Environment
29