Dimerized linear mimics of a natural cyclopeptide (TMC-95A) are potent noncovalent inhibitors of the eukaryotic 20S proteasome

Article abstract of DOI:10.1021/jm4002007

Noncovalent proteasome inhibitors introduce an alternative mechanism of inhibition to that of covalent inhibitors used in cancer therapy. Starting from a noncovalent linear mimic of TMC-95A, a series of dimerized inhibitors using polyaminohexanoic acid spacers has been designed and optimized to target simultaneously two of the six active sites of the eukaryotic 20S proteasome. The homodimerized compounds actively inhibited chymotrypsin-like (K_i = 6-11 nM) and trypsin-like activities, whereas postacid activity was poorly modified. The noncovalent binding mode was ascertained by X-ray crystallography of the inhibitors complexed with the yeast 20S proteasome. The inhibition of proteasomal activities in human cells was evaluated. The use of the multivalency inhibitor concept has produced highly efficient and selective noncovalent compounds (no inhibition of calpain and cathepsin) that have potential therapeutic advantages compared to covalent binders such as bortezomib and carfilzomib.

Full text of DOI:10.1021/jm4002007

Subscriber access provided by BALL STATE UNIV
Article
Dimerized Linear Mimics of a Natural Cyclopeptide (TMC-95A) Are
Potent Noncovalent Inhibitors of the Eukaryotic 20S Proteasome
Audrey Desvergne, Emilie Genin, Xavier Marechal, Nerea Gallastegui, Laure
Dufau, Nicolas Richy, Michael Groll, Joelle Vidal, and Michèle REBOUD-RAVAUX
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/jm4002007 • Publication Date (Web): 29 Mar 2013
Downloaded from http://pubs.acs.org on April 16, 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 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Dimerized Linear Mimics of a Natural Cyclopeptide
(TMC-95A) Are Potent Noncovalent Inhibitors of
the Eukaryotic 20S Proteasome
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
Audrey Desvergne†, Emilie Genin‡, Xavier Maréchal†, Nerea Gallastegui§, Laure Dufau†,
Nicolas Richy‡, Michael Groll§, Joëlle Vidal‡,*, Michèle Reboud-Ravaux†,*
†Enzymologie Moléculaire et Fonctionnelle, UR4, University Paris 6-Pierre et Marie Curie,
UPMC-Sorbonne Universités, case 256, 7 quai Saint Bernard, 75252 Paris Cedex 05, France
‡Université de Rennes 1, Chimie et photonique moléculaires, CNRS-UMR 6510, Bâtiment 10A,
Campus de Beaulieu, 35042 Rennes Cedex, France
§Center for Integrated Protein Science at the Department Chemie Lehrstuhl für Biochemie,
Technische Universität München, Lichetenbergstr. 4, 85747 Garching, Germany
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 45
1
2
3
4
5
ABSTRACT
6
7
8
9
Noncovalent proteasome inhibitors introduce an alternative mechanism of inhibition to that of
covalent inhibitors used in cancer therapy. Starting from a noncovalent linear mimic of TMC-
95A, a series of dimerized inhibitors using polyaminohexanoic acid spacers has been designed
and optimized to target simultaneously two of the six active sites of the eukaryotic 20S
proteasome. The homodimerized compounds actively inhibited chymotrypsin-like (Ki = 6-11 nM)
and trypsin-like activities whereas post-acidic activity was poorly modified. The noncovalent
binding mode was ascertained by X-ray crystallography of the inhibitors complexed with the
yeast 20S proteasome. The inhibition of proteasomal activities in human cells was evaluated. The
use of the multivalency inhibitor concept has produced highly efficient and selective noncovalent
compounds (no inhibition of calpain and cathepsin) that have potential therapeutic advantages
compared to covalently binding bortezomib and carfilzomib.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
ACS Paragon Plus Environment

Page 3 of 45
Journal of Medicinal Chemistry
1
2
3
4
INTRODUCTION
5
6
7
8
The ubiquitin-proteasome system plays an essential role in nuclear and cytosolic protein
degradation in eukaryotic cells by controlling numerous essential cellular processes including
turnover and quality control of proteins, cell cycle and apoptosis, transcription and cell signaling,
immune response and inflammation.1,2 The degradation of proteins is mediated by the
constitutive 26S multicatalytic proteasome that consists of a 20S catalytic core and two 19S
regulatory complexes. The catalysis of proteolysis is accomplished by the 20S proteasome
particle that has been the major target to identify and develop inhibitors as potential drugs able to
modulate proteasomal activity. The clinical value of proteasome inhibition has become evident
since the FDA approval of bortezomib for the treatment of relapsed multiple myeloma and
mantle cell lymphoma.3-5 This drug is a highly selective peptide boronate.6 Nevertheless, due in
particular to the presence of the boronic acid moiety, prolonged treatments induce toxicity and
severe adverse effects including neurological disorders.7-9 Resistances have also been reported.10
The main second-generation proteasome inhibitors that are in clinical trials are carfilzomib11
(approved in July 2012 for treatment of multiple myeloma) (PR-171), marizomib12
(salinosporamide or NPI-0052), ixazomib13 (MLN-9708), oprozomib14 (ONX-0912 or PR-047)
and delanzomib15 (CEP-18770). As with bortezomib, they are all covalent binding inhibitors
reacting with the catalytic threonine of the active sites to form a covalent adduct.16 Noncovalent
inhibitors, which are devoid of a reactive group, do not have the drawbacks generally associated
with the presence of a warhead, such as lack of specificity, instability and excessive reactivity,
though epoxyketones are the most specific applied proteasome inhibitors to date. Since toxicity
will always be an issue for proteasome inhibitors, because of the huge variety of functions
implicating this proteolytic machinery, noncovalent proteasome inhibitors may be an alternative
to covalent ones.4,17 This less documented category of inhibitors includes the natural cyclopeptide
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
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 45
1
2
3
4
5
6
7
8
9
1 (TMC-95A)18 (Figure 1A) and simpler cyclic19-21 and linear mimics22,23 obtained by rational
drug design. Several series of peptide and pseudopeptide noncovalent inhibitors were identified
by in vitro screenings of small or very large synthetic chemical libraries coupled with chemical
optimization.24-29 Noncovalent hydroxyureas have also been recently identified by in vitro
screening30 and in silico screening of a large collection of organic compounds led to several
structurally distinct noncovalent inhibitors.31 A growing number of studies suggest that
proteasome inhibitors may become valuable drugs for the treatment of non-tumorous diseases
when used in conditions in which the cellular functions are modulated without induction of cell
death.2,32 Diverse beneficial effects are expected for example in inflammation,33
neurodegenerative diseases,34 muscular dystrophies35,36 and cachexia.37 Moreover, they have
therapeutic potentialities as antiparasitics in malaria,38,39 sleeping syndrome,40-42 and anti-
microbial agents (tuberculosis).43 The selective inhibition of immunoproteasomes predominantly
expressed in immune cells appears to be a promising rationale for the treatment of autoimmune
disorders.32
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The constitutive 26S proteasome contains a barrel-shaped 20S catalytic particle composed of
28 subunits capped on either end by a regulatory particle called 19S. The complete subunit
architecture of the yeast regulatory 19S particle has been recently described providing insight into
recognition and processing of polyubiquitinated substrates.44 The architecture of constitutive
eukaryotic 20S proteasomes was reported several years ago45, 46 and recently compared to that of
immunoproteasome.47 The 28 subunits ( and  are arranged in four stacked rings (1-71-71-
71-7) with a 2-fold axial symmetry (Figure 1B).45 Three distinct types of catalytic activities are
assigned in constitutive proteasomes to two 1 subunits (caspase-like or post-acid activity, PA),
two 2 subunits (trypsin-like activity, T-L), and two 5 subunits (chymotrypsin-like activity,
4
ACS Paragon Plus Environment

Page 5 of 45
Journal of Medicinal Chemistry
1
2
3
4
ChT-L). Immunoproteasomes are generated by replacing 1, 2 and 5 subunits by their
5
6
7
immuno-homologues 1i, 2i, and 5i whose synthesis is induced by IFN-.48, 49
8
9
Targeting the ChT-L active site has long been considered as sufficient to develop drugs,
especially anti-cancer ones. Nevertheless, inhibition of multiple active sites is required to
markedly decrease protein degradation.50 Recent data suggest that for anti-cancer drugs it is also
important to target the PA in addition to the ChT-L active site.51, 52 Moreover, inhibitors of T-L
activity sensitize multiple myeloma cells to inhibitors of the ChT-L active site.53 Proteasome
inhibitors targeting the T-L activity also appear as the rational choice for anti-trypanosomal drug
development.54
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
Consequently, we developed new noncovalent proteasome inhibitors presenting within the
same molecule the possibility to inhibit several active sites. The proteasome catalytic particle
presents symmetry related multiple catalytic sites. Thus, chemically symmetrical drugs involving
identical inhibitory heads are of significance to provide enhanced inhibitory power. The
application of the principle of multivalency has been applied to design potent multivalent
inhibitors.55 Its application was successful in the protease field56 and reported for covalent
proteasome inhibitors57, 58 and noncovalent ones.59 In both cases, two ligands were chemically
connected by a PEG spacer to target two active sites in proteasomes. But polydispersity is one of
the drawbacks of pegylation,59 and access to monodisperse PEG is a challenge.60 Here, we
describe the design, synthesis and inhibitory efficacies of dimers 4 (Figure 1C) formed of two
inhibitory heads derived from the noncovalent proteasome inhibitors 2, linked by spacers of
controlled length to optimally target two selected active sites (Figure 1B). In order to mimic the
unfolded protein chain of proteasome substrates that enter the catalytic core, oligomers of amino
hexanoic acid (Ahx) and adipic acid derived from 5 or 6 (Figure 1D) were chosen as the spacers.
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 45
1
2
3
4
5
6
7
8
9
The different moieties constituting the bivalent molecules 4 such as inhibitors 2, conjugated
monomers 3, and poly Ahx derivatives 5 and 6, were evaluated independently. The crystal
structures of several inhibitors in complex with yeast 20S proteasome were obtained and the
mechanism of inhibition determined. Furthermore, the selectivity of action towards challenging
proteases such as calpain I and cathepsin B is examined.
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
RESULTS
Design and Synthesis of Inhibitors. The design of the bivalent inhibitors relied on the X-ray
structure of the yeast 20S proteasome from which distances between the catalytic Thr1O of the
different active sites have been determined (Figure 1B).45, 61 The two chosen identical inhibitory
heads were the tripeptidic sequence found in the noncovalent inhibitors 2a and 2b that are linear
mimics of the natural product 1 (Figure 1A, C). Their binding to the ChT-L active site of the
yeast 20S proteasome was previously characterized by crystallography as well as by their
inhibitory potency (IC50 of 1.5 M and 9.4 M for 2a and 2b, respectively).23 Compounds 2a and
2b displayed identical antiparallel  sheet binding to the S4-S3-S2-S1 proteasomal specificity
pocket of the ChT-L active site (Figure S1 in Supporting Information). The C-terminal benzyl
group was inserted into the S1 pocket. The reduced activity of the N-terminal deprotected peptide
2b compared to its Boc analog 2a was attributed to the disruption of the interaction between 6
His98 and the Boc protecting group. So as a bulky lipophilic R group is favorable, we chose to
replace the R substituent in 1 by derivatives of Ahx trimer 5a or tetramer 5b to generate the new
monovalent molecules 3a and 3b that could be considered as reference compounds to study the
influence of the poly Ahx on the activity of the inhibitory head (Figure 1C, D). Then we built
dimers 4 by linking two identical moieties 2b to oligomers of Ahx (p+q varying from 3 to 8) and
adipic acid, in order to simultaneously target two active sites. As the length of extended Ahx and
6
ACS Paragon Plus Environment

Page 7 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
adipoyl units can be evaluated to ca. 10 Å,62 the maximum distance between the two N-terminus
tripeptide nitrogens of dimers 4a-e may vary discontinuously from 40 to 90 Å (Figure 1C, Tables
S1 and S2 in Supporting Information). The lengths of the different spacers are in principle
sufficient to potentially bind simultaneously two distinct active sites (distance between two
catalytic Thr1O from 30 to 65 Å, Figure 1B, C). The diprotected spacers 5 and 6 were prepared
from aminohexanoic acid and methyl adipoyl chloride using stepwise conventional peptide
synthesis techniques (see scheme for their synthesis in Figure S3, Supporting Information).
Saponification of diprotected spacer 5 followed by coupling to tripeptide 2b23 afforded reference
monovalent compounds 3 in 34-55% overall yield (Scheme 1). Two ways were used to obtain
dimers 4. The first one started from diprotected spacer 6a-c and 2b23 and led to 4a-c in 22-44%
yield. Boc deprotection of monovalent compounds 3 and subsequent coupling to 0.5 equivalent of
adipoyl dichloride yielded 4d-e (25-28%). Target compounds 3 and 4 were characterized by
NMR, high resolution mass spectrometry and HPLC.
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 vitro inhibitory efficiency and mechanism. We estimated the capacity of compounds 2a-b,
3a-b, and 4a-e to inhibit the ChT-L activity of constitutive human 20S proteasome (fluorogenic
substrate Suc-LLVY-AMC), its PA (Z-LLE--NA), and its T-L activities (Boc-LRR-AMC) by
determining the corresponding IC50 values (Figure S2 in Supporting Information, Table 1).
Conversely, no inhibition was observed with poly Ahx derivatives 5d and 6a. The deprotection of
the N-terminal end of the tyrosine residue (R = H, 2b) compared to 2a (R = Boc) resulted in an
increase of the IC50 value by a factor of 3.6 (ChT-L activity) and 4.4 (PA activity) whereas the T-
L activity was abolished. The replacement of the Boc group in 2a by the longer chains Boc(Ahx)n
(n = 3 for 3a and 4 for b) increased noticeably the inhibitory potency towards the three types of
activity: the IC50 values for 3a were 7-, 2- and 8-times smaller than the IC50 values for the
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 45
1
2
3
4
5
6
7
8
9
reference 2a (ChT-L, PA and T-L activities, respectively). These values were more significantly
decreased for dimeric inhibitors 4a-e with factors of 123-167 (ChT-L activity), 19-34 (PA
activity), and 355-577 (T-L activity) (compared with 2a), respectively.
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
To elucidate the mechanism of inhibition of the ChT-L activity by dimeric inhibitors 4
compared to conjugated monomers 3 and starting molecules 2a-b, kinetic studies were conducted
at a fixed enzyme concentration and various substrate (Suc-LLVY-AMC) and inhibitor
concentrations. The tested compounds fall into distinct classes of inhibitors when analyzed by
Lineweaver-Burk plots (Figure 2): mixed (2a, 3a-b), non competitive (2b) and competitive (4a-
e). For mixed and non competitive inhibitors, the inhibitors bind to the free (E) and bound
enzyme (ES), respectively: E + I  EI (inhibitor constant Ki) and ES + I  ESI (inhibitor
constant K’i) with Ki = [E][I]/[EI] and K’i = [ES][I]/[ESI]. For competitive inhibitors, they bind
only to the free enzyme. N-terminal Boc (2a) instead of H (2b) increased binding into the ChT-L
active site (factor of 10 for Ki and 4.8 for K’i). In the presence of a long Boc(Ahx)n chain (n = 3
for 3a and n = 4 for 3b) instead of Boc (2a), the binding to the free enzyme was increased by a
factor of 24 and 42, respectively. The binding to the ES complex was also favored (factors of 16-
25). These results clearly highlighted the favorable role of the Boc(Ahx)n chain for inhibitor
efficacy, as observed with the natural products fellutamide B,63 glidobactin A64 and cepafungin
I65 which all harbor an aliphatic tail that improves the ligand's affinity to the proteasome.
Remarkably, the introduction of a second inhibitory head yielding compounds 4a-e led to
competitive inhibition (Figure 2). The dimeric molecules (4a-e) were significantly more potent
on the ChT-L activity than 2a and 2b with Ki = 6 nM for 4e and Ki = 10.3-11.4 nM for 4a-d
whereas Ki = 2,430 nM and K’i = 5,030 nM for 2a and Ki = K’i = 24,500 nM for 2b, respectively.
The ratio of free energies of interaction for the bivalent and monovalent binding (= Kibi/Kimono
)
8
ACS Paragon Plus Environment

Page 9 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
is large (4,083 for 4e compared to 2b), indicating that the binding of the bivalent ligand is
entropically favored.56 Interestingly, the Dixon graphs established for 2a and 3a appeared to be
non linear, but it was linear for the bivalent compound 4b. The shape of the plots obtained for 2a
and 3a suggested parabolic inhibition in which two molecules of inhibitor are bound per
molecule of enzyme.66 In contrast, only one molecule was susceptible to bind when two
inhibitory heads are covalently linked within the bivalent inhibitor 4b in agreement with linear
Dixon plots.
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
Structural basis of inhibition. The crystal structures of 20S yeast proteasome complexed with
3a, 4a and 4e were determined at 2.9, 3.1 and 2.8 Å with an Rfree = 22.7%, 22.3% and 24.1%,
respectively (Table 3S, Supporting Information). Well defined electron density maps were
obtained for all three compounds bound to the ChT-L active site; compound 3a was also seen in
the T-L active site. The tripeptide scaffold revealed low Debye-Waller factors indicating a high
occupancy. As depicted in Figure 3, the inhibitors bound to the ChT-L active site exhibited an
identical positioning of the tripeptide scaffold with an extended conformation and maintaining
each amino acid side chain in the same binding pockets as for the original 2a ligand (Figure S1 in
Supporting Information). However, as observed for the PEG linkers used in covalent dimerized
inhibitors,57 the poly Ahx linkers of these three compounds were only partially defined due to
their flexible nature.
The superimposition of 2a, 3a and 4e showed no concerted movement of any of the side chains
in the active site and the respective inhibitors. However, the complex structure of the yeast
proteasome with 4a displays a concerted movement of Try96 and His98, thus generating by its
spacious S4 pocket appropriate space for the linker chain.
These results prove that it is predominantly the bivalency of this set of inhibitors which gives
rise to the high binding preference, though all linkers are defined in the electron density map only
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 45
1
2
3
4
5
6
7
8
9
to the first nitrogen atom. Hence, the analysis of the crystal structure complexes demonstrates
multiple conformations for the linker, reiterating the flexibility of the spacer elements. Moreover,
a MES molecule was observed in all three structures occupying the oxyanion hole Gly47NH and
performing Van der Waals interactions with the inhibitors. Notably, there is 48 Å distance
between the two Tyr NH of 3a bound to the 5/5'subunits, 54 Å for 22', 47 Å for intra-ring
5/2, and 25 Å for inter-ring 5/2', thus allowing simultaneous binding of the bivalent
inhibitors to two different active sites, explaining the low IC50 values.
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
Effect of bivalent inhibitors in human cells. We treated HEK-293 cells with the various
compounds for 2 h and determined their proteasome inhibition profile (Figure 4, Table 2).
Monovalent 3a and bivalent molecules (4a-e) inhibited the ChT-L activity at submicromolar
concentrations (IC50 of 252-756 nM). Much higher concentrations of the parent molecule 2a were
required (IC50 = 7.97 µM). The PA activity was poorly inhibited (IC50 of about 20 µM). The
reference molecule epoxomicin appeared to be a stronger inhibitor of both ChT-L ad PA
activities but only by a factor of 18 (ChT-L activity) and 15 (PA activity) when exemplarily
compared to 4c.
Specificity and metabolic stability. All synthesized compounds were analyzed for their effect
on two representative cellular proteases, cytosolic calpain I and lysosomal cathepsin B. No
significant inhibition was observed at 20 µM. Furthermore, the stability of compounds 2a, 3a and
4a in the biological medium used in cellular studies was determined in order to evaluate their
susceptibility to enzymatic hydrolysis. They were incubated at 37°C in the culture medium
(RPMI) in the presence of 20% fetal calf serum for different periods of time. Kinetics were
followed by HPLC. The HPLC results showed that the peak area of compounds 2a and the
10
ACS Paragon Plus Environment

Page 11 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
bivalent inhibitor 4a did not change or deviate over time, indicating stability within 50 h
compared to the monovalent compound 3a (degradation half-time of 19 h).
7
8
9
DISCUSSION
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The principle of polyvalency is ubiquitously exploited in nature.55,56 It led medicinal chemists
to take advantage of multivalent binding to design new molecules with increased binding affinity
and selectivity. This alternative approach is particularly pertinent for the multicatalytic 20S
proteasome and has been used to generate homo- and heterobivalent proteasome covalent
binders.57,58 Homodimers of tripeptidic aldehydes linked by a polydisperse PEG spacer led to
efficient covalent inhibitors of ChT-L activity which potency was increased 123-fold.57 Here we
present homobivalent molecules formed by two noncovalent inhibitory heads linked by a
monodisperse poly Ahx spacer.
Inspection of the in vitro inhibition data in conjunction with the crystallographic data revealed
interesting features of the new bivalent inhibitors. Compared to the reference molecule 2a, the
dimeric inhibitors 4a-e displayed efficacies noticeably increased by factors of 133-167 (ChT-L
activity), 19-34 (PA activity) and 355-577 (T-L activity). Out of the substances prepared, five
show inhibitory activity against human proteasome ChT-L site comparable to the parent cyclic
molecule 1: Ki of 6.0-11.4 nM for 4a-e in the range of that obtained by 1 (Ki = 2.3 of nM).
Remarkably, the T-L activity was more efficiently blocked by the bivalent molecules than by 1
(factors of 10-16 between the IC50 values). Smaller differences were observed for the PA activity
(6-10). Importantly, the mechanism of inhibition of the ChT-L activity is shifted from mixed for
the starting linear mimics 2a and poly Ahx monomers 3a-b to competitive inhibition for the
bivalent compounds 4a-e. This highlights on the one hand the influence on the mechanism of the
substituent at the N-terminal end of the peptide, and on the other hand, the importance of the
presence of a second inhibitory head. The influence of a (Ahx)3 N-terminal extension on both
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 45
1
2
3
4
5
6
7
8
9
efficiency and selectivity of proteasome inhibitors was previously noticed for covalent peptide
vinyl sulfones.67 Finally, the reference starting molecule (2a), monovalent (3a) and bivalent (4a)
compounds showed a promising stability in cell culture medium with half-lives of over 60 h (2a
and 4a) and 19 h (3a), respectively.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The proteasome is a valid drug target for cancer but also for sleeping sickness.40,41 It was
demonstrated that the proteasome from Trypanosoma brucei is particularly sensitive to inhibition
of its T-L activity.54,68,69 Recent research has shown that T-L activities of trypanosome and
mammalian proteasomes may have distinct sensitivities for inhibitors.70 Nevertheless, it is
reasonable to assume that bivalent inhibitors targeting simultaneously T-L and ChT-L activities
of mammalian proteasomes may provide a new opportunity for the development of drugs to be
implemented in treating sleeping sickness disease.
CONCLUSION
Proteasome inhibitors have a vast therapeutic potential, but resistance and poor efficacy of
clinical covalent inhibitors on some solid tumors demonstrate the need for molecules with a new
mechanism of action. We pursue a noncovalent strategy with inhibitors theoretically free from
the inherent drawbacks associated to the presence of a reactive group by designing dimerized
linear mimics of 1. To optimally target two selected active sites, monodisperse poly Ahx spacers
were used to avoid the polydispersity of PEG spacers or detergent-like properties of compounds
with long alkyl chains. We have generated a new set of dimerized noncovalent inhibitors for
which blockage of the proteasome in vitro or in cellular assays is strongly enhanced. Their
mechanism of action is corroborated by crystallographic studies. Crossing simple chemistry and
taking advantage of the proteasome C2-symmetry lead to inhibitor activity similar to the reference
natural product 1. Remarkably the specificity of inhibition was significantly improved by the
synthesized bivalent ligands. Thus, the presented results open new possibilities for efficient and
12
ACS Paragon Plus Environment

Page 13 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
controlled inhibition of two types of active sites using a single probe. Synergistic effects to
decrease protein degradation may be induced favoring the inhibitor action in several pathologies.
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
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 45
1
2
3
4
EXPERIMENTAL SECTION
5
6
7
8
Chemistry. Commercially available reagents were used without further purification unless
otherwise stated. Dry solvents were distilled from the appropriate drying reagents immediately
before use. Yields refer to chromatographically and spectroscopically homogeneous materials,
unless otherwise stated. Reactions were monitored by thin-layer chromatography carried out on
silica gel aluminium sheets (60F-254) using UV light as a visualizing agent, 15% ethanolic
phosphomolybdic acid and heat or 0.2% ethanolic ninhydrin and heat as developing agent.
Column chromatography was performed on silica gel 60, 0.063-0.200 mm. MPLC (medium
pressure liquid chromatography) was performed on a Flashsmart two apparatus, using C18 silica
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
1
13
flash column (prepacked column, 4.3g C18 silica, AIT chromato, 40 -60 m). H NMR and C
NMR spectra were recorded at room temperature on a BRUCKER Avance 300 or Avance 500
spectrometers. Chemical shifts were reported in ppm ( units) and residual non deuterated solvent
was used as internal reference. High-resolution mass spectra were recorded on a Waters Q-TOF 2
or a Bruker MicrO-TOF Q II apparatus using an electrospray source (ESI). Melting points were
measured on a Stuart melting point apparatus. Microanalysis and mass spectra were performed by
the Centre régional des mesures physiques de l’Ouest (CRMPO, Rennes, France). Anhydrous
hydrogen chloride 3.3 M in anhydrous methanol was generated by reaction of acetyl chloride
(2.34 mL, 33 mmol) with anhydrous methanol (10 mL). Tripeptides 2a-b were previously
described.23 Aminohexanoic acid derivatives HCl, H-Ahx-OMe,71 Boc-Ahx-OH72 and Boc-Ahx-
OSu73 were prepared according known procedures. Scheme for the synthesis of compounds 3-6 is
given in Supporting Information (Figure S3). Purity of building blocks 5a-e and 6a-c was
checked by elemental analysis. Data for C, H, and N were within 0.4% of the theoretical values.
Purity of compounds 3-4, was determined by reverse phase HPLC using a C18 column (5 μm, 4.6
14
ACS Paragon Plus Environment

Page 15 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
mm X 250 mm, ACE), detection at 254 nm, a 1 mL per minute flow rate and solvent A (50:50
MeOH/water) for 2 min then a linear gradient from 100% of solvent to 100% of solvent B (85:15
MeOH/water) for 20 min then solvent B for 16 min. Purity was higher than 95% for all
compounds except for compounds 4c (88 %) and 4d (78%).
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
Monovalent inhibitor 3a: Tripeptide hydrochloride 2b (135 mg, 0.197 mmol) was
dissolved in water (4 mL) and treated with aqueous 4 M NaOH (1 mL). The aqueous solution
was extracted three times by ethyl acetate. The combined organic phases were dried over sodium
sulfate and concentrated in vacuo to afford basic tripeptide 2b (121,1 mg, 0.186 mmol). A
solution of Boc-(Ahx)3-OH (obtained from 0.154 mmol Boc-(Ahx)3-OMe 5a) in anhydrous DMF
(1 mL) was reacted under Ar overnight at room temperature with HOBt (26.10 mg, 0.17 mmol),
EDC, HCl (32.05 mg, 0.167 mmol), Et3N (0.026 mL, 0.184 mmol) and a solution of tripeptide 2b
(0.186 mmol) in anhydrous DMF (2 mL). The mixture was concentrated in vacuo and the
resulting solid was washed by water, 0.1 M hydrochloric acid and then water. After
chromatography through silica gel (13 g, eluent 5 % to 10 % MeOH in CH2Cl2) compound 3a
1
(90.84 mg, yield 55 %) was afforded as a white solid. Rf = 0.19 (10 % MeOH in CH2Cl2). H
NMR (300 MHz, DMSO-d6)  1.04-1.46 (m, 30 H), 2.00 (t, J = 7 Hz, 6H), 2.09 (dd, J = 14.5 Hz,
J = 9 Hz, 1H), 2.38 (m, 1H), 2.62 (m, 1H), 2.82-3.00 (m, 7H), 4.04-4.27 (m, 4H), 4.45 (m, 1H),
5.02 (s, 2H), 6.01 (s, 1H), 6.75 (d, J = 7.5 Hz, 2H), 6.87-6.95 (m, 3H), 7.08-7.43 (m, 18 H), 7.71
(m, 2H), 7.76 (d, J = 8 Hz, 1H), 7.91-7.98 (m, 2H), 8.08 (d, J = 6.5 Hz, 1H), 10.12 (s, 1H).
HPLC: Rt. 26.1 min, area percent 97% at 254 nm. HRMS (ESI) calcd for C60H80N8O11Na
[(M+Na)+] 1111.5844, found 1111.5841.
Monovalent inhibitor 3b: A solution of Boc-(Ahx)4-OH (115.3 mg, 0.202 mmol, obtained
from Boc-(Ahx)4-OMe 5b) in anhydrous DMF (2 mL) was reacted overnight under Ar and at
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 45
1
2
3
4
5
6
7
8
9
room temperature with HOBt (34 mg, 0.22 mmol), EDC, HCl (42 mg, 0.219 mmol), Et3N (0.034
mL, 0.24 mmol) and basic tripeptide 2b (0.26 mmol) in anhydrous DMF (2 mL). The mixture
was concentrated in vacuo and the resulting solid was washed by water, 0.1 M hydrochloric acid
then water. After medium pressure chromatography through C-18 reversed phase silica gel
(eluent 0.001 M aqueous HCl /MeOH 100/0 to 0/100) compound 3b (82.4 mg, yield 34 %) was
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
afforded as a white solid. Rf = 0.20 (10 % MeOH in CH2Cl2). H NMR (500 MHz, CD3OD) 
1.19-1.61 (m, 36H), 2.13-2.25 (m, 8H), 2.46 (dd, J = 9.6 Hz, J = 14 Hz, 1H), 2.59 (dd, J = 3.1
Hz, J = 14 Hz, 1H), 2.78 (dd, J = 9.5 Hz, J = 14 Hz, 1H), 3,00-3.33 (m, 9H), 4.16 (q, J = 7.3 Hz,
1H), 4.19 (dd, J = 2.8 Hz, J = 9.5 Hz, 1H), 4.26 and 4.39 (JAB = 15.0 Hz, 2H), 4.47 (dd, J = 5.2
Hz, J = 9.6 Hz, 1H), 5.06 (s, 2H), 6.85 (d, J = 7.0 Hz, 1H), 6.90 (d, J = 8.5 Hz, 2H), 7.04 (t, J =
13
8.0 Hz, 1H), 7.12 (d, J = 8.5 Hz, 2H), 7.18-7.42 (m, 12H). C NMR (125.77 MHz, CD3OD) 
17.2 (CH3), 23.4 (CH2), 26.4 (CH2), 26.6 (CH2), 26.7 (CH2), 26.8 (CH2), 27.4 (CH2), 27.5 (CH2),
27.6 (CH2), 28.8 (CH3), 30.0 (CH2), 30.1 (CH2), 30.7 (CH2), 36.6 (CH2), 37.0 (CH2), 37.8 (CH2),
37.4 (CH2), 39.7 (CH2), 40.2 (CH2), 40.3 (CH2), 44.1 (CH2), 51.3 (CH), 51.5 (CH), 56.6 (CH),
71,0 (CH2), 76.4 (C), 79.6 (C), 111.4 (CH), 115.9 (CH), 123.9 (CH), 125.5 (CH), 128.1 (CH),
128.4 (CH), 128.6 (CH), 128.8 (CH), 129.4 (CH), 129.5 (CH), 130.6 (C), 130.9 (CH), 131.3
(CH), 131.8 (C), 138.8 (C), 139.7 (C), 142.9 (C), 158.5 (C), 159.1(C), 172.2 (C), 173.5 (C),
174.8 (C),175.9 (C),176.0 (C),176.1 (C), 176.6 (C),181.5 (C). HPLC: Rt. 25.7 min, area percent
95% at 254 nm. HRMS (ESI) calcd for C66H91N9O12Na [(M+Na)+] 1224.6685, found 1224.6681.
Bivalent inhibitor 4a: (p+q = 3, p=0) A solution of MeO-Adip-Ahx3-OMe 6a (218.8 mg,
0.426 mmol) in THF (1.5 mL) was reacted for 3.5 h at room temperature with 1 M LiOH (1.5
mL, 1.5 mmol) and then treated by concentrated HCl. The white precipitate was filtered, washed
by water, and dried to afford HO-Adip-Ahx3-OH (154.7 mg, yield 75 %) which was used without
16
ACS Paragon Plus Environment

Page 17 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
further purification. A solution of HO-Adip-(Ahx)3-OH (10.89 mg, 0.0224 mmol) in anhydrous
DMF (0.4 mL) was reacted under Ar for 72 h at room temperature with HOBt (7.62 mg, 0.056
mmol), EDC, HCl (10.83 mg, 0.0565 mmol), Et3N (0.008 mL, 0.057 mmol) and basic tripeptide
2b (39.12 mg, 0.0602 mmol). The mixture was concentrated in vacuo and the resulting solid was
washed by water, 0.1 M hydrochloric acid and then water. After purification by medium pressure
chromatography through C-18 reversed phase silica gel (eluent 0.001 M aqueous HCl /MeOH,
gradient from 100/0 to 0/100) dimeric compound 4a (17.56 mg, yield 44 %) was afforded as a
white solid. Rf = 0.40 (15 % MeOH in CH2Cl2). 1H NMR (500 MHz, DMSO-d6)  1.08-1.35 (m,
24H), 1.46 (m, 4H), 2.01 (m, 10H), 2.11 (dd, J = 9 Hz, J = 14.5 Hz, 2H), 2.45 (m, 2H), 2.65 (m,
2H), 2.98 (m, 8H), 4.07-4.26 (m, 8H), 4.46 (m, 2H), 5.04 (s, 4H), 6.03 (s, 2H), 6.76 (d, J = 7 Hz,
2H), 6.89 (d, J = 8 Hz, 4H), 6.94 (t, J = 8 Hz, 2H), 7.14-7.44 (m, 28H), 7.68 (m, 3H), 7.76 (dd, J
= 7 Hz, J = 1.5 Hz, 2H), 7.93 (t, J = 6 Hz, 2H), 7.98 (m, 2H), 8.10 (m, 2H), 10.13 (s, 2H). HPLC:
Rt. 27.2 min, area percent > 99% at 254 nm. HRMS (ESI) calcd for C98H117N13O17Na [(M+Na)+]
1770.8588, found 1770.8586.
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
Bivalent inhibitor 4b: (p+q = 4, p=0) Similarly, a solution of HO-Adip-(Ahx)4-OH (12.33
mg, 0.0206 mmol, prepared from MeO-Adip-(Ahx)4-OMe 6b) in anhydrous DMF (0.5 mL) was
reacted under Ar for 72 h at room temperature with HOBt (7.74 mg, 0.051 mmol), EDC, HCl
(13.37 mg, 0.0697 mmol), Et3N (0.008 mL, 0.057 mmol) and basic tripeptide 2b (37.7 mg,
0.0581 mmol). The mixture was concentrated in vacuo and the resulting solid was washed by
water, 0.1 M hydrochloric acid and then water. After purification by medium pressure
chromatography through C-18 reversed phase silica gel (eluent 0.001 M aqueous HCl /MeOH,
gradient from 100/0 to 0/100) dimeric compound 4b (15.28 mg, yield 40 %) was afforded as a
1
white solid. H NMR (500 MHz, DMSO-d6)  1.08-1.36 (m, 28H), 1.46 (m, 6H), 2.01 (m, 12H),
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 45
1
2
3
4
5
6
7
8
9
2.11 (dd, J = 9 Hz, J = 14.5 Hz, 2H), 2.45 (dd, J = 4 Hz, J = 14.5 Hz, 2H), 2.65 (m, 2H), 3.00 (m,
10H), 4.07-4.26 (m, 8H), 4.46 (m, 2H), 5.04 (s, 4H), 6.04 (s, 2H), 6.77 (d, J = 7 Hz, 2H), 6.89 (d,
J = 8 Hz, 4H), 6.94 (t, J = 7.5 Hz, 2H), 7.14-7.44 (m, 28H), 7.68-7.72 (m, 4H), 7.76 (dd, J = 9
Hz, J = 4 Hz, 2H), 7.94 (t, J = 6 Hz, 2H), 7.99 (m, 2H), 8.10 (m, 2H), 10.13 (s, 2H). HPLC: Rt.
27.1 min, area percent 95% at 254 nm. HRMS (ESI) calcd for C104H128N14O18Na [(M+Na)+]
1883.9429, found 1883.9453.
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
Bivalent inhibitor 4c: (p+q = 5, p=0) Similarly, a solution of HO-Adip-(Ahx)5-OH (14.56
mg, 0.0205 mmol, prepared from MeO-Adip-(Ahx)5-OMe 5c) in anhydrous DMF (0.5 mL) was
reacted under Ar for 72 h at room temperature with HOBt (7.29 mg, 0.047 mmol), EDC, HCl
(20.33 mg, 0.106 mmol), Et3N (0.008 mL, 0.057 mmol) and basic tripeptide 2b (37.7 mg, 0.0581
mmol). The mixture was concentrated in vacuo and the resulting solid was washed by water, 0.1
M hydrochloric acid and then water. After purification by medium pressure chromatography
through C-18 reversed phase silica gel (eluent 0.001 M aqueous HCl /MeOH, gradient from
1
100/0 to 0/100) dimeric compound 4c (9.02 mg, yield 22 %) was afforded as a white solid. H
NMR (500 MHz, DMSO-d6)  1.17-1.36 (m, 32H), 1.46 (m, 8H), 2.02 (m, 14H), 2.12 (dd, J =
8.5 Hz, J = 14 Hz, 2H), 2.42 (m, 2H), 2.62 (m, 2H), 3.00 (m, 12H), 4.07-4.26 (m, 8H), 4.46 (m,
2H), 5.04 (s, 4H), 6.04 (br s, 2H), 6.77 (d, J = 7.5 Hz, 2H), 6.89 (d, J = 7 Hz, 4H), 6.94 (t, J = 7.5
Hz, 2H), 7.14-7.44 (m, 28H), 7.71-7.72 (m, 5H), 7.76 (m, 2H), 7.94 (t, J = 6 Hz, 2H), 7.99 (m,
2H), 8.10 (m, 2H), 10.13 (s, 2H). HPLC: Rt. 26,8 min, area percent 88% at 254 nm. HRMS (ESI)
calcd for C110H139N15O19Na [(M+Na)+] 1997.0264, found 1997.0351.
Bivalent inhibitor 4d: (p+q = 6, p = 3) A solution of 3a (86.85 mg, 0.0809 mmol) in
MeOH (2 mL) at 0 °C was reacted for 6 min with 12 M anhydrous HCl in MeOH (1.8 mL, 25
mmol) and then concentrated in vacuo without heating. The resulting solid was dissolved in water
18
ACS Paragon Plus Environment

Page 19 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(3 mL) and reacted with aqueous 4 M NaOH (1 mL). Extraction by ethyl acetate quantitatively
yielded basic deprotected 3a. This product was dissolved in anhydrous DMF (0.5 mL) and was
reacted under Ar with Et3N (0.008 mL, 0.057 mmol) and adipoyl dichloride (3.2 L, 0.022
mmol) for 72 h at room temperature. The mixture was concentrated in vacuo and the resulting
solid was purified by medium pressure chromatography through C-18 reversed phase silica gel
(eluent 0.001 M aqueous HCl /MeOH, gradient from 100/0 to 0/100). Dimeric compound 4d
(12.9 mg, yield 28 %) was afforded as a white solid. 1H NMR (500 MHz, DMSO-d6)  1.09-1.36
(m, 36H), 1.44 (m, 10H), 2.02 (m, 16H), 2.11 (dd, J = 9 Hz, J = 14 Hz, 2H), 2.45 (m, 2H), 2.64
(m, 2H), 3.00 (m, 14H), 4.09-4.23 (m, 8H), 4.47 (m, 2H), 5.04 (s, 4H), 6.03 (s, 2H), 6.77 (d, J =
7.5 Hz, 2H), 6.89 (d, J = 7 Hz, 4H), 6.94 (t, J = 7.5 Hz, 2H), 7.14-7.44 (m, 28H), 7.71-7.72 (m,
6H), 7.77 (d, J = 7.5 Hz, 2H), 7.94 (t, J = 6 Hz, 2H), 7.98 (d, J = 8 Hz, 2H), 8.10 (d, J = 7 Hz,
2H), 10.13 (s, 2H). HPLC: Rt. 26,3 min, area percent 78% at 254 nm. HRMS (ESI) calcd for
C116H150N16O20Na [(M+Na)+] 2110.1100, found 2110.1130.
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
Bivalent inhibitor 4e: (p+q = 8, p = 4) A solution of 3b (66 mg, 0.055 mmol) in MeOH (2
mL) at 0 °C was reacted for 6 min with 12 M anhydrous HCl in MeOH (0.26 mL, 3.1 mmol) and
then concentrated in vacuo without heating. The resulting solid was dissolved in anhydrous DMF
(0.34 mL) and was reacted under Ar with Et3N (0.052 mL, 0.358 mmol) and adipoyl dichloride
(5.0 L, 0.033 mmol) for 72 h at room temperature. The mixture was concentrated in vacuo and
the resulting solid was purified by medium pressure chromatography through C-18 reversed
phase silica gel (eluent 0.001 M aqueous HCl /MeOH, gradient from 100/0 to 0/100) followed by
a second purification (eluent water/MeOH, gradient from 50/50 to 0/100, fraction analysis by
HPLC). Dimeric compound 4e (16 mg, yield 25 %) was afforded as a white solid after
lyophilisation. 1H NMR (500 MHz, CD3OD)  1.19-1.65 (m, 58H), 2.19-2.25 (m, 20H), 2.48 (dd,
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 45
1
2
3
4
5
6
7
8
9
J = 9.5 Hz, J = 14 Hz, 2H), 2.61 (dd, J = 3.1 Hz, J = 14 Hz, 2H), 2.80 (dd, J = 9.5 Hz, J = 14 Hz,
2H), 3.09 (dd, J = 5 Hz, J = 14 Hz, 2H), 3,14-3.21 (m, 16H), 4.16 (q, J = 7.5 Hz, 2H), 4.21 (dd, J
= 3 Hz, J = 9.5 Hz, 2H), 4.28 and 4.41 (JAB = 15.5 Hz, 4H), 4.48 (dd, J = 5.5 Hz, J = 9.5 Hz, 2H),
5.06 (s, 4H), 6.87 (d, J = 7.5 Hz, 2H), 6.92 (d, J = 8.5 Hz, 4H), 7.06 (t, J = 8.0 Hz, 2H), 7.14 (d, J
= 8.5 Hz, 4H), 7.19-7.44 (m, 24H). 13C NMR (125.77 MHz, CD3OD)  17.2 (CH3), 26.3 (CH2),
26.4 (CH2), 26.6 (CH2), 26.7 (CH2), 27.4 (CH2), 27.5 (CH2), 27.6 (CH2), 30.0 (CH2), 36,5 (CH2),
36.6 (CH2), 36.7 (CH2), 37.4 (CH2), 39,7 (CH2), 40,3 (CH2), 40,4 (CH2), 40.5 (CH2), 44.1 (CH2),
51.4 (CH), 51.6 (CH), 56.7 (CH), 71,0 (CH2), 76.4 (C), 111.4 (CH), 115.9 (CH), 123.9 (CH),
125.5 (CH), 128.1 (CH), 128.4 (CH), 128.6 (CH), 128.8 (CH), 129.4 (CH), 129.5 (CH), 130.6
(C), 130.9 (CH), 131.3 (CH), 131.8 (C), 138.8 (C), 139.7 (C), 142.9 (C), 159.1(C), 173.2 (C),
174.5 (C), 174.8 (C),176;0 (C),176.3 (C),176.4 (C), 176.6 (C),181.5 (C). HPLC: Rt. 27.4 min,
area percent 96% at 254 nm. HRMS (ESI) calcd for C128H172N18O22Na [(M+Na)+] 2336.2791,
found 2336.2758.
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
Enzyme activity and inhibition assays using purified enzymes. Purified human constitutive
20S proteasome from erythrocyte was obtained from Boston Biochem (Euromedex, France). All
substrates (Suc-LLVY-AMC, Boc-LRR-AMC and Z-LLE-BNA) were obtained from Bachem
(Weil am Rhein, Germany).
Proteasome activities were determined in vitro by monitoring the hydrolysis of the 5 (ChT-L)
substrate Suc-LLVY-AMC, 1 (PA) substrate Z-LLE-NA, and 2 (T-L) substrate Boc-LRR-
AMC for 45 min 37°C in the presence of untreated proteasome (control) or proteasome that had
been previously incubated during 15 min in the presence of the applied compound at the chosen
concentration. The excitation wavelengths were 360 nm (AMC substrates) and 340 nm (NA
substrate) whereas the emission wavelengths were 460 nm (AMC substrates) and 405 nm (NA
20
ACS Paragon Plus Environment

Page 21 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
substrate), respectively. Substrate and compounds were previously dissolved in DMSO. The final
DMSO concentration was kept constant at 2% (v/v) in all assays. Buffers were composed of 20
mM Tris (pH 7.5), 10% (v/v) glycerol, 0.01% (w/v) SDS (ChT-L and PA activities); 20 mM Tris
(pH 8.0), 10% (v/v) glycerol, 0.01% (w/v) for T-L activity. The final enzyme concentrations were
0.3 nM (20S proteasome) using 20 µM Suc-LLVY-AMC (ChT-L), 50 µM Z-LLE-NA (PA) and
20 µM Boc-LRR-AMC (T-L). Initial rates determined in control experiments (V0) were
considered to be 100% of the peptidase activity while initial rates below 100% were considered
to be inhibitions. The values of IC50 (compound concentrations resulting in 50% enzyme
inhibition) were calculated by fitting the experimental data to the equation 1 or equation 2.
100  IC50
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
%inhibition 100  (1 Vi /V0) 
(eq. 1)
IC50  [I]0
nH
100  [I]0
%inhibition 
(eq. 2)
nH
nH
IC50 [I]0
Reaction rates determined at fixed enzyme concentration and varying substrate and inhibitor
concentrations were fitted by non-linear regression to the following Michaelis-Menten-type
equations; equation 3 for competitive and equation 4 for non competitive (Ki = K’i) and mixed
inhibitions (Ki  K’i):
Vmax [S]0
[S]0  Km (1  [I]/Ki)
Vi 
(eq. 3)
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 45
1
2
3
4
5
6
7
8
Vmax [S]0
(1  [I]/K'i )
Vi 
(eq. 4)
Km (1  [I]/Ki)
(1  [I]/K'i )
[S]0 
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
with Km, Michaelis constant; Vmax, maximum initial rate; Ki, dissociation constant of EI; Ki’,
dissociation constant of ESI. Lineweaver-Burk (1/Vi = f(1/[S]0) and Eadie-Hostee (Vi = f(Vi/[S]0)
representations were also used in routine to conveniently visualize the mechanism of proteasomal
blockage and determine the corresponding kinetic constants Ki or K’i. Dixon graphs (V0/Vi =
f([I]) were also constructed.
The activity of human calpain I in the absence and in the presence of a tested compound was
determined by monitoring the hydrolysis of the fluorogenic substrate Suc-LLVY-AMC (50 µM)
for 30 min in 50 mM Tris HCl, 2 mM CaCl2, 10 mM DTT ([E]0 = 81 nM ; (exc = 360 nm and
em = 460 nm ; pH 7.2 ; 25°C). Human cathepsin B activity ([E]0 = 0.55 nM) was received by
monitoring the hydrolysis of the fluorogenic substrate Z-RR-AMC (50 µM ; exc = 360 nm, em
=
460 nm) for 30 min at 37°C in the presence of untreated proteasome (control) or proteasome that
had been incubated for 5 min with the appropriate compound. The buffer was: 352 mM KH2PO4,
48 mM Na2HPO4, 1 mM EDTA, 1 mM DTT (pH 6.0). Substrate was previously dissolved in
water and all compounds in DMSO. Calpain I (or cathepsin B) and the respective inhibitor were
incubated for 5 min (or 15 min) before the measurement of the remaining enzyme activity. The
IC50 values were obtained by adjusting the experimental points to eq 1 or 2.
Cell-based Proteasome-GloTM β1, β2, and β5 assay. The effects of the compounds on ChT-
L, PA and T-L activities of HEK-293 cells were determined using a chemiluminescent assay
(Proteasome Glo Cell-Based Assay, Promega). HEK-293 cells were dispensed into white opaque
384-well microtiter plates (2,500 cells/well, 50 µL culture medium per well). 0.5 µL of each
22
ACS Paragon Plus Environment

Page 23 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
compound previously dissolved in DMSO was added. An equivalent volume of DMSO was used
in controls for a final DMSO concentration of 1% (v/v). After an incubation period of 2 h, 25 µL
assay buffer containing a recombinant firefly luciferase and the luminogenic proteasome
substrate Suc-LLVY-GloTM (ChT-L activity), Z-nLPnLD-GloTM (PA activity) or Z-LRR-GloTM
(T-L activity) was added. After a further incubation at room temperature for 10 min,
chemiluninescence expressed as RLU (Relative Light Units), was measured on a BMG Labtech
Fluostar optima plate reader with a measuring time of 12 s for each well. The final concentrations
of tested compounds varied between 2 - 20 µM.
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
Remaining ChT-L, PA or T-L activities were determined according to equation 5:
n



n 


Activityi
Activity
100 [I]0
%Activity 100  1
100 

(eq. 5)


IC n  [I]0

0 

50
Stability. Stock solutions of compounds dissolved in DMSO were prepared (5 mM) and
diluted in the biological medium RPMI containing 20% fetal serum. The final concentration of
each sample was 50 µM with 1% DMSO. These samples were left in an incubator at 37°C for 47
h. Aliquots were withdrawn at various times (0, 17, 23, 41 and 47 h); proteins were precipitated
by ethanol (final concentration 20%). After centrifugation (10,000 rpm, 10 min), the supernatant
was analyzed by HPLC (C18-reverse phase column, 5 μm, 4.6 mm x 250 mm, isochratic
60/40/0.1 CH3CN/H2O/TFA, 1 mL per min, detection at 254 nm) to determine the amount of the
respective compound present in each aliquot at 0, 17, 23, 41 and 47 hours. Finally, peak areas
were plotted against the time.
Crystallization and structure determination. Crystals of the 20S proteasome from S.
Cerevisiae were grown in hanging drops at 24 °C as previously described74,75 and incubated for at
least 48 h with their respective 1 derivatives. 40 mg/mL protein concentration was used for
crystallization in Tris-HCl (10 mM, pH 7.5) and EDTA (1 mM). The drops contained 3 µL of
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 45
1
2
3
4
5
6
7
8
9
protein and 2 µL of the reservoir solution, consisting of 30 mM of Magnesium Acetate (MgAc2),
100 mM of morpholino-ethane-sulphonic acid (MES) (pH 6.8) and 10% of 2-methyl-2,4-
pentandiol (MPD). Crystals were soaked in a cryoprotecting buffer (30% MPD, 20 mM MgAc2,
100 mM MES pH 6.8), frozen in a stream of liquid nitrogen gas at 100 K (Oxford Cryo Systems)
for data collection. The space group of CP: 1 derivative complexes is P21 with unit cell
parameters of approximately a=136 Å, b=301 Å, c= 145 Å and b=113 (Table S3 in Supporting
Information). Datasets were collected using synchrotron radiation at the X06SA-beamline in
SLS/Villingen/Switzerland at =1.0 Å. X-ray intensities and data reduction were evaluated by
using the program package XDS.76 The anisotropy of diffraction was corrected by comparing
observed and calculated structure amplitudes by the program CNS (including a bulk solvent
correction).77 Electron density maps were improved by averaging and back transforming the
reflections 10 times over the two fold non-crystallographic symmetry axis using the program
package MAIN.78 Conventional crystallographic rigid body and positional refinements were
carried out with CNS using coordinates of the yeast 20S proteasome structure (RCSB; 1RYP) as
a starting model.45 Model building including the distinct ligands was performed with the program
COOT and final TLS refinement was carried out in REFMAC5. Apart from the bound inhibitor
molecules, structural changes were only noted in the specificity pockets. Temperature factor
refinement indicates full occupancies of all inhibitor binding sites. All models were completed
with final Rfree values below 22.7% and root-mean-square deviation (rmsd) bond and angle
values of lower than 0.005 Å and 0.9°, respectively (Table S3).
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
ASSOCIATED CONTENT
Supporting Information Available: Supplementary figures S1-S2, supplementary tables S1-S4,
scheme for the synthesis of compounds 3-6, experimental procedures for the synthesis of
24
ACS Paragon Plus Environment

Page 25 of 45
Journal of Medicinal Chemistry
1
2
3
4
compounds 5a-d and 6a-c, copies of NMR spectra. This material is available free of charge via
5
6
the Internet at http://pubs.acs.org.
7
8
PDB ID Codes: yeast proteasome:3a (4JSU); yeast proteasome:4a (4JT0); yeast proteasome:4e
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
(4JSQ)
AUTHOR INFORMATION
Corresponding Authors Information
*Phone: + 33 2 23 23 57 33. Fax + 33 2 23 23 69 55. E-mail: joelle.vidal@univ-rennes1.fr
*Phone: +33 1 44 27 50 78. Fax: 33 1 44 27 51 40. E-mail: michele.reboud@upmc.fr
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
X. M. received financial support from the Ministère de l’Enseignement Supérieur et de la
Recherche. We thank Prof. Michel Morange (Ecole Normale Supérieure and UPMC, Paris) for
fruitful discussions. This study was financed by grant n° 14999 from the ‘Association Française
contre les Myopathies’ (AFM) and the German-Israeli Foundation for Scientific Research and
Development (GIF 23 1102/2010 to M.G.).
ABBREVIATIONS USED
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 45
1
2
3
4
Ahx, aminohexanoic; AMC, 7-amino-4-methylcoumarin; ChT-L, chymotrypsin-like; EDC, 1-
5
6
Ethyl-3-(3-dimethylaminopropyl)carbodiimide,
IFN-,
interferon
gamma;
HOBt,
7
8
9
hydroxybenzotriazole; MES, morpholinoethane sulfonate; NA, beta-naphthylamine; PA, post-
acid; Su, succinimydyl; Suc, succinyl; RPMI, Roswell Park Memorial Institute medium; T-L,
trypsin-like
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
REFERENCES
(1) Borissenko, L.; Groll, M., 20S proteasome and its inhibitors: Crystallographic knowledge
for drug development. Chem. Rev. 2007, 107, 687-717.
(2) Meiners, S.; Ludwig, A.; Stangl, V.; Stangl, K., Proteasome inhibitors: Poisons and
remedies. Med. Res. Rev. 2008, 28, 309-327.
(3) Richardson, P. G.; Mitsiades, C.; Hideshima, T.; Anderson, K. C., Bortezomib:
proteasome inhibition as an effective anticancer therapy. Annu. Rev. Med. 2006, 57, 33-47.
(4) Genin, E.; Reboud-Ravaux, M.; Vidal, J., Proteasome inhibitors: Recent advances and
new perspectives in medicinal chemistry. Curr. Top. Med. Chem. 2010, 10, 232-256.
(5) Adams, J.; Kauffman, M., Development of the proteasome inhibitor Velcade
(bortezomib). Cancer Invest. 2004, 22, 304-311.
(6) Groll, M.; Berkers, C. R.; Ploegh, H. L.; Ovaa, H., Crystal structure of the boronic acid-
based proteasome inhibitor bortezomib in complex with the yeast 20S proteasome. Structure
2006, 14, 451-456.
(7) Chauhan, D.; Hideshima, T.; Anderson, K. C., Proteasome inhibition in multiple
myeloma: Therapeutic implication. Annu. Rev. Pharmacol. Toxicol. 2005, 45, 465-476.
26
ACS Paragon Plus Environment

Page 27 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(8) Argyriou, A. A.; Iconomou, G.; Kalofonos, H. P., Bortezomib-induced peripheral
neuropathy in multiple myeloma: a comprehensive review of the literature. Blood 2008, 112,
1593-1599.
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
(9) Demo, S. D.; Kirk, C. J.; Aujay, M. A.; Buchholz, T. J.; Dajee, M.; Ho, M. N.; Jiang, J.;
Laidig, G. J.; Lewis, E. R.; Parlati, F.; Shenk, K. D.; Smyth, M. S.; Sun, C. M.; Vallone, M. K.;
Woo, T. M.; Molineaux, C. J.; Bennett, M. K., Antitumor activity of PR-171, a novel irreversible
inhibitor of the proteasome. Cancer Res. 2007, 67, 6383-6391.
(10) Naujokat, C.; Fuchs, D.; Berges, C., Adaptive modification and flexibility of the
proteasome system in response to proteasome inhibition. BBA Mol. Cell Res. 2007, 1773, 1389-
1397.
(11) O'Connor, O. A.; Stewart, A. K.; Vallone, M.; Molineaux, C. J.; Kunkel, L. A.;
Gerecitano, J. F.; Orlowski, R. Z., A phase 1 dose escalation study of the safety and
pharmacokinetics of the novel proteasome inhibitor carfilzomib (PR-171) in patients with
hematologic malignancies. Clin. Cancer Res. 2009, 15, 7085-7091.
(12) Potts, B. C.; Albitar, M. X.; Anderson, K. C.; Baritaki, S.; Berkers, C.; Bonavida, B.;
Chandra, J.; Chauhan, D.; Cusack, J. C.; Fenical, W.; Ghobrial, I. M.; Groll, M.; Jensen, P. R.;
Lam, K. S.; Lloyd, G. K.; McBride, W.; McConkey, D. J.; Miller, C. P.; Neuteboom, S. T. C.;
Oki, Y.; Ovaa, H.; Pajonk, F.; Richardson, P. G.; Roccaro, A. M.; Sloss, C. M.; Spear, M. A.;
Valashi, E.; Younes, A.; Palladino, M. A., Marizomib, a proteasome inhibitor for all seasons:
Preclinical profile and a framework for clinical trials. Curr. Cancer Drug Targets 2011, 11, 254-
284.
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 45
1
2
3
4
5
6
7
8
9
(13) Kupperman, E.; Lee, E. C.; Cao, Y.; Bannerman, B.; Fitzgerald, M.; Berger, A.; Yu, J.;
Yang, Y.; Hales, P.; Bruzzese, F.; Liu, J.; Blank, J.; Garcia, K.; Tsu, C.; Dick, L.; Fleming, P.;
Yu, L.; Manfredi, M.; Rolfe, M.; Bolen, J., Evaluation of the proteasome inhibitor MLN9708 in
preclinical models of human cancer. Cancer Res. 2010, 70, 1970-1980.
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
(14) Zhou, H.-J.; Aujay, M. A.; Bennett, M. K.; Dajee, M.; Demo, S. D.; Fang, Y.; Ho, M. N.;
Jiang, J.; Kirk, C. J.; Laidig, G. J.; Lewis, E. R.; Lu, Y.; Muchamuel, T.; Parlati, F.; Ring, E.;
Shenk, K. D.; Shields, J.; Shwonek, P. J.; Stanton, T.; Sun, C. M.; Sylvain, C.; Woo, T. M.;
Yang, J., Design and synthesis of an orally bioavailable and selective peptide epoxyketone
proteasome inhibitor (PR-047). J. Med. Chem. 2009, 52, 3028-3038.
(15) Piva, R.; Ruggeri, B.; Williams, M.; Costa, G.; Tamagno, I.; Ferrero, D.; Giai, V.; Coscia,
M.; Peola, S.; Massaia, M.; Pezzoni, G.; Allievi, C.; Pescalli, N.; Cassin, M.; di Giovine, S.;
Nicoli, P.; de Feudis, P.; Strepponi, I.; Roato, I.; Ferracini, R.; Bussolati, B.; Camussi, G.; Jones-
Bolin, S.; Hunter, K.; Zhao, H.; Neri, A.; Palumbo, A.; Berkers, C.; Ovaa, H.; Bernareggi, A.;
Inghirami, G., CEP-18770: a novel, orally active proteasome inhibitor with a tumor-selective
pharmacologic profile competitive with bortezomib. Blood 2008, 111, 2765-2775.
(16) Groll, M.; Huber, R.; Potts, B. C. M., Crystal structures of salinosporamide A (NPI-0052)
and B (NPI-0047) in complex with the 20S proteasome reveal important consequences of beta -
lactone ring opening and a mechanism for irreversible binding. J. Am. Chem. Soc. 2006, 128,
5136-5141.
(17) Gräwert, M. A.; Groll, M., Exploiting nature's rich source of proteasome inhibitors as
starting points in drug development. Chem. Commun. 2012, 48, 1364-1378.
28
ACS Paragon Plus Environment

Page 29 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(18) Koguchi, Y.; Kohno, J.; Nishio, M.; Takahashi, K.; Okuda, T.; Ohnuki, T.; Komatsubara,
S., TMC-95A, B, C, and D, novel proteasome inhibitors produced by Apiospora montagnei Sacc.
TC 1093. Taxonomy, production, isolation, and biological activities. J. Antibiot. 2000, 53, 105-
109.
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
(19) Groll, M.; Götz, M.; Kaiser, M.; Weyher, E.; Moroder, L., TMC-95-based inhibitor
design provides evidence for the catalytic versatility of the proteasome. Chem. Biol. 2006, 13,
607-614.
(20) Lin, S.; Yang, Z.-Q.; Kwok, B. H. B.; Koldobskiy, M.; Crews, C. M.; Danishefsky, S. J.,
Total synthesis of TMC-95A and -B via a new reaction leading to Z-enamides. Some preliminary
findings as to SAR. J. Am. Chem. Soc. 2004, 126, 6347-6355.
(21) Kaiser, M.; Groll, M.; Siciliano, C.; Assfalg-Machleidt, I.; Weyher, E.; Kohno, J.;
Milbradt, A. G.; Renner, C.; Huber, R.; Moroder, L., Binding mode of TMC-95A analogues to
eukaryotic 20S proteasome. ChemBioChem 2004, 5, 1256-1266.
(22) Basse, N.; Piguel, S.; Papapostolou, D.; Ferrier-Berthelot, A.; Richy, N.; Pagano, M.;
Sarthou, P.; Sobczak-Thépot, J.; Reboud-Ravaux, M.; Vidal, J., Linear TMC-95-based
proteasome inhibitors. J. Med. Chem. 2007, 50, 2842-2850.
(23) Groll, M.; Gallastegui, N.; Maréchal, X.; Le Ravalec, V.; Basse, N.; Richy, N.; Genin, E.;
Huber, R.; Moroder, L.; Vidal, J.; Reboud-Ravaux, M., 20S Proteasome inhibition: designing
noncovalent linear peptide mimics of the natural product TMC-95A. ChemMedChem 2010, 1701-
1705.
29
ACS Paragon Plus Environment