Identification of a β1/β2-Specific Sulfonamide Proteasome Ligand by Crystallographic Screening

Article abstract of DOI:10.1002/anie.201505054

The proteasome represents a validated drug target for the treatment of cancer, however, new types of inhibitors are required to tackle the development of resistant tumors. Current fluorescence-based screening methods suffer from low sensitivity and are limited to the detection of ligands with conventional binding profiles. In response to these drawbacks, a crystallographic screening procedure for the discovery of agents with a novel mode of action was utilized. The optimized workflow was applied to the screening of a focused set of compounds, resulting in the discovery of a β1/β2-specific sulfonamide derivative that noncovalently binds between subunits β1 and β2. The binding pocket displays significant differences in size and polarity between the immuno- and constitutive proteasome. The identified ligand thus provides valuable insights for the future structure-based design of subtype-specific proteasome inhibitors.

Full text of DOI:10.1002/anie.201505054

Angewandte
Chemie
International Edition: DOI: 10.1002/anie.201505054
German Edition: DOI: 10.1002/ange.201505054
Proteasome Inhibition
Identification of a b1/b2-Specific Sulfonamide Proteasome Ligand by
Crystallographic Screening
Philipp Beck, Mich›le Reboud-Ravaux, and Michael Groll*
Abstract: The proteasome represents a validated drug target
for the treatment of cancer, however, new types of inhibitors are
required to tackle the development of resistant tumors. Current
fluorescence-based screening methods suffer from low sensi-
tivity and are limited to the detection of ligands with conven-
tional binding profiles. In response to these drawbacks,
a crystallographic screening procedure for the discovery of
agents with a novel mode of action was utilized. The optimized
workflow was applied to the screening of a focused set of
compounds, resulting in the discovery of a b1/b2-specific
sulfonamide derivative that noncovalently binds between
subunits b1 and b2. The binding pocket displays significant
differences in size and polarity between the immuno- and
constitutive proteasome. The identified ligand thus provides
valuable insights for the future structure-based design of
subtype-specific proteasome inhibitors.
Recently, the discovery of N-hydroxyureas and alkaloids
as proteasome inhibitors broadened the range of potential
scaffolds.[12,13] In addition to the scientific literature, some
innovative scaffolds have been revealed in patents,[14–18] but
apart from these advances, the general structure of validated
CP inhibitors has historically been limited to peptidic
pharmacophores.[19,20] Concurrently, the development of new
proteasome blockers continues to be hampered by a lack of
structural knowledge from ligands that target novel binding
sites since common fluorescence-based screening methods
suffer from two major limitations. First, the assays are
susceptible to fluorescence quenching artifacts, thus resulting
in a high number of false-positive or false-negative results.
Second, the employed chromogenic substrates are unsuitable
for the detection of weakly binding fragments or ligands that
populate allosteric sites. In addition, screening with biophys-
ical methods like isothermal calorimetry, surface plasmon
resonance, or NMR are hampered by the large size and
demanding architecture of the CP.
The fight against oncological diseases requires a permanent
search for chemical agents that block cellular pathways that
are indispensable for the survival of cancer cells. The 20S
proteasome (core particle; CP) was identified as such a target
owing to its central role in protein quality control, protein
turnover, cell-cycle regulation, cell differentiation, and apop-
tosis. Indeed, FDA approval of the CP inhibitors bortezomib
(BTZ, Velcade) and carfilzomib (CFZ, Kyprolis) confirmed
that blocking proteasomal activity is beneficial for the
treatment of multiple myeloma and mantle cell lymphoma
(Figure S1 in the Supporting Information).[1,2] The two agents
share a similar peptidic scaffold with an electrophilic warhead
for covalent binding to the catalytic Thr1 residues of the
protease, and the development of resistance against both
agents highlights the need for non-peptidic inhibitors.[3–8] In
addition, the development of the next generation of protea-
some blockers has to focus on selectively addressing the
constitutive CP (cCP) and tissue-specific proteasomal sub-
types that play crucial roles in autoimmune or inflammatory
diseases (immunoproteasome, iCP) as well as T cell develop-
ment in the thymus (thymoproteasome, tCP).[9–11]
To address these drawbacks, we applied a crystallographic
screening procedure for the rapid evaluation of focused
compound libraries (Scheme 1). The purification and crystal-
lization of yeast CPs (yCP) is well documented and results in
an abundance of high-quality crystals for compound soak-
ing.[21] By using recent advances in synchrotron automation
and dataset evaluation software, we established an acceler-
ated workflow that includes crystal mounting, dataset collec-
tion, data reduction, refinement, and subsequent automated
ligand search in less than 10 min per dataset (Scheme 1).[22]
For the evaluation of our crystallographic screening
approach, we assembled a focused set of compounds that
were expected to show novel binding modes and furthermore
fulfilled the following criteria (Scheme S2 and Table S1 in the
Supporting Information):
a) Representation of a vast variety of chemotypes with
predominantly non-peptidic scaffolds.
b) Compounds that have previously been reported to act as
CP inhibitors in vitro or in cell culture studies.
c) Lack of structural data for the respective ligands bound to
the CP.
d) Simple chemical synthesis, commercial availability, or
availability through collaborators (Figure S2 in the Sup-
porting Information).
[*] Dr. P. Beck, Prof. Dr. M. Groll
Center for Integrated Protein Science Munich (CIPSM)
Department of Chemistry, Technische Universität München
Lichtenbergstraße 4, 85748 Garching (Germany)
E-mail: michael.groll@tum.de
Following our setup, yeast 20S proteasome crystals were
soaked with the screening compounds and diffraction data
were recorded to a maximum resolution of about 2.6 Š within
3 min. Rigid-body refinement using yCP as starting coordi-
nates (PDB ID 1RYP)[21] yielded Rfree values below 25%
within another 3 min. Subsequent automated searching of the
Prof. Dr. M. Reboud-Ravaux
Sorbonne UniversitØs, UPMC Univ Paris 06, UMR/CNRS 8256
Case 256, 7 Quai St-Bernard, 75252 Paris Cedex 05 (France)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201505054.
Angew. Chem. Int. Ed. 2015, 54, 11275 –11278
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
11275

.
Angewandte
Communications
Scheme 1. Workflow of the optimized crystallographic screening approach for identifying novel proteasome ligands.
resulting FoÀFc map visualized the presence of any ligand-
related positive electron density (< 1 min). Our method was
verified with the identification of the non-peptidic sulfona-
mide 1 as a hit, a compound that has been reported as
a putative CP inhibitor in an in silico screening study for novel
b5 inhibitors.[23] However, the ligand turned out to selectively
block the b1 and b2 activities, while b5 remained unaffected.
This unusual inhibition profile led to its incorporation into the
compound set and the application of 1 in our screening
workflow to give a yCP:1 cocrystal structure (Rfree = 21.4%,
PDB ID 5BOU).
Surprisingly, inspection of the crystallographic data
revealed that the non-primed substrate binding channels of
all of the catalytic sites lacked any ligand-related electron
density. Instead, 1 was located at the primed site of subunit b2,
which represents a hitherto unobserved binding pocket in the
proteasome (Figure 1). The ligand bridges the entire length of
the primed b2 substrate binding channel, forming extensive
hydrophilic and hydrophobic interactions. In particular, the
sulfonyl oxygen atoms of 1 are tightly coordinated by b2-
+
Thr1OgH, b2-Thr1NH3 , and b2-Gly47NH, thus disrupting
key residues that stabilize the tetrahedral transition state of
substrates during catalytic peptide-bond cleavage (Figure 1).
In addition, the 2,4-substituted quinoline ring system pro-
trudes deeply into a large cavity that is formed by residues of
both the b2 and b1 subunits. The binding of 1 is thereby
favored by van-der-Waals interactions with L127, Y97, and
H114 (all b2), as well as b1-Y25, while a single polar
interaction between b1-T22OgH and the quinoline nitrogen
atom provides additional stabilization.
Interestingly, structural superposition of yCP:1 and yCP in
Figure 1. Crystal structure of yCP in complex with 1. a) 2FOÀFC
complex with the tetrapeptidic inhibitor carfilzomib revealed
that the 4’-ethoxyphenyl moiety of 1 is in close contact with
b2-H116, which participates in the formation of the S4
substrate specificity pocket of subunit b1 (Figure 1 and
Figure 2).[24] These findings explain why 1 is able to inhibit
both the b1 and b2 catalytic sites (inhibitor concentration
giving 50% inhibition (IC50) ꢀ 100 and 50 mm, respectively),
while the chymotrypsin-like activity remains unaffected.[23]
First, the sulfonyl moiety of the ligand forms hydrogen bonds
with b2-Thr1 and thus impairs peptide bond hydrolysis at the
b2 subunit. Second, the biaryl system of 1 stretches into the
substrate binding channel of the b1 subunit, thereby affecting
substrate turnover. Notably, the reported high micromolar
IC50 values for 1 are only a rough indication of the binding
strength since the applied peptidic substrates do not compete
against ligand binding in the primed site. As a result,
electron density map (blue mesh, 1s, PDB ID 5BOU) of 1 (yellow)
located at the primed site of the b2 substrate binding channel (gray).
The sulfonyl group of 1 forms a network of hydrogen bonds (black
dashed lines) with b2-Thr1OgH, b2-Thr1NH3+, and b2-Gly47NH,
thereby disrupting the hydrogen-bonding network at Thr1 and the
oxyanion hole, respectively. b) Schematic overview of (a). The interac-
tion distances are shown in Š. The amino acids of b1 and b2 are
colored in teal and gray, respectively.
structure–activity-guided ligand optimization is limited by
the shortcomings of current in vitro CP activity assays that
make use of non-natural substrate mimetics.
Next, we aimed to employ a crystallographic approach to
determine the minimal requirements for ligand binding in the
newly discovered site. The sulfonamide compound 1 consists
of a restrained, conjugated scaffold with a limited number of
11276 www.angewandte.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 11275 –11278

Angewandte
Chemie
c) Variation of the quinoline ring motif (blue) enables
selectivity regarding either b2c or b2i.
Taken together, the herein presented crystallographic
screening procedure resulted in the identification of sulfona-
mide 1 as a promising noncovalent proteasome ligand. The
newly discovered binding site is located between the b2-S’
primed and b1 non-primed substrate binding channel and
displays significant differences in size and polarity between
iCP and cCP. The highly constrained ligand results in
a predominantly enthalpy-controlled binding event, which
in general is a driving force for target selectivity.[28] Moreover,
the absence of the pharmacokinetically unfavorable reactive
headgroups and peptidic scaffold of traditional CP inhibitors
qualifies the ligand for further optimization towards a highly
active non-peptidic proteasome inhibitor.
Figure 2. Bivalent targeting of subunits b1 and b2. Structural super-
position of 1 (yellow) and CFZ (gray)[24] bound to the b1 and b2 active
sites of yCP illustrates the combined effect of the ligand to compet-
itively and noncovalently block both catalytic activities. The sulfonyl
group of 1 is in close contact with Thr1 (blue arrow), while its 4’-
ethoxyphenyl moiety protrudes into the S4 specificity pocket of subunit
b1 (red arrow).
rotatable bonds, which minimizes the entropic
penalty that has to be overcome during the binding
process. As a result, most of the contributions to
yCP:1 complex formation must originate from
favorable enthalpic gains from the interactions
depicted in Figure 1b. Indeed, for the known
sulfonamide antibiotic sulfacetamide 2 as well as
the synthesized ligand fragments 3 and 4 (Fig-
ure S3 in the Supporting Information), no defined
electron density for the compounds could be
obtained in the recorded datasets. Therefore,
only the combination of the biaryl quinoline and
the sulfacetamide motif gives a sufficient gain in
enthalpic energy for complex formation, thus
representing an important precondition for a frag-
ment evolution approach.
In spite of recent advances, the development of
iCP-specific ligands is hampered by the limited
structural diversity of inhibitor binding sites
Scheme 2. Schematic representation of the essential structural characteristics of
1 that contribute to its selective binding affinity. The b2-S’ pocket of cCP and iCP are
represented as a solid black or dashed green line, respectively.
among the proteasomal subtypes.[25,26] However, comparison
of the primed substrate binding channels of b2c and b2i
highlights extraordinary structural differences between these
two isoforms.[11,27] This is illustrated by superposition of the
binding site of 1 with the murine b2c and b2i subunits, which
reveals that the respective b2-S’ pockets display significant
differences in size and polarity (Scheme 2). Specifically,
amino acid replacements in the iCP, including b2-S131Q
and b1-A27V, result in a smaller and more polar S’ pocket.
This knowledge will enable future efforts toward the struc-
ture-based design of specific ligands.
Acknowledgements
This work was funded by the DFG GR 1861/10-1. We are
grateful to our collaborators who helped us to assemble the
screening set. We thank the staff of the beamline X06SA at
the Paul Scherrer Institute, Swiss Light Source, Villigen
(Switzerland) for assistance during data collection and R.
Feicht for large-scale purification and crystallization of yeast
20S proteasomes.
Keywords: crystallographic screening · drug discovery ·
In summary, the introduced crystallographic screening
method identified a sulfonamide derivative as a novel pro-
teasome ligand. Its unique binding properties provide poten-
tial for further optimization in the following ways (Scheme 2):
a) The sulfonyl group forms a tight hydrogen-bonding net-
work with residues involved in catalysis at subunit b2.
b) Ligand growing is possible towards the b1 substrate
binding channel through extension of the 4’-ethoxyphenyl
group that already protrudes into the S4 pocket of subunit
b1.
proteasome · reversible inhibition · sulfonamides
How to cite: Angew. Chem. Int. Ed. 2015, 54, 11275–11278
Angew. Chem. 2015, 127, 11428–11431
[1] P. G. Richardson, T. Hideshima, K. C. Anderson, Cancer Control
2003, 10, 361 – 369.
[2] S. D. Demo, C. J. Kirk, M. A. Aujay, T. J. Buchholz, M. Dajee,
M. N. Ho, J. Jiang, G. J. Laidig, E. R. Lewis, F. Parlati et al.,
Cancer Res. 2007, 67, 6383 – 6391.
Angew. Chem. Int. Ed. 2015, 54, 11275 –11278
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 11277

.
Angewandte
Communications
[3] M. Groll, C. R. Berkers, H. L. Ploegh, H. Ovaa, Structure 2006,
14, 451 – 456.
[4] M. Groll, K. B. Kim, N. Kairies, R. Huber, C. M. Crews, J. Am.
Chem. Soc. 2000, 122, 1237 – 1238.
[5] R. Z. Orlowski, D. J. Kuhn, Clin. Cancer Res. 2008, 14, 1649 –
1657.
[6] V. Cheriyath, B. S. Jacobs, M. A. Hussein, Drugs R&D 2007, 8,
1 – 12.
[18] “Substituted Triazole And Imidazole Compounds”: S. M. Lynch,
R. E. Martin, W. Neidhart, J.-M. Plancher, T. Schulz-Gasch,
WO2014/086663A1, 2014.
[19] A. F. Kisselev, W. a van der Linden, H. S. Overkleeft, Chem.
Biol. 2012, 19, 99 – 115.
[20] C. Blackburn, C. Barrett, J. L. Blank, F. J. Bruzzese, N. Bump,
L. R. Dick, P. Fleming, K. Garcia, P. Hales, M. Jones et al.,
MedChemComm 2012, 3, 710.
[7] D. J. McConkey, K. Zhu, Drug Resist. Updates 2008, 11, 164 –
179.
[21] M. Groll, L. Ditzel, J. Lowe, D. Stock, M. Bochtler, H. D.
Bartunik, R. Huber, Nature 1997, 386, 463 – 471.
[8] W. Harshbarger, C. Miller, C. Diedrich, J. Sacchettini, Structure
2015, 23, 418 – 424.
[22] M. Groll, R. Huber, Methods Enzymol. 2005, 398, 329 – 336.
[23] N. Basse, M. Montes, X. MarØchal, L. Qin, M. Bouvier-Durand,
E. Genin, J. Vidal, B. O. Villoutreix, M. Reboud-Ravaux, J. Med.
Chem. 2010, 53, 509 – 513.
[24] E. M. Huber, W. Heinemeyer, M. Groll, Structure 2015, 23, 407 –
417.
[25] C. Dubiella, H. Cui, M. Gersch, A. J. Brouwer, S. A. Sieber, A.
Krüger, R. M. J. Liskamp, M. Groll, Angew. Chem. Int. Ed. 2014,
53, 11969 – 11973; Angew. Chem. 2014, 126, 12163 – 12167.
[26] G. de Bruin, E. M. Huber, B. Xin, E. J. van Rooden, K. Al-Ayed,
K. Kim, A. F. Kisselev, C. Driessen, M. van der Stelt, G. A.
van der Marel et al., J. Med. Chem. 2014, 57, 6197 – 6209.
[27] E. M. Huber, M. Groll, Angew. Chem. Int. Ed. 2012, 51, 8708 –
8720; Angew. Chem. 2012, 124, 8838 – 8850.
[9] M. Groettrup, R. Kraft, S. Kostka, S. Standera, R. Stohwasser, P. -
M. Kloetzel, Eur. J. Immunol. 1996, 26, 863 – 869.
[10] S. Murata, K. Sasaki, T. Kishimoto, S.-I. Niwa, H. Hayashi, Y.
Takahama, K. Tanaka, Science 2007, 316, 1349 – 1353.
[11] E. M. Huber, M. Basler, R. Schwab, W. Heinemeyer, C. J. Kirk,
M. Groettrup, M. Groll, Cell 2012, 148, 727 – 738.
[12] N. Gallastegui, P. Beck, M. Arciniega, R. Huber, S. Hillebrand,
M. Groll, Angew. Chem. Int. Ed. 2012, 51, 247 – 249; Angew.
Chem. 2012, 124, 251 – 254.
[13] P. Beck, T. A. Lansdell, N. M. Hewlett, J. J. Tepe, M. Groll,
Angew. Chem. Int. Ed. 2015, 54, 2830 – 2833; Angew. Chem.
2015, 127, 2872 – 2875.
[14] S. M. Lynch, W. Neidhart, J. M. Plancher, T. Schulz-Gasch, WO
2014086664, 2014.
[28] H. Gohlke, G. Klebe, Angew. Chem. Int. Ed. 2002, 41, 2644 –
2676; Angew. Chem. 2002, 114, 2764 – 2798.
[15] “Substituted Thiazole Compounds”: R. C. Hawley, S. M. Lynch,
WO 2014114603A1, 2014.
[16] “Substituted Thiazole Compounds”: M. Alam, R. C. Hawley,
S. M. Lynch, A. Narayanan, WO 2014086701A1, 2014.
[17] “Substituted Thiazole Compounds”: Stephen M. Lynch, A.
Narayanan, WO 2014086697A1, 2014.
Received: June 3, 2015
Revised: July 3, 2015
Published online: August 4, 2015
11278 www.angewandte.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 11275 –11278