2-Sulfonylpyrimidines Target the Kinesin HSET via Cysteine Alkylation

Article abstract of DOI:10.1002/ejoc.201900586

Supernumerary centrosomes are a source of aneuploidy, and cells have adopted different mechanisms to avoid multipolar mitoses. The kinesin HSET is required for pseudo-bipolar mitoses in cancer cells with amplified centrosomes and suppression of HSET activ

Full text of DOI:10.1002/ejoc.201900586

A Journal of
Accepted Article
Title: 2-Sulfonylpyrimidines target the kinesin HSET via cysteine
alkylation
Authors: Herbert Waldmann, Tim Förster, Erchang Shang, Kenshiro
Shimizu, Emiko Sanada, Beate Schölermann, Mylene
Hübecker, Gernot Hahne, Maria Pascual Lopez Alberca,
Petra Janning, Nobumoto Watanabe, Sonja Sievers, Fabrizio
Giordanetto, Takeshi Shimizu, Slava Ziegler, and Hiroyuki
Osada
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Eur. J. Org. Chem. 10.1002/ejoc.201900586
Link to VoR: http://dx.doi.org/10.1002/ejoc.201900586
Supported by

European Journal of Organic Chemistry
10.1002/ejoc.201900586
2
-Sulfonylpyrimidines target the kinesin HSET via cysteine alkylation
Dr. Tim Förster[a,b], Dr. Erchang Shang , Kenshiro Shimizu , Emiko Sanada[c,e], Beate
[a]
[c]
[
a]
[a,d]
[a]
Schölermann , Mylene Huebecker , Dr. Gernot Hahne , Dr. Maria Pascual López-
Alberca[a,e], Dr. Petra Janning , Prof. Dr. Nobumoto Watanabe , Dr. Sonja Sievers , Dr.
[a]
[e]
[a]
Fabrizio Giordanetto[f,g], Dr. Takeshi Shimizu , Dr. Slava Ziegler , Prof. Dr. Hiroyuki
[c]
[a]
Osada[c,e], Prof. Dr. Dr. h.c. Herbert Waldmann*[a,b]
[
[
[
[
[
a] Max Planck Institute of Molecular Physiology, Department of Chemical Biology, Otto-
Hahn-Str. 11, 44227 Dortmund, Germany
b] Technical University of Dortmund, Faculty of Chemistry and Chemical Biology, Otto-
Hahn-Str. 6, 44227 Dortmund, Germany
c] Chemical Biology Research Group, RIKEN Center for Sustainable Resource Science,
2
-1 Hirosawa, Wako, Saitama 351-0198, Japan
d] Current address: University of Oxford, Department of Pharmacology, Oxford OX1
QT, UK
3
e] RIKEN-Max Planck Joint Research Division for Systems Chemical Biology, Center for
Sustainable Resource Science, 2-1 Hirosawa, Wako, Saitama 351-0198, Japan
f] Taros Chemicals GmbH & Co. KG, Emil-Figge-Str. 76, 44227 Dortmund
g] Current address: D. E. Shaw Research, 120 West, 45th Street, New York, NY 10036,
USA
[
[
*
corresponding author: Herbert Waldmann, email: Herbert.Waldmann@mpi-
dortmund.mpg.de; URL: https://www.mpi-dortmund.mpg.de/research-groups/waldmann
1
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
Email addresses:
Tim Förster: tim.foerster@mpi-dortmund.mpg.de
Erchang Shang: shangech@163.com
Kenshiro Shimizu: kenshirou.shimizu.bs@gmail.com
Emiko Sanada: sanadae@riken.jp
Beate Schölermann: beate.schoelermann@mpi-dortmund.mpg.de
Mylene Hübecker: mylene.huebecker@pharm.ox.ac.uk
Gernot Hahne: gernot.hahne@web.de
Maria Pascual Lopez Alberca: mpascualla@gmail.com
Petra Janning: petra.janning@mpi-dortmund.mpg.de
Nobumoto Watanabe: nwatanab@riken.jp
Sonja Sievers: sonja.sievers@mpi-dortmund.mpg.de
Fabrizio Giordanetto: fabrizio.giordanetto@deshawresearch.com
Takeshi Shimizu: tshimizu@riken.jp
Slava Ziegler: slava.ziegler@mpi-dortmund.mpg.de
Hiroyuki Osada: cb-secretary@ml.riken.jp
Herbert Waldmann: herbert.waldmann@mpi-dortmund.mpg.de
Keywords: inhibitors, Protein modifications, Biological activity, Antitumor agents
2
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
Abstract
Supernumerary centrosomes are a source of aneuploidy and cells have adopted different
mechanisms to avoid multipolar mitoses. The kinesin HSET is required for pseudo-bipolar
mitoses in cancer cells with amplified centrosomes and suppression of HSET activity is
regarded a potential anti-cancer approach. We report the identification of 2-
sulfonylpyrimidine inhibitors of HSET enzymatic activity. HSET inhibition results in
establishment of multipolar mitoses and simultaneous inhibition of the kinesin Eg5 restored
bipolar spindle formation. Correlation of structure to activity revealed that the 2-
sulfonylpyrimidinyl group is required for the activity of the compound class and that 2-
sulfonylpyrimidines covalently modify HSET. In addition, these electrophiles react with
glutathione, thereby causing oxidative stress. This general reactivity needs to be taken into
account if 2-sulfonylpyrimidines will be employed in the development of biologically active
small molecules.
Introduction
Covalent inhibition of protein function and activity has proven to be a viable option in drug
[
1]
discovery, e.g., in the development of novel anti-infectives and anti-cancer drugs. In
addition, nature employs covalent targeting by endogenous covalent ligands to regulate
[
2]
proteins, e.g., IKKβ and PPAR, and various natural products covalently modulate their
[
3]
targets. Effective and safe covalent inhibitors must have tuned reactivity to avoid unspecific
[
4]
toxicity resulting from promiscuous reactions or metabolic inactivation , which poses
particular challenges for their development. For structure-based design of covalent inhibitors,
in general a reversible inhibitor is equipped with an electrophilic moiety (‘warhead’) that
targets an amino acid, which is proximal to the binding site. Most frequently, cysteine is
targeted, followed by lysine and other nucleophilic amino acids.[5]
3
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
Modulation of mitosis has been considered as a promising anti-cancer approach. However,
the use of anti-mitotic drugs is limited by severe side effects mostly due to simultaneous
targeting of non-cancerous cells. Thus, strategies that exploit differences between tumor
cells and normal cells are in demand. Supernumerary centrosomes have been described for
[
6]
several cancers including breast, pancreas and prostate tumors. Cancer cells with amplified
centrosomes employ strategies to suppress multipolar mitosis, e.g., extra copies of
chromosomes are extruded, inactivated, segregated or clustered to facilitate pseudo-bipolar
[
7]
mitosis. Suppression of multipolar mitoses may offer novel opportunities for anti-cancer
research, since small molecules interfering with this process may preferentially target cancer
cells.[6a] The kinesin HSET (also termed KIFC1 or KNSL2) is important for cancer cells but
dispensable in normal cells.[6a] HSET is a member of the kinesin-14 subfamily of minus-end
[
8]
directed kinesins that function as spindle pole organizing components. HSET is involved in
[
9]
cross-linking and bundling of microtubules. In normal cells, HSET is required for bipolar
spindle assembly under acentrosomal conditions, while it is not essential for spindle
assembly in the presence of functionally dominant centrosomes.[9b] In cancer cells, HSET is
critical for clustering extra centrosomes and ensures pseudo-bipolar cell division.[6a] In
addition, in tumor cells with correct centrosome number acentrosomal microtubule
polymerization generates acentrosomal poles and requires HSET for proper clustering and
focusing of these acentrosomal microtubule organizing centers for bipolar spindle
assembly.[10] Therefore, HSET is an emerging target in anti-cancer drug discovery and
recently the first inhibitors of HSET were identified.[11]
Here we report the discovery of 2-sulfonylpyrimidines as covalent HSET inhibitors that
impair progression through mitosis due to the formation of multipolar mitotic spindles. Due to
their electrophilic nature 2-sulfonylpyrimidines also react with glutathione (GSH) and cause
oxidative stress. This general reactivity needs to be taken into account in the use of 2-
sulfonylpyrimidines for bioactive compound development.
4
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
Results
To identify inhibitors of HSET, a collection of 148,000 compounds was screened for
inhibition of HSET’s ATPase activity in the presence of microtubules by means of the ADP-
Glo assay.[12] Hit compounds were assessed for inhibition of luciferase and for influence on
microtubules to reveal false positives. This led to the identification of 2-sulfonylpyrimidines as
a prominent class of HSET inhibitors (Table 1).
5
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
Table 1 Hits identified in screening for HSET inhibition. The activity of HSET was
determined in the presence of microtubules using the ADP-Glo system. Data are mean
values (n≥2) ± SD.
Entry No.
Structure
IC50 ± s.d. [µM]
1
2
1
2
3.4 ± 0.9
3.8 ± 1.1
3.9 ± 1.1
3
4
3
4
5.8 ±4.4
5
6
7
5
6
7
6.6 ± 3.8
14.8 ± 9.9
16.5 ± 4.5
6
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
We designed and synthesized analogues based on the general structure depicted in
Scheme 1A to establish a structure-activity relationship (SAR) and explored (i) the role of the
1
2
trifluoromethyl group (R ), (ii) replacement of the thiophene (R ), (iii) the importance of
sulfone-, sulfoxide- and thioether substituents (X), (iv) the length of the carbon linker between
3
pyrimidine and amide (n), and (v) variation of amides (R ). Scheme 1 shows two general
2
strategies for the synthesis of the compounds with different R group and different amides
respectively (Scheme 1B-1C). The syntheses of further analogues were achieved with minor
modifications of these synthetic routes (see the Experimental section for details). To
2
investigate different R substituents, aryl chloride 11 was synthesized by means of an
efficient one-pot reaction. Deprotection of thioacetate 8 with potassium carbonate led to a
solution of thiol 9, which was added dropwise to a solution of pyrimidine 10 in methanol at
2
0
°C to give the chloride 11. With this compound in hand, different R groups were introduced
by Suzuki-Miyaura coupling. The resulting thioethers 12 were further oxidized with meta-
chloroperoxybenzoic acid (mCPBA) to the target sulfones 13 (Scheme 1B). Different amides
were introduced while retaining the trifluoromethyl- and thiophene groups. As shown in
Scheme 1C, the condensation of diketone 14 and thiourea 15 led to pyrimidine 16, which
reacted with 3-chloropropionic acid 17 to yield acid 18. Formation of amides 19 followed by
the mCPBA mediated oxidation afforded mixtures of sulfoxides 20 and sulfones 21, which
were separated and subjected to biochemical investigation.
7
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
Scheme 1. General structure of the 2-sulfonylpyrimidines. (A) (B-C) Synthetic routes for
2
analogues with different substituents at R position (B) and analogues with different amides
(
C). (D) Structure-activity relationship for 2-sulfonylpyrimidines.
8
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
All compounds were tested for inhibition of HSET in a manual assay, which resulted in
higher IC50 values for compound 2 as compared to the automated HTS. The trifluoromethyl-
1
(
R ) as well as the pyrimidine ring are required for inhibition of HSET motor domain ATPase
activity (22 - 25, see Table 2, entries 9 - 12). Replacement of the sulfone with sulfoxide or
thioether reduced the activity of the compound (19a, 20a, see Table 2, entries 2 and 3).
Variation of the carbon linker revealed that the three-carbon linker without side chain is most
favorable (21l - 21p, see Table 2, entries 4 - 8).
Compared to compound 2, replacement of the 2-thiophene with 2-furan or phenyl group
retained the activity (13d, 13e, see Table 3, entries 5 and 6). However, replacement of the 2-
thiophene with 3-thiophene or 3-methoxyphenyl resulted in a twofold increase of activity
(
13a, 13f, see Table 3, entries 2 and 7). Introduction of a chlorine at the meta-position of the
aromatic ring led to three- to fivefold enhanced potency (13g; see also 13c as comparable
thiophene derivative; Table 3, entries 4 and 8), while analogues with small mono-substituents
slightly increased the activity (13h - 13l, see Table 3, entries 9 - 13). Moreover, 3,5-di-
trifluoromethyl substitution clearly increased the potency (13o, see Table 3, entry 16), while
introduction of several substituents or bulkier derivatives led to decreased activity (13n, 13p,
13r, 13s, see Table 3, entries 15, 17, 19, 20). However, the potency of compound 13q is
comparable to the parent compound (see Table 3, entry 18). Thus, a 3-chlorophenyl or 3,5-
ditrifluoromethylphenyl is most favourable as R2.
3
Analogues with different amides (R ) were also prepared based on the synthetic route
shown in Scheme 1C. As shown in Table 4, 2,4-dichlorophenyl- (21a, see Table 4, entry 2),
3,5-difluorophenyl- (21b, see Table 4, entry 3) and phenethyl amides (21j, see Table 4, entry
11) led to nearly threefold enhanced HSET inhibition, while derivatives based on different
amides (21c - 21e, 21h - 21i, see Table 4, entries 4 - 6, 9 - 10) retained or displayed only
slightly improved potency. Moreover, some heterocyclic amides had reduced activity (21f,
21k, see Table 4, entries 7 and 12). Pleasingly, the replacement of the aromatic amide with
9
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
an n-hexanyl amide resulted in a fourfold increase in activity to an IC50 of 5.5 µM (21g, see
Table 4, entry 8).
A summary of the SAR study is shown in Scheme 1D: a sulfone and an electron
withdrawing group (trifluoromethyl) in the pyrimidine ring are essential for activity, while the
2
3
thiophene (R ), linker and amide (R ) tolerate changes.
1
0
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
Table 2: Structure-activity relationship studies of 2-sulfonylpyrimidines. IC50 values for
inhibition of HSET activity by compounds with variation of the pyrimidine ring, the
1
2
trifluoromethyl (R ), thiophene (R ) and sulfonyl groups (X) as well as the linker length were
determined by means of the ADP-Glo™ Kinase Assay (n=2-3).
Entry
No.
2
R1
R2
X
Z
Linker
IC50 ± s.d. [µM]
1
CF
3
3
3
3
3
3
3
3
SO
2
N
19.6 ± 6.2
2
3
4
5
6
7
19a
20a
21l
CF
CF
CF
S
N
N
N
N
N
N
N
> 50
SO
SO
SO
SO
SO
SO
> 50
2
2
2
2
2
> 30
21m CF
5.9 ± 2.8
> 30
21n
21o
21p
CF
CF
CF
10.4 ± 3.0
13.9 ± 0.5
8
9
22
23
CH
3
SO
SO
2
2
N
N
> 50
> 50
10
1
1
2
24
25
H
H
H
H
SO
SO
2
2
N
> 50
> 50
1
CH
1
1
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
Table 3: Structure-activity relationship of 2-sulfonylpyrimidines with different
2
substitutents at R . Determination of IC50 values for inhibition of HSET activity was
performed by means of ADP-Glo assay (n=2-3).
IC50 ± s.d.
IC50 ± s.d.
[µM]
Entry
No.
R2
Entry
No.
R2
[
µM]
1
2
3
2
19.6 ± 6.1
11
12
13
13j
13k
13l
9.6 ± 0.7
13a
13b
10.2 ± 2.5
21.5 ± 5.1
11.0 ± 5.2
8.2 ± 0.5
4
5
6
13c
13d
13e
5.9 ± 1.3
18.8 ± 3.1
22.2 ± 6.2
14
15
16
13m
13n
13o
26.4 ± 1.6
>50
5.1 ± 1.7
7
8
9
13f
9.7 ± 0.3
3.8 ± 0.1
17
18
13p
13q
>50
13g
12.2 ± 5.4
13h
13i
8.2 ± 1.8
9.4 ± 1.8
19
20
13r
>50
>50
10
13s
1
2
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
Table 4: Structure-activity relationship of 2-sulfonylpyrimidines with substitutions at
3
R . Determination of IC50 values for inhibition of HSET activity was performed by means of
ADP-Glo™ Assay (n=2-3).
IC50 ± s.d.
µM]
IC50 ± s.d.
[µM]
Entry
No.
R3
Entry No.
R3
[
1
2
2
19.6 ± 6.1
8.3 ± 1.3
7
8
21f
>30
21a
21g
5.5 ± 1.4
3
21b
7.0 ± 0.8
9
21h
18.0 ± 6.8
4
5
6
21c
21d
21e
10.3 ± 0.9
14.3 ± 6.6
14.9 ± 5.8
10
11
12
21i
21j
21k
16.6 ± 4.4
7.9 ± 1.1
>30
Impairment of HSET function by small molecules should cause defective mitosis in cells
with formation of additional centrosomes due to failure in centrosome clustering. Thus, the
influence of the compound collection on mitosis, particularly focusing on mitotic spindle
morphology, was assessed in the human cervix cancer cell line HeLa and breast carcinoma
BT-549 cells. HeLa cell populations contain a small number of cells with amplified
centrosomes [6a, 11b], while BT-549 is widely used to study HSET activity since 45% of the
1
3
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
cells display centrosome amplifications.[6a] Several compounds reduced the cell number or
were cytotoxic (see Figure S1and S2). Interestingly, only compound 5 showed specific
cellular activity and induced the accumulation of mitotic cells with multipolar mitotic spindles
(
Figure 1A and 1B, Figure S3A-S3B). This resulted in a 12-fold increased amount of mitotic
cells compared to vehicle-treated cells (Figure 1C and 1D). In BT-549 breast carcinoma
cells, treatment with 5 led to a threefold increase in the amount of mitotic cells (Figure S3C-
S3D). This phenotype is prominent for small molecules that target microtubules.[13] However,
compound 5 did not affect the in vitro tubulin polymerization (Figure 1E) and thus, most likely
does not target tubulin/microtubules.
1
4
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
Figure 1. Influence of compound 5 on mitosis in HeLa cells. (A) HeLa cells were treated
with 5 (25 µM) or DMSO as a control for 24 h prior to staining with an anti-α-tubulin antibody
1
5
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
(
green) and DAPI (blue). Scale bar: 50 µm. (B) Compound 5 causes the formation of
multipolar mitotic spindles. HeLa cells were treated were treated with 5 (25 µM) or DMSO as
a control for 24 h prior to staining with an anti-α-tubulin antibody (green), anti-pericentrin
antibody (red) and DAPI (blue). Scale bar: 50 µm. Arrows indicate inlets. (C and D)
Compound 5 leads to mitotic arrest. HeLa cells were treated with 5 (25 µM) or DMSO as a
control for 24 h prior to staining with an anti-phospho-histone H3 (pH3) antibody (red) or
DAPI (blue). Images (C) and image-based quantification (D) of mitotic cells are shown. Scale
bar: 50 µm. Data are percentage values of mitotic cells of 70 (DMSO) and 87 (compound 5)
images (n≥3 biological replicates) ± s.d. (E) Influence of 5 on in vitro tubulin polymerization.
50 µM compound 5 and 1 µM of the control substances were mixed with 10 µM tubulin
followed by application of 500 µM GTP to start the reaction. Absorption at 340 nm was
measured for 60 min and normalized to the value at t=0 for each probe. (F) Cell growth in the
presence of compound 5. HeLa cells were treated with 2, 10 and 25 µM of 5 and confluence
was quantified as a measure of viability/cell growth for 72 h using real-time live-cell
microscopy. Confluence was normalized to the value at the time of seeding for each probe.
Arrow indicates time point of compound addition. Data is representative of three biological
replicates.
The minus-end directed activity of HSET is antagonistic to the plus-end directed activity of
the kinesin Eg5.[9b] Whereas Eg5 inhibition results in the formation of monoastral mitotic
spindles, simultaneous inhibition of Eg5 and HSET can rescue the monoaster phenotype.[9b,
11b, 14]
To assess the ability of compound 5 to revert the monopolar mitotic spindle phenotype,
HeLa cells were co-treated with the Eg5 inhibitor S-trityl-L-cysteine (STLC) [15] in the
presence or absence of 25 µM of 5 and the mitotic spindle architecture was analyzed (Figure
2). While single treatment with STLC or compound 5 led to the formation of monopolar and
multipolar mitotic spindles respectively, mitotic spindles resembled a bipolar morphology
1
6
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
when both inhibitors were present (Figure 2). This finding is in line with HSET inhibition by
derivative 5 in cells.
Figure 2. Compound 5 rescues Eg5 inhibition. HeLa cells were treated with 25 µM
compound 5 with or without 2 µM STLC for 24 h. Cells were fixed and permeabilized prior to
staining of DNA, centrosomes and microtubules using DAPI (blue), a pericentrin antibody
(
red) and an alpha-tubulin antibody (green). Representative images (A) and quantification (B)
are shown. (B) Data are mean values ± s.d. of four individual experiments (n=4; 37-139 cells
per condition). Scale bar: 50 µm. Arrows indicate inlets.
1
7
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
Sulfones are good leaving groups when linked to electron-withdrawing aromatic rings.[16]
Recently, 2-sulfonylpyrimidines were reported as mild alkylating agents, which react with
highly nucleophilic cysteines.[17] Thus, we assumed a covalent binding mode for compound 5
as well. To this end, we monitored formation of covalent adducts of compound 5 and a
peptide containing nucleophilic amino acids in an in vitro peptide trapping assay using mass
spectrometry.[18] Upon incubation of 5 with the peptide Ac-YACAKASAHA, a new peak (m/z
1262.27) was detected in addition to the peptide peak (m/z 1034.45), thus proving the
formation of peptide-compound conjugate. The mass increase (Δ m/z 227.82, measured)
corresponds to the transfer of the 4-(thiophen-2-yl)-6-(trifluoromethyl)pyrimidine moiety to the
test peptide (Δ m/z 228.00, theoretical) (Figure 3A and B). To explore the putative covalent
binding of compound 5 to HSET, nano-HPLC-MS/MS measurements upon incubation of
HSET with 50 µM 5, followed by tryptic digestion of the protein and a database search were
performed. We considered the proposed nucleophilic aromatic substitution (S Ar) for 2-
N
sulfonylpyrimidines[17], elimination of substituted sulfinic acid and a transfer of 4-(thiophen-2-
yl)-6-(trifluoromethyl)pyrimidine to nucleophilic amino acids (Figure 3A). Oxidation of
methionine, carbamindomethylation of cysteine and 4-(thiophen-2-yl)-6-
(
trifluoromethyl)pyrimidine addition to cysteine, serine, lysine, or histidine, respectively, were
defined as variable modifications. The modification of two prominent cysteine residues (Cys
85 and Cys 663) with 4-(thiophen-2-yl)-6-(trifluoromethyl)pyrimidine according to the
4
reaction proposed by Bauer et al.[17] were clearly identified (localization probability of 1 in
both cases, PEP 1.1E-19 for Cys 485 and 2.1E-30 for Cys 663). Two further modifications
(
His 370 and Cys 405) were assumed but the localization and/or identification probabilities
were worse (His 370: localization probability 0.999972, PEP 0.007; Cys 405: localization
probability 0.864, PEP 1.21E-09). The MS/MS spectrum of the tryptic peptide KGQGGECEIR
modified at position Cys 485 is shown in Figure 3C (for the remaining MS/MS spectra see
Figure S4) The modified residues were mapped near the protein surface, i.e., not the active
site (Figure 3D). Thus, their alkylation is most likely due to favorable environment, i.e.,
1
8
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
accessibility and nucleophilicity. These results confirm 5 as an alkylator and the covalent
modification of HSET as the likely cause for HSET inhibition by 5.
1
9
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
2
0
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
Figure 3. Compound 5 alkylates HSET. (A) Proposed mechanism of protein alkylation by 5.
B and C) The model peptide Ac-YACAKASAHA was incubated with 5 (50 µM) and was then
(
subjected to mass spectrometry. MS spectra for the peptide (B) and peptide + 5 (C) are
shown. (D) Formation of compound-HSET conjugate. HSET was incubated with 5 (50 µM)
followed by MS/MS analysis. The spectrum for the alkylated peptide 479KGQGGECEIR488 is
shown. The localization probability resulting from MaxQuant search is mentioned in brackets
behind the modified residue. (E) Mapping of the alkylated residues in the structure of HSET.
PDB: 5WDH.
The reactivity of sulfonyl pyrimidines may also be directed towards other cellular thiols than
cysteines such as glutathione (GSH). Since sulfonyl-containing small molecules have been
shown to react with GSH,[19] we analyzed the stability of compound 5 in the presence of 5
mM GSH, which corresponds to the estimated thiol concentration in cells.[20] Compound 5
was added at a concentration of 25 µM, resulting in a 200-fold excess of GSH. Mass
spectrometry analysis of an aqueous solution of compound 5 revealed one peak. In contrast,
three peaks were detected upon incubation of 5 in the presence of 5 mM GSH (Figure 4A).
Mass analysis confirmed the nucleophilic substitution of the sulfonyl part of 5 and the
formation of the 4-(thiophen-2-yl)-6-(trifluoromethyl)pyrimidine-2-yl-GSH conjugate (Figure
4A). Quantification of the unreacted compound mass peak revealed partial scavenging (51.7
±
16.7 %) of compound 5 by thiols (Figure 4B). However, in the presence of GSH a large
amount of unreacted compound remained, which indicates slow substitution of the sulfonyl
by GSH. Thus, a sufficient amount of intact compound is likely to reach its target protein in
cells under physiological conditions.
2
1
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
Figure 4. Compound 5 conjugates to GSH. (A und B) 5 mM GSH was incubated with 5 (25
µM) for 24 h at 37 °C prior to ESI-MS analysis. Representative spectra are shown. (B) Upon
GSH treatment spectra were quantified for intact compound abundance and related to pure
compound in H O. Data are mean values (n=5) ± SD. (C) Influence of 5 on cellular
2
GSH/GSSG levels. HeLa cells were incubated with the compound (10 µM) or DMSO as a
2
2
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
control for 24 h prior to determination of GSH and GSSG levels. Data are mean values (n=6)
±
SD.
GSH is the most abundant scavenger of reactive oxygen species and its depletion may
cause oxidative stress and eventually cell death. Conjugation of reactive compounds to GSH
in cells reduces cellular GSH levels. The ratio of the levels of GSH and its oxidized form
GSSG is a measure of oxidative stress [21]. To assess a possible influence of compound 5 on
the GSH/GSSG ratio, U2OS cells were treated with 5 for 24h before determining the levels of
each GSH derivative. Although 5 did not change the level of reactive oxygen species in HeLa
cells, it reduced the GSH/GSSG ratio as compared to cells, which were treated with DMSO
(
10.8 for 5 vs. 22 for the control, see Figure 4C), which indicates induction of oxidative stress
by 5.
Discussion
The kinesin HSET is essential only in cells with amplified centrosomes[6a] and selective
inhibition of HSET opens up an opportunity to selectively target malignant cells with
overduplicated centrosomes, while sparing healthy cells. HSET is overexpressed in a broad
range of cancer types, including lung, breast, colon and cervical tumors,[22] which all exhibit
centrosome amplifications.[6b] HSET has been proposed as biomarker for breast tumors,
since HSET overexpression correlates with cancer development and aggressiveness as well
as accelerated proliferation.[22] Similar observations have been made in ovarian cancer,
where HSET overexpression is associated with poor prognosis in high-grade serous
tumors.[23] To date, only a few small-molecule inhibitors of HSET have been reported.[11, 24]
Here, we identified 2-sulfonylpyrimidines as a prominent class of hit compounds in a screen
for HSET inhibition. A compound collection based on the 2-sulfonylpyrimidine scaffold was
prepared to explore the structure-activity relationship. Introducing substituents like furan or
2
phenyl instead of thiophene at R retained the activity. However, replacement of the 2-
2
3
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
thiophene with 3-thiophene or 3-methoxyphenyl resulted in a twofold increase of activity.
Introduction of a chlorine at the meta-position of the aromatic ring led to three- to fivefold
enhanced potency, whereas bulkier groups were not tolerated. A three carbon linker was
shown to be most advantageous and n-hexanyl amide was most appropriate as R3
substituent. The 2-sulfonyl group as well as the 4-trifluoromethyl substituent in the pyrimidine
ring are essential for HSET inhibition.
The sulfonyl-pyrimidine derivative 5 bearing a thiophene and 3-methyl-1-phenyl-1H-
pyrazole moieties increased the amount of mitotic cells and provoked multipolar spindle
formation in HeLa and BT-549 cells. The compound prevented the formation of monopolar
mitotic spindles induced by Eg5 inhibition which is in line with inhibition of HSET and the
opposing activity of HSET and Eg5 on the spindle.[9b] In vitro peptide trapping identified the
substance as covalent binder of at least two cysteine residues in HSET and suggested the
binding of the 4-trifluoromethyl-pyrimidine scaffold to cysteines. Reactivity towards cysteines
is related also to reactivity to other cellular thiols, e.g., GSH. Derivative 5 is only partially
GSH-reactive as a substantial amount of the compound remained unreacted upon 24 h of
incubation in thiol-containing solution. Thus, the compound is likely to reach its target in the
intracellular environment. However, compound 5 reduced the GSH/GSSG levels in cells,
which is indicative of oxidative stress and of an additional mode-of-action along with HSET
inhibition. Oxidative stress has been linked to mitotic arrest. [25] Therefore, reactive
compounds that target GSH may also lead to mitotic arrest. However, rescue of Eg5-induced
spindle monopolarity was observed for compound 5, thus suggesting a more complex
regulation of mitotic arrest than only through oxidative stress. Search for known bioactivity of
the 3-methyl-1-phenyl-1H-pyrazole moiety revealed a link to kinesin inhibition [26] (see Figure
S5A), whereas a different 3-methyl-1-phenyl-1H-pyrazole -bearing compound was toxic to
L12102 and K562 cells[27] (see Figure S5B). In addition, various compounds bearing the 1-
phenyl-1H-pyrazole moiety were among the hits from the initial screening for HSET
modulators, which inhibited the HSET activity >50% at 50 µM. Thus, 3-methyl-1-phenyl-1H-
pyrazole in compound 5 may guide the compound to mitotic kinesins, e.g. HSET, whereas
2
4
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
the 2-sulfonylpyrimidine part may mediate the reactivity to proteins and GSH. Sulfonyl group-
bearing compounds have been described as small-molecule modulators of biological
[
5]
[17, 19, 28]
processes including substituted 2-sulfonylpyrimidines.
Sulfones are good leaving
groups when linked to electron withdrawing aromatic rings[19] and for various compounds
conjugation to GSH has been confirmed.[17, 19, 29] Thus, this general reactivity needs to be
taken into account if 2-sulfonylpyrimidines will be employed in the development of biologically
active small molecules.
2
5
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
Experimental Section
General Techniques
All solvents were reagent grade. Acetone, acetonitrile (MeCN), benzene, chloroform (CHCl
3
),
dichloromethane (CH Cl ), dimethyl sulfoxide (DMSO), ethanol (EtOH), ethyl acetate (EtOAc)
2
2
methanol (MeOH), tetrahydrofuran (THF) N,N-dimethylformamide (DMF), pyridine and
toluene were purchased from Sigma-Aldrich, FUJIFILM Wako Pure Chemical Corporation,
and Tokyo Chemical Industry Co., Ltd. Compounds 1 - 7 were purchased from ChemDiv.
Reactions were monitored by thin layer chromatography (TLC) with 0.25 mm E. Merck
precoated silica gel plates with F-254 indicator. E. Merck 0.5 mm precoated silica gel was
used for preparative thin layer chromatography (PLC). Merck 60 (40-63 µm) silica gel and
Kanto chemical silica gel 60 N (spherical, neutral) (40-50 µm) was used for flash
1
13
chromatography. H NMR and C NMR spectra were recorded on Bruker DRX400 (400
MHz), Bruker DRX500 (500 MHz), INOVA500 (500 MHz), JEOL JNM-ECA-500 (500 MHz),
and INOVA600 (600 MHz) spectrometer using CDCl or CD OD as solvents. Data are
3
3
reported in the following order: chemical shifts (δ) were reported in parts per million (ppm)
1
with the solvent resonance as internal standard (CDCl
3
: δ = 7.26 ppm for H, δ = 77.16 ppm
13
1
13
for C; CD OD: δ = 3.31 ppm for H, δ = 49.00 ppm for C); multiplicities are indicated as
3
following: s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet, br = broad; coupling
constants (J) are given in Hertz (Hz). High resolution mass spectra were recorded on a LTQ
Orbitrap mass spectrometer coupled to a Accela HPLC-System (HPLC column: Hypersyl
GOLD, 50 mm x 1 mm, particle size 1.9 μm, ionization method: electron spray ionization) and
Waters IMS Vion QTof system coupled to a Waters UPLC H-class (column: Waters UPLC
BEH C18, 50 mm x 2.1 mm, particle size 1.7m, ionization method: electron spray
ionization).
2
6
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
Subcloning, Expression and purification of HSET
pGEX-F/A-HSETΔ228 containing the kinesin motor domain of human HSET and
corresponding to amino acids 229-673 as well as EG5_pQE80 containing the motor domain
of Eg5 were kind gifts of T. U. Mayer, University of Konstanz, Germany. HSETΔ228 was
subcloned by the Dortmund Protein Facility into pOPIN-CHis or pOPIN-NHis-EGFP plasmid
by amplifying HSETΔ228 cDNA using 5’-
CAAAGGAGATATACCATGGAGCTTCAGAAAAAACAGGTGGAATTGC-3’/ 5’-
GATGGTGATGGTGATGTTTCTTCCTGTTGGCCTGAGCAGTAC-3’ or 5’-
GAAGTTCTGTTTCAGGGTCCCGAGCTTCAGAAAAAACAGGTGGAATTGC-3’/5’-
TAAACTGGTCTAGAAAGCTTTACTTCCTGTTGGCCTGAGCAGTAC-3’ respectively
followed by sequence and ligation independent cloning[30] to obtain HSETΔ228 with a C-
terminal His
8
tag (HSETΔ228-His) or HSETΔ228 with N-terminal His -EGFP tag (His-EGFP-
8
HSETΔ228). pOPIN-HSETΔ228-His and EG5_pQE80 were transformed in E. coli strain
BL21 DE3 Codon Plus RIL. Cells were grown overnight in auto-induction medium (TB
medium supplemented with 0.01 % lactose and 2 mM MgSO ). Cells were diluted 1:100 and
4
were incubated for 4 h at 37 °C and 200 rpm prior to reduction of the temperature to 25 °C
and further incubation for 20 h. Cells were harvested and lysed in buffer containing 50 mM
Hepes, 300 mM NaCl, 20 mM imidazol, 1 mM TCEP, 0.5 mM 4-(2-Aminoethyl)
benzenesulfonyl fluoride hydrochloride (AEBSF) and 40 U/ml DNase using a cell disruptor.
After centrifugation at 75,600 rcf at 10°C for 30 min the supernatant was loaded on a Ni-
HisTrap FF crude column. After washing with 50 mM Hepes (pH 8.0), 300 mM NaCl, 30mM
imidazol and 1 mM TCEP proteins were eluted with 50 mM Hepes (pH 8.0), 300 mM NaCl,
500 mM imidazol and 1 mM TCEP. Eluted proteins were further purified by size exclusion
chromatography using HiLoad 16/60 Superdex 75.
Determination of kinesin activity using ADP-Glo™
2
7
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
In vitro HSET inhibition by sulfonylpyrimidines was measured by ADP generation using
ADP-Glo™ assay (Promega) in presence of premade microtubules. Premade microtubules
(
Cytoskeleton) were resuspended in 15 mM PIPES pH 7.0, 1 mM MgCl and 4 µM taxol and
2
incubated at room temperature for 15 min until further use. Master mix containing 1 µM
purified HSET and 2 µM premade microtubules was prepared in 40 mM Tris pH 7.5, 20 mM
MgCl and 0.1 mg/ml BSA. Compounds existed in stock solutions of 10 mM and were pre-
2
diluted in DMSO. 0.5 µl of the compound pre-dilution was added to 50 µl master mix and
shortly vortexed to mix. 3 µl were added to a 384 well plate (3673, Corning) in
quadruplicates. ATP was added to each sample containing well to generate a final
concentration of 400 µM.
All steps were performed at room temperature. The samples were incubated for 60 min to
allow the ATPase reaction, followed by addition of 5 µl Glo reagent to each well. After further
40 min incubation to deplete remaining ATP and convert ADP to ATP, 10 µl detection
reagent were applied to all wells to carry out the luciferase reaction. Luminescence was
determined using a microplate reader (Infinite M200, Tecan). Settings: Shaking Linear
Duration 5 s, Shaking Linear Amplitude 1.5 mm, Waiting time 5 s, Mode Luminescence,
Attenuation NONE, Integration Time 1000 ms, Settle Time 5 ms.
Cell culture
BT-549 and HeLa cells were obtained from DSMZ, U2OS cells were purchased from CLS
GmbH. All cells were cultured in Dulbecco’s modified Eagle’s medium with L-glutamine
(
DMEM, PAN Biotech) supplemented with 10% fetal bovine serum (FBS, Life Technologies),
sodium pyruvate, non-essential amino acids, 100 µg/ml streptomycin and 100 U/ml penicillin
PAN Biotech) at 37°C and 5% CO in humidified atmosphere.
(
2
Immunofluorescence
2
8
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.201900586
5.000 cells / well were seeded in a black 96 well plate with transparent bottom and
incubated for 24 h at 37°C and 5 % CO in humidified atmosphere. Cells were treated with
2
compounds for additional 24 h. Fixation and permeabilization were carried out using 3.7 %
formaldehyde and 0.1 % (v/v) Triton X-100 for 30 min prior to blocking by 2 % BSA (w/v) in
TBS-T for 1 hour. Immunocytochemical staining with FITC-labeled anti-α-tubulin- (Sigma-
Aldrich) and pericentrin-antibodies (Abcam) was performed overnight at 4 °C, followed by
application of Alexa-594 coupled antibody (Thermo Scientific) for 1 hour at room
temperature. Images were subsequently acquired at an automated microscope (Axiovert
M200, Zeiss).
Peptide trapping
Selected compounds were evaluated for covalent binding to HSET and Ac-YACAKASAHA
peptide by in vitro peptide trapping approach using ESI-MS technique. For binding studies
with the Ac-YACAKASAHA peptide, 50 µM compound was added to 0.4 mM peptide in 75 µl
Tris (25 mM, pH 7.4) in nitrogen atmosphere and incubated for 2 h at room temperature. The
sample was centrifuged for 10 min at 5000 g at 4°C. The supernatant was 1:10 diluted in 95
%
(v/v) H O/5 % (v/v) acetonitrile prior to measurement. Samples were subjected to flow-
2
injection analysis using an Agilent 1100 HPLC system coupled to an LTQ-XL electrospray
ionization mass spectrometer (Thermo Scientific). The flow rate was set to 300 µL/min; the
eluent composition was water/acetonitrile 1/1 acidified with 0.1% formic acid; a mass range
from 700 to 2000 m/z was recorded; and a source fragmentation of 20 V was used.
57 µg purified HSET in 25 mM Tris pH 7.4 were mixed with 50 µM compound or 0.5 %
DMSO as vehicle control, followed by 30 min incubation at room temperature. The samples
were concentrated by centrifugation using Amicon® Ultra 0.5ml filters (Merck Millipore) and
digested using 1.5 µg trypsin overnight. Trypsinized samples were stage-tip purified [31] and
subsequently measured using UltiMateTM 3000 RSLCnano-System coupled to a Q
2
9
This article is protected by copyright. All rights reserved.