Tozasertib Analogues as Inhibitors of Necroptotic Cell Death

Article abstract of DOI:10.1021/acs.jmedchem.7b01449

Receptor interacting protein kinase 1 (RIPK1) plays a crucial role in tumor necrosis factor (TNF)-induced necroptosis, suggesting that this pathway might be druggable. Most inhibitors of RIPK1 are classified as either type II or type III kinase inhibitors. This opened up some interesting perspectives for the discovery of novel inhibitors that target the active site of RIPK1. Tozasertib, a type I pan-aurora kinase (AurK) inhibitor, was found to show a very high affinity for RIPK1. Because tozasertib presents the typical structural elements of a type I kinase inhibitor, the development of structural analogues of tozasertib is a good starting point for identifying novel type I RIPK1 inhibitors. In this paper, we identified interesting inhibitors of mTNF-induced necroptosis with no significant effect on AurK A and B, resulting in no nuclear abnormalities as is the case for tozasertib. Compounds 71 and 72 outperformed tozasertib in an in vivo TNF-induced systemic inflammatory response syndrome (SIRS) mouse model.

Full text of DOI:10.1021/acs.jmedchem.7b01449

Subscriber access provided by UNIV OF DURHAM
Article
Tozasertib analogues as inhibitors of necroptotic cell death
Sam Hofmans, Lars Devisscher, Sofie Martens, Dries Van Rompaey, Kenneth
Goossens, Tatyana Divert, Wim Nerinckx, Nozomi Takahashi, Hans De Winter,
Pieter Van der Veken, Vera Goossens, Peter Vandenabeele, and Koen Augustyns
J. Med. Chem., Just Accepted Manuscript • Publication Date (Web): 13 Feb 2018
Downloaded from http://pubs.acs.org on February 13, 2018
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 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 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 100
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Tozasertib analogues as inhibitors of necroptotic
cell death
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
a
b,c
a
Sam Hofmans , Lars Devisscher , Sofie Martens , Dries Van Rompaey , Kenneth
a
b,c
d,e,
b,c
a
Goossens , Tatyana Divert , Wim, Nerinckx , Nozomi Takahashi , Hans De Winter ,
a
b,c
b,c,f
a
Pieter Van Der Veken , Vera Goossens , Peter Vandenabeele , Koen Augustyns
a
Laboratory of Medicinal Chemistry, University of Antwerp, Universiteitsplein 1, Wilrijkꢀ
Antwerp 2610, Belgium
b
Molecular Signaling and Cell Death unit, VIB Center for Inflammation Research,
Technologiepark 927, ZwijnaardeꢀGhent 9052, Belgium
c
Department of Biomedical Molecular Biology (DBMB), Ghent University, Technologiepark
927, ZwijnaardeꢀGhent 9052, Belgium
d
Unit for Medical Biotechnology, Center for Medical Biotechnology, VIB, Technologiepark
927, ZwijnaardeꢀGhent 9052, Belgium
e
Laboratory for Protein Biochemistry and Biomolecular Engineering, Department of
Biochemistry and Microbiology, Ghent University, K.L.ꢀLedeganckstraat 35, Ghent 9000,
Belgium
f
Methusalem Program, Ghent University, Ghent 9000, Belgium
Keywords: necroptosis, regulated necrosis
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 100
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ABSTRACT
Receptor interacting protein kinase 1 (RIPK1) plays a crucial role in tumor necrosis factor (TNF)ꢀ
induced necroptosis, suggesting that this pathway might be druggable. Most inhibitors of RIPK1 are
classified as either type II or type III kinase inhibitors. This opened up some interesting perspectives
for the discovery of novel inhibitors that target the active site of RIPK1. Tozasertib, a type I panꢀ
aurora kinase (AurK) inhibitor, was found to show a very high affinity for RIPK1. Since tozasertib
presents the typical structural elements of a type I kinase inhibitor, the development of structural
analogues of tozasertib is a good starting point for identifying novel type I RIPK1 inhibitors. In this
paper we identified interesting inhibitors of mTNFꢀinduced necroptosis with no significant effect on
AurK A and B, resulting in no nuclear abnormalities as is the case for tozasertib. Compouds 71 and 72
outperformed tozasertib in an in vivo TNFꢀinduced systemic inflammatory response syndrome (SIRS)
mouse model.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
INTRODUCTION
Cell death upon stimulation by tumor necrosis factor (TNF) mainly occurs by two major forms,
namely apoptosis and necrosis. These two different forms of cell death are each characterized by their
1,2
own typical morphological features. Apoptosis is the major cell death pathway used to remove
3
unwanted and harmful cells in an immunologically “silent”. The main executioners of the apoptotic
pathway are the soꢀcalled caspases, a class of proteolytic enzymes that can cleave various intracellular
4
proteins. Since the execution of apoptosis is highly dependent on caspaseꢀactivity, it can be inhibited
5,6
by using the panꢀcaspase inhibitor zVAD.fmk. In contrast, necroptosis is characterized by a lack of
caspase activation, cytoplasmic and mitochondrial swelling, random DNA degradation and irreversible
plasma membrane damage. This in turn results in the leakage of intracellular content and the soꢀcalled
“damage associated molecular patterns” (DAMPs) in the surrounding tissue which then initiate an
7
,8
inflammatory response.
2
ACS Paragon Plus Environment

Page 3 of 100
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Necrosis has historically been considered to be an uncontrolled form of cell death that is refractory to
therapy. However, cell death research was revitalized by the understanding that necrosis can occur in a
9
programmed and tightly regulated manner leading to the introduction of the term necroptosis.
Multiple forms of regulated necrosis have been characterized so far, including necroptosis, ferroptosis,
oxytosis and pyroptosis. One of the most studied forms of regulated necrosis is tumour necrosis factor
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
9
receptorꢀ1 (TNFR1) mediated necroptosis which can be induced by TNF. This process of TNFꢀ
mediated necroptosis is highly dependent on the function of receptor interacting protein kinase 1
1
0–14
(
RIPK1) which was identified as a crucial kinase in the necroptotic core machinery.
It should be
noted that stimulation of TNFR1 by TNF does not irrevocably result in cell death. The binding of TNF
to its receptor results in the formation of a receptor proximal complex I in which RIPK1 fulfills an
important scaffolding function. Postꢀtranslational effects to this complex I can result in the
downstream activation of the transcription factor NFꢀkB. This in turn results in the upregulation of
various proꢀsurvival genes. To maintain a proꢀsurvival status of the cell, the kinase activity of RIPK1
is supressed. However, when the kinase activity of RIPK1 is not supressed, multiple cellular outcomes
such as apoptosis and necroptosis become possible. This makes RIPK1 an important enzyme that
functions at the crossroads between cell death and survival, having a scaffolding function that
promotes cell survival while its kinase function is important for cell death signalling (both apoptosis
7,14–17
and necroptosis).
Since the discovery that TNFꢀinduced necroptosis is dependent on the kinase
activity of RIPK1, it became clear that this pathway might be druggable. The design of chemical
inhibitors of RIPK1 could further elucidate the complex signalling cascade that drives this form of
regulated necrosis.
The first inhibitors of necroptosis were discovered in 2005 in a phenotypical screening to inhibit
14,18–20
necroptotic cell death.
This lead to the identification of both necrostatinꢀ1 (Necꢀ1, 1, Figure 1A)
and necrostatinꢀ1 stable (Necꢀ1s, 2, Figure 1A) which were able to inhibit the kinase activity of
RIPK1. No other molecule has been used as extensively as Necꢀ1s as an important tool compound in
cell death research. More importantly, the discovery of Necꢀ1s demonstrated that the inhibition of
20–27
RIPK1 could ameliorate certain conditions and diseases.
Following the success of Necꢀ1s,
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 100
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
significant followꢀup research has been conducted which lead to the identification of multiple other
20,28–30
molecules that were subsequently classified as necrostatins.
Despite their excellent kinase
selectivity, all of the reported necrostatins struggle with multiple drawbacks such as a narrow SAR
1
4,18,28,31–33
profile, moderate potency and nonꢀideal pharmacokinetic properties.
underline the need for the discovery of novel inhibitors of necroptosis.
These findings
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Cocrystallization of Necꢀ1s with the RIPK1 kinase domain showed that it can be classified as a type
III kinase inhibitor by occupying an allosteric hydrophobic pocket created by the DLGꢀout
2
8,34–37
conformation without interacting with any of the residues in the hinge region of RIPK1.
Other
cocrystallization studies that revealed that all necrostatins bind in a similar manner and can therefore
28
all be classified as type III inhibitors.
In addition to the necrostatins, another very promising RIPK1 inhibitor was recently discovered by
GlaxoSmithKline (GSK). Screening of RIPK1 against GSK’s property collection of DNAꢀencoded
smallꢀmolecule libraries identified GSK’481 (3, Figure 1A) which was able to selectively inhibit the
kinase activity of RIPK1. Upon investigation of the binding mode of GSK’481 (3) it was found that
the benzylic moiety occupied the same allosteric lipophilic pocket as the necrostatins while the
benzoxazepinone moiety was located near the ATP binding site. It is thus difficult to classify this
38
compound as either a pure type II or type III kinase inhibitor. Lead optimization of GSK’481
resulted in the development of GSK2982772 (4, Figure 1A) which is currently in phase 2a clinical
39
studies for psoriasis, rheumatoid arthritis and ulcerative colitis.
Less work has been reported considering type II RIPK1 inhibitors, however some compounds that use
this binding mode have been reported by various research groups together with some very interesting
28,35
in vivo activity. (5ꢀ7, Figure 1B)
In an attempt to discover novel necroptosis inhibitors, we decided to use a drug repurposing approach.
The investigation whether wellꢀcharacterized drugs can be used in an entirely different context has
received increased interest over the last years, mainly due to the rising costs of traditional drug
development. Also drugs selected for drug repurposing have often been through several stages of
4
ACS Paragon Plus Environment

Page 5 of 100
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
clinical development and therefore have a wellꢀcharacterized safety profile and pharmacokinetic
4
0–42
profile.
A paper by Davis et al. published a kinase inhibitor selectivity study back in 2011
reporting a screening of 72 preclinical and clinical kinase inhibitors against 442 kinases, covering
more than 80% of the human catalytic protein kinome. In this screening tozasertib (also known as VXꢀ
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
680 or MKꢀ0457) had the strongest affinity for RIPK1 of all tested kinase inhibitors (8, Figure 1C)
43
with a K of 20 nM. In addition to its high affinity for RIPK1 we also considered tozasertib to be
d
opportune due to its scaffold that is divergent from currently used type II and type III RIPK1
37
inhibitors. The development and screening of structural analogues of tozasertib may thus lead to new
insights in the synthesis of novel RIPK1 inhibitors. A first objective in the framework of drug
repurposing was thus to evaluate if tozasertib and structural analogues of tozasertib were indeed able
to effectively inhibit necroptosis.
Tozasertib was originally identified as a pan auroraꢀkinase (AurK) inhibitor. Since these kinases play
an important role during mitosis and meiosis it is obvious that these enzymes are integral for correct
cell proliferation. Inhibition of AurK by tozasertib subsequently leads to cellular abnormalities which
can be observed as an inhibition of cellular growth and changes in nuclear morphology and DNA
44–46
content of the cells (increased nuclear area).
However in the context of necroptosis inhibition these
effects are unfavourable and should thus be preferably removed. The second objective in the
framework of drug respurposing was thus to investigate whether it was possible to decrease or even
completely remove these cellular and nuclear effects associated with AurK inhibition by introducing
structural variations to the tozasertib scaffold. Therefore a library of tozasertib analogues with a 2,4,6ꢀ
trisubstituted pyrimidine or triazine core was synthesized.
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 100
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 1. Reported inhibitors of necroptosis: (A) Type III RIPK1 inhibitors; (B) Type II RIPK1 inhibitors; (C)
Tozasertib, a type I panꢀaurora kinase inhibitor
6
ACS Paragon Plus Environment

Page 7 of 100
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
RESULTS AND DISCUSSION
Compound design
Prior to the synthesis of the first library of inhibitors, the potential of tozasertib to inhibit TNFꢀinduced
necroptosis was validated. Next, a library of 30 compounds that are structurally similar to tozasertib
was synthesized in order to investigate whether novel analogues of tozasertib were still able to inhibit
TNFꢀinduced necroptosis in a doseꢀdependent manner. On the other hand this also helped to identify
which structural modifications were preferred to possibly increase the potency of these analogues. The
structural variations were introduced at three different positions. Firstly, the amideꢀlinked aliphatic
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
substituent at R was varied between different cyclic and aliphatic groups with divergent steric
1
properties. Secondly, the pyrazole moiety at the R
2
position was preserved but its substitution pattern
was varied with different cyclic amines.
was varied. Finally, the solubilityꢀimproving moiety at R
3
(Figure 2) These 30 compounds were tested in a phenotypical assay in which their inhibitory potential
against TNFꢀinduced necroptosis was studied. (Supporting information S1ꢀS3) The results from this
initial series allowed us to formulate a preliminary structureꢀactivity relationship (SAR) to see which
substituents in R , R and R were beneficial for improving the inhibitory potency of this type of
1
2
3
compounds. This newly found SAR was then used as a guideline in the design of a second library of
tozasertib analogues. Next to these most preferred functional groups in R , R and R , additional
1
2
3
variations were introduced to the central heterocyclic core and linker atom connecting the two sixꢀ
membered aromatic systems. (Figure 2, respectively shown as X and Y). The introduction of these
modifications provided a set of 46 final compounds 8, 43, 44, 46, 48, 50, 52, 53, 55, 57, 70ꢀ81, 89ꢀ
100, 107ꢀ118. These compounds were evaluated for their potential to inhibit mTNFꢀinduced
necroptosis in the presence of zVAD.fmk. Since tozasertib analogues can still potentially interfere
with cell proliferation and induce nuclear abnormalities, the effect of these 46 compounds on cellular
growth and nuclear area was also investigated.
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 100
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. Overview of tozasertib analogues. The results from the phenotypical assay of the first series of
inhibitors (Supplementary info S1ꢀS3) were used to narrow down the selection of substituents for the second
series of compounds. For this second series the most preferred functional groups were used and extra variations
were additionally introduced in X and Y.
Chemistry
All compounds in this study were prepared following the general strategies in Schemes 1ꢀ6. In
general, nearly all final compounds were prepared using nucleophilic aromatic substitution reactions.
Prior to the synthesis of the first library of 30 tozasertib analogues, several Nꢀ(4ꢀ
mercaptophenyl)amide intermediates 10ꢀ14 were synthesized (Scheme 1). These molecules were
synthesized from the commercially available 4ꢀaminobenzothiol 9 which was acylated using
appropriate acylchlorides to form the corresponding aromatic amides.
8
ACS Paragon Plus Environment

Page 9 of 100
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 1: Reagents and conditions: (a) 4ꢀaminobenzothiol, substituted acylchloride, triethylamine,
4
7
tetrahydrofuran, 2h, 0 °C
Treatment of 4,6ꢀdichloroꢀ2ꢀ(methylsulfonyl)pyrimidine 15 with intermediates 10ꢀ14 resulted in
compounds 16ꢀ20. These compounds were subsequently treated with 5ꢀmethylꢀ1Hꢀpyrazolꢀ3ꢀamine or
4,5ꢀdimethylꢀ1Hꢀpyrazolꢀ3ꢀamine, providing compounds 21ꢀ27. Target compounds 8 and 28ꢀ57 were
obtained by substituting the 6ꢀposition of the pyrimidine scaffold with a a piperidine, morpholine,
piperazine, Nꢀmethylꢀ or Nꢀethylpiperazine.
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 100
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Scheme 2: Reagents and conditions: (a) 4,6ꢀdichloroꢀ2ꢀ(methylsulfonyl)pyrimidine, triethylamine, acetonitrile,
2h; ꢀ10 °C to rt; (b) 5ꢀmethylꢀ1Hꢀpyrazolꢀ3ꢀamine or 4,5ꢀdimethylꢀ1Hꢀpyrazolꢀ3ꢀamine, DIPEA, DMF, 18h; 90
47–49
°C; (c) piperidine, morpholine, piperazine, Nꢀmethylpiperazine or Nꢀethylpiperazine, 2h, 110 °C;
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
In order to investigate the effects of the heteroatom that links both aromatic sixꢀmembered rings in
tozasertib, we prepared a set of various Nꢀ(4ꢀaminophenyl)cycloalkyl carboxamide intermediates
which are described in Scheme 3. These compounds were prepared by Nꢀacylation of 58 using either
cyclopropanecarbonyl chloride or cyclohexanecarbonyl chloride to afford the corresponding amide
analogues 59 and 60 respectively. Catalytic hydrogenation of the nitroꢀfunctional group resulted in
compounds 61ꢀ62.
2 2
Scheme 3: Reagents and conditions: (a) substituted acylchloride, DIPEA, DCM, 3h, rt; (b) Pd(OH) , H ,
5
0
methanol, 17h, rt
For the sake of clarity the second series can be divided into 3 smaller groups of 12 compounds, of
which the structures can be described as 2,4,6ꢀtriaminosubstituted pyrimidines, 2,6ꢀdiaminoꢀ4ꢀ
sulfosubstituted triazines and 2,4,6ꢀtriaminosubstituted triazines. The synthesis of these compounds is
discussed in Schemes 4ꢀ6 respectively.
1
0
ACS Paragon Plus Environment

Page 11 of 100
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The synthesis of 2,4,6ꢀtriaminosubstituted pyrimidines is described in Scheme 4. The pyrimidine
derivates were synthesized starting from 2,4,6ꢀtrichloropyrimidine 63. Firstly the 2ꢀposition was
substituted by treating 63 with the corresponding 5ꢀmethylꢀ1Hꢀpyrazolꢀ3ꢀamine or 4,5ꢀdimethylꢀ1Hꢀ
pyrazolꢀ3ꢀamine to give intermediates 64ꢀ65. These compounds were in turn coupled to the previously
prepared substituents 61ꢀ62 to give 66ꢀ69. Finally compounds 66ꢀ69 were treated with either Nꢀ
methylpiperazine, morpholine or piperidine using the same conditions as described earlier to yield
final compounds 70ꢀ81.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 4: Reagents and conditions: (a) 5ꢀmethylꢀ1Hꢀpyrazolꢀ3ꢀamine or 4,5ꢀdimethylꢀ1Hꢀpyrazolꢀ3ꢀamine,
5
1
52
dioxane, 12h, rt; (b) 61ꢀ62, pꢀtoluene sulfonic acid monohydrate, nꢀbutanol, 12h, reflux; (c) piperidine,
4
9
morpholine, Nꢀmethylpiperazine, 2h, 110°C
The synthesis of 2,6ꢀdiaminoꢀ4ꢀsulfosubstituted triazines is described in Scheme 5. All derivatives of
this type were synthesized from the 2,4,6ꢀtrichlorotriazine 82. Firstly, intermediates 13 and 14 were
1
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 100
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
introduced to the triazine core which resulted in the intermediates 83ꢀ84. Subsequently these
intermediates were treated with either 5ꢀmethylꢀ1Hꢀpyrazolꢀ3ꢀamine or 4,5ꢀdimethylꢀ1Hꢀpyrazolꢀ3ꢀ
amine resulting in intermediates 85ꢀ88. Once again, the Nꢀmethylpiperazine, morpholine or piperidine
moiety was introduced at the 6ꢀposition of the triazine core of compounds 85ꢀ88 which resulted in the
formation of the final compounds 89ꢀ100.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 5: Reagents and conditions: (a) Compound 13 or 14, DIPEA, 0 °C, dropwise addition (b) 5ꢀmethylꢀ1Hꢀ
pyrazolꢀ3ꢀamine, DIPEA, dioxane, 16 h, 120 °C (c) 4,5ꢀdimethylꢀ1Hꢀpyrazolꢀ3ꢀamine, DIPEA, dioxane, 16 h, rt
(d) piperidine, morpholine or Nꢀmethylpiperazine, 30 min, rt
The synthesis of 2,4,6ꢀtriaminosubstituted triazines is described in Scheme 6. The triazine derivates
were synthesized from 2,4,6ꢀtrichlorotriazine 82. In the first step intermediates 61ꢀ62 were coupled to
82 to give compounds 101ꢀ102. . Next, overnight treatment of 101ꢀ102 with the appropriate 5ꢀmethylꢀ
1Hꢀpyrazolꢀ3ꢀamine or 4,5ꢀdimethylꢀ1Hꢀpyrazolꢀ3ꢀamine yielded compounds 103ꢀ106. Treatment of
1
2
ACS Paragon Plus Environment

Page 13 of 100
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
03ꢀ106 with either Nꢀmethylpiperazine, morpholine or piperidine using conditions as described in
previous schemes afforded target triazines 107ꢀ118.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 6: Reagents and conditions: (a) 61ꢀ62, K CO , dioxane, rt; (b) 5ꢀmethylꢀ1Hꢀpyrazolꢀ3ꢀamine or 4,5ꢀ
2
3
dimethylꢀ1Hꢀpyrazolꢀ3ꢀamine, DIPEA, dioxane, 16h, 120° C; (c) piperidine, morpholine, Nꢀmethylpiperazine,
4
9
2h, 110°C
1
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 100
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Evaluation and characterization of the synthesized tozasertib analogues
Phenotypical screening of the compounds in L929 cells
The research was initiated by synthesizing a preliminary library of 30 compounds in order to
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1 2 3
investigate which substituents in R , R and R were favourable. Exact IC50ꢀvalues and assay
conditions are reported in table S1 (Supporting information S1ꢀS3). In general, it was observed that
compounds containing a cyclopropyl and cyclohexyl moiety in R showed the best potential towards
1
2
the inhibition of TNFꢀinduced necroptosis. Regarding R , both hydrogen and methyl substituents were
well tolerated, but generally the compounds containing a hydrogen substituent instead of an extra
methyl group showed to be more potent for inhibiting TNFꢀinduced necroptosis. At R it was clear that
3
compounds which contained an Nꢀmethylpiperazine, morpholine or piperidine moiety were most
potent in inhibiting TNFꢀinduced necroptosis. (Figure 3)
Figure 3. Summary of the results from the first phenotypical assay in determining the preferred substituents in
1 2 3
R , R and R .
With these preferred groups at R , R and R in place, additional structural variations were introduced
1
2
3
to the central heterocyclic core and the linker connecting the two aromatic sixꢀmembered rings. A
1
4
ACS Paragon Plus Environment

Page 15 of 100
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
second compound library was synthesized in which nearly all possible and preferred aforementioned
combinations were introduced. This library consists of 46 compounds of which 8, 43, 44, 46, 48, 50,
52, 53, 55, 57 were reꢀused from the initial series, and of which 70ꢀ81, 89ꢀ100, 107ꢀ118 were
synthesized additionally. These 46 compounds were evaluated in a phenotypical assay for their
potential to inhibit necroptosis stimulated by mTNF. Additional data on different cell lines and the
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
3
mechanism of action of selected compounds will be published elsewhere.
As mentioned earlier, treatment of cells with tozasertib results in an inhibition of TNFꢀinduced
necroptosis, but also affects cellular growth and induces nuclear abnormalities with similar IC50
ꢀ
values. With this second library, we investigated whether it was possible for a molecule to be a potent
inhibitor of TNFꢀinduced necroptosis without affecting the cellular growth and/or the nuclear area.
The results of this assay are reported in table 1. The 95% confidence interval for the active compounds
is reported in the supporting information (Table S2, S5ꢀS9).
IC50 ꢀvalues
Core Linker
(µM)
Compound
R
1
R
2
R
3
Y
X
a
b
Cell c Nuclear
ꢀ0.5h
ꢀ24h
d
growth
0.97
area
8,
Tozasertib
Reference)
CH
S
Cyclopropyl
H
NMe
0.98
1.02
1.06
(
4
3
4
CH
CH
S
S
Cyclopropyl
Cyclopropyl
H
H
CH
2
0.75
0.39
1.74
0.81
1.32
1.00
1.00
1.04
4
O
1
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 100
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
47
48
50
52
53
55
57
70
71
72
73
74
75
76
77
78
79
80
81
89
90
91
92
93
94
95
96
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
N
S
S
Cyclohexyl
Cyclohexyl
Cyclohexyl
H
H
H
CH
2
1.75
0.38
>3
5.04
0.36
>3
0.813
0.214
<0.6
>3
>3
<0.6
>3
>3
>3
>3
>3
>3
>3
>3
>3
>3
>3
>3
>3
>3
>3
>3
>3
>3
>3
0.80
>3
>3
>3
>3
>3
O
S
NMe
S
Cyclohexyl Me CH
2
>3
>3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
S
Cyclohexyl Me
O
>3
>3
>3
S
Cyclohexyl Me NMe
Cyclopropyl Me NEt
>3
>3
>3
S
>3
>3
>3
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
S
Cyclopropyl
Cyclopropyl
Cyclopropyl
H
H
H
CH
2
0.53
0.62
1.04
>3
0.68
0.43
0.64
>3
>3
O
>3
NMe
>3
Cyclopropyl Me CH
Cyclopropyl Me
Cyclopropyl Me NMe
2
>3
O
>3
>3
>3
>3
>3
>3
Cyclohexyl
Cyclohexyl
Cyclohexyl
H
H
H
CH
2
>3
1.94
1.50
1.75
>3
>3
O
0.64
0.91
>3
2.25
>3
NMe
Cyclohexyl Me CH
Cyclohexyl Me
Cyclohexyl Me NMe
2
>3
O
>3
>3
>3
>3
>3
>3
Cyclopropyl
Cyclopropyl
Cyclopropyl
H
H
H
CH
O
>3
>3
>3
2
N
S
0.70
>3
1.45
1.14
>3
>3
N
S
NMe
0.44
>3
N
S
Cyclopropyl Me CH
Cyclopropyl Me
Cyclopropyl Me NMe
>3
2
N
S
O
>3
>3
>3
N
S
>3
>3
>3
N
S
Cyclohexyl
Cyclohexyl
H
H
CH
O
>3
>3
>3
2
N
S
>3
1.63
2.17
1
6
ACS Paragon Plus Environment

Page 17 of 100
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
9
9
9
7
8
9
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
S
Cyclohexyl
Cyclohexyl Me CH
Cyclohexyl Me
Cyclohexyl Me NMe
H
NMe
>3
>3
>3
>3
>3
1.06
>3
>3
>3
>3
>3
>3
>3
>3
>3
>3
>3
>3
>3
>3
>3
2.64
>3
>3
>3
>3
>3
>3
>3
>3
>3
>3
0.33
>3
>3
>3
>3
>3
>3
>3
>3
>3
>3
2.30
>3
>3
>3
>3
0.40
>3
>3
>3
>3
>3
>3
>3
>3
>3
>3
>3
>3
>3
>3
>3
S
2
S
O
100
107
108
109
110
111
112
113
114
115
116
117
118
S
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
Cyclopropyl
Cyclopropyl
Cyclopropyl
H
H
H
CH
2
O
NMe
Cyclopropyl Me CH
Cyclopropyl Me
Cyclopropyl Me NMe
2
O
Cyclohexyl
Cyclohexyl
Cyclohexyl
H
H
H
CH
2
O
NMe
Cyclohexyl Me CH
Cyclohexyl Me
Cyclohexyl Me NMe
2
O
Table 1. Evaluation of the antiꢀnecroptotic activity and observation of possible effects on cellular growth and
nuclear area of the synthesized tozasertib analogues. Compound is classified as inactive when less than 50%
inhibition is observed at a concentration of 3 ꢁM of inhibitor. The IC50ꢀvalue is then shown as >3 ꢁM in table
4
.1.
a
Inhibition of mTNFꢀinduced necroptosis in the presence of panꢀcaspase inhibitor zVAD.fmk. L929sAhFAS
cells have been preꢀtreated with compound for 0.5h followed by stimulation with mTNF (2500 IU/ml) and
zVAD.fmk (1 µM) for 3h in the presence of Hoechst (1 µM) and propidium iodide (PI) (3 µM). IC50ꢀvalues were
calculated through a doseꢀresponse curve with the percentage of cell death as quantitative readꢀout.
b
Inhibition of mTNFꢀinduced necroptosis in the presence of panꢀcaspase inhibitor zVAD.fmk. L929sAhFAS
cells have been preꢀtreated with compound for 24h followed by stimulation with mTNF (2500 IU/ml) and
zVAD.fmk p(1 µM) for 3h in the presence of Hoechst (1 µM) and propidium iodide (PI) (3 µM). IC50ꢀvalues
were calculated through a doseꢀresponse curve with the percentage of cell death as quantitative readꢀout.
1
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 100
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
c
Value is calculated via doseꢀresponse curve of the concentration of the individual compound in function of the
cellular growth after 24h. For curve fitting and IC50 determination, the bottom value equals the total amount of
cells in the highest concentration of tozasertib and the top value equals the total amount of cells in untreated
conditions.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
d
Value is calculated via doseꢀresponse curve of the concentration of the individual compound in function of the
mean nuclear area. For curve fitting and IC50ꢀdetermination, this results in a sigmoidal curve of which the bottom
value equals the normal nuclear area and the top value equals the maximal nuclear area at the highest
concentration of tozasertib.
Because tozasertib was initially identified as a panꢀAurK inhibitor to treat multiple types of leukemia,
it is to be expected that novel analogues that are derived from tozasertib can still display similar
4
8,49
properties.
As can be viewed in table 1, tozasertib clearly reduced cellular growth (IC50, growth = 0.97
µM) and also interfered with normal cell division which can be verified by morphological analysis of
the nuclei of the cells (IC50, nuclear area = 1.06 µM). Tozasertib is able to inhibit necroptosis in the low
micromolar range (IC50, ꢀ0.5h = 0.98 µM and IC50, ꢀ24h = 1.02 µM ). However due to its effects on cell
division, tozasertib can be classified as an aselective compound for the inhibition of necroptosis. In
order to meet the aforementioned goals, it was important that tozasertib analogues were free from
these undesirable effects on cell growth and nuclear area whilst still maintaining a potent inhibition of
TNFꢀinduced necroptosis.
Out of the 46 compounds that were tested under these conditions, seven compounds (70, 71, 72, 77,
7
8, 90 and 108) had the potential to inhibit TNFꢀinduced necroptosis without affecting cell growth
and/or nuclear area. Out of these seven compounds 70ꢀ72 were able to inhibit necroptosis in a slightly
more potent manner than tozasertib after both 0.5h and 24h preꢀtreatment. Compounds 77, 78, 90 and
1
08 also provided the desired inhibitory profile albeit in a slightly less significant manner than
compounds 70ꢀ72. In addition to compounds that favour necroptosis inhibition without affecting cell
growth and/or nuclear area, there were also compounds with the opposite effect. These compounds did
not inhibit TNFꢀinduced necroptosis, but did inhibit cellular growth and nuclear area in a very similar
1
8
ACS Paragon Plus Environment

Page 19 of 100
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
manner as tozasertib. This is most clearly exemplified by compound 97 which affects cell growth and
nuclear area in a more potent fashion than tozasertib.
The results of this assay led to another interesting remark that can be made towards the nature of the
linker atom X between the two aromatic sixꢀmembered rings. In compounds 70ꢀ72, which were
deemed the most interesting compounds regarding both potency and absence of nuclear abnormalities,
the sulfur linker connecting the two sixꢀmembered aromatic systems was replaced by a nitrogen linker.
In general, most compounds in which the sulfur linker was replaced by a nitrogen no longer affect
cellular growth and nuclear area. This strongly suggests that the choice of the linker atom could be
important for diminishing the unwanted cellular effects that are classically associated with AurK
inhibition by tozasertib. Overall a trend was observed that compounds containing a central triazine
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
core, a 4,5ꢀdimethylated pyrazole moiety at R
growth and nuclear area. This was clearly exemplified by compounds 73ꢀ75, 79ꢀ81, 92ꢀ94, and 110ꢀ
12. It should be noted however that these compounds do no longer possess any activity in the
2
or the combination of both, no longer affected cell
1
phenotypical assays.
In order to validate the findings of the phenotypical assay, the most favourable necroptosis inhibitors
70ꢀ72 were first evaluated in an enzymatic assay against recombinant hAurK A and B, and hRIPK1.
Compound 97 was also included in this assay since it showed a strong potential for induction of
nuclear abnormalities without showing any potency to inhibit TNFꢀinduced necroptosis. In this
enzymatic assay some interesting observations were seen regarding the nature of the linker atom
between the two sixꢀmembered aromatic rings. In order to investigate this further, the structural effects
of the substitution of a sulfur to a nitrogen linker were studied through a molecular modelling study.
Evaluation of the compounds against the recombinant enzymes
In addition to compounds 70ꢀ72 and 97, Necꢀ1s (2) and tozasertib (8) were also included as reference
compounds. (Table 2)
1
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 100
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
In this enzymatic assay Necꢀ1s (2) potently inhibited hRIPK1 but did not affect the hAurK enzymes.
Tozasertib (8) on the other hand was shown to be a potent inhibitor of all the enzymes included in this
assay.
In this experiment compounds 70ꢀ72 were able to inhibit hRIPK1 with low micromolar IC50ꢀvalues.
These compounds are even 2ꢀ to 4ꢀfold more potent for inhibition of hRIPK1 than Necꢀ1s (2).
Tozasertib (8) was also shown to be a very potent inhibitor of both hAurK A and hAurK B. It should
be noted that the most potent compounds 70ꢀ72 still inhibited both hAurK enzymes, nevertheless in a
significantly less potent manner than tozasertib. Compound 97, which did not demonstrate any
potency towards the inhibition of TNFꢀinduced necroptosis in the aforementioned phenotypic assay,
showed to be a significantly less potent inhibitor for hRIPK1 in the enzymatic assay. Compound 97
still maintained its inhibition of the hAurK enzymes, but in a slightly more potent manner in
comparison to compound 70ꢀ72. In general it can be concluded that a correlation is observed between
the results of both the phenotypic and the enzymatic assay.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
hAurK A
µM)
hAurK B
(µM)
>3
hRIPK1
(µM)
Compound
(
Necꢀ1s (2)
>3
0.754
0.208
0.295
0.167
0.178
2.468
Tozasertib (8)
0.030
0.310
0.430
0.293
0.210
0.068
0.959
0.853
1.345
0.229
70
71
72
97
Table 2. IC50ꢀvalues from the enzymatic assay. Compounds were evaluated for their potency to inhibit
recombinant hAurK A, hAurK B and hRIPK1. The results were obtained through a luminescenceꢀbased ADPꢀ
Glo kinase
™ assay. When a compound was inactive the IC ꢀvalue is shown as >3 ꢁM. The confidence intervals
50
and statistical validation of the results presented in table 2 are reported in the supplementary information table
5
3
S3.
2
0
ACS Paragon Plus Environment

Page 21 of 100
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Possible effects of the nature of the linking atom X
Inspection of the tozasertib:aurora kinase A:TPX2 cocrystal structure (PDB 3E5A) reveals that the
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
linker atom in tozasertib (8) is rather solventꢀexposed and does not engage in highly specific
54
interactions with the protein. It thus seems unlikely that contacts with the protein are responsible for
the difference in assay outcomes. However, the linker atoms do differ substantially in electronic
properties. The nitrogen linker is more strongly conjugated to the aromatic ring systems, which may
increase the rigidity of the general structure and influence the conformational preferences of the
system. In order to investigate this effect, two model systems were designed: Nꢀphenylpyrimidinꢀ2ꢀ
amine and 2ꢀ(phenylthio)pyrimidine. The two torsion angles around the nitrogen or sulfur linker atoms
controlling the orientation of the two rings are examined through relaxed dihedral scans at the
5
5–57
58,59
60
PW6B95ꢀD3(BJ)
/def2ꢀTVZP
level of theory in ORCA. For the first dihedral, depicted in
figure 4 (upper image), both profiles are similar in shape. Mimima are located at 0° (or 360°) and
180°. In the tozasertib:aurora kinase A:TPX2 cocrystal structure, this dihedral is approximately 336°.
As this is quite close to a minimum, we do not anticipate this dihedral to be responsible for the
differences observed in the cellular assay. For the second dihedral, illustrated in figure 4 (lower
image), marked differences between the energy profiles are observed. For the sulfurꢀlinked system,
mimima are located at 90° and 270°. The angle in the crystal structure is 98°, corresponding to a
minimum. In the most stable conformation of the sulfurꢀlinked system, the two aromatic rings are
approximately peripendicular to one another. In the nitrogenꢀlinked system, energetic minima are
located at 0° (and 360°) and 180° degrees. The 98° torsion angle in the crystal structure of tozasertib
(
8) is highly unfavourable for this system. In contrast to the sulfurꢀlinked system, the most stable
conformation of the nitrogenꢀlinked system is a fully planar conformation. We thus propose that the
nitrogenꢀlinked compounds are not able to bind in the same conformation as tozasertib (8), due to the
difference in their respective conformational preferences. This may explain the lower affinity of the
nitrogenꢀlinked analogs for AurK A.
2
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 100
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4: DFT relaxed surface scan of both dihedrals. (Upper image) Relaxed surface scan for the first dihedral.
(Lower Image) Relaxed surface scan for the second dihedral. The energies for the nitrogenꢀlinked model
compound are shown in blue, the energies for the sulfurꢀlinked model compound are shown in orange. The arrow
denotes the dihedral angle being scanned. Dihedral angle values corresponding to the crystal structure are
marked by a dotted line.
2
2
ACS Paragon Plus Environment

Page 23 of 100
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Investigation of the binding mode of tozasertib in RIPK1
A docking experiment was performed to propose a possible binding mode for tozasertib in RIPK1
(
Figure 5). Tozasertib was docked into a cocrystal structure 4NEU of RIPK1 complexed with a 1ꢀ
35
aminoisoquinoline inhibitor 4NEU. The proposed binding mode is very similar to the binding mode
observed in the tozasertib:AurA cocrystal structure. The three hydrogen bonds to the hinge motif are
conserved and the phenyl ring interacts with the Gꢀloop.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 5. Predicted binding mode of tozasertib on RIPK1. The protein is shown in grey cartoon representation.
Key hinge backbone residues are shown in stick representation. The ligand is shown in yellow ballꢀandꢀstick
representation. Key hydrogen bonds with the hinge are denoted with a yellow dashed line.
2
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 100
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
In vivo investigation of tozasertib and selected analogues in a TNFꢀinduced SIRS model
Tozasertib 8 and compounds 70ꢀ72 and 97 were tested in the TNFꢀinduced systemic inflammatory
response syndrome (SIRS) model in order to evaluate their potential to protect against RIPK kinaseꢀ
27,61–63
driven inflammation in vivo, a model that relies on RIPK1 and RIPK3 kinase activity.
Based on
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
the in vitro results (both phenotypical and enzymatic assays), compounds 70ꢀ72 were selected as more
TNFꢀinduced necroptosis specific inhibitors, as well as compound 97 that favours AurK over
necroptosis inhibition. This experiment was conducted to investigate whether the most interesting
tozasertib analogues 70ꢀ72 have an improved profile over tozasertib for inhibition of TNFꢀinduced
necroptosis in vivo. All compounds were administered by oral gavage (50 mg/kg) 1.5h before mTNF
treatment (i.v., 500 µg/kg). Tozasertib protected mice significantly from hypothermia and death
caused by TNF (p=0.0035 compared to vehicle, p= 0.0014 compared to 97), though in a less potent
manner than Necꢀ1s. (Figure 6) The mice preꢀtreated with compounds 71 and 72, survived the TNFꢀ
treatment better than those treated with tozasertib (p72=0.0133 and p71=0.0199 compared to vehicle
and p72=0.0195 and p71=0.0174 compared to 97). Suprisingly, compound 70 did not exert any
protective effect.
2
4
ACS Paragon Plus Environment

Page 25 of 100
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 6. Tozasertib and its analogues 71 and 72 protect against TNFꢀinduced SIRS. Mice were challenged with
10 µg mTNF (500 µg/kg) in the presence or absence of tozasertib/analogue. All compounds were given by oral
gavage 1.5h before mTNF challenge at a dose of 50 mg/kg. MantelꢀCox test was performed as statistical analysis
on the survival curves. Pꢀvalues <0.05 = *, Pꢀvalues <0.005 = **, Pꢀvalues <0.0005 = **** (compared to
vehicle). Temperature curves represent mean and standard error of the mean. Data from two independent
experiments are shown.
2
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 100
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
CONCLUSION
Tozasertib, a panꢀaurora kinase inhibitor, was originally developed to treat multiple forms of
4
9
leukemia. Due to AurK inhibition tozasertib induces nuclear abnormalities like inhibition of cell
proliferation and increased nuclear area. In a kinase selectivity screen tozasertib was reported to have a
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
43
high affinity for RIPK1 (K
d
= 20 nM.). In the framework of drug repurposing we started this research
with the confirmation that tozasertib indeed potently inhibited TNFꢀinduced necroptosis in a doseꢀ
dependent manner using a cellular assay. This prompted the design and synthesis of a set of 30
compounds with very strong structural similarities to tozasertib to investigate the potential of this
scaffold for the design of novel inhibitors of TNFꢀinduced necroptosis. Different variations at multiple
positions were introduced in order to investigate which structural modifications of the original
tozasertib scaffold were well tolerated. Upon analysis of the results of this first set of molecules we
were able to formulate a preliminary SAR that was used as a base for a followꢀup series. After the
most desirable substituents were identified, a second library of 46 tozasertib analogues was
synthesized. These 46 novel compounds were once again tested in a phenotypical screen using mTNF
as a stimulus of necroptotic cell death whilst preventing the cells from undergoing apoptosis by using
the panꢀcapase inhibitor zVAD.fmk. Whilst compounds 70ꢀ72 were considered to be good inhibitors
for TNFꢀinduced necroptosis, they did not result in visible malfunctions in normal cell division which
is a characteristic for treatment with AurKꢀinhibitors such as tozasertib. In order to validate these
findings, compounds 70ꢀ72 were tested in an enzymatic assay against recombinant hAurK A, hAurK
B and hRIPK1. Necꢀ1s (2) and tozasertib (8) were used as reference molecules. In this assay
compounds 70ꢀ72 were able to inhibit recombinant hRIPK1 in the low micromolar range. These
compounds were even shown to be 2ꢀ to 4ꢀfold more potent for inhibition of hRIPK1 than Necꢀ1s (2),
a compound extensively used in necrotic cell death research. Another interesting feature of compounds
70ꢀ72 was that they inhibited both hAurK in a significantly less potent manner than the reference
compound tozasertib (8). For compounds 70ꢀ72 the linker atom between the two sixꢀmembered
aromatic rings was modified from a sulfur to a nitrogen. Molecular modelling studies suggested that
nitrogenꢀlinked compounds are not able to bind with AurK in the same conformation as tozasertib (8)
2
6
ACS Paragon Plus Environment

Page 27 of 100
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
due to the difference in their respective conformational preferences. In order to get a better insight of
the binding mode of tozasertib, it was docked into a cocrsytal structure of RIPK1 complexed with a 1ꢀ
aminoisoquinoline inhibitor. This showed that three hydrogen bonds to the hinge motif are conserved
and the phenyl ring interacts with the Gꢀloop. Finally, the in vivo effect of tozasertib and four
tozasertib analogues (70ꢀ72 and 97) was studied in a TNFꢀinduced SIRS model. In this study,
compounds 71 and 72 had an increased survival to TNFꢀtreatment compared to tozasertib.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Concludingly, we were able to report for the first time that tozasertib analogues can potently inhibit
TNFꢀinduced necroptosis. By applying some structural modifications to the tozasertib scaffold, we
were able to fineꢀtune the inhibitory profile of this type of compound in order to reduce the offꢀtarget
effects that are commonly associated with AurK inhibition. The more selective compounds 71 and 72
were able to inhibit TNFꢀinduced necroptosis in vitro and in vivo in a similar or even more potent
manner than tozasertib (8).
2
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 100
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
EXPERIMENTAL SECTION
Unless otherwise stated, laboratory reagent grade solvents were used. Reagents were obtained from
SigmaꢀAldrich, Acros Organics or Fluorochem and were used without further purification.
1
13
1
13
Characterization of all compounds was done with H and C NMR and mass spectrometry. H and C
NMR spectra were recorded on a 400 MHz Bruker Avance III Nanobay spectrometer with Ultrashield
and analysed by use of MestReNova analytical chemistry software. Chemical shifts are in ppm, and
coupling constants are in hertz (Hz). ES mass spectra were obtained from an Esquire 3000plus ion trap
mass spectrometer from Bruker Daltonics. Purities were determined with two diverse HPLC systems
based either on mass determination or on UV detection. A Waters acquity UPLC system coupled to a
Waters TQD ESI mass spectrometer (HPLC System A) or a Waters SQD ESI mass spectrometer
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
HPLC system B) was used both in combination with a Waters TUV detector. The same methods on
both HPLC system A and B was used for compound detection and purity determination. Water (A)
and CH CN (B) were used as eluents. Waters Acquity UPLC BEH C18 1.7 ꢁm, 2.1 mm × 50 mm
3
column was used. Solvent A consisted of water with 0.1% formic acid. Solvent B consisted of
acetonitrile with 0.1% formic acid. Method I involved the following: 0.15 min 95% A, 5% B, then in
1.85 min from 95% A, 5% B to 95% B, 5% A, then 0.25 min (0.350 mL/min), 95% B, 5% A. The
wavelength for UV detection was 254 nm. Method II involved the following: flow 0.4 mL/min, 0.25
min 95% A, 5% B, then in 4.75 min to 95% B, 5% A, then 0.25 min 95% B, 5% A, followed by 0.75
min 95% A, 5% B. The wavelength for UV detection was 214 nm. Where necessary, flash purification
was performed on a Biotage ISOLERA One flash system equipped with an internal variable
dualwavelength diode array detector (200−400 nm). For normal phase purifications SNAP cartridges
(
10−340 g, flow rate of 10−100 mL/min) were used, and reversed phase purifications were done
making use of KPꢀC18 containing cartridges. Dry sample loading was done by selfꢀpacking samplet
cartridges using silica and Celite 545, respectively, for normal and reversed phase purifications.
Gradients used varied for each purification. Mouse and rat plasma came from Innovative Research.
The turbidity in the kinetic solubility experiments was measured using the UV/vis spectrophotometer
Synergy MX, Biotek with Gen5.
2
8
ACS Paragon Plus Environment

Page 29 of 100
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The following section comprises the synthetic procedures and analytical data for all compounds
reported in this manuscript. Several synthesis procedures that were used in the preparation of
intermediates and final products are summarized here as “General Procedures”. The purities of all final
products were found to be > 95%.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
General procedure A for compounds 10ꢀ14
To a solution of 4ꢀaminobenzothiol (1 equiv.) in THF was added triethylamine (2.2 equiv). After
cooling down the mixture to 0°C, the appropriate acyl chloride (2.2 equiv.) was added dropwise in
order to keep the temperature below 10°C. The reaction mixture was stirred at 0°C for 20 minutes then
warmed up to room temperature for 1 hour. The obtained solid was filtered off and the filtrate was
concentrated under reduced pressure. The concentrated residue was treated with NaOH (3 equiv.)
which was first dissolved in a mixture of 6:1 water:ethanol. The reaction mixture was heated to 100°C
for 1 hour and cooled down to room temperature. The crude was filtered and concentrated under
reduced pressure. The residue was diluted with water and filtered through a path of celite. The
obtained filtrate was acidified with concentrated HCl upon reaching a pH of 3 and a white precipitate
was obtained. The resulting precipitate was filtered, dissolved in ethyl acetate and washed with brine.
4
The organic phase was dried over MgSO and concentrated under reduced pressure to the appropriate
amidobenzothiol in good yields.
General procedure B for compounds 16ꢀ20
A solution of 4,6ꢀdichloroꢀ2ꢀ(methylsulfonyl)pyrimidine (1 equiv.) and the appropriate Nꢀ(4ꢀ
mercaptophenyl)amide analogue (1.05 equiv.) in ACN was cooled down to ꢀ10°C. Next triethylamine
(
1 equiv.) was added dropwise over 20 minutes while maintaining the temperature at ꢀ10°C. Once
added, the solution was stirred at that temperature for 20 more minutes. Afterwards the reaction
mixture was allowed to warm up to room temperature while stirring and concentrated. Water was
added to the crude and a precipitate was formed. This solid was collected by filtration and dried by
2
9
ACS Paragon Plus Environment