N?-Acryloyllysine Piperazides as Irreversible Inhibitors of Transglutaminase 2: Synthesis, Structure-Activity Relationships, and Pharmacokinetic Profiling

Article abstract of DOI:10.1021/acs.jmedchem.8b00286

Transglutaminase 2 (TGase 2)-catalyzed transamidation represents an important post-translational mechanism for protein modification with implications in physiological and pathophysiological conditions, including fibrotic and neoplastic processes. Consequently, this enzyme is considered a promising target for the diagnosis of and therapy for these diseases. In this study, we report on the synthesis and kinetic characterization of N^?-acryloyllysine piperazides as irreversible inhibitors of TGase 2. Systematic structural modifications on 54 new compounds were performed with a major focus on fluorine-bearing substituents due to the potential of such compounds to serve as radiotracer candidates for positron emission tomography. The determined inhibitory activities ranged from 100 to 10?000 M^-1 s^-1, which resulted in comprehensive structure-activity relationships. Structure-activity correlations using various substituent parameters accompanied by covalent docking studies provide an advanced understanding of the molecular recognition for this inhibitor class within the active site of TGase 2. Selectivity profiling of selected compounds for other transglutaminases demonstrated an excellent selectivity toward transglutaminase 2. Furthermore, an initial pharmacokinetic profiling of selected inhibitors was performed, including the assessment of potential membrane permeability and liver microsomal stability.

Full text of DOI:10.1021/acs.jmedchem.8b00286

Subscriber access provided by UNIV OF NEW ENGLAND ARMIDALE
#
N-Acryloyllysine piperazides as irreversible inhibitors of transglutaminase
2
– synthesis, structure-activity relationships and pharmacokinetic profiling
Robert Wodtke, Christoph Hauser, Gloria Ruiz-Gomez, Elisabeth Jäckel, David Bauer, Martin
Lohse, Alan Wong, Johanna Pufe, Friedrich-Alexander Ludwig, Steffen Fischer, Sandra
Hauser, Dieter Greif, Maria Teresa Pisabarro, Jens Pietzsch, Markus Pietsch, and Reik Löser
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b00286 • Publication Date (Web): 17 Apr 2018
Downloaded from http://pubs.acs.org on April 17, 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.
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 112
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
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 112
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
ε
N ꢀAcryloyllysine piperazides as irreversible inhibitors of transglutaminase 2 –
‡
synthesis, structureꢀactivity relationships and pharmacokinetic profiling
a,e,f
b
c
a,e
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
Robert Wodtke , Christoph Hauser , Gloria RuizꢀGómez , Elisabeth Jäckel ,
a,f
a,e
a
a
David Bauer , Martin Lohse , Alan Wong , Johanna Pufe ,
d
d
a
e
FriedrichꢀAlexander Ludwig , Steffen Fischer , Sandra Hauser , Dieter Greif ,
c
a,f
b,*
a,f,*
M. Teresa Pisabarro , Jens Pietzsch , Markus Pietsch , Reik Löser
[
a] HelmholtzꢀZentrum DresdenꢀRossendorf, Institut für Radiopharmazeutische Krebsforschung,
Bautzner Landstraße 400, 01328 Dresden, Germany
b] Zentrum für Pharmakologie, Medizinische Fakultät, Universität zu Köln, Gleueler Straße 24, 50931
Köln, Germany
[
[
c] Structural Bioinformatics, BIOTEC, TU Dresden, Tatzberg 47ꢀ51, 01307 Dresden, Germany
d] HelmholtzꢀZentrum DresdenꢀRossendorf, Institut für Radiopharmazeutische Krebsforschung,
[
Forschungsstelle Leipzig, Permoserstraße 15, 04318 Leipzig, Germany
e] Fakultät Naturꢀ und Umweltwissenschaften, Hochschule Zittau/Görlitz, TheodorꢀKörnerꢀAllee 16,
2763 Zittau, Germany
f] Fakultät Chemie und Lebensmittelchemie, Technische Universität Dresden, Mommsenstraße 4,
[
0
[
0
1062 Dresden, Germany
‡
th
th
Dedicated to Profs. Drs. Michael Gütschow and Jörg Steinbach, on the occasion of their 60 and 65
birthdays, respectively, in gratefulness for their past and continuing support.
ACS Paragon Plus Environment

Page 3 of 112
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Abstract:
Transglutaminase 2 (TGase 2)ꢀcatalysed transamidation represents an important
posttranslational mechanism for protein modification with implications in physiological and
pathophysiological conditions including fibrotic and neoplastic processes. Consequently, this
enzyme is considered a promising target for the diagnosis and therapy of these diseases. In
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
ε
this study, we report on the synthesis and kinetic characterisation of N ꢀacryloyllysine
piperazides as irreversible inhibitors of TGase 2. Systematic structural modifications on 54
new compounds were performed with a major focus on fluorineꢀbearing substituents due to
the potential of such compounds to serve as radiotracer candidates for positron emission
ꢀ1 ꢀ1
tomography. The determined inhibitory activities ranged from 100ꢀ10 000 M s , which
resulted in comprehensive structureꢀactivity relationships. Structureꢀactivity correlations
using various substituent parameters accompanied by covalent docking studies provide an
advanced understanding of the molecular recognition for this inhibitor class within the active
site of TGase 2. Selectivity profiling of selected compounds for other transglutaminases
demonstrated an excellent selectivity towards transglutaminase 2. Furthermore, an initial
pharmacokinetic profiling of selected inhibitors was performed including the assessment of
potential membrane permeability and liver microsomal stability.
2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 112
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
Introduction
1
Transglutaminases have been discovered 60 years ago in guinea pig liver. Even though
initially identified in mammals, their occurrence has been confirmed for all kingdoms of
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
ꢀ4
eukaryotic organisms including plants and fungi. In addition, transglutaminases are found
in bacteria, which indicates the fundamental importance of these enzymes for biological
processes.
The human transglutaminases constitute a family of nine homologues, which comprises the
transglutaminases 1ꢀ7, the blood coagulation factor XIIIa and the catalytically inactive
erythrocyte protein band 4.2. All human family members represent multidomain proteins
consisting of a βꢀsandwich, the central α/β transamidase domain which harbours the acyl
transferase active site and exhibits similarity to papainꢀlike cysteine proteases, and two Cꢀ
5
terminal βꢀbarrels.
Transglutaminase 2 (TGase 2), which is also referred to as tissue transglutaminase,
represents the most intensively studied family member. Its acyltransferase site is constituted
by the residues Cys277, His335, Asp358 and Trp241 (numbering according to the human
enzyme), which directly participate in catalysis. The central α/β transamidase domain
catalyses the γꢀglutamyl transfer between proteinꢀbound glutamine residues and various
nucelophiles comprising a variety of primary amines (Scheme 1) and also water under
certain conditions. In particular, proteinꢀbound lysine residues, lowꢀmolecular weight
polyamines, e.g. cadaverine, spermidine and spermine, and biogenic monoamines can
comprise the latter category of substrates. Beside its eponymous acyl transferase activity,
TGase 2 can bind to fibronectin and exhibits GTPase activity, functions which have been
ascribed to the Nꢀterminal βꢀsandwich and the Cꢀterminal βꢀbarrel domains, respectively.
Due to its multidomain structure the enzyme exists in two distinguished conformational
3
ACS Paragon Plus Environment

Page 5 of 112
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
states: a stretched open one and a compact closed conformation. In the closed state, which
is induced and stabilised by GTP, the Cꢀterminal βꢀbarrels block the access to the active site
of the transamidase domain. The open conformation, in which TGase 2 is fully active with
2
+
6ꢀ7
regards to its acyl transferase activity, is induced and stabilised by Ca ions.
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. Transamidation reaction catalysed by TGase 2
TGase 2 is ubiquitously expressed and is mainly located in the cytoplasm under
physiological conditions. This implies that under these conditions the enzyme is mainly
2
+
inactive due to the high levels of GTP and GDP and the low concentration of Ca ions in the
cytosol. Under certain circumstances, TGase 2 can occur in other cell organelles such as the
nucleus and mitochondria and can be released to the extracellular milieu. Despite higher
2
+
levels of Ca and low concentrations of GTP and GDP, the enzyme is largely inactive in that
latter compartment as well. Activation can occur under the influence of certain pathogenic
8
9ꢀ10
stimuli, such as local injury, proꢀinflammatory mediators or hypoxia.
Extracellular
inactivity of the enzyme has been mainly attributed to the formation of a vicinal disulphide
bridge between Cys370 and Cys371, which renders TGase 2 inactive regarding acyl transfer
catalysis. Thioredoxinꢀcatalysed reduction of this disulphide bond can activate TGase 2,
11
which is stimulated by inflammatory signals such as the cytokine interferonꢀγ.
Even though strictly regulated and largely supressed under physiological conditions,
increased acyl transferase activity by TGase 2 has been shown to contribute to the
4
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 112
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
2ꢀ14
pathogenesis of several diseases such as such as celiac disease,
neurodegenerative
While the extent of
15ꢀ16
17ꢀ20
21ꢀ23
disorders,
diseases related to fibrotic processes,
and cancer.
17
γꢀglutamyl transfer catalysis clearly correlates to the disease state for kidney fibrosis, the
function of TGase 2 in neoplastic diseases is less clear. However, it becomes more and
more apparent that TGase 2 is a key player for the progression of several kinds of cancer,
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
24
25
26
including breast carcinoma , renal cell carcinoma and malignant melanoma . In this
context, the elevated levels of TGase 2 expression are directly correlated with poor
prognostic indicators for survival, e.g. metastatic phenotype and drug resistance of the
cancer cells. Moreover, recent studies identified TGase 2 as a survival factor for cancer stem
27ꢀ31
cells from different tumour entities,
even though its detailed function at the molecular
level is not completely understood. Direct evidence for the involvement of extracellular
TGase 2ꢀcatalysed acyl transfer in tumour progression has been obtained by studying the
enzyme’s function in pancreatic ductal adenocarcinoma. According to this work, the
expression of TGase 2 is highly increased in the malignant pancreatic tissue and
ε
immunohistochemical staining for the enzyme is accompanied by N ꢀ(γꢀglutamyl)ꢀlysine
isopeptide crossꢀlinks in the extracellular matrix and basement membrane. Furthermore,
investigations at the cellular level have demonstrated the efficient crosslinking of collagen by
TGase 2 secreted from cancer cells, which stimulates fibroblasts for further matrix
deposition. In turn, proliferation of tumour cells through mechanical stimuli is elicited. In line
with these results, siRNAꢀmediated knockdown of TGase 2 led to the attenuation of tumour
3
2
growth in an orthotopic xenograft mouse model of pancreatic cancer. Similar results were
3
3
obtained for the invasive ductal carcinomaꢀtype of breast cancer. Besides extracellular
crosslinking, the posttranslational modification of potentially oncogenic intracellular proteins,
3
4ꢀ35
36
37
such as IκBα,
phospholipase A2
and aconitase 2,
by TGase 2ꢀmediated
transamidation is discussed in the context of tumour progression. Moreover, the release of
ammonia through TGase 2ꢀcatalysed hydrolytic deamidation of proteinꢀbound glutamine
5
ACS Paragon Plus Environment

Page 7 of 112
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
residues was identified as a potential mechanism for counteracting the acidification of the
38
tumour microenvironment associated with increased aerobic glycolysis of tumour cells.
Further tumourꢀpromoting functions of TGase 2 independent of its acyl transferase activity
39
have been attributed to its capability of acting as intracellular GTPꢀbinding protein and its
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
40
role as extracellular adapter protein, both of which involve the closed conformation of the
7
enzyme.
In consequence of its obvious involvement in the progression of different tumours and further
pathogenic processes, TGase 2 represents an interesting target for pharmacological
inhibition to treat these diseases. Furthermore, considering the remaining open questions
with regards to its detailed functions – particularly in cancer – compounds that can be
employed as molecular probes for imaging this enzyme in vivo are highly desirable and their
41ꢀ42
development receives increasing interest.
In addition, appropriate imaging agents which
allow for quantitative assessment of target occupancy and dose selection will support the
clinical translation of inhibitors. Based on its quantitative character and its sensitivity,
positronꢀemission tomography (PET) is often considered as the most advantageous imaging
43ꢀ47
modality for such purposes.
18
Based on the motivation to develop Fꢀlabelled radiotracers for PET studies, the present
study focusses on the synthesis and kinetic characterisation of fluorinated irreversible
inhibitors for targeting TGase 2. Irreversible inhibitors ranging from small molecules to
peptidic compounds with different electrophilic warheads, which form a covalent bond with
the active cysteine residue, represent the most intensively studied class of TGase 2
48ꢀ49
inhibitors.
Radiotracers derived from such irreversibly acting inhibitors appear to be
favourable for imaging purposes due to the persistent labelling of the target tissue, which
potentially results in a high signal/noise ratio as demonstrated for PET tracers targeting
50ꢀ51
monoamine oxidase B and fatty acid amide hydrolase.
Although directed against the α/β
transamidase domain, irreversible TGase 2 inhibitors have been shown to attenuate tumour
6
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 112
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
5
2ꢀ53
proliferation even when acyl transferaseꢀindependent functions are mainly involved.
This
finding is attributed to the presence of a dynamic equilibrium between the open and closed
54
conformations of TGase 2.
An important subclass of irreversible inhibitors is represented by Michael acceptors, which
react with the active cysteine residue according to a 1,4ꢀaddition. In this context, Signorini et
al. observed the inhibition of the transamidase activity of TGase 2 from erythrocytes by
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
5
acrylamide. Later, Marrano et al. incorporated the acrylamide moiety as electrophilic
ε
warhead in different diaminocarboxylic acid derivatives and identified N ꢀacryloyllysines as
5
6ꢀ57
potent inhibitors of TGase 2.
Based on these results, Wityak et al. investigated structureꢀactivity relationships (SARs),
which led to the identification of the phenylacetyl and arylpiperazine moieties as favourable
5
8
substituents at the αꢀamino and carboxyl group of lysine, respectively. The most potent
compound of this study is compound 6a, which exhibits an excellent selectivity for TGase 2
within the transglutaminase family. Furthermore, this compound is amenable to structural
modifications with fluorine, particularly via its phenylacetyl and pyridylpiperazine moieties
(
see Figure 1). Therefore, 6a was chosen as lead compound for the development of potential
fluorinated radiotracers in the present study.
7
ACS Paragon Plus Environment

Page 9 of 112
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 1. Course of development within acrylamideꢀbased TGase 2 inhibitors and
selected fluorinated derivatives envisaged in this study
In addition to fluorination, further structural modifications were considered to uncover novel
SARs. On the one hand, this will support radiotracer design by combining reactivityꢀ
conferring and fluorinated moieties. On the other hand, SAR studies supported by in silico
molecular docking could give insights into the molecular basis for recognition of these
ε
inhibitors. To assess the inhibitory potential of the synthesised N ꢀacryloyllysines, two
5
9ꢀ60
TGase 2 activity assays recently developed by us were used.
Furthermore, covalent
docking studies together with structureꢀactivity relationships concluded from the kinetic data
ε
shed light on the binding mode of the N ꢀacryloyllysines. Regarding the suitability of
compounds as potential radiotracers, both their reactivity towards the desired molecular
target and their pharmacokinetic properties are important. Therefore, an initial in vitro
ε
pharmacokinetic characterisation of the novel N ꢀacryloyllysines was carried out. These
studies included the assessment of the compound’s potential membrane permeability, their
stability towards oxidative metabolism and their potential unspecific binding, using
appropriate analytical assays and computational methods.
8
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 112
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
Results and Discussion
ε
Chemistry for the synthesis of N ꢀacryloyllysine piperazides
α
ε
Synthesis of N ꢀBocꢀN ꢀacryloyllysine
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
Structural modifications of lead compound 6a were primarily focused on the introduction of
fluorine at different sites, but also other functional groups were considered to uncover SARs.
According to retrosynthetic considerations, compound 6a can be fragmented into three main
ε
parts: the central N ꢀacryloylꢀ
Lꢀlysine, the Nꢀterminal phenyl acetic acid and the Cꢀterminal 6ꢀ
methylpyridinꢀ2ꢀylpiperazine. For chemical modifications, the Cꢀ and Nꢀterminal moieties
seem to be most appropriate. Based on the principles of amino acid and peptide chemistry,
incorporation of the phenylacetyl group at the αꢀamino group should be conducted after
ε
incorporation of the piperazinyl moiety at the αꢀcarboxyl group (Scheme 2). Furthermore, N ꢀ
acryloylation was accomplished as first step, which allows for a modular synthetic approach.
As the positions for structural modfications are at two different moieties of the inhibitors, the
inhibitory potential can be finally further increased by combination of advantageous
substituents. It should be mentioned that the synthetic strategy for the inhibitors in this study
basically corresponds to the published method, but it varies in the conditions for distinct
58,61
steps.
9
ACS Paragon Plus Environment

Page 11 of 112
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
α
ε
Scheme 2. General synthesis steps for the N ꢀacylꢀN ꢀacryloyllysine piperazides using the
example of reference compound 6a
Reagents and conditions: a) TEA, CH
3
OH, 2 h; b) nꢀBuOH, 130°C, 5 d; c) PyBOP, DIPEA, THF, 5 h; d)
TFA/CH Cl (1/1, v/v), 2 h; e) phenylacetyl chloride, TEA, CH
2
2
2
Cl , 5 h. This synthetic approach is basically
2
followed for the preparation of all other final compounds. The respective piperazine building blocks used in
place of 3a are shown Scheme 3. The identity of the moieties attached to the αꢀamino group is disclosed in
the Tables within the section ‘Structureꢀactivity relationships for human TGase 2’. The corresponding
ε
acylating agents required for conversion of the 5aꢀderived N ꢀacryloyllysine piperazides are detailed within
the compound characterisation data in the Supporting Information.
α
ε
ε
For the synthesis of N ꢀBocꢀN ꢀacryloyllysine, N ꢀacryloylation was performed by the use of
Nꢀacryloxysuccinimide (1a), which was synthesised from acryloyl chloride and Nꢀ
6
2ꢀ64
hydroxysuccinimide according to published methods.
In this context, Nꢀ
acryloxysuccinimide appeared to be advantageous over acryloyl chloride, as the latter one
α
ε
α
led to the formation of the condensation product of N ꢀBocꢀN ꢀacryloyllysine and N ꢀBocꢀ
lysine as side product. The high solubility of the desired product in water caused challenges
during processing and purification of the reaction mixture. Therefore, initially the crude
product was used for the subsequent coupling reaction with the piperazine building blocks.
Residual Nꢀhydroxysuccinimide and triethylamine should not have detrimental effects on that
reaction. However, ongoing attempts to purify the crude product (without processing)
2 2
revealed a method for column chromatography which uses CH Cl with gradually increasing
1
0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 112
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
content of isopropanol (from 5 to 8%) containing a constant portion of acetic acid (2%) as
eluent. To avoid formation of acetamide during the subsequent coupling reaction, residual
acetic acid was removed by repeated azeotropic evaporations of the product in the presence
65
of toluene. Compound 2a was obtained in 50% yield by this method.
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
Syntheses of the piperazine building blocks
Most of the piperazine building blocks used in this study were synthesised by the reaction of
piperazine or Bocꢀpiperazine with the respective pyridine, benzoic acid and sulfonic acid
derivatives (Scheme 3). For compounds 3a and 3b, the thermal substitution at 2ꢀ
6
6
67
chloropicoline (in nꢀbutanol ) and 3ꢀbromoꢀ6ꢀnitropyridine (in acetonitrile ), respectively,
with piperazine was used without the addition of a further base. Due to the high reactivity of
2,6ꢀdifluoropyridine, the procedure of Prante et al. was adopted for the synthesis of Bocꢀ3d,
which uses Bocꢀpiperazine instead of piperazine, triethylamine as base and DMF instead of
68
nꢀbutanol as solvent. A series of pyridylꢀsubstituted piperazines (Bocꢀ3iꢀBocꢀ3n) was
obtained by BuchwaldꢀHartwig amination of Bocꢀpiperazine with the respective bromoꢀ
69
substituted pyridine derivatives, with the exception of the 6ꢀnitropyridinꢀ2ꢀyl derivative,
which was synthesised by the use of 2ꢀchloroꢀ6ꢀnitropyridine. The coupling reactions
furnished the desired piperazine building blocks in good yields (40ꢀ76%) after purification by
column chromatography. 2ꢀBromoꢀ6ꢀtertꢀbutylpyridine (Brꢀ3n), which was needed for the
synthesis of Bocꢀ3n, was accessible by a novel Cu(I)ꢀcatalysed coupling reaction of tertiary
70ꢀ71
Grignard reagents with azacyclic electrophiles.
The other pyridine derivatives were
commercially available. A structural variation, which was not intended arose during the
reaction of Bocꢀpiperazine with 2ꢀchloroꢀ6ꢀtrichloromethylpyridine under the conditions of a
BuchwaldꢀHartwig amination. Instead of the intended 6ꢀtrichloromethylpyridyl moiety, the
incorporation of the 6ꢀchloropicolinoyl moiety occurred, which was confirmed by NMR and
72
ESIꢀMS analyses (compound Bocꢀ3h). The substitution of the ring chlorine is known,
3 3
however, even under different conditions (NaOH/TEA as bases and CH CN/DMF/CH OH as
1
1
ACS Paragon Plus Environment

Page 13 of 112
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
solvents), the formation of the picolinic acid derivative was exclusively observed. Such
reaction products have been identified for the reaction of βꢀtrichloromethylpyridines with
7
2ꢀ73
nucleophiles as well as for the hydrolysis of 2ꢀchloroꢀ6ꢀtrichloromethylpyridine.
For the selective monoꢀacylation with the 4ꢀfluorobenzoylꢀ and 4ꢀnitrobenzoyl groups, Bocꢀ
piperazine was reacted with the respective benzoyl chlorides. After processing of the
reaction mixtures (especially washing with 1 M HCl), a purification step for both compounds
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
(
Bocꢀ3e and Bocꢀ3f) could be omitted. The incorporation of the dansyl moiety was managed
according to the procedure by Sashuk et al. using piperazine and dansyl chloride without the
7
4
need for purification (3c). The synthesis of the 4ꢀfluorobenzylꢀsubstituted piperazine
Bocꢀ3g was realised by reductive amination between Bocꢀpiperazine and 4ꢀ
75
fluorobenzaldehyde using sodium triacetoxyborohydride as reducing agent.
After
purification by column chromatography, this piperazine derivative was obtained in moderate
yield (42%).
Some variations in the piperazine building blocks were achieved by direct modification of
pyridylꢀsubstituted piperazines (Scheme 3). In this context, the fluoroethoxy group was
introduced via nucleophilic aromatic substitution of 1ꢀBocꢀ4ꢀ(6ꢀfluoropyridinꢀ2ꢀyl)piperazine
76
and 2ꢀfluoroethanol. Worthy of note, the fluoroethoxyꢀsubstituted product Bocꢀ3o could not
be separated from its fluoroꢀsubstituted educt Bocꢀ3d, neither by RPꢀHPLC nor by
conventional normalꢀphase column chromatography. Analysis of the crude product by NMR
revealed a proportion of 40% of the product in the mixture. As the extent of conversion could
2
not be increased by different conditions (CaH instead of NaH and longer reaction time),
subsequent Boc deprotection of the crude product was performed followed by purification via
RPꢀHPLC. The commercially available 1ꢀBocꢀ4ꢀ(6ꢀbromopyridinꢀ2ꢀyl)piperazine served as
starting material for two further piperazine building blocks. The 1ꢀBocꢀ6ꢀiodopyridinꢀ2ꢀ
ylpiperazine (Bocꢀ3p) was obtained quantitatively via a cooperꢀcatalysed “aromatic
77
Finkelstein reaction“ with NaI. For the synthesis of 1ꢀBocꢀ6ꢀphenylpyridinꢀ2ꢀylpiperazine
1
2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 112
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
(
Bocꢀ3q, 59%), a Suzuki coupling with phenylboronic acid in aqueous environment was
selected, according to the procedure for the Suzuki coupling of βꢀchloroacroleins with boronic
78
acid described by Hesse and Kirsch.
Starting from the final inhibitor bearing a picolinic acid methyl ester moiety (compound 8f),
two further structural variations were obtained. Hydrolysis of the ester bond furnished the
respective picolinic acid derivative 8e, which was subsequently used for the conversion to
the primary amide 8g.
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
3
ACS Paragon Plus Environment

Page 15 of 112
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 3. Synthetic methods for the preparation of the piperazine building blocks 3
Boc removal from Bocꢀpiperazine derivatives Bocꢀ3dꢀBocꢀ3q to yield 3dꢀ3q was performed using a mixture
of TFA/CH
Compounds 3rꢀ3x are not shown in the scheme as they are commercially available. Reagents and
conditions: a ) 2ꢀchloropicoline, nꢀbutanol, 130°C, 5 d; a ) 3ꢀbromoꢀ6ꢀnitropyridine, CH CN, 95°C, 6 h; a
dansyl chloride, CH ) 2,6ꢀdifluoropyridine, TEA, DMF, 150°C, 24 h; b ) 4ꢀfluorobenzoyl
chloride, TEA, CH Cl ) 4ꢀfluorobenzaldehyde, sodium
triacetoxyborohydride, THF, N (dba) , Ar, 70°C, 24 h; c)
respective bromopyridine derivative, Cs , THF, Ar, 70°C, 5ꢀ12 h; d) CuI, tertꢀ
2 2
Cl as outlined in the main text and in the Chemistry Section in Supporting Information.
1
2
3
3
)
2
Cl
2
, 15 min; b
1
2
2
2
, 2 h; b
) 4ꢀnitrobenzoyl chloride, TEA, THF, 2 h; b
3 4
2
, 5 h; b
5
) nitrapyrin, Cs
2
CO
3
, xantphos, Pd
2
3
2
CO , xantphos, Pd
3
2
(dba)
3
butylmagnesium chloride, THF, Ar, <0°C ꢀ 30°C, 3 h; e) fluoroethanol, NaH, DMSO, Ar, 100°C, 2 d; f) CuI,
NaI, TMEDA, 1,4ꢀdioxane, Ar, 100°C, 44 h; g) phenylboronic acid, Pd(OAc) , (C NBr, K CO , H O/THF
5:2, v/v), 100°C, 6h.
2
4
H
9
)
4
2
3
2
(
1
4
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 112
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 the case of the Bocꢀprotected piperazine derivatives, the Boc group had to be removed
α
ε
prior to coupling with N ꢀBocꢀN ꢀacryloyllysine. To this end, the Bocꢀpiperazine derivatives
were treated with a mixture of TFA/CH Cl (1:1, v/v) for 2 h. To avoid excessive use of base
2
2
for the amide coupling reactions, the resulting TFA salts were neutralised with aqueous
NaOH and extracted into the organic phase to obtain the free secondary amines. This
procedure was omitted for the picolinic acid methyl ester derivative Bocꢀ3l due to its
susceptibility to basic hydrolysis.
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
α
ε
Amide coupling between N ꢀBocꢀN ꢀacryloyllysine and the piperazine building blocks
The formation of the amide bond was realised using PyBOP as a coupling agent in the
presence of DIPEA as base. All coupling products (compounds 4) were purified by column
α
ε
chromatography. The conversion of the crude product of N ꢀBocꢀN ꢀacryloyllysine resulted in
only slightly diminished yields compared to the purified product (20ꢀ90% and 50ꢀ90%), which
ε
justifies the initial approach omitting the purification after N ꢀacryloylation.
Boc removal and functionalisation at the αꢀamino group
The removal of the Boc group from the coupling products was conducted as described for
2 2
the Bocꢀprotected piperazine building blocks (TFA/CH Cl (1:1, v/v), 2 h). However, the
neutralisation of the TFA salts (compounds 5) was not accomplished. Therefore, additional
equivalents of base had to be considered for the next reaction step. The last step to the final
inhibitors (compounds 6ꢀ21) comprised the functionalisation at the αꢀamino group. The
incorporation of the different substituents was achieved by three different ways: reaction of
activated carboxylic acid derivatives or analogues, PyBOPꢀmediated reaction with carboxylic
acids and reductive amination.
In the course of the last reaction step several side reactions were observed. When possible,
the respective side products (6c, 16aꢀd, 14nꢀq) were separated and characterised by NMR
and ESIꢀMS. The results are discussed in the Supporting Information (Discussion S1).
1
5
ACS Paragon Plus Environment

Page 17 of 112
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Transglutaminase assays
Fluorimetric assay method
ε
For the characterisation of the N ꢀacryloyllysines towards inhibition of hTGase 2 (human
transglutaminase 2), a continuous fluorimetric assay method was used, which was recently
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
59
developed by our group. This assay is based on the measurement of an increase in
fluorescence over time due to the TGase 2ꢀcatalysed hydrolysis of the waterꢀsoluble
fluorogenic acyl donor ZꢀGlu(HMC)ꢀGlyꢀOH resulting in the formation of HMC and ZꢀGluꢀGlyꢀ
OH. Due to significant spontaneous hydrolysis of ZꢀGlu(HMC)ꢀGlyꢀOH at higher pH values,
kinetic characterisation of inhibitors was performed at pH 6.5 using six different
concentrations of inhibitor. Analysis of the progress curves was done considering the
spontaneous reactivity of substrate 5b. This phenomenon results in steady state rates
different from zero, even though the enzymatic activity is completely inhibited by the
irreversible inhibitors (see Figure S1 in Supporting Information). Therefore, equation VII (see
Experimental Section) was applied which is usually used for the evaluation of reversible
slowꢀbinding inhibitors. This allowed for a correct fitting of the progress curves to obtain
59
values of kobs and finally to derive values for the performance constants kinact/K
I
.
In case of
hyperbolically decreasing initial velocities with increasing inhibitor concentration, the
dissociation constants (K) of the nonꢀcovalent enzymeꢀinhibitor complex were also derived.
i
In addition to the characterisation of the desired drugꢀtarget interaction, assessment of
specificity and selectivity of drug candidates is fundamental. As the familiy of
transglutaminases consists of eight enzymatically active proteins, the development of
selective hTGase 2 inhibitors is challenging. To address the issue of selectivity, assay
methods for the characterisation of transglutaminases beside hTGase 2 are necessary. For
the characterisation of the selectivity (based on kinact/K
I
values) of irreversible inhibitors with
79
3ꢀbromoꢀ4,5ꢀdihydroisoxazol warhead, Klöck et al. utilised the wellꢀknown GDHꢀcoupled
8
0
assay method with appropriate small glutamineꢀcontaining peptides for the respective
1
6
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 112
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
transglutaminases (ZꢀQGꢀOH for hTGase 2 and mTGase 2, ZꢀQSꢀOH for hTGase 1 and
hTGase 3, HꢀQEQVSPLSLLKꢀOH for hfXIIIa). Inspired by the results of this study, we
hypothesised that the fluorogenic acyl donor ZꢀGlu(HMC)ꢀGlyꢀOH, which is a good substrate
59
for gpTGase 2 (guinea pig transglutaminase 2) and hTGase 2, could be also used to
characterise further TGases. Thus, the enzymatic hydrolysis of ZꢀGlu(HMC)ꢀGlyꢀOH by
mTGase 2 (murine transglutaminase 2), hTGase 1, hTGase 3, hTGase 6 and hfXIIIa was
investigated at pH 6.5. Worthy of note, acyl donor ZꢀGlu(HMC)ꢀGlyꢀOH is also a suitable
substrate for all other TGases studied (Figure 2 and Table 1). While the kinetic parameter
towards mTGase 2 are comparable to gpTGase 2 and hTGase 2, the substrate properties
towards hTGase 1, hTGase 3, hTGase 6 and hfXIIIa are significantly diminished. The
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
m
performance constants (kcat/K ) for these TGases are reduced to 1/18 or less compared to
hTGase 2. Considering the offꢀtarget TGases, ZꢀGlu(HMC)ꢀGlyꢀOH exhibits the most
favourable properties towards hTGase 6, followed by hTGase 1, hfXIIIa and hTGase 3. In a
8
1
recent study, Akbar et al. kinetically characterised the wellꢀknown TGase 2 substrate Zꢀ
ꢀ1 ꢀ1
Glu(pNP)ꢀGlyꢀOH (k /K =11 833 M s for hTGase 2) towards hTGase 1 and hTGase 6 at
cat
M
pH=6.5. Similar to its coumarinylester analogue, this compound is also a suitable substrate
for these TGases even though the performance constants are significantly reduced
compared to hTGase 2. In contrast to ZꢀGlu(HMC)ꢀGlyꢀOH, ZꢀGlu(pNP)ꢀGlyꢀOH is better
ꢀ1 ꢀ1
ꢀ1 ꢀ1
M M
tolerated by hTGase 1 (kcat/K =1 333 M s ) than by hTGase 6 (kcat/K =500 M s ), which
might indicate differences in the recognition of these substrates within the active sites of the
different TGases.
1
7
ACS Paragon Plus Environment

Page 19 of 112
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 2. Enzymatic hydrolysis of ZꢀGlu(HMC)ꢀGlyꢀOH by different TGases at pH 6.5
Plots of v0total=f([ZꢀGlu(HMC)ꢀGlyꢀOH]) for gpTGase 2, hTGase 2 and mTGase 2 and plots of v0corr=f([Zꢀ
Glu(HMC)ꢀGlyꢀOH]) for hTGase 1, hTGase 3, hTGase 6 and hFXIIIa including nonlinear regressions
according to equation VI (Experimental Section). Data shown are mean values ±SEM of three separate
experiments, each performed in duplicate. When not apparent, error bars are smaller than the symbols. Plots
59
for gpTGase 2 and hTGase 2 have already been shown in a previous study. Conditions: pH 6.5, 30 °C, 5%
DMSO, 500 ꢁM TCEP, 3 ꢁg/mL TGase.
1
8
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 112
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 1. Kinetic parameters for the enzymatic hydrolysis of acyl donor ZꢀGlu(HMC)ꢀGlyꢀOH by
different TGases at pH 6.5 and 30 °C
regression analysis
numerical integration
ꢀ1
ꢀ1 ꢀ1
ꢀ1
ꢀ1 ꢀ1
Enzym
K
m
(ꢁM)
k
cat (s )
kcat/K
m
(M s )
K
m
(ꢁM)
k
cat (s )
m
kcat/K (M s )
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
gpTGase 2*
hTGase 2*
mTGase 2
hTGase 1
2.53 (0.18)
6.60 (1.06)
6.10 (0.26)
0.33 (0.01)
0.32 (0.02)
0.29 (0.01)
0.17 (0.02)
130 000
48 500
47 500
929
2.67 (0.20)
5.78 (1.03)
5.55 (0.34)
0.34 (0.01)
0.32 (0.02)
0.28 (0.01)
129 000
57 600
50 500
183 (39)
190 (52)
0.18 (0.03)
947
#
#
#
226
0.19
841
hTGase 3
202 (25)
0.07 (0.01)
347
ꢀ
ꢀ
ꢀ
#
#
#
5
06
0.06
119
2 820
683
hTGase 6
hFXIIIa
46.1 (2.7)
0.13 (0.01)
0.34 (0.08)
44.1 (2.4)
0.14 (0.01)
0.33 (0.07)
3 170
748
498 (148)
4
41 (104)
#
#
#
7
72
0.48
622
For details on the calculation of the kinetic parameters, see Experimental Section. Data shown are mean
values (±SEM) of three separate experiments, each performed in duplicate. Active concentrations of
mTGase 2, hTGase 1, hTGase 3, hTGase 6 and hFXIIIa were calculated from Zedira’s activity data. *Kinetic
59 #
data for gpTGase 2 and hTGase 2 have previously been published.
Data obtained by the method of
82
CornishꢀBowden and Eisenthal.
59
The same data sets were also analysed by numerical regression providing kinetic
parameters that are in good agreement to those of the classical regression analyses (Table
1). However, no fit traces could be obtained for the data of hTGase 3. The high proportion of
59
spontaneous reaction of ZꢀGlu(HMC)ꢀGlyꢀOH might be problematic for the accurate
analysis of the data for this enzyme by numerical integration.
Due to the low rates of substrate conversion by hTGase 1, hTGase 3, hTGase 6 and hfXIIIa,
the progression curves for each substrate concentration are linear. Therefore, acyl donor Zꢀ
1
9
ACS Paragon Plus Environment

Page 21 of 112
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Glu(HMC)ꢀGlyꢀOH can be applied for the kinetic characterisation of irreversible inhibitors
within its solubility limit (<250 ꢁM) towards these TGases, even though the enzymes are not
saturated by the substrate.
Fluorescence anisotropy assay
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
60
Another assay method, which was recently described by our group, is based on the
measurement of an increase in fluorescence anisotropy (FA) during the TGase 2ꢀcatalysed
transamidation of a fluorophoreꢀconjugated cadaverine derivative with DMC. Out of the three
investigated cadaverine derivatives towards gpTGase 2ꢀcatalysed transamidation,
rhodamine Bꢀisonipecotylꢀcadaverine (RꢀIꢀCad) exhibits the most favourable substrate
properties. With respect to the potential usefulness of this assay method for characterising
cellular TGase 2’s transamidation activity, the characterisation of the substrate pair RꢀIꢀCad
and DMC towards hTGase 2 was envisaged. Furthermore, the characterisation of selected
inhibitors by this assay method was sought to compare the results with those of the
ε
fluorimetric assay method. This enabled the evaluation of the inhibitory properties of the N ꢀ
acryloyllysines at different pH values (6.5 versus 8.0) as well as towards different enzymatic
activities (hydrolase versus transamidase activity).
Initial investigations on the dependence between the activity and concentration of hTGase 2
provided similiar results compared to gpTGase 2 as linearity predominates over a wide range
of enzyme concentrations (0.5ꢀ5 ꢁg/mL, Figure S2 in Supporting Information). The results
obtained by varying the concentration of one substrate in the presence of the other substrate
at a fixed concentration are shown in Table 2 and Figure S3 in Supporting Information.
Compared to gpTGase 2, a similar kinetic behaviour of both substrates can be observed
towards hTGase 2, especially with regards to the maximum velocities. However, the K
of DMC towards hTGase 2 is significantly reduced (2.78 ꢁM) compared to gpTGase 2
27.5 ꢁM). Concerning the characterisation of irreversible inhibitors, the potential of the FA
m
value
(
60
assay method was also recently demonstrated. Due to the high rate of hTGase 2ꢀcatalysed
2
0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 112
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
incorporation of RꢀIꢀCad in DMC it was decided to reduce the enzyme concentration from
5
ꢁg/mL to 2 ꢁg/mL for inhibitor characterisation. Compared to gpTGase 2, the time for
preincubation of enzyme and inhibitor had to be reduced for hTGase 2 from 30 to 5 min due
ε
to the significantly higher reaction rate of most of the N ꢀacryloyllysines towards the human
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
enzyme (see below).
Table 2. Kinetic parameters for the substrate pairs DMC/FꢀCad towards gpTGase 2 and
DMC/RꢀIꢀCad towards gpTGase 2 and hTGase 2
ꢀ1
Enzym
K
m
(ꢁM)
V
max (mA s )
K
i
(ꢁM)
DMC
3.05 (0.44)
2.22 (1.40) ×10
27.5 (6.4)
0.027 (0.002)
0.018 (0.001)
0.176 (0.028)
0.113 (0.013)
0.132 (0.001)
0.121 (0.002)
ꢀ
gpTGase 2
ꢀ
4
FꢀCad
9.15 (3.11)
DMC
135 (32)
gpTGase 2
hTGase 2
ꢀ4
RꢀIꢀCad
2.42 (0.69) ×10
2.78 (0.10)
7.14 (2.20)
DMC
174 (16)
ꢀ4
RꢀIꢀCad
3.43 (0.02) ×10
6.81 (0.45)
For details on the calculation of the kinetic parameters, see Experimental Section. Data shown are
mean values (±SEM) of three (gpTGase 2) or two (hTGase 2) experiments, each performed in
duplicate (hTGase 2) or triplicate (gpTGase 2). Kinetic data for gpTGase 2 have previously been
60
published.
Taken together, two reliable assay methods based on fluorescence intensity and
fluorescence anisotropy were established for inhibitor characterisation, which facilitate
investigations of the inhibitory behaviour at different pH values towards the acyl transferase
activity for different substrates and for different TGase isoforms. Due to the known similarity
between gpTGase 2 and hTGase 2, which has often justified the use of gpTGase 2 as costꢀ
8
3
efficient alternative to the human variant in the past, there was a special interest on the
ε
behaviour of the N ꢀacryloyllysines towards both enzymes. Even though the settings of both
2
1
ACS Paragon Plus Environment

Page 23 of 112
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
assay methods were basically identical for these TGases, it should be mentioned that for
gpTGase 2 fluoresceinꢀlabelled cadaverine (FꢀCad) was chosen as acyl acceptor for the
measurements by the FA assay method while RꢀIꢀCad was used for experiments on
hTGase 2 (kinetic plots and data are also shown for the purpose of comparison in Table 2
and Supporting Information). This difference in the composition of the assay mixture should,
however, be unproblematic, as FꢀCad and RꢀIꢀCad were shown to provide reasonably
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
6
0
comparable results for the characterisation of irreversible inhibitors.
ε
To confirm the irreversibility of the inhibition of gpTGase 2 and hTGase 2 by the N ꢀ
8
4
acryloyllysines, a jumpꢀdilution experiment according to Copeland
was exemplarily
performed for the inhibition of gpTGase 2 by lead compound 6a. As expected, no enzymatic
activity has been restored after 1:100 dilution of a preꢀincubated mixture of enzyme and
inhibitor (both 100ꢀfold concentrated) into the assay mixture, which clearly confirms the
ε
irreversible inhibition of TGase 2 by N ꢀacryloyllysines (Figure S4 in Supporting Information).
In order to determine the active enzyme concentrations, the FAꢀbased assay has recently
been shown to be a reliable and fast method for activeꢀsite titration of gpTGase 2 by using
60
iodoacetamide in a concentration range stoichiometric to the enzyme. However, when
TCEP instead of DTT was used as antioxidant for hTGase 2, the activeꢀsite titration with
iodoacetamide led to false results for that enzyme due to the known reactivity of the inhibitor
85
ε
with TCEP. Therefore, one of the most potent N ꢀacryloyllysines in this study, compound 8d
(
see below), was applied for the activeꢀsite titration hTGase 2. By using a preincubation time
of 40 min, the concentration of catalytically active hTGase 2 could be determined (with Rꢀ
IꢀCad and DMC as substrates; Figure S5 in Supporting Information). Furthermore, a
comparable result was obtained for hTGase 2 with the fluorimetric assay following the
hydrolysis of ZꢀGlu(HMC)ꢀGlyꢀOH in the presence of 8d after a preincubation period of
40 min (Figure S5).
2
2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 112
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
Molecular docking studies
Computational methods that are available for the modelling of covalent enzymeꢀinhibitor
8
6
complexes were considerably improved over the recent years. Within this study, covalent
docking was carried out to predict the binding mode of the synthesised irreversible covalent
inhibitors towards hTGase 2 and rationalise the differences in enzymatic activity which were
experimentally observed at atomic detail. The crystal structure of the open conformation of
hTGase 2 in complex with a peptideꢀlike irreversible inhibitor developed by Khosla’s group
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
8
7
was used for this purpose (PDB ID 2Q3Z). In this structure, two hydrophobic pockets are
well defined. The entrance to pocket 1 through the catalytic site is flanked by Lys176 and
Met252 and its tunnelꢀlike shape involves the residues Gln169, Ile178, Trp180, and and the
backbone of Trp254. Pocket 2 is deeper and comprises Asn302, Ala304, Leu312, Ile313,
Phe316, Met330, Ile331 and Leu420. In the crystal structure, there is no electron density for
residues 307ꢀ308 and 319ꢀ327, which are located proximal to the catalytic site. Due to the
potential role of these missing residues on defining proximal binding sites for the newly
developed inhibitors, they were modelled before performing the docking studies (see
Experimental Section, Figure S6 in Supporting Information). A similar approach was
88
previously undertaken by Badarau et al. with good results regarding the predictions. The
energyꢀrefined 3D molecular model of hTGase 2 in the open conformation suggested an
additional pocket 3 proximal to the pocket 2 flanked by Arg317 and Asn308 (Figure 3A),
whose existence has been proposed before even though no experimental or computational
49
proof was provided. Covalent docking studies were first performed with the lead compound
6a. Binding poses were predicted either with the phenylacetyl group or the pyridylpiperazinyl
moiety occupying the hydrophobic pocket 2. The first scenario with the phenylacetyl in
pocket 2 (Figure 3A) has revealed three H bonds involving the hTGase 2 residues Trp241,
Ile331 and Asn333. In this case, the methyl group of the pyridylpiperazinyl moiety was
forming vanꢀderꢀWaals contacts with Arg317. In the second scenario with the
2
3
ACS Paragon Plus Environment

Page 25 of 112
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
pyridylpiperazinyl moiety located in pocket 2 (Figure S7 in Supporting Information), multiple
8
9
binding modes were predicted in which the phenyl group could either be oriented towards
Arg317, Phe320 or Phe334. The latter binding mode was more favourable than the other
two, being stabilised by three H bonds (Trp241, Cys277 and Asn333). Hence, the hTGase 2ꢀ
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
6a complex might be represented by a dynamic ensemble of multiple binding poses with the
phenylacetyl group occupying pocket 2 or interacting with Phe334. Residues Arg317 and
Phe320 might act as intermediate anchor points for the dynamic interconversion between the
two binding modes.
2
4
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 112
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 3. Molecular modelling of the interaction of hTGase 2 with covalent inhibitors
The enzyme is shown in gray surface with relevant residues coloured by atom type and inhibitors are shown
in sticks and colored by atom type. Intermolecular H bonds and cationꢀπ interactions are depicted by black
and green dashed lines, respectively. A) 6a (green), B) 7e (yellow, iodine atom highligted in magenta
surface) superimposed with 6a (methyl substitution in pyridine ring highlighted in green surface), C) 7f
90
(
orange) and D) 7i (light orange). Figure generated in Maestro (Schrödinger) using the model generated
87
from PDB ID 2Q3Z .
Structureꢀactivity relationships for human TGase 2
ε
Enantiomeric purity of the N ꢀacryloyllysine piperazides
Concerning the enantiomeric purity of the inhibitors, the amide coupling between the lysine
mainꢀchain carbonyl and the piperazine building blocks is the most crucial step. However,
due to the low electron withdrawing effect of the lysine side chain as well as the protection of
the αꢀamino group as urethane, enantiomerisation is strongly hindered. To further support
25
ACS Paragon Plus Environment

Page 27 of 112
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
the conservation of the configuration at the C atom, the moderately strong base DIPEA and
α
9
1
the moderately polar solvent THF were used. Finally, the enantiomeric purity of the
inhibitors was analysed exemplarily for compound 6a by CD spectroscopy and HPLC using a
chiral stationary phase (chiral HPLC). In addition to the appropriate analytical method, both
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
enantiomers are needed. Therefore, compound 6b was synthesised starting from BocꢀDꢀ
lysine. The CD spectra confirmed that both compounds are enantiomers (Figure S8 in
Supporting Information). Analysis of mixtures of 6a with increasing amounts of 6b by chiral
HPLC revealed that the enantiomeric purity is greater than 98% (Figure S9 in Supporting
Information). Therefore, the stereochemical integrity was not affected during the synthesis of
the compounds. As the procedure for the amide coupling was similarly performed for all
compounds, the enantiomeric purity of 6a can be considered as representative for all
inhibitors.
α
Influence of configuration at C
ε
Even though N ꢀacryloyllysines are known for a long time to act as potent inhibitors of
56
α
TGase 2, the influence of the configuration at C on their inhibitory behaviour has not been
investigated so far. The focus on the ꢀenantiomer is probably a result from studies on
L
substrates for gpTGase 2 where peptides containing Lꢀglutamine exhibit a significantly higher
92
substrate potential than peptides with Dꢀglutamine. However, to confirm this assumption the
D
ꢀenantiomer (6b) of compound 6a was kinetically characterised towards hTGase 2 (Table
). The change of the configuration leads to a marked decrease in the inhibitory potential.
The ratio of the inactivation constants of both enantiomers ((kinact/K , eudismic
ratio) is 24. Consequently, this verifies the focus on the ꢀconfiguration in the past and
justifies the synthesis of all inhibitors in this work starting from Bocꢀ ꢀlysine.
3
I S I R
) /(kinact/K)
L
L
To shed light on the molecular bases for recognition that could explain the different inhibitory
potential of both enantiomers, further covalent docking studies were carried out on 6b, which
revealed to be challenging considering the hydrophobic nature of the binding pockets. As for
26
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 112
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
6
a, either the phenylacetyl or the pyridylpiperazinyl moieties were predicted to be located in
pocket two. The top ranked pose resembled the binding mode predicted for 6a with the
pyridylpiperazinyl moiety making contacts to Arg317 and phenylacetyl group being located in
pocket two but oriented towards Phe316 (Figure S7B). In contrast to 6a, binding poses with
the pyridylpiperazinyl moiety of 6b in pocket two are less favourable. Here, two binding
modes were predicted with the phenylacetyl group either oriented towards Tyr245 and
Met252, supported by two H bonds (Trp241 and Phe334), or positioned towards Ser171 and
Ser303, being stabilised by just one H bond (Trp241) (Figure S7C). The latter differences in
binding modes found for both enantiomers could offer a plausible explanation for the much
lower inhibitory potential of 6b. Conclusively, results from docking studies cannot
comprehensively explain the found differences in the inhibitory activity of the enantiomers 6a
and 6b. For an advanced understanding of the differential mechanisms for molecular
recognition of these compounds, QM/MM (quantumꢀmechanics/molecularꢀmechanics)
studies should be performed in the future.
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
α
Table 3. Influence of the configuration at C on the inhibition of hTGase 2
ꢀ
1 ꢀ1
compound
configuration
kinact/K (M s )
K (ꢁM)
i
I
#
6a
S
R
4 880 (20)
206 (8)
5.73 (0.77)
ꢀ
6b
Data shown are mean values (
±
SEM) of two separate experiments, each performed in duplicate.
#
59
Kinetic data for compound 6a have previously published.
27
ACS Paragon Plus Environment

Page 29 of 112
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Substitution at position 6 of pyridineꢀ2ꢀylpiperazines
As discovered by Wityak et al., the methyl group in position 6 of the pyridine ring significantly
increases the inhibitory potential compared to the unsubstituted pyridine (as investigated for
compounds 4o and 4p by Wityak et al., which differ from 7a and 6a in the way 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
α
58
phenylacetyl group is replaced by a Z group as N ꢀsubstituent). Therefore, we initially
focussed on the systematic variation of the substituent in this position. The kinetic data of the
respective series of analogues are shown in Table 4.
28
ACS Paragon Plus Environment