Development of potent and selective tissue transglutaminase inhibitors: Their effect on TG2 function and application in pathological conditions

Article abstract of DOI:10.1016/j.chembiol.2015.08.013

Potent-selective peptidomimetic inhibitors of tissue transglutaminase (TG2) were developed through a combination of protein-ligand docking and molecular dynamic techniques. Derivatives of these inhibitors were made with the aim of specific TG2 targeting to the intra- and extracellular space. A cell-permeable fluorescently labeled derivative enabled detection of in situ cellular TG2 activity in human umbilical cord endothelial cells and TG2-transduced NIH3T3 cells, which could be enhanced by treatment of cells with ionomycin. Reaction of TG2 with this fluorescent inhibitor in NIH3T3 cells resulted in loss of binding of TG2 to cell surface syndecan-4 and inhibition of translocation of the enzyme into the extracellular matrix, with a parallel reduction in fibronectin deposition. In human umbilical cord endothelial cells, this same fluorescent inhibitor also demonstrated a reduction in fibronectin deposition, cell motility, and cord formation in Matrigel. Use of the same inhibitor in a mouse model of hypertensive nephrosclerosis showed over a 40% reduction in collagen deposition.

Full text of DOI:10.1016/j.chembiol.2015.08.013

Article
Development of Potent and Selective Tissue
Transglutaminase Inhibitors: Their Effect on TG2
Function and Application in Pathological Conditions
Graphical Abstract
Authors
Eduard Badarau, Zhuo Wang,
Dan L. Rathbone, ..., Said El Alaoui,
Milda Barkeviciute, Martin Griffin
Correspondence
m.griffin@aston.ac.uk
In Brief
Badarau et al. design and develop high-
potency TG2-specific irreversible
inhibitors that show reactivity with the
intracellular active form of TG2, leading to
inhibition of its translocation into the
extracellular matrix. The compounds are
effective in inhibiting in vitro angiogenesis
and hypertensive nephrosclerosis in
animal models.
Highlights
d
d
d
d
Irreversible tissue transglutaminase (TG2) inhibitors of high
selectivity and potency
Target the Ca2+ activated form of TG2 in the intra- and/or
extracellular space
Reactive with intracellular TG2 and block its translocation
into the matrix
Effective in blocking hypertensive nephrosclerosis in animal
models
Badarau et al., 2015, Chemistry & Biology 22, 1–15
October 22, 2015 ª2015 Elsevier Ltd All rights reserved
http://dx.doi.org/10.1016/j.chembiol.2015.08.013

Please cite this article in press as: Badarau et al., Development of Potent and Selective Tissue Transglutaminase Inhibitors: Their Effect on TG2 Func-
tion and Application in Pathological Conditions, Chemistry & Biology (2015), http://dx.doi.org/10.1016/j.chembiol.2015.08.013
Chemistry & Biology
Article
Development of Potent and Selective Tissue
Transglutaminase Inhibitors: Their Effect on TG2
Function and Application in Pathological Conditions
Eduard Badarau,1,4,5 Zhuo Wang,1,4 Dan L. Rathbone,1 Andrea Costanzi,1 Thomas Thibault,1 Colin E. Murdoch,2
Said El Alaoui,3 Milda Barkeviciute,2 and Martin Griffin1,
*
1School of Life and Health Sciences, Aston University, Aston Triangle, Birmingham B4 7ET, UK
2Aston Medical Research Institute, Aston Medical School, Aston University, Aston Triangle, Birmingham B4 7ET, UK
3Covalab France, 11 avenue Albert Einstein, 69100 Villeurbanne, France
4Co-first author
5Present address: European Institute of Chemistry and Biology, 2 Rue Robert Escarpit, 33607 Pessac, France
*Correspondence: m.griffin@aston.ac.uk
http://dx.doi.org/10.1016/j.chembiol.2015.08.013
SUMMARY
specifically inhibiting TG2 crosslinking activity in physiological/
pathological conditions. TG2 is involved in a host of human dis-
Potent-selective peptidomimetic inhibitors of tissue eases such as celiac disease, cancer, fibrosis, multiple sclerosis,
and neurodegeneration (Wang and Griffin, 2012). As a conse-
quence, TG2 has become a potential therapeutic target for
transglutaminase (TG2) were developed through a
combination of protein-ligand docking and molecu-
lar dynamic techniques. Derivatives of these inhibi-
tors were made with the aim of specific TG2 targeting
to the intra- and extracellular space. A cell-perme-
able fluorescently labeled derivative enabled detec-
tion of in situ cellular TG2 activity in human umbilical
cord endothelial cells and TG2-transduced NIH3T3
the treatment of these pathological conditions (Badarau et al.,
2013a). Moreover, since TG2 is found both in the intra- and extra-
cellular space, both targeted specificity for TG2 itself and spe-
cific targeting to a particular subcellular space are needed.
We recently described inhibitors aimed at targeting only the
extracellular space (Badarau et al., 2013b; Griffin et al., 2008).
These molecules incorporate two key polar groups in their
structure in order to assure their high hydrophilic character: a
carboxylate moiety and a charged warhead represented either
cells, which could be enhanced by treatment of cells
with ionomycin. Reaction of TG2 with this fluorescent
inhibitor in NIH3T3 cells resulted in loss of binding by a dimethylsulfonium (Griffin et al., 2008) or by an imidazolium
function (Badarau et al., 2013b). Despite their activities in the
low micromolar range, these derivatives have great potential
of TG2 to cell surface syndecan-4 and inhibition of
translocation of the enzyme into the extracellular
matrix, with a parallel reduction in fibronectin depo-
sition. In human umbilical cord endothelial cells,
this same fluorescent inhibitor also demonstrated a
reduction in fibronectin deposition, cell motility, and
cord formation in Matrigel. Use of the same inhibitor
in a mouse model of hypertensive nephrosclerosis
because of their high aqueous solubility and their excellent
cellular toxicity profile. In fact, since their intracellular access is
restrained, these inhibitors show very little toxicity in cells up to
750 mM over 72 hr (Wang and Griffin, 2013). They have also
shown considerable success in proof-of-concept studies in
animal models of kidney fibrosis over 120 days without any signs
of toxicity but with significant improvements in kidney function
showed over a 40% reduction in collagen deposition.
(Huang et al., 2009; Johnson et al., 2007). However, the major
drawback with these extracellular acting TG inhibitors is potency
and specificity. Therefore, the first objective of our work was to
design potent, highly specific TG2 inhibitors that could be used
for targeting the intra- and/or extracellular space, were not toxic
INTRODUCTION
Transglutaminases (TGs) belong to a family of protein crosslink- in cellular systems, and had the potential to be used in preclinical
ing enzymes that catalyze an acyl transfer reaction between studies.
the g-carboxamide group of peptide-bound glutamine and the
In the design of this new family of inhibitors, we were also very
ε-amino group of peptide-bound lysine (or a suitable primary much aware that, once deposited into the extracellular matrix
amine) resulting in formation of a ε(g-glutamyl)lysine isopeptide (ECM), TG2 is tightly associated with its potential substrates
bond (Griffin et al., 2002). In mammals, eight catalytically and is not easily accessible; and that the secretion and translo-
active TGs, each Ca2+-dependent, are found in this family, of cation of TG2 onto the cell surface and into the extracellular
which tissue transglutaminase (TG2) is probably the most ubiq- matrix ECM, which is independent of the conventional ER/Golgi
uitous in human tissues. The current challenges in studying pathway, requires the binding of TG2 to the cell surface heparan
the role of TG2 are to (1) develop inhibitors that target TG2 sulfate proteoglycan syndecan-4. Importantly, the binding of
and not other members of the TG family (specificity); (2) investi- TG2 to this proteoglycan is highly favored by its compact closed
gate independently the intracellular and extracellular roles of conformation (Wang et al., 2012; Lortat-Jacob et al., 2012).
the enzyme (cell permeability); (3) characterize the effect of Hence, a further objective in this work was to assess whether
Chemistry & Biology 22, 1–15, October 22, 2015 ª2015 Elsevier Ltd All rights reserved
1

Please cite this article in press as: Badarau et al., Development of Potent and Selective Tissue Transglutaminase Inhibitors: Their Effect on TG2 Func-
tion and Application in Pathological Conditions, Chemistry & Biology (2015), http://dx.doi.org/10.1016/j.chembiol.2015.08.013
suitably labeled irreversible TG2 specific inhibitors that can cord endothelial cells (HUVECs). As very little cellular toxicity
permeate the cell can also bind TG2 inside the cell thus both was observed after 72 hr at concentrations up to 100 mM
inhibiting enzyme activity and locking it into its open confir- (Figure S1), this new molecule (1a) represented the starting
mation. This is particularly important when considering TG2 as point for subsequent modifications. To rule out the effect of
a therapeutic target since TG2 in its open confirmation is likely any potential TG2 and FXIII present in the serum from the cell cul-
to be blocked in its translocation to the cell surface and into ture medium, which could eliminate the effect of the compounds
the ECM because it is unable to bind effectively to cell surface in the toxicity assay, the HUVECs were cultured in serum-free
heparan sulfates (Wang et al., 2012). Moreover, if it does manage endothelial cell medium in the presence of compounds 3h, 1h,
to reach the ECM, e.g. from cell lysis, it reaches it in an inactive and 3e up to 72 hr, with no significant toxicity detected at the
form. Such inhibitors are likely to have a broad spectrum of appli- concentrations used (Figure S2).
cation in TG2-mediated human pathologic conditions in which
TG2 is acting either inside or outside the cell.
Optimization of Inhibitor Potency
The first step of our modification strategy focused on the central
amino acid region of the newly designed ligands. In order to
evaluate the effect of steric hindrance that could be generated
RESULTS
In developing the TG2 inhibitors, the work was divided into two by additional substituents in this position, derivatives bearing
interlinked processes: in silico molecular modeling to first design new core amino acids were synthesized (Figure 1C, Schemes
and then synthesize the inhibitors and then a series of studies 1 and 2). As summarized in Table 1A, the activity was lost
on their characterization. The latter included enzyme potency, when using bulky substituents arising from phenylalanine (1l,
specificity, cell permeability, solubility, and in vitro ADME (ab- 1k), the same but more moderate trend being observed for small
sorption, distribution, metabolism, excretion) characterization substituents (1i, 1j). It should also be mentioned that the chirality
followed by an assessment of their potential application in bio- of the stereogenic carbon atom seemed important, the analog
logical systems using cell-based and in vivo preclinical models. in the L-Ala series (1i) being favored over its D isomer (1j).
Our attention next focused on the electrophilic warhead re-
Generation of New Peptidomimetic TG2 Inhibitors
gion. Different groups were tested in order to evaluate their influ-
Molecular modeling was used to yield insights into the probable ence on the binding process. As a possible means of targeting
binding pattern of our previous peptidic water-soluble deriva- the extracellular space by limiting cell permeability, highly polar
tives. The modeling approach was based on the Michaelis- warheads were attached to the newly designed skeleton. As
Menten complexes generated by docking of these compounds shown in Table 1B, the dialkylsulfonium moieties retained a
with the open conformation of TG2 (PDB: 2Q3Z). These were good inhibition range (1a, 1q), while the imidazolium warhead
then subsequently ‘‘relaxed’’ by molecular dynamics conducted induced a complete loss of activity (1s). It should be noted
in explicit water at 300 K (see Supplemental Experimental that, in our initial water-soluble peptidic inhibitors, the same
Procedures).
imidazolium warhead maintained a good TG2 inhibition range
The trajectories generated provided a rich source of informa- (Badarau et al., 2013b), suggesting the binding patterns of the
tion that helped in the development of the new TG2 inhibitors. two classes of inhibitors are very different.
One of the successfully exploited pathways relied on the obser-
Our next approach was to design inhibitors containing lipo-
vation that, during the dynamics trajectory, the water-soluble philic warheads, which were more likely to be cell permeable
ligands tend to adopt a specific conformation in the central and could be used to target the intracellular space. As the cata-
amino acid region. In fact, the dihedral N1-C2-C3-N4 empha- lytic CYS residue was revealed by the different available crystal
sized in Figure 1A stabilized around 0ꢀ with a limited variation structures of TG2 to be hosted in a narrow tunnel (PDB: 2Q3Z,
of 20ꢀ. This key observation was the starting point for the Pinkas et al., 2007; PDB: 1KV3, Liu et al., 2002; PDB: 3LY6,
design of a new series of inhibitors.
Han et al., 2010), our choice was focused on terminal Michael
In order to mimic the observed conformation for the central acceptors such as acrylamides and vinylsulfonamides. Such
N1-C2-C3-N4 region, we next introduced a piperazine scaffold, warheads possess an additional carbon atom in the electrophilic
as indicated in Figure 1B. The same three-atom linker was pre- region of the ligand, and it is likely that this will induce a different
served between N4 and the carbonyl (vicinal to the warhead), conformation of the warhead around the catalytic CYS residue
however, in order to facilitate the subsequent chemical modifi- compared with the sulfonium-based inhibitors. Table 1B shows
cations, a new nitrogen atom was introduced. One of the first that both of these new warheads preserved a good inhibition
structures synthesized to validate this ‘‘restrained conformation’’ range (2a, 3a), the vinylsulfonamide having a more favorable
hypothesis exhibited the same activity range as the initial water- effect on binding.
soluble substrates (1a, Figure 1B). The irreversibility of the inhib-
Other modifications were performed in the carbamate region
itors was first confirmed by prior incubation of the compounds (Table 1C). As this moiety was predicted by modeling to bind
with recombinant purified human TG2 and then subsequent in the hydrophobic pocket of the binding site (as revealed by
dilution prior to assay of the incubated enzyme as previously the crystal structure of the open-form conformation; Pinkas
described (Griffin et al., 2008). An indication of the hydrophilic et al., 2007), different-size aromatic carbamates were tested in
character of this derivative was observed experimentally during the first instance. In this case, the corresponding bromide pre-
the isolation step by freeze-drying the aqueous layer after cursors originally developed for inhibiting the serine proteases
separation from the organic layer. In addition, an indication of (Schoellmann and Shaw, 1963), and which have high reactivity
its cellular toxicity profile was undertaken in human umbilical with the enzymatic nucleophile, were also biologically evaluated.
2
Chemistry & Biology 22, 1–15, October 22, 2015 ª2015 Elsevier Ltd All rights reserved

Please cite this article in press as: Badarau et al., Development of Potent and Selective Tissue Transglutaminase Inhibitors: Their Effect on TG2 Func-
tion and Application in Pathological Conditions, Chemistry & Biology (2015), http://dx.doi.org/10.1016/j.chembiol.2015.08.013
Figure 1. Design and Synthesis of Lead TG2 Inhibitor Precursor
(A) Examples of dihedral angles defined by N1-C2-C3-N4 recorded during the molecular dynamics trajectory.
(B) Design of new peptidomimetic derivatives starting from the water-soluble inhibitors, R = SMe2.
(C) Scheme 1 for the synthesis of the amine precursors. (a) p-NO2-phenylchloroformate, N-methylmorpholine (NMM)/dichloromethane (DCM), 0ꢀC, 2 hr; (b)
piperazine, triethylamine (TEA)/dimethylformamide, 0ꢀC to room temperature, 12 hr; (c) tert-butyloxycarbonyl (t-Boc)-Glu-OH, 1-ethyl-3-(3-dimethylaminopropyl)
carbodiimide, HOBt, NMM/DCM, room temperature, 12 hr; (d) trifluoroacetic acid/DCM, room temperature, 24 hr. Commercially available alcohols were acti-
vated as p-nitrophenyl carbonates 7 and subsequently reacted with piperazine at room temperature in order to give the desired mono-carbamates 8. A peptidic
coupling step with t-Boc-protected amino acids, followed by the acidic deprotection of the amines conducted to the key intermediates 5.
(D) Scheme 2 for the synthesis of the final derivatives. (a) Acryloyl chloride, TEA/acetonitrile, 0ꢀC to room temperature, 3 hr; (b) bromoacetyl bromide or
chloroacetyl chloride, TEA/DCM, À78ꢀC to room temperature, 12h r; (c) dimethylsulfide/MeOH, room temperature, 24 hr; (d) acetyl bromide, TEA/DCM, À78ꢀC,
2 hr; (e) 2-chloroethylsulfonyl chloride, TEA/DCM, À60ꢀC–0ꢀC, 4 hr. Starting from the previously obtained amines 5, various acylation agents were used in order to
obtain the final derivatives 1–4 as summarized in Scheme 2. In the case of polar dimethylsulfonium warhead, an additional nucleophilic substitution step was
conducted on the halomethylamide precursor 6.
(E) Toxicity profile of the fluorescent derivative 3h on HUVEC lines, after 24 hr (white), 48 hr (gray), and 72 hr (black) as described in the Experimental Procedures.
CNTL, control.
If for the dimethylsulfonium and acrylamide warheads the synthesized. The best median inhibitory concentration (IC50) in
inhibition varies in the lower micromolar range, for the corre- the ‘‘lipophilic’’ series was obtained for the acrylamide warhead
sponding halomethylketones the potency was increased to (3e), while in the ‘‘hydrophilic’’ series, the tert-butyl carbamate
the lower nanomolar range. However, their high cellular toxicity (1m) outperformed its analogs (Table 1C).
profile for the halide precursors, particularly the more reactive
bromides, confirmed their limited pharmacological use.
Design and Characterization of a Fluorescent Probe
A new set of inhibitors was prepared following the same and Further Characterization of Lead Compounds
hypothesis of targeting the hydrophobic pocket of the active A dansyl fluorescent tag was attached to the central piperazine
site. An adamantyl moiety was chosen as the reference substit- scaffold to fully exploit the biological potential of this new class
uent and its distance from the piperazine ring was systematically of TG2-specific inhibitors, with the view to being able to pin point
varied in order to find an optimum for the interaction with the and visualize the presence of the inhibitors, most importantly the
enzyme. The corresponding tert-butyl carbamate was also presence of the active enzyme located at a cellular level. Analogs
Chemistry & Biology 22, 1–15, October 22, 2015 ª2015 Elsevier Ltd All rights reserved
3

Please cite this article in press as: Badarau et al., Development of Potent and Selective Tissue Transglutaminase Inhibitors: Their Effect on TG2 Func-
tion and Application in Pathological Conditions, Chemistry & Biology (2015), http://dx.doi.org/10.1016/j.chembiol.2015.08.013
Table 1. TG2 Inhibition Data on Varying the Central Amino Acid, the Warhead on the Initial Hit, and the Substituents in the Carbamate
Region
(A) Central Amino Acid:
(B) Warhead on the Initial Hit:
O
O
N
O
Aa
N
S
Br
N
H
O
Lig
Compound
Aa
IC50 (mM SD)
Compound
1a
R
IC50 (mM SD)
1a
Gly
1.4 0.5
1.4 0.5
1j
1i
D-Ala
6.8 2.0
1q
1s
1.85
100
0
L-Ala
0.7 0.3
1l
D-Phe
L-Phe
>100
>100
2a
3a
1.725 0.11
5.925 0.11
1k
(C) Substituents in the Carbamate Region:
Lig
Compound
6c
R1
R2
IC50 (mM SD)
0.029 0.019
1c
3c
1.07 0.1
0.44 0.057
(Continued on next page)
4
Chemistry & Biology 22, 1–15, October 22, 2015 ª2015 Elsevier Ltd All rights reserved

Please cite this article in press as: Badarau et al., Development of Potent and Selective Tissue Transglutaminase Inhibitors: Their Effect on TG2 Func-
tion and Application in Pathological Conditions, Chemistry & Biology (2015), http://dx.doi.org/10.1016/j.chembiol.2015.08.013
Table 1. Continued
6d
0.015 0.007
1d
3d
1.5 0.4
2.1
0
6n
6b
0.00665 0.00021
0.0021 0.0004
1b
3b
3.3 0.3
6.3 2.9
6e
6o
0.0039 0.0004
Lig
0.00875 0.00035
1e
3e
0.889 0.2
0.125 0.066
2e
6f
0.875 0.035
0.00629 0.0005
Lig
O
1f
3f
0.775 0.035
1.625 0.035
(Continued on next page)
Chemistry & Biology 22, 1–15, October 22, 2015 ª2015 Elsevier Ltd All rights reserved
5

Please cite this article in press as: Badarau et al., Development of Potent and Selective Tissue Transglutaminase Inhibitors: Their Effect on TG2 Func-
tion and Application in Pathological Conditions, Chemistry & Biology (2015), http://dx.doi.org/10.1016/j.chembiol.2015.08.013
Table 1. Continued
1g
3.15 0.92
O
Lig
3g
0.9 0.14
1m
0.273 0.11
IC50 values accompanied by the SD were averaged from at least three independent experiments; they were otherwise obtained in a single experiment.
in both water-soluble and lipophilic series were obtained and tion assay. All derivatives from this series proved selective over
evaluated for their activity on a battery of TGs (Table 2A). Com- the TG1, TG3, and FXIIIa TGs with selectivity shown against
pound 3h had an excellent IC50 of ꢁ6 nM for TG2, as well as a TG6 for compounds 1e, 1f, and 3f although less selectivity was
good toxicity profile when tested up to 100 mM (Figure 1E), and shown for TG6 with compound 3e (Table 2B). A moderate TG1
showed selectivity for TG2 when assayed against TG1, TG3, inhibition trend was observed for the halomethylamides. In the
TG6, and Factor XIIIa (FXIIIa). A comparable TG selectivity profile case of TG2, however, the inhibition activities are in the lower
was also shown for 1h with an IC50 value of 380 nM, again with a nanomolar range (>100-fold selectivity ratio). To rule out any
good toxicity profile (Figure S1).
non-specific reaction of the inhibitors with TG1, TG3, and FXIIIa,
Measurement of the Ki and Kinact of 3h and 1h using the gluta- the dansylated probes 3h and 1h were incubated with these
mate dehydrogenase coupled assay indicated high potency, three isoforms in the presence of Ca2+ and with EDTA, enzymes
particularly 3h, with a Kinact of 0.387 minÀ1, a Ki of 1.31 mM, were then separated by denaturing SDS gel electrophoresis
and a Kinact/Ki of 297,692 minÀ1
M
À1. This was comparable and western blotted for the detection of the dansyl group,
with the other high-potency non-dansylated inhibitor 3e which showed negligible reactivity of the TG isoforms with either
(IC50 125 nM), which had a calculated Kinact of 0.275 minÀ1, a inhibitor (Figure S3).
Ki of 1.63 mM, and a Kinact/Ki of 171,875 minÀ1 MÀ1. The other
fluorescent inhibitor, 1h, in line with its lower IC50 (380 nM), had a Further Characterization of the TG2 Inhibitors
Ki of 57 mM, a Kinact of 0.337 minÀ1, and a Kinact/Ki of Toxicity Studies
5,830 minÀ1
M
À1. Given the potential application of 3h and 1h The toxicity profile of some of these new potent inhibitors was
in TG2 conformational studies, it was important to demonstrate evaluated in HUVECs using the XTT-based assay as described
that their reactivity was Ca2+ dependent. For this, recombinant previously (Wang and Griffin, 2013). The toxicity varied between
human TG2 was incubated for 30 min at 4ꢀC with either Ca2+ 25 and 100 mM (Table S1 and Figure S1). As expected, in relation
(1 mM) or GTP (1 mM), and then for a further 30 min in the pres- to their high reactivity, the halomethylketones (6n, 6o) were the
ence of compound 3h (10 nM) or 1h (1 mM). Spin columns were most toxic compounds. The more hydrophilic inhibitors bearing
then used to remove excess inhibitor. Evaluation of TG2 activity the dimethylsulfonium warhead were well tolerated by the
after removal of the excess inhibitor indicated that only the same cell lines with toxicity thresholds of <100 mM. Encouraging
enzyme incubated with GTP was active (Figure 2A), thus con- results were also obtained for the compounds designed to be
firming the requirement of Ca2+ for the inhibitors 3h and 1h to more lipophilic, especially 3e, which has an inhibition activity
access the transamidating active site.
for TG2 in the nanomolar range (IC50 = 0.125
0.066 mM).
To rule out potential off-target effects of the inhibitors, inhibi- Different toxicity profiles were observed between the acryl-
tion of the cysteine proteases Caspase 3 and 7 was undertaken. amide- and vinylsulfonamide-based derivatives (3e vs 2e), which
HUVECs were induced into apoptosis and the caspase activity might be explained in part by their different chemical reactivities.
measured in situ in the presence and absence of 3h and 1h.
No inhibition of Caspase 3 or 7 was shown (Figure 2B). The Confirmation of Covalent Interaction and Targeting
same results were obtained for the inhibitors 1e and 3e, which the Active Site Cys277 by 1h and 3h
carry the dimethylsulfonium and the acrylamide warheads, The fluorescent inhibitor probes allowed us to demonstrate
respectively.
that these compounds were targeting the active site Cys277
Other potent compounds were also tested for their specificity in their covalent reactivity. Pure recombinant wild-type (wt)
toward TG2 compared with other TG family members. Bio- TG2 and its active site mutant Cys277Ser were incubated with
tinylated TG-specific glutamine-containing peptides specific the compound 1h in the presence of Ca2+ for 3 hr. As a positive
for TG family members, including TG1, FXIIIa, TG3, and TG6, control, the wt enzyme was incubated with the competitive
were used in these assays (CovTests; Perez Alea et al., 2009). amine substrate monodansylcadaverine (MDC), which becomes
Comparable IC50 values of these inhibitors for TG2 were de- enzymatically incorporated into available g-glutamyl residues
tected in both the CovTest and the biotin-cadaverine incorpora- on neighboring TG2 molecules.
6
Chemistry & Biology 22, 1–15, October 22, 2015 ª2015 Elsevier Ltd All rights reserved

Please cite this article in press as: Badarau et al., Development of Potent and Selective Tissue Transglutaminase Inhibitors: Their Effect on TG2 Func-
tion and Application in Pathological Conditions, Chemistry & Biology (2015), http://dx.doi.org/10.1016/j.chembiol.2015.08.013
Table 2. Inhibition profiles (IC50) for different TGs for both fluorescent derivatives and other selected potent TG2 inhibitors
(A)
O
O
S
N
N
O
N
N
H
R
O
2
˚
Compound
3h
FXIIIa (mM)
TG1 (mM)
TG3 (mM)
TG6 (mM)
TG2 (mM)
PSA (A )
R
>100
25
20
5
0.0061 0.00042
90
Lig
1h
>100
37
>100
2.2
>100
11
>10
2
0.38 0.057
0.125 0.066
112
70
S
Lig
Lig
3e
(B) Compound
IC50
TG1 (mM)
>100
>100
>100
5
TG3 (mM)
>100
>100
>100
65
TG6 (mM)
NT
FXIIIa (mM)
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
1a
1c
3c
6d
1d
3d
1b
6e
1e
6g
1f
NT
NT
NT
>100
>100
>100
1
>100
>100
>100
70
NT
NT
NT
NT
>100
3
>100
50
>10
NT
>100
>100
>100
>100
>100
>100
>100
>100
>10
>10
NT
3f
3g
1m
NT
For (A) in the case of TG2, the inhibition data (IC50) accompanied by standard deviation (SD) were averaged from at least three independent
experiments using the biotin cadaverine incorporation assay. For TG1, TG3, TG6 and Factor XIIIa the commercial TG-CovTest assay was used (Hitomi
et al., 2009; Fukui et al., 2013) with data obtained from one experiment. The polar surface area (PSA) values were calculated as described in the
Supplemental Experimental Procedures (B) Shows mean IC50 values obtained in a single experiment undertaken in duplicate using the commercial
TG-CovTest assay. NT = not tested.
Following incubation, TG2 was separated on denaturing SDS- based on the ability of compounds to permeate the cell and
PAGE and western blotted using an anti-dansyl antibody (Fig- inhibit intracellular TG2 activity after its activation by the eleva-
ure 2C). Bands were seen in the positive control containing the tion of intracellular Ca2+ using ionomycin. As initially designed,
MDC and in the wt enzyme where the active site Cys277 is intact, the assay confirmed that acrylamide-based inhibitors 3e, 3f
indicating covalent interaction of the inhibitor with TG2. In the and the fluorescent probe 3h penetrate the cell and inhibit intra-
Cys277Ser mutant, no incorporation of MDC was detected in cellular TG2 activity, contrary to the charged compounds 1e and
a band that was indicative for TG2. Presence of the inhibitor 1f (Figure 2D). As a means of validating the permeability assay
was not detected with the Cys277Ser mutant, suggesting that used in these studies, the cell permeability of compounds 3h,
fluorescent inhibitor 1h is specifically targeting the active site 3e, and 1h was also determined in MDCKII-MDR1 cells. The
Cys277.
assay was conducted in the presence of the efflux transporter
inhibitor GF120918. Using this assay, the passive permeability
(P exact) of compounds 3h, 1h, and 3e was 649, 1.3, and
Cellular Permeability
The cellular permeability profile for a subset of compounds was 676 nm/s respectively, confirming that the acrylamide inhibitors
investigated using the irreversible imidazolium-based inhibitor 3h and 3e are cell permeable. In contrast, the positively charged
R283 (Freund et al., 1994), which is cell permeable, and the 1h is cell impermeable. Measurement of the thermodynamic sol-
peptidic inhibitor R281, which is cell impermeable, as reference ubility of these same three compounds in PBS (pH 7.4) over 24 hr
as described previously (Baumgartner et al., 2004). This assay is was 24 mM for 3h, 1594.6 mM for 1h, and 698 mM for 3e.
Chemistry & Biology 22, 1–15, October 22, 2015 ª2015 Elsevier Ltd All rights reserved
7

Please cite this article in press as: Badarau et al., Development of Potent and Selective Tissue Transglutaminase Inhibitors: Their Effect on TG2 Func-
tion and Application in Pathological Conditions, Chemistry & Biology (2015), http://dx.doi.org/10.1016/j.chembiol.2015.08.013
Figure 2. Characterization of the Lead Compounds for their Specificity and Cell Permeability
(A) Columns 2–5 show Ca2+-dependent binding of 3h with rhTG2. rhTG2 pre-treated with Ca2+ or GTP was incubated with 3h (10 nM) or 1h (1 mM) for 30 min at 4ꢀC.
Following removal of excessive inhibitors using 30-kDa cut off spin columns, bars 6 and 7 show the efficiency of the spin columns, where reaction mixes with
inhibitor alone (either 3h or 1h) were prepared and filtered through the spin columns prior to assay. Bars 8–10 show the efficacy of GTP binding, where enzyme was
incubated with 250 mM Ca2+ to show activation of rhTG2 (bar 8) in the absence of GTP and showing that 1 mM Ca2+ activates TG2 with 250 mM GTP present (bar 10).
Samples with 1 mM GTP and 250 mM calcium were prepared in duplicate and one sample was processed through the spin column (bar 10) and the other sample
was not (bar 9) to evaluate whether the spin column affects GTP binding. A positive control incubated without inhibitor was used as 100% reference (bar 1).
(B) Effect of TG2 inhibitors 3h, 1h, 3e, and 1e on Caspase 3/7 activity. Following induction of apoptosis in HUVECs by staurosporine, caspase activity was
measured as described in the Experimental Procedures. Ctrl represents the negative control without any inhibitor; addition of the caspase inhibitor Z-VAD-FMK to
the assay represents the positive control.
(C) Western blotting showing covalent interaction of compound 1h with TG2 and demonstrating targeting of the active site Cys277 by 1h as described in the
Experimental Procedures.
(D) Cell permeability of TG2 inhibitors 3h, 1f, 3f, 3e, 1e in primary HUVECs. The assay measures the inhibition of the TG2-mediated incorporation of
N-(5-aminopentyl)biotinamide into intracellular proteins. R281, which is not cell permeable, was used as the 100% control. Experiments were undertaken as
described in the Experimental Procedures. *p < 0.05.
(E) Detection of dansyl inhibitor 3h in NIH3T3 cells. NIH3T3 cells transduced with TG2 containing Lentivirus were treated with 3h in the presence or absence
of ionomycin and 1e. Co-IP using the cell cytosol fractions was performed to detect the interaction between TG2 and 3h as described in the Experimental
Procedures.
(F) Non-denaturing gel electrophoresis. rhTG2 pre-treated with Ca2+ or GTP (concentrations as indicated in the figure), followed by incubation with 3h or 1h as
described in the Experimental Procedures. The samples were separated by non-denaturing PAGE and visualized using gel imaging.
In Vitro ADME Properties of Compounds 3h, 1h, and 3e
man serum albumin and a-glycoprotein indicated that none of
Further characterization of lead compounds 3h, 1h and 3e by them had very high binding.
in vitro ADME assays are shown in Table S2. Microsomal stability
assessed in human microsomes indicated that compounds Assaying for the Presence of Potential Intracellular
3h and 1h exhibited low and high stability, respectively, with In Situ TG2 Activity/Open Conformation Using Dansyl
compound 3e demonstrating moderate stability. In contrast in Derivative 3h
the cell health screen assessed in the hepatic cell line HepG2, The measured cell permeability of the fluorescent inhibitor
no significant effect was observed for any of the compounds 3h, which requires TG2 to be in its Ca2+ bound open confor-
on mitochondrial potential, nuclear morphology, and membrane mation prior to its irreversible binding, should enable the
permeability. Importantly, no significant effect was observed for detection of TG2 in its open/relaxed conformation inside
hERG liability with any of the compounds and their binding to hu- the cell when combined with fluorescent microscopy and/or
8
Chemistry & Biology 22, 1–15, October 22, 2015 ª2015 Elsevier Ltd All rights reserved

Please cite this article in press as: Badarau et al., Development of Potent and Selective Tissue Transglutaminase Inhibitors: Their Effect on TG2 Func-
tion and Application in Pathological Conditions, Chemistry & Biology (2015), http://dx.doi.org/10.1016/j.chembiol.2015.08.013
co-immunoprecipitation (co-IP) studies. NIH3T3 cells trans- sclerosis. Therefore, we next studied the effect of inhibitor 3h
duced with wt TG2 were used to confirm whether the intracellular on the binding of TG2 to syndecan-4 and its subsequent trans-
reaction of 3h with TG2 could take place. Figure 2E shows that location into the ECM.
inhibitor 3h permeates into NIH3T3 cells and reacts with intracel-
It has been reported that TG2 binding to syndecan-4 is favored
lular TG2, which can be increased further in the presence of the by the compact conformation (Wang et al., 2012) of the enzyme.
Ca2+ ionophore ionomycin (Figure 2E). This is demonstrated by It has been shown in Figure 2F that the inhibitor 3h locks TG2 into
co-IP of the inhibitor/TG2 complex in the cytosolic fraction of its open conformation, which would have a significant impact
cells using an anti-dansyl antibody and TG2 revealed in this com- on the binding of the enzyme to syndecan-4. Therefore, we
plex using an anti-TG2 antibody by western blotting. Importantly, next studied the effect of 3h binding to TG2 on the enzyme’s
the reactivity of 3h with intracellular TG2 was reduced by incu- binding to cell surface heparan sulfates. Figure 3C shows that
bating in the presence of the cell-permeable inhibitor 3e in NIH3T3 cells, previously used to show TG2 binding to cell
(Figure 2E).
surface syndecan-4, 3h significantly reduced the interaction
To confirm that the newly developed fluorescent inhibitors 3h between TG2 and cell surface syndecan-4 following co-IP of
or 1h only interact with TG2 in its open Ca2+ bound conformation TG2 with anti-syndecan-4 antibody as previously documented
and lock the enzyme into this conformation, wt recombinant TG2 (Wang et al., 2012). Importantly, in these same cells, binding of
pre-treated with Ca2+ or GTP was incubated with inhibitor 3h TG2 to 3h led to a dramatic reduction in the presence of the
and then separated by native PAGE. As shown in Figure 2F, 3h enzyme on the cell surface and in the ECM (Figure 3D), which
and 1h in the presence of Ca2+ bind to TG2 and hold it in its paralleled a dramatic reduction in the deposition of fibronectin
open conformation, while the binding of GTP, which holds TG2 (FN) into the ECM (Figure 3D).
in its closed conformation, blocks the interaction of 3h or 1h
with TG2 and gives rise to increased mobility.
Effects of Fluorescent TG2 Inhibitor 3h on Endothelial
HUVECs, which contain high levels of endogenous TG2, were Cell Function
also used to verify our observations in TG2-transduced NIH3T3 In our previous report, using our first generation of peptidic TG
cells. Cells were incubated in the absence or presence of the inhibitors, we showed that extracellular TG2 could be a potential
Ca2+ ionophore ionomycin (positive control). Mouse endothelial target in the treatment of VEGF-induced angiogenesis due to
cells isolated from TG2À/À mice were used as the negative its ability to block the deposition of FN, which is one of the impor-
control together with HUVECs incubated with 3h in the pres- tant ECM proteins required for matrix-bound VEGF signaling
ence of the cell-permeable non-dansylated inhibitor 3e. After (Wang et al., 2013). Given the ability of compound 3h to react
fixation of the cells and removal of any excess inhibitor by with both intracellular and extracellular HUVEC TG2, we tested
washing, fluorescence microscopy revealed a faint intracellular the effects of compound 3h on different but related endothelial
dansyl signal in the HUVECs without ionomycin treatment, cell functions. Figure 4A shows that inhibitor 3h reduced the
which was increased with ionomycin and reduced in the pres- deposition of FN into the ECM by HUVECs, which agrees with
ence of inhibitor 3e. No dansyl signal was detected in the nega- its effect in NIH3T3 cells (Figure 3D). One further role of TG2 in
tive control mouse endothelial cells, which are null for TG2 after endothelial cell angiogenesis is in cell migration during tubule
fixing and washing out any excess inhibitor (Figure 3A). Further formation (Wang et al., 2013), which is dependent on the cross-
confirmation that the dansyl inhibitor is targeting TG2 was ob- linking activity of the enzyme. Figures 4B and 4C demonstrate
tained by co-IP of the dansyl-labeled proteins in the cell cytosol that compound 3h inhibited HUVEC migration in the wound-
fractions of cell lysates and the detection of TG2 by western healing assay and in Matrigel cord formation assays.
blotting. TG2 was found to be present in the HUVECs with or
without ionomycin. The signal was reduced in the cells treated Potency of Inhibitor 3h in an In Vivo Model of
with the inhibitor 3e and, importantly, as found with immunoflu- Hypertensive Nephrosclerosis
orescent microscopy, totally absent in the TG2À/À mouse As a further means of demonstrating the potential application of
endothelial cells (Figure 3B). Hence, in both TG2-transduced these new peptidomimetic inhibitors in human disease where
NIH3T3 cells and in HUVECs, TG2 appears to be present in TG2 has been shown to act extracellularly, we next looked at
the intracellular environment in an open conformation and the effects of inhibitor 3h, which blocks both intracellular and
able to react with inhibitor 3h, suggesting that the enzyme could extracellular TG2 activity, on angiotensin II (AngII)-induced hy-
be active if suitable substrates are available to it, or, dependent pertensive nephrosclerosis in mice. Before undertaking these
on its environment, is simply alternating between the open and studies, we first gained an approximate cell IC50 for 3h by looking
closed conformation.
at the inhibition of FN deposition in HUVECs and in mouse endo-
thelial cells, which in both cases gave an approximate IC50 for 3h
of 0.3 mM in human HUVECs and 0.4 mM in mouse endothelial
cells (Figure S4).
Effect of 3h on TG2 Interaction with Syndecan-4
and its Translocation to the Cell Surface and ECM
The interaction between TG2 and syndecan-4 has been re-
AngII and 3h were delivered using a subcutaneously implanted
ported to be not only essential for the externalization of the osmotic minipump (Azlet) over 14 days. Mice receiving AngII
enzyme onto the cell surface and deposition into the ECM alone together with inhibitor vehicle showed signs of increased
(Wang et al., 2012) but also for its pathological roles in fibrosis collagen deposition around the arterioles and, in some cases,
(Scarpellini et al., 2014). Previous work from us and others in the glomeruli and interstitial tubular space. In contrast, animals
suggested the importance of TG2 crosslinking activity in patho- receiving AngII plus compound 3h showed significantly reduced
logical conditions such as fibrosis, angiogenesis, and multiple (approx. 40%) collagen deposition in all these regions (Figure 5A)
Chemistry & Biology 22, 1–15, October 22, 2015 ª2015 Elsevier Ltd All rights reserved
9

Please cite this article in press as: Badarau et al., Development of Potent and Selective Tissue Transglutaminase Inhibitors: Their Effect on TG2 Func-
tion and Application in Pathological Conditions, Chemistry & Biology (2015), http://dx.doi.org/10.1016/j.chembiol.2015.08.013
Figure 3. Intracellular Reactivity of Fluorescent Compound 3h with TG2 and its Effect on Enzyme Externalization
(A) Detection of in situ TG2 activity using the dansyl derivative 3h, HUVECs and TG2À/À mouse endothelial cells (top panel) were treated as described in the
Experimental Procedures. The cells were pre-treated with (a–i) or without ionomycin (j–l), in the presence (g, h) or absence (a–f and j–l) of non-dansylated cell-
permeable inhibitor 3e, while TG2 antigen was visualized using anti-TG2 antibody Cub7402 (d–l). DAPI is shown in the top right hand corner of each image.
(B) Detection of TG2-dansyl derivative 3h complex in HUVECs and mouse TG2À/À endothelial cells after co-IP and western blotting. HUVECs were treated
as described in (A). Cytosol fractions were collected and co-IP was undertaken using an anti-dansyl antibody as described in the Experimental Procedures.
The presence of TG2 was detected via western blotting. Rabbit IgG and TG2 knockout endothelial cells were used as the negative controls.
(C) Effect of inhibitor 3h on the interaction between TG2 and syndecan-4. Swiss3T3 cells transduced with wt TG2 were pre-treated with inhibitor 3h. Co-IP was
performed using anti-syndecan-4 antibody to pull down the immune complex and the presence of TG2 was detected via western blotting. The empty vector
transduced cells (vector) or the heparan sulfate binding mutant (HS2) TG2-transduced cells were used as the controls.
(D) Effect of 3h on TG2 externalization and deposition. The presence of TG2 in Swiss3T3 cells from (C) was detected via biotinylation of cell surface (CS) proteins
followed by western blotting. The matrix proteins were collected as described in the Experimental Procedures and the presence of TG2 was measured using
anti-TG2 antibody Cub7402 via western blotting. WCL, whole-cell lysate.
with around 45% reduction around the blood vessels (Figure 5B) et al., 2010, 2011; Pardin et al., 2008a, 2008b; Prime et al.,
in the kidneys of inhibitor-treated animals. No obvious signs of 2012; Schaertl et al., 2010). Their pharmacological characteriza-
toxicity were observed in the animals after inhibitor treatment.
tion in terms of potency, specificity, binding-site targeting,
toxicity, and cellular permeability still remains to be revealed.
In addition, a more recent report screening existing TG2 inhibi-
tors based on a weak electrophilic 3-bromo-4,5-dihydroisoxa-
DISCUSSION
In this paper, we report a new series of small-molecule peptido- zole (DHI) scaffold showed poor selectivity against TG1 (Klock
mimetic inhibitors that are selective for TG2 activity. Importantly, et al., 2014). Constructed on a piperazine core, this new scaffold
we also demonstrate the biological impact of these inhibitors and presented here has the advantage of an excellent selectivity
demonstrate their potential application in both cell and animal profile at least against four other members of the TG family
models for the targeting of TG2 in diseases such as fibrosis that were evaluated during our studies (TG1, TG3, TG6, and
and pathological angiogenesis.
FXIIIa). It should be noted, however, that another two groups
Contrary to the classic family of TG2 inhibitors represented (Prime et al., 2012; Wityak et al., 2012) have reported pipera-
mainly by the peptidic inhibitors, this new series adds a new zine-based derivatives with interesting TG2 activities; however,
member to the modest, but continuously increasing, family of Prime et al. (2012) reported that their selectivity over the other
non-peptidic TG2 inhibitors. Sporadic examples have been TGs (especially over FXIIIa) requires further optimization. Wityak
disclosed so far (Duval et al., 2005; Klock et al., 2011; Ozaki et al. (2012) reported that their lead inhibitor had an IC50 for TG2
10 Chemistry & Biology 22, 1–15, October 22, 2015 ª2015 Elsevier Ltd All rights reserved

Please cite this article in press as: Badarau et al., Development of Potent and Selective Tissue Transglutaminase Inhibitors: Their Effect on TG2 Func-
tion and Application in Pathological Conditions, Chemistry & Biology (2015), http://dx.doi.org/10.1016/j.chembiol.2015.08.013
Figure 4. Effect of Compound 3h on Endo-
thelial Cell Migration and Tubule Formation
(A) Representative images showing the effect of 3h
on FN deposition by HUVECs. 7 3 104 HUVECs
were seeded into chamber slides in the presence
or absence of 3h treatment for 72 hr. Immunoflu-
orescence staining was performed to detect the
extracellular FN as described in the Experimental
Procedures. Fluorescence images were captured
using an epifluorescence microscope from three
separate experiments.
(B) Representative images showing the effect of
3h on HUVEC migration in the wound-healing
assay. HUVECs were seeded into 12-well plates
until confluency. Wounds were introduced into
the monocell layer using 200-ml tips. The cells
were allowed to migrate for 6 hr and the closure of
the wounds was analyzed using ImageJ software
(n = 3). *p < 0.05.
(C) Representative images demonstrating the
inhibition of HUVEC cord formation on Matrigel by
3h. 96-well plates were pre-coated with Matrigel
and HUVECs were allowed to form cord structures
for 6 hr. The images from three separate experi-
ments were taken using a Nikon digital camera.
(Wang and Griffin, 2012; Wang et al.,
2013). The development of a highly potent
cell-permeable fluorescent inhibitor spe-
cific for TG2, allowed us to test for the
of 14 nM with selectivity over TG1, TG3, TG6, and FXIIIa but no first time whether TG2, when inside the cell, can indeed exist in
cell or animal studies were undertaken on these compounds both its potentially active open conformation as well as its
although in vitro DMPK studies indicated they had low toxicity; closed/compact conformation. Our data suggest that reactivity
the authors indicated that improved ADME profiling was still of TG2 with the fluorescent inhibitor probe can take place in
required for in vivo applications because of their low plasma the low Ca2+ intracellular environment of both HUVECs and
stability. Importantly, the straightforward synthetic route of the NIH3T3 cells, which can be further enhanced by using the Ca2+
compounds presented in this report gives easy access to various ionophore ionomycin. Given that the reported Km app for Ca2+
modulations, allowing us to improve the toxicity profile of our for TG2 is around 3 mM (Hand et al., 1985), this suggests that
inhibitors up to 100 mM/72 hr, while controlling their cellular the conformation of the enzyme may also be regulated by other
access. However, like the compounds reported by Wityak binding factors in addition to Ca2+, which permits flexibility be-
et al. (2012), further improvements on ADME characteristics tween the open and compact conformation and which, when in
are still necessary for some of our lead compounds, in particular the open conformation, allows the irreversible reaction with the
improvements in microsomal stability and aqueous solubility fluorescent TG2 inhibitor probe. Earlier fluorescence resonance
for 3h, although solubility may not be such an issue for in vivo po- energy transfer studies with TG2 have also suggested that,
tency given that the solubility of 3h is >3,000 times the IC50 of 3h. in certain localities of the cell, the enzyme may exist in its
Moreover, the poor microsomal stability of 3h is unlikely to be open conformation (Pavlyukov et al., 2012). This may explain a
due to the piperazinyl moiety as suggested by Wityak et al. number of observations where TG2 has been reported to regu-
(2012) given that compound 1h, which is a comparable structure late a several cellular processes either through conformational
apart from the warhead, is microsomal stable.
Physiologically, TG2 is an enzyme that exists both in the intra- via its transamidating activity (Colak et al., 2011).
and extracellular environment. Inside the cell, its Ca2+-depen-
As referred to earlier, TG2 can be secreted into the ECM by a
changes that regulate protein binding (Zhang et al., 2013) or
dent crosslinking activity is thought to be tightly controlled by non-conventional mechanism, which requires the cell surface
the binding of GTP/GDP, which holds the enzyme in a compact binding of heparan sulfates, such as syndecan-4, for its trans-
closed inactive conformation (Bergamini et al., 2011; Wang and location into the ECM. This binding of TG2 to syndecan-4 at
Griffin, 2012). However, TG2 is reported to be involved in a the cell surface is favored by the enzyme being in its compact
number of human diseases where it acts both intracellularly closed conformation, suggesting that the enzyme is secreted
(e.g. cystic fibrosis, cancer progression, epidermal growth in this conformation or that this conformation is assumed once
factor receptor recycling) (Bergamini et al., 2011; Zhang et al., it reaches the cell surface (Wang et al., 2012). Therefore,
2013) and/or extracellularly, including matrix deposition, tissue our next question was to ask if the binding of an irreversible inhib-
fibrosis, VEGF-dependent angiogenesis, and multiple sclerosis itor inside the cell, which would hold the enzyme in its open
Chemistry & Biology 22, 1–15, October 22, 2015 ª2015 Elsevier Ltd All rights reserved 11

Please cite this article in press as: Badarau et al., Development of Potent and Selective Tissue Transglutaminase Inhibitors: Their Effect on TG2 Func-
tion and Application in Pathological Conditions, Chemistry & Biology (2015), http://dx.doi.org/10.1016/j.chembiol.2015.08.013
Figure 5. TG2 Inhibition Attenuates Angiotensin II-Induced Renal Fibrosis
Two groups of animals were used. In the first group, five C57BL/6 male mice (14 weeks old) were infused with AngII 1.1 mg/kg/day in 50% DMSO in PBS (pH 7.4)
for 2 weeks without TG2 inhibitor via a subcutaneously implanted osmotic pump (Alzet 1002) as previously described (Murdoch et al., 2014). In the second group,
compound 3h was infused in a similar manner at 25 mg/kg/day in 50% DMSO in PBS (pH 7.4) together with AngII over 14 days.
(A) Representative images of kidneys from the inhibitor- (lower panel) and non-inhibitor-(upper panel) treated animals.
(B and C) Averaged data of Sirius red/collagen staining in kidney sections showing staining (B) around arterioles (n = 4) and (C) total collagen staining (n = 5).
*p < 0.05.
conformation, would interfere with both its translocation to the regions of the kidney with around a 45% reduction in collagen
cell surface and its deposition into the ECM. deposition around the kidney arterioles when compared with
Using NIH3T3 cells, we clearly show that treatment of these mice receiving AngII and vehicle.
cells with the fluorescent permeable inhibitor 3h reduces the
Hence, our findings demonstrate that our new peptidomimetic
presence of TG2 at the cell surface and nearly completely blocks TG2 small-molecule inhibitors, which can react with TG2 inside
translocation of the enzyme into the ECM, a finding that is the cell and prevent export of the enzyme as well as block its
accompanied by the reduced deposition of FN. In HUVECs, activity, provide an ideal therapeutic avenue for the treatment
where we know TG2 also reacts with our fluorescent probe in- of human diseases involving TG2.
side the cell, we also observed a large reduction in the deposition
of FN in the presence of this TG2 inhibitor. In relation to biological SIGNIFICANCE
functions of endothelial cells, this intracellular acting probe is
also a potent inhibitor of endothelial cord formation on Matrigel TG2 has been shown to be essential in a number of human
and a potent inhibitor of endothelial motility, which is required disease states, including fibrosis and angiogenesis. How-
for endothelial tube formation (Wang et al., 2013). As a further ever, other transglutaminase family members sharing a
means of demonstrating the potential application of the new similarity in their structures generate obstacles in gener-
peptidomimetic inhibitors in in vivo studies in a disease state ating TG2-specific inhibitors with high potency and low
where TG2 acts extracellularly, we next looked at the effects toxicity. Using molecular dynamics techniques, a series of
of inhibitor 3h on AngII-induced hypertensive nephrosclerosis new peptidomimetic inhibitors against TG2 were produced
in mice. Mice receiving AngII plus compound 3h showed sig- with excellent selectivity when tested against four other
nificantly reduced (approx. 40%) collagen deposition across all members of the transglutaminase family (TG1, TG3, TG6,
12 Chemistry & Biology 22, 1–15, October 22, 2015 ª2015 Elsevier Ltd All rights reserved

Please cite this article in press as: Badarau et al., Development of Potent and Selective Tissue Transglutaminase Inhibitors: Their Effect on TG2 Func-
tion and Application in Pathological Conditions, Chemistry & Biology (2015), http://dx.doi.org/10.1016/j.chembiol.2015.08.013
the presence of polyamine substrates immobilized onto 96-well microplates.
and FXIIIa). A highly potent TG2 cell-permeable fluorescent
probe (IC50 = 6 nM; Kinact/Ki of 297,692 min–1 M–1) was de-
The incorporated biotinylated peptides were measured using horseradish
peroxidase-conjugated streptavidin and then measured using o-phenylenedi-
signed and exploited in several biological studies. We
amine dihydrochloride substrates. The absorbance was measured at 490 nm
demonstrate that the reactivity of the inhibitor can take
place with TG2 in the low Ca2+ intracellular environment
using a microplate reader.
of HUVECs and NIH3T3 cells, which is enhanced by using
the Ca2+ ionophore ionomycin. Importantly, reaction of
TG2 with this fluorescent inhibitor led to a loss of binding
of TG2 to cell surface syndecan-4 in NIH3T3 cells and inhibi-
tion of TG2 translocation into the ECM, which was paralleled
by a reduction in the deposition of matrix FN. Further studies
with HUVECs treated with this cell-permeable fluorescent
inhibitor revealed a significant reduction in FN deposition,
cord formation on Matrigel, and cell motility. As a means
of demonstrating the potential application of the new pepti-
domimetic inhibitors in vivo, where TG2 acts extracellularly
(Huang et al., 2009; Johnson et al., 2007), the effects of inhib-
itor 3h on Angiotensin II (AngII)-induced hypertensive neph-
rosclerosis in mice was investigated. Mice receiving AngII
plus compound 3h showed significantly reduced (approx.
40%) collagen deposition across all regions of the kidney
with around 45% reduction around blood vessels. Hence,
this new series of TG2-specific cell-permeable peptidomi-
metic inhibitors may be highly effective in targeting human
diseases involving extracellular TG2 via their ability to block
both protein crosslinking activity and enzyme export.
Glutamate Dehydrogenase Coupled Assay Used for Kinact/Ki
Calculations
TG2 inhibitors were assayed in accordance with the method of Choi et al.
(2005) using the glutamate dehydrogenase coupled assay. Briefly, a reaction
mixture containing 4 mM CaCl2, 10 mM a-ketoglutarate, 0.2 U/ml glutamate
dehydrogenase, 0.12 mM NADH, 20 mg TG2, and 10 mM Cbz-Gln-Gly in
200 mM 3-(N-morpholino)ethanesulfonic acid (pH 7.1) was used in the assay.
The enzyme reaction was initiated by the addition of TG2 and the consumption
of NADH was monitored using different concentrations of inhibitor (10–500 nM)
by UV spectroscopy (340 nm, ε = 6,220 cmÀ1 MÀ1). Kinetic parameters (Kinact
and Ki) were obtained using the Dynafit 4.05.129 software (Kuzmic, 1996,
2009).
Cell Permeability Assay
By Measuring Inactivation of Intracellular TG2
To induce intracellular TG2 activity, HUVECs with high endogenous TG2 were
incubated with fresh growth medium with 0.5% FBS containing 1 mM ionomy-
cin and 1 mM biotin-cadaverine (Zedira) in the presence or absence of 50 mM
of TG2 inhibitors 3h, 1f, 3f, 3e, and 1e. The non-cell-permeable inhibitor R281
and the cell-permeable inhibitor R283 were used as the negative and positive
controls, respectively. After a 3-hr incubation, the cells were collected in
homogenization buffer (50 mM Tris-HCl [pH 7.5], 150 mM NaCl, 1 mM
EDTA) and sonicated on ice. The cell membrane was pelleted at 60,000 3 g
for 60 min and the supernatant (the cytosol fraction) was used for the assay.
50 mg of total protein in 100 ml of the homogenization buffer was used in the
biotin-cadaverine incorporation assay as described previously by Wang
et al. (2012).
EXPERIMENTAL PROCEDURES
By the In Vitro Passive Cellular Permeability Assay
Synthetic and Analytical Methodology
The permeability (P exact) of 3h, 1h, 3e across an MDCK-MDR1 cell monolayer
was measured at a starting concentration of 3 mM in the presence of
GF120918, an efflux inhibitor. The pH of the donor and receiver compartments
was 7.4 (Hank’s balanced salt solution).
The newly synthesized derivatives were characterized by IR spectroscopy,
proton and carbon-13 nuclear magnetic resonance spectroscopy (1H- and
13C-NMR), melting points, high resolution mass spectrometry, and had prop-
erties consistent with the proposed structures (see Supplemental Experi-
mental Procedures).
Incubations were carried out in an atmosphere of 5% CO2 with a relative hu-
midity of 95% at 37ꢀC for 60 min. Apical and basolateral samples were diluted
for analysis by liquid chromatography-tandem mass spectrometry. The integ-
rity of the monolayers throughout the experiment was checked by monitoring
Lucifer yellow permeation using fluorimetric analysis (Tran et al., 2004).
Biological Methodology
General experimental procedures were performed as described previously
(Wang et al., 2012, 2013, Wang and Griffin, 2012) and experimental information
is detailed in the Supplemental Experimental Procedures.
Western Blotting for the Dansyl Inhibitor Interaction with rhTG2
1 mg of recombinant human TG2 (rhTG2), wt or the active site C277S mutant
protein, was incubated with the dansylated TG2 inhibitor 1h at 100 mM and
50 mM in the presence of 10 mM Ca2+ and 1 mM DTT in Tris-HCl buffer
(pH 7.4) for 30 min at room temperature. The primary amine substrate MDC,
which is incorporated into the available g-glutamyl residues of TG2 by the
enzyme, was used as the positive signal control sample, while non-labeled
rhTG2 was used as the negative control. The reaction was stopped by diluting
the samples into 23 Laemmli buffer and boiling at 95ꢀC for 5 min. The pres-
ence of dansyl was detected via western blotting (Wang et al., 2012, 2013)
by using a specific anti-dansyl antibody (Invitrogen).
Reagents and Antibodies
The general reagents were purchased from Sigma-Aldrich (UK), unless stated
otherwise.
Chemical Synthesis of the Inhibitors
The synthesis of the inhibitors is described in the Supplemental Experimental
Procedures.
Cell Culture
HUVECs (Lonza) (Wang et al., 2013) and NIH3T3 cells (Wang et al., 2012) were
cultured as described previously (Supplemental Experimental Procedures).
The Targeting of In Situ TG2 Activity by Fluorescence Microscopy
Using the Dansyl Inhibitor 3h
TG2 Activity Assays
For measurement of TG2 activity, the incorporation of biotinylated cadaverine
into immobilized N,N-dimethylcasein was used as described previously (Wang
et al., 2012).
HUVECs and TG2À/À mouse endothelial cells were used in this study.
HUVECs were treated with 1 mM ionomycin in the presence or absence of
50 mM of 3e to pre-block the intracellular active TG2 for 3 hr, and the cells
were further treated with 5 mM 3h for another 2 hr. TG2À/À endothelial cells
were treated in a comparable manner but were not exposed to the inhibitor
3e. For detection of endogenous TG2 activity, HUVECs were incubated
without ionomycin. After the incubation, the cells were washed three times
with pre-chilled methanol and fixed with methanol for 20 min at À20ꢀC.
Following blocking with 3% BSA in PBS (pH 7.4) for 30 min at 37ꢀC, the
cells were incubated with rabbit anti-dansyl antibody and revealed using a
For determination of the activity of TG1, TG3, TG6, and FXIIIa, commercial
microassays were used (TG-CovTest; Covalab) (Hitomi et al., 2009), according
to the manufacturer’s instructions. For comparable purposes, a number of
TG2 assays were also undertaken using this assay. Briefly, TG-specific bio-
tinylated peptides, including pepF11KA (FXIII pre-activated with thrombin),
pepK5 (TG1), pepT26 (TG2), pepE51 (TG3), or pepY25 (TG6) (Hitomi et al.,
2009; Fukui et al., 2013) were incubated with suitable TG family members in
Chemistry & Biology 22, 1–15, October 22, 2015 ª2015 Elsevier Ltd All rights reserved 13

Please cite this article in press as: Badarau et al., Development of Potent and Selective Tissue Transglutaminase Inhibitors: Their Effect on TG2 Func-
tion and Application in Pathological Conditions, Chemistry & Biology (2015), http://dx.doi.org/10.1016/j.chembiol.2015.08.013
secondary tetramethylrhodamineisothiocyanate-conjugated antibody for a
2-hr incubation at 37ꢀC. The fluorescence signal was visualized using a Zeiss
epifluorescent microscope using a 603 objective.
research grant (‘‘TRANSPATH’’, FP7 No 289964). None of the authors of this
manuscript have a financial interest related to this work.
Received: April 9, 2015
Revised: August 4, 2015
Accepted: August 18, 2015
Published: October 8, 2015
Detection of TG2 Activity Using Co-IP
Immunoprecipitation of the intracellular TG2-dansyl inhibitor complex was
performed following the treatment of cells with 3h with or without pre-treat-
ment of 50 mM 3e (as described above). Cytosol fractions of HUVECs or
NIH3T3 cells transduced with TG2 were collected as described above.
Anti-TG2 antibody was used to pull down the TG2 immuno-complex and
SDS-PAGE and western blotting were then performed to detect the presence
of the dansyl group of 3h. TG2 knockout endothelial cells or non-TG2
transduced NIH3T3 cells, as well as a mouse IgG negative control antibody,
were used as the negative controls.
REFERENCES
Badarau, E., Collighan, R.J., and Griffin, M. (2013a). Recent advances in the
development of tissue transglutaminase (TG2) inhibitors. Amino Acids 44,
119–127.
Badarau, E., Mongeot, A., Collighan, R.J., Rathbone, D.L., and Griffin, M.
(2013b). Imidazolium-based warheads strongly influence activity of water-sol-
uble peptidic transglutaminase inhibitors. Eur. J. Med. Chem. 66, 526–530.
Animal Study
Animal studies were undertaken with full ethical approval and all procedures
carried out under license according to regulations laid down by Her Majesty’s
Government, United Kingdom (Animals Scientific Procedures Act, 1986) and
undertaken in accordance with the 3Rs. Two groups of five animals were
used. In the first group, C57BL/6 male mice (14 weeks old) were infused
with AngII 1.1 mg/kg/day in 50% DMSO in PBS (pH 7.4) for 2 weeks without
TG2 inhibitor via a subcutaneously implanted osmotic pump (Alzet 1002) as
previously described (Murdoch et al., 2014). In the second group, compound
3h was infused in a similar manner at 25 mg/kg/day together with AngII in 50%
DMSO in PBS (pH 7.4) over 14 days. After 14 days, kidneys from the treated
mice were frozen in OCT (CellPath) at an optimal temperature (À150⁰C) using
isopentane; subsequently, 6-mm sections were prepared. Picrosirius red
staining was preformed to measure collagen staining as previously reported
(Murdoch et al., 2014). The collagen content was quantified as the percentage
area in a minimum of ten images per mouse using Image-Pro Plus version
5.0 (Media Cybernetics). The concentration (270 mM) of inhibitor used in the
pump was approximated from computer simulations to give the steady state
concentration in the mouse assuming an extensive distribution of the drug at
5 mM with a half-life of 1 hr.
Baumgartner, W., Golenhofen, N., Weth, A., Hiiragi, T., Saint, R., Griffin, M.,
and Drenckhahn, D. (2004). Role of transglutaminase 1 in stabilisation of
intercellular junctions of the vascular endothelium. Histochem. Cell Biol. 122,
17–25.
Bergamini, C.M., Collighan, R.J., Wang, Z., and Griffin, M. (2011). Structure
and regulation of type 2 transglutaminase in relation to its physiological
functions and pathological roles. Adv. Enzymol. Relat. Areas Mol. Biol. 78,
1–46.
Colak, G., Keillor, J.W., and Johnson, G.V. (2011). Cytosolic guanine nucledo-
tide binding deficient form of transglutaminase 2 (R580a) potentiates cell death
in oxygen glucose deprivation. PLoS One 6, e16665.
Choi, K., Siegel, M., Piper, J.L., Yuan, L., Cho, E., Strnad, P., Omary, B., Rich,
K.M., and Khosla, C. (2005). Chemistry and biology of dihydroisoxazole
derivatives: selective inhibitors of human transglutaminase 2. Chem. Biol.
12, 469–475.
Duval, E., Case, A., Stein, R.L., and Cuny, G.D. (2005). Structure-activity rela-
tionship study of novel tissue transglutaminase inhibitors. Bioorg. Med. Chem.
Lett. 15, 1885–1889.
Freund, K.F., Doshi, K.P., Gaul, S.L., Claremon, D.A., Remy, D.C., Baldwin,
J.J., Pitzenberger, S.M., and Stern, A.M. (1994). Transglutaminase inhibition
by 2-[(2-oxopropyl)thio]imidazolium derivatives: mechanism of factor XIIIa
inactivation. Biochemistry 33, 10109–10119.
Statistical Analysis
Data are expressed as the mean SD for at least three independent replicate
experiments (n R 3) unless otherwise stated. Statistical tests were undertaken
using InStat (GraphPad). Statistical analysis of the results was undertaken
using one-way ANOVA using a post-test depending on the requirement.
Fukui, M., Kuramoto, K., Yamasaki, R., Shimizu, Y., Itoh, M., Kawamoto, T.,
and Hitomi, K. (2013). Identification of a highly reactive substrate peptide
for transglutaminase 6 and its use in detecting transglutaminase activity in
the skin epidermis. FEBS J. 280, 1420–1429.
SUPPLEMENTAL INFORMATION
Griffin, M., Casadio, R., and Bergamini, C.M. (2002). Transglutaminases:
Supplemental Information includes Supplemental Experimental Procedures,
four figures, and two tables and can be found with this article online at
http://dx.doi.org/10.1016/j.chembiol.2015.08.013.
nature’s biological glues. Biochem. J. 368, 377–396.
Griffin, M., Mongeot, A., Collighan, R., Saint, R.E., Jones, R.A., Coutts, I.G.,
and Rathbone, D.L. (2008). Synthesis of potent water-soluble tissue transglu-
taminase inhibitors. Bioorg. Med. Chem. Lett. 18, 5559–5562.
AUTHOR CONTRIBUTION
Han, B.G., Cho, J.W., Cho, Y.D., Jeong, K.C., Kim, S.Y., and Lee, B.I. (2010).
Crystal structure of human transglutaminase 2 in complex with adenosine
triphosphate. Int. J. Biol. Macromol. 47, 190–195.
M.G., Z.W., and D.R. designed the experiments. E.B. performed the com-
pound synthesis and the modeling, with D.R. Z.W., A.C., T.T., C.M., M.B.,
and S.E.L. performed the biological experiments, and all authors analyzed
the data. M.G., Z.W., E.B., and D.R wrote the manuscript.
Hand, D., Bungay, P.J., Elliott, B.M., and Griffin, M. (1985). Activation of
transglutaminase at calcium levels consistent with a role for this enzyme as
a calcium receptor protein. Biosci. Rep. 5, 1079–1086.
ACKNOWLEDGMENTS
Hitomi, K., Kitamura, M., and Sugimura, Y. (2009). Preferred substrate
sequences for transglutaminase 2: screening using a phage-displayed peptide
library. Amino Acids 36, 619–624.
The authors acknowledge EPSRC National Mass Spectrometry Centre, Swan-
sea (UK) for recording the high resolution mass spectra. To GlaxoSmithKline,
UK, for undertaking the in vitro DMPK assays. Prof Soichi Kojima’s group
(Riken, Japan) for kindly providing the TG2 knockout endothelial cells. To Dr
Raj K. Singh Badhan (Aston University) for help in calculating the inhibitor
concentration used in the in vivo experiments following pharmokinetic simula-
tions. Fruitful discussions with Dr Alexandre Mongeot and Dr Russell Collighan
(Aston University) are also acknowledged. M.G. was the major grant holder
that funded the work. E.B. was supported by a Marie Curie research grant
(‘‘TRANSCOM’’, FP7 No: 251506). A.C. was supported by a Marie Curie
Huang, L., Haylor, J.L., Hau, Z., Jones, R.A., Vickers, M.E., Wagner, B., Griffin,
M., Saint, R.E., Coutts, I.G., El Nahas, A.M., et al. (2009). Transglutaminase
inhibition ameliorates experimental diabetic nephropathy. Kidney Int. 76,
383–394.
Johnson, T.S., Fisher, M., Haylor, J.L., Hau, Z., Skill, N.J., Jones, R., Saint, R.,
Coutts, I., Vickers, M.E., El Nahas, A.M., et al. (2007). Transglutaminase inhibi-
tion reduces fibrosis and preserves function in experimental chronic kidney
disease. J. Am. Soc. Nephrol. 18, 3078–3088.
14 Chemistry & Biology 22, 1–15, October 22, 2015 ª2015 Elsevier Ltd All rights reserved

Please cite this article in press as: Badarau et al., Development of Potent and Selective Tissue Transglutaminase Inhibitors: Their Effect on TG2 Func-
tion and Application in Pathological Conditions, Chemistry & Biology (2015), http://dx.doi.org/10.1016/j.chembiol.2015.08.013
Klock, C., Jin, X., Choi, K., Khosla, C., Madrid, P.B., Spencer, A., Raimundo,
B.C., Boardman, P., Lanza, G., and Griffin, J.H. (2011). Acylideneoxoindoles:
a new class of reversible inhibitors of human transglutaminase 2. Bioorg.
Med. Chem. Lett. 21, 2692–2696.
Pinkas, D.M., Strop, P., Brunger, A.T., and Khosla, C. (2007).
Transglutaminase 2 undergoes a large conformational change upon activa-
tion. PLoS Biol. 5, 2788–2795.
Prime, M.E., Andersen, O.A., Barker, J.J., Brooks, M.A., Cheng, R.K.,
Toogood-Johnson, I., Courtney, S.M., Brookfield, F.A., Yarnold, C.J.,
Marston, R.W., et al. (2012). Discovery and structure-activity relationship
Klock, C., Herrera, Z., Albertelli, M., and Khosla, C. (2014). Discovery of potent
and specific dihydroisoxazole inhibitors of human transglutaminase 2. J. Med.
Chem. 57, 9042–9064.
of potent and selective covalent inhibitors of transglutaminase
2 for
Huntington’s disease. J. Med. Chem. 55, 1021–1046.
Kuzmic, P. (1996). Program DYNAFIT for the analysis of enzyme kinetic data:
application to HIV proteinase. Anal. Biochem. 237, 260–273.
Scarpellini, A., Huang, L., Burhan, I., Schroeder, N., Funck, M., Johnson, T.S.,
and Verderio, E.A. (2014). Syndecan-4 knockout leads to reduced extracellular
transglutaminase-2 and protects against tubulointerstitial fibrosis. J. Am. Soc.
Nephrol. 25, 1013–1027.
Kuzmic, P. (2009). DynaFit–a software package for enzymology. Methods
Enzymol. 467, 247–280.
Liu, S., Cerione, R.A., and Clardy, J. (2002). Structural basis for the guanine
nucleotide-binding activity of tissue transglutaminase and its regulation of
transamidation activity. Proc. Natl. Acad. Sci. USA 99, 2743–2747.
Schaertl, S., Prime, M., Wityak, J., Dominguez, C., Munoz-Sanjuan, I., Pacifici,
R.E., Courtney, S., Scheel, A., and Macdonald, D. (2010). A profiling platform
for the characterization of transglutaminase 2 (TG2) inhibitors. J. Biomol.
Screen. 15, 478–487.
`
Lortat-Jacob, H., Burhan, I., Scarpellini, A., Thomas, A., Imberty, A., Vives,
R.R., Johnson, T., Gutierrez, A., and Verderio, E.A. (2012). Transglutaminase-
2 interaction with heparin: identification of a heparin binding site that regulates
cell adhesion to fibronectin-transglutaminase-2 matrix. J. Biol. Chem. 287,
18005–18017.
Schoellmann, G., and Shaw, E. (1963). Direct evidence for the presence of
histidine in the active center of chymotrypsin. Biochemistry 2, 252–255.
Tran, T.T., Mittal, A., Gales, T., Maleeff, B., Aldinger, T., Polli, J.W., Ayrton, A.,
Ellens, H., and Bentz, J. (2004). Exact kinetic analysis of passive transport
across a polarized confluent MDCK cell monolayer modeled as a single barrier.
J. Pharm. Sci. 93, 2108–2123.
Murdoch, C.E., Chaubey, S., Zeng, L., Yu, B., Ivetic, A., Walker, S.J.,
Vanhoutte, D., Heymans, S., Grieve, D.J., Cave, A.C., et al. (2014). Edothelial
NADPH oxidase-2 promotes interstitial cardiac fibrosis and diastolic dysfunc-
tion through proinflammatory effects and endothelial-mesenchymal transition.
J. Am. Coll. Cardiol. 63, 2734–2741.
Wang, Z., and Griffin, M. (2012). TG2, a novel extracellular protein with multiple
functions. Amino Acids 42, 939–949.
Ozaki, S., Ebisui, E., Hamada, K., Goto, J., Suzuki, A.Z., Terauchi, A., and
Mikoshiba, K. (2010). Potent transglutaminase inhibitors, aryl beta-aminoethyl
ketones. Bioorg. Med. Chem. Lett. 20, 1141–1144.
Wang, Z., and Griffin, M. (2013). The role of TG2 in regulating S100A4-medi-
ated mammary tumour cell migration. PLoS One 8, e57017.
Wang, Z., Collighan, R.J., Pytel, K., Rathbone, D.L., Li, X., and Griffin, M.
(2012). Characterization of heparin-binding site of tissue transglutaminase:
its importance in cell surface targeting, matrix deposition, and cell signaling.
J. Biol. Chem. 287, 13063–13083.
Ozaki, S., Ebisui, E., Hamada, K., Suzuki, A.Z., Terauchi, A., and Mikoshiba, K.
(2011). Potent transglutaminase inhibitors, dithio beta-aminoethyl ketones.
Bioorg. Med. Chem. Lett. 21, 377–379.
Pardin, C., Pelletier, J.N., Lubell, W.D., and Keillor, J.W. (2008a). Cinnamoyl in-
hibitors of tissue transglutaminase. J. Org. Chem. 73, 5766–5775.
Wang, Z., Perez, M., Caja, S., Melino, G., Johnson, T.S., Lindfors, K., and
Griffin, M. (2013). A novel extracellular role for tissue transglutaminase in
matrix-bound VEGF-mediated angiogenesis. Cell Death Dis. 4, e808.
Pardin, C., Roy, I., Lubell, W.D., and Keillor, J.W. (2008b). Reversible and
competitive cinnamoyl triazole inhibitors of tissue transglutaminase. Chem.
Biol. Drug Des. 72, 189–196.
Wityak, J., Prime, M.E., Brookfield, F.A., Courtney, S.M., Erfan, S., Johnsen,
S., Johnson, P.D., Li, M., Marston, R.W., Reed, L., et al. (2012). SAR develop-
Pavlyukov, M.S., Antipova, N.V., Balashova, M.V., and Shakhparonov, M.I.
(2012). Detection of transglutaminase 2 conformational changes in living cell.
Biochem. Biophys. Res. Commun. 421, 773–779.
ment of lysine-based irreversible inhibitors of transglutaminase
2 for
Huntington’s disease. ACS Med. Chem. Lett. 3, 1024.
Perez Alea, M., Kitamura, M., Martin, G., Thomas, V., Hitomi, K., and El Alaoui,
S. (2009). Development of an isoenzyme-specific colorimetric assay for tissue
transglutaminase 2 cross-linking activity. Anal. Biochem. 389, 150–156.
Zhang, J., Antonyak, M.A., Singh, G., and Cerione, R.A. (2013). A mechanism
for the upregulation of EGF receptor levels in glioblastomas. Cell Rep. 3, 2008–
2020.
Chemistry & Biology 22, 1–15, October 22, 2015 ª2015 Elsevier Ltd All rights reserved 15