3-Substituted imidazo[1,2-d][1,2,4]-thiadiazoles: A novel class of factor XIIIa inhibitors

Article abstract of DOI:10.1021/jm049221w

A new class of selective FXIIIa inhibitors with a bicyclic [1,2,4]-thiadiazole pharmacophore is described. At 160 μM, compound 8 caused 50% reduction in fibrin γ-chain crosslinking and suppressed the polymerization of a?chains in platelet-depleted human p

Full text of DOI:10.1021/jm049221w

2266
J. Med. Chem. 2005, 48, 2266-2269
Chart 1. Examples of Non-Peptide Inhibitors of Factor
3-Substituted
Imidazo[1,2-d][1,2,4]-thiadiazoles: A
Novel Class of Factor XIIIa Inhibitors
XIIIa via Modification of Active Site Cysteine Thiol13-15
Regis Leung-Toung,† Tim F. Tam,*,†
Jolanta M. Wodzinska,*,‡ Yanqing Zhao,†
Jayme Lowrie,‡ Craig D. Simpson,†
Khashayar Karimian,† and Michael Spino†
Chart 2. Inhibition of Thiol-Dependent Enzymes by
3-Substituted Imidazo[1,2-d][1,2,4]-thiadiazoles
Departments of Medicinal Chemistry and Preclinical
Discovery, Apotex Research Inc., 400 Ormont Drive,
Toronto, ON M9L 1N9, Canada
Received September 24, 2004
Abstract: A new class of selective FXIIIa inhibitors with a
bicyclic [1,2,4]-thiadiazole pharmacophore is described. At 160
µM, compound 8 caused 50% reduction in fibrin γ-chain cross-
linking and suppressed the polymerization of R chains in
platelet-depleted human plasma clots. Fibrinolysis rates in
response to tissue plasminogen activator were directly pro-
portional to the concentration of 8 in plasma at the time of
clotting.
matrix proteins such as fibronectin, vitronectin, and
collagen, thereby anchoring the clot to the blood vessel
wall.2,3
Factor XIIIa inhibitors do not prevent fibrin clot
formation, but they impede the formation of highly
cross-linked, polymeric “hard” clots. Therefore, specific
FXIIIa inhibitors can potentially provide a safer alter-
native to anticoagulants for the therapeutic treatment
of thrombosis,7 atherosclerosis,8 and coronary heart
diseases.9 Several FXIIIa inhibitors (e.g., the polypep-
tide tridegin,10 the low molecular weight non-peptide
ZG-1400,11 cerulenin,12 and alutacenoic acids A and
B13,14) have been discovered through screening of natu-
ral products. Surprisingly, very few specific non-peptide
inhibitors of FXIIIa have been reported to date.13-17
Interestingly, 115,16 and 213,14 (Chart 1) inactivate FXIIIa
through modification of the active site thiol.
Factor XIII (FXIII) is involved in the final step of the
blood coagulation cascade, and its main function is to
cross-link adjacent fibrin fibers into fibrin clot in
response to an external or an internal injury to the blood
vessels.1-3 FXIII exists as a heterotetramer of two A and
two B subunits in the circulating plasma and as a dimer
of two A subunits in platelets, placenta, and other
tissues. The X-ray crystal structures of several forms
of FXIII A subunit revealed a conserved Cys314-His373-
Asp396 triad in the active site.4,5 The active site cysteine
is not accessible for catalysis in the zymogen form.5 In
the presence of Ca2+, thrombin mediates the conversion
of FXIII to the active form FXIIIa. The large change in
protein conformation that occurs upon enzyme activa-
tion allows access to the active site and hence the active
site cysteine. Both the platelet and the plasma enzymes
are identical upon activation.2-6
We18 and others19 have recently disclosed the mono-
cyclic [1,2,4]-thiadiazoles as cysteine protease inhibitors.
It has been postulated that the active site Cys314 of
FXIIIa could be accessed from a long groove that runs
along the subunit-subunit interface or by bending along
the â-sandwich domain.5 A binding form model of FXIIIa
was also built using data from the X-ray crystal
structure of the zymogen.14 Furthermore, monotosylca-
daverine (Tos-NH(CH2)5NH2) and its derivatives are
known potent pseudo substrate inhibitors of trans-
glutaminases including FXIIIa.20,21 We reasoned that
compounds (Chart 2) that incorporated the [1,2,4]-
thiadiazole pharmacophore18,22 substituted at the 3
position with an amino anchor followed by a hydropho-
bic (CH2)n (n ) 3-10) spacer and terminated with an
amino sulfone moiety could be promising FXIIIa inhibi-
tors.
Herein, we describe specific inhibitors of FXIIIa based
on a novel class of thiol-trapping pharmacophore, the
3-substituted imidazo[1,2-d]-1,2,4-thiadiazole, with a
(CH2)6 spacer and report our initial findings on their
inhibitory activities.
Compounds 4-10 were prepared uneventfully as
shown in Scheme 1. These new compounds did not react
with the zymogen FXIII but inhibited the activated
FXIIIa enzyme. These bicyclic [1,2,4]-thiadiazoles 5-10
inhibited pure FXIIIa in a time-dependent, irreversible
manner, and the loss of enzymatic activity followed
FXIIIa catalyzes the formation of amide bonds be-
ꢀ
tween γ-carboxamide of glutamine and N -lysine resi-
ꢀ
dues of proteins and/or peptides to form the N -(γ-
glutamyl)lysine bond with release of ammonia.1-3
Similarly, FXIIIa cross-links the γ chains of fibrin to
form γ dimers and causes the polymerization of the R
chains into high molecular weight structures. In addi-
tion, FXIIIa catalyzes the cross-linking of R2-antiplas-
min, a potent inhibitor of fibrinolysis, to the R chains
of fibrin. The γ-chain dimerization and the cross-linking
of R2-antiplasmin are relatively fast processes and are
completed within a few minutes after clot formation,
while the polymerization of the R chains is much slower
and takes several hours. As a result of FXIIIa action,
the mechanical strength of the clot is improved, and its
stability against fibrinolytic enzymes is enhanced. FXI-
IIa is also capable of cross-linking fibrin to extracellular
* To whom correspondence should be addressed. For T.F.T.: phone,
416-749-9300 extension 7384; fax, 416-401-3845; e-mail, ttam@apotex.ca.
For J.M.W.: phone, 416-749-9300 extension 2362; fax, 416-401-3845;
e-mail, jwodzins@apotex.ca.
† Department of Medicinal Chemistry.
‡ Department of Preclinical Discovery.
10.1021/jm049221w CCC: $30.25 © 2005 American Chemical Society
Published on Web 02/22/2005

Letters
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 7 2267
Table 2. Inhibitory Activities of Bicyclic [1,2,4]-Thiadiazole
Derivatives toward FXIIIa and Guinea Pig Liver
Transglutaminase and Their Aqueous Solubilities at pH 7.4
Scheme 1. Synthesis of 3-Substituted Bicyclic
[1,2,4]-Thiadiazole Sulfonamide Derivativesa
FXIIIa
Tgase
IC50
ratiod
Tgase/
FXIIIa
IC50
ki/Ki
solubilitye
(mg/mL)
b,c
compd
(µM)a
(M-1 s-1
)
(µM)a
1f
0.10
0.13
0.11
0.42
0.62
23000
17000
21000
6700
0.25
8.1
6.2
32
2.5
62
56
76
27
NAg
8
0.013
0.004
0.081
0.380
14
16
17
4500
17
a IC50 values were generated by linear regression of percent
inhibition vs log inhibitor concentration for at least three different
concentrations. Standard errors of the slope and the intercept were
below 6%. b The second-order rate constants for inactivation were
determined using the same assay procedure but varying the
incubation time and inhibitor concentration as described by Kitz
and Wilson (see ref 24). Standard errors of the slope and the
intercept of linear regression of apparent rate constants for
inhibition at a given inhibitor concentration vs inhibitor concen-
tration were below 5%. c See ref 25. d The number in the table
represents the preference of the inhibitor for FXIIIa over guinea
pig liver transglutaminase and was calculated as a ratio of the
appropriate IC50 values. e Solubilities of the inhibitors were mea-
sured at pH 7.4. f Compound 1 was prepared using a procedure
described in ref 26. g Not available.
a Reagents and conditions: (a) 5 equiv of 1,6-hexanediamine, 2
equiv of Et3N, DMF; (b) 2 equiv of ArSO2Cl, pyridine or 1.1 equiv
of ArSO2Cl, 2 equiv of solid K2CO3, acetone.
Table 1. In Vitro Activity Data of the New Bicyclic
[1,2,4]-Thiadiazole Derivatives toward Purified Factor XIIIa
compd
R1
R2
R3
IC50 (µM)a
4
11.0
2.3
5
H
CH3
H
NO2
H
H
H
H
6
H
H
0.7
7
H
H
NO2
H
CH3
H
H
NO2
2.2
8
0.13
1.9
9
10
0.91
a IC50 values were generated by linear regression of percent
inhibition vs log inhibitor concentration for at least three different
concentrations. Standard errors of the slope and the intercept were
below 6%. Enzyme activation procedure and the assay conditions
were as described earlier (see ref 15).
Scheme 2. Synthesis of 2-Nitrobenzenesulfonamide
3-Substituted Bicyclic [1,2,4]-Thiadiazole Derivativesa
Chart 3. Examples of 2-Nitrobenzenesulfonamide
Derivatives without the Thiadiazole Moiety and Their
Associated Activity Data toward Factor XIIIa
a Reagents and conditions: (a) 1.5 equiv of Boc2O, 0.5 equiv of
DMAP, ACN; (b) for 14, excess CH3I, solid K2CO3, powder KOH,
cat. n-Bu4N+I-, toluene; for 15, excess BrCH2CO2CH3, solid K2CO3,
powder KOH, cat. n-Bu4N+I-, toluene; (c) HCl(g), MeOH, then
neutralize; (d) concentrated NH3, MeOH; (e) 3 N NaOH, MeOH,
then HCl.
pseudo-first-order kinetics. The inactivated enzyme did
not regain its activity upon exhaustive dialysis. How-
ever, the enzymatic activity could be restored by incu-
bation with dithiothreitol (DTT), a disulfide bond re-
ducing agent. These observations are consistent with
the proposed mechanism (Chart 2) of active-site directed
inactivation of the enzyme brought about by the forma-
tion of a disulfide bond between the active site cysteine
and the sulfur atom of the thiadiazole.18
The in vitro inhibitory results of 4-10 toward FXIIIa
are shown in Table 1. Compound 4 inhibited FXIIIa
with an IC50 of 11.0 µM.15,23 The incorporation of a
phenylsulfone moiety resulted in inhibitor 5, and inter-
estingly its potency toward FXIIIa has now increased
by almost 5-fold compared to 4. Moreover, the substitu-
tion pattern on the aryl ring attached to the sulfone
moiety strongly influenced the inhibitory activities of
these compounds (6-10) because inactivators bearing
ortho substituents were the most effective against
FXIIIa. In this series of compounds, inhibitor 8 with an
IC50 of 130 nM was identified as the lead. Interestingly,
11 and 12 (Chart 3), which do not possess the bicyclic
[1,2,4]-thiadiazole moiety, did not inhibit FXIIIa at up
to 100 µM.
than Merck’s L-722,15115 (1, Table 2) but showed
considerably higher selectivity against guinea pig liver
transglutaminase23 (62-fold) and cysteine-dependent
proteases, caspase-1 (65-fold) and caspase-3 (35-fold)
(results not shown). Furthermore, 8 inactivated pure
FXIIIa with a second-order rate constant (ki/Ki) of 17 000
M-1 s-1, which is 2500 times greater than the second-
order rate constant for its reaction with glutathione (6.8
M-1 s-1).24,25
The aqueous solubility of inactivator 8 at pH 7.4
(0.013 mg/mL) was disappointing. Through a synthetic
effort, 14-17 were prepared by regioselective alkylation
of the amino group at the N3 position of the thiadiazole
moiety via a protection-alkylation-deprotection pro-
tocol (Scheme 2). The structures of the new compounds
14-17 were characterized by 2D NMR, including the
HSQC and HMBC experiments.
Introduction of a methyl substituent at the N3 posi-
tion of the thiadiazole moiety resulted in 14, which
displayed inhibitory activities comparable to that of
inhibitor 1 toward FXIIIa but exhibited a 56-fold prefer-
ence over transglutaminase (Table 2). However, a 3-fold
loss in solubility relative to inhibitor 8 was observed for
14 on substituting H for Me. Functionalization of the
methyl substituent (CH2-H) in 14 with a more polar
This class of small nonpeptidic compounds displayed
selectivity for FXIIIa against the closely related trans-
glutaminase. For example, 8 was slightly less potent

2268 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 7
Letters
Figure 1. Composition of platelet-depleted human plasma
clots made in the presence of varying concentrations of the
bicyclic [1,2,4]-thiadiazole inhibitor (8). Each bar represents
an average of three experiments. Mean values that are
significantly different from the mean of an appropriate control
(Student’s t test, p < 0.003) are indicated by an asterisk (/).
Figure 2. Acceleration of fibrinolysis of platelet-depleted and
platelet-rich human plasma clots made in the presence of
varying concentrations of inhibitor (8). Citrated human plasma
(0.3 mL) spiked with 125I-fibrinogen (200 µCi/mg) was clotted
in the presence or absence of inhibitor by the addition of CaCl2
(50 mM) and thrombin (5 units/mL). After a 30-min incubation
at 37 °C, clots were harvested by winding on a bamboo stick
and washed three times in 50 mM Tris-HCl buffer, pH 7.4,
containing 100 mM NaCl and 1 mM EDTA. Clots were placed
in 2 mL of 50 mM Tris-HCl buffer, pH 7.4, containing 100 mM
NaCl, 1 mM EDTA, and 300 IU/mL human t-PA and were
lysed at 37 °C for 22 h. The degree of fibrinolysis was measured
as a percent of 125I-fibrinogen solubilized from clots. Each bar
represents an average of three experiments except for 40 µM
8 in platelet-depleted plasma where n ) 2. Mean values that
are significantly different from the mean of an appropriate
control (Student’s t test, p < 0.0007) are indicated by an
asterisk (/).
group such as CH2-CONH2 (16) or CH2-COOH (17)
caused a 3- to 4-fold decrease in potency for these
inhibitors toward FXIIIa (Table 2). However, the acid
derivative 17 showed an almost 30-fold increase in
aqueous solubility at pH 7.4 (0.38 mg/mL) relative to
inactivator 8 (0.013 mg/mL). The second-order rate
constant (ki/Ki ) 4500 M-1 s-1) measured for 17 was
very similar to that reported for inhibitor 2 (ki/Ki ) 5083
M-1 s-1) shown in Chart 1.13
We next turned our attention to the effect of our
compounds on the composition of platelet-depleted
plasma clots (Figure 1). Plasma spiked with 125I-
fibrinogen was clotted at 37 °C for 2 h in the presence
and the absence of varying concentrations of inhibitor
8. The clots were washed, dissolved in 8 M urea with
10% SDS and 10% mercaptoethanol, and analyzed using
SDS-PAGE.15 Electrophoretic bands were quantitated
by autoradiography. Increasing inhibitor concentration
resulted in a decrease in the formation of γ dimers and
higher molecular weight fibrins. The formation of R
polymers was completely suppressed at 160 µM concen-
tration of 8. At that same inhibitor concentration, the
amount of γ dimers has been halved relative to the no-
inhibitor control. Clots, formed from 125I-fibrinogen
spiked platelet-depleted and platelet-rich human plasma
in the presence of varying concentrations (40, 80, and
160 µM) of inhibitor 8, were subjected to in vitro
fibrinolysis by human tissue plasminogen activator (t-
PA).15 The amount of lysed fibrin was measured after
22 h at 37 °C. Plasma clots were carefully washed to
remove the inhibitor and were placed in a fresh solution
containing t-PA. This design of fibrinolysis studies
precludes the possibility of inhibitors directly affecting
the interaction between t-PA and plasminogen activator
inhibitor-1 (PAI-1) or plasmin and R2-antiplasmin be-
cause it is not present in solution at the time of
fibrinolysis.
8 strongly suggest that the acceleration of in vitro
fibrinolysis is due to the inhibition of FXIIIa-catalyzed
processes.
Detailed results on the biological studies of this new
class of FXIIIa inhibitors will be reported in due course.
Supporting Information Available: Synthesis and spec-
tral data of the new inhibitors, including 2D-COSY, 2D-HSQC,
and 2D-HMBC spectra of inhibitor 17. This material is
available free of charge via the Internet at http://pubs.acs.org.
References
(1) McDonath, J. In Haemostasis and Thrombosis: Basic Principles
and Clinical Practice, 3rd ed.; Coleman, R. W., Hirsh, J., Marder,
V. J., Salzman, E. W., Eds.; Lippincott Company: Philadelphia,
PA, 1994; pp 301-313.
(2) For a review, see the following. Sicker, T.; Hilgenfeld, R. Blood
coagulation factor XIII: activation, substrates and structure of
a transglutaminase. Hamostaseologie 2002, 22, 20-27.
(3) For a review, see the following. Lorand, L. Factor XIII: structure,
activation, and interactions with fibrinogen and fibrin. Ann. N.
Y. Acad. Sci. 2001, 936, 291-311.
(4) Yee, V. C.; Le Trong, I.; Bishop, P. D.; Pedersen, L. C.; Stenkamp,
R. E.; Teller, D. C. Structure and function studies of factor XIIIa
by X-ray crystallography. Semin. Thromb. Hemostasis 1996, 22,
377-384.
(5) Yee, V. C.; Pedersen, L. C.; Le Trong, I.; Bishop, P. D.; Stenkamp,
R. E.; Teller, D. C. Three-dimensional structure of a trans-
glutaminase: human blood coagulation factor XIII. Proc. Natl.
Acad. Sci. U.S.A. 1994, 91, 7296-7300.
(6) Curtis, C. G.; Brown, K. L.; Credo, R. B.; Domanik, R. A.; Gray,
A.; Stenberg, P.; Lorand, L. Calcium-dependent unmasking of
active center cysteine during activation of fibrin stabilizing
factor. Biochemistry 1974, 13, 3774-3780.
(7) Leidy, E. M.; Stern, A. M.; Friedman, P. A.; Bush, L. R. Enhanced
thrombolysis by a factor XIIIa inhibitor in a rabbit model of
femoral artery thrombosis. Thromb. Res. 1990, 59, 15-26.
(8) Kloczko, J.; Wojtukiewicz, M.; Bielawiec, M.; Zuch, A. Alterations
of haemostasis parameters with special reference to fibrin
stabilization, factor XIII and fibronectin in patients with oblit-
erative atherosclerosis. Thromb. Res. 1988, 51, 575-581.
Figure 2 illustrates the acceleration of fibrinolysis
with increasing inhibitor concentration relative to no-
inhibitor control. These results and the analysis of
composition of the clots made in the presence of inhibitor

Letters
(9) Francis, C. W.; Connaghan, D. G.; Scott, W. L.; Marder, V. J.
Increased plasma concentration of cross-linked fibrin polymers
in acute myocardial infarction. Circulation 1987, 75, 1170-1177.
(10) Finney, S.; Seale, L.; Sawyer, R. T. Wallis, R. B. Tridegin, a new
peptidic inhibitor of factor XIIIa, from the blood-sucking leech
Haementeria ghilianii. Biochem. J. 1997, 324, 797-805.
(11) West, R. R.; Martinez, T.; Franklin, H. R.; Bishop, P. D.; Rassing,
B. R. U.S. Patent 5,710,174, 1998.
(12) Tymiak, A. A.; Tuttle, J. G.; Kimball, S. D.; Wang, T.; Lee, V. G.
A simple and rapid screen for inhibitors of factor XIIIa. J.
Antibiot. (Tokyo) 1993, 46, 204-206.
(13) Kogen, H.; Kiho, T.; Tago, K.; Miyamoto, S.; Fujioka, T.; Otsuka,
N.; Suzuki-Konagi, K.; Ogita, T. Alutacenoic Acids A and B, Rare
Naturally Occurring Cyclopropenone Derivatives Isolated from
Fungi: Potent Non-Peptide Factor XIIIa Inhibitors. J. Am.
Chem. Soc. 2000, 122, 1842-1843.
(14) Iwata, Y.; Tago, K.; Kiho, T.; Kogen, H.; Fujioka, T.; Otsuka,
N.; Suzuki-Konagai, K.; Ogita, T.; Miyamoto, S. Conformational
analysis and docking study of potent factor XIIIa inhibitors
having a cyclopropenone ring. J. Mol. Graphics Modell. 2000,
18, 591-599, 602-604.
(15) Freund, K. F.; Doshi, K. P.; Gaul, S. L.; Claremon, D. A.; Remy,
D. C.; Baldwin, J. J.; Pitzenberger, S. M.; Stern, A. M. Trans-
glutaminase inhibition by 2-[(2-oxopropyl)thio]imidazolium de-
rivatives: mechanism of factor XIIIa inactivation. Biochemistry
1994, 33, 10109-10119.
(16) Shebuski, R. J.; Sitko, G. R.; Claremon, D. A.; Baldwin, J. J.;
Remy, D. C.; Stern, A. M. Inhibition of factor XIIIa in a canine
model of coronary thrombosis: effect on reperfusion and acute
reocclusion after recombinant tissue-type plasminogen activator.
Blood 1990, 75, 1455-1459.
(17) For a recent review, see the following. Prasa, D.; Sturzebecher,
J. Inhibitors of factor XIIIa. Hamostaseologie 2002, 22, 43-47.
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 7 2269
(18) Leung-Toung, R.; Wodzinska, J.; Li, W.; Lowrie, J.; Kukreja, R.;
Desilets, D.; Karimian, K.; Tam, T. F. [1,2,4]-Thiadiazole: A
Novel Cathepsin B Inhibitor. Bioorg. Med. Chem. 2003, 11,
5529-5537.
(19) Marrano, C.; de Macedo, P.; Gagnon, P.; Lapierre, D.; Gravel,
C.; Keillor, J. W. Synthesis and evaluation of novel dipeptide-
bound 1,2,4-thiadiazoles as irreversible inhibitors of guinea pig
liver transglutaminase. Bioorg. Med. Chem. 2001, 9, 3231-3241.
(20) Stenberg, P.; Ljunggren, C.; Nilsson, J. L.; Lunden, R.; Eriksson,
O. Fibrin-stabilizing factor inhibitors. 5. Primary amines related
to monotosylcadaverine. J. Med. Chem. 1972, 15, 674-675.
(21) Lorand, L.; Parameswaran, K. N.; Stenberg, P.; Tong, Y. S.;
Velasco, P. T.; Jonsson, N. A.; Mikiver, L.; Moses, P. Specificity
of guinea pig liver transglutaminase for amine substrates.
Biochemistry 1979, 18, 1756-1765.
(22) Tam, T.; Leung-Toung, R.; Li, W.; Spino, M.; Karimian, K.
Medicinal Chemistry and Properties of 1,2,4-Thiadiazoles. Mini-
Rev. Med. Chem., in press.
(23) Curtis, C. G.; Lorand, L. Fibrin-stabilizing factor (factor XIII).
Methods Enzymol. 1976, 45, 177-191.
(24) ki is the rate constant for conversion of noncovalent complex to
covalent adduct, and Ki is the dissociation constant for nonco-
valent enzyme-inactivatior complex. Kitz, R.; Wilson, I. B.
Esters of methanesulfonic acid as irreversible inhibitors of
acetylcholinesterase. J. Biol. Chem. 1962, 237, 3245-3249.
(25) The second-order rate constants (ki/Ki) for FXIIIa inactivation
were measured to better characterize the series of compounds
in Table 2. The Ki values were much higher than the concentra-
tions of inhibitors used and therefore could not be determined.
(26) Jones, W. J., Jr. Ph.D. Thesis, Duke University, 1966 (University
Microfilms, No. 66-13691).
JM049221W