Synthesis of potent water-soluble tissue transglutaminase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2008.09.006

Dipeptide-based sulfonium peptidylmethylketones derived from 6-diazo-5-oxo-l-norleucine (DON) have been investigated as potential water-soluble inhibitors of extracellular transglutaminase. The lead compounds were prepared in four steps and exhibited pote

Full text of DOI:10.1016/j.bmcl.2008.09.006

Bioorganic & Medicinal Chemistry Letters 18 (2008) 5559–5562
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis of potent water-soluble tissue transglutaminase inhibitors
b
a
c
Martin Griffin a, , Alexandre Mongeot , Russell Collighan , Robert E. Saint ,
*
Richard A. Jones c, , Ian G.C. Coutts c, Daniel L. Rathbone a
a School of Life and Health Sciences, Aston University, Aston Triangle, Birmingham B4 7ET, UK
b X-LINK Ltd, BIOCITY, Pennyfoot St. Nottingham NG1 1GF, UK
c School of Science and Technology, Nottingham Trent University, Clifton Campus, Nottingham NG11 8NS, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Dipeptide-based sulfonium peptidylmethylketones derived from 6-diazo-5-oxo-L-norleucine (DON) have
Received 16 June 2008
Revised 29 August 2008
Accepted 3 September 2008
Available online 6 September 2008
been investigated as potential water-soluble inhibitors of extracellular transglutaminase. The lead com-
pounds were prepared in four steps and exhibited potent activity against tissue transglutaminase.
Ó 2008 Elsevier Ltd. All rights reserved.
Keywords:
Transglutaminase
Inhibitors
Dipeptide
Transglutaminases (TGases) catalyse the crosslinking of pro-
teins by formation of an isopeptide bond between a glutamyl car-
dihydroisoxazole isomers were better inhibitors of human TG2
than the corresponding 5-(R) stereoisomers.17
boxamide in one protein and a lysyl
e
-amino group of another
Sulfonium peptidylmethylketones are known to inhibit prote-
ases irreversibly by transfer of the peptidyl portion to key active
site cysteine sulphur atoms.18 Compounds of structure Cbz–Phe–
NH(CH2)nCOCH2S+(CH3)2, have been found to be potent inactiva-
tors of transglutaminases, the chain length of the –(CH2)n– spacer
moiety, n = 3, being optimum.19 Here, we present a set of related
protein. In mammals the activity is Ca2+-dependent and in some
family members such as TG2 (tissue transglutaminase) is inhibited
by GDP/GTP binding.1,2 Protein modification mediated by TG2 has
been implicated in numerous diseases and processes3 such as
fibrosis and scarring,4 Huntington’s disease and Alzheimer’s dis-
ease,5 coeliac disease,6 thrombosis7 and cancer.8 It can therefore
be predicted that potent and selective TG2 inhibitors would find
wide therapeutic potential. Importantly, in pathologies such as
fibrosis and scarring, thrombosis and coelic disease, TG2 is likely
to be acting predominantly outside, rather than inside, the cell
where high Ca2+ concentrations lead to its activation.9
inhibitors of TG2 that are derived from 6-diazo-5-oxo-L-norleucine
(DON)20 and terminate in the dimethylsulphonium methylketone
moiety. The compounds to be described incorporate an extra car-
boxyl group compared to the previously presented sulfonium pept-
idylmethylketones, designed to potentially contribute to both the
structure–activity profile and enhancement of their water solubil-
ity. Importantly, these inhibitors, which are permanently charged
at physiological pH, have an increased likelihood of remaining out-
side cells, thus targeting the extracellular pool of TG2 rather than
crossing the cell membrane and interacting with other intracellular
transglutaminases, a factor which could be important in reducing
their in vivo toxicity.21
Several irreversible inhibitors of tissue transglutaminase (TG)
have been published based upon, or related to, N-benzyloxycar-
bonyl-protected phenylalanine or N-benzyloxycarbonyl-protected
L
-glutaminylglycine, a commonly used dipeptide acyl-donor sub-
strate.10 These compounds generally include a spacer group termi-
nating in an electrophilic warhead capable of covalently modifying
the sulphur atom of the key cysteine residue at the active site.
The target compounds (except 4c and 4j) were prepared from
commercially available N-protected amino acids in four steps
(Scheme 1 and Table 1). The acids were converted to the corre-
sponding N-hydroxysuccinimide active esters 1 by means of N,N0-
disuccinimidyl carbonate in pyridine–acetonitrile.22
Examples of the warheads include maleimides,11,12
a,b-unsatu-
rated amides,12–14 chloroacetamide,12 epoxides,13,14 thiadiazoles15
and dihydroisoxazoles.16 Unlike the other warheads, the dihydrois-
oxazole group contains a chiral centre. It was found that the 5-(S)-
The active esters were combined with 6-diazo-5-oxo-L-norleu-
cine (‘DON’ purchased from SAFC) in THF:water, 1:1 in the pres-
ence of triethylamine at 0 °C for 2 h. The solvent was evaporated
at room temperature to afford the diazo intermediates 2. These
* Corresponding author. Tel.: +44 121 204 3942.
E-mail address: m.griffin@aston.ac.uk (M. Griffin).
Deceased.
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.09.006

5560
M. Griffin et al. / Bioorg. Med. Chem. Lett. 18 (2008) 5559–5562
O
(89%) that was taken through to 4c. Phenylalanine was reacted
(i)
(ii)
R
AA
OH
O
R
AA
N
with the N-hydroxysuccinimide active ester of phenyl-acetic acid
to give the shortened version of Cbz–phenylalanine (92%). This
was carried forward to compound 4j. Using Cbz–phenylalanine
as a scaffold, compounds 5–7 were prepared to vary the nature
of the electrophilic warhead (Fig. 1). Compound 5 was prepared
in the same manner as described above but replacing dimethyl sul-
phide by diethyl sulphide in the final step (85%). 1,3,4,5-Tetrame-
N
N
H
N
82-96%
H
O
O
O
1
O
O
+
N
H
N
H
N
(iii)
R
AA
Br
R
AA
N
N
H
68-95%
(two steps)
H
thylimidazole-2(3H)-thione25
was
alkylated
with
the
O
O
O
OH
O
OH
bromomethyl ketone 3 (R = Cbz and AA = phenylalanine) to give
the imidazolium compound 6 (88%). Compound 7 was prepared
following published procedures.17 The known TG2 inhibitor 826
was included for reference. In addition, to investigate the role of
the carboxyl group in compounds of type 4, a methyl ester version
was prepared in two steps from the Cbz phenylalaninyl bromo-
methyl ketone 3: treatment with trimethylsilyldiazomethane27 in
THF–methanol at 0 °C for 3 h gave the corresponding methyl ester;
the subsequent action of dimethyl sulphide in dry methanol gave
the target compound 4n.
3
2
(v)
(iv) 73-92%
O
82%
O
Br
H
N
H
R
AA
Br
R
AA
N
S+
N
H
N
H
O
O
O
O
O
OH
4a-m
(iv)
94%
O
Br
The compounds were characterised by melting point, infra-red
spectroscopy, mass spectrometry (+ve electrospray accurate mass),
proton NMR spectroscopy and carbon-13 NMR spectroscopy and
had properties consistent with the proposed structures.
H
N
R
AA
S+
N
H
O
O
O
4n
TG2 activity was measured using an enzyme linked sorbent
assay to measure the covalent incorporation of biotinyl-5-pentyl-
amine into N,N0-dimethylcasein as described previously.28 Briefly,
a 96-well microtitre plate was coated with N,N0-dimethylcasein
and a reaction mix containing TG2 (2 ng mlꢀ1), biotinyl-5-pentyl-
Scheme 1. Reagents and conditions: (i) N,N0-disuccinimidyl carbonate/pyridine/
acetonitrile; (ii) DON/THF/water/Et3N/0 °C; (iii) HBr/AcOH/EtOAc; (iv) Me2S/MeOH;
(v) TMS–CH2N2/THF/MeOH/0 °C.
amine (132 lM), CaCl2 (5 mM) and DTT (5 mM) with varying con-
were redissolved immediately in ethyl acetate and treated with
HBr:acetic acid 1:1. Work-up gave the bromomethyl ketones 3.
These were dissolved in dry methanol and treated with dimethyl
sulphide (5 equiv) for 2 days at room temperature. After removal
of the solvent the residue was dissolved in water, washed with
ethyl acetate and the water layer was freeze-dried to give the tar-
get compounds 4 as white powders.23 The N-protected phenylala-
nine building blocks required for compounds 4c and 4j were
prepared as follows: 2-Phenyl ethanol was converted into the cor-
responding 4-nitrophenyl carbonate17 (79%). This was reacted with
phenylalanine to give the stretched Cbz-protected phenylalanine24
centrations of inhibitor, was added to each well and incubated for
1 h at 37 °C. The amount of incorporated biotinyl-5-pentylamine
was quantitated by reaction with ExtravidinÒ peroxidase and
colour development with o-phenylenediamine. The IC50 was
expressed as the inhibitor concentration at which 50% inhibition
of TG2 activity was observed. The inhibition data against guinea
pig liver TG2 for compounds 4–8 are summarised in Table 1. Paral-
lel studies were also undertaken using recombinant human TG2
(rhTG), using compounds 4a and 4d, which gave IC50s comparable
to those of guinea pig TG2. This was not unexpected, given the very
highly conserved structures of their active sites.
To confirm the irreversible modification of TG2 with this ser-
ies of compounds, compound 4a was pre-incubated with TG2 in
the presence of Ca2+ for different time periods up to 60 min at
20 °C. It was then diluted 100-fold and residual TG2 activity
determined using the above assay. At the IC50 concentration of
Table 1
Inhibition data for compounds prepared as TG2 inhibitors
a
Compound
R
AA
IC50
Aqueous
solubility
(mg/mL)
(l
M)
5 lM, compound 4a reduced activity by 77 7% after 60 min
preincubation (data not shown). It was noted that Ca2+ was
required for reaction between TG2 and inhibitors, consistent
with the requirement of this cofactor for the transglutaminase
activity of TG2.
4a
4b
4c
4d
4e
4f
Cbz
Cbz
Proline
Glycine
5
8
8
10
12
1
1
1
500
1000
PhCH2CH2OC(@O) Phenylalanine
Cbz
Cbz
Cbz
Phenylalanine
Tryptophan
(2S,4R)-4-hydroxy-2- 18
pyrrolidine
carboxylic acid
Alanine
Aspartic acid
Phenylalanine
Phenylalanine
Lysine
Phenylalanine
Isoleucine
2
2
3
167
Compounds 4 were indeed found to be water-soluble (Table 1;
data presented for 4a, 4b, 4d and 4n), very much more so than
compound 7 and even comparable to or exceeding that of the ref-
erence compound 8. As expected, the methyl ester 4n was some-
what less soluble compared to its parent compound 4d.
Using the phenylalanine compound 4d as a reference point it
can be seen that increasing the size of the dimethyl sulphonium
group to the diethyl analogue 5 halved the activity, presumably
through steric effects. In contrast, employing the tetramethylimi-
dazolium group (analogue 6) increased the activity to give the
most potent compound in this study. Given that this moiety is at
least as large as the diethyl sulphonium group it may be that a dif-
ference in chemical reactivity between the two electrophilic war-
heads reflects the difference in observed activity. Compound 7
contains the dihydroisoxazole pioneered by Khosla’s group.16 It
has the electrophilic group on a shorter tether compared with
4g
4h
4i
4j
4k
4l
4m
4n
5
Cbz
Cbz
20
28
30
3
5
4
Fmoc
PhCH2C(@O)
Cbz
t-BOC
Cbz
110 15
150 25
300 40
350 42
Cbz
Phenylalanine
5
20
3
1
67
See Figure 1
2
6
7
8
See Figure 1
See Figure 1
See Figure 1
1
>200
<1
200
4
1
a
IC50 values are mean values standard errors from
a minimum of three
experiments.

M. Griffin et al. / Bioorg. Med. Chem. Lett. 18 (2008) 5559–5562
5561
O
O
Br
H
N
H
N
S+
S
N+
Br
CbzNH
CbzNH
O
O
N
O
OH
O
O
OH
Cl
6
5
O
N+
N
S
H
N
Br
CbzNH
N
O
8
7
Figure 1. Variation in the electrophilic warhead.
4d, 5 and 6 and so it is not a direct comparison of the warheads’
efficacy. Nevertheless, in our assay it proved to be relatively
inactive.
In a recent study to test the efficacy of 4d in an animal model of
renal scarring, a 90% reduction in scarring and significant protec-
tion of kidney function was observed. Compound 4d demonstrated
no obvious signs of toxicity.29 Given the increasing importance of
TG2 as a medicinal target and the work demonstrated here and
by others, it is not unreasonable to suggest these studies may even-
tually lead to an effective clinical treatment of those disorders
where TG2 has a role.
From the compounds derived from phenylalanine it is clear that
the optimum derivatisation on the N-terminus is a carbamate of
the Cbz kind. This is in agreement with the work of Chica et al.1
who demonstrated that Cbz-Gln-Gly, but not BOC-Gln-Gly is an
acyl donor substrate for guinea pig liver Tgase. Moving to Fmoc
(4i), t-BOC (4l) or shortening the Cbz to the amide version (4j)
increased the IC50. Lengthening the Cbz group by one methylene
unit (4c) gave a slight improvement in the IC50 value over that
for the Cbz compound 4d. Swapping the phenyl group of 4d for
the larger indole ring 4e made little change to the activity. Remov-
ing the phenyl ring entirely from 4d to give the alanine analogue
4g halved the activity and extension of the alkyl side chain to that
of the isoleucine derivative 4m effectively abolished the activity. In
contrast, removal of the whole sidechain to give the glycine ana-
logue 4b made no change to the activity. Blocking the carboxyl
group of 4d as the methyl ester not only reduced its water solubil-
ity but also led to a small increase in activity, suggesting that the
ionised carboxyl group is not playing a role in the efficacy of the
inhibitors. Restraint of the flexibility of the inhibitor in the form
of the proline analogue 4a gave a small increase in activity com-
pared to the glycine and phenylalanine compounds (4b and 4d).
Addition of a hydroxyl group (the hydroxyproline analogue 4f)
reduced the activity.
Spectroscopic data for N-benzyloxycarbonyl-L-prolinyl-6-dim-
ethylsulphonium-5-oxo- -norleucine bromide (4a)
L
1H NMR (250 MHz, DMSO-d6, d DMSO = 2.50 ppm): 1.69–1.89
(m, 4H, –NCH2CH2CH2– and NCH2CH2CH2–), 2.01–2.23 (m, 2H,
–CHCH2CH2–), 2.61–2.83 (m, 2H, –CHCH2CH2–), 2.91 (s, 6H, –
S+(CH3)2), 3.32–3.52 (m, 2H, –NCH2CH2CH2–), 4.13-4.33 (m, 2H,
–N(CH2)3CH– and –CHCH2CH2–), 4.70–4.92 (m, 2H, –C(@O)
CH2S+(CH3)2), 5.01–5.16 (m, 2H, PhCH2O–), 7.29–7.43 (m, 5H, Ph),
8.36 (d, J 7.8 Hz, 1H, –NH–), 12.80 (br s, 1H, –COOH) ppm. 13C NMR
(63 MHz, DMSO-d6, d DMSO = 39.5 ppm): 23.1, 23.9 and 24.4 ((–
S+(CH3)2, –NCH2CH2CH2– and –NCH2CH2CH2–), 30.0, 31.1 and 37.7
(–NCH2CH2CH2–, –CHCH2CH2– and –CHCH2CH2–), 46.6, 50.5 and
53.6 ((–C(@O)CH2S+(CH3)2, –CHCH2CH2– and N(CH2)3CH–), 65.9
(PhCH2O–), 127.1, 127.7 and 128.5 (Ar CH), 137.0 (Ar C), 154.1
(PhCH2OC(@O)–), 172.2 and 173.1 (COOH and –C(@O)NHCH-
COOH–), 201.6 (–C(@O)CH2S+(CH3)2). High-resolution MS (+electro-
spray): m/z: found 437.1747; calcd for C21H29N2O6S, 437.1746.
Spectroscopic data for N-benzyloxycarbonyl-L-phenylalanyl-6-
These results are consistent with the compounds in this study
adopting two different binding modes to the enzyme. It is pro-
posed that the Cbz-related protected analogues having aromatic
sidechains (4c, 4d and 4e) interact with a hydrophobic pocket
suitable for flat aromatic ligands. The loss of the aryl group
(4g) relinquishes this stabilisation. Its conversion to a bulky ali-
phatic group (4m) destabilises the complex. The proline com-
pound 4a is constrained to a more angular overall shape that
would naturally be adopted by the other compounds and it is
possible that this drives a different binding mode in this case.
The glycine analogue 4b is one of the most active compounds
in this study but the complex with the enzyme cannot attain
the hydrophobic stabilisation proposed for 4c, 4d and 4e. It is,
however, the most flexible of all the compounds and may there-
fore be able to engage in the binding mode preferred by the pro-
line compound.
dimethylsulphonium-5-oxo- -norleucine bromide (4d)
L
1H NMR (250 MHz, DMSO-d6, d DMSO = 2.50 ppm): 1.77–1.93
(m, 1H, –CHCH2CH2–), 2.01–2.17 (m, 1H, –CHCH2CH2–), 2.65–2.81
(m, 3H, –CHCH2CH2– and PhCH2CH–), 2.90 (s, 6H, –S+(CH3)2), 3.00–
3.09 (m, 1H, PhCH2CH–), 4.21–4.35 (m, 2H, PhCH2CH– and
–CHCH2CH2–), 4.70–4.85 (m, 2H, –C(@O)CH2S+(CH3)2), 4.88–5.01
(m, 2H, PhCH2O–), 7.15–7.40 (m, 10H, Ph), 7.57 (d, J 8.5 Hz, 1H,
ZNH–), 8.45 (d, J 8.3 Hz, 1H, –C(@O)NHCHCOOH–), 12.90 (br s, 1H,
–COOH) ppm. 13C NMR (63 MHz, DMSO-d6, d DMSO = 39.5 ppm):
24.5 (–S+(CH3)2), 37.3, 37.6 and 40.4 (–CHCH2CH2–, PhCH2CH– and
–CHCH2CH2–), 50.9, 53.6 and 56.1 (–C(@O)CH2S+(CH3)2,
–CHCH2CH2– and PhCH2CH–), 65.3 (PhCH2O–), 126.3, 127.5, 127.8,
128.1, 128.4 and 129.3 (Ar CH), 137.0 and 138.1 (Ar C), 156.0
(PhCH2OC(@O)–), 172.0 and 173.0 (COOH and –C(@O)NHCH-
COOH–), 201.5 (–C(@O)CH2S+(CH3)2). High-resolution MS (+electro-
spray): m/z: found 487.1906; calcd for C25H31N2O6S, 487.1903.
The use of reference compound 4d has so far shown no signs of
cell toxicity in many of the studies it has been cited. Incubation of
the compound with endothelial plasma membrane monolayers
indicated no penetration of the compound over a 5-h period.21
Their lack of cellular penetration may be a contributing factor to
their lack of toxicity in that only the extracellular transglutaminas-
es are targeted.
Acknowledgments
The authors acknowledge the support of the Marie Curie
Research and Training network ‘Transglutaminases: role in patho-
genesis, diagnosis and therapy—TRACKS’ (MRTN-CT-2006-03032)
[postdoctoral research fellowship for A.M.] and the Wellcome Trust

5562
M. Griffin et al. / Bioorg. Med. Chem. Lett. 18 (2008) 5559–5562
15. Marrano, C.; de Macédo, P.; Gagnon, P.; Lapierre, D.; Gravel, C.; Keillor, J. W.
Bioorg. Med. Chem. 2001, 9, 3231.
16. Choi, K.; Siegel, M.; Piper, J. L.; Yuan, L.; Cho, E.; Strnad, P.; Omary, B.; Rich, K.
‘‘Control of central venous catheter obstruction and infection
by selective inhibition of transglutaminase (Grant Reference
GR080885MA).
M.; Khosla, C. Chem. Biol. 2005, 12, 469.
17. Watts, R. E.; Siegel, M.; Khosla, C. J. Med. Chem. 2006, 49, 7493.
18. Shaw, E. J. Biol. Chem. 1988, 263, 2768.
19. Pliura, D. H.; Bonaventura, B. J.; Pauls, H. W.; Killackey, J. F.; Krantz, A. J.
Enzymes Inhib. Med. Chem. 1992, 6, 181.
References and notes
1. Chica, R. A.; Gagnon, P.; Keillor, J. W.; Pelletier, J. N. Protein Sci. 2004, 13, 979.
2. Griffin, M.; Casadio, R.; Bergamini, C. M. Biochem. J. 2002, 368, 377.
3. Aeschlimann, D.; Thomazy, V. Connect. Tissue Res. 2000, 41, 1.
4. Johnson, T. S.; Skill, N. J.; El Nahas, A. M.; Oldroyd, S. D.; Thomas, G. L.;
Douthwaite, J. A.; Haylor, J. L.; Griffin, M. J. Am. Soc. Nephrol. 1999, 10, 2146.
5. Citron, B. A.; SantaCruz, K. S.; Daviesi, P. J. A.; Festoff, B. W. J. Biol. Chem. 2001,
276, 3295.
6. Marzari, R.; Sblattero, D.; Florian, F.; Tongiorgi, E.; Not, T.; Tommasini, A.;
Ventura, A.; Bradbury, A. J. Immunology 2001, 166, 4170.
7. Ariëns, R. A.; Lai, T. S.; Weisel, J. W.; Greenberg, C. S.; Grant, P. J. Blood 2002, 100,
743.
20. Coutts, I. G. C.; Saint, R. E. Tetrahedron Lett. 1998, 39, 3243.
21. Baumgartner, W.; Golenhofen, N.; Weth, A.; Hiiragi, T.; Saint, R.; Griffin, M.;
Drenkhahn, D. Histochem. Cell Biol. 2004, 122, 17.
22. Ogura, H.; Kobayashi, T.; Shimizu, K.; Kawabe, K.; Takeda, K. Tetrahedron Lett.
1979, 20, 4745.
23. Griffin, M.; Coutts, I. G. C.; Saint, R. E. World patent WO2004113363,
2004.
24. Madec-Lougerstay, R.; Florent, J.-C.; Monneret, C. J. Chem. Soc., Perkin Trans. 1
1999, 1369.
25. Kuhn, N.; Kratz, T. Synthesis 1993, 561.
26. Baldwin, J. J.; Remy, D. C. US patent US5098707, 1991.
27. Kuehnel, E.; Laffan, D. D. R.; Lloyd-Jones, G. C.; del Campo, T. M.; Shepperson, I.
R.; Slaughter, J. L. Angew. Chem., Int. Ed. 2007, 46, 7075.
28. Slaughter, T. F.; Achyuthan, K. E.; Lai, T. S.; Greenberg, C. S. Anal. Biochem. 1992,
205, 166.
29. Johnson, T. S.; Fisher, M.; Haylor, J. L.; Hau, Z.; Skill, N. J.; Jones, R. SaintR.;
Coutts, I.; Vickers, M. E.; El Nahas, A. M.; Griffin, M. J. Am. Soc. Nephrol. 2007, 18,
3078.
8. Mangala, L. S.; Mehta, K. Prog. Exp. Tumor Res. 2005, 38, 125.
9. Smethurst, P. A.; Griffin, M. Biochem. J. 1996, 313, 803.
10. Folk, J. E.; Chung, S. I. Methods Enzymol. 1985, 113, 358.
11. Halim, D.; Caron, K.; Keillor, J. W. Bioorg. Med.. Chem. Lett. 2007, 17, 305.
12. Pardin, C.; Gillet, S. M. F. G.; Keillor, J. W. Bioorg. Med. Chem. 2006, 14, 8379.
13. Marrano, C.; de Macédo, P.; Keillor, J. W. Bioorg. Med. Chem. 2001, 9, 1923.
14. de Macédo, P.; Marrano, C.; Keillor, J. W. Bioorg. Med. Chem. 2002, 10, 355.