New substrate analogue furin inhibitors derived from 4-amidinobenzylamide

Article abstract of DOI:10.1016/j.bmcl.2011.06.091

A series of new peptidomimetic furin inhibitors was synthesized, which was derived from our previously described lead structure phenylacetyl-Arg-Val-Arg-4- amidinobenzylamide (1). Substitution of Val by other amino acid residues revealed several highly po

Full text of DOI:10.1016/j.bmcl.2011.06.091

Bioorganic & Medicinal Chemistry Letters 21 (2011) 4695–4697
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
New substrate analogue furin inhibitors derived from 4-amidinobenzylamide
⇑
Gero L. Becker, Kornelia Hardes, Torsten Steinmetzer
Department of Pharmaceutical Chemistry, Philipps University Marburg, Marbacher Weg 6, D-35032 Marburg, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of new peptidomimetic furin inhibitors was synthesized, which was derived from our previously
described lead structure phenylacetyl-Arg-Val-Arg-4-amidinobenzylamide (1). Substitution of Val by
other amino acid residues revealed several highly potent furin inhibitors with Ki values of less than
2 nM, containing guanidinoalanine, Ile, Phe or Tyr in the P3 position. The replacement of the P2 Arg by
Received 25 May 2011
Revised 17 June 2011
Accepted 19 June 2011
Available online 25 June 2011
Lys was also well accepted, whereas the incorporation of D-amino acids at various positions resulted in
poor inhibitors. The use of the 4-amidinobenzylamide group provides convenient synthetic access to sta-
ble proprotein convertase inhibitors and derivatives as biochemical tools and for further studies in cell
culture.
Keywords:
Furin
Proprotein convertase
Protease inhibitor
Peptidomimetic inhibitor
Serine protease
Ó 2011 Elsevier Ltd. All rights reserved.
The proprotein convertase furin is a Ca2+-dependent cellular
serine endoprotease with a strong preference for cleavage at mult-
ibasic consensus sequences Arg-X-Arg/Lys-Arg;X.1 It is involved in
the maturation of many proproteins in the secretory pathway,
including prohormones, proenzymes and proforms of receptors
or extracellular matrix proteins, and therefore has an important
function in embryogenesis and homeostasis.2,3 A comprehensive
overview about potential furin substrates provides a recently
developed database.4 In addition to normal physiological pro-
cesses, furin is most likely involved in diseases, such as cancer,5
dementia, neurodegenerative disorders, various viral and bacterial
infections, and has therefore emerged as a potential target for drug
design.6 Meanwhile, in first proof-of-concept studies it could be
demonstrated that synthetic furin inhibitors might have therapeu-
tic benefit in living animals.7 Various types of synthetic furin inhib-
itors have been reviewed in the past.8,9 Among these are pure
PC5/6 and PACE4, whereas PC2 and PC7 or trypsin-like serine pro-
teases were poorly affected.15
O
O
H
N
H
N
N
H
N
H
O
O
NH2
NH
NH
NH
HN
NH2
HN
NH2
1
Modeling of the binding mode of these inhibitors15 based on the
X-ray structure of mouse furin inhibited by the irreversible inhib-
itor decanoyl-Arg-Val-Lys-Arg-chloromethyl ketone (1p8j.pdb)16
revealed that the side chain of the P3 Val is directed away from
the enzyme surface towards the solvent and has only loose contact
with the carboxyl group of the Glu257 side chain (distance ꢀ 3.5–
4 Å). Therefore, in this study, we investigated the modification of
this P3 residue and prepared additional analogues within this lead
structure for comparison.
peptides, such as nona-D ; substrate-analogue peptide
-Arg10
mimetics containing non-cleavable ketone-derived peptide bond
replacements between the P1 and P10 residues11; and various
non-peptidic derivatives based on highly potent guanylated 2,5-
dideoxystreptamines,12 with Ki values <10 nM. Additional non-
peptidic lead structures have also been reported; however, most
of these compounds exhibit reduced potency.13,14
We have recently reported the first furin inhibitors containing a
C-terminal 4-amidinobenzylamide as a decarboxylated arginine
mimetic. The most potent of these inhibitors, Phac-Arg-Val-Arg-
4-Amba 1, inhibits furin with a Ki value of 0.81 nM and has a sim-
ilar inhibitory potency against the proprotein convertases PC1/3,
All inhibitors were prepared by a combination of solid phase
and solution synthesis according to the strategy described previ-
ously.15 Briefly, the side-chain protected P5-P2-segment was syn-
thesized by
a Fmoc-protocol on 2-chloro-tritylchloride resin
using HBTU as a coupling reagent. After weak acidic cleavage from
the resin the intermediates were coupled to unprotected 4-amidin-
obenzylamine ꢁ 2HCl using PyBOP/DIPEA in the presence of 6-Cl-
HOBT, followed by final side chain deprotection (see Scheme 1
for the synthesis of inhibitors 2 and 7).
⇑
Corresponding author. Tel.: +49 6421 2825900; fax: +49 6421 2825901.
E-mail address: steinmetzer@staff.uni-marburg.de (T. Steinmetzer).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.06.091

4696
G. L. Becker et al. / Bioorg. Med. Chem. Lett. 21 (2011) 4695–4697
Scheme 1. Synthesis of inhibitors 2 and 7 on Fmoc-Arg(Pbf)-2-chlorotrityl resin. Reagents and conditions: (a) standard Fmoc SPPS; (b) 1% TFA in DCM, 2 ꢁ 30 min; (c) (i)
1.5 equiv 4-amidinobenzylamine ꢁ 2HCl, 1.1 equiv PyBOP, 3 equiv 6-Cl-HOBt, 3 equiv DIPEA in DMF, 2 h, (ii) TFA/TIS/H2O (95/2.5/2.5, v/v/v), 3 h, 35 °C; (d) 5 equiv. 1H-
pyrazole-1-carboxamidine ꢁ HCl in 1 NaCO3, 48 h. All final compounds were purified by reversed phase HPLC and obtained as lyophilized powders.
Inhibition constants varied approximately 400-fold (between 1
and 400 nM) among derivatives containing one of the 20 proteino-
genic amino acids in the P3 position (Fig. 1). Compared to lead 1,
nearly identical inhibitory activity was found for the Ile inhibitor
3 (Ki of 0.84 nM), whereas the structurally similar Leu derivative
14 was, surprisingly, more than 10-fold less potent. We do not be-
lieve that this drop in affinity is related to a simple sterical hin-
drance of the Leu side chain, because a relatively high potency
was also found for the sterically more demanding Phe and Tyr ana-
logues 4 and 5. In contrast, poor activity (Ki >100 nM) was found
for the Glu and Asp inhibitors 21 and 22, most likely due to electro-
static repulsion from residue Glu257 and the negatively charged
active site of furin. Although we expected some benefit from the
incorporation of an additional basic amino acid (Arg, Lys) in the
P3 position, the resulting inhibitors 6 and 8 showed slightly de-
creased potency compared to 1. Based on our model15 with the
Val in P3 position we assume that the side chains of Arg or Lys
are also directed towards the solvent and might be too long and
flexible to enable any direct interactions with Glu257. To reduce
Gly255, which is present in the X-ray structure with a P3 Val,16
but cannot be formed from a prolyl residue due to the lack of an
amide hydrogen.
In a second series (Table 1) we have modified other residues,
mainly in the P4 and P2 positions. The vast majority of furin sub-
strates contain an Arg at P4 and P1 positions as a minimal recogni-
tion requirement.1,17 However, a study regarding processing of the
wild type pro-insulin-like growth factor 1A demonstrated, that
also Lys can be accepted as an alternative P4 residue, although
even in that case an Arg residue in a mutated analogue is fa-
voured.18 Therefore, compound 23 containing a P4 Lys was pre-
pared, but it was found to be a poor furin inhibitor with a 300-
fold reduced Ki-value compared to compound 1, containing an
Arg at this position. This drop in the affinity was not surprising, be-
cause it is known from the furin crystal structure (1p8j) that the P4
guanidino group is involved in a complex network of salt bridges to
the carboxyl groups of Asp264 and Glu236 in a relatively solvent
excluded environment and forms also an additional hydrogen bond
to the OH of Tyr308. Therefore, it seems to be impossible that all of
these interactions can be maintained with a Lys residue in P4 posi-
tion. Interestingly, the complete elimination of the basic P4 side
chain, as present in the neutral urea-containing citrulline deriva-
tive 24 and the protected Lys(Cbz) inhibitor 25, provided com-
pounds with similar potencies as the Lys inhibitor. This also
clearly indicates that a Lys cannot be used as an alternative basic
P4 residue in these substrate analogue inhibitors.
their side chain length we also incorporated the shorter a,b-diami-
nopropionic acid residue (Dap) in the P3 position and converted its
side chain amine into a guanidino group. Whereas the Dap inhibi-
tor 7 exhibited inhibitory activity similar to the Arg compound, its
guanylated analogue 2 was more potent, inhibiting furin with a Ki
value of 0.83 nM.
Relatively weak potencies, with Ki values of 40 and 36 nM, were
found for the P3 Gly and Pro inhibitors, respectively. This might be
explained by the lack of any P3 side chain in the case of the P3 Gly
inhibitor, and by the elimination of the hydrogen bond between
the P3 backbone with the carbonyl oxygen of the furin residue
Peptidic compounds often suffer from limited stability in bio-
logical systems due to proteolytic degradation. One strategy to
overcome this problem is the incorporation of amino acids in
configuration, which was applied also for the design of nona-
D
-
-
D
Arg. Therefore, some analogues containing
D-amino acids were pre-
pared. However, inhibitors 26 and 27, containing
D
Val or Ala in
D
the P3 position, were approximately 1000-fold less potent than
Table 1
Inhibition of furin by inhibitors of the type P5-P4-P3-P2-4-Amba (compound 1 is
included as reference).
No.
P5
P4
P3
P2
Ki (nM)
1
Phac
Phac
Phac
Phac
Phac
Arg
Lys
Cit
Lys(Cbz)
Arg
Val
Val
Val
Val
Arg
Arg
Arg
Arg
Arg
0.81
285
238
702
1110
23
24
25
26
DVal
27
28
29
Phac
Phac
Phac
Arg
Arg
Arg
Arg
1385
970
D
Ala
Val
D
Arg
Val
7340
DArg
30a
–
–
Acetyl
Phac
Val
Arg
Arg
2390
3200
31a
DArg
32b
33
34
Phac
Phac
Phac
Arg
Arg
Arg
Val
Val
Dap
N
(Me)Arg
142
1.5
3.7
a
Figure 1. Inhibition of h-furin by inhibitors of the type Phac-Arg-P3-Arg-4-
amidinobenzylamide. Ki measurements were performed as described previously
and are the average of at least two measurements.13,15 Inhibitors with basic and
acidic P3 residues are shown with dark gray and black bars, respectively. Other
potent analogues having Ki values <2 nM are presented with white bars and the
remaining derivatives in gray.
Lys
Lys
a
These analogues are missing any P5 residue.
This inhibitor was synthesized using commercially available Fmoc-N (me-
b
a
thyl)Arg(Mtr)-OH (Bachem) for loading of the 2-chloro-tritylchloride resin.

G. L. Becker et al. / Bioorg. Med. Chem. Lett. 21 (2011) 4695–4697
4697
inhibitor 1. It should be noted that the inversion of the stereo-
chemistry is not simply an exchange of the hydrogen and side
In summary, we have synthesized several new substrate ana-
logue furin inhibitors with a 4-amidinobenzylamide as the P1 res-
idue; various derivatives have inhibition constants in the low
nanomolar range. Due to their close structural similarity to our
previously described inhibitor 1, which is also a highly potent
inhibitor of PC1/3, PC5/6 and PACE4, we assume that at least some
of these analogues should also inhibit other PCs (data not avail-
able). The further optimization of the P5 position within this inhib-
itor type will be described in a following publication.
chain attached to the C -carbon of an amino acid, but also has a
a
strong impact on the backbone / and
w
dihedral angles. In the
Ramachandran plot the torsion angles of
D-amino acid residues dif-
fer by a 180° rotation from that of L
-amino acids.19 We assume that
a change in the backbone conformation completely disrupts
important interactions between the Arg side chain and the S4
pocket of furin. Indeed, similar Ki values were found for derivative
30, which is missing the P4 Arg residue, and for compounds 25 and
28 which contain a Lys(Cbz) or
reduced potency was found for compound 29, which contains a P4
Val in combination with Arg in the P3 position. This inhibitor
might be useful as a negative control in cell culture studies for
detection of non-specific effects, because this structure should
have a similar overall physicochemical profile as compound 1. Ana-
D
Arg in that position. Even further
Acknowledgment
D
The authors would like to thank Iris Lindberg for the generous
gift of furin, as well as for reading and correcting the English ver-
sion of the manuscript.
logue 30, which is missing the P4 Arg and contains
D
Arg as the P3
References and notes
residue, was even slightly more potent than compound 29. It
should be noted that there are relatively small differences in the
Ki values between inhibitors 25–31 (ꢀfactor of 10); therefore, we
assume that they are still able to bind with their P2–P1 segment,
which acts as a kind of anchor that provides a basic affinity to furin
in the micromolar range.
1. Rockwell, N. C.; Krysan, D. J.; Komiyama, T.; Fuller, R. S. Chem. Rev. 2002, 102,
4525.
2. Thomas, G. Nat. Rev. Mol. Cell Biol. 2002, 3, 753.
3. Seidah, N. G. Ann. N.Y. Acad. Sci. 2011, 1220, 149.
4. Tian, S.; Huang, Q.; Fang, Y.; Wu, J. Int. J. Mol. Sci. 2011, 12, 1060.
5. De Vos, L.; Declercq, J.; Rosas, G. G.; Van Damme, B.; Roebroek, A.; Vermorken,
F.; Ceuppens, J.; van de Ven, W.; Creemers, J. Int. J. Oncol. 2008, 32, 1073.
6. Fugere, M.; Day, R. Trends Pharmacol. Sci. 2005, 26, 294.
7. Sarac, M. S.; Peinado, J. R.; Leppla, S. H.; Lindberg, I. Infect Immun. 2004, 72, 602.
8. Basak, A. J. Mol. Med. 2005, 83, 844.
9. Bontemps, Y.; Scamuffa, N.; Calvo, F.; Khatib, A. M. Med. Res. Rev. 2007, 27, 631.
10. Kacprzak, M. M.; Peinado, J. R.; Than, M. E.; Appel, J.; Henrich, S.; Lipkind, G.;
Houghten, R. A.; Bode, W.; Lindberg, I. J. Biol. Chem. 2004, 279, 36788.
11. Angliker, H. J. Med. Chem. 1995, 38, 4014.
12. Jiao, G. S.; Cregar, L.; Wang, J.; Millis, S. Z.; Tang, C.; O’Malley, S.; Johnson, A. T.;
Sareth, S.; Larson, J.; Thomas, G. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 19707.
13. Sielaff, F.; Than, M. E.; Bevec, D.; Lindberg, I.; Steinmetzer, T. Bioorg. Med. Chem.
Lett. 2011, 21, 836.
14. Komiyama, T.; Coppola, J. M.; Larsen, M. J.; van Dort, M. E.; Ross, B. D.; Day, R.;
Rehemtulla, A.; Fuller, R. S. J. Biol. Chem. 2009, 284, 15729.
15. Becker, G. L.; Sielaff, F.; Than, M. E.; Lindberg, I.; Routhier, S.; Day, R.; Lu, Y.;
Garten, W.; Steinmetzer, T. J. Med. Chem. 2010, 53, 1067.
An additional strategy to stabilize peptides against proteolytic
degradation is the modification of peptide bonds, for example, by
a
the incorporation of N (methyl)amino acids. As one example Ingles
and Knowles demonstrated that the methylation of the P2–P1
amide bond has a dramatic influence on the
lyzed hydrolysis of acetyl-N (methyl)-tyrosine methylester, which
a
-chymotrypsin-cata-
a
was more than 100,000-fold less efficiently cleaved (based on kcat
/
Km values) compared to the non-methylated analogue.20 We have
previously used this approach to stabilize hirudin-derived bivalent
thrombin inhibitors against thrombin degradation by incorpora-
tion of N(methyl)arginine as the P1 residue.21 However, the incor-
a
poration of N (methyl)Arg as P2 residue resulted in a nearly 175-
16. Henrich, S.; Cameron, A.; Bourenkov, G. P.; Kiefersauer, R.; Huber, R.; Lindberg,
I.; Bode, W.; Than, M. E. Nat. Struct. Biol. 2003, 10, 520.
17. Remacle, A. G.; Shiryaev, S. A.; Oh, E. S.; Cieplak, P.; Srinivasan, A.; Wei, G.;
Liddington, R. C.; Ratnikov, B. I.; Parent, A.; Desjardins, R.; Day, R.; Smith, J. W.;
Lebl, M.; Strongin, A. Y. J. Biol. Chem. 2008, 283, 20897.
fold loss in potency of inhibitor 32 as compared to 1. This was
not expected, because based on the X-ray structure of decanoyl-
Arg-Val-Lys-Arg-CMK-inhibited mouse furin we did not expect
any steric hindrance, and the P2 backbone NH was neither involved
in a hydrogen bridge to furin nor to a surrounding water molecule.
In contrast to the P4 position a P2 Lys is well accepted in many
furin substrates1,17 and substrate-analogue inhibitors.11,22 This was
confirmed in this study with inhibitors 33 and 34, which exhibit
only slightly decreased potency compared to the P2 Arg analogues
1 and 7.
18. Duguay, S. J.; Milewski, W. M.; Young, B. D.; Nakayama, K.; Steiner, D. F. J. Biol.
Chem. 1997, 272, 6663.
19. Horng, J. C.; Kotch, F. W.; Raines, R. T. Protein Sci. 2007, 16, 208.
20. Ingles, D. W.; Knowles, J. R. Biochem. J. 1968, 108, 561.
21. Steinmetzer, T.; Batdordshjin, M.; Pineda, F.; Seyfarth, L.; Vogel, A.; Reissmann,
S.; Hauptmann, J.; Stürzebecher, J. J. Biol. Chem. 2000, 381, 603.
22. Hallenberger, S.; Bosch, V.; Angliker, H.; Shaw, E.; Klenk, H. D.; Garten, W.
Nature 1992, 360, 358.