Discovery of new C3aR ligands. Part 1: Arginine derivatives

Article abstract of DOI:10.1016/j.bmcl.2007.04.022

The synthesis and in vitro binding of several new arginine-containing C3aR ligands are reported. DMPK properties and functional activities of selected compounds have been evaluated. One compound is shown to be active in an in vivo model of airway inflamma

Full text of DOI:10.1016/j.bmcl.2007.04.022

Bioorganic & Medicinal Chemistry Letters 17 (2007) 3258–3261
Discovery of new C3aR ligands. Part 1: Arginine derivatives
a,
b
a
a
b
*
´ ´
Frederic Denonne, Sophie Binet, Maggi Burton, Philippe Collart, Alan Dipesa,
Tanmoy Ganguly,b Alexander Giannaras,b Seema Kumar,b Timothy Lewis,b
Florence Maounis,b Jean-Marie Nicolas,a Tamsin Mansley,b Patrick Pasau,a
Dorin Preda,b Karin Stebbins,b Alexander Volosovb and Dong Zoub
aUCB Global Chemistry, Chemin du Foriest R4, 1420 Braine l’Alleud, Belgium
bUCB Research Inc., Cambridge, MA 02139, USA
Received 7 February 2007; revised 31 March 2007; accepted 4 April 2007
Available online 10 April 2007
Abstract—The synthesis and in vitro binding of several new arginine-containing C3aR ligands are reported. DMPK properties and
functional activities of selected compounds have been evaluated. One compound is shown to be active in an in vivo model of airway
inflammation after aerosol administration.
ꢀ 2007 Elsevier Ltd. All rights reserved.
The complement system is at the heart of the inflamma-
tory process. This complex system encompasses approx-
imately 30 components working together to recruit
phagocytes for the destruction of pathogens. It is acti-
vated by the immune complex and complements C1 acti-
vators. This results in C3 being cleaved to release C3a,
which in turn leads to the cleavage of C5 into C5a and
C5b. The latter binds together with C6–C9 to form the
membrane attack complex, responsible for cell lysis.
When left unchecked this very efficient system favors
the development of chronic inflammatory diseases such
as asthma, rheumatoid arthritis, multiple sclerosis, etc.
C3a in particular has been linked to various pro-inflam-
matory processes linked to allergic airway diseases, such
as contraction of human parenchymal strips,1 increase
of vascular permeability,2 and release of vasoactive
amines.3 Moreover, C3aR knock-out mice are charac-
terized by decreased airway hyperresponsiveness4 and
KO guinea pigs show decreased allergic response.5 In
humans, asthmatic patients have elevated levels of C3a
in the broncho-alveolar fluid6 and in the plasma.7
the periphery and the brain. It has been shown that
the C-terminus LPLPR sequence is essential for C3aR
activation.8 Disrupting this interaction seems a good
way to block the inflammatory processes linked to C3a.
HN NH2
Cl
Cl
Cl
NH
N
N
N
H
O S OO
N
Me
Cl
N
H
CO2H
2
1
Me
NH2
NH
HN
HN NH2
NH
O
Ph
O
O
O
N
CO2H
Ph
N
CO2H
H
H
3
4
5
This central role of C3a makes it an ideal drug target.
C3a is a 77-amino acid peptide which activates a G-pro-
tein coupled receptor called C3aR, widely expressed in
HN NH2
NH
Ph
O
O
Ph
N
H
CO2H
Keywords: C3a; Inflammation; Arginine.
*
Corresponding author. Tel.: +32 2 386 3272; fax: +32 2 386
2704; e-mail: frederic.denonne@ucb-group.com
Figure 1. Small molecule ligands of the C3a receptor.
0960-894X/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.04.022

F. Denonne et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3258–3261
3259
This approach has been followed by several others be-
fore us, and as a result, all known ligands feature an
arginine (compounds 1, 3, 4, and Fig. 1)9 or a bioisostere
in the case of 2.10 We selected 4 (pIC50 = 6.3) as a start-
ing point and we reasoned that we could gain some affin-
ity by rigidifying its ether chain. We were delighted to
see that compound 5 in which the ether chain is locked
within a furan ring had a much improved affinity for
C3aR indeed, with a pIC50 of 7.2.
‘Eastern’ substitution (data not shown) arose from
chemical modifications of protected amino-acids and
coupling with the carboxylic acid of interest (D or I,
Scheme 1).
All compounds were tested in a C3a binding assay. Var-
ious aryl (Table 1) or diarylmethine groups (Table 2)
were introduced on the Western side. In the first series,
it quickly became apparent that the 3-substitution of
the phenyl gave the best improvements in affinities com-
pared to 2- and 4-substituted phenyls (data not shown).
After some experimentation, a chlorine atom (12) was
found optimal. Additional substituents did not yield
any improvements (13–15). Attempts at replacing the
furan linker with a benzene (18 and 19) or a thiophene
(20) highlighted the essential role of the oxygen atom.
In the second series, the linker seemed as critical as in
the first series, as highlighted by close analogues of 5:
21 and 22. However, here again, the introduction of hal-
ogen atoms at the 3-position of phenyls gave rise to
more potent ligands 23 and 24. The most active com-
pounds were profiled internally against a panel of 20 dif-
ferent GPCRs and showed no additional affinity.
With this good starting point in hand we decided to
investigate its structure–activity relationships.
‘Western’ substitution was varied starting from 5-bro-
mofuran-2-carboxylates (B or G, Scheme 1). The ester
G was coupled under Suzuki conditions with a range
of aryl boronic acids F. It was then hydrolyzed and cou-
pled with arginine. On the other hand, carboxylic acid B
was treated with three equivalents of tert-butyllithium
and the resulting dianion was reacted with substituted
benzophenones A. The resulting tertiary alcohols C were
then reduced with triethylsilane and finally coupled with
arginine.
O
c
d
Ar
HO
Ar
Ar
OH
OH
OH
OH
Br
+
O
e
e
O
O
Ar Ar
O
O
O
O
O
Ar
H
D
A
B
C
N
R'
R
O
O
a
b
E
OMe
OMe
Ar B(OH)2
+
Ar
Ar
Br
O
O
I
O
O
F
H
G
Scheme 1. Synthetic route toward Western side modifications. Reagents and conditions: (a) Pd(PPh3)4, Na2CO3, DME, 100 ꢁC; (b) LiOH, THF/
H2O; (c) t-BuLi (3 equiv), THF, ꢀ70 to 22 ꢁC; (d) TFA, Et3SiH, DCM (Ar = Ph, 42% two steps); (e) (COCl)2, THF then R0NH2, K2CO3, THF/H2O.
Table 1. SAR data for biaryl-substituted arginines
O
CO H
2
H
Ar
N
NH
NH2
linker
N
H
a
Compound
Aryl
Linker
C3a pIC50
6
7
3-Fluorophenyl
3-Methylphenyl
2,5-Furyl
2,5-Furyl
2,5-Furyl
2,5-Furyl
2,5-Furyl
2,5-Furyl
2,5-furyl
6.5
6.9
6.5
6.7
6.1
6.7
7.1
6.7
6.1
5.8
6.5
5.4
4.0
6.1
5.2
8
3-Methoxyphenyl
3-Ethoxyphenyl
9
10
11
12
13
14
15
16
17
18
19
20
3-Hydroxymethylphenyl
3-Cyanophenyl
3-Chlorophenyl
3-Chloro-4-fluorophenyl
3-Chloro-6-methylphenyl
3,6-Dichlorophenyl
3-Benzothiophenyl
6-Dibenzofuryl
2,5-Furyl
2,5-Furyl
2,5-Furyl
2,5-Furyl
2,5-Furyl
1,4-Phenyl
1,3-Phenyl
2,5-Thienyl
3-Chlorophenyl
3-Chlorophenyl
3-Chlorophenyl
a Values are means of three experiments. Assays were performed in 96-well format using wheat germ agglutinin SPA beads (GE). Various con-
centrations of compounds were added, followed by 125I-C3a (2200 Ci/mmol, PE) at 0.02 nM in 120 ll of 50 mM Tris/2 mM MgCl2 gC/0.5% BSA.
Non-specific binding was determined using C3a at 1 nM. Pre-coupled beads and membranes prepared from HMC-1 cells (stably expressing both
Aequorin and the human C3a receptor) diluted in 80 ll buffer were then added. Plates were sealed and incubated at rt for 3 h. Bound C3a was
determined by scintillation counting on a Wallac Beta Trilux scintillation counter.

3260
F. Denonne et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3258–3261
Table 2. SAR data for triarylmethine-substituted arginines
Ar
O
CO H
H
2
N
NH
NH
Ar
N
linker
H
2
a
Compound
Diarylmethine
Linker
C3a pIC50
21
22
5
Phenyl
Phenyl
Phenyl
1,4-Piperazinyl
1,4-Phenyl
2,5-Furyl
4.3
6.1
7.2
7.5
7.8
6.8
23
24
25
3-Fluorophenyl
3-Chlorophenyl
4-Chlorophenyl
2,5-Furyl
2,5-Furyl
2,5-Furyl
a Values are means of three experiments.
On the other side of the molecule, it proved impossible
to replace the original arginine while maintaining a good
affinity: the three-carbon chain, the chirality, and the
carboxylic acid were all found essential for C3aR recog-
nition. Moreover, all modifications of the guanidine
function to make it less basic also led to compounds
with drastically reduced affinities.
Figure 2. Hepatic extraction of 5 after oral administration (rats).
ance (Clint of 4 lL/min/mg prot in NADPH-fortified
rat liver microsomes and 5 lL/min/106 cells in rat hepa-
tocytes) and low potential to inhibit human CYP2D6/
3A4 (12 and 0% inhibition at 50 lM, respectively). How-
ever, 5 showed a low apical-to-basolateral permeability
in Caco-2 assay (0.4 · 10ꢀ6 cm/s). In line with this find-
ing, the oral bioavailability in male rats was found to be
ca. <1% (single oral and iv dosing at 1 mg/kg). The mea-
sured concentrations in the portal vein and the vena
cava further confirmed that only a small fraction of
the dose is absorbed, while the hepatic extraction is min-
imal. Overall this suggests that all our arginine com-
pounds suffer from restricted intestinal absorption,
presumably because of high hydrogen bonding and/or
The functional activity of our best ligands was then
investigated (Table 3). In the first series, most of the li-
gands behaved as antagonists, such as 19 and 20. How-
ever, 12, our best ligand in this series, turned out to be a
full agonist. To our dismay this tendency was confirmed
in the other series, where our strongest ligand 24 showed
agonistic properties, while its p-dichloro analogue 23
was a partial agonist. Compound 5 was the best example
of this series to keep an antagonistic profile.
Our compounds were also evaluated for their in vitro
DMPK properties. Compound 5, representative of the
chemical family, showed low in vitro metabolic clear-
2
˚
high basicity. The high PSA (above 130 A on average)
of all these compounds might also account for their low
permeability11 (Fig. 2).
Table 3. Functional activities of selected compounds
Compound 5 was also tested in an in vivo model of
inflammation.12 Mice were treated with 5 either intra-
peritoneally or were exposed to inhale an aerosol of 5
for 30 min. The eosinophil count in the broncho-alveo-
Compound
SPA
binding
pIC50
Antagonist
activitya
pEC50
Agonist
activityb
pEC50
5
12
19
20
23
24
25
7.2
7.1
6.1
5.2
6.4
7.9
6.8
6.8
—
—
6.3
5.4
4.8
5.9
—
100
90
80
70
60
50
40
30
20
10
0
—
7.7
6.9
—
a Following the 2-h incubations with Coelenterazine, the cells were
plated and treated with compounds, dilutions (200 ll final volume,
1% DMSO) for 1 h at 37 ꢁC. Fifty microliters of C3a agonist (final
concentration of 5 nM) was then injected. Emitted light was imme-
diately recorded for 30 s using Novostar (BMG). Maximal inhibition
is determined using 4 as the reference antagonist (20 lM final
concentration).
100
30
10
1
3 mg/ml Ae
Dose(mg/kg, IP)
b HMC-1 cells stably expressing both Aequorin and the human C3a
receptor were loaded with 10 lM Coelenterazine in HBSS/Hepes
20 mM/BSA 0.2%, pH 7.4, for 2 h at 37 ꢁC. They were then diluted
four times, placed in 96-well plates (200 ll final volume, 1% DMSO),
and incubated for 1 h at 37 ꢁC. Fifty microliters of compound dilu-
tions was then injected to get a final volume of 250 ll at 1% DMSO
final concentration. Emitted light was immediately recorded for 30 s
using Novostar (BMG). The maximal signal was determined in the
presence of C3a agonist (Advanced Research Technologies, 50 nM).
Figure 3. Evaluation of 5 in vivo in an ovalbumin-triggered model of
airway inflammation. BalbC mice (n = 5) were treated by intra-
peritoneal injection of ovalbumin (OVA 8 lg + aluminum hydroxide
2 mg per mouse) at day 0 and day 14 and by aerosol (OVA 0.5% for
30 min) at days 21–23. Compound 5 was injected ip 1 h before and 4 h
after the aerosol on days 21–23 or inhaled for 30 s at 3 mg/ml. The
broncho-alveolar fluid was sampled on day 25 and the eosinophils
counted.

F. Denonne et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3258–3261
3261
lar fluid revealed that inhalation of 5 (3 mg/ml aerosol)
References and notes
was equally effective to a 30-mg/kg dose ip. The mea-
sured peak concentrations of 5 in the plasma were of
33,800 and 27 ng/ml for ip and aerosol administration,
respectively. Aerosol administration thus leads to a
thousandfold reduced plasma exposure with equal effec-
tiveness in preventing airway inflammation. In another
study, IL-13 levels in the broncho-alveolar fluid of mice
treated with 30 mg/kg of 5 were reduced to levels com-
parable to mice treated with 10 mg/kg ip of montelukast.
This in vivo proof of concept confirmed our hypothesis
that C3a was a valid target to treat airway inflammation
and that aerosol administration is a promising mode of
administration to circumvent the PK issues of this class
of compounds (Fig. 3).
1. Stimler, N. P. et al. Immunopharmacology 1983, 5, 251.
2. Marceau, F.; Hugli, T. E. J. Pharm. Exp. Ther. 1984, 230,
749.
3. Legler, D. F.; Loetscher, M.; Jones, S. A.; Dahinden, C.
A.; Arock, M.; Moser, B. Eur. J. Immunol. 1996, 26, 753.
4. Drouin, S. M.; Corry, D. B.; Hollman, T. J.; Kildsgaard,
J.; Wetsel, R. A. J. Immunol. 2002, 169, 5926.
5. Bautsch, W. et al. J. Immunol. 2000, 165, 5401.
6. (a) Humbles, A. A. et al. Nature 2000, 406, 998; (b) Krug,
N. et al. Am. J. Resp. Crit. Care Med. 2001, 164, 1841.
7. Nakano, Y.; Morita, S.; Kawamoto, A.; Suda, T.; Chida,
K.; Nakamura, H. J. Allergy Clin. Immunol. 2003, 112, 525.
8. Takenaka, Y.; Utsumi, S.; Yoshikawa, M. Biosci. Bio-
technol. Biochem. 2001, 65, 1202.
9. (a) Ames, R. S.; Lee, D.; Foley, J. J.; Jurewicz, A. J.;
Tornetta, M. A.; Bautsch, W.; Settmacher, B.; Klos, A.;
Erhard, K. F.; Cousins, R. D.; Sulpizio, A. C.; Hieble, J.
P.; McCafferty, J.; Ward, K. W.; Adams, J. L.; Bondinell,
W. E.; Underwood, D. C.; Osborn, R. R.; Badger, A. M.;
Sarau, H. M. J. Immunol. 2001, 166, 6341; (b) Therien, A.
G. J. Immunol. 2005, 174, 7479; (c) Lee, D.; Bondinell, W.
E.; Jurewicz, A. J. PCT Patent WO2000009129, 2000; (d)
Lee, D. PCT Patent WO9915490, 1999.
In conclusion, rigidification of known scaffold 4 led to
the discovery of new potent C3aR ligands. Those com-
pounds had poor bioavailability due to the presence of
the arginine side-chain. Attempted modification of this
essential part led to almost inactive compounds. Minute
structural modifications on the ‘Western’ part of the
molecule were shown to exert a drastic influence on
the functional activity. Antagonist 5 was shown to be ac-
tive in vivo, with aerosol administration leading to much
reduced plasma concentrations of active compound.
This suggests that such compounds could be useful for
the treatment of chronic and acute airway inflammatory
diseases.
10. Grant, E. B.; Guiadeen, D.; Singer, M.; Argentieri, D.;
Hlasta, D. J.; Wachter, M. Bioorg. Med. Chem. Lett. 2001,
2817.
11. Martin, Y. C. J. Med. Chem. 2005, 48, 3164.
12. Kung, T. T.; Jones, H.; Adams, G. K., 3rd; Umland, S. P.;
Kreutner, W.; Egan, R. W.; Chapman, R. W.; Watnick,
A. S. Int. Arch. Allergy Immunol. 1994, 105, 83.