Benzodiazepines as potent and selective bradykinin B1 antagonists

Article abstract of DOI:10.1021/jm034020y

Antagonism of the bradykinin B₁ receptor was demonstrated to be a potential treatment for chronic pain and inflammation. Novel benzodiazepines were designed that display subnanomolar affinity for the bradykinin B₁ receptor (Ki = 0.59 nM) and high selectivity against the bradykinin B₂ receptor (K_i > 10 nM). In vivo efficacy, comparable to morphine, was demonstrated for lead compounds in a rodent hyperalgesia model.

Full text of DOI:10.1021/jm034020y

J . Med. Chem. 2003, 46, 1803-1806
1803
a nociceptive spinal reflex assay in the rabbit demon-
strated the analgesic efficacy of a different peptide B1
antagonist,8 B9858,9 under conditions modeling chronic
pain and inflammation. The lack of efficacy of B9858
in the absence of prior inflammation induced by com-
plete Freund’s adjuvant (CFA) implies that a selective
bradykinin B1 antagonist should restrict its activity to
sites of prior insult, thereby limiting undesirable side
effects. While these known peptidic antagonists have
helped validate bradykinin B1 receptors as a target, they
do not have the physical properties or pharmacokinetic
characteristics required of an orally active pharmaceuti-
cal. Herein, we present the development of selective,
small-molecule, non-peptidic bradykinin B1 antagonists
with demonstrated in vivo efficacy against inflammatory
pain.
Screening of our in-house sample collection identified
benzodiazepine 1 as an antagonist of the human brady-
kinin B1 receptor (Ki ) 28.0 nM; fluorescence imaging
plate reader (FLIPR) IC50 ) 164 nM) with excellent
selectivity over the bradykinin B2 receptor (Ki > 10
µM).10 Related compounds were therefore synthesized.
Ch em istr y. The compounds appearing in Table 1
were prepared according to Scheme 1. The benzodiaze-
pine cores, if not commercially available, were prepared
from 2-aminobenzonitrile and the requisite Grignard
reagent.11 The isocyanates were prepared in a straight-
forward manner from 4-fluoronitrobenzene and com-
mercially available amines via a three-step sequence.
This entailed nucleophilic aromatic fluoride displace-
ment followed by a Pd/C catalyzed hydrogenation of the
nitro group to form the aniline, which was converted in
situ to the corresponding isocyanate with triphosgene.12
Biologica l Resu lts a n d Discu ssion . Ki values (nM)
were determined radiometrically using the appropriate
radioligand and Chinese hamster ovary (CHO) cells
stably expressing the human bradykinin B1, rat B1, or
human B2 receptor. In vitro functional activity was
assessed in standard FLIPR experiments (IC50, nM)
employing functionally active human B1 or rat B1
receptors. Full details for the above experiments are
described in the Supporting Information.
The optimization of lead compound 1 initially focused
on increasing the binding affinity to the human brady-
kinin B1 receptor. Cognizant of the high frequency with
which basic moieties appear in ligands for G-protein-
coupled receptors, we believed that the dimethylamino
terminus was important for binding. Consistent with
this hypothesis, replacement of the dimethylamino
moiety with a hydrogen provided a compound (2) devoid
of bradykinin B1 receptor binding. Since removal of the
basic side chain was clearly deleterious, modification of
the dimethylamino group was pursued. These efforts
resulted in the introduction of a piperidine terminus
that more than doubled the affinity and functional
activity over the initial lead compound (comparing a
racemic mixture of 3 and 4 to the racemate 1). Separa-
tion of enantiomers13 demonstrated that the (R) enan-
tiomer (3) was 2 orders of magnitude more potent than
the (S) enantiomer (4). Assignment of the (R) configu-
ration as the higher affinity enantiomer was based in
Ben zod ia zep in es a s P oten t a n d Selective
Br a d yk in in B1 An ta gon ists
Michael R. Wood,*,† J une J . Kim,† Wei Han,†
Bruce D. Dorsey,†,# Carl F. Homnick,†
Robert M. DiPardo,† Scott D. Kuduk,†
Tanya MacNeil,| Kathy L. Murphy,‡ Edward V. Lis,‡
Richard W. Ransom,‡ Gary L. Stump,‡
J oseph J . Lynch,‡ Stacey S. O’Malley,‡
Patricia J . Miller,‡ Tsing-Bau Chen,‡
Charles M. Harrell,‡ Raymond S. L. Chang,‡
Punam Sandhu, J oan D. Ellis, Peter J . Bondiskey,
Douglas J . Pettibone,‡ Roger M. Freidinger,† and
Mark G. Bock†
Departments of Medicinal Chemistry and
Neuroscience, Merck Research Laboratories,
P.O. Box 4, West Point, Pennsylvania 19486
Received J anuary 28, 2003
Abstr a ct: Antagonism of the bradykinin B1 receptor was
demonstrated to be a potential treatment for chronic pain and
inflammation. Novel benzodiazepines were designed that
display subnanomolar affinity for the bradykinin B1 receptor
(Ki ) 0.59 nM) and high selectivity against the bradykinin B2
receptor (Ki > 10 µM). In vivo efficacy, comparable to mor-
phine, was demonstrated for lead compounds in a rodent
hyperalgesia model.
In tr od u ction . Bradykinin B1 and B2 receptors are
G-protein-coupled receptors that function in pain and
inflammation pathways.1 The peptides, bradykinin and
kallidin, act as the physiological agonists for the con-
stitutively expressed bradykinin B2 receptor to evoke
acute pain immediately after tissue injury.2 After physi-
cal trauma, bradykinin (BK) and kallidin are metabo-
lized to [des-Arg9]BK and [des-Arg10]kallidin, which
serve as the natural agonists for the bradykinin B1
receptor.3 While this receptor is not widely expressed
in nondiseased states, it is induced upon injury and is
believed to be active during persistent pain and inflam-
mation. Recent studies suggest that the bradykinin B1
receptor is constitutively expressed in the central
nervous system of mice4 and rats,5 indicating a poten-
tially central role for these receptors in addition to the
accepted peripheral mode of action.
The therapeutic potential for a selective bradykinin
B1 antagonist has been supported by recent studies
demonstrating that B1 receptor knockout mice are
outwardly normal and fertile while displaying a reduced
sensitivity to various types of noxious, painful stimuli
as well as inflammatory agents, such as carrageenan.6
Additionally, known peptidic antagonists, [des-Arg9,
Leu8]BK and [des-Arg10,Leu9]kallidin, have shown prom-
ising results in chronic pain animal models.7 Recently,
* To whom correspondence should be addressed. Phone: 215-652-
9451. Fax: 215-652-3971. E-mail: michael_wood2@merck.com.
† Department of Medicinal Chemistry.
#
Current address: Locus Discovery, Inc., Four Valley Square, 512
Township Line Road, Blue Bell, PA 19422.
| Department of Obesity and Metabolic Research, Merck Research
Laboratories, P.O. Box 2000, Rahway, NJ 07065.
Department of Drug Metabolism.
‡ Department of Neuroscience.
10.1021/jm034020y CCC: $25.00 © 2003 American Chemical Society
Published on Web 04/03/2003

1804 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 10
Letters
Ta ble 1. Bradykinin B1 Receptor Binding Affinities and Rat Pharmacokinetics for the Benzodiazepines
b
b
human Ki
human FLIPR
IC50 (nM)
rat Ki
(nM)
rat FLIPR
IC50 (nM)
rat %F,
t1/2 (h)
b
b
compda
R
X
Y
(nM)
(()-1
(()-2
(R)-3
(S)-4
(R)-5
(R)-6
(R)-7
(()-8
(()-9
(()-10
(()-11
(R)-12
c-hexyl
c-hexyl
c-hexyl
c-hexyl
phenethyl
p-tolyl
c-hexyl
c-hexyl
c-hexyl
c-hexyl
phenethyl
phenethyl
CH
CH
CH
CH
CH
CH
N
N
N
N
CH
N
N,N-dimethylamino
H
28.0
>10000
5.73
1023
1.37
13.4
1.33
85.0
164
1-piperidinyl
1-piperidinyl
1-piperidinyl
1-piperidinyl
4-pyridinyl
3-pyridinyl
2-pyridinyl
phenyl
24.7
2200
4.83
64.5
8.00
18.3
59.0
10((3), 3.2 ((0.7)
1.97
17.0
7.25
27.0
24.2
4.9((4.5), 2.4((0.3)
32((14), 4.3((0.9)
2.5((0.4), 2.3((0.6)
97.5
>10000
55.0
4-pyridinyl
4-pyridinyl
0.59
1.90
0.92
2.80
3.4((0.5), 2.7((0.5)
a
b
All compounds were >95% pure by HPLC and characterized by 1H NMR and HRMS. Values represent the numerical average of at
least two experiments. Interassay variability was (10% for the binding assays and (25% for the FLIPR experiments.
Sch em e 1
F igu r e 1. Relocation of the basic side chain from the C-3
position to the end of the N-1 alkyl group provides a new class
of achiral compounds that retain significant bradykinin B1
affinity.
analogy to C-5 phenyl, N-1-methylbenzodiazepine ana-
logues of known configuration (not shown).
Further enhancement of the inhibitory activity of
these compounds was achieved through modification of
racemic 11 to the single enantiomer 12) could be
attributed to a combination of effects. The reduced
the benzodiazepine C-5 substituent. In general, alkyl
groups extending less distance than cyclohexyl (methyl,
phenyl, tert-butyl) at this location afforded substantially
decreased affinity (data not shown). However, elongat-
ing the C-5 substituent from c-hexyl to phenethyl (5)
enhanced affinity by 4-fold at the human receptor and
9-fold at the rat bradykinin B1 receptor (compare to 3).
Correspondingly, the bradykinin B1 functional activity
of 5 was improved 5-fold and 8-fold at the human and
rat receptors, respectively. Supporting the hypothesis
that extended alkyl groups were preferred at this
location, the addition of just a single methyl group at
the 4-position on a phenyl ring (6) restored potency to
within roughly 2-fold of 3.
basicity of the pyridinyl nitrogen concomitant with a
significant change in conformation represents the most
likely causes for the decreased binding affinity of 11.
Combining the potency-enhancing modifications de-
scribed above into a single antagonist provided com-
pound 12. Consistent with the concept of additive
structure-activity relationships (SAR), this compound
possessed subnanomolar affinity for both the human
and rat bradykinin B1 receptors while also displaying
excellent functional activity.
Figure 1 depicts an extensive structural modification
of antagonist 3 that still allows for high affinity at the
bradykinin B1 receptor. In this achiral molecule 13, the
entire basic side chain has been relocated to the end of
the N-1 alkyl group (of optimized chain length) and
attached by an amide. Surprisingly, only a 25-fold
reduction in affinity is observed, relative to 3. Examina-
tion of models reveals that, because of the flexibility of
the n-alkyl linker at the N-1 position, the basic func-
tionality still has the ability to access a similar region
of space compared to previous analogues where the
attachment occurs at the C-3 position. However, the
possibility of a second basic-binding pocket on the
receptor cannot be ruled out. An alternative explanation
for this compound’s B1 receptor affinity could be that
the basic side chain has not altered its location in the
binding site. Instead, the benzodiazepine portion may
have moved into a different position. Regardless of
which explanation is correct, this example and the
preceding modifications highlight the importance of
having a properly configured basic side chain.14
Another breakthrough in affinity occurred when the
piperidinyl-piperidinyl portion of 3 was replaced with
a 4-pyridinylpiperazinyl motif (compound 7). This modi-
fication imparted a 4-fold increase in affinity relative
to its predecessor 3. However, the position of the
pyridine nitrogen proved to be critical. Movement of this
nitrogen to either the 3- or the 2-position (racemic 8 and
9, respectively) precipitated a drop in affinity of over
30-fold relative to 7 (single enantiomer). As with com-
pound 2, deletion of the basic pyridine nitrogen to
provide 10 abrogated affinity for the bradykinin B1
receptor. The remaining aniline nitrogen in 10 is either
not sufficiently basic to bind to the receptor or not in
the proper location to bind effectively. Also detrimental
to binding, albeit less so, was the replacement of the
distal piperazine nitrogen (X ) CH), resulting in
compound 11. This 50-fold drop in affinity (comparing

Letters
In Vivo Ch a r a cter iza tion . We were encouraged to
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 10 1805
potential for a bradykinin B1 antagonist in the treat-
ment of chronic inflammatory pain.
find that the first compound tested in rat, 4, displayed
10% oral bioavailability with an intravenous (iv) half-
life of 3.2 h. Unfortunately, with the introduction of the
potency-enhancing phenethyl group at the C-5 position
in 5, there was a 2-fold reduction in bioavailability and
a drop in half-life (Table 1). However, while introduction
of a p-tolyl group at this C-5 position decreased affinity
only slightly, there was an improvement in the rat
pharmacokinetics for 6 to the highest levels observed:
32% oral bioavailability with a 4.3 h iv half-life. Modi-
fication of the basic side chain by introduction of the
other potency enhancer, the 4-pyridinylpiperazine found
in 7, reduced bioavailability 4-fold relative to 3. Follow-
ing this downward trend, the bioavailability for 12, now
containing both potency-enhancing modifications, was
in line with the values obtained for its closest analogues
(5 and 7). Although the bioavailabilities were not
acceptable for oral dosing, compounds 3 and 12 were of
sufficient potency to be tested in a rat pain model.
Ack n ow led gm en t. We thank Sarah Grimwood,
Mark Levorse, Glenn Mason, Pawel Richards, and
J anine Webb for their preliminary work with compound
1. We thank Merryl Cramer for her work on the rat
bioavailability studies.
Su p p or tin g In for m a tion Ava ila ble: Spectral data (1H
NMR and HPLC), HRMS data of new compounds, a repre-
sentative description for the preparation of 12, experimental
details for the bradykinin B1/B2 binding and FLIPR assays,
and the protocol for rat pharmacokinetics. This material is
available free of charge via the Internet at http://pubs.acs.org.
Refer en ces
(1) (a) Bock, M. G.; Longmore, J . Bradykinin antagonists: new
opportunities. Curr. Opin. Chem. Biol. 2000, 4, 401-406. (b)
Couture, R.; Harrisson, M.; Vianna, R. M.; Cloutier, F. Kinin
receptors in pain and inflammation. Eur. J . Pharmacol. 2001,
429, 161-176.
(2) Regoli, D.; Barabe´, J . Pharmacology of bradykinin and related
kinins. Pharmacol. Rev. 1980, 32, 1-46.
(3) Marceau, F. Kinin B1 receptors: a review. Immunopharmacology
1995, 30, 1-26.
The model chosen was a carrageenan-induced hyper-
algesia assay in the rat.15 In brief, hyperalgesia is
induced in the hind paw of a Sprague-Dawley rat by
the intraplantar injection of carrageenan. The ability
of a compound to inhibit this hyperalgesia is then
measured by the latency of the inflamed hind paw to
withdraw from the application of increasing pressure,
where suppressed hyperalgesia is expressed in terms
of a percentage relative to untreated animals. Com-
pound 3 (rat Ki ) 18.3 nM; FLIPR IC50 ) 59.0 nM)
administered via intraperitoneal (ip) injection was inef-
fective at the 1 mg/kg level. However, when dosed at 3
and 10 mg/kg, 27 ( 15% and 80 (10% inhibition of
hyperalgesia (n ) 8/group) was observed in a dose-
dependent fashion. As anticipated, the 20-fold more
(4) Seabrook, G. R.; Bowery, B. J .; Heavens, R.; Brown, N.; Ford,
H.; Sirinathsinghi, D. J . S.; Borkowski, J . A.; Hess, J . F.; Strader,
C. D.; Hill, R. G. Expression of B1 and B2 bradykinin receptor
mRNA and their functional roles in sympathetic ganglia and
sensory dorsal root ganglia neurones from wild-type and B2
receptor knockout mice. Neuropharmacology 1997, 36, 1009-
1017.
(5) (a) Ma, Q.-P.; Hill, R. G.; Sirinathsinghji, D. J . Basal expression
of bradykinin B1 receptor in peripheral sensory ganglia in the
rat. NeuroReport 2000, 11, 4003-4005. (b) Wotherspoon, G.;
Winter, J . Bradykinin B1 receptor is constitutively expressed
in the rat sensory nervous system. Neurosci. Lett. 2000, 294,
175-178.
(6) Pesquero, J . B.; Arau´jo, R. C.; Heppenstall, P. A.; Stucky, C. L.;
Silva, J . A.; Walther T.; Oliveira, S. M.; Pesquero, J . L.; Paiva,
A. C.; Calixto, J . B.; Lewin, G. R.; Bader, M. Hypoalgesia and
altered inflammatory responses in mice lacking kinin B1 recep-
tors. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 8140-8145.
(7) (a) For a review see the following. Rupniak, N. M. J .; Longmore,
J .; Hill, R. G. Elucidation of the role of bradykinin B1 and B2
receptors in nociception and inflammation using selective an-
tagonists and transgenic mice. In Molecular Basis of Pain
Induction 2000; Wood, J . J ., Ed.; Wiley Press: New York, 2000;
pp 149-174. (b) Ferreira, J .; Campos, M. M.; Arau´jo, R.; Bader,
M.; Pesquero, J . B.; Calixto, J . B. The use of kinin B1 and B2
receptor knockout mice and selective antagonists to characterize
the nociceptive responses caused by kinins at the spinal level.
Neuropharmacology 2002, 43, 1188-1197.
(8) Mason, G. S.; Cumberbatch, M. J .; Hill, R. G.; Rupniak, N. M.
J . The bradykinin B1 receptor antagonist B9858 inhibits a
nociceptive spinal reflex in rabbits. Can. J . Physiol. Pharmacol.
2002, 80, 264-268.
(9) Stewart, J . M.; Gera, L.; Chan, D. C.; Whalley, E. T.; Hanson,
W. L.; Zuzack, J . S. Potent, long-acting bradykinin antagonists
for a wide range of applications. Can. J . Physiol. Pharmacol.
1997, 75, 719-724.
(10) Subsequent compounds described in this manuscript showed no
affinity for the bradykinin B2 receptor (Ki > 10 µM). However,
benzodiazepine scaffolds can be part of bradykinin B2 agonists
and antagonists. (a) Amblard, M.; Daffix, I.; Berge´, G.; Calme`s,
M.; Dodey, P.; Pruneau, D.; Paquet, J .-L.; Luccarini, J .-M.;
Be´lichard, P.; Martinez, J . Synthesis and characterization of
bradykinin B2 receptor agonists containing constrained dipeptide
mimics. J . Med. Chem. 1999, 42, 4193-4201. (b) Artis, D. R.;
Brotherton-Pleiss, C.; Pease, J . H. B.; Lin, C. J .; Ferla, S. W.;
Newman, S. R.; Bhakta, S.; Ostrelich, H.; J arnagin, K. Structure-
Based Design of Six Novel Classes of Nonpeptide Antagonists
of the Bradykinin B2 Receptor. Bioorg. Med. Chem. Lett. 2000,
10, 2421-2425. (c) Dziadulewicz, E. K.; Brown, M. C.; Dunstan,
A. R.; Lee, W.; Said, N. B.; Garratt, P. J . The design of non-
peptide human bradykinin B2 receptor antagonists employing
the benzodiazepine peptidomimetic scaffold. Bioorg. Med. Chem.
Lett. 1999, 9, 463-468.
potent analogue 12 (rat Ki ) 0.92 nM; FLIPR IC50
)
2.80 nM) performed even better. At a dose of 1 mg/kg
(ip), 27 ( 22% inhibition was observed, and at a dose of
3 mg/kg (ip), 89 ( 14% of the carrageenan-induced
hyperalgesia (n ) 6/group) was suppressed. By com-
parison, morphine dosed (ip) at 0.3, 1.0, and 3.0 mg/kg
resulted in 24 ( 19%, 74 ( 24%, and 108 ( 20%
inhibition (n ) 6/group), respectively, with the highest
dose showing early signs of hypoalgesia. Importantly,
the IC50 values for 12 against rodent opioid receptors
were all greater than 1 µM.16 Thus, the in vivo efficacy
of benzodiazepine 12 and morphine was of comparable
strength despite their different target receptors. While
the in vivo antihyperalgesic activities of 3 and 12 are
consistent with bradykinin B1 antagonism, this assay
is not specific for the B1 mechanism, and therefore,
contribution of potential unknown off-target activities
cannot be ruled out. It is considered, however, that
selective bradykinin B1 antagonists will be efficacious
against chronic pain and inflammation, similar to
morphine, but without the deleterious side effects
associated with the opiates.
In conclusion, novel bradykinin B1 receptor antago-
nists have been identified, with the most potent mem-
bers demonstrating subnanomolar binding affinity and
low-nanomolar functional activity. These compounds
have enabled the demonstration of in vivo efficacy
comparable to morphine for the suppression of hyper-
algesia. The above observations support the therapeutic
(11) Bock, M. G.; DiPardo, R. M.; Evans, B. E.; Rittle, K. E.; Veber,
D. F.; Freidinger, R. M.; Hirshfield, J .; Sringer, J . P. Synthesis
and resolution of 3-amino-1,3-dihydro-5-phenyl-2H-1,4-benzo-
diazepin-2-ones. J . Org. Chem. 1987, 52, 3232-3239.

1806 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 10
Letters
(12) A more detailed description for the preparation of these com-
pounds can be found in the Supporting Information. Serving as
a representative example, the preparation of compound 12 is
elaborated.
(13) Enantiomers were separated and analyzed for enantiomeric
excess (in all cases >98%) on L-leucine Pirkle type columns
(preparative and analytical) using chloroform and methanol
(0.1% diethylamine) as eluants.
(14) Non-peptidic bradykinin B1 antagonists containing basic side
chains have been reported. (a) Altamura, M.; Meini, S.; Quar-
tara, L.; Maggi, C. A. Nonpeptide antagonists for kinin receptors.
Regul. Pept. 1999, 80, 13-26. (b) Ferrari, B.; Gougat, J .;
Muneaux, C.; Muneaux, Y.; Perreaut, P. Patent WO 025315 A1,
1997. (c) Ferrari, B.; Gougat, J .; Muneaux, Y.; Perreaut, P.;
Sarran, L. Patent WO 076964 A1, 2002.
(15) Boyce, S.; Chan, C.-C.; Gordon, R.; Li, C.-S.; Rodger, I. W.; Webb,
J . K.; Rupniak, N. M. J .; Hill, R. G. L-745,337: a selective
inhibitor of cyclooxygenase-2 elicits antinociception but not
gastric ulceration in rats. Neuropharmacology 1994, 33, 1609-
1611.
(16) Additionally, 12 shows high selectivity (>100×) against a panel
of assays (Panlabs, MDS Pharma Services, Bothell, WA) repre-
senting 156 enzymes, receptor, and transporters (including
cannabinoid, dopamine, histamine, opiate, and GABAA benzo-
diazepine receptors and transporters) with the following excep-
tions: human tachykinin NK2 > 30×, human leukotriene B4
>
60×, human muscarinic M1 > 70×, and human muscarinic M4
> 90×.
J M034020Y