Discovery of novel, orally active dual NK1/NK2 antagonists

Article abstract of DOI:10.1016/S0960-894X(01)00572-8

Exploration of the SAR around selective NK₂ antagonists, SR48968 and ZD7944, led to the discovery that naphth-1-amide analogues provide potent dual NK₁ and NK₂ antagonists. ZD6021 inhibited binding of [³H]-NKA or [³H]-SP to human NK₁ and NK₂ receptors, with high-affinity (K_i=0.12 and 0.62 nM, respectively). In functional assays ZD6021 had, at 10^-7 M, in human pulmonary artery pK_B=8.9 and in human bronchus pK_B=7.3, for NK₁ and NK₂, respectively. Oral administration of ZD6021 to guinea pigs dose-dependently attenuated ASMSP induced extravasation of plasma proteins, ED₅₀=0.5 mg/kg, and NK₂ mediated bronchoconstriction, ED₅₀=13 mg/kg.

Full text of DOI:10.1016/S0960-894X(01)00572-8

Bioorganic & Medicinal Chemistry Letters 11 (2001) 2769–2773
Discovery of Novel, Orally Active Dual NK1/NK2 Antagonists
Peter R. Bernstein,* David Aharony, Jeffrey S. Albert, Donald Andisik,
Herbert G. Barthlow, Russell Bialecki, Timothy Davenport, Robert F. Dedinas,
Bruce T. Dembofsky, Gerard Koether, Benedict J. Kosmider, Karin Kirkland,
Cyrus J. Ohnmacht, William Potts, William L. Rumsey, Lihong Shen, Ashok Shenvi,
Scott Sherwood, David Stollman and Keith Russell
CNS Discovery Research, AstraZeneca Pharmaceuticals LP, 1800 Concord Pike, PO Box 15437, Wilmington, DE 19850-5437, USA
Received 13 June 2001; accepted 9 August 2001
Abstract—Exploration of the SAR around selective NK2 antagonists, SR48968 and ZD7944, led to the discovery that naphth-1-
amide analogues provide potent dual NK1 and NK2 antagonists. ZD6021 inhibited binding of [3H]-NKA or [3H]-SP to human NK1
and NK2 receptors, with high-affinity (Ki=0.12 and 0.62 nM, respectively). In functional assays ZD6021 had, at 10ꢀ7 M, in human
pulmonary artery pKB=8.9 and in human bronchus pKB=7.3, for NK1 and NK2, respectively. Oral administration of ZD6021 to
guinea pigs dose-dependently attenuated ASMSP induced extravasation of plasma proteins, ED50=0.5 mg/kg, and NK2 mediated
bronchoconstriction, ED50=13 mg/kg. # 2001 Elsevier Science Ltd. All rights reserved.
The neurokinins, Substance P (SP), neurokinin A
(NKA), and neurokinin B (NKB), also known as the
tachykinins, are a family of closely related peptides.
They have been associated with many pathophysiologi-
cal conditions including: asthma, arthritis, cough,
emesis, anxiety, depression, and inflammatory bowel
disease. Because of this linkage much effort has gone
into the development of agents that can inhibit their
effects. Since the neurokinins act, respectively, through
three G-protein coupled receptors (called NK1, NK2,
and NK3) the development of selective tachykinin
receptor antagonists has been the focus of the bulk of
this effort.1,2
the NK2 selective antagonists, SR489686 and ZD79447
(Fig. 1). The results we obtained are the subject of this
report.
Figure 1. Functional activity (pKB) of lead NK2 selective antagonists,
as determined on rabbit pulmonary artery (RPA).
Both NK1 and NK2 receptors appear to be involved in
pulmonary pathophysiology.3 Therefore the hypothesis
has been advanced that treatment with a dual antago-
nist, that blocked both of these receptors, might provide
a more effective treatment for asthma.4 This hypothesis
has been based, in part, on models that showed either
additive or synergistic efficacy upon combining selective
NK1 and NK2 receptor antagonists.5 As our approach
to novel dual antagonists, as potential anti-asthmatic
agents, we chose to explore structural modification of
To begin, a broad array of compounds (1) were pre-
pared. In this set R2 was varied among several sub-
stituted piperidines8 and R1 was chosen from over 100
aryl, heteroaryl and arylalkyl groups. This array was
prepared using as the final synthetic stage, either an
acylation or a reductive amination (see Scheme 1).
*Corresponding author. Tel.:+1-302-886-8966; fax:+1-302-886-4989;
e-mail: peter.bernstein@astrazeneca.com
Scheme 1. General synthetic approach.
0960-894X/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(01)00572-8

2770
P. R. Bernstein et al. / Bioorg. Med. Chem. Lett. 11 (2001) 2769–2773
While this work was underway, several groups described
identification of dual acting antagonists starting from
selective NK2 antagonists.9ꢀ11 Alternatively, other
groups succeeded in producing dual acting compounds
starting with selective NK1 antagonists.12 The most
potent, dual acting compound to come out of this array
was naphthamide 2a (see Table 1). In this compound,
the piperidine substituent was the same as that found in
ZD7944, a potent and selective NK2 antagonist. Keep-
ing the naphthamide constant we explored a larger
group of piperidines13 and found that only the corre-
sponding sulfone 2h and pyridyl 2m analogues retained
good dual activity. The other piperidine groups led to
loss of activity either at the NK1 receptor or at both the
NK1 and NK2 receptors.
drawing group (e.g., CF3) is an important structural
feature of many NK1 antagonists.14 Commercially
available 3-nitronaphth-1-oic acid was of interest, both
directly for the 3-nitro analogue, and as a precursor to
other 3-substituted naphth-1-oic acids via reduction to
the amine and Sandmeyer type reactions. The nitro
analogue 3b proved particularly potent, both in vitro
and following intravenous testing in vivo in a guinea pig
pulmonary mechanics model.15 However because of
concern over its possible metabolic transformation to a
b-naphthyl amine derivative, alternative 3-substituents
were explored (Table 2).
Since substituent constants (p, MR, and s1) for cyano
(ꢀ0.57, 6.3, and 0.53) highlighted it as a replacement for
nitro (ꢀ0.28, 7.36, and 0.67), and as such arylnitriles are
not implicated as potential toxicophores, the 3-cyano
analogue (3d, ZD6021) was specifically targeted. The
nitro 3b and cyano 3d analogues proved to be the most
potent dual NK1/NK2 antagonists in this set. Unfortu-
nately, efforts to define a broader QSAR relationship
To follow-up the discovery of this naphthamide we
explored the effect of substituents on the naphthalene
ring. Substituents at the 3-position were a particularly
interesting target. This is because an aryl group sub-
stituted in the meta-position with an electron-with-
Table 1. Exploration of piperidines while keeping naphthamide constant
a
Compd
pKB
NK1
pKB
NK2
Compd
pKB
NK1
pKB
NK2
2, R2N–=
2, R2N–=
2a
7.89ꢁ0.08
8.18ꢁ0.28
2i
6.78ꢁ0.01
8.23ꢁ0.27
2b
2c
7.34ꢁ0.017
7.40ꢁ0.36
7.92ꢁ0.34
6.76ꢁ0.31
2j
6.42ꢁ0.08
6.82ꢁ0.27
8.44ꢁ0.26
8.41ꢁ0.36
2k
2d
7.18ꢁ0.15
8.18ꢁ0.38
2l
6.98ꢁ0.09
NAb
2e
2f
7.30ꢁ0.26
7.3ꢁ0.13
8.21ꢁ0.32
7.32ꢁ0.04
2m
2n
7.69ꢁ0.04
6.98ꢁ0.12
7.82ꢁ0.05
8.10ꢁ0.06
2g
2h
6.92ꢁ0.24
7.81ꢁ0.20
7.54ꢁ0.12
7.55ꢁ0.07
2o
2p
6.03ꢁ0.25
6.28ꢁ0.11
NA
NA
apKB values were determined on rabbit pulmonary artery on nꢂ2 tissues. For NK1 receptor antagonism ASMSP [Ac-[Arg6,Sar9,Met11(O2)]SP(6–
11)] was used as the agonist and for NK2 antagonism BANK [b-Ala8]NKA(4–10)] was the agonist.
bNA, not active (shift in dose–response curve <2) at an antagonist concentration of 1 mM.

P. R. Bernstein et al. / Bioorg. Med. Chem. Lett. 11 (2001) 2769–2773
2771
between substituent constants for the different groups at
position 3 and in vitro activity were unsuccessful.
Broader profiling of ZD6021 was undertaken (Table 5),
and comparisons made with a reference dual NK1/NK2
antagonist, MDL105212.9 Binding assays showed that
ZD6021 was slightly more potent than MDL105212 at
the cloned human receptors.16 However, in both animal
and human functional assays, it is much more potent.
Also, it is significantly more potent in vivo in the guinea
pig model. In the latter model an ED50 of 13 mg/kg was
determined for orally dosed ZD6021 against BANK-
induced (NK2) bronchoconstriction. Orally dosed
ZD6021 was also examined in an ASMSP-induced
(NK1) guinea pig model of extravasation of plasma
proteins, which showed an ED50=0.5 mg/kg.
Table 2. Exploration of varying substituents in 3-X-naphthamides
a
a
Compd
3 X¼
pKB
NK1
pKB
NK2
Dose ratiob
Because of this favorable profile, ZD6021 was chosen
for additional studies on its selectivity, pharmacokinetic
and toxicologic profiles. Selectivity of ZD6021 was
NK1 NK2
2a
3b
3c
3d
H
NO2
Br
CꢃN
7.89ꢁ0.08 8.18ꢁ0.28
8.16ꢁ0.10 9.03ꢁ0.18
8.15ꢁ0.34 7.67ꢁ0.24
52
50
43
262
321
34
8.98ꢁ0.17 8.26ꢁ0.10 144
74
Table 4. Relationship of stereochemistry to pharmacologic activitya
(ZD6021)
3e
3f
3g
3h
3i
3j
3k
3l
3m
3n
3o
3p
3q
SO2CH3
Cl
OMe
CO2H
CH3
7.43ꢁ0.25 7.35ꢁ0.04
7.15ꢁ0.12 7.10ꢁ0.09
7.95ꢁ0.04 7.70ꢁ0.06
22
13
47
28
31
77
ND
123
39
156
75
7
ND
34
52
ND
5.68ꢁ0.14 6.86ꢁ0.11 NDc
8.03ꢁ0.04 7.29ꢁ0.21
26
CH2CN
Ac
8.42ꢁ0.24 6.99ꢁ0.06 133
7.41ꢁ0.35 7.17ꢁ0.13
41
31
C(¼CH2)CH3 7.24ꢁ0.19 7.24ꢁ0.29
SO2NH2
CON(Me)2
CꢃCH
F
7.54ꢁ0.04 7.02ꢁ0.21 170
5.17ꢁ0.22 7.31ꢁ0.33 ND
Compd
Stereochemistry, 4
pKB
NK1
pKB
NK2
Dose ratio
7.71ꢁ0.14 7.44ꢁ0.22
7.90ꢁ0.07 8.15ꢁ0.23
23
12
Methine
Aryl
sulfoxide
NK2 NK1
CF3
7.84ꢁ0.07 6.45ꢁ0.25 ND
aIn vitro pKB values were determined as described in Table 1 (footnote
a).
ZD6021
4b
4c
S
S
R
S
R
S
8.98ꢁ0.17 8.26ꢁ0.10 72
8.36ꢁ0.21 8.39ꢁ0.06 51
144
100
bIn vivo efficacy was determined in the anesthetized guinea pig and
reported as the shift in the dose ratio for bronchoconstriction induced
by a selective agonist. Animals were dosed, at 10 mmol/kg, 10 min prior
to administration of the agonist, ASMSP for NK1 and BANK for
NK2.
NA
7.10ꢁ0.25 ND ND
aSee legend in Table 2.
cND, not determined.
Table 5. Broader profiling of ZD6021 and comparison to
MDL105212
To further elucidate what the effect of the cyano-sub-
stituent was on in vitro activity, the cyano group was
moved to several other positions on the naphthyl ring
system (Table 3). The results indicated that the best
mixed activity was retained with the original 3-position.
Compound KI (nM)
in hNKxa
ZD6021
MDL105212
NK1
NK2
NK3
0.1
0.6
74
0.3
1.0
200
b
pKB
To explore the impact of absolute stereochemistry on
activity, isomers of ZD6021 at the aryl methine and the
sulfoxide were prepared (Table 4). Previously, for both
SR48968 and ZD7944, potent NK2 antagonist activity
had been linked with (S)-stereochemistry at the aryl
methine. For ZD6021 there was an even greater effect
on NK1, than on NK2 antagonism, as the methine ste-
reochemistry was changed from S to R.
RPA NK1
RPA NK2
HuPA NK1
HuBr NK2
9.0
8.3
8.9
7.5
7.4
6.7
6.9
6.4
PO Dose ratioc
NK1
NK2
23
18
5
1
aBinding affinity (KI) was measured against the human receptor cloned
and expressed in MEL cells.
bFunctional activity was determined on: rabbit pulmonary artery
(RPA) using as selective NK1 and NK2 agonists ASMSP and BANK,
respectively; and on: human pulmonary artery (HuPA) for NK1
antagonism and human bronchus (HuBr) for NK2 antagonism, using
ASMSP and NKA as the respective agonists.
Table 3. Effect of moving naphthyl cyano substituent in 3 (X¼CN)a
Compd
Position of
cyano group
pKB NK1
pKB NK2
cIn vivo efficacy was determined in the anesthetized guinea pig and
reported as the shift in the dose ratio for bronchoconstriction induced
by a selective agonist. Animals were dosed orally, at 30 mmol/kg, 120
min prior to administration of the agonist, ASMSP for NK1 and
BANK for NK2.
3r
3s
3t
7
6
4
8.13ꢁ0.27
8.55ꢁ0.14
6.31ꢁ0.09
8.51ꢁ0.22
7.62ꢁ0.24
7.34ꢁ0.42
aSee legend in Table 2.

2772
P. R. Bernstein et al. / Bioorg. Med. Chem. Lett. 11 (2001) 2769–2773
Table 6. Pharmacokinetics (PK) of ZD6021 in the rat and dog
PK Parameter
Rat
Dog
IV dose (umol/kg)
Formulation
10
20% HPBCDa
1
20% HPBCD
Scheme 2. Attempted use of Sandmeyer type chemistry.
AUC-IV(0-i) (ng h/mL)
CLp (mL/min/kg)
Vdss (L/kg)
3200
32.6
3.7
1100
9.5
2.0
3.8
t1/2 (h)
1.8
Oral dose (umol/kg)
Formulation
Cmax (ng/mL)
Tmax (h)
AUC-PO(0-i) (ng h/mL)
Bioavailability
t1/2 (h)
100
Aq suspension
10
75% PEG400/saline
approach to the 3-nitro naphthoic acid.17 Most of the
remaining 3-substituted naphthoic acids needed for the
compounds in Table 2, were prepared from this inter-
mediate by standard techniques. For example, the
cyano analogue was prepared, as illustrated, by reac-
tion with CuCN. Literature approaches were used for
the naphthoic acids needed for 3g18 and 3p.19 The 4,
5, and 6 cyano-substituted naphthoic acids needed for
compounds 3r, 3s, and 3t were prepared as illu-
strated.
421
2
3010
9%
NDb
364
2
2000
18%
2.3
aHPBCD, hydroxypropyl-beta-cyclodextrin.
bND, not determined, data inadequate for characterization of t1/2
.
assayed in a broad array of tests, performed at Pan-
Labs1. Assays with IC50 values <1 mM were (IC50 in
mM): l-type benzothiazepine Ca2+ channel (0.58),
muscarinic M4 (0.34), muscarinic M5 (0.65), and Sigma,
s non-selective (0.75). The pharmacokinetic profile of
ZD6021 was determined in rat and dog. These studies
(Table 6) showed bioavailability of 9 and 18%, respec-
tively.
ZD6021 is a potent, high-affinity and selective dual
NK1/NK2 receptor antagonist. It shows excellent func-
tional activity in both animal and human tissue and
shows good efficacy in vivo in animal models, against
the effects of neurokinin agonists. It was selected for
detailed preclinical evaluation and for preliminary tox-
icological evaluation. It may play an important part in
defining the role of neurokinin antagonists in patho-
physiologic conditions.
Several synthetic approaches to SR48968 and related
compounds have been previously reported.9,10 The key
new synthetic need for the efforts described here was
access to selectively substituted naphthoic acids. Ori-
ginally we tried a Sandmeyer type approach, from 3-
nitronaphthoic acid. However, decomposition of the
intermediate naphthyl diazonium salts (Scheme 2) did
not work well. For example, with CuBr2 the pre-
dominant product was 3,4-dibromonaphthoic acid
(X=Br) and not the desired 3-bromonaphthoic acid.
Acknowledgements
We thank J. Hulsizer and W. Moss and their teams in
the AstraZeneca Large Scale and Process Research labs,
respectively, for intermediates. We also thank Dr. B.
Undem, Johns Hopkins University for determining
functional activity on human tissues.
As an alternate key intermediate to the 3-substituted
analogues (Scheme 3), we chose to utilize methyl 3-
bromonaphth-1-oate. This compound was prepared
from phthalic anhydride by a variation of a known
Scheme 3. Reagents and conditions: (a) NaOH, H2O; (b) HgO (yellow), glacial HOAc, reflux 4 days; (c) 5 M aq HCl, reflux 4 h; (d) recrystallize
from glacial HOAc (2ꢄ); (e) (COCl)2, cat DMF, CH2Cl2; (f) MeOH; (g) CuCN, DMF, pyridine, 180 ꢅC; (h) LiOH, THF, MeOH, H2O; (i) Tf2O,
Et3N, CH2Cl2; (j) Zn(CN)2, (Ph3P)4Pd, DMF; (k) Br2, HOAc; (l) (i) nBuLi, (ii) CO2, (iii) HCl.

P. R. Bernstein et al. / Bioorg. Med. Chem. Lett. 11 (2001) 2769–2773
2773
References and Notes
10. Shah, S. K.; Hale, J. J.; Qi, H.; Miller, D. J.; Dorn, C. P.;
Mills, S. G.; Sadowski, S. J.; Cascieri, M. A.; Metzger, J. M.;
Eiermann, G. J. Abstracts of Papers, 212th National Meeting
of the American Chemical Society, Orlando, FL, Aug 25–29,
1996; MEDI 136.
1. Swain, C. N.; Rupniak, M. J. In Annual Reports in Medic-
inal Chemistry; Doherty, A., Ed.; Academic: New York, 1999;
Vol. 34; pp 51–60.
2. Gao, Z.; Peet, N. P. Curr. Med. Chem. 1999, 6, 375.
3. Barnes, P. J.; Baraniuk, J. B.; Belvisi, M. G. Am. Rev.
Respir. Dis. 1991, 144, 1187.
4. Advenier, C.; Joos, G.; Molinard, M.; Lagente, V.; Pau-
wels, R. Clin. Exp. Allergy 1999, 29, 579.
11. Ting, P. C.; Anthes, J. F. L.; Shih, N.-Y.; Piwinski, J. J.
Bioorg. Med. Chem. Lett. 2000, 10, 2333.
12. For an excellent review on ‘Dual neurokinin NK1/NK2
receptor antagonists’ see Gerspacher, M.; von Sprecher, A.
Drugs Future, 1999, 24, 883
5. Turner, C. R.; Andresen, C. J.; Pattersen, D. K.; Keir, R.
F.; Obach, S.; Lee, P.; Watson, J. W. Am. J. Resp. Crit. Care
1996, 153 (abstr. A160).
6. Emonds-Alt, X.; Vilain, P.; Goulaouic, P.; Proletto, V.;
Van Broeck, D.; Advenier, C.; Naliine, E.; Neliat, G.; Le Fur,
G.; Breliere, J. C. Life Sci. 1992, 50, 101
7. Shenvi, A. B.; Aharony, D.; Brown, F. J.; Buckner, C. K.;
Campbell, J. B.; Dedinas, R. F.; Gero, T. W.; Green, R. C.;
Jacobs, R. T.; Kusner, E. J.; Miller, S. C.; Ohnmacht, C. J.;
Palmer, W. E.; Smith, R. W.; Steelman, G. B.; Ulatowski, T.
G.; Veale C. A.; Walsh, S. A. Abstracts of Papers, Part 1,
214th National Meeting of the American Chemical Society,
Las Vegas, NV, Sept 7–11, 1997; MEDI 264
13. Weak dual NK1/NK2 activity in unsubstituted naphtha-
mides has been reported by Shah et al.10
14. Hale, J.; Mills, S. G.; MacCoss, M.; Finke, P. E.; Cascieri,
M. A.; Sadowski, S.; Ber, E.; Chicchi, G.; Kurtz, M.; Metzger,
J.; Eiermann, G.; Tsou, N.; Tattersall, F. D.; Rupniak,
N. M. J.; Williams, A. R.; Rycroft, W.; Hargreaves, R.;
MacIntyre, D. E. J. Med. Chem. 1998, 41, 4607.
15. Buckner, C. K.; Liberati, N.; Dea, D.; Lengel, D.; Stin-
son-Fisher, C.; Campbell, J.; Miller, S.; Shenvi, A.; Krell,
R. D. J. Pharmacol. Exp. Ther. 1993, 267, 1168.
16. Aharony, D.; Buckner, C. K.; Ellis, J. L.; Ghanekar, S. V.;
Graham, A.; Kays, J. S.; Little, J.; Meeker, S.; Miller, S. C.;
Undem, B. J. J. Pharmacol. Exp. Ther. 1995, 274, 1216.
17. Kice, J. L.; Lotey, H. J. Org. Chem. 1989, 54, 3596.
18. Bell, K. H.; McCaffery, L. F. Aust. J. Chem. 1993, 46,
731.
8. The piperidines used in the initial exploration are illustrated
by compounds 2a, 2b, and 2d.
9. Burkholder, T. P.; Kudlacz, E. M.; Le, T.-B.; Knippenberg,
R. W.; Shatzer, S. A.; Maynard, G. D.; Webster, M. E.; Hor-
gan, S. W. Bioorg. Med. Chem. Lett. 1996, 6, 951.
19. Adcock, W.; Dewar, M. J. S. J. Am. Chem. Soc. 1967, 89,
386.