The discovery of potent, selective, and orally bioavailable hNK1 antagonists derived from pyrrolidine

Article abstract of DOI:10.1016/j.bmcl.2007.06.085

SAR studies on amides, ureas, and vinylogous amides derived from pyrrolidine led to the discovery of several potent hNK₁ antagonists. One particular vinylogous amide (45b) had excellent potency, selectivity, pharmacokinetic profile, and functio

Full text of DOI:10.1016/j.bmcl.2007.06.085

Bioorganic & Medicinal Chemistry Letters 17 (2007) 5191–5198
The discovery of potent, selective, and orally bioavailable
hNK1 antagonists derived from pyrrolidine
Peter Lin,a,* Lehua Chang,a Robert J. DeVita,a Jonathan R. Young,a Ronsar Eid,a
Xinchun Tong,a Song Zheng,a Richard G. Ball,a Nancy N. Tsou,a Gary G. Chicchi,b
Marc M. Kurtz,b Kwei-Lan C. Tsao,b Alan Wheeldon,c Emma J. Carlson,c WaiSi Eng,d
H. Donald Burns,d Richard J. Hargreavesd and Sander G. Millsa
aDepartment of Medicinal Chemistry, Merck & Co. Inc., PO Box 2000, Rahway, NJ 07065-0900, USA
bDepartment of Metabolic Disorders, Merck & Co. Inc., PO Box 2000, Rahway, NJ 07065-0900, USA
cMerck Sharp and Dohme—Neuroscience Research Centre, Terlings Park, Eastwick Road, Harlow, Essex CM20 2QR, UK
dImaging Research, Merck & Co. Inc., West Point, PA 19486, USA
Received 4 May 2007; revised 25 June 2007; accepted 28 June 2007
Available online 4 July 2007
Abstract—SAR studies on amides, ureas, and vinylogous amides derived from pyrrolidine led to the discovery of several potent
hNK1 antagonists. One particular vinylogous amide (45b) had excellent potency, selectivity, pharmacokinetic profile, and functional
activity in vivo. An in vivo rhesus macaque brain receptor occupancy PET study for compound 45b revealed an estimated
Occ90 ꢀ 300 ng/ml.
Ó 2007 Elsevier Ltd. All rights reserved.
Interest in the agonist neuropeptide substance P (SP)
began in 19311 when the 11-amino acid peptide was first
isolated from extracts of horse intestines and brains, and
its pharmacological action demonstrated by stimulating
atropine resistant contraction of rabbit ileum. It was
several years more before other functions of SP in mam-
malian biochemistry became known and the existence of
a G protein-coupled receptor, the neurokinin-1 (NK1)
receptor, was established.2 SP is purportedly involved
in a number of biochemical pathways which include
the transmission of the sensation of pain and an over-
expression of SP has been implicated in several medical
conditions: emesis, CNS disorders, and urinary inconti-
nence.3 Consequently, this sparked interest in finding
neurokinin antagonists for medicinal use. The first dis-
closure of a selective small molecule NK1 antagonist oc-
curred in 19914 and since then potent clinical candidates
based on several different scaffolds have been discov-
ered.5 There is currently one drug for use in man which
works through blockade of the human NK1 (hNK1)
receptor avaiÒlable in the medicinal formulary: aprepit-
ant6 [Emend (1)]. EmendÒ was approved for use in
conjunction with cancer treatment drugs to alleviate
chemotherapy-induced nausea and vomiting7 (CINV)
in 2003 and for post-operative nausea and vomiting
(PONV) in 2006.
We wish to describe our initial synthetic work and new
structure–activity relationships (SAR) on a scaffold
based on a simple pyrrolidine ether 3. Some work in a
cyclopentane series, such as 2, related to the pyrrolidine
described below was disclosed recently.8 We have
extended that previous work to investigate the effect of
incorporation of the nitrogen into the five-membered
cyclic system, but still b to the phenyl, and which offers
a different orientation of substituents (Fig. 1).
The pyrrolidine compounds initially used in our studies9
were synthesized as shown in Scheme 1. An epoxidation
of Boc-protected D3,4 pyrroline 4 with mCPBA gave the
expected meso-epoxide 5 and then a copper(I) iodide
assisted ring opening of epoxide 5 with phenylmagnesium
bromide afforded a racemic mixture of the pyrrolidine
alcohols 6a. The racemic mixture was resolved into alco-
hols 7a and 13a by HPLC on a Chiralpak-AD column.
Both of these alcohol enantiomers, separately, were
Keywords: NK1; Neurokinin NK1 antagonist; Tachykinin NK1 antag-
onist; Substance P antagonist.
*
Corresponding author. Tel.: +1 732 594 4646; fax: +1 732 594
5790; e-mail: peter_lin@merck.com
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.06.085

5192
P. Lin et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5191–5198
Figure 1. Evolution of SAR studies from aprepitant and the biological data for compounds 1 and 2.
Scheme 1. Reagents and conditions: (a) mCPBA, CH2Cl2, À78 °C to rt; (b) PhMgBr or 4-F-C6H4MgBr, CuI, THF, 0 °C; (c) imidate 19, heptane-
(CH2Cl)2, À30 °C to rt; (d) TFA, anisole, CH2Cl2, rt; (e) AcCl, NEt3, CH2Cl2, rt.
taken through the following reaction sequence described
here for alcohol 7a. Reaction of the homochiral alcohol
7a with (S)-imidate10 19 gave a 1:3 mixture of epimeric
ethers (8a) and subsequent treatment with TFA
removed the Boc-protecting group to reveal the amines
9a. The 1:3 mixture of pyrrolidines 9a was reacted with
acetyl chloride and the two epimeric amides 10a thus
obtained were resolved into the homochiral amides
11a and 12a by HPLC on a Chiralcel-OD column:
amide 12a having the (R)-ether being the major isomer
as expected.10 Similarly, the amides 17a and 18a in
homochiral form were obtained from the pyrrolidine

P. Lin et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5191–5198
5193
alcohol 13a. The corresponding 4-fluorophenyl analogs
17b and 18b were synthesized using 4-fluorophenylmag-
nesium bromide in a similar reaction and separation
sequence as shown in Scheme 1.
deprotonation were obtained using LDA as the base,
while products from thermodynamic deprotonation
were obtained with LiHMDS.12
The hNK1 binding affinities for the pyrrolidine analogs
were determined by measuring the ability of the com-
pounds to displace [125I]-SP from the hNK1 receptor sta-
bly expressed in CHO cells.13,14 Using the hNK1 binding
assay results, the ranking in terms of potency of the iso-
mers 11a, 12a, 17a, and 18a was established, the most
potent being 18a (IC50 = 0.09 nM) then, in order of
activity 12a (IC50 = 32 nM), 17a (IC50 = 80 nM), and
11a (18% I at 100 nM). From these binding data and
work done in the cyclopentane series15 isomer 18a was
tentatively assigned as being derived from the (3R,4S)-
pyrrolidine, the carbon bearing the benzylic methyl
was (R) based on the mechanism10 of the etherification
reaction while the less potent epimer, 17a, was the
(3R,4S)-pyrrolidine with the (S)-chirality adjacent to
the benzylic ether oxygen. Consequently, 11a and 12a
were (3S,4R)-pyrrolidine derivatives. The major product
from the etherification, 12a, had the (R)-configuration
next to the oxygen, while the minor ether 11a had an
(S)-center. Later the tentative stereochemistry assign-
ment of 18a was confirmed through determining the rel-
ative stereochemistry of the three chiral centers in urea
28b (IC50 = 0.11 nM) (Fig. 2) in a single crystal X-ray
analysis.16 Since the benzylic carbon of the ether may
be reasonably assigned as (R) from the aforementioned
Since the major isomer, 18a, derived from 14a was the
most active (see below), in later work (Scheme 2) the
major isomers in the epimeric mixture of ethers (14a
and 14b) were separated by HPLC using a Chiralpak-
AD column and the Boc-protecting group was then re-
moved using TFA. These two resulting pyrrolidines
20a and 20b served as starting materials for the prepara-
tion of various simple derivatives on the pyrrolidine
nitrogen.
Reaction of the pyrrolidines 20a and 20b with acid chlo-
rides or acids and EDC led to the formation of pyrroli-
dine amides 21a and 21b, and with isocyanates or
carbamoyl chlorides, or phosgene and amine, to afford
ureas 22a and 22b. Some cyclic vinylogous amides (23
and 24) were synthesized (Scheme 3) by reacting the
pyrrolidines 20a and 20b with cyclic 1,3-diketones and
a catalytic amount of acid. The vinylogous amides with
heteroatoms incorporated within the ring as in 25 were
made similarly from tetronic acid or from N-methyltet-
ramic acid. Methylation or hydroxylation of vinylogous
amide 23 was achieved by treatment with base and then
reaction with methyl iodide or with Vedejs’s reagent11
MoOPH (Scheme 3). The products derived from kinetic
Scheme 2. Reagents and conditions: (a) RCOCl, NEt3, CH2Cl2, 0 °C to rt; (b) RCO2H, EDC, HOBt, NEt3, CH2Cl2, rt; (c) RNCO, NEt3, CH2Cl2,
0 °C to rt; (d) R1R2NCOCl, NEt3, CH2Cl2, rt; (e) COCl2 in toluene 0 °C, then R1R2NH, NEt3, CH2Cl2 rt.
Scheme 3. Reagents and conditions: (a) cyclopentane-1,3-dione or 2-methylcyclopentane-1,3-dione, TsOH, toluene, 80 °C; (b) tetronic or N-
methyltetramic acid, TsOH, toluene, 80 °C; (c) LiHMDS (0.95 equiv), THF, HMPA, À78 °C then MeI (R2 = Me) or MoOPH (R2 = OH); (d) LDA
(1.1 equiv), THF À78 °C then MeI (R2 = Me) or MoOPH (R2 = OH).

5194
P. Lin et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5191–5198
F6
C4
The corresponding 4-fluoro analogs 17b (IC50 = 40 nM)
F2
and 18b (IC50 = 0.07 nM) had similar IC50 values to 17a
and 18a (see above) of the phenyl series in the hNK1
binding assay. In the gerbil foot tapping (GFT) assay
that has been previously described as an in vivo pharma-
codynamic model for CNS penetrant NK1 com-
pounds17,18 both amides 18a and 18b were efficacious.
The 4-fluoro analog 18b inhibited GFT 50% after 4 h
with a 2 mpk iv dose demonstrating that the amide
was brain penetrant after a short time interval: the phe-
nyl analog 18a had 98% inhibition at 24 h with a 6 mpk
iv dose. The phenyl analog 18a was titrated in the gerbil
and it had an ID50 = 0.6 mpk (iv) at t = 0 h and
ID50 = 3.2 mpk (iv) at 24 h. Encouraged by the GFT
data, the rat pharmacokinetic parameters of amide 18b
were determined and indicated good bioavailability
(F = 65%) and half-life (t1/2 = 5 h).
F4
C2
C22
F5
C3
F1
F3
F7
C1
C6
C16
C17
C5
C15
C8
C14
C7
C18
C13
O1
C9
C12
C11
C10
N1
C19
O2
N2
Having established potency in hNK1 receptor binding,
GFT efficacy, and bioavailability with amide 18b, other
simple amides and groups that had previously shown
in vivo activity in the cyclopentane series were prepared
and tested. All the amides (Table 1) had sub-nanomolar
hNK1 binding and two of the amides 31b and 32a inhib-
ited GFT ꢀ 50% at 24 h with a 6 mpk iv dose. Three of
the heterocyclic derived amides 33b, 34b, and 35b had
similar or better potencies at a lower dose with lactam
34b being the most effective with 100% inhibition at a
C20
Figure 2. Perspective view (ORTEP) of 28b showing the crystallo-
graphic numbering scheme; non-hydrogen atoms are drawn as 20%
probability envelopes.
reason,10 the absolute stereochemistry of the two
centers on the pyrrolidine ring can then be assigned as
(3R,4S).
Table 1. Synthesis and hNK1 binding and gerbil foot tapping assay data for pyrrolidine amide compounds
CF3
Me
CF3
O
R N
Ar
Compounda
R
Preparation methodb
GFTd
c
hNK1 IC50 (nM)
% I at 24 h
iv dose mpk
29be
18af
18b
30b
31b
32a
CHO
0.24
0.09
0.07
0.09
0.15
0.18
COMe
COMe
COEt
COiPr
COtBu
b
b
b
b
b
98
26
0
6
6
6
6
6
43
45
O
O
33b
c
c
0.07
0.12
50
3
3
N
Ac
34bg
100
N
Me
O
N
N
O
35b
36b
c
c
0.12
0.34
45
13
3
3
Me
O
H
N
O
O
a a refers to Ar = phenyl and b refers to Ar = 4-F-phenyl series.
b See Scheme 2 for preparation methods.
c See Ref. 13 and note 14.
d See Refs. 6, 17, and 18 for details.
e Compound 29b was prepared by heating pyrrolidine 20b in ethyl formate.
f ID50 = 0.6 mpk at 0 h; ID50 = 3.2 mpk at 24 h.
g ID50 = 0.6 mpk at 0 h; ID50 = 1.2 mpk at 24 h.

P. Lin et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5191–5198
5195
3 mpk iv dose and an ID50 = 1.2 mpk at 24 h and an
ID50 = 0.6 mpk at 0 h (iv). The amide compounds in
general still had weaker efficacy in the GFT assay at
24 h compared to previous structures 1 and 2 (see
Fig. 1).
bond (46b) or at other positions around the five-mem-
bered ring (47b–49b) afforded compounds with potent
hNK1 binding, but all of these were less potent in the
GFT assay than 45b. The two epimeric hydroxylated
analogs 50b and 51b were both potent and more active
in the GFT assay than the methylated analogs, but were
still less potent than vinylogous amide 45b. Vinylogous
amides 52b and 53b derived from tetronic and N-meth-
yltetramic acid and the cyclohexenyl examples 54a and
55a were also sub-nanomolar in the hNK1 binding assay
but lacked significant efficacy in the GFT assay even at a
3 mpk iv dose at 24 h. The most efficacious vinylogous
amide in GFT remained the parent compound 45b.
A series of ureas (28b, 37b–41b) and a set of heterocyclic
derived ureas 42b–44b listed in Table 2 were assayed
and, like the amides, were all potent, sub-nanomolar
compounds in the hNK1 binding assay. Some of the ur-
eas had similar efficacy in the GFT assay to the amides
18a and 34b at 24 h (see Table 1 footnotes). Three of the
most potent ureas 28b, 39b, and 40b were titrated in the
gerbil, giving ID50 values in the range 0.8–3 mpk at 24 h.
The two ureas 39b and 40b inhibited GFT ꢀ 60% at a
dose of 1 mpk (iv) and 28b had an ID50 = 1.0 mpk at
t = 0 h. From these data it was inferred that the three ur-
eas 28b, 39b, and 40b were rapidly brain penetrant and
the compounds had reasonably long half-lives in gerbil.
Urea 28b also showed good bioavailability (F = 70%)
with a half-life (t1/2 = 1.25 h) in a rat PK study.
Further evaluation of the properties of vinylogous
amide 45b (Fig. 3) showed that this antagonist was selec-
tive for the hNK1 receptor over the other known hNK2
and hNK3 sub-types.20 Promising PK parameters were
obtained in rat, dog, and rhesus measurements: 45b
was highly bioavailable and had reasonable half-lives
in all three animal species consistent with once-daily
dosing in man. Also, vinylogous amide 45b performed
well in two animal models of NK1 antagonism after oral
dosing. When compound 45b was dosed orally in the
GFT assay, it was efficacious at 1 h (ID50 = 0.9 mpk)
and 4 h (ID50 = 0.25 mpk). In a guinea pig separation
induced vocalization experiment21 an ID50 = 0.17 mg/
kg at 4 h (po) was obtained (Fig. 4). Vinylogous amide
45b was also evaluated in vivo in a rhesus positron emis-
sion tomography (PET) study22 to determine brain NK1
receptor occupancy after a 4-h iv infusion. This study
afforded an estimated 50% occupancy of NK1 receptors
(Occ50) at a plasma steady state concentration of 33 ng/
In attempting to broaden the SAR of polar substituents
on the pyrrolidine nitrogen, two vinylogous amides 45a
and 45b19 were prepared. Both the phenyl and 4-fluor-
ophenyl analogs were sub-nanomolar in hNK1 binding.
4-Fluorophenyl analog 45b (Table 3) distinguished itself
as being superior to the phenyl example 45a (and indeed
more active than all the other analogs described above)
in the GFT assay at both the t = 0 and 24 h time points.
The SAR around vinylogous amide 45b was explored by
strategically introducing methyl and hydroxyl groups
onto the cyclopentenyl ring. Methylation on the double
Table 2. Synthesis and hNK1 binding and gerbil foot tapping assay data for pyrrolidine urea compounds
CF3
Me
CF3
O
R N
Ar
Compounda
R
Preparation methodb
GFTd % I at iv dose (ID50
24 h
)
c
hNK1 IC50 (nM)
0 h
28b
37b
CONHMe
CONMe2
O
c
0.11
0.38
100% at 2 mpk (1.0 mpk)
nd
77% at 6 mpk (3.0 mpk)
nd
d
38a
d
0.16
40% at 1 mpk
N
O
39b
40b
41b
CONHEt
CONHiPr
CONHtBu
c
c
c
0.10
0.06
0.14
61% at 1 mpk
64% at 1 mpk
28% at 1 mpk
93% 6 mpk (1.0 mpk)
86% at 6 mpk (0.8 mpk)
6% at 6 mpk
O
N
N
42b
e
0.30
57% at 3 mpk
N
Ac
O
N
N
CF3
43b
44b
e
e
0.47
0.58
66% at 3 mpk
24% at 3 mpk
N
O
N
NH
N
nd, not determined at t = 0 and 24 h.
a a refers to Ar = phenyl and b refers to Ar = 4-F-phenyl series.
b See Scheme 2 for preparation methods.
c See Ref. 13 and note 14.
d See Refs. 6, 17, and 18 for details.

5196
P. Lin et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5191–5198
Table 3. Synthesis and hNK1 binding and gerbil foot tapping assay data for pyrrolidine vinylogous amide compounds
CF3
Me
CF3
O
R N
Ar
d
hNK1 IC50 (nM)
Compounda
R
Preparation methodb,c
GFTe % I at iv dose (ID50
24 h
)
0 h
45a
45b
a
a
0.07
0.11
100% at 3 mpk (0.9 mpk)
100% at 1 mpk (0.1 mpk)
100% at 3 mpk (1.1 mpk)
100% at 3 mpk (0.045 mpk)
O
O
Me
46b
47b
a
c
0.13
0.16
67% at 3 mpk
85% at 3 mpk
O
O
Me
diastereomer A
O
48b
c
0.09
2% at 3 mpk
Me
diastereomer B
O
49b
50b
d
c
0.11
0.22
79% at 3 mpk
Me
two diastereomer mix
O
100% at 3 mpk (0.5 mpk)
100% at 3 mpk (1.1 mpk)
100% at 3 mpk (0.7 mpk)
HO
diastereomer A
O
51b
52b
c
0.30
0.16
95% at 3 mpk (0.5 mpk)
67% at 3 mpk
HO
diastereomer B
O
O
b
O
53b
54a
55a
b
a
a
0.12
0.18
0.13
8% at 3 mpk
12% at 3 mpk
5% at 3 mpk
N
Me
O
Me
Me
O
Me
Me
a a refers to Ar = phenyl and b refers to Ar = 4-F-phenyl series.
b See Scheme 3 for preparation methods.
c Compounds 54a and 55a were obtained from the appropriate cyclohexa-1,3-diones and pyrrolidine 20a.
d See Ref. 13 and note 14.
e See Refs. 6, 17, and 18 for details.
ml and a 90% occupancy (Occ90) at a concentration of
300 ng/ml (Fig. 5). Overall the vinylogous amide 45b
had excellent potency, selectivity, pharmacokinetic pro-
file, and functional activity in vivo.
amides showed potent in vitro hNK1 binding, but most
were modestly or weakly active in vivo in the GFT func-
tional assay at 24 h—the exceptions were amides 18a
and 34b. A series of ureas was then prepared and evalu-
ated and these compounds were also sub-nanomolar in
hNK1 binding. The best three ureas 28b, 39b, and 40b
were comparable or were more potent in the GFT assay
at 24 h (iv) and slightly less potent at t = 0 h than the
amides 18a and 34b. These then led to preparation of
the vinylogous amide derivatives. These vinylogous
amides in both the phenyl and 4-fluorophenyl series
Amide, urea, and a vinylogous amide derived from a
new pyrrolidine scaffold design were prepared as hNK1
receptor antagonists. The hNK1 binding results from
amides 11a, 12a, 17a, and 18a, and the X-ray structure
of urea 28a allowed the stereochemistry of the active ser-
ies to be identified. The active series of pyrrolidine

P. Lin et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5191–5198
5197
Binding assay results (in vitro)
Acknowledgments
CF3
hNK1 IC50 = 0.11 nM
hNK2 IC50 = 1730nM
hNK3 IC50 > 3000 nM
The authors thank the Comparative Medicine group for
dosing the animals used in pharmacokinetic experi-
ments, the Synthetic Services team for scale-up and large
scale separation of synthetic intermediates, George Doss
for the NMR characterization of vinylogous amide 45b,
and the staff in Imaging Research for PET tracer synthe-
sis and the imaging studies. Help from Paul Finke, Shre-
nik Shah, and Scott Edmondson in preparing this
manuscript is also gratefully acknowledged.
Me
CF3
Gerbil data
O
O
N
0 h 100% at 1 mpk (iv)
0 h titration ID50 = 0.1 mpk (iv)
24 h 100% at 3 mpk(iv)
45b
24 h titration ID50 = 0.045 mpk (iv)
F
1 h titration ID50 = 0.9 mpk (po)
4 h titration ID50 = 0.25 mpk (po)
Pharmacokinetic Parameters
Rat PK Dog PK
Rhesus PK
Clp = 19.6 mL/min/Kg
Vd = 6.2 L/Kg
1/2 = 3.6 h
= 100%
Clp = 1mL/min/Kg
Vd = 1.6 L/Kg
Clp = 4.5 mL/min/Kg
Vd = 2.3 L/Kg
t
t
F
1/2 = 20 h
t1/2 = 5.7 h
References and notes
F
= 100%
F
= 68%
1. von Euler, U. S.; Gaddum, J. H. J. Physiolol. (Lond.)
1931, 72, 74.
Figure 3. Data summary for vinylogous amide 45b.
2. Strand, F. L. In Neuropeptides: Regulators of Physiological
Processes; Bradford Book-MIT Press: Cambridge, USA,
1999; pp 365–381.
100
75
3. (a) Quatara, L.; Maggi, C. A. Neuropeptides 1998, 32, 1;
(b) Rupniak, N. M. J.; Kramer, M. S. Trends Pharmacol.
Sci. 1999, 20, 485; (c) Dando, T. M.; Perry, C. M. Drugs
2004, 64, 777; (d) Green, S. A.; Alon, A.; Ianus, J.;
McNaughton, K. S.; Tozzi, C. A.; Reiss, T. F. J. Urol.
2006, 176, 2535; (e) Pendergrass, K.; Hargreaves, R.;
Petty, K. J.; Carides, A. D.; Evans, J. K.; Horgan, K. J.
Drugs Today 2004, 40, 853.
50
25
0
ID50 = 0.17 mg/kg
Veh
0.01
0.10
1.00
4. Sneider, R. M.; Constantine, J. W.; Lowe, J. A., III;
Longo, K. P.; Lebel, W. S.; Woody, H. A.; Drozda, E. E.;
Desai, M. C.; Vinick, F. J.; Spencer, R. W.; Hess, H.-J.
Science 1991, 251, 437.
Dose (mg/kg p.o.)
Figure 4. Inhibition of separation induced vocalizations in guinea pig
pups for 45b at t = 4 h.
5. Seward, E. M.; Swain, C. J. Exp. Opin. Ther. Patents 1999,
9, 571.
100
80
6. Hale, J. J.; Mills, S. G.; MacCoss, M.; Finke, P. E.;
Cascieri, M. A.; Sadowski, S.; Ber, E.; Chicchi, G. G.;
Kurtz, M.; Metzger, J.; Eiermann, G.; Tsou, N. N.;
Tatersall, F. D.; Rupniak, N. M. J.; Williams, A. R.;
Rycroft, W.; Hargreaves, R.; MacIntyre, D. E. J. Med.
Chem. 1998, 41, 4607.
60
7. Gale, J. D.; O’Neill, B. T.; Humphrey, J. M. Exp. Opin.
Ther. Patents 2001, 11, 1837.
40
8. (a) Finke, P. E.; Meurer, L. C.; Levorse, D. A.; Mills, S. G.;
MacCoss, M.; Sadowski, S.; Cascieri, M. A.; Tsao, K.-L.;
Chicchi, G. G.; Metzger, J. M.; MacIntyre, D. E. Bioorg.
Med. Chem. Lett. 2006, 16, 4497; (b) Meurer, L. C.; Finke,
P. E.; Owens, K. A.; Tsou, N. N.; Ball, R. G.; Mills, S. G.;
MacCoss, M.; Sadowski, S.; Cascieri, M. A.; Tsao, K.-L.;
Chicchi, G. G.; Egger, L. A.; Luell, S.; Metzger, J. M.;
MacIntyre, D. E.; Rupniak, N. M. J.; Williams, A. R.;
Hargreaves, R. J. Bioorg. Med. Chem. Lett. 2006, 16, 4504.
9. All compounds gave satisfactory proton NMR and mass
spectroscopic data. Experimental details for the prepara-
tion of the compounds described in this Letter may be
found in DeVita, R. J.; Mills, S. G.; Young, J. R.; Lin, P.
World Patent 2005/032464, 2005.
10. Kuethe, J. T.; Marcoux, J.-F.; Wong, A.; Wu, J.; Hillier,
M. C.; Dormer, P. G.; Davies, I. W.; Hughes, D. L.
J. Org. Chem. 2006, 71, 7378.
11. Vedejs, E.; Engler, D. A.; Telschow, J. E. J. Org. Chem.
1978, 43, 188.
12. (a) Chen, Y. L.; Mariano, P. S.; Little, G. M.; O’Brien, D.;
Huesmann, P. L. J. Org. Chem. 1981, 46, 4643; (b)
Lakhvich, F. A.; Lis, L. G.; Pap, A. A. Zhur. Org. Khim.
1988, 24, 1388.
13. Cascieri, M. A.; Ber, E.; Fong, T. M.; Sadowski, S.;
Bansal, A.; Swain, C. J.; Seward, E. M.; Frances, B.;
Burns, D.; Strader, C. D. Mol. Pharmacol. 1992, 42, 458.
20
Fit to 3 -parameter Hill equation
nH ~ 0.9
0
1
10
100
1000
10000
Plasma Steady State Concentration (ng/mL)
Figure 5. NK1 receptor occupancy in rhesus brain determined by PET
for compound 45b.
were sub-nanomolar in binding but the parent 45b was
the most potent of all the analogs prepared in the gerbil
assay at t = 0 and 24 h (iv). This particular vinylogous
amide 45b was evaluated in rat, dog, and rhesus PK,
and was highly bioavailable with good half-lives in three
preclinical species. The efficacy of 45b in oral drug form
was evaluated in two efficacy models: the gerbil foot tap-
ping and guinea pig separation induced vocalization
assays. The NK1 receptor occupancy in the rhesus
macaque brain was also determined in a PET study indi-
cating plasma steady state concentrations of 45b of
Occ50 ꢀ 33 mg/ml and Occ90 ꢀ 300 mg/ml. From this
work a series of pyrrolidine ether hNK1 antagonists with
a unique vinylogous amide pharmacophore was identi-
fied that afforded excellent in vitro and in vivo activity.

5198
P. Lin et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5191–5198
14. The assay-to-assay replicate variability in the human NK1
binding assay was typically less than 2-fold from the
average of 2–6 independent determinations.
19. Vinylogous amide 45b appears to exist as two rotamers
in solution, in CD3CN, resulting from restricted rota-
tion around the C–N bond. A variable temperature
NMR experiment was conducted on a Varian Unity
Inova 500 MHz spectrometer to estimate the energy
barrier to rotation. The two vinylic proton signals of
the rotamers coalesced at 60 °C and the exchange rate
constant (k) was 56 sÀ1 at the coalescence temperature.
The energy barrier (DGà) is estimated to be 16.9 kcal/
mol.
15. The potent series of cyclopentane analogs described in
8a
Ref.
has the stereochemistry as shown in Figure 1,
compound 2; the pyrrolidine series that corresponds with
this cyclopentane series is the (3R,4S)-pyrrolidine with the
(R)-stereochemistry next to the benzylic ether oxygen.
16. The structure is shown with the crystallographic atom-
numbering scheme. Crystallographic data (excluding
structure factors) for the urea structure have been depos-
ited with the Cambridge Crystallographic Data Centre as
supplementary publication numbers CCDC [645923]. A
copy of the data may be obtained free of charge from
CCDC, 12 Union Road, Cambridge CB2 1EZ, UK. [fax:
+44 (0)1223 336033 or e-mail: deposit@ccdc.cam.ac.uk.
17. Rupniak, N. M. J.; Williams, A. R. Eur. J. Pharmacol.
1994, 265, 179.
20. Results for the hNK2 and hNK3 binding assays were
obtained by MDS Pharma Services-Discovery.
21. Tattersall, F. D.; Rycroft, W.; Francis, B.; Pearce, D.;
Merchant, K.; MacLeod, A. M.; Ladduwahetty, T.;
Keown, L.; Swain, C.; Baker, R.; Cascieri, M.; Ber, E.;
Metzger, J.; MacIntyre, E.; Hill, R. G.; Hargreaves, R. J.
Neuropharmacology 1996, 35, 1121.
22. Unpublished results obtained by Merck Research Labo-
ratories, Imaging Research WP44C-2, West Point, PA
19486. The results were obtained in rhesus macaques using
the PET tracer [18F]-SPARQ as described by Burns, H.
D.; Hamill, T. G.; Gibson, R.E.U.S. Patent 6,241,964,
2001.
18. Rupniak, N. M. J.; Tattersall, F. D.; Williams, A. R.;
Rycroft, W.; Carlson, E. J.; Cascieri, M. A.; Sadowski, S.;
Ber, E.; Hale, J. J.; Mills, S. G.; MacCoss, M.; Seward, E.;
Huscroft, I.; Owen, S.; Swain, C. J.; Hill, R. G.;
Hargreaves, R. J. Eur. J. Pharmcol. 1997, 326, 201.