Novel lactam NK1 antagonists with anti-emetic activity

Article abstract of DOI:10.1016/j.bmcl.2005.11.111

A series of 4,4-disubstituted cyclohexylamine NK₁ antagonists containing a lactam ring is described. The compounds are brain penetrant and activity is demonstrated in a ferret emesis model.

Full text of DOI:10.1016/j.bmcl.2005.11.111

Bioorganic & Medicinal Chemistry Letters 16 (2006) 1197–1201
Novel lactam NK1 antagonists with anti-emetic activity
b
a
c
Gregory J. Hollingworth,a, Emma J. Carlson, Jose L. Castro, Gary G. Chicchi,
*
´
Natalie Clark,b Laura C. Cooper,a Olivier Dirat,a Jerry Di Salvo,c Jason M. Elliott,a
Ruth Kilburn,c Marc M. Kurtz,c Wayne Rycroft,b F. David Tattersall,b
Kwei-Lan Tsaoc and Christopher J. Swaina
aDepartment of Medicinal Chemistry, Merck Sharp & Dohme Research Laboratories, The Neuroscience Research Centre,
Terlings Park, Eastwick Road, Harlow, Essex CM20 2QR, UK
bDepartment of Pharmacology, Merck Sharp & Dohme Research Laboratories, The Neuroscience Research Centre, Terlings Park,
Eastwick Road, Harlow, Essex CM20 2QR, UK
cDepartment of Biochemistry, Merck Research Laboratories, Rahway, NJ 07065, USA
Received 28 October 2005; revised 23 November 2005; accepted 23 November 2005
Available online 27 December 2005
Abstract—A series of 4,4-disubstituted cyclohexylamine NK1 antagonists containing a lactam ring is described. The compounds are
brain penetrant and activity is demonstrated in a ferret emesis model.
Ó 2006 Elsevier Ltd. All rights reserved.
The utility of neurokinin-1 (NK1) antagonists for the
H
CF3
treatment of chemotherapy-induced emesis has now
been established; Aprepitant (EmendÒ) is currently the
only commercially available drug in this class.1 More-
over, their potential utility for a range of therapeutic
conditions such as depression, inflammatory bowel dis-
ease and asthma has been widely reported.2 The interest
in this receptor, to which the neuropeptide substance P
normally binds, has led to the disclosure of selective
NK1 antagonists by several groups including our
own.2,3 In two recent communications, we described a
novel structural class of cyclohexylamine NK1 antago-
nists exemplified by the lead compound 1.4
N
O
CF3
N
1
(±)
hNK1 IC50 0.34 nM
O
was developed to give optimal in vitro binding to the
NK1 receptor (hNK1 IC50 0.34 nM) and binding selec-
tivity over the IKr cardiac potassium channel (hIKr Ki
0.8 lM); early compounds in the series had shown poor
selectivity over this potassium channel.4b
The cyclohexyl core of 1 is geminally substituted with a
phenyl ring and a 2-[3,5-bis(tri-fluoromethyl)phen-
yl]propionamide, the key elements of the NK1 pharma-
cophore. A pendant amine substituent in the 4-position,
trans to the amide, locks the cyclohexane ring in the pre-
ferred binding conformation with the phenyl ring axially
disposed.4a The receptor is tolerant of a variety of amine
substituents and the 2-oxa-8-azaspiro[5.4]decane group
Compound 1 was also shown to be brain penetrant since
it was active in the gerbil foot-tapping assay.5 Central
infusion of the NK1 agonist GR73632 in gerbils causes
a foot-tapping response in control animals. Intravenous
administration of 1 immediately prior to this infusion
showed an inhibition of this centrally mediated response
(ID50 1.4 mg/kg iv). Moreover, the compound had sim-
ilar activity when dosed with a 2 h or 24 h pretreatment
time (ID50 1.1 mg/kg iv and 1.6 mg/kg iv, respectively)
indicating good central duration of action. The com-
pound was also active after oral dosing at the 24 h
time-point (ID50 2.2 mg/kg po).
Keywords: NK1; Neurokinin; Tachykinin; Substance P; Emesis; Anti-
emetic; Lactam.
*
Corresponding
greg_hollingworth@merck.com
author.
Tel.:
+44
1279
440468; e-mail:
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.11.111

1198
G. J. Hollingworth et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1197–1201
While this is a good profile, the activity in the gerbil as-
say at early time-points is weaker than expected by com-
parison with other series of NK1 antagonists. It was
reasoned that reducing the high polarity of the amide
might confer better brain penetration on these com-
pounds and improve activity in this key in vivo assay.
phenylcyclohexan-1-one.4 Ketone protection under
standard conditions led to amides 2a and 2b, respective-
ly. Ozonolysis with reductive work-up then bromination
of the resulting alcohol led to the bromides 3a/3b. Cyc-
lisation of the bromides to the lactams 4a/4b was
achieved using sodium hydride as base in moderate
yield. Ketal deprotection then reductive amination and
removal of the less active cis-isomer by chromatography
furnished the final compounds 5 as the pure trans-iso-
mers.6 The cyclisation reaction of the bromides 3a/3b
using sodium hydride was complicated by the propensity
of the lactam to hydroxylate at the benzylic position,
presumably by reaction of the anion (produced in the
presence of excess base) with trace amounts of oxygen.
This side reaction, particularly problematic on a small
scale, could be suppressed by rigorous degassing with
argon, although a more convenient and reproducible
cyclisation was found to occur using one equivalent of
sodium hexamethyldisilazide as base at room tempera-
ture, producing a 86% yield of 4a from 3a. The oxida-
tion could be used to advantage if allowed to run to
completion, the oxidised product 6 allowing synthesis
of hydroxylated analogues 7.
A series of lactams was therefore proposed which do not
have an amide N–H bond and as a result are less polar.
Whilst the formation of a lactam also limits the confor-
mational degrees of freedom, modelling studies had sug-
gested that these molecules could adopt a suitable
conformation for binding to the NK1 receptor. Ana-
logues containing five- and six-membered lactam rings
were synthesised according to Scheme 1.
Mono-alkylation of the dianion derived from 3,5-bis(tri-
fluoromethyl)phenylacetic acid using either allyl bro-
mide or 4-bromobut-1-ene as electrophile was followed
by amide coupling of the acid product with 4-amino-4-
( )n
H
N
CF3
In the lead series, the a-methyl substituent had been
shown to be important.4b For this reason, the a-meth-
ylated lactam was also targeted. Simple methylation
of 5b on a small scale was again complicated by the
propensity of the benzylic anion to oxidise. Thus, in
addition to the desired product, the corresponding
hydroxylated and methoxylated compounds were pro-
duced (Scheme 2).
CF3
HO2C
i, ii, iii, iv
v, vi
O
CF3
CF3
2a (n=1)
2b (n=2)
O
O
Br
( )n
( )n
H
CF3
N
A more satisfactory synthesis, amenable to analogue
production, was developed to obviate this problem
(Scheme 3). A slurry of sodium hydride in DMF was
added to a mixture of lactam 4a and iodomethane in
DMF. Under these conditions, the sole product in
97% yield was the desired methylated compound which
CF3
N
vii
O
O
CF3
O
O
CF3
O
O
4a (n=1)
4b (n=2)
3a (n=1)
3b (n=2)
( )n
CF3
N
viii, ix
R
O
CF3
CF3
N
N
CF3
i
O
O
NR2
5
CF3
CF3
N
N
7a (R=OH)
8a (R=Me)
9a (R=OMe)
OH
OH
CF3
N
O
CF3
5b
N
O
O
O
viii, ix
O
Scheme 2. Reagents: (i) NaH, MeI, DMF, combined yield 90%.
CF3
O
CF3
NR2
6
7
Scheme 1. Reagents and conditions: (i) 2 equiv n-BuLi,
Br(CH2)nCH@CH2, THF, 79%; (ii) (COCl)2 100%; (iii) 4-amino-4-
phenylcyclohexan-1-one, py, CH2Cl2, 60%; (iv) ethylene glycol, p-TSA,
PhCH3, 98%; (v) O3, then NaBH4, MeOH/CH2Cl2, 83%; (vi) CBr4,
PPh3, 2,6-di-tert-butyl-4-picoline, CH2Cl2, rt, 76%; (vii) NaH, THF,
88% or NaHMDS, THF, 86%; (viii) HCl, acetone-water, 97%; (ix)
HNR2, NaBH(OAc)3, DCE, ca. 80%, then separate cis- and trans-
isomers (Notes: yields quoted are for n = 1 sequence; all numbered
compounds are racemic).
CF3
CF3
N
N
iii
i, ii
O
4a
O
CF3
CF3
NR2
O
10
8
Scheme 3. Reagents: (i) MeI, DMF, then NaH, DMF; (ii) HCl,
acetone–water; (iii) HNR2, NaBH(OAc)3, DCE, ca. 80% then separate
isomers.

G. J. Hollingworth et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1197–1201
1199
Table 1. NK1 receptor binding affinity, gerbil foot-tapping results and IKr channel activity for the lactam series. Selected a-methyl amides are shown
for comparison
( )n
R
CF3
N
O
CF3
(±)
NR2
Gerbil FTb ID50
(mg/kg)
a
hNK1 IC50 (nM)
c
hIKr Ki (nM)
Compound
NR2
n
R
t = 0
t = 24 h
5a
5b
1
1
H
H
6.2
—
—
984
327
N
O
0.50
0.35
62% at 3
34% at 3
2.10
N
N
O
5c
5d
5e
7a
8a
8b
8c
9a
2
1
2
1
1
1
1
1
H
H
H
0.80
1.1
—
798
1996
6529
862
N
O
N
O
N
N
0.09
—
1.5
—
O
O
OH
Me
0.66
0.33
0.48
0.48
—
5% at 3
0.24
N
N
—
364
N
O
N
N
Me
0.22
—
0.60
1946
2987
947
N
Me
57% at 3
29% at 3
Cl O
O
OMe
0.48
—
N
H
CF3
N
O
CF3
(±)
NR2
Gerbil FTb ID50
(mg/kg)
a
hNK1 IC50 (nM)
c
hIKr Ki (nM)
Compound
NR2
t = 0
t = 24 h
O
1
0.34
0.62
1.44
1.62
1.20
787
N
N
O
N
11
0.13
3706
a Displacement of [125I]-labelled substance P from the cloned hNK1 receptor expressed in CHO cells (mean, n = 3).8
b Inhibition of GR73632-induced foot-tapping in gerbils.5 All NK1 antagonists given intravenously. Where ID50 value is not determined, % inhibition
at 3 mg/kg is quoted.
c Displacement of labelled MK-499 from the cloned channel expressed in HEK cells (mean, n = 3).9

1200
G. J. Hollingworth et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1197–1201
underwent deketalisation in 99% yield to give ketone 10.
This was elaborated to final compounds 8 as before.
1, indicating improved brain penetration in agreement
with the original hypothesis. For the inherently more
brain penetrant aryl piperazinone series, it was pleasing
that the lactams 5d and 8b retained the good activity of
the lead 11.
A series of lactams was synthesised predominantly
where the pendant amine substituent is a piperidine spi-
roether (as is the case for the lead compound 1) or a 1-
aryl-piperazin-2-one. The latter are based on the related,
more potent lead 11.7 The biological data for these lac-
tams are shown in Table 1.
Having demonstrated good brain penetration for the
lactam series, several NK1 antagonists were tested in
the gerbil assay after 24 h pretreatment to assess their
central duration of action. Both 5b and 5d were interest-
ing compounds, with activity being observed, although
they were poorer in this protocol than their analogous
amides 1 and 11, respectively. Substitution at the benzyl-
ic position with hydroxy or methoxy groups (7a and 9a)
was detrimental, no improvement over the lead being
seen. In contrast, substitution with a methyl group gave
compounds (8a and 8b) that did show a significantly im-
proved central duration of action over their non-methyl-
ated counterparts (5b and 5d) and also over the lead
compounds 1 and 11, respectively.
It was pleasing to find that high NK1 receptor binding
affinity was observed across the whole series. As seen
for the a-methyl amide series,4,7 the piperidine spiro-
ethers and 1-arylpiperazin-2-ones do give a real benefit
over the simple dimethylamino group of 5a. Both five-
and six-membered lactam rings are well tolerated (5b–
e) and hydroxy, methyl, and methoxy substituents all
appear to be acceptable at the benzylic position (7a,
8a–c and 9a).
The lactams were also tested for their binding selectivity
for NK1 over the hERG cardiac potassium channel
responsible for the IKr current. The selectivities are sim-
ilar to those observed for the secondary amide ana-
logues and range from modest (650-fold for 5b) to
excellent (4300-fold for 5e). The piperazinones appear
to give a slightly better selectivity than the correspond-
ing piperidine spiroether analogues, however, it would
be desirable to improve further upon this selectivity
window.
The excellent activity of these compounds in the gerbil
assay was an extremely encouraging result, and two
compounds (5b and 8a) were further evaluated in our
ferret emesis model.10 In this assay, the inhibition by
an NK1 antagonist of cis-platin-induced emesis is mea-
sured. This functional assay is clearly relevant to the
use of an NK1 antagonist for the treatment of chemo-
therapy-induced nausea and vomiting. The emetic re-
sponse is centrally mediated and in our experience
only highly active, rapidly brain penetrant compounds
show activity in this assay. It was therefore gratifying
to find that both of these compounds did show a posi-
tive effect.
Several compounds were profiled in the gerbil foot-
tapping assay at the t = 0 time-point. The piperidine
spiroether 5b showed improved activity over the lead
240
200
160
120
80
240
200
160
120
*
80
40
0
40
*
0
Vehicle
5b (1mg/kg iv)
Vehicle
8a (1mg/kg iv)
Treatment
Treatment
30
25
20
15
10
5
30
25
20
15
10
5
*
0
0
Vehicle
5b (1mg/kg iv)
Vehicle
8a (1mg/kg iv)
Treatment
Treatment
Figure 1. Effects of 5b (blue) and 8a (red) on cis-platin induced emesis in ferrets.

G. J. Hollingworth et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1197–1201
1201
Carlson, E.; Harrison, T.; Haworth, K. E.; Herbert, R.;
Kelleher, F. J.; Kurtz, M. M.; Moseley, J.; Owen, S. N.;
Owens, A. P.; Sadowski, S. J.; Swain, C. J.; Williams, B. J.
Bioorg. Med. Chem. Lett. 2002, 12, 2515.
Ferrets were dosed intravenously with either 1 mg/kg of
NK1 antagonist or vehicle immediately prior to admin-
istration of 10 mg/kg iv cis-platin. The animals were
then observed over a 4 h period and the retching and
vomiting recorded. The results are shown in Figure 1.
4. (a) Elliott, J. M.; Castro, J. L.; Chicchi, G. G.; Cooper, L.
C.; Dinnell, K.; Hollingworth, G. J.; Ridgill, M. P.;
Rycroft, W.; Kurtz, M. M.; Shaw, D. E.; Swain, C. J.;
Tsao, K.-L. Bioorg. Med. Chem. Lett. 2002, 12, 1755; (b)
Cooper, L. C.; Carlson, E. J.; Castro, J. L.; Chicchi, G. G.;
Dinnell, K.; Di Salvo, J.; Elliott, J. M.; Hollingworth, G.
J.; Kurtz, M. M.; Ridgill, M. P.; Rycroft, W.; Tsao, K-L.;
Swain, C. J. Bioorg. Med. Chem. Lett. 2002, 12, 1759.
5. Rupniak, N. M. J.; Tattersall, F. D.; Williams, A. R.;
Rycroft, W.; Carlson, E.; 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. Pharmacol. 1997, 326, 201.
6. Reductive amination reactions were typically unselective
or slightly trans-selective (ratios 1:1 to 2.5:1). 1H NMR
experiments (400 MHz, CD3OD) show that, in both
isomers, the cyclohexyl ring adopts a chair conformation
with the C-1 hydrogen axial. In cis-isomers, NOEs are
seen from the o-hydrogens of the 4-phenyl group to both
axial and equatorial C-3 hydrogens, indicating that the 4-
phenyl group is equatorial. In trans-isomers, NOEs are
seen from the o-hydrogens of the 4-phenyl group to the
equatorial C-3 hydrogens and to the axial C-2 hydrogens,
indicating that the 4-phenyl group is axial.
It can be seen that a 50% inhibition of both retching and
vomiting was observed for 5b, which is an encouraging
result. Methylation, however, appears to make a signif-
icant difference, compound 8a being superior in this
assay with an excellent 90% block of the emesis.11
In summary, a new class of high affinity cyclohexyl-
amine NK1 antagonists containing five- or six-mem-
bered lactam rings has been identified, with modest to
excellent binding selectivity over the hIKr potassium
channel. These compounds are brain penetrant (demon-
strated by activity in a gerbil foot-tapping assay) and by
judicious choice of substitution at the benzylic position,
a long central duration of action can be achieved. The
most promising compound 8a shows good activity in a
ferret emesis model when dosed intravenously with
ID90 of ꢀ1 mg/kg.
References and notes
7. Ridgill, M. P., manuscript in preparation.
8. 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.
9. IKr affinity is measured using the MK-499 binding assay
which is performed on membrane preparations from
human embryonic kidney (HEK) cells constitutively
expressing hERG. For a detailed protocol, see Ref. 4b.
10. 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.
1. Dando, T. M.; Perry, C. M. Drugs 2004, 64, 777.
2. (a) Leroy, V.; Mauser, P.; Gao, Z.; Peet, N. P. Expert
Opin. Investig. Drugs 2000, 9; (b) Seward, E. M.; Swain, C.
J. Exp. Opin. Ther. Patents 1999, 9; (c) Swain, C. J.;
Rupniak, N. M. J. Annu. Rep. Med. Chem. 1999, 33, 51;
(d) Holzer, P.; Holzer-Petsche, U. Curr. Opin. Pharm.
2001, 1, 583, and references cited therein.
3. (a) Cooper, L. C.; Chicchi, G. G.; Dinnell, K.; Elliott, J.
M.; Hollingworth, G. J.; Kurtz, M. M.; Locker, K. L.;
Morrison, D.; Shaw, D. E.; Tsao, K.-L.; Watt, A. P.;
Williams, A. R.; Swain, C. J. Bioorg. Med. Chem. Lett.
2001, 11, 1233; (b) Dinnell, K.; Chicchi, G. G.; Dhar, M.
J.; Elliott, J. M.; Hollingworth, G. J.; Kurtz, M. M.;
Ridgill, M. P.; Rycroft, W.; Tsao, K.-L.; Williams, A. R.;
Swain, C. J. Bioorg. Med. Chem. Lett. 2001, 11, 1237; (c)
Williams, B. J.; Cascieri, M. A.; Chicchi, G. G.; Harrison,
T.; Owens, A. P.; Owen, S. N.; Rupniak, N. M. J.;
Tattersall, F. D.; Williams, A. R.; Swain, C. J. Bioorg.
Med. Chem. Lett. 2002, 12, 2719; (d) Seward, E. M.;
11. Spectroscopic data for compound 8a: m/z (ES+) 609
1
(M+1); H NMR (400 MHz, CDCl3) d 7.73 (2H, s), 7.71
(1H, s), 7.51 (2H, d, J 6), 7.34 (2H, t, J 6), 7.25 (1H, t, J 6),
3.80 (2H, t, J 7), 3.48 (2H, s), 3.27–3.18 (1H, m), 3.15–3.08
(1H, m), 3.00–2.85 (2H, m), 2.50–2.20 (6H, m), 2.06–1.97
(1H, m), 1.87–1.75 (2H, m), 1.70–1.50 (8H, m), 1.47–1.32
(2H, m), 1.40 (3H, s).