Discovery and evolution of phenoxypiperidine hydroxyamide dual CCR3/H 1 antagonists. Part i

Article abstract of DOI:10.1016/j.bmcl.2012.09.113

The discovery of potent small molecule dual antagonists of the human CCR3 and H₁ receptors is described for the treatment of allergic diseases, for example, asthma and allergic rhinitis. Optimizing in vitro potency and metabolic stability, star

Full text of DOI:10.1016/j.bmcl.2012.09.113

Bioorganic & Medicinal Chemistry Letters 22 (2012) 7702–7706
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery and evolution of phenoxypiperidine hydroxyamide dual CCR3/H1
antagonists. Part I
Mark Furber d, , Lilian Alcaraz a, Christopher Luckhurst a, Ash Bahl b, Haydn Beaton a, Keith Bowers b,
⇑
John Collington c, Rebecca Denton c, David Donald a, Elizabeth Kinchin a, Cathy MacDonald c,
Aaron Rigby a, Rob Riley c, Matt Soars c, Brian Springthorpe a, Peter Webborn c
a Medicinal Chemistry, AstraZeneca R&D Charnwood, Loughborough, Leicestershire LE11 5RH, UK
b Discovery Biosciences, AstraZeneca R&D Charnwood, Loughborough, Leicestershire LE11 5RH, UK
c Physical and Metabolic Sciences, AstraZeneca R&D Charnwood, Loughborough, Leicestershire LE11 5RH, UK
d AstraZeneca Respiratory & Inflammation iMED, Pepparedsleden 1, 431 83 Mölndal, Sweden
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online 23 October 2012
The discovery of potent small molecule dual antagonists of the human CCR3 and H1 receptors is described
for the treatment of allergic diseases, for example, asthma and allergic rhinitis. Optimizing in vitro
potency and metabolic stability, starting from a CCR1 lead compound, led to compound 20 with potent
dual CCR3/H1 activity and in vitro metabolic stability.
Keywords:
CCR3
H1
Ó 2012 Elsevier Ltd. All rights reserved.
Receptor antagonist
Dual inhibitor
Chemokine antagonist
Allergic disease
Chemokines are a class of more than forty small peptides that
act on subsets of leukocytes via G protein-coupled receptors
(GPCRs).1,2 The chemokines control migration, recruitment and
activation of leukocytes during inflammation. CCL11 (eotaxin)
and two related chemokines, CCL24 (eotaxin-2) and CCL26 (eotax-
in-3) all act solely through a single receptor, CCR3. Eotaxin is pro-
duced by lung fibroblasts, alveolar macrophages and eosinophils
and its release causes the recruitment and activation of eosino-
phils, basophils and Th2 lymphocytes. Lung eosinophilia is a char-
acteristic symptom of asthma that is thought to relate to disease
severity.3 Inhibition of CCR3 is therefore expected to have a bene-
ficial effect on patients who suffer from asthma4–6 and several
groups have reported the discovery of antagonists of CCR3.7–11
Histamine is a significant mediator of inflammation and is re-
leased from the granules of mast cells following binding of IgE to
the mast cell after antigen binding. Histamine contributes to bron-
choconstriction in asthma and to the symptoms of allergic rhinitis
experienced by atopic subjects through an increase in vascular per-
meability. Many asthmatic patients take H1 antagonists in addition
to anti-asthmatic medication.12
and mast cell degranulation. Here we describe the identification
and development of a novel series of phenoxypiperidine hydroxya-
mides as dual CCR3/H1 antagonists.13
With this aim, a new chemical program was initiated following
an observation that a subset of compounds originally designed as
CCR1 antagonists had good cross-over activity at the CCR3 receptor
and also displayed some level of H1 activity.14 An example of such a
compound is shown in Figure 1.
The 2-acylamino aryl functionality was known to be an impor-
tant determinant of activity at the CCR1 receptor, and so removal
or replacement of this moiety, or introduction of additional substit-
uents, was expected to reduce CCR1 activity.15 Most of the com-
pounds lacking this functionality displayed low CCR1 potency; as
a result CCR1 screening was subsequently only performed on se-
lected compounds. In order to understand the structural features
which would increase CCR3 and H1 potency whilst further reduc-
ing CCR1 activity, a library of 900 compounds was targeted via a
O
Cl
Cl
O
OH
HN
Dual CCR3/H1 antagonists were therefore seen as potentially
offering a combined beneficial effect on eosinophil recruitment
N
O
1: CCR1 pKi 7.8, CCR3 pKi 7.0, H1 pIC50 5.3
Figure 1. CCR1 antagonist with CCR3 activity.
⇑
Corresponding author. Tel.: +46 31 7761062.
E-mail address: mark1.furber@astrazeneca.com (M. Furber).
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2012.09.113

M. Furber et al. / Bioorg. Med. Chem. Lett. 22 (2012) 7702–7706
7703
parallel synthesis approach. The compounds were accessed
CN
Cl
Cl
N
through the reaction of phenols with (S)-glycidyl nosylate 8 follow-
ing treatment of the resulting epoxides with six amine scaffolds
(2–7) selected based upon previous project experience (Scheme
1). In addition, one sub-library of compounds was synthesised
using the (R)-enantiomer of 8 and amine 2, although compounds
in this library proved ca 1 log unit less potent in CCR3 binding than
equivalent products derived from (S)-8. In total, 800 compounds
were isolated after HPLC purification.
N
H
O
OH
9: CCR3 6.9, H1 6.3, Clint HLM <3
Cl
Cl
N
S
N
N
H
O
OH
The purified compounds were spot-tested at a concentration of
1 lM in the CCR3 binding assay. Active compounds (>70% inhibi-
10: CCR3 6.8, H1 6.5, Clint HLM <3
tion) were then selected for pKi determination at CCR3 and H1.16
Thirty compounds from each of the three series derived from 2, 5
and 6 were found to have CCR3 pKi values between 6 and 7; com-
pounds derived from 3 and 4 were less active and not pursued fur-
ther. Aminopiperidine 6 afforded compounds with modest activity
at both CCR3 and H1 receptors, and compounds were stable in sim-
ple in vitro PK assays, for example, human liver microsomes
(examples 9 and 10, Fig. 2). However, it was felt that increasing po-
tency at both receptors simultaneously would be a greater task
than were the project to focus on a series already possessing high
activity at one of the two receptors. For this reason, the series of
compounds derived from amine 2 was identified as having greater
promise for optimization, given it already had high H1 potency
within the series (Table 1, entry 1).
Single point modifications were investigated to understand the
effect of chain length and the nature of the linker on CCR3 activity
and the degree of CCR3 versus H1 selectivity. For reasons discussed,
the 3,4-dichlorophenoxypiperidine 2 was chosen as the amine
scaffold and 3-cyanophenyl as the sidechain aryl moiety, from
the best lead to have emerged from the library exercise (Table 1,
entry 1).
Figure 2. Two compounds based upon scaffold
clearance Clint human liver microsome (HLM) data.
6 with potency and intrinsic
and the hydroxyl group led to an appreciable gain in CCR3 binding
that was accompanied by a large increment in H1 potency to a pKi
value of 9.5 (entry 6). This was a general observation with related
examples, and could be compared to the increased potency
achieved through incorporation of the amide linker, which is also
one atom longer. Inclusion of the amide linker also resulted in a
further reduction in lipophilicity (>1 log reduction in measured
logD) whilst simultaneously increasing CCR3 potency by 1 log unit
versus the corresponding ether. Although not having the very high
H1 potency of example 6 this change gave balanced dual CCR3/H1
potency such that further chemistry investigation became focused
around compound 15.
Exploration of the amide aryl group was performed by parallel
synthesis using a selected range of acids such that compounds with
acceptable logD values (<3.5) would be obtained (Scheme 2).
Limiting the lipophilicity of compounds of this series proved to
be challenging.
Removal of the hydroxyl functionality decreased CCR3 binding
marginally (Table 1, entry 2), but presence of this hydroxyl group
was seen to increase metabolic stability by reducing lipophilicity,
although this was still higher than typically measured. The ether
oxygen could be replaced with sulfur or sulfonamide functionality
without affecting CCR3 activity, whilst H1 activity diminished
(Table 1, entries 3 and 4). Substituting the ether oxygen for an
amide group afforded a selective increase of one log unit potency
at CCR3 without increasing H1 potency, to yield an equipotent
compound at both receptors (Table 1, entry 5). Introduction of an
extra methylene group in the linker between the amine nitrogen
Most of the resulting compounds displayed similar activities at
both CCR3 and H1 receptors, in the 100 nM IC50 range. Some typical
examples are shown in Table 2 (entries 1 and 2). However, two
compounds (entries 3 and 4) displayed significantly improved po-
tency at CCR3 with pKi values of 8.4 and 8.8, respectively.
One consideration in further optimizing the series was activity
at the hERG channel and the potential for QT prolongation leading
to cardiovascular issues in vivo. Some compounds within this series
exhibited activity at the hERG channel (e.g., compound 19, hERG
pIC50 6.6).17 Gratifyingly, a number of compounds displayed low
hERG potency, notably compound 20 combined advantages of high
H
Cl
Cl
N
Cl
Cl
O
N
N
NH
NH
NH
NH
Cl
Cl
Cl
Cl
2
3
6
4
Cl
Cl
O
Cl
Cl
NH2
NH2
5
8
7
O
O
S
NO2
O O
O
Ar150 OH
O
Ar150
Cl
Cl
W
X
OH
Y
O
Z
Ar150
800 compounds after purification
Scheme 1. Parallel synthesis approach for the reaction of phenols with (S)-epoxynosylate and reaction of the resulting epoxides with amine scaffolds.

7704
M. Furber et al. / Bioorg. Med. Chem. Lett. 22 (2012) 7702–7706
Table 1
Single point modifications to the phenoxypiperidine-based hydroxy ethers: effect of sidechain structure on CCR3 and H1 potencies, and microsomal stability (intrinsic clearance
Clint: H/RLM: human/rat liver microsomes
l
L/min/mg protein)
Cl
O
V
N
X
R
Cl
X
Entry
Compound
V
W
X
R
CCR3 binding pKi16
H1 binding pKi16
HLM Clint
RLM Clint
LogD
1
2
3
4
5
6
11
12
13
14
15
16
OH
H
OH
OH
OH
OH
CH2
CH2
CH2
CH2
CH2
CH2–CH2
O
O
S
CN
CN
OMe
CN
CN
6.6
6.1
6.9
6.8
7.7
7.5
8.5
7.9
7.7
7.1
7.7
9.5
<3
3
17
7
4
<3
23
129
49
13
25
22
>4.5
>4.5
>4.5
3.4
3.5
>4.5
NHSO2
NHCO
O
CN
Cl
Cl
O
of compounds a number of interesting trends were observed. High
CCR3 activity was retained in compounds that possessed an appro-
priately positioned hydrogen bond donor (N–H) in the heterocyclic
ring system, for example within an isoquinolinone (entries 1 and 6),
indazole (entry 2), indole (entry 3), phthalazinone (entry 4) or
benzimidazolone (entry 5) ring system. Partial charge calculations
also showed very similar values for the positive charge at the N–
H functionality. Reduced CCR3 activity was observed when this
N–H group was either moved or removed (entries 7 and 8). High
hERG activity was a characteristic of this series in general, with
the exception of a number of examples such as compounds 20
and 27. Compounds able to form a five- or six-membered hydro-
gen-bond between the amide N–H of the linker and the heterocyclic
nitrogen atom (entries 2 and 4) or oxygen atom (entry 5) of the het-
erocycle tended to show higher hERG activity, possibly through
favouring a hydrogen-bonded planar conformation.
OH
N
NH2
ArCOOH
HATU, DMF, 0 °C
Cl
Cl
O
OH
H
N
N
Ar
O
Scheme 2. Parallel synthesis of hydroxyamide derivatives.
CCR3 and H1 potency with a low hERG pIC50 value of 5.1. This com-
pound displayed expectedly low CCR1 activity pIC50 5.3. Compared
to other sidechain units, the isoquinolinone ring system offered an
advantage in terms of its high CCR3 potency and its reduced hERG
activity. To capitalise on this, a number of compounds with closely
analogous ring systems were prepared (Table 3). Within this series
This was supported by calculations that indicated a significant
barrier to rotation around the amide carbonyl to heterocycle bond
of 20 with both planar conformations being of high energy. Devia-
tion from planarity in 20 is presumably a consequence of the bicy-
Table 2
Data for compounds prepared by the route shown in Scheme 2 (Clint: RLM: rat liver microsomes
tested)
lL/min/mg protein, RH/HH: rat/human hepatocytes l
L/min/106 cells; NT = not
Ar
Entry
Compound
logD
CCR3 binding pKi16
h1 binding pKi16
Clint RLM
Clint RH
Clint HH
O
F
N
1
17
3.3
6.7
7.8
19
7
NT
N
O
O
SO2Me
2
3
18
19
2.9
3.1
6.6
8.4
7.0
7.2
20
18
3
6
NT
<2
S
NH2
OMe
O
H
N
O
4
5
20
21
3.3
2.3
8.8
7.8
8.4
6.7
20
8
<3
<3
NT
<2
O
O
SO2Me

M. Furber et al. / Bioorg. Med. Chem. Lett. 22 (2012) 7702–7706
7705
Table 3
Replacements for the isoquinolinone ring system in compound 20
Cl
Cl
O
OH
X
N
Ar
N
O
Ar
17
Entry
Compound
X
logD
CCR3 binding pKi16
H1 binding pKi16
hERG pIC50
O
H
N
O
1
20
H
3.3
8.8
8.4
5.1
O
O
O
H
N
N
2
3
22
23
H
H
2.9
3.9
8.8
8.6
8.1
7.5
6.6
5.7
H
N
H
N
O
N
4
5
6
24
25
26
H
3.7
4.5
3.2
8.8
8.6
8.6
8.2
7.7
8.4
6.4
6.3
5.7
O
O
O
H
N
H
N
H
N
O
Me
O
O
N
NH
7
8
27
28
H
H
3.5
3.3
8.1
7.9
8.4
7.9
<5
O
O
N
5.2
clic nature of the heterocyclic ring system in this compound and the
steric interactions that would preclude a planar conformation.
Having established some understanding of the factors driving
potency at CCR3 and H1 with the reduction of hERG activity, atten-
tion turned to pharmacokinetic properties of the compounds. As
discussed for compounds in Table 2, in vitro DMPK properties of
these were generally observed to be acceptable in rat liver hepato-
cytes (RH) and rat liver microsomal (RLM) preparations. However,
compounds were observed to be cleared rapidly in vivo in the rat,
illustrated by PK data for two representative compounds: 19 (Cl
compared to 30 and 30 compared to 21, Table 4). This is probably
a result of the chain length already being one atom longer than that
in the phenoxypiperidine series. H1 activity was essentially unaf-
fected by this change, but was already intrinsically lower in this
series. The disconnect between in vitro and in vivo clearance in
the aminopiperidine series was even more significant.
Within this chemical series, the lower logD translated into im-
proved in vitro metabolic stability. Unfortunately, when pro-
gressed in vivo in the rat, clearance and volume of distribution
for compound 30 were high (Cl 54 mL/min/kg, Vss 9 L/kg, t 2.3 h).
½
56 mL/min/kg, Vss 25 L/kg, t 6.5 h, F 30%) and 20 (Cl 60 mL/min/
In light of the high in vivo clearance of compounds in this series,
exemplified by compound 20, the second part to this communica-
tion will highlight steps to reduce in vivo clearance of compounds
and address the poor in vitro to in vivo scaling in the rat.
The provision of intermediates of high chemical and enantio-
meric purity was required en route to the synthesis of compound
20 (Scheme 3, conditions i). In addition, a route needed to be devel-
oped which would be suitable for scale-up and easy access of inter-
mediates (Scheme 3, conditions ii).
½
kg, Vss 12 L/kg, t 2.8 h, F 11%). Given that in vivo clearance ap-
½
peared to be considerably higher than that expected from the
in vitro Clint data, lowering in vivo clearance in this series of com-
pounds became the focus of further efforts.
Interestingly, a similar exercise performed in the aminopiperi-
dine series (i.e., based on scaffold 6) gave somewhat different re-
sults whereby replacement of the ether oxygen with an amide
group gave a smaller increase in CCR3 binding, (compound 29

7706
M. Furber et al. / Bioorg. Med. Chem. Lett. 22 (2012) 7702–7706
Table 4
Data for compounds 29 and 30 (intrinsic clearance Clint: RLM: rat liver microsomes
lL/min/mg protein; RH: rat hepatocytes l
L/min/106 cells)
Cl
N
N
H
O
CN
Cl
OH
OH
29
Cl
N
O
SO2Me
N
N
Cl
H
H
30
Compound
logD
CCR3 binding pKi16
H1 binding pKi16
RLM Clint
RH Clint
29
30
3.1
1.2
6.5
7.1
6.3
5.9
23
<3
7
4
H
N
Cl
O
O
Cl
Cl
O
Cl
Cl
O
a,b
c
OH
OH
H
N
N
NH2
NH
N
Cl
O
2
31
20
Scheme 3. Reagents and conditions: (i) (a) (R)-glycidylnosylate 8, DMF, Et3N, 60 °C, then NaN3, DMF, 60 °C; (b) Ph3P, THF, water, 60 °C; (c) ArCOCl, i-Pr2EtN, DCM.; (ii) (a) (S)-
epichlorohydrin, EtOH; then MeOH, NaOH; (b) aq NH3; (c) ArCOCl, i-Pr2EtN, DCM.
Both routes start from dichlorophenoxypiperidine 2, which was
obtained in a single step from piperidin-4-ol and 1,2-dichloro-4-
fluorobenzene.13 In the route outlined using reagent conditions
(a) in Scheme 3 (i), glycidylnosylate 8 was the source of the 3-car-
bon side-chain; the dichlorophenoxy piperidine 2 displaced the
nosylate group to afford an amine epoxide in the first instance. This
was opened in situ with sodium azide and the resulting hydroxyaz-
ide was reduced to afford amine 31. Determination of enantiomeric
purity at this stage through a 1H NMR chiral shift reagent method
indicated only 0.7% of the enantiomeric impurity (ee 98.6%). Acyl-
ation and crystallization of the final product afforded a multi-gram
batch of compound 20 (ee 98.5%). The main limitations of this
route were principally around potential hazards associated with
the epoxynosylate and with the use of sodium azide, but also the
cost and availability of the (R)-glycidylnosylate reagent.
A second route was developed (Scheme 3, conditions ii) that
overcame the concerns highlighted above and provided material
of equivalent enantiomeric purity. In this process it was the epox-
ide ring of (S)-epichlorohydrin which was first opened by amine 2.
The chlorohydrin product from the first step was isolated and
underwent base-mediated cyclization. Opening of the epoxide
with ammonia obviated the requirement for sodium azide and di-
rectly afforded the primary amine intermediate 31, which was
then converted to compound 20 as described.
References and notes
1. Kunkel, S. L.; Standiford, T. J.; Strieter, R. M.; Kasama, T. Lung Biol. Health Dis.
1995, 80, 579.
2. Teran, L. M.; Velazquez, J. R. Inflamm. Allergy Drug Des. 2011, 2, 253.
3. Nakamura, H.; Weiss, S. T.; Israel, E.; Luster, A. D.; Drazen, J. M.; Lilly, C. M. Am.
J. Respir. Crit. Care Med. 1952, 1999, 160.
4. Pease, J. E. Methods Princ. Med. Chem. 2011, 46, 339 (Chemokine Receptors as
Drug Targets).
5. Lukacs, N. W.; Hogaboam, C. M.; Kunkel, S. L. Curr. Drug Targets Inflamm. Allergy
2005, 4, 313–317.
6. Hansel, T. T.; Barnes, P. J. Prog. Respir. Res. 2010, 39, 3.
7. Bahl, A. K.; Springthorpe, B.; Riley, R. Prog. Respir. Res. 2010, 39, 153.
8. Willems, L. I.; Ilzerman, A. P. Med. Chem. Rev. 2010, 30, 778–817.9; Gong, L.;
Wilhelm, R. S. Exp. Opin. Ther. Pat. 2009, 19, 1109–1132.
9. Grundl, M.; Dollinger, H.; Giovannini, R.; Hoenke, C.; Hoffmann, M.;
Kriegl, J.; Martyres, D.; Rast, G.; Seither, P. PCT Int. Appl. 2010, WO
2010115836.
10. Ly, T. W.; Forrester, J. A.; Potter, G. T. PCT Int. Appl. 2011, WO2011116161.
11. Ly, T. W. U.S. Pat. Appl. Publ. 2012, US 20120088769.
12. Wilson, A. M. Treat. Respir. Med. 2006, 5, 149–158.
13. Alcaraz, L.; Furber, M.; Purdie, M.; Springthorpe, B. WO2003068743.
14. Eriksson, T.; Henriksson, K. WO2001098272.
15. Unpublished results.
16. Human CCR3 pKi was determined in Chinese hamster ovary cells expressing
human recombinant CCR3 receptors as described in WO2007102768, Example
13. Human H1 pKi was determined in competition binding experiments as
described in WO2006126948, Example 6.
17. hERG pIC50 was determined in human embryonic kidney cells via ion flux
electrophysiology.