Discovery and evolution of phenoxypiperidine hydroxyamide dual CCR3/H 1 antagonists. Part II: Optimising in vivo clearance

Article abstract of DOI:10.1016/j.bmcl.2012.09.112

The second part of this communication focuses on the resolution of issues surrounding the series of hydroxyamide phenoxypiperidine CCR3/H₁ dual antagonists described in Part I. This involved further structural exploration directed at reducing m

Full text of DOI:10.1016/j.bmcl.2012.09.112

Bioorganic & Medicinal Chemistry Letters 22 (2012) 7707–7710
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 II: Optimising in vivo clearance
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 10 October 2012
The second part of this communication focuses on the resolution of issues surrounding the series of
hydroxyamide phenoxypiperidine CCR3/H1 dual antagonists described in Part I. This involved further
structural exploration directed at reducing metabolism and leading to the identification of compound
60 with a greatly improved in vivo pharmacokinetic profile.
Keywords:
CCR3
H1
Ó 2012 Elsevier Ltd. All rights reserved.
Receptor antagonist
Dual inhibitor
Chemokine antagonist
Allergic disease
As described in the first part of this communication, dual CCR3/
H1 antagonists were considered as having potential benefit in the
treatment of asthma.1 Further progress in the identification of phe-
noxypiperidine hydroxy amides as dual CCR3/H1 antagonists is de-
scribed in this second part. Although presenting a very promising
in vitro profile in terms of potency at both CCR3 and H1 receptors,
low affinity at hERG channel and encouraging in vitro ADME prop-
erties, one of the major issues with the lead compound 20 (Fig. 1)
identified in Part I, was the poor in vitro to in vivo correlation
(ivivc), specifically the high clearance observed in vivo.
This poor scaling may be connected with the high volumes of
distribution associated with these compounds that in turn may
be attributed to their lipophilicity. Two principal approaches were
investigated to reduce in vivo clearance in rat within the series.
The first focused on reducing lipophilicity, with the dichlorophenyl
moiety being identified as problematic in this respect. The second
focused on identifying and blocking sites of metabolism.
compound 32 displayed low hERG activity (pIC50 <5).3 Much of this
reduction in CCR3 affinity was compensated by a reduction in hu-
man plasma protein binding from 98% to 79%; a 10-fold increase in
free drug concentration. Furthermore, the compound displayed
low in vivo clearance in rat (8.6 ll/min/kg) and a low volume of
distribution (Vss 0.5 L/kg) which was more in line with expecta-
tions from in vitro rat hepatocyte data (Clint 2 ml/min/1E6 cells).
Unfortunately, this compound demonstrated poor oral bioavail-
ability in rat (F <5%); likely due to poor intestinal absorption pre-
dicted from low calculated permeability, and perhaps not
unexpected for a primary carboxamide that possesses a number
of additional hydrogen bond donor/acceptor groups. Furthermore,
structure–activity proved very tight around this molecule and did
not allow modification of the primary carboxamide group. Whilst
this molecule itself could not be progressed, it did indicate that
with appropriate modification of physicochemical properties a bal-
ance of dual CCR3/H1 potency and suitable PK properties might be
achievable.
Table 1 also illustrates the effect of replacing one or both Cl
groups with a more polar cyano substituent, with compound 34
affording very similar CCR3 activity (pKi 8.6) to compound 20 itself.
Unfortunately, in vitro stability of this compound in a rat hepato-
cyte system was disappointing. It should be noted that simple re-
moval of both Cl substituents results in a greater than 2log drop
in CCR3 potency. However, combining the 3-chloro-4-cyano
substituted aryl ring system with methanesulfonyl substitution
In the first instance, attention was directed toward introducing
polarity into the dichlorophenyl portion of the molecule (Table 1).
Introduction of a primary amide group into the dichlorophenoxy
ring (compound 32) resulted in a significant drop in measured
logD, but only a 10-fold drop in CCR3 affinity (pKi 7.7).2 In addition,
⇑
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.2012.09.112

7708
M. Furber et al. / Bioorg. Med. Chem. Lett. 22 (2012) 7707–7710
H
N
compounds had very high in vivo clearance, approaching rat liver
blood flow.
This inability to predict in vivo clearances from in vitro data due
to a complete lack of ivivc precluded further progression of this
isoquinolinone subseries.
O
O
Cl
Cl
OH
H
N
N
O
7
20
Poor ivivc is not infrequently seen with relatively lipophilic
amines and is not always understood. Consequently, a more prag-
matic approach was undertaken to concentrate efforts in those
subseries whose in vivo clearance was predictable from in vitro
measurements, even though this could not be rationalised at the
time. This change in direction ultimately provided compounds
with lower in vivo clearance. To this aim, we reassessed a number
of compounds that that had been synthesised during the course of
the project but had not progressed for a range of reasons and
elected to revisit compound 53.
CCR3 pKi 8.8, H1 pKi 8.4, hERG pIC50 5.2
Clint rat mics 19µL/min/mg; rat heps 4µL/min/106 cells;
dog heps 3µL/min/106cells; human heps 4µL/min/106 cells.
in vivo PK:
rat: Cl 60 mL/min/kg; Vss 12 L/kg; t1/2 2.8 h; F 11%
dog: Cl 18 mL/min/kg; Vss 5.5 L/kg; t1/2 6.0 h; F 40%
logD 3.3; hu ppb 98%; pKa 7.7 (base); 10.9 (acid)
Figure 1. Properties and site of metabolism of compound 20.
When dosed iv to rat, compound 53 displayed an in vivo clear-
ance of 22 mL/min/kg, a Vss of 10 L/kg, and a t1/2 of 6 h. These re-
sults were in line with that expected from the in vitro data.
When dosed orally in the rat this compound exhibited good plasma
exposure and reasonable bioavailability (F 36%). Initially the com-
pound had not been progressed owing to a low hERG margin. In
light of the ivivc issues outlined above with other series, close ana-
logues of compound 53 were prepared to explore the possibility of
reducing clearance and hERG activity within this series (Table 3).
To overcome the hERG issue with these compounds it was in-
tended to replace the pyrazole ring system in 53 with the pyridone
ring present in compound 20 in the hope that the combined fea-
tures of lower in vivo clearance and low hERG activity could be
achieved in one molecule (Table 4). Three compounds were pre-
pared with different aryloxy substituents but all possessing a tri-
fluoromethyl substituted pyridone ring. All three compounds
displayed high CCR3 and H1 potencies, low hERG activity and high
in vitro metabolic stability. Investigation of 57 and 58 in vivo
showed the compounds scaled from in vitro results and gave good
plasma exposure both in rat and in dog. Dosed orally, compound 57
was highly bioavailable in dog (63%) but much less so in rat (<10%).
Reasoning that permeability might be an issue for the NH pyridone
series, the N-methyl pyridone analogue of 57 (compound 60) was
prepared and finally achieved the balance of properties the project
sought. Compound 60 retained both CCR3 and H1 potencies (pKi
8.4 and 8.3) and low hERG binding efficacy (pIC50 4.9), whilst also
exhibiting low clearance and good oral bioavailability in rat (Cl
on the isoquinolone ring afforded compound 35 which had a fur-
ther reduced logD, and was found to be significantly more stable
in vitro. Unfortunately, as seen previously, this compound did
not scale when progressed to in vivo PK measurements, displaying
very high clearance in rat (Cl 70 mL/min/kg).
The second approach adopted to improve in vivo PK properties
was to block sites of metabolism in compound 20. The principal
route of metabolism of 20 as determined from in vitro and
in vivo rat and in vitro human systems was oxidation in the 7-
position of the isoquinolone ring system, with further glucuronida-
tion of the newly introduced hydroxyl group in hepatocytes.
Accordingly, compounds were prepared wherein the isoquinoli-
none ring system was substituted at or adjacent to the site of
metabolism with either a fluorine or a polar group such as
methanesulfone or a sulfonamide substituent (Table 2). In some
cases this provided the added benefit of lowering logD with the
possibility of reducing tissue distribution of the drug.
Within this series of compounds some very high CCR3 and H1
activities were observed. Compound 49, in particular, was found
to be an extremely potent dual CCR3/H1 antagonist. Nevertheless,
potency was not the main goal in this exercise and compounds
36, 40, 43 and 49 were dosed in vivo in rat to ascertain if improved
ivivc and lower in vivo clearance could be obtained by this ap-
proach. Unfortunately, the outcome was unfavourable; despite
lower Clint values in vitro and lower observed volumes of distribu-
tion in vivo (Vss 1.8, 6.4, 4.4, 0.6 L/kg respectively) all four
Table 1
LogD-lowering phenoxy and isoquinolinone substitutions
Compound
Structure
logD
1.5
CCR3 Binding pKi2
H1Binding pKi2
Clint RH4
Clint HLM4
O
NH2
O
H
N
Cl
O
OH
OH
OH
OH
32
7.7
6.6
2
18
H
N
N
Cl
O
H
O
O
O
N
O
H
33
34
NT
2.6
7.5
8.4
6.3
6.9
6
18
22
N
N
NC
O
H
Cl
N
O
H
10
N
N
N
NC
O
H
N
Cl
O
H
N
35
1.4
7.2
6.6
<3
<3
NC
O
SO2Me

M. Furber et al. / Bioorg. Med. Chem. Lett. 22 (2012) 7707–7710
7709
Table 2
Modifications to the substitution of the aryloxy and isoquinolinone ring systems
R1
H
N
R2
O
O
OH
H
N
R6
R5
N
R3
O
R4
R1
R2
R3
R4
R5
R6
LogD
CCR3 pKi2
H1 pKi2
Clint RH4
Clint HLM4
hERG binding pIC50
3
Compound
36
37
38
39
40
41
42
43
44
45
46
47
48
H
Cl
Me
H
H
Cl
H
Cl
Cl
Cl
Me
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
H
H
H
SO2Me
SO2Me
SO2Me
H
F
F
H
SO2NHCH2CH2OH
SO2NH2
H
SO2NHMe
SO2NHc-Pr
SO2NMe2
H
H
H
H
H
H
H
H
H
H
H
H
H
2.6
2.8
2.4
NT
3.2
NT
3.9
2.1
2.2
2.8
2.8
3.7
3.1
8.1
8.6
7.8
8.2
8.7
8.4
8.5
8.0
7.6
9.0
8.4
8.4
9.1
7.7
7.3
7.1
7.2
8.1
7.9
8.2
7.3
7.7
7.2
7.6
8.3
7.9
3
12
7
6
5.2
<4.5
5.7
<4.5
NT
NT
5.9
5.3
<5.2
<4.5
5.7
Cl
Me
Me
H
Me
H
H
H
Cl
H
H
13
14
7
8
9
SO2Me
H
H
F
H
H
8
<3
<3
5
8
6
5
<3
<3
<3
6
<3
<3
14
21
86
SO2Me
H
H
H
6.1
6.0
H
10
N
S
49
H
H
Cl
Cl
H
H
3.0
9.5
8.6
7
3
<3
6.0
5.5
O
O
OH
N
O
50
Cl
Cl
H
H
2.5
7.9
7.6
18
S
O
H
H
51
52
H
Me
Cl
H
Cl
Cl
H
H
F
F
3.2
3.0
8.7
9.1
8.6
8.6
<3
7
6
9
5.7
<4.5
Table 3
Pyrazole hydroxyamides
Compound
Structure
CCR3/H1 binding pKi2
hERG pIC50 (binding)3
5.5–6.5 (variable) (5.5)
PPB (human) %
Clint RH4
Clint HLM4
H
O
O
Cl
OH
OH
N
H
N
N
N
N
53
8.2/7.8
98
96
3
<3
Cl
O
O
CF3
CH3
H
N
N
H
N
54
8.5/7.6
5.9 (5.5)
5
<3
Cl
CF3
Cl
H
N
N
O
O
OH
OH
H
N
55
56
9.1/7.4
8.8/7.9
(5.7)
(5.6)
NT
96
6
<3
9
N
N
Cl
O
O
CF3
H
Cl
Cl
N
H
N
N
<3
9 mL/min/kg, Vss7.3 L/kg, t1/2 9.7 h, F 53%) and dog (Cl 8 mL/min/
kg, Vss 6.4 L/kg, t1/2 10.8 h, F 81%).
The 7-methanesulfonyl-substituted analogue was obtained by
chlorosulfonylation of isoquinolinone-4-carboxylic acid followed
by reduction and methylation of the resulting sulfonyl chloride
(Scheme 1).
The 6-methanesulfonyl, 6-fluoro and 7-fluoro substituted ana-
logues were synthesised as shown in Scheme 2.
The key step in this sequence was a copper catalysed coupling
reaction between an appropriately substituted ortho-halo benzoic
acid (61) and a malonate ester. The coupling conditions were found
to be substrate dependent and needed to be optimised individu-
ally. The hydrolysis-decarboxylation step was initially performed
using concentrated hydrochloric acid and was later replaced by a
saponification–decarboxylation procedure under basic conditions,
In summary, issues surrounding the series of hydroxyamide
phenoxypiperidines described in Part I had inspired structural
explorations culminating in the identification of compound 60.
The enhanced in vivo pharmacokinetic profile of this compound
has encouraged further studies in this series and ultimately led
to the discovery of a suitable clinical candidate for the project.
Synthetic chemistry
The synthetic routes towards substituted isoquinolinones
shown in Tables 1 and 2 are described below (Schemes 1 and 2).

7710
M. Furber et al. / Bioorg. Med. Chem. Lett. 22 (2012) 7707–7710
Table 4
Pyridone hydroxyamides
Compound
Structure
CCR3/H1 binding pKi2
hERG pIC50 (binding)3
5.0 (5.3)
PPB human (rat)%
89(91)
Clint RH4
Clint HLM4
H
N
O
O
O
O
Cl
Cl
OH
OH
H
N
57
8.5/8.4
4
<3
N
N
O
O
CF3
H
N
H
N
58
59
8.7/8.0
8.9/8.0
5.0 (5.0)
(5.3)
86
96
<3
3
3
6
Cl
Cl
CF3
Cl
H
N
O
O
O
O
OH
OH
H
N
N
N
O
CF3
CH3
N
Cl
Cl
H
N
60
8.4/8.3
(4.9)
93(92)
<3
5
O
O
CF3
H
H
H
N
N
N
O
O
a
b, c
HO
HO
HO
O
O
O
SO2Cl
SO2Me
Scheme 1. (a) ClSO3H, 100 °C; (b) Na2SO3, NaHCO3; (c) MeI, KHCO3.
RO
O
O
COOH
COOH
COOH
X
RO
HO
a
b
c
O
R2
R2
R2
R1
R1
R1
61
62
63
H
N
O
O
O
O
O
O
g
d
e,f
MeO
MeO
HO
TARGETS
O
O
R2
R2
R2
R1
R1
R1
64
65
66
Scheme 2. (a) For X = Br, R1 = H, R2 = F or R1 = F, R2 = H: (MeO2C)2CH2, NaH, CuBr, 80 °C; for X = Cl, R1 = SO2Me, R2 = H: (EtO2C)2CH2, EtOH, 80 °C; (b) conc HCl, 110 °C or NaOH,
then HCl; (c) (MeO)3 CH, acetic anhydride, 110 °C; (d) conc H2SO4, MeOH; (e) NH4OAc, AcOH, 80 °C; (f) NaOH; (g) CDI, DMF, then amine.
with shorter reaction times. Treatment of the diacid 63 with trim-
ethylorthoformate yielded the corresponding cyclised anhydride
64 that underwent rearrangement to the lactone 65. Finally, forma-
tion of the desired isoquinolinone acid was achieved by reaction
with ammonium acetate followed by hydrolysis of the resulting
methyl esters. It is worth noting that the entire sequence could
be performed without the need for chromatography since all of
the steps proceeded in high yield and the intermediates were iso-
lated as clean solids (>95% purity determined by reverse phase
HPLC) at each stage.
References and notes
1. Luckhurst, C. A.; Furber, M.; Alcaraz, L.; Bahl, A.; Beaton, H.; Bowers, K.;
Collington, J.; Denton, R.; Donald, D.; Kinchin, E.; MacDonald, C.; Rigby, A.; Riley,
R.; Soars, M.; Springthorpe, B.; Webborn, P. Bioorg. Med. Chem. Lett. 2012. http://
dx.doi.org/10.1016/j.bmcl.2012.09.113.
2. 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.
3. hERG pIC50 was determined in human embryonic kidney cells via ion flux
electrophysiology. hERG binding was performed using the procedure described
in WO2005037052.
Final coupling with the amine 31 described in Part I of this com-
munication1 was carried out under standard coupling conditions,
for example, using CDI in DMF.
4. Clint: HLM = human liver microsomes
lL/min/mg protein, RH = rat hepatocytes
l
L/min/106 cells; NT = not tested.