Inhalation by design: Novel tertiary amine muscarinic M3 receptor antagonists with slow off-rate binding kinetics for inhaled once-daily treatment of chronic obstructive pulmonary disease

Article abstract of DOI:10.1021/jm200884j

A novel tertiary amine series of potent muscarinic M₃ receptor antagonists are described that exhibit potential as inhaled long-acting bronchodilators for the treatment of chronic obstructive pulmonary disease. Geminal dimethyl functionality pr

Full text of DOI:10.1021/jm200884j

ARTICLE
pubs.acs.org/jmc
Inhalation by Design: Novel Tertiary Amine Muscarinic M3 Receptor
Antagonists with Slow Off-Rate Binding Kinetics for Inhaled
Once-Daily Treatment of Chronic Obstructive Pulmonary Disease
Paul A. Glossop,*,† Christine A. L. Watson,*,† David A. Price,†,# Mark E. Bunnage,† Donald S. Middleton,†
Anthony Wood,†,f Kim James,†,Δ Dannielle Roberts,† Ross S. Strang,† Michael Yeadon,‡
^
Christelle Perros-Huguet,‡,§ Nicholas P. Clarke,‡,3 Michael A. Trevethick,‡ Ian Machin,‡ Emilio F. Stuart,‡
||
||
Steven M. Evans,‡ Anthony C. Harrison,§ David A. Fairman,§, Balaji Agoram,§, Jane L. Burrows,^
^
Neil Feeder,^ Craig K. Fulton,^ Barry R. Dillon,^ David A. Entwistle,^ and Fiona J. Spence3
†Department of Worldwide Medicinal Chemistry, ‡Allergy and Respiratory Research Unit, §Department of Pharmacokinetics, Dynamics
^
and Metabolism, ^Department of Pharmaceutical Sciences, and 3Department of Drug Safety, Pfizer Global Research and Development,
Sandwich Laboratories, Ramsgate Road, Kent CT13 9NJ, U.K.
S Supporting Information
b
ABSTRACT: A novel tertiary amine series of potent muscarinic M3 receptor antagonists are
described that exhibit potential as inhaled long-acting bronchodilators for the treatment of
chronic obstructive pulmonary disease. Geminal dimethyl functionality present in this series
of compounds confers very long dissociative half-life (slow off-rate) from the M3 receptor
that mediates very long-lasting smooth muscle relaxation in guinea pig tracheal strips.
Optimization of pharmacokinetic properties was achieved by combining rapid oxidative
clearance with targeted introduction of a phenolic moiety to secure rapid glucuronidation. Together, these attributes minimize
systemic exposure following inhalation, mitigate potential drugꢀdrug interactions, and reduce systemically mediated adverse
events. Compound 47 (PF-3635659) is identified as a Phase II clinical candidate from this series with in vivo duration of action
studies confirming its potential for once-daily use in humans.
’ INTRODUCTION
receptors (M1ꢀM5) are known in humans, three of which are
present in human lung (M1ꢀM3). The M3 receptor is located on
airway smooth-muscle cells where it mediates the contractile
response.4
Chronic obstructive pulmonary disease (COPD) is character-
ized by airflow limitation that is both progressive and not fully
reversible, resulting in deterioration of lung function and chronic
dyspnea.1 The airflow limitation is associated with an abnormal
inflammatory response to noxious particles, often developing in
smokers. COPD is currently the fourth largest cause of death in
the U.S. and is projected to become the third leading cause of
death worldwide by 2020.1 Bronchodilator drugs are commonly
used to manage the symptoms of COPD, maximize lung func-
tion, and reduce exacerbation rates. Inhaled muscarinic recep-
tor antagonists such as ipratropium bromide and tiotropium
bromide2 (Chart 1) are examples of one class of bronchodilators
prescribed for symptom control. Ipratropium bromide is a short-
acting agent administered four times per day, whereas tiotropium
bromide (Spiriva) is administered once-daily, for example, from
the HandiHaler capsule-based dry powder inhaler (DPI). Tio-
tropium is classed as a long-acting muscarinic antagonist (LAMA)
or long-acting anticholinergic (LAAC).
A major area of current effort in the respiratory field is
identification of novel long-acting muscarinic antagonists that
provide inhaled once-daily bronchodilatory efficacy with poten-
tial for improved therapeutic index (TI) versus systemically
driven adverse events such as tachycardia and dry mouth.5,6 An
agent of this type would have value as either a stand-alone
therapy or for use in combination products with other once-daily
bronchodilator and/or anti-inflammatory agents.7 Reports in the
literature have described numerous chemical series that offer
promise against these objectives, although to date, none of these
compounds have received regulatory approval as new once-daily
inhaled agents.6,8 Novel quaternary ammonium agents from
these efforts include darotropium bromide (GSK233705B)9
and aclidinium bromide (LAS-34273).10,11 Marketed quaternary
ammonium oral agents reformulated for inhaled delivery include
glycopyrronium bromide (NVA-237)6,7,12 and trospium chloride
(ALKS-27).6 These four clinical agents have progressed to phase
II or phase III trials in COPD (Chart 2).
By definition, anticholinergic agents are particularly effective
because they prevent acetylcholine-induced airway smooth mus-
cle contraction, an overall effect manifested in the lungs as
bronchodilation. Cholinergic tone is therefore highly relevant
and acknowledged as the major reversible component of airflow
obstruction in COPD.3 Five distinct muscarinic acetylcholine
Received: July 6, 2011
Published: August 29, 2011
r
2011 American Chemical Society
6888
dx.doi.org/10.1021/jm200884j J. Med. Chem. 2011, 54, 6888–6904
|

Journal of Medicinal Chemistry
ARTICLE
In this paper we describe the application of an “inhalation by
design” philosophy to the discovery of a novel tertiary amine series of
muscarinic receptor antagonists that combine the pharmacology
required for once-daily inhaled administration, with pharmacokinetics
designed to minimize systemic exposure and provide potential for
improved therapeutic index versus systemically driven adverse events.
One approach to the synthesis of gem-dimethyl pyrrolidines is
shown in Scheme 2. In this case, commercially available homo-
chiral phenoxy-substituted pyrrolidines were coupled to an
activated ester of 2, to give amides (R)- and (S)-5 and (S)-6.
The dimethyl moiety was subsequently introduced with a
modified Bouveault reaction.13 Amides (R)- and (S)-5 and (S)-6
were treated with excess zirconium tetrachloride or titanium
tetrachloride, followed by Grignard reagent, to deliver com-
pounds (R)- and (S)-7 and (S)-8 in good yield. Hydrolysis of
these nitriles with potassium hydroxide delivered amides (R)-
and (S)-9 and (S)-10. Demethylation of (S)-10 with boron
tribromide delivered phenol (S)-11 in good yield.
An alternative strategy toward the synthesis of the pyrrolidine
derivatives is shown in Scheme 3. In this approach, a late stage
pyrrolidine alcohol intermediate was prepared to allow rapid
analogue generation. Both (R)- and (S)-enantiomers of amide 12
could be prepared by coupling (R)- or (S)-3-benzyloxypyrroli-
dine to acid 2 under standard conditions. Modified Bouveault
reaction followed by nitrile hydrolysis gave amides (R)- and
(S)-13 in good yield. Subsequent benzyl ether deprotection
proceeded in good yield to give alcohols (R)- and (S)-14 as key
intermediates. Mitsunobu reaction was then performed with
inversion of configuration to give a broad range of phenyl ether
derivatives. In some cases, a biphenol was used in this reaction
to provide phenol test compounds directly ((S)-15, (S)-16).
In other cases, where a methyl ether protecting group was
employed, test compounds were prepared by Mitsunobu
reaction to give ethers ((R)-10, (S)-17, (R)-18) followed
by boron tribromide deprotection to give phenols ((R)-11,
(S)-19, (R)-20).
’ CHEMISTRY
Our carboxamide target compounds were prepared across
three chemical series: pyrrolidines, piperidines, and azetidines.
The pyrrolidine lead 4 was prepared by initial nucleophilic addi-
tion of diphenylacetonitrile to tert-butyl acrylate (Scheme 1).
Subsequent ester deprotection gave acid 2, a key intermediate in
the synthesis of all our target compounds. Aldehyde formation
(3), followed by reductive amination and nitrile hydrolysis,
delivered amide 4 in good overall yield.
Chart 1
Chart 2
Azetidine 25 and piperidine 26 were prepared using a similar
approach to that described in Scheme 2. Thus, commercially
available phenoxy-substituted azetidine 21 and piperidine 22
were coupled to acid 2 to give amides 23 and 24 in good yield
(Scheme 4). Subsequent Bouveault reaction and nitrile hydro-
lysis yielded compounds 25 and 26.
Phenol-substituted piperidine 32 was prepared according to
the route shown in Scheme 5. Boc-protected piperidine alcohol
27 was alkylated with 3-bromobenzyl bromide to give 28, which
was deprotected under acidic conditions to give amine 29 in
good overall yield. Derivatization to gem-dimethyl intermediate
30 was accomplished under the amide coupling, Bouveault, and
nitrile hydrolysis conditions already described. The benzyl ether
of 30 was removed via hydrogenation to give alcohol 31, followed
by Mitsunobu reaction with resorcinol to give phenol 32.
The ortho-substituted azetidine 35 was prepared from acid 2
and 3-(methanesulfonyl)azetidine as shown in Scheme 6. Initial
coupling via the acid chloride gave mesylate 33 that was used to
Scheme 1. Synthesis of Pyrrolidine 4a
a Reagents: (a) tButyl acrylate, KOH, MeOH, tBuOH, 50 ꢀC, 5 h, 90%; (b) 4 M HCl in 1,4-dioxane, RT, 18 h, 77%; (c) (i) CDI, THF, RT, 18 h;
(ii) NaBH4, RT, 18 h, 40%; (d) DessꢀMartin periodinane, CH2Cl2, RT, 3 h, 90%; (e) (S)-3-phenoxypyrrolidine, Na(OAc)3BH, AcOH, Et3N, CH2Cl2,
RT, 5 h, 47%; (f) KOH, 3-methyl-3-pentanol, reflux, 18 h, 62%.
6889
dx.doi.org/10.1021/jm200884j |J. Med. Chem. 2011, 54, 6888–6904

Journal of Medicinal Chemistry
ARTICLE
Scheme 2. Synthesis of Pyrrolidines (R)- and (S)-9 and (S)-11a
a Reagents: (a) (S)-3-phenoxypyrrolidine or (R)-3-phenoxypyrrolidine or (S)-3-(3-methoxyphenoxy)pyrrolidine, 1-hydroxybenzotriazole, 1-(3-
dimethylaminopropyl)-3-ethyl carbodiimide, Et3N, CH2Cl2, RT, 18 h, 41ꢀ100%; (b) TiCl4, THF, ꢀ10 ꢀC, 0.5 h; MeMgBr, ꢀ5 ꢀCꢀRT, 18 h, 54%, or
ZrCl4, THF, ꢀ20 ꢀC, 1 h; MeMgCl, ꢀ10 ꢀC, 2 h, 59ꢀ78%; (c) KOH, 3-methyl-3-pentanol, reflux, 18 h, 62ꢀ96%; (d) BBr3, CH2Cl2, 0 ꢀC, 0.5 h, 60%.
Scheme 3. Synthesis of Pyrrolidines (R)-11, (S)-15, (S)-16, (S)-19, and (R)-20a
a Reagents: (a) (R)-3-benzyloxypyrrolidine, CDI, THF, RT, 18 h, 83%, or (S)-3-benzyloxypyrrolidine, 1-hydroxybenzotriazole, 1-(3-dimethy-
laminopropyl)-3-ethyl carbodiimide, Et3N, CH2Cl2, RT, 18 h, 70%; (b) (i) ZrCl4, THF, ꢀ20 ꢀC, 1 h; MeMgCl, ꢀ10 ꢀC, 2 h, 76% (ii) KOH,
3-methyl-3-pentanol, reflux, 24 h, 78ꢀ89%; (c) FeCl3, CH2Cl2, RT, 3 h, 60%, or 20% Pd(OH)2/C, 1 M HCl, EtOH, H2, 50 psi, 50 ꢀC, 28 h, 90%;
(d) ArOH, PPh3, DIAD, THF, 0 ꢀCꢀRT, 2ꢀ22 h, 75ꢀ7%; (e) BBr3, CH2Cl2, RT, 18 h, 46ꢀ68%.
Scheme 4. Synthesis of Azetidine 25 and Piperidine 26a
a Reagents: (a) 2, 1-hydroxybenzotriazole, 1-(3-dimethylaminopropyl)-3-ethyl carbodiimide, Et3N, DMF, RT, 18 h, 75ꢀ95%; (b) (i) ZrCl4, THF,
ꢀ20 ꢀC, 1 h; (ii) MeMgCl, ꢀ10 ꢀC, 2 h, 38ꢀ41%; (c) KOH, 3-methyl-3-pentanol, reflux, 18 h, 88ꢀ93%.
alkylate 2-(benzyloxy)phenol. Subsequent Bouveault reaction
proceeded in poor yield in this case, but still afforded gem-
dimethyl azetidine 34. Nitrile hydrolysis and hydrogenation of
the benzyl protecting group afforded phenol 35.
The low yield of the modified Bouveault reaction led us to seek
an alternative method for the introduction of the gem-dimethyl
group in the azetidine series. Synthesis began with Boc-protected
oxathiazinane dioxide 36 (Scheme 7).14 Ring-opening of protected
6890
dx.doi.org/10.1021/jm200884j |J. Med. Chem. 2011, 54, 6888–6904

Journal of Medicinal Chemistry
ARTICLE
Scheme 5. Synthesis of Piperidine 32a
a Reagents: (a) NaH, 3-bromobenzyl bromide, THF, 0 ꢀCꢀRT, 18 h, 61%; (b) 4 M HCl in 1,4-dioxane, RT, 2.5 h, 98%; (c) 2, 1-hydroxybenzotriazole, 1-(3-
dimethylaminopropyl)-3-ethyl carbodiimide, Et3N, DMF, RT, 18 h, 95%;(d) (i) ZrCl4, THF, ꢀ20 ꢀC, 1 h; (ii) MeMgCl, ꢀ10 ꢀC, 2 h, 33%; (e) KOH, 3-methyl-
3-pentanol, reflux, 18 h, 99%; (f) 20% Pd(OH)2/C, 1 M HCl, EtOH, H2, 50 psi, 50 ꢀC, 18 h, 95%; (g) resorcinol, PPh3, DIAD, THF, 0 ꢀCꢀRT, 12 h, 14%.
Scheme 6. Synthesis of Azetidine 35a
a Reagents: (a) (i) oxalyl chloride, DMF, CH2Cl2, 2 h; (ii) 3-(methanesulfonyl)azetidine, Et3N, CH2Cl2, ꢀ78 ꢀC, 1 h, 97%; (b) (i) 2-(benzyloxy)
phenol, Cs2CO3, DMF, 80 ꢀC, 18 h, 77%; (ii) ZrCl4, THF, ꢀ35 ꢀC, 1 h; (iii) MeMgCl, ꢀ20 ꢀC, 1 h, 10%; (c) (i) KOH, 3-methyl-3-
pentanol, reflux, 20 h, 55%; (ii) NH4HCO2, 20% Pd(OH)2/C, EtOH, reflux, 4 h, 100%.
Scheme 7. Synthesis of Azetidines 47ꢀ53 and 55a
a Reagents: (a) (i) diphenylacetonitrile, KOtBu, THF, RT, 4 h; (ii) 4 M HCl in dioxane, 50 ꢀC, 2 h, 87%; (b) epichlorohydrin, EtOH, 70 ꢀC, 18 h, 79%;
(c) MsCl, pyridine, ꢀ15ꢀ5 ꢀC, 3 h, 100%; (d) ArOH, Cs2CO3, DMF, 80 ꢀC, 18 h, 68ꢀ90%; (e) KOH, 3-methyl-3-pentanol, reflux, 18 h, 72ꢀ96%;
(f) BBr3, CH2Cl2, 0 ꢀC, 2 h, 25ꢀ100%, or 4 M HCl in dioxane, 60 ꢀC, 0.5 h, 61%; (g) 1-allyloxy-3-bromomethylbenzene, NaH, DMF, 0 ꢀC, 1 h, 54%;
(h) 4 M HCl in dioxane, 85 ꢀC, 0.5 h, 70%.
6891
dx.doi.org/10.1021/jm200884j |J. Med. Chem. 2011, 54, 6888–6904

Journal of Medicinal Chemistry
ARTICLE
Scheme 8. Synthesis of Azetidine 64a
a Reagents: (a) (i) HCOOH, 0 ꢀC, 0.5 h; (ii) 2-cyclopentyl ethanol, pyridine, CH2Cl2, < 7 ꢀC, 2 h, 95%; (b) rhodium acetate dimer, magnesium
oxide, iodobenzene diacetate, CH2Cl2, 18 h, 75%; (c) di-tert-butyl dicarbonate, Et3N, 4-dimethylaminopyridine, CH2Cl2, 3 h, 46%; (d) (i)
diphenylacetonitrile, KOtBu, DMF, 18 h; (ii) 4 M HCl in dioxane, 40 ꢀC, 2 h, 67%; (e) epichlorohydrin, MeOH, 60 ꢀC, 48 h, 52%; (f) MsCl,
pyridine, ꢀ15 ꢀC, 2 h, 59%; (g) 3-methoxyphenol, Cs2CO3, DMF, 80 ꢀC, 18 h, 73%; (h) KOH, 3-methyl-3-pentanol, reflux, 20 h, 48%; (i) BBr3,
CH2Cl2, 0 ꢀC, 2 h, 53%.
sulfamate esters similar to 36 has been reported using a broad
range of nucleophiles including amines, acetates, azides,15 and
carbon nucleophiles such as malonate derivates and aryl-
substituted enolates.16 Gratifyingly, deprotonation of diphe-
nylacetonitrile under basic conditions, followed by nucleophi-
lic attack on sulfamate 36, gave ring-opened intermediate 37,
after deprotection of the Boc group. Reaction of amine 37 with
racemic epichlorohydrin delivered azetidine 38 in moderate
yield, which could be converted to mesylate 39 under standard
conditions. Mesylate 39 was a key intermediate that furnished a
broad range of azetidine derivatives. Thus, reaction of 39 with a
panel of phenols, suitably protected where appropriate, gave
compounds 40ꢀ46 in high yield. Hydrolysis of the nitrile and
removal of any protecting groups gave carboxamide derivatives
47ꢀ53. Alternatively, alkylation of alcohol 38 with 1-allyloxy-
3-bromomethylbenzene under basic conditions led to a ben-
zylated intermediate that gave amide 54 after treatment with
potassium hydroxide. Interestingly, potassium hydroxide led to
isomerization of the allyl protecting group to give an enol ether
that could be cleaved readily under acid conditions to give
phenol 55. In general, the protected phenol derivatives em-
ployed in Scheme 7 were commercially available; however,
when synthesis was required, they could be prepared by
monoprotection of the corresponding biphenol under stan-
dard conditions.17
The route to cyclopentyl derivative 64 is shown in Scheme 8.
The synthesis began with conversion of 2-cyclopentyl ethanol to
sulfamate ester 56 using chlorosulfonyl isocyanate and formic
acid. Treatment of 56 with rhodium acetate in the presence of
magnesium oxide and an oxidant resulted in CꢀH insertion to
produce oxathiazine dioxide 57.15 Boc protection proceeded
under standard conditions to deliver compound 58 that was
subsequently converted to phenol 64 via intermediates 59ꢀ63,
using similar conditions to those described for the corresponding
gem-dimethyl derivative 47 (Scheme 7).
’ RESULTS AND DISCUSSION
The objective of our research was to identify a long-acting
muscarinic M3 receptor antagonist with suitability for once-daily
inhaled administration and pharmacokinetics consistent with
minimal systemic exposure. We were confident that we would
be able to apply our “inhalation by design” experience to
maximize TI versus systemic muscarinic effects through careful
optimization of the pharmacokinetic properties, specifically, by
incorporation of a metabolically labile functional group. In
contrast, strategies to incorporate long duration of action
(DoA) are far less well understood. One approach reported in
the field of antimuscarinics is quaternization of the tertiary amino
functional group, for example, as applied in the design of
aclidinium bromide.10 However, this approach does not guaran-
tee long DoA, as evidenced by the short-acting nature of the
quaternary agent ipratropium bromide. Both aclidinium and
tiotropium are known to exhibit an off-rate from the muscarinic
M3 receptor slower than that of ipratropium, and this is thought
to be important in driving their long-acting effects in vivo.18
Consequently, we decided to use targeted screening to identify a
compound with a long dissociation half-life from the M3 receptor
as a strategy to secure intrinsic, sustained DoA in the lung.
Previous efforts in our laboratories had focused on the delivery
of an oral muscarinic antagonist for the treatment of urinary urge
incontinence and overactive bladder. This culminated in the
identification of the carboxamide clinical candidate darifenacin.19
At the start of our program to identify an inhaled candidate, we
rescreened a subset of legacy muscarinic antagonists from our
compound file, this time looking specifically for slow dissociation
from the M3 receptor, measured in a dilution filtration bind-
ing assay.
Under our assay conditions, both tiotropium and ipratropium
had high affinity for the M3 receptor (binding Ki = 0.075 and
0.29 nM, respectively, Table 1). The dissociation half-life of
6892
dx.doi.org/10.1021/jm200884j |J. Med. Chem. 2011, 54, 6888–6904

Journal of Medicinal Chemistry
ARTICLE
Table 1. M3 Functional Potency, M3 and M2 Binding Affinity, M3 Dissociation Half-Life, clogP and log D, and HLM Clearance
Data for Compounds 4, (S)- and (R)-9, 25, and 26
a Cell-based functional potency in CHO cells expressing recombinant human M3 receptors. Carbamoyl choline (CCh) and test compounds were
coadministered and allowed to equilibrate for 4 h. Compound inhibition of CCh stimulated calcium currents was measured. All compounds appeared to
be full antagonists in this assay. Determinations were the mean of at least two replicates. b Binding affinity determined in human recombinant M3
receptors expressed in CHO cells following incubation with [3H]-NMS and test compounds for 2 h. Determinations were the mean of at least two
replicates. c Binding affinity determined in human recombinant M2 receptors expressed in CHO cells following incubation with [3H]-NMS and test
compounds for 2 h. Determinations were the mean of at least two replicates. d Test compound was incubated with CHO cells expressing the recombinant
human M3 receptor for 2 h at 100-fold its hM3 binding Ki. 90-fold dilution with buffer containing excess [3H]-NMS followed. The dissociation half-life of
e
the compound was measured by monitoring the association rate of [3H]-NMS. Determinations were the mean of at least four replicates. Intrinsic
clearance in human liver microsomes. f ND: not determined.
tiotropium was >1440 min, whereas ipratropium dissociated
rapidly (17 min), in agreement with literature data.2 Legacy
carboxamide 4 had good binding affinity (Ki = 1.9 nM) but a
relatively rapid dissociation half-life (38 min). However, gem-
dimethyl substitution adjacent to the basic amine moiety had a
profound effect on the rate of dissociation. Compound (S)-9 had
a similar binding affinity (Ki = 1.8 nM) to 4 but greatly extended
dissociation half-life (>1440 min), similar to tiotropium. We
speculate that dimethyl substitution may restrict conformational
freedom in close proximity to the amine functional group (a key
binding motif), requiring significant reordering of the ligand and
receptor for binding to occur, resulting in reduced rates of
association and dissociation. The enantiomer of (S)-9 had a
similar, exciting profile ((R)-9), albeit slightly less potent. The
achiral azetidine (25) and piperidine (26) analogues also looked
interesting; it appeared that the gem-dimethyl effect on dissocia-
tion half-life translated across different ring systems. Azetidine 25
had excellent, tiotropium-like binding affinity (Ki = 0.11 nM) and
retained an encouraging dissociation half-life (481 min), while
the piperidine had a profile similar to that of the pyrrolidines (S)-
and (R)-9.
Both tiotropium and ipratropium are nonselective muscarinic
antagonists, binding to the M1ꢀM5 receptors with approxi-
mately equal affinity.18 Compounds representing our novel
carboxamide series were also shown to be nonselective muscari-
nic antagonists when tested against the M2 receptor (Table 1),
which we deemed to be the most relevant initial counter-screen
due to the potential for M2-mediated tachycardia. Data against
the M1, M4, and M5 receptors were not gathered at this stage. All
compounds tested were full M3 antagonists in a functional cell-
based assay where inhibition of carbamyl choline-stimulated
calcium currents was measured. The potency of the carboxamide
series in this functional assay was found to be broadly equivalent
to affinity in the filtration binding experiment (Table 1).
Compounds (S)- and (R)-9, 25, and 26 gave us confidence
that slow dissociation from the M3 receptor and potential for
6893
dx.doi.org/10.1021/jm200884j |J. Med. Chem. 2011, 54, 6888–6904

Journal of Medicinal Chemistry
ARTICLE
Table 2. M3 Functional Potency, M3 Dissociation Half-Life, clogP and log D, and in Vitro Clearance Data for Phenolic
Compounds (R)- and (S)-11, (S)-15, (S)-19, 32, 35, 47, 55, and 64
a Cell based functional potency in CHO cells expressing recombinant human M3 receptors. Carbamoyl choline (CCh) and test compounds were
coadministered and allowed to equilibrate for 4 h. Compound inhibition of CCh stimulated calcium currents was measured. All compounds appeared to
be full antagonists in this assay. Determinations were the mean of at least two replicates. b Test compound was incubated with CHO cells expressing the
recombinant human M3 receptor for 2 h at 100-fold its hM3 binding Ki. 90-fold dilution with buffer containing excess [3H]-NMS followed. The
dissociation half-life of the compound was measured by monitoring the association rate of [3H]-NMS. Determinations were the mean of at least four
replicates. c Intrinsic clearance in human liver microsomes activated with NADPH. d Intrinsic clearance in human hepatocytes. e Intrinsic clearance in
human liver microsomes treated with Brij-58 and UDPGA. f ND: not determined.
long DoA was achievable in our gem-dimethyl-substituted car-
boxamides. Attention then turned to optimization of the phar-
macokinetic properties to maximize TI. Inhaled anticholinergics
are associated with a range of side effects related to systemic
exposure including dry mouth, constipation, glaucoma, urinary
retention, and cardiovascular effects. Consequently, we aimed to
6894
dx.doi.org/10.1021/jm200884j |J. Med. Chem. 2011, 54, 6888–6904

Journal of Medicinal Chemistry
ARTICLE
minimize such systemic effects through reduced oral bioavail-
ability (of the swallowed fraction of the dose) and high clearance
of the inhaled dose.
remained high (HLM Clint > 440 (μL/min)/mg), despite the
lower log D (2.4 vs 2.9). The enantiomer (R)-11 had a very
similar profile in terms of both pharmacology and oxidative
clearance. Interestingly, the location of the phenol substituent
(ortho-, meta-, or para-position) had little impact on M3 func-
tional potency in the pyrrolidine series ((S)-15, (S)-11, (S)-19,
respectively). However, ortho- and para-substitution ((S)-15,
(S)-19) resulted in a drop in the dissociation half-life (387 and
229 min, respectively).
Quaternary muscarinic antagonists have low oral bioavailabil-
ity, largely as a consequence of reduced absorption due to their
polarity.6 Both tiotropium and ipratropium have log D < 1
(Table 1), resulting in reduced membrane permeability and
low oral absorption that drives low systemic exposure of the
swallowed fraction of the dose. In contrast, our carboxamides
leads (S)- and (R)-9, 25, and 26 possessed high clogP and
moderate log D and would therefore be expected to have good
membrane permeability and be well absorbed. Compound (S)-9
was typical, with a moderate cell permeability and no evidence of
transporter-mediated efflux (Papp(AꢀB) = 16 ꢁ 10ꢀ6 cm/s,
Papp(BꢀA) = 13 ꢁ 10ꢀ6 cm/s), measured by apical to basal flux
rate in Caco-2 cells. It is widely understood that absorption is
only one component of oral bioavailability; first pass metabolism
is also important. We therefore aimed to achieve a good TI by
designing in high hepatic clearance in addition to reduced
permeability. Encouragingly, (S)-9 and 25 were turned over
rapidly in HLM (Clint > 440 (μL/min)/mg); as expected,
tiotropium and ipratropium showed high levels of microsomal
stability due to their polarity, but are subject to rapid renal
clearance in humans.2
Both the rate and route of clearance are important in deter-
mining TI across a patient population. For example, are the
metabolites pharmacologically active? Which enzymes are re-
sponsible for drug clearance? A compound metabolized by only
cytochrome P450 (CYP) 3A4 could be particularly vulnerable to
drugꢀdrug interactions, where a patient exposed to a CYP3A4
inhibitor (e.g., ketoconazole) would suffer from increased sys-
temic exposure and increased muscarinic receptor-driven side
effects. Interpatient variability can also be caused by genetic
polymorphism, for example, CYP2D6 substrates are cleared at a
significantly lower rate in certain subsets of the population
(6ꢀ8% of Caucasians do not express CYP2D6).20 We therefore
sought to maximize clearance by targeting multiple clearance
routes with high capacity and the lowest risk of saturation or
inhibition.
Early on, we decided to target glucuronidation as a significant
route of clearance for our compounds. There is little evidence of
clinically relevant drug-related inhibition of glucuronidation so
the risk of drugꢀdrug interactions is low.21 This is due to the low
affinity of any inhibitors to the uridine-50-diphosphoglucurono-
syl transferase (UGT) enzymes responsible for glucuronidation
and the potential for multiple UGTs to effect the glucuronidation
process. Glucuronidation results in the conjugation of glucuronic
acid from UDPGA (uridine-50-diphosphoglucuronic acid) to a
reactive functional group in the substrate. In this way, a lipophilic
drug molecule is derivatized to a polar, water-soluble glucuronide
which can be cleared by biliary or renal elimination. The chance
of a pharmacologically active metabolite is therefore low. Phenols
and carboxylic acids are the main substrates for UGTs, but our
experience in the design of β2-adrenoreceptor agonists led us to
favor glucuronidation of phenols as our phase II clearance
route.22
We were encouraged that our phenols were still cleared rapidly
by oxidative metabolism but wanted to establish whether they
were also glucuronidation substrates. One method frequently
used to assess clearance via glucuronidation is the measurement
of turnover of a compound in human hepatocytes. However,
hepatocytes support both phase I and phase II metabolism,
making it more difficult to determine the degree of non-P450-
mediated clearance for compounds with rapid turnover in the
standard HLM assay. Therefore, we also employed a detergent-
treated (Brij-58) HLM preparation where UDPGA was added as
a glucuronic acid donor to assess specific UGT metabolic
activity.23 In contrast to our standard HLM assay, no NADPH
was provided in the preparation, so phase I metabolism was not
supported. Brij-58 was added to allow access to the active site of
the UGTs, as the metabolizing enzymes lie on the luminal side of
the endoplasmic reticulum. The SAR for clearance via glucur-
onidation (Gluc Clint) is shown in Table 2. Once again, the
enantiomeric pair (R)- and (S)-11 had similar profiles, both
demonstrating moderate clearance (52 and 82 (μL/min)/mg,
respectively). In contrast, o- and p-phenols (S)-15 and (S)-19 did
not undergo significant turnover in this assay (13 and <4 (μL/
min)/mg, respectively), indicating that glucuronidation would
be a minor clearance process for these compounds. The clear-
ance in human hepatocytes revealed the same trend, with (S)-15
showing a turnover (30 (μL/min)/million) lower than that for
(R)- and (S)-11 (67 and 65 (μL/min)/million, respectively).
The m-phenol-substituted piperidine 32 exhibited signifi-
cantly reduced M3 potency (Ki = 2.3 nM). In contrast, the
azetidine core showed excellent potency and dissociation half-life
with the meta-substituted phenol 47 and the benzyl analogue 55
having subnanomolar potency (Ki = 0.20 and 0.18 nM, re-
spectively). Phenol substitution was not always tolerated by
the receptor; for example, ortho-substituted azetidine 35 lost
significant (55-fold) activity relative to its meta-substituted
analogue (47). Interestingly, cyclization of the gem-dimethyl
moiety into a cyclopentyl ring gave another potent derivative
64 (Ki = 0.73 nM) that also retained a long dissociation half-life
(>1440 min). While all of the azetidine derivatives showed high
levels of oxidative clearance, it was the meta-substituted phenyl
ethers (47, 64) that again showed the most encouraging turnover
in the glucuronidation assay. Compound 47 had an excellent
profile in all the in vitro assays. It possessed high affinity for the
M3 receptor (Ki = 0.20 nM), a long dissociation half-life (>1440
min), similar to tiotropium, rapid clearance in human liver
microsomes (Clint > 282 (μL/min)/mg) and human hepatocytes
(Clint = 96 (μL/min)/million), and very high turnover in our
glucuronidation assay (Clint = 221 μL/min/mg).
Phenolic substitution appeared to be well tolerated in some of
our initial examples (Table 2). For example phenol (S)-11
displayed similar, excellent M3 functional potency (Ki = 0.21
nM) to its nonphenolic counterpart (S)-9 (Ki = 0.25 nM).
Importantly, the dissociation half-life of (S)-11 (>1440 min)
also appeared to be unaffected and the phase I clearance rate
The human hepatocyte assay could be used to rank com-
pounds for phase II clearance to some degree, but the glucur-
onidation HLM assay clearly showed that SAR existed for
glucuronidation. It was found that certain phenol derivatives
were not significantly turned over by the UGTs. In some cases
this might be explained by steric hindrance, for example, as with
6895
dx.doi.org/10.1021/jm200884j |J. Med. Chem. 2011, 54, 6888–6904

Journal of Medicinal Chemistry
ARTICLE
Table 3. The Effect of Halogen Substitution on M3 Functional Potency, M3 Dissociation Half-Life, clogP and log D, in Vitro
Clearance, and Calculated Acidic pKa of Phenolic Compounds (S)-16, (R)-20, and 48ꢀ53
a Cell-based functional potency in CHO cells expressing recombinant human M3 receptors. Carbamoyl choline (CCh) and test compounds were
coadministered and allowed to equilibrate for 4 h. Compound inhibition of CCh-stimulated calcium currents was measured. All compounds appeared to
be full antagonists in this assay. Determinations were the mean of at least two replicates. b Test compound was incubated with CHO cells expressing the
recombinant human M3 receptor for 2 h at 100-fold its hM3 binding Ki. 90-fold dilution with buffer containing excess [3H]-NMS followed. The
dissociation half-life of the compound was measured by monitoring the association rate of [3H]-NMS. Determinations were the mean of at least four
replicates. c Intrinsic clearance in human liver microsomes activated with NADPH. d Intrinsic clearance in human hepatocytes. e Intrinsic clearance in
human liver microsomes treated with Brij-58 and UDPGA. f Most acidic pKa value calculated at 25 ꢀC and zero ionic strength in aqueous solutions, using
ACD/Physchem Batch 9.03 software from Advanced Chemistry Development, Inc. g ND: not determined.
the low clearance of o-phenol (S)-15. In other cases, it appeared
to be related to structure more broadly. For example, there was a
general trend for the azetidine series to exhibit higher clearance
via glucuronidation than the pyrrolidine series (e.g., 47 vs (R)-11).
The low clearance of benzyl derivative 55 versus phenyl
analogue 47 seemed to indicate that even relatively lipophilic
and sterically accessible phenols could be quite stable with
respect to glucuronidation. UGTs are reported to turn over a
wide array of chemical structures as a result of their important role
in detoxification of endogenous compounds and xenobiotics.24
Our data showed that structural features could be used to tune
the rate of clearance up or down. Such information could also be
illuminating in drug discovery programs where low clearance
compounds are desired.
We were keen to investigate further structural features that
could affect the rate of turnover by UGTs. Structural studies have
indicated that the transfer of glucuronic acid to the phenol
functional group is a base-catalyzed process where conserved
6896
dx.doi.org/10.1021/jm200884j |J. Med. Chem. 2011, 54, 6888–6904

Journal of Medicinal Chemistry
ARTICLE
Table 4. Additional in Vitro and in Vivo Characterization of Selected Compounds (R)- and (S)-11, 47, and 52
guinea pig trachea
b
M3 binding Kia
M3 dissociation t1/2
human bronchus
guinea pig in vivo
e
compd
(R)-11
(nM)
(min)
IC50c (nM)
T25d (h)
pKb
IC50f (μg)
0.201
0.180
0.128
0.163
0.0098
0.213
>1440
>1440
>1440
>1440
>1370
11.8
3.87 (n = 3)
4.30 (n = 4)
1.53 (n = 3)
3.48 (n = 3)
0.41 (n = 16)
NDg
>16 (n = 3)
>16 (n = 2)
>16 (n = 3)
>16 (n = 2)
>12.9 (n = 4)
1.47 (n = 2)
8.0 (n = 1)
9.3 (n = 2)
9.6 (n = 3)
8.8 (n = 2)
10.2 (n = 3)
NDg
20 (n = 3)
10 (n = 3)
3 (n = 6)
4 (n = 4)
0.4 (n = 10)
NDg
(S)-11
47
52
tiotropium
ipratropium
a Binding affinity at human recombinant M3 receptors expressed in CHO cells following incubation with [3H]-NMS and test compound for 24 h.
Determinations were the mean of at least six replicates. b Test compound was incubated with CHO cells expressing the recombinant human M3 receptor
for 24 h at 100-fold its hM3 binding Ki. 90-fold dilution with buffer containing excess [3H]-NMS followed. The dissociation half-life of the compound was
c
measured by monitoring the association rate of [3H]-NMS. Determinations were the mean of at least six replicates. Potency in guinea pig trachea
following incubation with test compound for 8 h. d Time taken for the EFS contractile response to recover by 25% of the inhibition induced by an Emax
concentration of the test compound incubated for 2 h. e Potency in isolated human bronchial ring following incubation with test compound for 4 h.
f Potency in an anaesthetized guinea pig model of ACh (iv)-induced bronchoconstriction, following cumulative intratracheal administration of test
compound at hourly intervals, as a solution in micelle. ACh (iv) challenge was administered hourly (prior to the next escalating dose of compound), and
the maximum increase in lung resistance recorded over a 5 min period, in order to produce a five-point doseꢀresponse curve. g ND: not determined.
carboxylate and histidine residues facilitate the deprotonation of
the phenol, enabling nucleophilic attack on UDPGA.24 We were
therefore keen to investigate whether the pKa of the phenol could
be a factor in determining the rate of clearance. As meta-
substituted phenols had shown the best balance of potency,
dissociation half-life, and clearance (Table 2), the properties
of halogen-substituted m-phenols were therefore explored
(Table 3). In the azetidine core, the M3 potency of chlorophenol
48 was reduced 10-fold relative to 47, although there was little
change in dissociation half-life. The additional chlorine atom
resulted in increased clogP, and HLM and hepatocyte clearance
remained high. In addition, the clearance by glucuronidation
increased to 430 (μL/min)/mg, possibly as a result of increased
clogP, but reduced pKa could also be a factor. The calculated pKa
of the phenol of 48 was significantly lower than that of 47 (7.9 vs
9.5). The addition of a m-chloro substituent (49) also resulted in
the retention of M3 potency (Ki = 0.22 nM), high oxidative
clearance (HLM Clint = 361 (μL/min)/mg), and rapid glucur-
onidation (Gluc Clint = 223 (μL/min)/mg). The corresponding
fluorophenol 50 had a similar profile, again demonstrating very
rapid glucuronidation (Clint = 403 (μL/min)/mg). Addition of a
second halogen substituent (51) resulted in retention of the high
clearance properties of the lead azetidine (47) but significant loss
of potency (37-fold) and dissociation half-life (165 min). Fluoro
substitution ortho to the phenol (52) resulted in a slightly
reduced rate of phase II clearance (Gluc Clint = 52 (μL/min)/
mg), but excellent potency (Ki = 0.26 nM) was retained.
Fluorophenol isomer 53 had a similar profile to its parent
azetidine (47), with good potency against the M3 receptor, long
dissociation half-life, and rapid phase I and phase II clearance.
In the pyrrolidine core, addition of a chloro substituent ((S)-
16) resulted in reduced potency (10-fold) relative to phenol (S)-
11 but did provide an approximately 3-fold increase in the rate of
glucuronidation (Clint = 146 (μL/min)/mg). This effect may
also be due to a combination of increased clogP and reduced pKa.
The m-fluoro derivative (R)-20 possessed good potency (Ki =
0.56 nM), a very long dissociation half-life (>1440 min), rapid
clearance in HLM, and high hepatocyte clearance that indicated
a high level of turnover by glucuronidation. To conclude, we
were able to affect the rate of clearance of compounds in our
glucuronidation assay by modification of the substituents on the
phenol ring.
A selection of promising compounds ((R)- and (S)-11, 47,
52), based on a combination of pharmacology and material
properties, was advanced to more comprehensive evaluation and
comparison against tiotropium and ipratropium (Table 4). Fol-
lowing observations that these agents dissociated very slowly
from the M3 receptor (slow off-rate, small koff), there was an
accompanying possibility that they would demonstrate slow
association (slow on-rate, small kon), a rate that is a function of
both ligand concentration (dose) and time.25 This property has
the potential to extend incubation times and concentrations
required for an assay to reach ligandꢀreceptor equilibrium and
accurately define the equilibrium dissociation constant (Kd). In
response to these observations, M3 binding affinity (Ki) was
determined for the selected compounds in revised assay condi-
tions requiring incubation with test compound for 24 h. Slightly
increased potency (up to 12-fold) was observed for (R)- and
(S)-11, 47, 52, and tiotropium, illustrating the importance of
incubation time when screening compounds that possess slow
kinetics of binding. As expected, ipratropium potency under
these conditions was unaffected due to its rapid binding kinetics.
M3 dissociation half-lives (t1/2) determined under this revised
incubation time (24 h) confirmed that all these agents (except
ipratropium) exhibited very long residence time and potential for
extended pharmacological effect.2 The rapid dissociation half-life
observed for ipratropium is in agreement with literature data.26
The effect of these agents on airway smooth muscle was
investigated in vitro using appropriate compound incubation
times in isolated guinea pig tracheal (GPT) strips27 and human
bronchial rings. Potency within 10-fold of tiotropium was
observed for (R)- and (S)-11, 47, and 52 in GPT, along with
very long-lasting DoA (T25 >16 h); remarkably, the time taken
for the contractile response to recover by 25% following inhibi-
tion with compounds (R)- and (S)-11, 47, and 52 was in excess of
16 h, indicating DoA that was at least tiotropium-like (T25 >12.9 h).
Ipratropium was confirmed to be short-acting (T25 = 1.47 h).
Human bronchus data indicated that 47 (pKb = 9.6) was the
most potent of these analogues and more importantly was also
within 10-fold of tiotropium (pKb = 10.2). This trend was also
6897
dx.doi.org/10.1021/jm200884j |J. Med. Chem. 2011, 54, 6888–6904

Journal of Medicinal Chemistry
ARTICLE
Figure 1. Lung resistance data in anaesthetized guinea pigs for (R)- and
(S)-11, 47, 52, and tiotropium bromide, following cumulative intra-
tracheal administration at hourly intervals, as solutions in micelle vehicle.
ACh (iv) challenge was administered hourly (prior to the next escalating
dose of compound) and the maximum increase in lung resistance
recorded over a 5 min period. Data are expressed as a percentage of
the pretreatment ACh (iv) control response (100%).
observed in vivo following lung resistance studies in anaesthe-
tized guinea pigs (Figure 1 and Table 4). Overall, 47 was
consistently the most attractive agent from these studies, posses-
sing tiotropium-like DoA and potency within approximately
10-fold.
Compound 47 was subsequently profiled across human
M1ꢀM5 receptors to build a comprehensive picture of muscari-
nic pharmacology (Table 5). Binding affinity (Ki) determined
with 24 h compound incubation demonstrated that 47 had
balanced, potent activity across all M1ꢀM5 receptors, similar
to ipratropium and tiotropium. Dissociation half-life determined
with 24 h compound incubation indicated that 47 possessed
tiotropium-like kinetic selectivity for M3 vs M2 receptors.2 This
kinetic selectivity could be receptor rather than compound
driven, as ipratropium has also been reported to show this kinetic
behavior, albeit over a much shorter, less clinically relevant time
frame.26 Furthermore, with the exception of M3 receptor dis-
sociation half-life previously discussed, it appeared that 47 had a
dissociation half-life shorter than that for tiotropium across
M1ꢀM5 receptors. Finally, the binding kinetics of 47 were
investigated according the method of Motulsky and Mahan.28
Interestingly, the off-rate (koff) for 47 in this experiment was
approximately 1 order of magnitude slower than that for
tiotropium, confirming a very slow dissociation half-life (t1/2
=
4998 min). The on-rate (kon) for 47 was also slower than that for
tiotropium and appeared consistent with previous in vitro
observations. The resulting binding affinity for 47 (Ki = 0.0132
nM) was within 3-fold of that for tiotropium (Ki = 0.0057 nM),
both of which were very potent in this assay and more potent
than ipratropium, which was confirmed to have fast on-rate and
off-rate.
With encouraging pharmacology in hand, we profiled 47 in a
conscious dog model of bronchoconstriction to assess in vivo
efficacy, DoA, and cardiovascular (CV) side effects (e.g., heart
rate).29 Dogs were treated with intratracheal (it.) solution
administration of either M3 antagonist or vehicle (micelle).
Methacholine (MCh) challenge was administered it. to induce
6898
dx.doi.org/10.1021/jm200884j |J. Med. Chem. 2011, 54, 6888–6904

Journal of Medicinal Chemistry
ARTICLE
Figure 2. Lung resistance data in conscious dogs for 47, ipratropium bromide, and tiotropium bromide, following intratracheal administration of a
single dose, as a solution in micelle vehicle. MCh (it.) challenge was administered at 0.5 h, 1 h, and subsequently 2 h intervals, for up to 32 h, to measure
the maximum increase in lung resistance, expressed as a percentage of the pretreatment MCh (it.) control response (100%).
Figure 3. Heart rate data in conscious dogs for 47, ipratropium bromide and tiotropium bromide following intratracheal administration of a single dose,
as a solution in micelle vehicle. MCh (it.) challenge was administered at 0.5 h, 1 h and subsequently 2 h intervals, for up to 32 h, to measure CV
parameters relative to the pretreatment MCh (it.) control response measured in that animal at the start of the study. Data are presented as heart rate, in
beats-per-minute (bpm).
Table 6. In Vitro Permeability and in Vivo Pharmacokinetic Data for Compound 47
in vitro permeability, MDCK-MDR1 Papp x 10ꢀ6 (cm/s)
rat in vivo PK, iv
ClT ((mL/min)/kg)
134
compd
AꢀB
BꢀA
Vd (L/kg)
t1/2 (h)
ppb (%)
93.5
47
7
64
8.4
0.9
an increase in lung resistance (RL). RL and CV parameters with
antagonist treatment were compared to the control MCh
response. In this model, 47 was found to be approximately 3-fold
a well validated in vivo model provided strong evidence that 47
would be able to deliver long-lasting (g24 h) bronchodilation in
humans from a projected once-daily inhaled dose of approxi-
mately 200 μg.
more potent (ID50 = 3.0 μg) than that for ipratropium (ID50
10.0 μg) but slightly less potent than that for tiotropium (ID50
=
<
In vitro observations illustrated that 47 met our criteria of
rapid CYP-mediated oxidative metabolism (Clint > 440 (μL/
min)/mg) and phenolic glucuronidation (Clint > 221 (μL/min)/
mg) (Table 2). Furthermore, in vitro permeability data in
MDCK-MDR1 cells revealed that while 47 displayed high
intrinsic permeability, it was also subject to very high transpor-
ter-mediated efflux (Papp(AꢀB) = 7 ꢁ 10ꢀ6 cm/s, Papp(BꢀA) =
64 ꢁ 10ꢀ6 cm/s), most likely as a substrate for P-glycoprotein
(P-gp) (Table 6). Interestingly, the phenolic moiety appeared to
be responsible for two of these observations: first, it introduced a
site for extensive glucuronidation, and second, it induced sig-
nificant P-gp efflux that was not observed for nonphenolic
analogues (e.g., 25; Papp(AꢀB) = 27 ꢁ 10ꢀ6 cm/s, Papp(BꢀA) =
28 ꢁ 10ꢀ6 cm/s). With desirable in vitro data predictive of low
oral bioavailability (due to transporter limited absorption
3.0 μg) (Figure 2). At the ID50 dose (3.0 μg), 47 displayed long-
lasting intrinsic DoA (>24 h) that was significantly superior to
ipratropium (10.0 μg). Furthermore, a 3-fold higher dose of 47
(10 μg) afforded greater efficacy and DoA that was comparable
to that for tiotropium (3.0 μg); both compounds were still
efficacious 32 h postdose. Slower onset of action was observed for
both doses of 47, when compared to that for ipratropium and
tiotropium. In vivo onset/DoA data for 47 appeared consistent
with binding on-rate/off-rate kinetics observed in vitro
(Table 5). Hemodynamic data from these studies confirmed
that, relative to the control response measured in each animal at
the start of the study, 47 (10 μg), ipratropium (10 μg), and
tiotropium (3.0 μg) had no detectable CV effects up to 32 h
postdose (Figure 3). These comparative data to clinical agents in
6899
dx.doi.org/10.1021/jm200884j |J. Med. Chem. 2011, 54, 6888–6904

Journal of Medicinal Chemistry
ARTICLE
Chart 3
through the gut lumen and high first pass metabolism), 47
was advanced to rat in vivo pharmacokinetic (PK) studies
(Table 6).30 Following iv administration, 47 exhibited rapid and
extensive distribution (Vdss = 8.4 L/kg), very high unbound and
total clearance (Clu = 2065 (mL/min)/kg, ClT = 134 (mL/
min)/kg) that exceeded liver blood flow, and a short terminal
half-life (t1/2 = 0.9 h). Oral administration to rats resulted in no
detectable systemic exposure, consistent with predictions from in
vitro data.
The metabolites of 47 were found to be similar across a range
of in vitro incubations in rat, dog, and human hepatocytes and
liver microsomes. The major phase I metabolite from micro-
somal preparations was the azetidinol, resulting from O-dearyla-
tion. This metabolite was found to possess >10-fold weaker M3
binding affinity (Ki = 1.64 nM) and much shorter M3 dissociation
half-life (t1/2 = 69 min) than that for 47, when assessed using
these assays in their 24 h incubation format. In contrast, the
major metabolite in hepatocytes was a phase II glucuronide, most
likely the O-glucuronide from conjugation of the phenol group.
Analysis of rat PK plasma samples showed that neither of these
metabolites circulated in vivo to any meaningful extent. The
primary enzyme-mediating phase I oxidative metabolism of 47
was found to be CYP3A4. Coincubation of human hepatocytes
with the CYP3A4 inhibitor ketoconazole led to only a minor
reduction (30%) in metabolism of 47, again confirming extensive
clearance of 47 by glucuronidation.
Overall, 47 exhibited an ideal pharmacokinetic profile for an
inhaled agent by virtue of incomplete oral absorption (of the
swallowed fraction) and high unbound clearance predicted in
humans via oxidative metabolism and glucuronidation. Systemic
exposure and associated adverse events in humans were therefore
predicted to be minimal.30 Furthermore, clinically relevant
drugꢀdrug interactions (DDI) were predicted to be minimal
as a consequence of multiple routes of metabolism.
Subsequently, the hydrochloride salt form of 47 was isolated
from ethanol and found to possess material properties that were
consistent with our solid form requirements.31 Analysis using
well-known techniques such as PXRD, DSC, TGA, and DVS
found 47 to be highly crystalline, anhydrous and nonhygroscopic
(0.1% water uptake at 90% relative humidity) with high melting
point (223 ꢀC). The solid-state stability of 47 with lactose
monohydrate (the most common carrier excipient used in
commercial DPI) was demonstrated following storage of materi-
al blended at a 1:100 ratio for 28 days at 40 ꢀC and 75% relative
humidity. These material properties confirmed that 47 was
compatible with inhaled administration from a DPI.
Figure 4. Phase I spirometry data in healthy human volunteers for 47,
following inhalation of dry-powder formulations. Data are expressed as
the mean change in FEV1 from baseline, by treatment.
cough model following it. solution administration of doses up to
2000 μg.32 These data provided preclinical confidence that 47
was unlikely to elicit cough in humans, thus addressing a key
attrition risk for inhaled compound development and endorsing
compound progression. Overall, 47 had a preclinical safety
profile that suggested it should be well tolerated in humans at
all clinically relevant inhaled doses.
’ CONCLUSION
In conclusion, we have described our “inhalation by design”
strategy to deliver a novel muscarinic M3 receptor antagonist
(47) that has potential for once-daily inhaled administration in
humans for the treatment of COPD. Geminal dimethyl function-
ality confers very long dissociative half-life (slow off-rate) from
the M3 receptor that mediates very long-lasting bronchodilation
in vivo (g24 h). Incorporation of the phenol moiety delivers
extensive glucuronidation “by design” and negligible oral absorp-
tion, features that are complemented by rapid oxidative metabo-
lism to provide 47 with an optimal preclinical PK profile for an
inhaled agent. Material properties and projected once-daily
inhaled dose size (∼200 μg) are fully compatible with adminis-
tration from a DPI delivery device. Furthermore, the preclinical
safety profile of 47 suggests that it will be well tolerated in
humans at all clinically relevant doses.
Successful delivery of this compelling profile ultimately
identified 47 (PF-3635659) as a phase II clinical candidate
(Chart 3). Spirometry data from healthy volunteers in phase I
clinical studies illustrate that 47 provides efficacious 24 h
bronchodilation (FEV1 > 100 mL) from a single inhaled dose,
thus confirming the suitability of 47 as a novel once-daily inhaled
muscarinic M3 receptor antagonist for the treatment of COPD
(Figure 4). Human pharmacokinetic data from multidose phase
I studies demonstrate very low levels of unbound systemic
exposure at all doses (e.g., day 7, 200 μg dose, Cmaxu = 0.008
nM), with no accumulation on repeat dosing, findings that are
consistent with preclinical predictions. Safety (tachycardia) and
toleration (dry mouth) data from multidose phase I studies
indicate an adverse event profile that is at least noninferior to
tiotropium bromide.
From a drug safety perspective, 47 demonstrated a safety
margin in excess of 100-fold when evaluated for broader phar-
macological activity against a wide ranging panel of in vitro assays
representing receptors, enzymes, and ion channels. In vitro
genetic toxicity testing and in vivo rat toleration studies were
also completed without observing any meaningful adverse
effects. Furthermore, 47 was inactive in a validated rabbit lung
6900
dx.doi.org/10.1021/jm200884j |J. Med. Chem. 2011, 54, 6888–6904

Journal of Medicinal Chemistry
ARTICLE
’ EXPERIMENTAL SECTION
were 100% nerve mediated and 100% cholinergic as confirmed by
blockade by 1 μM tetrodotoxin or 1 μM atropine. Tissues were then
repeatedly stimulated at 2 min intervals until the responses were
reproducible.
(i) To assess potency (IC50), the study compound was added to the
tissue baths, with each tissue receiving a single concentration of
compound and allowed to equilibrate for 8 h. At 8 h postaddition,
the inhibition of the EFS response was recorded and IC50 curves
were generated using a range of compound concentrations over
tracheal strips from the same animal.
(i) To assess duration of action, the study compound was added for
2 h, the tissues were then rapidly washed, and the 1 mL/min
perfusion with Krebs solution was re-established. Tissues were
stimulated for a further 16 h, and recovery of the EFS response
was recorded. At the end of the 16 h, 10 μM histamine was added
to the baths to confirm tissue viability. The just maximal con-
centration (tested concentration giving a response >70% inhibi-
tion but less than 100%) of antagonist was identified from the
IC50 curve, and the time to 25% recovery of the induced
inhibition (T25) was calculated in tissues receiving this concen-
tration. Compounds are typically tested at n = 2ꢀ5 to estimate
duration of action.
Biology. In Vitro Potency at Recombinant Human M3
Receptors Expressed in CHO Cells. M3 potency was determined
in CHO-K1 cells transfected with the NFAT-Betalactamase gene. The
cells were resuspended at 2 ꢁ 105 cells/mL in growth medium, and
20 μL of this cell suspension was added to each well of a 384-well, black,
clear-bottomed plate. The assay buffer used was PBS, supplemented
with 0.05% Pluronic F-127 and 2.5% DMSO. Muscarinic M3 receptor
signaling was stimulated, in the presence and absence M3 receptor
antagonists, using 80 nM carbamyl choline incubated with the cells for 4
h at 37 ꢀC/5% CO2 and monitored at the end of the incubation period
using a Tecan SpectraFluor+ plate reader (λ, excitation 405 nm,
emission 450 and 503 nm). Inhibition curves were plotted and IC50
values generated using a four-parameter sigmoid fit and converted to Ki
values using the ChengꢀPrusoff correction and the KD value for
carbamyl choline in the assay.
In Vitro Binding Affinity at Recombinant Human M1ꢀM5
Receptors Expressed in CHO Cells. Affinity and selectivity at
human recombinant M1ꢀM5 receptors, separately expressed in Chinese
hamster ovary (CHO) cell membranes, were assessed using an N-methyl
[3H] scopolamine ([3H]-NMS) filter binding assay similar to that
previously described.18 Briefly, CHO cell membranes were incubated
In Vivo Potency and Duration of Action in the Conscious
Dog. Conscious male and female beagle dogs (10ꢀ15 kg) which had
previously undergone a tracheostomy and esophagotomy were im-
planted with telemetry to continuously monitor arterial blood pressure
(ABP), heart rate, myocardial contactility (change in left ventricular
pressure (LVP) with time), and ECG. Flow was measured via a Fleisch
tube and validyne attached to a trach tube placed in the tracheostomy.
Pleural pressure was measured using a balloon catheter placed in the
esophagotomy, linked to a pressure transducer. Total lung resistance and
dynamic lung compliance (derived from flow and pulmonary pressure
(PP)) were measured after each intratracheal methacholine (MCh)
challenge. Following instrumentation and a 30 min stabilization period,
nebulized MCh (1.0 mg/mL) was given to provide cardiovascular (CV)
and respiratory parameters as a control response.
(i) For the doseꢀresponse study, doses of test compound or vehicle
were given via intratracheal (it.) administration at cumulative 1 h
intervals using a Penn-Century insufflator. Intratracheal MCh
challenge was repeated 1 h posttest compound administration to
compare airway resistance and CV parameters to the control
MCh response.
3
in the presence or absence of test compounds and H-NMS at room
temperature for 2 or 24 h with shaking. The assay was terminated by
rapidly filtering through GF/B Unifilter plates which were read on an
NXT Topcount. Percent specific binding versus test compound con-
centration was plotted to determine an IC50 from a sigmoid curve using
an in-house data analysis program. IC50 values were corrected to Ki
values by applying the ChengꢀPrussoff equation.
In Vitro Assessment of Dissociation Half-Life at Recombi-
nant Human M1ꢀM5 Receptors. Duration of action at M1ꢀM5
receptors was assessed using a dilutionꢀoffset filter binding assay.33
Briefly, the receptor was maximally occupied with the compound of
interest, by incubating the membrane for 2 or 24 h with test compound
at 100 ꢁ Ki. At various time points thereafter (between 15ꢀ1440 min)
the mixture was diluted (90-fold), reducing the test compound con-
centration to 1 ꢁ Ki. This dilution was performed with assay buffer
containing [3H]-NMS, to give a final assay concentration of 12.5 ꢁ KD.
Therefore, the test compound was encouraged to dissociate from the
receptor by both dilution and competition with [3H]-NMS. Nonspecific
binding was defined by 1 μM atropine, and total binding by equivalent
DMSO (0.01%). Following 1440 min incubation, the assay was termi-
nated by rapid filtration and the bound radioactivity quantified as
described above. The association of [3H]-NMS to the receptor was
fitted to a generalized hyperbola, fixing the maximum to the maximum
specific binding for the assay plate. The offset of the test compound was
inferred from the association of [3H]-NMS, expressed as the time taken
to reach 50% of total [3H]-NMS binding, for solvent-treated mem-
branes. Offset values (t1/2) were expressed as arithmetic means with 95%
confidence intervals (CI). When the time to 50% recovery was estimated
outside of the time points measured, results were quoted as >1440 or
<15 min.
(ii) For the duration of action studies, one dose of test compound or
vehicle was administered intratracheally and MCh challenge (it.)
repeated at 0.5 h, 1 h, and subsequently 2 h intervals for up to
32 h to compare airway resistance and CV parameters to the
control MCh response.
Chemistry. Unless otherwise indicated, all reactions were carried
out under a nitrogen atmosphere, using commercially available anhy-
drous solvents. Thin-layer chromatography was performed on glass-
backed precoated Merck silica gel (60 F254) plates, and compounds
were visualized using UV light, 5% aqueous potassium permanganate, or
chloroplatinic acid/potassium iodide solution. Silica gel column chro-
matography was carried out using 40ꢀ63 μm silica gel (Merck silica gel 60).
Ion exchange chromatography was performed using Isolute strong
cation exchange resin (SCX-2) cartridges which had been prewashed
with methanol. Proton NMR spectra were measured on a Varian Inova
400 or Varian Mercury 400 spectrometer in the solvents specified. In
these NMR spectra, exchangeable protons that appeared distinct from
solvent peaks are also reported. Low resolution mass spectra were
recorded using the following system: PAL CAT sample injector; Agilent
1100 stack comprising of solvent degasser, pump and UV detector;
Micromass ZQ 2000 MS detector with a scanning range of m/z 150ꢀ900.
Test compound purity g95% was determined using combustion
In Vitro Potency and Duration of Action in Guinea Pig
Isolated Trachea. Tracheal strips (removed from male Dunkinꢀ
Hartley guinea pigs) were suspended in 5 mL of tissue baths under an
initial tension of 1 g and bathed in warmed (37 ꢀC), aerated (95% O2/
5% CO2) Krebs solution containing 3 μM indomethacin and 10 μM
guanethidine. At the end of the equilibration (1 h), tissues were
electrically field stimulated (EFS) using the following parameters:
10 V, 10 Hz 0.1 ms pulse width with 10 s trains every 2 min. In each
tissue a voltage response curve was constructed over the range 10ꢀ30 V
(keeping all other stimulation parameters constant) to determine a just
maximal stimulation. Using these stimulation parameters EFS responses
6901
dx.doi.org/10.1021/jm200884j |J. Med. Chem. 2011, 54, 6888–6904

Journal of Medicinal Chemistry
ARTICLE
analysis conducted by Exeter Analytical U.K. Ltd., Uxbridge, Middlesex,
U.K. Tiotropium bromide was synthesized according to literature
procedures.34 Ipratropium bromide was purchased from Sigma.
5-Amino-5-methyl-2,2-diphenylhexanenitrile Hydrochloride
Salt (37). A solution of diphenylacetonitrile (25.6 g, 132.3 mmol) in
tetrahydrofuran (200 mL) was treated with potassium tert-butoxide
(14.8 g, 132.3 mmol) and the resulting solution stirred for 10 min.
A solution of tert-butyl 4,4-dimethyl-1,2,3-oxathiazinane-3-carboxylate
2,2-dioxide (36)14 (35.1 g, 132.3 mmol) in tetrahydrofuran (300 mL)
was then added over 15 min. The resulting solution was stirred for 4 h
and then concentrated in vacuo. The residue was treated with 4 M
hydrochloric acid in dioxane (165 mL, 661.5 mmol) and heated at 50 ꢀC
for 2 h. The solution was concentrated in vacuo to low volume, basified
with saturated aqueous sodium carbonate solution, and extracted with
dichloromethane (150 mL). The organic layer was dried (sodium
sulfate) and concentrated in vacuo. The residue was treated with 4 M
hydrochloric acid in dioxane (50 mL), concentrated in vacuo, and tritu-
rated with diethyl ether to give 37 as a solid, 36.2 g, 87% yield. 1H NMR
(400 MHz, CD3OD) δ: 1.35 (s, 6H), 1.70ꢀ1.74 (m, 2H), 2.56ꢀ2.61
(m, 2H), 7.32ꢀ7.47 (m, 10H) ppm. LRMS: m/z 279 [M + H]+.
5-(3-Hydroxyazetidin-1-yl)-5-methyl-2,2-diphenylhexane-
nitrile (38). A solution of amine hydrochloride salt 37 (22.1 g, 70.2
mmol) in dichloromethane (250 mL) was washed with saturated
aqueous sodium carbonate solution (100 mL). The organic layer was
dried (sodium sulfate) and concentrated in vacuo. The residue was
dissolved in ethanol (250 mL) and treated with 2-chloromethyloxirane
(7.8 g, 84.2 mmol), and the resulting solution was heated at 70 ꢀC for
18 h. The mixture was cooled to room temperature and concentrated in
vacuo. The residue was crystallized from toluene/pentane to give 38 as a
solid, 18.1 g, 79% yield. 1H NMR (400 MHz CDCl3) δ: 0.92 (s, 6H),
1.31ꢀ1.36 (m, 2H), 2.42ꢀ2.46 (m, 2H), 2.88ꢀ2.97 (m, 2H),
3.29ꢀ3.37 (m, 2H), 4.31ꢀ4.38 (m, 1H), 7.27ꢀ7.43 (m, 10H) ppm.
LRMS: m/z 335 [M + H]+.
Methanesulfonic Acid 1-(4-Cyano-1,1-dimethyl-4,4-di-
phenylbutyl)azetidin-3-yl Ester (39). A solution of hydroxyazeti-
dine 38 (15.0 g, 44.9 mmol) in pyridine (150 mL) at ꢀ15 ꢀC was treated
dropwise with methanesulfonyl chloride (10.4 mL, 134.6 mmol) over 15
min. The resulting solution was stirred at 0ꢀ5 ꢀC for 3 h. The mixture
was quenched with saturated aqueous sodium hydrogen carbonate
solution (150 mL) and extracted with ethyl acetate (2 ꢁ 150 mL).
The combined organic layers were dried (sodium sulfate) and concen-
trated in vacuo. The residue was redissolved in toluene (100 mL) and the
solvent removed in vacuo to give 39 as a gum, 18.6 g, 100% yield. 1H
NMR (400 MHz, CDCl3) δ: 0.92 (s, 6H), 1.29ꢀ1.33 (m, 2H),
2.41ꢀ2.45 (m, 2H), 2.98 (s, 3H), 3.21ꢀ3.24 (m, 2H), 3.37ꢀ3.41 (m,
2H), 4.96ꢀ5.01 (m, 1H), 7.26ꢀ7.40 (m, 10H) ppm. LRMS: m/z 413
[M + H]+.
(13.4 g, 30.4 mmol) in 3-methylpentan-3-ol (150 mL) was treated with
potassium hydroxide (34.1 g, 608 mmol), and the resulting solution was
stirred at 120 ꢀC for 18 h. The mixture was cooled to room temperature
and concentrated in vacuo, and the residue was partitioned between
ethyl acetate (200 mL) and water (200 mL). The organic layer was
separated, washed with water (200 mL), dried (sodium sulfate), and
concentrated in vacuo. The residue was purified by column chromatog-
raphy on silica gel, eluting with ethyl acetate/methanol/0.88 specific
gravity ammonia (90:10:1, by volume) to give 5-[3-(3-methoxyphenoxy)-
azetidin-1-yl]-5-methyl-2,2-diphenylhexanamide as a white foam, 10.7 g,
1
77% yield. H NMR (400 MHz, CDCl3) δ: 0.92 (s, 6H), 1.12ꢀ1.23
(m, 2H), 2.40ꢀ2.50 (m, 2H), 3.10ꢀ3.25 (m, 2H), 3.44ꢀ3.58 (m, 2H),
3.78 (s, 3H), 4.62ꢀ4.72 (m, 1H), 6.30ꢀ6.38 (m, 2H), 6.47ꢀ6.54
(m, 1H), 7.10ꢀ7.18 (m, 1H), 7.22ꢀ7.45 (m, 10H) ppm. LRMS: m/z
459 [M + H]+.
A solution of 5-[3-(3-methoxyphenoxy)azetidin-1-yl]-5-methyl-2,2-
diphenylhexanamide (9.0 g, 19.6 mmol) in dichloromethane (1.25 L) at
0 ꢀC was treated dropwise with 1 M boron tribromide in dichloro-
methane (58.9 mL, 58.9 mmol). The resulting mixture was warmed to
room temperature with stirring over 2 h. The mixture was recooled to
0 ꢀC, quenched with addition of 1 M aqueous sodium hydroxide solution
(200 mL), and allowed to warm to room temperature over 1 h. The
aqueous layer was separated and extracted with ethyl acetate (2 ꢁ
200 mL). The combined organic layers were dried (sodium sulfate) and
concentrated in vacuo. The residue was purified by column chromatog-
raphy on silica gel, eluting with (dichloromethane/methanol/0.88
specific gravity ammonia (90:10:1, by volume))/pentane (50:50, by
volume) to give 47 as a white foam, 3.40 g, 39% yield. 1H NMR (400
MHz, CDCl3) δ: 1.10 (s, 6H), 1.22ꢀ1.34 (m, 2H), 2.42ꢀ2.55 (m, 2H),
3.28ꢀ3.40 (m, 2H), 3.65ꢀ3.88 (m, 2H), 4.70ꢀ4.80 (m, 1H),
5.55ꢀ5.70 (bs, 2H), 6.23ꢀ6.36 (m, 2H), 6.45ꢀ6.53 (m, 1H),
7.03ꢀ7.12 (m, 1H), 7.19ꢀ7.39 (m, 10H) ppm. LRMS: m/z 443
[M ꢀ H]ꢀ, 445 [M + H]+. Anal. (C28H32N2O3 1.0H2O) C, H, N. HPLC
3
analysis: g98.5% purity.
’ ASSOCIATED CONTENT
S
Supporting Information. Biology experimental section
b
for the in vitro human bronchus assay, in vivo anaesthetized
guinea pig model, and in vivo rabbit lung Aδ-fiber cough model;
chemistry experimental details and analytical data for intermedi-
ates 1ꢀ3, (R)- and (S)-5, (S)-6, (R)- and (S)-7, (S)-8, (R)- and
(S)-10, (R)- and (S)-12ꢀ14, (S)-17, (R)-18, 23, 24, 28ꢀ31, 33,
34, 41ꢀ46, 54, 56ꢀ63, and test compounds 4, (R)- and (S)-9,
(R)- and (S)-11, (S)-15, (S)-16, (S)-19, (R)-20, 25, 26, 32, 35,
48ꢀ53, 55, 64. This material is available free of charge via the
Internet at http://pubs.acs.org.
5-[3-(3-Methoxyphenoxy)azetidin-1-yl]-5-methyl-2,2-di-
phenylhexanenitrile (40). A solution of mesylate 39 (12.3 g, 29.8
mmol) in acetonitrile (150 mL) was treated with 3-methoxyphenol
(3.9 mL, 35.8 mmol) and potassium carbonate (10.3 g, 74.5 mmol). The
resulting mixture was stirred at 80 ꢀC for 3.5 h. The mixture was cooled
to room temperature and concentrated in vacuo, and the residue was
partitioned between ethyl acetate (250 mL) and saturated aqueous
sodium hydrogen carbonate solution (150 mL). The aqueous layer was
separated and extracted with ethyl acetate (2 ꢁ 200 mL). The combined
organic layers were washed with brine (300 mL), dried (sodium sulfate),
and concentrated in vacuo to give 40 as a brown oil, 16.2 g, 69% yield. 1H
NMR (400 MHz, CDCl3) δ: 0.95 (s, 6H), 1.31ꢀ1.44 (m, 2H),
2.41ꢀ2.56 (m, 2H), 3.07ꢀ3.24 (m, 2H), 3.42ꢀ3.54 (m, 2H), 3.77 (s,
3H), 4.63ꢀ4.74 (m, 1H), 6.28ꢀ6.38 (m, 2H), 6.48ꢀ6.55 (m, 1H),
7.26ꢀ7.49 (m, 11H) ppm. LRMS: m/z 441 [M + H]+.
’ AUTHOR INFORMATION
Corresponding Author
*(P.G.) Phone: +441304812980; e-mail: paglossop@gmail.com.
(C.W.) Phone: +441223847005; e-mail: christine.watson100@
gmail.com.
Present Addresses
#Pfizer Global Research and Development, Eastern Point Rd.,
Groton, CT 06340.
fPfizer, 2002 Executive Suite, 200 Cambridge Park Dr., Cambridge,
MA 02140.
ΔPeakdale Chemistry Services, Ramsgate Rd, Sandwich, Kent
CT13 9NJ, U.K.
MedImmune, Milstein Building, Granta Park, Cambridge, CB21
6GH, U.K.
5-[3-(3-Hydroxyphenoxy)azetidin-1-yl]-5-methyl-2,2-dip-
henylhexanamide (47, PF-3635659). A solution of nitrile 40
6902
dx.doi.org/10.1021/jm200884j |J. Med. Chem. 2011, 54, 6888–6904

Journal of Medicinal Chemistry
ARTICLE
’ ACKNOWLEDGMENT
19, 4560–4562. (b) Lainꢀe, D. I.; Wan, Z.; Yan, H.; Zhu, C.; Xie, H.; Fu,
W.; Busch-Petersen, J.; Neipp, C.; Davis, R.; Widdowson, K. L.; Blaney,
F. E.; Foley, J.; Bacon, A. M.; Webb, E. F.; Luttmann, M. A.; Burman, M.;
Sarau, H. M.; Salmon, M.; Palovich, M. R.; Belmonte, K. Design,
Synthesis, and StructureꢀActivity Relationship of Tropane Muscarinic
Acetylcholine Receptor Antagonists. J. Med. Chem. 2009, 52, 5241–
5252.
(10) Prat, M.; Fernꢀandez, D.; Buil, M. A.; Crepso, M. I.; Casals, G.;
Ferrer, M.; Tort, L.; Castro, J.; Monleꢀon, J. M.; Gavaldꢁa, A.; Miralpeix,
M.; Ramos, I.; Domꢀenech, T.; Vilella, D.; Antꢀon, F.; Huerta, J. M.;
Espinosa, S.; Lꢀopez, M.; Sentellas, S.; Gonzꢀalez, M.; Albertí, J.; Segarra,
V.; Cꢀardenas, A.; Beleta, J.; Ryder, H. Discovery of Novel Quaternary
Ammonium Derivatives of (3R)-Quinuclidinol Esters as Potent and
Long-Acting Muscarinic Antagonists with Potential for Minimal Sys-
temic Exposure after Inhaled Administration: Identification of (3R)-
3-{[Hydroxy(di-2-thienyl)acetyl]oxy}-1-(3-phenoxypropyl)-1-azonia-
bicyclo[2.2.2]octane Bromide (Aclidinium Bromide). J. Med. Chem.
2009, 52, 5076–5092.
(11) Cazzola, M. Aclidinium bromide, a novel long-acting muscari-
nic M3 antagonist for the treatment of COPD. Curr. Opin. Invest. Drugs
2009, 10, 482–490.
(12) Haeberlin, B.; Stowasser, F.; Wirth, W.; Baumberger, A.; Abel,
S.; Kaerger, S.; Kieckbusch, T. Compositions of glycopyrronium salt for
inhalation. PCT Int. Appl. Publ. WO 2008000482, 2008.
(13) Denton, S. M.; Wood, A. A Modified Bouveault Reaction for
the Preparation of α,α-Dimethylamines from Amides. Synlett 1999,
1, 55–56.
We thank Michele Coghlan, Stuart Marshall, Sheena Patel, Jo
Ann Rhodes, and Jessica Watson for their contributions to in
vitro biology studies, and John Adcock, Tim Davies, James Philip,
Louise Sladen, and Karen Wright for in vivo biology studies. We
also thank Katie Bainbridge, Trish Costello, David Cox, Steve
Denton, David Fengas, Ben Greener, Tim Hobson, Simon
Mantell, Louise Marples, and Keith Reeves for assistance with
the chemistry program. Dzelal Serdarevic is also thanked for the
provision of clinical data.
’ ABBREVIATIONS USED
ACh, acetyl choline; CCh, carbamyl choline; CHO, Chinese
hamster ovary; COPD, chronic obstructive pulmonary disease;
CYP, cytochrome P450; DDI, drugꢀdrug interaction; DoA,
duration of action; DPI, dry powder inhaler; DSC, differential
scanning calorimetry; DVS, dynamic vapor sorption; EFS, elec-
trical field stimulation; FEV1, forced expiratory volume in 1 s;
GPT, guinea pig trachea; IT, intratracheal; LAAC, long-acting
anticholinergic; LAMA, long-acting muscarinic antagonist;
MCh, methacholine; NADPH, reduced form of nicotinamide
adenine dinucleotide phosphate; NMS, N-methylscopolamine;
PXRD, powder X-ray diffraction; RL, lung resistance; TGA,
thermogravimetric analysis; TI, therapeutic index; UDPGA,
uridine-50-diphosphoglucuronic acid; UGT, uridine-50-dipho-
sphoglucuronosyl transferase
(14) Boehringer, M.; Hunziker, D.; Kuehne, H.; Loeffler, B. M.;
Sarabu, R.; Wessel, H. P. N-substituted Pyrrolidine Derivatives as
Dipeptidyl Peptidase IV Inhibitors. PCT Int. Appl. Publ. WO 2003037327,
2003.
’ REFERENCES
(15) Espino, C. G.; Wehn, P. M.; Chow, J.; Du Bois, J. Synthesis of
1,3-Difunctionalized Amine Derivatives through Selective CꢀH Bond
Oxidation. J. Am. Chem. Soc. 2001, 123, 6935–6936.
(1) Rabe, K. F.; Hurd, S.; Anzueto, A.; Barnes, P. J.; Buist, S. A.;
Calverley, P.; Fukuchi, Y.; Jenkins, C.; Rodriguez-Roisin, R.; van Weel,
C.; Zielinski, J. Global Strategy for the Diagnosis, Management, and
Prevention of Chronic Obstructive Pulmonary Disease: GOLD Execu-
tive Summary. Am. J. Respir. Crit. Care Med. 2007, 176, 532–555.
(2) Hansel, T. T.; Barnes, P. J. Tiotropium bromide: A novel once-
daily anticholinergic bronchodilator for the treatment of COPD.
Drugs Today 2002, 38, 585–600.
ꢂ
(16) Bower, J. F.; Svenda, J.; Williams, A. J.; Charmant, J. P. H.;
Lawrence, R. M.; Szeto, P.; Gallagher, T. Cyclic Sulfamidates as Vehicles
for the Synthesis of Substituted Lactams. Org. Lett. 2004, 6, 4727–4730.
(17) Glossop, P. A.; Mantell, S. J.; Strang, R. S.; Watson, C. A. L.;
Wood, A. Carboxamide derivatives as muscarinic receptor antagonists.
PCT Int. Appl. Publ. WO 2007034325, 2007.
(18) (a) Casarosa, P.; Bouyssou, T.; Germeyer, S.; Schnapp, A.;
Gantner, F.; Pieper, M. Preclinical Evaluation of Long-Acting Muscari-
nic Antagonists: Comparison of Tiotropium and Investigational Drugs.
J. Pharmacol. Exp. Ther. 2009, 330, 660–668. (b) Gavaldꢁa, A.; Miralpeix,
M.; Ramos, I.; Otal, R.; Carre~no, C.; Vi~nals, M.; Domꢀenech, T.;
Carcasona, C.; Reyes, B.; Vilella, D.; Gras, J.; Cortijo, J.; Morcillo, E.;
Llenas, J.; Ryder, H.; Beleta, J. Characterization of Aclidinium Bromide, a
Novel Inhaled Muscarinic Antagonist, with Long Duration of Action and
a Favorable Pharmacological Profile. J. Pharmacol. Exp. Ther. 2009,
331, 740–751.
(19) (a) Cross, P. E.; Mackenzie, A. R. Preparation of pyrrolidine
derivatives as muscarinic receptor antagonists. Eur. Pat. Appl. Publ.
EP388054, 1990. (b) Alabaster, V. A. Discovery and development of
selective M3 antagonists for clinical use. Life Sci. 1997, 60, 1053–1060.
(20) Lee, J. T.; Kroemer, H. K.; Silberstein, D. J.; Funck-Brentano,
C.; Lineberry, M. D.; Wood, A. J.; Roden, D. M.; Woosley, R. L. The
Role of Genetically Determined Polymorphic Drug Metabolism in the
Beta-Blockade Produced by Propafenone. New Engl. J. Med. 1990,
322, 1764–1768.
(21) Williams, J. A.; Hyland, R.; Jones, B. C.; Smith, D. A.; Hurst, S.;
Goosen, T. C.; Peterkin, V.; Koup, J. R.; Ball, S. E. Drug-Drug
Interactions for UDP-Glucuronosyltransferase Substrates: A Pharmaco-
kinetic Explanation for Typically Observed Low Exposure (AUCI/
AUC) Ratios. Drug Metab. Dispos. 2004, 32, 1201–1208.
(22) Glossop, P. A.; Lane, C. A. L.; Price, D. A.; Bunnage, M. E.;
Lewthwaite, R. A.; James, K.; Brown, A. D.; Yeadon, M.; Perros-Huguet,
C.; Trevethick, M. A.; Clarke, N. P.; Webster, R.; Jones, R. M.; Burrows,
J. L.; Feeder, N.; Taylor, S. C. J.; Spence, F. J. Inhalation by Design:
(3) Gross, N. J.; Skorodin, M. S. Role of the Parasympathetic System
in Airway Obstruction Due to Emphysema. N. Eng. J. Med. 1984,
311, 421–425.
(4) Coulson, F. R.; Fryer, A. D. Muscarinic acetylcholine receptors
and airway diseases. Pharmacol. Ther. 2003, 98, 59–69.
(5) Joos, G. F. Potential for long-acting muscarinic antagonists in
chronic obstructive pulmonary disease. Expert. Opin. Investig. Drugs
2010, 19, 257–264.
(6) Lainꢀe, D. I. Long-acting muscarinic antagonists for the treatment
of chronic obstructive pulmonary disease. Expert Rev. Clin. Pharmacol.
2010, 3, 43–53.
(7) Cazzola, M.; Matera, M. G. Emerging inhaled bronchodilators:
an update. Eur. Respir. J. 2009, 34, 757–769.
(8) Villetti, G.; Pastore, F.; Bergamaschi, M.; Bassani, F.; Bolzoni,
P. T.; Battipaglia, L.; Amari, G.; Rizzi, A.; Delcanale, M.; Volta, R.;
Cenacchi, V.; Cacciani, F.; Zaniboni, M.; Berti, F.; Rossoni, G.; Harrison,
S.; Petrillo, P.; Santoro, E.; Scudellaro, R.; Mannini, F.; Geppetti, P. A.;
Razzetti, R.; Patacchini, R.; Civelli, M. Bronchodilator Activity of (3R)-
3-[[[(3-Fluorophenyl)[(3,4,5-trifluorophenyl)methyl]amino]carbonyl]-
oxy]-1-[2-oxo-2-(2-thienyl)ethyl]-1-azoniabicyclo[2.2.2]octane Bro-
mide (CHF5407), a Potent, Long-Acting, and Selective Muscarinic M3
Receptor Antagonist. J. Pharmacol. Exp. Ther. 2010, 335, 622–635.
(9) (a) Wan, Z.; Laine, D. I.; Yan, H.; Zhu, C.; Widdowson, K. L.;
Buckley, P. T.; Burman, M.; Foley, J. J.; Sarau, H. M.; Schmidt, D. B.;
Webb, E. F.; Belmonte, K. E.; Palovich, M. Discovery of (3-endo)-3-(2-
cyano-2,2-diphenylethyl)-8,8-dimethyl-8-azoniabicyclo[3.2.1]octane
bromide as an efficacious inhaled muscarinic acetylcholine receptor
antagonist for the treatment of COPD. Bioorg. Med. Chem. Lett. 2009,
6903
dx.doi.org/10.1021/jm200884j |J. Med. Chem. 2011, 54, 6888–6904

Journal of Medicinal Chemistry
ARTICLE
Novel Ultra-Long-Acting β2-Adrenoreceptor Agonists for Inhaled
Once-Daily Treatment of Asthma and Chronic Obstructive Pulmonary
Disease That Utilize a Sulfonamide Agonist Headgroup. J. Med. Chem.
2010, 53, 6640–6652.
(23) Boase, S.; Miners, J. O. In vitro-in vivo correlations for drugs
eliminated by glucuronidation: Investigations with the model substrate
zidovudine. Br. J. Clin. Pharmacol. 2002, 54, 493–503.
(24) Magdalou, J.; Fournel-Gigleux, S.; Ouzzine, M. Insights on
membrane topology and structure/function of UDP-glucuronosyltrans-
ferases. Drug Metab. Rev. 2010, 42, 159–166.
(25) Copeland, R. A.; Pompliano, D. L.; Meek, T. D. Drug-target
residence time and its implications for lead optimization. Nature Rev.
Drug Discov. 2006, 5, 730–739.
(26) Disse, B.; Reichl, R.; Speck, G.; Traunecker, W.; Rominger,
K. L.; Hammer, R. Ba 679 BR, A novel long-acting anticholinergic
bronchodilator. Life Sci. 1993, 52, 537–544.
(27) Nials, A. T.; Sumner, M. J.; Johnson, M.; Coleman, R. A.
Investigations into factors determining the duration of action of the beta
2-adrenoceptor agonist, salmeterol. Br. J. Pharmacol. 1993, 108, 507–
515.
(28) Motulsky, H. J.; Mahan, L. C. The Kinetics of Competitive
Radioligand Binding Predicted by the Law of Mass Action. Mol.
Pharmacol. 1984, 25, 1–9.
(29) Wright, K.; Davies, T.; Clarke, N.; Yeadon, M.; Perros-Huguet,
C. PF-00610355 ꢀ an inhaled β2-adrenoceptor agonist with superior
potency and duration of action (DoA) to salmeterol in the conscious
dog model of bronchoconstriction. Eur. Respir. J. 2009, 34 (S53),
Abstract P2060.
(30) Taburet, A. M.; Schmit, B. Pharmacokinetic Optimization of
Asthma Treatment. Clin. Pharmacokinet. 1994, 26, 396–418.
(31) Glossop, P. A.; James, K. The hydrochloride salt of 5-(3-(3-
hydroxyphenoxy)azetidin-1-yl)-5-methyl-2,2-diphenylhexanamide and
its use as a medicament. PCT Int. Appl. Publ. WO 2008135819, 2008.
(32) Sladen, L.; Wright, K.; Adcock, J.; Chaffe, P.; Clarke, N.;
Yeadon, M.; Perros-Huguet, C. Effects of intra tracheal PF-00610355,
an inhaled β2-adrenoceptor agonist, in anaesthetised animals. Eur.
Respir. J. 2009, 34 (S53), Abstract P2020.
(33) Watson, J.; Strawbridge, M.; Brown, R.; Campany, K.; Coghlan,
M.; Trevethick, M. Offset rates of tiotropium and ipratropium at human
recombinant muscarinic M1-M5 receptors using a dilution-offset meth-
odology. Fundam. Clin. Pharmacol. 2008, 22 (S2), 78, Abstract P102.
(34) Banholzer, R.; Bauer, R.; Reichl, R. Preparation of anticholi-
nergic scopine, (nor)tropine, and granatoline esters of thienylcarboxylic
acids and their quaternary salts. Eur. Pat. Appl. Publ. EP 418716, 1991.
6904
dx.doi.org/10.1021/jm200884j |J. Med. Chem. 2011, 54, 6888–6904