Discovery of AZD6642, an inhibitor of 5-lipoxygenase activating protein (FLAP) for the treatment of inflammatory diseases

Article abstract of DOI:10.1021/jm501531v

A drug discovery program in search of novel 5-lipoxygenase activating protein (FLAP) inhibitors focused on driving a reduction in lipophilicity with maintained or increased ligand lipophilic efficiency (LLE) compared to previously reported compounds led to the discovery of AZD6642 (15b). Introduction of a hydrophilic tetrahydrofuran (THF) ring at the stereogenic central carbon atom led to a significant shift in physicochemical property space. The structure-activity relationship exploration and optimization of DMPK properties leading to this compound are described in addition to pharmacokinetic analysis and an investigation of the pharmacokinetic (PK)-pharmacodynamic (PD) relationship based on ex vivo leukotriene B₄ (LTB₄) levels in dog. AZD6642 shows high specific potency and low lipophilicity, resulting in a selective and metabolically stable profile. On the basis of initial PK/PD relation measured, a low dose to human was predicted.

Full text of DOI:10.1021/jm501531v

Article
pubs.acs.org/jmc
Discovery of AZD6642, an Inhibitor of 5‑Lipoxygenase Activating
Protein (FLAP) for the Treatment of Inflammatory Diseases
,†
†
§
†
†
̈
Malin Lemurell,* Johan Ulander, Susanne Winiwarter, Anders Dahlen, Ojvind Davidsson,
́
Hans Emtenas,† Johan Broddefalk,† Marianne Swanson,† Daniel Hovdal,† Alleyn T. Plowright,†
̈
Anna Pettersen,∥ Marie Ryden-Landergren,‡ Jonas Barlind,† Antonio Llinas,‡ Margareta Herslof,†
́
̈
Tomas Drmota,‡ Kalle Sigfridsson,∥ Sara Moses,§ and Carl Whatling†
†Cardiovascular & Metabolic Diseases iMed, ‡Respiratory, Inflammation & Autoimmune Diseases iMed, §Drug Safety & Metabolism,
∥
and Pharmaceutical Development, AstraZeneca, Pepparedsleden 1, Molndal, 43183, Sweden
̈
S
* Supporting Information
ABSTRACT: A drug discovery program in search of novel 5-lipoxygenase acti-
vating protein (FLAP) inhibitors focused on driving a reduction in lipophilicity
with maintained or increased ligand lipophilic efficiency (LLE) compared to pre-
viously reported compounds led to the discovery of AZD6642 (15b). Introduc-
tion of a hydrophilic tetrahydrofuran (THF) ring at the stereogenic central
carbon atom led to a significant shift in physicochemical property space. The
structure−activity relationship exploration and optimization of DMPK properties
leading to this compound are described in addition to pharmacokinetic analysis
and an investigation of the pharmacokinetic (PK)−pharmacodynamic (PD) relationship based on ex vivo leukotriene B4 (LTB4)
levels in dog. AZD6642 shows high specific potency and low lipophilicity, resulting in a selective and metabolically stable profile.
On the basis of initial PK/PD relation measured, a low dose to human was predicted.
our laboratories. At the time, two classes of compounds had
been described. The first class is represented herein by the
phase II clinical candidate AM-803 (asthma indication, now
GSK2190915).8 The second class of compounds was described
in a patent family represented herein by P-1.9−11
INTRODUCTION
■
Leukotrienes are lipid mediators derived from arachidonic acid
that have a variety of inflammatory and vasoactive actions.1 They
are synthesized by the 5-lipoxygenase (5-LO) pathway through
the common precursor leukotriene A4 (LTA4) followed by
conversion to leukotriene B4 (LTB4) via the action of the enzyme
LTA4 hydrolase or to the cysteinyl leukotriene (cysLT) LTC4 via
the action of the enzyme LTC4 synthase. The principal source of
leukotrienes is white blood cells (leukocytes). However, vascular
cells that do not express 5-LO can also take part in transcellular
leukotriene production following the uptake of LTA4 released by
leukocytes in proximity.2,3 Leukotrienes play a role in innate
immunity but have also been implicated in the pathology of a
variety of diseases that have an inflammatory component. In
particular, the role of leukotrienes in asthma has been identified
and receptor antagonists that block the actions of cysLTs on the
CysLT1 receptor are used inthe clinic.4 More recently, interest has
also developed around the role of leukotrienes in cardiovascular
disease, leading to the initial evaluation of a number of leukotriene
inhibitors in patients with coronary artery disease.5−7
Two crystal structures of FLAP in complex with a forerunner
to AM-803, MK-591, and its iodine analogue have been
published.12,13 While the resolution in the crystal structures is
relatively low (4.0 Å for 2q7r and 4.25 Å for 2q7m) with limited
electron density for both ligands and protein residues, the
presence of an iodine with intrinsic high electron density in one
of the structures strengthened the resulting ligand binding mode
hypothesis and encouraged the use of structural information for
prospective design of compounds.
To generate LTA4, 5-LO needs to interact with 5-lipoxygenase
activating protein (FLAP) in the nuclear membrane. FLAP
facilitates the transfer of arachidonic acid released from
membrane phospholipids via the action of a cytoplasmic
phospholipase to the active site of 5-LO. Both 5-LO and FLAP
are essential for leukotriene production, and inhibitors of either
protein can block leukotriene production.
Both AM-803 and P-1 reported good potency for inhibition of
FLAP. However, both compounds display high lipophilicity
which results in potential liabilities such as low solubility and the
A program to identify novel inhibitors of FLAP with a profile
suitable for chronic oral once daily administration was initiated in
Received: October 5, 2014
© XXXX American Chemical Society
A
dx.doi.org/10.1021/jm501531v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
ab c
, ,
Scheme 1. Late Stage Variation of the Right-Hand Side To Provide Derivatives 12−15
a
Reagents and conditions: (a) N,O-dimethylhydroxylamine·HCl, DIPEA, CDI, 2-MeTHF, 67%; (b) (4-(benzyloxy)phenyl)magnesium bromide,
THF, 81%; (c) tert-butylmagnesium bromide, THF, 41%; (d) DAST, DCM, CHCl3, 80%; (e) CO (g), Pd(dppf)Cl2, Et3N, DMF, MeOH, 70 °C,
71−98%; (f) NH4HCO2, Pd(OH)2, MeOH, 120 °C, 100%; (g) (3-(dimethylamino)propyl)magnesium chloride, THF; (h) BuBr, NaH, xylene, 75%
(over two steps); (i) H2, Pd(OH)2, MeOH, 89%; (j) 2-(chloromethyl)pyridine hydrochloride, KI, Cs2CO3, DMF, 69−76%.; (k) Tf2O, DMAP,
DCM, 0 °C, 99%; (l) (1) (PinB)2, Pd(dppf)Cl2, KOAc, dioxane, 80 °C; (2) ArBr, Pd(dppf)Cl2, K2CO3, dioxane, H2O, 80 °C; or ArBPin,
b
Pd(dppf)Cl2, K3PO4, DME, EtOH, H2O, 135 °C (m) LiOH, THF, H2O; (n) R1-NH2, TBTU, NMM, DMF. See Table 1 for detailed substitution
pattern in compounds 12−15. The final compounds were separated by chiral chromatography and the most active enantiomers are the ones further
described and for which data are given, unless otherwise stated. See Supporting Information for details.
c
potential for increased off-target effects.14−16 The putative
binding site of the FLAP protein is located in the central
transmembrane region, and it is important to identify com-
pounds with binding affinity that is not governed by general
partitioning into the membrane. Our strategy was to drive design
with a strong emphasis on improving ligand lipophilic efficiency
(LLE) and to identify novel potent FLAP inhibitors with reduced
lipophilicity compared to earlier known inhibitors. Such com-
pounds, being within a more optimal physicochemical property
space and with nonhydrophobically driven affinity, are likely
to enhance the chances for successful development of a well
tolerated oral drug for long-term treatment in inflammatory
indications.
Analysis of the property space covered by AM-803 and P-1
showed that P-1 had both lower molecular weight and
lipophilicity, and hence, the structure of P-1 was deemed a
more favorable starting point. Herein, we describe the
identification of a novel scaffold that significantly shifted the
molecular property space of the chemical series and led us to
discover AZD6642 as a drug candidate.
RESULTS AND DISCUSSION
■
Chemistry. The strategy during the lead generation phase
was to simplify the synthesis of screening compounds by
designing routes that allowed late stage diversification. As a
result, variations, either on the right-hand side (Scheme 1) or on
B
dx.doi.org/10.1021/jm501531v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
ab c
, ,
was then obtained via a one-step benzyl deprotection/
defluorination reaction with 3 using transfer hydrogenation
with ammonium formate and Pd(OH)2 in MeOH with micro-
wave heating. In comparison, attempts to hydrogenate 3 with
Pd/C at room temperature gave only partially defluorinated
mixtures.
Scheme 2. Late Stage Variation of the Left-Hand Side
To construct the tetrahydrofuran (THF) analogue 7, the
crucial intermediate 5 was first prepared via addition of
(3-(dimethylamino)propyl)magnesium chloride to 1 followed
by treatment with NaH and butyl bromide to induce ring-
closure.18 Compound 7 was then obtained by palladium
catalyzed carbonylation17 to give 6 followed by hydrogenation
using Pd(OH)2 as catalyst.
The key intermediates 4 and 7 were alkylated with 2-
(chloromethyl)pyridine to give 8 and 10, respectively. However,
to obtain the Suzuki coupled compounds 9 and 11, the phenols 4
and 7 were converted to the corresponding triflates and then
either coupled directly with heteroaryl pinacolboranes or treated
first with bis(pinacolato)diboron followed by reaction with
heteroaryl bromides.
Finally, the desired compounds 12−15 could be synthesized
by hydrolysis of the methyl ester to provide the carboxylic acids,
which were treated with a range of amines, the coupling reagent
TBTU, and N-methylmorpholine (NMM) to give the amide
derivatives 12−15. The racemic mixtures of 12−15 were
separated by chiral chromatography to give both enantiomers.
The most active enantiomers are the ones further described in
this article unless otherwise stated.
a
Reagents and conditions: (a) LiOH, THF, H2O, 100%.; (b) R1-NH2,
TBTU or HATU, NMM, DMF; (c) H2 (g), Pd(OH)2, MeOH; (d)
NH4HCO2, Pd(OH)2, MeOH, 120 °C; (e) 2-(chloromethyl)pyridine
hydrochloride, KI, Cs2CO3, DMF; (f) Tf2O, DMAP, DCM, −78 °C;
(g) (1) (PinB)2, Pd(dppf)Cl2, KOAc, dioxane, 80 °C; (2) ArBr,
Pd(dppf)Cl2, K2CO3, dioxane, H2O, 80 °C; or ArBPin, Pd(dppf)Cl2,
b
K2CO3, dioxane, H2O, 80 °C. See Table 1 for detailed substitution
c
pattern in compounds 12−15. The final compounds were separated
by chiral chromatography and the most active enantiomers are the
ones further described and for which data are given, unless otherwise
stated. See Supporting Information for details.
the left-hand side (Scheme 2) as the final step, were developed
in the following way. Ketone 1 was prepared in large scale
by converting 6-bromonicotinic acid to the corresponding
Weinreb amide followed by addition of (4-(benzyloxy)phenyl)-
magnesium bromide. Addition of tert-butylmagnesium bromide
to 1 gave a dibenzylic alcohol, which was sequentially fluorinated
with diethylaminosulfur trifluoride (DAST) to give compound 2.
The bromide 2 was then converted to the methyl picolinate 3 by
palladium catalyzed carbonylation.17 The key building block 4
To introduce the left-hand side of the molecules as the final
step, the N-substituted picolinamide intermediates 16 and 17
were first prepared via hydrolysis of the intermediates 3 and 6,
a b
,
Scheme 3. Synthesis of Oxadiazole Derivatives 19−22
a
Reagents and conditions: (a) N2H4·H2O, MeOH, reflux, 99%; (b) CDI, DMF, rt, 19%; (c) ethyl 2-chloro-2-oxoacetate, DIPEA, DCM, rt, 5 min,
then TosCl, rt, 2 h, 87%; (d) MeMgBr, THF, rt, 5 min, 58%; (e) BrCN, NaHCO3, dioxane/H2O, rt, 32%; (f) HC(OEt)3, toluene/AcOH, 100 °C,
b
100%. The final compounds were separated by chiral chromatography and the most active enantiomers are the ones further described and for which
data are given, unless otherwise stated. See Supporting Information for details.
C
dx.doi.org/10.1021/jm501531v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a b
,
Scheme 4. Ring Closing Metathesis Route to Tetrahydropyran and Substituted Tetrahydrofuran Rings
a
Reagents and conditions: (a) allylic- or vinylic-MgBr, THF, reflux; (b) allylic bromide, NaH, THF, 40 °C, 40−85% (over two steps); (c) Grubbs
b
second, DCM or toluene, reflux, 90−95%; (d) procedures according to Schemes 1 and 2, 55−60% (over seven steps). The final compounds were
separated by chiral chromatography and the most active enantiomers are the ones further described and for which data are given, unless otherwise
stated. See Supporting Information for details.
a b
,
Scheme 5. Synthesis of the Difluorotetrahydrofuran Derivative 37
a
Reagents and conditions: (a) allyl-MgBr, THF, reflux, 90%; (b) K2OsO4, NMO, THF/H2O, acetone, rt, 90%; (c) PTSA, CHCl3, reflux, 90%; (d)
b
DMP, DCM, rt, 99%; (e) Deoxo-Fluor, DCM, reflux, 80%; (f) procedures according to Schemes 1 and 2, 46% (over seven steps). The final
compounds were separated by chiral chromatography and the most active enantiomers are the ones further described and for which data are given,
unless otherwise stated. See Supporting Information for details.
respectively, followed by amide couplings and finally depro-
tection of the phenol (Scheme 2). The same protocols as
described in Scheme 1 were then utilized to finalize the target
compounds 12−15.
During the optimization of the side chains, we found that
oxadiazole rings served as good amide isosteres. These
derivatives were synthesized via the semicarbazide intermediates
18a−c (Scheme 3), followed by reaction with orthoesters,19
carbonylimidazole,20 bromocyanide,21 or ethyl 2-chloro-2-
oxoacetate and Grignard addition to give the oxadiazole deriv-
atives 19−22.
Preparation of tetrahydropyran (THP) and methyl substituted
THF derivatives was perfomed via a general synthetic route
consisting of Grignard addition of the appropriate vinylic or
allylic magnesium reagent to ketone 1 followed by allylation of
the transitional hydroxyl group to give the intermediates 23−25
(Scheme 4). Ring-closing metathesis (RCM) with Grubbs
second generation catalyst then provided 26−28. The seven final
steps to install the desired side chains were continually per-
formed using the same methodology as described in Schemes 1
and 2 to give the derivatives 29−31. A precursor to compound
32, in which the double bond was still intact, was isolated as a
byproduct after the hydrogenation step and was further
developed to a screening compound using the procedure
described for 29−31.
Although the RCM route gave us the opportunity to
synthesize substituted THF and THP derivatives, it was not a
suitable route to construct heterocyclic rings containing fluoride
substituents (Scheme 5). Several routes were investigated to find
a good synthesis toward the difluorosubstituted THF scaffold
starting from ketone 1, e.g., Knoevenagel reactions, Wittig
reactions, or allyl Grignard addition followed by Wacker
oxidation, epoxidation, or hydroboration−oxidation of the allyl
double bond. Unfortunately, we soon learned that steric bulk,
unfavorable selectivities, or unintentional elimination of water
dramatically limited our options. However, the most straight-
forward method to 37 was to first prepare 33 by addition of
allylmagnesium reagent to 1, followed by dihydroxylation of the
double bond. The desired THF intermediate 34 was then
obtained by SN1-type ring-closure with p-toluenesulfonic acid.
Dess−Martin periodinane oxidation gave ketone 35, which was
sequentially fluorinated with Deoxo-Fluor to give scaffold 36.
Finally, compound 37 could be synthesized with the preferred
left-hand and right-hand side substituents using the methods
described previously (e.g., see compound 15b, Schemes 1 and 2)
in 27% overall yield over 12 steps.
In order to substitute the methylene group adjacent to the
oxygen in the THF ring, the lactone 38 was synthesized by addi-
tion of a suitable pyridyl Grignard reagent to the commercially
available methoxyphenyl keto ester (Scheme 6). Ring-opening of
D
dx.doi.org/10.1021/jm501531v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
ab
,
Scheme 6. Synthesis of the Dimethyl- and Dideuterotetrahydrofuran Analogues
a
Reagents and conditions: (a) 5-iodo-2-bromopyridine, i-PrMgCl, THF, 0 °C to rt, 44%; (b) LiAlD4, THF, 0 °C or MeMgBr, THF, rt; (c) PTSA,
CHCl3, reflux, 74% (over two steps, R1 = D), respectively 27% (over two steps, R1 = Me); (d) procedures according to Schemes 1 and 2 except for
b
using NaSMe, DMF, 120 °C instead of H2, Pd(OH)2, MeOH to deprotect the ether, 18−20% (over seven steps, R1 = D or Me). The final
compounds were separated by chiral chromatography and the most active enantiomers are the ones further described and for which data are given,
unless otherwise stated. See Supporting Information for details.
a b
,
Scheme 7. Synthesis of the 3,3-Disubstituted Tetrahydrofuran Analogue 47
a
Reagents and conditions: (a) H2SO4, EtOH, reflux, 95%; (b) p-fluoronitrobenzene, NaH, DMF, 0 °C, 67%; (c) methyl bromoacetate, NaH, DMF,
rt, 55%; (d) LiAlH4, Et2O, 0 °C, 80%; (e) HBr (aq), reflux, 54%; (f) CO (g), Pd(dppf)Cl2, Et3N/DMF/MeOH, 70 °C, 65%; (g) CuBr, HBr (aq),
NaNO2, −10 °C, 40%; (h) LiOH, THF/H2O, rt, 100%; (i) R1-NH2, EDCI, HOBt, Et3N, DMF, rt, 69%; (j) (PinB)2, Pd(dppf)Cl2, KOAc, dioxane,
b
80 °C; (k) 5-bromopyrazin-2-amine, Pd(dppf)Cl2, K2CO3, dioxane/H2O, 80 °C, 33% (over two steps). The final compound was separated by
chiral chromatography, and the most active enantiomer is the one further described and for which data are given. See Supporting Information for
details.
the lactone, either by LiAlD4 or MeMgBr, generated a diol, which
after subsequent acidic treatment formed the desired disub-
stituted THF ring containing derivatives 39 and 40 via an SN1-
type ring-closure.22 The geminal substituted THF derivatives
were then subjected to the same synthetic sequence described for
compound 15b (described in Schemes 1 and 2) to install the
desired substituents to give 41 and 42.
The synthetic route to the 3,3-disubstituted THF analogue,
where the oxygen atom was repositioned one atom in the ring,
required major changes of the original route (Scheme 7). These
changes were made primarily because of problems with low
reactivity of the benzylic position, steric bulk, or incompatibility
of the protecting groups. Initial Fischer esterification of a suitable
pyridyl acetic acid followed by nucleophilic aromatic substitution
using p-fluoronitrobenzene allowed the construction of the
central parts of the scaffold as shown in 43. Deprotonation of the
benzylic position followed by treatment with methyl bromoace-
tate resulted in an advanced diester intermediate 44. Subsequent
reduction of 44 with LiAlH4 gave a diol intermediate which was
treated with aqueous HBr to give the desired THF intermediate
45. The left- and right-hand side chain substituents were then
installed according to the procedures described in Schemes 1 and
2 (see compound 15b) to give compound 47 with the exception
that the left-hand side was installed via a bromide instead of a triflate.
Finally, in an attempt to introduce a more flexible scaffold,
in which the THF ring had been opened, ketone 1 was treated
with MeMgCl followed by SOCl2 and MeOH to give the
methoxymethyl core structure 48 (Scheme 8). The final steps
were once again carried out as described in Schemes 1 and 2 to
give the required compound 49.
Determination of the absolute stereochemistry for the
enantiomers in the THF series proved challenging. Similar
E
dx.doi.org/10.1021/jm501531v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a b
,
Scheme 8. Synthesis of the Methoxymethyl Analogue 49
a
Reagents and conditions: (a) MeMgCl, THF, −78 °C, 90%; (b) SOCl2, DCM/DMF, rt, 10 min, then MeOH, 5 min, 50%; (c) procedures
b
according to Schemes 1 and 2, 43% (over seven steps). The final compound was separated by chiral chromatography and the most active
enantiomer is the one further described and for which data is given. See Supporting Information for details.
3
vibrational circular dichroism (VCD) spectra were obtained for
15b and its corresponding less active enantiomer 15b′, and no
suitable crystals for X-ray diffraction measurements could be
obtained. Hence, our attention turned to the racemic bromo
intermediate 5 (Scheme 1). Use of chiral chromatography
provided the enantiomers of 5, and to our delight VCD could be
used to establish the absolute stereochemistry (Figure 1). The
binding assay with H-MK-591 as the ligand, implying that
we would optimize inhibitors binding to the same site in the
transmembrane region of FLAP as this compound. The more
potent compounds were then evaluated in a secondary assay that
measured the effect on LTB4 production in human whole blood
(hWB). Because of the effect of nonspecific binding, the free
potency is in general better suited for comparison between
assays. This not only is the case for in vitro systems but also needs
to be considered for translational studies, since the free
concentration in the body’s effect compartments typically
drives the pharmacological response.23,24 Thus, IC50(hWBfree
)
was a primary optimization parameter, estimated according to
IC50(hWBfree) = IC50(hWB)·fu, where fu is the compound
specific fraction unbound in human plasma. Upon correction of
nonspecific binding in the binding assay, the correlation between
the binding assay and free whole blood potency was sufficient
(R2 = 0.8) to establish that the effect on LTB4 production is
driven by FLAP inhibition.
The initial focus was to reduce lipophilicity by introduction of
alternative heterocycles as well as more polar substituents on
both sides of the molecule. The stereogenic alkyl substituted
central carbon, on the other hand, was considered to provide a
distinct conformation critical for inhibition of FLAP. Limitations
in synthetic speed and feasibility with the quarternary carbon
center led us to undertake SAR exploration using the tertiary tert-
butyl substituted carbon analogues as exemplified with com-
pound 12. Combinations of aromatic and heteroaromatic rings
were explored at the positions adjacent to the chiral carbon
center with limited success in achieving any improvement in
potency and physicochemical property space compared to our
starting point. Phenyl in combination with 3-pyridyl was one of
the more preferred subseries identified. On the left-hand side
(R1) a relatively narrow SAR limited the opportunities for
modulation of lipophilicity. However, aminopyrimidine was
identified as a functional group for which potency was retained
despite contributing to a clearly lower lipophilicity (see com-
parison in cLogP and potency for the racemate 13a vs 12) (Table 1).
The right-hand side (R2) was tolerant to structural variation,
including amides,11 providing a useful handle to modulate the
properties of this compound class and allowed for introduction
of a wide span of hydrophilic groups.
Overall, the desired significant shift in compound property
space was not achieved with this initial strategy. Alternatively,
we investigated more polar central fragments. SAR analyses
suggested the requirement of a pharmacophore with a lipophilic
core and with restricted conformational freedom around the
central carbon. Our hypothesis was to embed the polarity with
retained conformational rigidity, and a set of five- and six-
membered aliphatic heterocyclic rings were explored. A THF
ring (14e and 15a) was discovered to retain potency with a sig-
nificant drop in log D, and hence, the desired shift in physicochemical
Figure 1. Vibrational circular dichroism spectra for the enantiomers of
compound 5: blue = experimental (−)-5; purple = experimental (+)-5;
red = simulated (R); green = simulated (S) (see Supporting Information
for experimental details).
enantiomer with negative optical rotation was determined to
have (R)-configuration. This configuration was also confirmed
using X-ray diffraction measurements (Figure 2). Starting from
Figure 2. X-ray diffraction structure of (−)-(R)-5 (see Supporting
Information for additional details).
the (−)-(R)-enantiomer of 5, the final compound was prepared
using the protocol for rac-15b (Scheme 1). The optical rotation
of the enantiomer obtained matched the most active enantiomer
15b; i.e., the absolute stereochemistry of 15b (AZD6642) was
thereby determined to have a (−)-(R)-configuration.
In Vitro Data and Structure−Activity Relationships. Com-
pounds binding to FLAP were evaluated using a competitive
F
dx.doi.org/10.1021/jm501531v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Table 1. Structure−Activity Relationship for the Described Series of FLAP Inhibitors
a
Data are given for the most active isomer of the enantiomeric pairs unless denoted (rac). Defined in Supporting Information. ND = not determined.
b
c
Rac = racemic. pIC50 in the FLAP binding assay, pIC50 in the human whole blood assay measuring the inhibition of the downstream LTB4
production. pIC50 in the hWB assay taking into account the plasma protein binding. LLE calculated from pIC50(hWBfree) minus log D7.4 hplc.
d
e
property space could be achieved (cf. 12 and 14e, respectively)
(Table 1).
SAR was found to be transferable to a large extent from pre-
core. Important SAR information using the THF core structure
is given for selected compounds in Table 1. Exploration of R2
with a variation of amides (14a−d) while keeping the picolyl
substituent as R1 revealed that log D could be reduced toward 2
viously gained knowledge using the tert-butyl substituted central
G
dx.doi.org/10.1021/jm501531v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Table 2. Structure−Activity Relationship Exploration at the Central Core A
a
Data are given for the most active isomer of the enantiomeric or diasteromeric pairs unless otherwise stated. Defined in Supporting Information.
b
c
ND = not determined. pIC50 in the FLAP binding assay. pIC50 in the human whole blood assay measuring the inhibition of the downstream LTB4
production. pIC50 in the hWB assay taking into account the plasma protein binding. LLE calculated from pIC50(hWBfree) minus log D7.4 hplc. The
d
e
f
least active enantiomer.
with retained potency (pIC50 = 6−7) in both the binding and the
hWB assays for a wide set of compounds. In addition, substituted
heteroaromatic rings were well tolerated, exemplified here with
two oxadiazoles (20, 21). In summary, both acidic and basic
moieties as well as neutral groups with hydrogen bond donors or
acceptors in R2 could lead to potent inhibitors. Modification of
R1 was focused on finding alternatives to the aminopyrimidine, a
fragment that had been identified to significantly improve
specific binding. Racemate 15a had an increase in LLE of around
1.5 units when compared to the picolyl variant 14e (most active
enantiomer). Aromatic amines are often associated with
genotoxicity, and 15a was flagged as potentially genotoxic in
database searches.25,26 We tested the corresponding biaryl R1
fragment in the in vitro AMES assay,27 and it was found to be
active. By systematically making close analogues (15b,c, 15d
(racemic)) we identified the aminopyrazine as in 15b to retain
the high LLE. This compound was tested in the AMES assay and
found to be negative and hence without the genotoxic liability.
Monomethylation of the primary amine (15e) was tolerated
but with some reduction in LLE, whereas dimethylation re-
sulted in a major loss of affinity. From matched pair analyses
we found that SAR in R1 was not directly transferable between
R2’s and vice versa. Compounds 15f and 22 were two of the
more interesting compounds in addition to 15b and were
further profiled.
Table 2 illustrates the SAR identified at the central core
showing matched pairs with variation only at the central chiral
carbon, including examples of chiral variation. In the t-Bu series, a
difference in affinity between the most active enantiomer 13b
and the least active enantiomer 13b′ is seen, but the impact of
the stereochemistry is much more pronounced in the THF series
(cf. 15b to its stereoisomer 15b′) with >1 unit difference in pIC50
between the enantiomers. This implies that the oxygen is in-
volved in a specific interaction, since the two THF enantiomers
H
dx.doi.org/10.1021/jm501531v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 3. In Vitro Data for Selected Compounds (Lipophilicity, Solubility, CYP Inhibition, Metabolic Stability in Human
Hepatocytes, Caco-2 Permeability, Reactive Metabolite (RM) Trapping, Contribution of CYP3A4 to Metabolism, and hERG
a
Inhibition)
pIC50
CYP
CYP
3A4
RM
RM
aq solub
b
inh
CYP
isoform
HH CLint
Caco2 Papp
(GSH)
(MeO-NH )
h2
contr
c
d
e
f
g
i
j
compd log D7.4 hplc
(μM)
(max)
((μL/min)/(1 × 106)) (1 × 10−6 cm/s) trapping
trapping
(%)
pIC50(hERG)
k
12
4.4
2
medium
5.2
5.1
4.8
<4.7
5.2
5.5
6.4
<4.7
<4.7
4.9
5.2
6.3
5.8
5.3
6.0
5.6
5.2
2C9
3A4
2C9
11
72
yes
no
no
99
95
87
93
95
64
ND
16
93
l
4.7
14b
14c
14d
14e
15b
15f
19
>1000
5.5
17.0
14
ND
100
61
yes
ND
yes
yes
no
<4.5
ND
<4.5
<4.5
<4.5
ND
ND
<4.5
<4.5
<4.5
ND
ND
ND
ND
ND
ND
k
2.3
2.3
3.1
1.8
1.7
2.3
2.2
1.5
2.4
2.5
2.5
2
high
ND
no
k
high
67
2C9
2C9
3A4
2C9
6.8
<4
62
no
230
ND
ND
73
no
m
<5
ND
ND
93
yes
ND
no
ND
ND
yes
yes
no
47.0
12
k
21
low
k
22
high
2C9, 3A4
2C9
8.3
5.0
2.8
3.3
2.3
3.2
5.0
14
59
no
k
29
high
46
no
73
ND
ND
ND
68
l
30
13
9
2C9
ND
ND
ND
44
no
ND
no
31
2C9
no
32
100
2C9
ND
no
no
k
37
2.1
1.8
2.9
high
2C9
no
41
210
140
2C9
ND
59
yes
ND
no
42
2C9
yes
60
a
b
c
ND, not determined. Experimental details for all assays are described in Supporting Information. Aqueous solubility. Inhibition of CYP450,
d
maximum value for a panel of six CYP isoforms (CYP1A2, CYP2C8, CYP2C9, CYP2C19, CYP2D6, and CYP3A4). Specification of CYP isoform
e
f
associated with the maximum inhibition value. Intrinsic clearance measured in human hepatocytes. Apparent permeability measured in Caco-2 cells.
g
h
Reactive metabolite trapping using gluthathione as trapping agent. Reactive metabolite trapping using methoxylamine as trapping agent.
i
Contribution of CYP3A4 to CYP mediated metabolism of a compound considering a panel of six CYP isoforms (CYP1A2, CYP2C8, CYP2C9,
j
k
CYP2C19, CYP2D6, and CYP3A4). Inhibition of the hERG ion channel. Measured in earlier solubility assay (solubility classified as low, medium,
and high: <10 μM, low; 10−50 μM, medium; >50 μM, high). No CYP3A4 CLint detected. CLint measured in human liver microsomes.
l
m
are close to symmetrical from a steric perspective (visualized in
Figure 5 Supporting Information).
compound was found to be primarily metabolized by CYP3A4,
which potentially could lead to a drug−drug interaction risk
when co-medicated with a CYP3A4 inhibitor.28 Changing the
central core from t-Bu to THF resulted in greater than a log unit
reduction in lipophilicity as measured by log D and improved
solubility and metabolic stability with only slight reduction of
the Caco-2 Papp (cf. 12 vs 14e). While no GSH adducts were
trapped for compound 14e in the reactive metabolite experi-
ment, it formed small amounts of metabolites that could be
trapped with methoxylamine, a trapping agent for reactive
carbonyl species.29 Most likely, oxidation at the 5-position in the
THF ring followed by ring opening gave an aldehyde metabolite
that was subsequently trapped with methoxylamine. Such a
metabolite was also seen in a metabolite identification experi-
ment (data not shown).
Keeping the picolyl side chain as R1 and varying the right-hand
side generated compounds with different profiles including those
with good solubility (e.g., 14b−d) and those with no CYP in-
hibition (e.g., 14d, 19, and 21). Compound 14b, a basic com-
pound containing a dimethylamine group in the side chain,
showed inhibition of both CYP3A4 and CYP2C9 isoforms
(pIC50 values of 5.1 and 4.9, respectively), whereas most other
compounds showed inhibition of only CYP2C9 or no CYP in-
hibition at all. These compounds all displayed good permeability,
with no indication of efflux as measured for a couple of com-
pounds in the Caco-2 cell line (data not shown). Metabolic
stability varied across the series (e.g., 14b had good stability,
whereas 19 was unstable). Most of the compounds were again
mainly metabolized by CYP3A4 except compound 19.
Interestingly, this metabolically unstable compound was found
Compounds with alternative positioning of the oxygen atom,
as in the extension of the THF ring to a THP ring (31), in 47 with
the oxygen moved to the 3-position of the THF ring, or in the
open analogue 49, all show a clearly lower potency (up to 30-fold
difference). In general, inhibition of FLAP was significantly
affected by the position and stereochemistry of any methyl
substitution in the THF ring (cf. 29 and 30). The unsaturated
analogue of 29 with a double bond in the THF ring (32) and a
rather distorted conformation showed a significant drop in
affinity. The 4,4-difluoro analogue 37, on the other hand, showed
improved potency and LLE in the hWB assay and was seen as a
potential alternative to the THF ring.
Selected compounds were profiled in vitro using additional
parameters such as solubility, permeability, and metabolic
stability, and the series was optimized on these parameters, as
the combination of these properties along with the compound’s
potency influences the predicted human dose (Table 3).
Additionally, properties that indicated potential risks for adverse
effects, e.g., CYP inhibition, reactive metabolite formation, and
hERG inhibition, were measured.
Compound 12 is lipophilic with a log D of 4.4 and showed
medium solubility, medium stability in human hepatocytes, but
good Caco-2 permeability. The compound also displayed
moderate inhibition of CYP2C9. Trapping experiments with
glutathione (GSH) indicated the potential formation of a
reactive metabolite, most likely a quinone-like structure formed
by dealkylation of the picolyl residue as observed in a metabolite
identification experiment (data not shown). In addition, this
I
dx.doi.org/10.1021/jm501531v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Table 4. Rat in Vivo PK Data and Allometric Human Dose Prediction
allometric dose from rat (3 × IC50 as average
dose from rat targeting IC80 as minimum
b
c
d
CL
compd (mL min−1 kg−1
t1/2
Vss
bioavailability
(%)
concn; fu species differences considered)
concentration (fu and EH species differences
e
e
)
(h) (L/kg)
(mg)
considered) (mg)
12
19.9
7.8
2.2
3.8
3.7
1.4
3.4
3.8
2.0
2.5
1.5
1.9
40
85
50
60
75
125
15
10
10
5
150
25
15b
21
10.3
15.5
7.6
15
22
270
41
15
a
b
c
d
e
Experimental details including dosing regimen in Supporting Information. Clearance. Half-life. Volume of distribution at steady state. Human
dose prediction per 70 kg body weight.
to have a pronounced contribution to its metabolism from
CYP2C8 instead. Compound 21 with its high hWBfree pIC50 was
deemed interesting from this compound set, despite its low
solubility.
Compounds containing the aminopyrazine ring in R1, 15b, 22,
and 15f, had promising properties. Compound 15b had clearly
better solubility and metabolic stability than its picolyl analogue
14e, and no reduction in permeability was observed. Because
of the high metabolic stability, no metabolites were trapped by
GSH or methoxylamine, and the contribution of CYP3A4 to
metabolism was reduced to below 70%.
Compound 22 had very low lipophilicity, good solubility, low
inhibition of both CYP3A4 and CYP2C9, medium metabolic
stability, good permeability, and no dependency on CYP3A4 for
metabolism. The replacement of the gem-dimethyl substituents
in 15b with a cyclopropyl ring (15f) to reduce lipophilicity was
successful to generate a potent and metabolically stable com-
pound. However, it was soon seen that the compound was
positive in the GSH trapping assay and also had a pronounced
CYP inhibition potential, especially of CYP3A4.
analogue of 15b, compound 42, nor the 3-methyl analogue 29
showed any methoxylamine adducts. However, 42 had reduced
potency and was also found to be metabolically less stable.
Compound 29, on the other hand, was metabolically quite stable
and had good potency. Interestingly, compound 30, with the
methyl group in the 4-position, not only was found to be less
potent than 29 but also showed significantly lower solubility and
higher CYP inhibition (10-fold for CYP2C9) compared with
29. Data are given for the most active stereoisomer of each
diastereomeric methyl analogue, but CYP inhibition and
metabolic stability differed only within experimental error be-
tween each of the enantiomeric pairs. The 4,4-difluoro substituted
THF compound (37), which was highlighted previously because
of its high affinity and LLE, was unfortunately found to be a strong
CYP2C9 inhibitor with a pIC50 value of 6.
Most compounds were found to have low hERG inhibition
(pIC50 values of <4.5). Three of the compounds, 12, 14e, and
15b, were also tested in a panel of four additional cardiac ion
channels (Nav1.5, Cav1.2, IKs, and K.v.4.3; see Supporting
Information), with pIC50 < 4.5.
Hydroxylation in the THF ring was observed in human, rat,
and dog hepatocytes, and in some cases also unsaturated
metabolites could be identified indicating oxidation and
elimination of water. Trapping experiments with methoxylamine
gave small amount of adducts for most THF compounds, con-
firmed to be at the THF ring and implying the formation of
aldehydes. In order to block metabolism various THF analogues
were explored as described above (see Table 2). The deuterium
substituted THF ring 41 was prepared in order to benefit from
both primary and secondary isotope effects to reduce metabolic
hydroxylation in the THF ring.30 A prolonged incubation in the
hepatocyte metabolism assay was required to be able to explore
and compare the metabolite pattern for the highly stable com-
pounds 15b and 41, which resulted in similar levels of ring
hydroxylation for both compounds. Since no methoxylamine
adducts were detected for 15b, we could not verify our hypo-
thesis, that deuteration hindered the specific metabolism, with
these two molecules but investigated another pair of molecules
for potential methoxylamine adduct differences. The 5,5-di-2H-THF
analogue of 21 (data not shown) was found to have very similar
properties to 21, but here the deuterated analogue was observed
to be somewhat more metabolically stable in human hepatocytes
(CLint = 6.5 compared with 11.5 μmol mL−1 kg−1). Still, some of
the metabolites could be trapped with methoxylamine, and
metabolite identification studies showed metabolism/hydrox-
ylation in the THF-ring. Deuteration did not stop metabolism
leading to methoxylamine adducts, even though metabolism
seemed to be slowed.
Selected compounds (12, 14e, 15b, 22, and 41) were profiled
in human blood assays to assess whether they had any effect on
eicosanoid (prostanoid) production by the COX-1 or COX-2
pathways, biochemical pathways similar to the 5-LO pathway
that also uses arachidonic acid as substrate. None of the tested
compounds had significant activity on the production of either
thromboxane B2 (TXB2) or prostaglandin E2 (PGE2), markers
of COX-1 or COX-2 pathway activity, respectively, indicating
≥100-fold selectivity of the compounds for the 5-LO pathway.
In Vivo Pharmacokinetic (PK) Data. The compounds
with the most promising in vitro profiles, 15b, 21, 22, and 41
were tested in rat in vivo PK studies (Table 4). In general, the
compounds had low clearance (CL), half-lives greater than 3 h,
and a steady state volume of distribution around 2 L/kg. The
most interesting tert-butyl compound, 12, was included in these
studies for reference. The in vivo rat CL for 12 was approximately
twice as high compared to most other compounds but still in the
low CL range (<50% of the liver blood flow in rat, defined as
72 mL min−1 kg−1). All compounds showed good bioavailability,
especially 15b and 41, which was in line with the high per-
meability observed in vitro.
Early dose predictions were targeting an average drug con-
centration of 3× IC50 (hWB assay), considering species dif-
ferences for protein binding (Table 5; eq 1, Supporting
Information). Dose prediction was refined to achieve a trough
concentration level corresponding to 80% inhibition of
leukotriene production in the circulation, using a one compart-
ment pharmacokinetic model assuming immediate absorption
(eq 2, Supporting Information).31 The target level of 80%
inhibition was selected based on published clinical data with
Other variations in the THF ring were more successful in
reducing this specific liability. Neither the 5,5-dimethyl THF
J
dx.doi.org/10.1021/jm501531v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Table 5. Additional in Vitro Data
c
d
e
f
g
h
mean
Hu PPB
(% free)
rat PPB
dog PPB
HH CLint
RH CLint
DH CLint
b
compd
pIC80
(% free)
(% free)
((μL/min)/(1 × 106))
((μL/min)/(1 × 106))
((μL/min)/(1 × 106))
12
6.2
6.9
7.2
7.3
7.2
1.6
10
4.8
10
5.8
13
11
1.3
13
11.4
<4
9.4
6.8
13.2
<6
18.8
5.5
15b
21
3.5
8.1
5.6
ND
12
11.5
8.3
ND
3.8
22
41
ND
5.0
2.9
5.1
a
b
ND = not determined. Experimental details for all assays are described in Supporting Information. Human whole blood inhibition assay,
c
d
e
f
concentration with 80% inhibition. Protein binding in human plasma. Protein binding in rat plasma. Protein binding in dog plasma. Intrinsic
clearance measured in human hepatocytes. Intrinsic clearance measured in rat hepatocytes. Intrinsic clearance measured in dog hepatocytes.
g
h
a
Table 6. Dog in Vivo PK Data and Allometric Human Dose Prediction
b
c
d
CL
t1/2
Vss
bioavailability
(%)
dose from dog targeting IC80 as minimum concentration
e
compd
(mL min−1 kg−1
)
(h)
(L/kg)
(fu and EH species differences considered) (mg)
12
2.2
3.3
3.2
11.3
8.1
2.0
2.3
2.7
40
100
65
240
15
5
15b
41
9.7
a
b
c
d
Experimental details including dosing regimen are described in Supporting Information. Clearance. Half-life. Volume of distribution at steady
e
state. Human dose prediction per 70 kg body weight.
Figure 3. Drug plasma concentration vs time and LTB4 production following ionophore stimulation vs time and reduction of LTB4 production
(indicating FLAP inhibition) vs plasma concentration profiles (only time points up to 24 h are shown): blue diamonds and red squares, oral dosing;
green triangles, two-step infusion.
earlier FLAP inhibitors32,33 and should also enable an
investigation of dose dependency on surrogate end points in
future clinical trials. Species differences in the hepatocyte
metabolism were observed (see Table 5), and the predicted
human CL used was corrected for species differences in hepatic
extraction (EH), based on the in vitro CLint data (see Table 5 for
additional in vitro data).34
By use of eq 1 (Supporting Information), all THF compounds
were predicted to have low doses in humans (∼10−20 mg once a
day dosing, Table 4), whereas the tert-butyl compound 12 was
likely to have a higher human dose due to its higher CL and lower
potency. Use of eq 2 (Supporting Information) resulted in a
clearer differentiation between the compounds; e.g., 22, which
seemed good enough in the simple dose prediction targeting an
average concentration, was found not to be able to cover the
required concentration during a longer period because of its
short half-life (<1.5 h in rat). The compound was also found to
possess some chemical instability in the oxadiazole ring, leading
to decomposition of the compound. Additionally, the identified
species difference in hepatic extraction added to a higher dose for
some of the compounds. Three of the compounds were selected
for dog in vivo PK studies: compounds 12, 15b, and 41 (Table 6).
Compound 21 was not pursued further because of its low solubility.
The dog PK studies indicated that both 15b and 41 had a
favorable PK profile, with good bioavailability and a long half-life.
Human dose estimates using eq 2 (Supporting Information)
adapted for dog indicated similar doses as found from rat, around
10 mg per day for compounds 15b and 41.
The compounds in this series showed limited rodent activity
but were potent in dogs, and dogs were utilized for an initial
pharmacokinetic/pharmacodynamic (PK/PD) analysis. The
collected blood samples were used to determine both the
compound’s plasma concentration and also to measure ex vivo
levels of LTB4 production following ionophore stimulation
(Supporting Information). These studies indicated that LTB4
production was inhibited in a concentration dependent manner,
and no time delay was observed for the compounds investigated.
However, in some experiments a baseline shift was noted,
especially at later time points. Figure 3 shows the plasma
concentration vs time and LTB4 production (% of predose) vs
time profiles for compound 15b for time points between 0 and
24 h. FLAP activity is clearly inhibited at early time points, where
the inhibitor concentration is high. However, a tendency for a
baseline shift can be seen for two of the experiments (LTB4
production of >100% at 24 h; later time points showed even
higher LTB4 production values). An estimation of the ex vivo
K
dx.doi.org/10.1021/jm501531v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
to 98% v/v CH3CN in H2O. All tested compounds were determined to
be ≥95% pure using the analytical method a or b described above based
on the peak area percentage.
potency of 15b using the data of the initial time points indicated
a similar ex vivo inhibition potency as determined in vitro
(pIC80 = 6.9) (Table 5).
Synthetic Procedure to AZD6642 (15b and Its Corresponding
Enantiomer 15b′): [4-(Benzyloxy)phenyl](6-bromopyridin-3-yl)-
methanone (1). 6-Bromopyridine-3-carboxylic acid (354 g, 1.75 mol)
was suspended in 2-Me-THF (2.5 L) at rt. and CDI (284 g, 1.75 mol)
was added in portions. After 1.5 h DIPEA (460 mL, 2.63 mol) was
added followed by N,O-dimethylhydroxylamine hydrochloride (171 g,
1.75 mol, finely grounded). The resulting reaction mixture was stirred
overnight. The reaction mixture was washed with 10% citric acid solu-
tion (2 × 1 L, until pH ≈ 3), 10% aqueous Na2CO3 solution (375 mL),
and finally water. The organic phase was dried (MgSO4) and the solvent
was removed in vacuo to give 6-bromo-N-methoxy-N-methylpyridine-3-
The overall properties in combination with the low predicted
dose and the target engagement for 15b shown ex vivo led to the
selection of this compound as AZD6642 for further develop-
ment.
CONCLUSION
■
Rational design with a focus on lipophilicity and LLE led to
the discovery of a novel series of FLAP inhibitors with a 2,2-
disubstituted THF ring core structure. Compared to the starting
point P-1 and the related tert-butyl analogues, a significant shift in
LLE, a measure of specific affinity, was observed for compounds
with the (R)-enantiomer of the THF ring at the central carbon
(Figure 4). This suggests that the oxygen in the THF ring is
engaged in a specific interaction with the receptor site.
1
carboxamide as an oil (289 g, 67%). H NMR (400 MHz, CDCl3) δ
8.65−8.70 (m, 1H), 7.85 (dd, J = 2.4, 8.3 Hz, 1H), 7.50 (dd, J = 0.7, 8.3
Hz, 1H), 3.50 (s, 3H), 3.32 (s, 3H); MS m/z 245 (M + H)+.
1-(Benzyloxy)-4-bromobenzene (202 g, 769 mmol), 20% solution in
THF, was heated with magnesium scrapings (26 g, 1.11 mol) to 60 °C
for 2 h to produce the Grignard reagent. 6-Bromo-N-methoxy-N-
methylpyridine-3-carboxamide (173 g, 769 mmol), prepared as
described above, was dissolved in THF (850 mL). The Grignard
solution was added to the 6-bromo-N-methoxy-N-methylpyridine-3-
carboxamide solution at 20−30 °C over a period of 1.75 h. After
completion the resulting reaction mixture was cooled to 0 °C and left
overnight. The reaction mixture was worked up using CH2Cl2, washed
with 2 × 1.5 L of citric acid, saturated NaHCO3 (aq), dried (MgSO4)
and the solvent was removed in vacuo to give compound 1 (234 g, 81%).
1H NMR (600 MHz, CDCl3) δ 8.69 (d, J = 2.0 Hz, 1H), 7.92 (dd, J = 2.4,
8.2 Hz, 1H), 7.80 (d, J = 8.8 Hz, 2H), 7.63 (d, J = 8.2 Hz, 1H), 7.39−7.46
(m, 4H), 7.33−7.38 (m, 1H), 7.06 (d, J = 8.8 Hz, 2H), 5.16 (s, 2H); MS
m/z 368 (M + H)+.
5-{2-[4-(Benzyloxy)phenyl]tetrahydrofuran-2-yl}-2-bromo-
pyridine (5). Isopropylmagnesium chloride (2 M in THF, 70.6 mL,
141.2 mmol) was added carefully to a slurry of dry 3-chloro-N,N-
dimethylpropan-1-amine hydrochloride (18.03 g, 114.1 mmol) in THF
(135 mL). Magnesium turnings (3.7 g, 152.1 mmol), activated with
iodine (1.45 g, 5.7 mmol) and gently heated with a heat gun, were added,
and the reaction mixture was heated at reflux for 1.5 h. The formed
Grignard reagent was decanted into a mixture of compound 1 (20.0 g,
54.3 mmol) in THF (315 mL). After 1.5 h the reaction was quenched by
addition of saturated NH4Cl (aq) (∼10 mL) and the mixture was
concentrated in vacuo. The residue was then partitioned between
CH2Cl2 and saturated NaHCO3 (aq), and the aqueous phase was
extracted with CH2Cl2. The combined organic phase was dried (phase
separator/Na2SO4) and concentrated in vacuo to give crude 1-[4-
(benzyloxy)phenyl]-1-(6-bromopyridin-3-yl)-4-(dimethylamino)-
butan-1-ol (24.44 g, 99%), which was used without further purifications.
MS m/z 455 (M + H)+.
Figure 4. pIC50(hWBfree) vs cLogP showing the change in ligand
lipophilic efficiency (LLE) throughout compound optimization.
Compounds are depicted according to structure (yellow circles, THF
series; green star, AZD6642 (15b); light blue diamonds, tert-butyl
series; dark blue square, starting point P-1).
Several compounds from the series with high potency and
good DMPK and physicochemical and in vitro selectivity prop-
erties could be identified which led to the selection of AZD6642
as a FLAP inhibitor for long-term treatment of inflammatory
diseases to progress into further development. In vivo PK/PD
investigations in dog with ex vivo measurement of the biomarker
LTB4 levels in plasma provide evidence for target engagement
and suggest a good correlation with the in vitro data. The
predicted human dose based on rat and dog in vivo PK data is
suggested to be in the range of 15−25 mg once daily dosing for
AZD6642 (15b).
Sodium hydride in mineral oil (7.08 g, 177.11 mmol) was added to a
solution of crude 1-[4-(benzyloxy)phenyl]-1-(6-bromopyridin-3-yl)-4-
(dimethylamino)butan-1-ol (24.44 g, 53.67 mmol), prepared as above,
in p-xylene (350 mL). The reaction mixture was heated to reflux and
kept at that temperature for 10 min. Then 1-bromobutane (14.71 mL,
139.54 mmol) was added and the mixture heated at reflux for 1 h. The
reaction mixture was allowed to reach rt, and the reaction was then
quenched by addition of 3.8 M HCl (aq) (∼5 mL). The mixture was
then concentrated in vacuo. The residue was partitioned between
EtOAc and saturated NaCl (aq) with a pH adjusted to ∼5 using 3.8 M
HCl (aq). The organic phase was washed with saturated NH4Cl (aq)
and saturated NaCl (aq), dried (Na2SO4), and concentrated in vacuo.
The residue was purified by flash column chromatography (0% → 20%
EXPERIMENTAL SECTION
■
Extensive experimental information can be found in Supporting
Information. Analytical HPLC/MS was conducted on a QTOF mass
spectrometer using a UV detector monitoring either at (a) 210 nm with
a BEH C18 column (2.1 mm × 100 mm, 1.7 μm, 0.7 mL/min flow rate),
using a gradient of 2% v/v CH3CN in H2O (ammonium carbonate
buffer, pH 10) to 98% v/v CH3CN in H2O, or at (b) 230 nm with an
HSS C18 column (2.1 × 100 mm, 1.8 μm, 0.7 mL/min flow rate), using
a gradient of 2% v/v CH3CN in H2O (ammonium formate buffer, pH 3)
1
EtOAc in heptane) to give compound 5 (16.42 g, 75%). H NMR
(500 MHz, CDCl3) δ 8.41−8.45 (m, 1H), 7.57 (dd, J = 2.6, 8.3 Hz, 1H),
7.36−7.44 (m, 5H), 7.28−7.35 (m, 3H), 6.9−6.96 (m, 2H), 5.04 (s,
2H), 4.04 (t, J = 7.1 Hz, 2H), 2.55−2.65 (m, 1H), 2.39−2.48 (m, 1H),
1.89−2.08 (m, 2H); 13C NMR (126 MHz, CDCl3) δ 158.0, 147.8, 141.7,
140.3, 137.1, 136.9, 136.5, 128.6, 128.0, 127.5, 127.4, 126.9, 114.7, 86.0,
70.0, 67.6, 38.6, 25.4. MS m/z 410 (M + H)+.
L
dx.doi.org/10.1021/jm501531v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Methyl 5-{2-[4-(Benzyloxy)phenyl]tetrahydrofuran-2-yl}-
pyridine-2-carboxylate (6). CO(g) was bubbled through a stirred
solution of 5 (7.00 g, 17.06 mmol) in DMF (100 mL), MeOH (100 mL),
and Et3N (20 mL) at rt for 10 min. PdCl2(dppf) (1.752 g, 2.17 mmol)
(complex with CH2Cl2 1:1) was then added, and the mixture was heated
at70 °C underCOatmosphere (balloon) overnight. The reaction mixture
was then concentrated in vacuo, and the residue was partitioned between
CH2Cl2 and saturated NaCl (aq). The aqueous phase was extracted with
CH2Cl2 and the combined organic phase was dried (phase separator) and
concentrated in vacuo. The residue was purified by flash column
chromatography (0% → 30% EtOAc in toluene) to give compound 6
(4.71 g, 71%). 1H NMR (600 MHz, CDCl3) δ 8.79 (d, J = 1.8 Hz, 1H),
8.03 (d, J = 8.2 Hz, 1H), 7.85 (dd, J = 2.2, 8.2 Hz, 1H), 7.34−7.42 (m,
4H), 7.28−7.33 (m, 3H), 6.90 (d, J = 8.8 Hz, 2H), 5.02 (s, 2H), 4.04 (t,
J = 7.1 Hz, 2H), 3.97 (s, 3H), 2.59−2.69 (m, 1H), 2.42−2.51 (m, 1H),
1.88−2.06 (m, 2H); 13C NMR (126 MHz, CDCl3) δ 165.6, 158.0, 147.6,
146.1, 146.0, 136.9, 136.8, 134.3, 128.9, 128.6, 128.0, 127.4, 127.0, 124.7,
114.7, 114.6, 86.2, 70.0, 67.6, 52.8, 38.6, 25.4. MS m/z 390 (M + H)+.
N-(2-Hydroxy-2-methylpropyl)-5-[2-(4-hydroxyphenyl)-
tetrahydrofuran-2-yl]pyridine-2-carboxamide (17a). A solution
of lithium hydroxide (0.259 g, 10.83 mmol) in water (12 mL) was added
to a suspension of compound 6 (3.51 g, 9.03 mmol) in THF (30.0 mL)
and MeOH (30.0 mL). The reaction mixture was stirred at rt overnight
(∼16 h). The reaction mixture was then concentrated in vacuo, and
the residue was partitioned between EtOAc and ∼0.05 M HCl (aq).
The aqueous phase was extracted with CH2Cl2 (×4), and the com-
bined organic phase was dried (Na2SO4 and phase separator) and
concentrated in vacuo to give crude 5-{2-[4-(benzyloxy)phenyl]-
tetrahydrofuran-2-yl}pyridine-2-carboxylic acid which was used without
further purifications. MS m/z 376 (M + H)+.
TBTU (3.77 g, 11.74 mmol) followed by a solution of 1-amino-2-
methylpropan-2-ol (2.42 g, 27.09 mmol) in CH2Cl2 (30 mL) was added
to a solution of crude 5-{2-[4-(benzyloxy)phenyl]tetrahydrofuran-2-yl}-
pyridine-2-carboxylic acid (3.39 g, 9.03 mmol) and 4-methylmorpholine
(1.99 mL, 18.06 mmol) in DMF (75 mL). The reaction mixture was
stirred at rt overnight. The reaction mixture was then partially evaporated,
and the residue was partitioned between EtOAc and 0.1 M HCl (aq). The
organic phase was washed with saturated NaHCO3 (aq), dried (Na2SO4),
and concentrated in vacuo. The residue was purified by flash column
chromatography (5% EtOH in toluene) to give 5-{2-[4-(benzyloxy)-
phenyl]tetrahydrofuran-2-yl}-N-(2-hydroxy-2-methylpropyl)pyridine-2-
carboxamide (4.04 g, 100%). 1H NMR (600 MHz, CDCl3) δ 8.61 (d, J =
1.6 Hz, 1H), 8.36 (t, J = 6.2 Hz, 1H), 8.07−8.13 (m, 1H), 7.87 (dd, J =
2.2, 8.2 Hz, 1H), 7.35−7.42 (m, 4H), 7.29−7.34 (m, 3H), 6.92 (d, J = 8.8
Hz, 2H), 5.03 (s, 2H), 4.01−4.11 (m, 2H), 3.47 (d, J = 6.4 Hz, 2H), 2.6−
2.71 (m, 1H), 2.52 (s, 1H), 2.42−2.51 (m, 1H), 1.9−2.06 (m, 2H), 1.27
(s, 6H). MS m/z 447 (M + H)+.
and the reaction mixture was then added to saturated NaHCO3 (aq)
(200 mL). The phases were separated, and the aqueous phase was
extracted with CH2Cl2. The combined organic phase was dried (phase
separator) and concentrated in vacuo. The residue was purified by
flash column chromatography (70% → 100% EtOAc in heptane) to
give 4-(2-{6-[(2-hydroxy-2-methylpropyl)carbamoyl]pyridin-3-yl}-
tetrahydrofuran-2-yl)phenyl trifluoromethanesulfonate (3.53 g, 90%).
1H NMR (600 MHz, CDCl3) δ 8.62 (d, J = 1.6 Hz, 1H), 8.35 (t, J =
6.3 Hz, 1H), 8.14 (d, J = 8.2 Hz, 1H), 7.89 (dd, J = 2.3, 8.2 Hz, 1H),
7.48−7.55 (m, 2H), 7.2−7.24 (m, 2H), 4.02−4.12 (m, 2H), 3.47 (d, J =
6.4 Hz, 2H), 2.60 (t, J = 7.2 Hz, 2H), 2.41 (s, 1H), 1.92−2.09 (m, 2H),
1.28 (s, 6H). MS m/z 489 (M + H)+.
Potassium acetate (3.55 g, 36.1 mmol), 4,4,4′,4′,5,5,5′,5′-octamethyl-
2,2′-bi(1,3,2-dioxaborolane) (2.39 g, 9.4 mmol), and PdCl2(dppf)
(1.05 g, 1.44 mmol) were added to a solution of 4-(2-{6-[(2-hydroxy-2-
methylpropyl)carbamoyl]pyridin-3-yl}tetrahydrofuran-2-yl)phenyl tri-
fluoromethanesulfonate (3.53 g, 7.2 mmol) in degassed dioxane (30 mL)
under Ar(g). The reaction mixture was heated at reflux for 50 min and was
then partitioned between EtOAc and H2O, and the aqueous phase was
extracted with EtOAc (×2). The combined organic phase was dried
(Na2SO4) and concentrated in vacuo. CH2Cl2 was added to the residue
and the mixture was filtered (Celite) and the filtrate was concentrated in
vacuo to give crude N-(2-hydroxy-2-methylpropyl)-5-{2-[4-(4,4,5,5-
tetramethyl-1,3,2-dioxaborolan-2-yl)phenyl]tetrahydrofuran-2-yl}-
pyridine-2-carboxamide, which was used without further purifications. MS
m/z 467 (M + H)+.
A solution of potassium carbonate (2.0 g, 14.4 mmol) in degassed
water (60 mL) was added to a solution of crude N-(2-hydroxy-2-
methylpropyl)-5-{2-[4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)-
phenyl]tetrahydrofuran-2-yl}pyridine-2-carboxamide (7.2 mmol), 5-
bromopyrazin-2-amine (1.51 g, 8.7 mmol), and PdCl2(dppf) (0.52 g,
0.72 mmol) in dioxane (60 mL). The reaction mixture was heated at
reflux for 50 min. The reaction mixture was then diluted with EtOAc and
filtered (Celite), and the filtrate was washed with saturated NaCl (aq).
The aqueous phase was extracted with EtOAc, and the combined
organic phases were dried (Na2SO4) and concentrated in vacuo. The
residue was purified by flash column chromatography (0% → 10%
EtOH in EtOAc) and by preparative HPLC (XBridge C18, 20% → 80%
MeCN in 0.2% NH3 (aq)) to give the racemate of the title compounds
(2.45 g, 78%). 1H NMR (400 MHz, DMSO-d6) δ 8.77 (d, J = 1.8 Hz,
1H), 8.46 (s, 1H), 8.41 (t, J = 6.1 Hz, 1H), 8.06 (dd, J = 2.1, 8.2 Hz, 1H),
7.99 (d, J = 8.2 Hz, 1H), 7.93 (d, J = 1.0 Hz, 1H), 7.86 (d, J = 8.4 Hz,
2H), 7.53 (d, J = 8.4 Hz, 2H), 6.55 (s, 2H), 4.68 (s, 1H), 4.00 (t, J =
7.1 Hz, 2H), 3.26 (d, J = 6.1 Hz, 2H), 2.58−2.72 (m, 2H), 1.84−2
(m, 2H), 1.09 (s, 6H). 13C NMR (126 MHz, DMSO-d6) δ 163.3, 154.9,
148.2, 145.9, 144.7, 144.2, 138.8, 138.6, 135.9, 134.7, 131.4, 125.7, 124.7,
121.4, 86.1, 69.1, 67.0, 49.6, 37.8, 27.2, 24.9. HRMS (ESI) m/z calcd for
C24H28N5O3 [M + H]+ 434.2192; found 434.2177.
Palladium hydroxide (0.317 g, 0.45 mmol, 5% on carbon) was added
to a solution of 5-{2-[4-(benzyloxy)phenyl]tetrahydrofuran-2-yl}-N-(2-
hydroxy-2-methylpropyl)pyridine-2-carboxamide (4.03 g, 9.03 mmol)
in MeOH (100 mL), and the reaction mixture was hydrogenated at
1 atm for 3.5 h. The catalyst was then filtered off, washed with MeOH,
and the filtrate was concentrated in vacuo. The residue was purified by
preparative HPLC (C8, 15% → 75% MeCN in 0.2% CH3CO2H (aq)).
Fractions containing the desired product were partially concentrated,
and the aqueous mixture was extracted with CH2Cl2 and then EtOAc
(×4). The combined organic phases were dried (Na2SO4) and con-
centrated in vacuo to give compound 17a (2.86 g, 89%). 1H NMR (400
MHz, CD3OD) δ 8.66 (d, J = 1.4 Hz, 1H), 8.00 (d, J = 8.2 Hz, 1H), 7.96
(dd, J = 2.2, 8.2 Hz, 1H), 7.22−7.30 (m, 2H), 6.68−6.76 (m, 2H), 3.94−
4.11 (m, 2H), 3.40 (s, 2H), 2.60−2.77 (m, 1H), 2.38−2.57 (m, 1H),
1.86−2.06 (m, 2H), 1.21 (s, 6H); 13C NMR (126 MHz, MeOD) δ
166.7, 157.7, 149.0, 147.7, 147.3, 137.0, 135.9, 128.1, 122.5, 116.1, 87.8,
71.4, 68.5, 50.9, 39.5, 27.2, 26.4; MS m/z 357 (M + H)+.
The enantiomers of the racemate (2.45 g, 5.7 mmol) were separated
by chiral chromatography on a Chiralcel OJ 250 mm × 50 mm, 20 μm
HPLC column. 125 mg (25 mg/mL in MeOH/CH3CN/HCO2H 12/
6/1) was injected and eluted with EtOH at a flow rate of 120 mL/min
and detected at 323 nm. The first eluted compound was collected
and evaporated to yield the most active enantiomer 15b (1.18 g, 98.1% ee).
[α]2D0 −23.5° (c 1.0, CH3OH). HRMS (ESI) m/z calculated for
C24H28N5O3 [M + H]+ 434.2192; found 434.2211. The second eluted
compound was collected and evaporated to yield the least active
enantiomer 15b′ (1.1 g, 99.7% ee). [α]2D0 +23.2° (c 1.0, CH3OH).
HRMS (ESI) m/z calculated for C24H28N5O3 [M + H]+ 434.2192;
found 434.2207.
ASSOCIATED CONTENT
* Supporting Information
■
S
Detailed synthetic chemistry procedures, absolute configuration
determination with VCD calculations and data generated from
X-ray, FLAP binding assay, human whole blood LTB4 assay,
whole blood COX-1 and COX-2 assays, assays related to dmpk
and safety profiling (HPLC log D, solubility, CYP inhibition,
hepatocyte CLint (human, rat, dog), Caco2 Papp assay, RM
5-{(2R and S)-2-[4-(5-Aminopyrazin-2-yl)phenyl]tetra-
hydrofuran-2-yl}-N-(2-hydroxy-2-methylpropyl)pyridine-2-
carboxamide (15b and 15b′). Trifluoromethanesulfonic anhydride
(1.7 mL, 10.4 mmol) was slowly added to a solution of compound 17a
(2.9 g, 8.01 mmol) and DMAP (4.9 g, 40.05 mmol) at −78 °C in
CH2Cl2 (130 mL). The reaction mixture was stirred at −78 °C for 7 h,
M
dx.doi.org/10.1021/jm501531v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Hutchinson, J. H.; Prasit, P.; Evans, J. F. Pharmacology of AM803, a
novel selective five-lipoxygenase-activating protein (FLAP) inhibitor in
rodent models of acute inflammation. Eur. J. Pharmacol. 2010, 640,
211−218.
(9) Ogawa, A.; Ujjainwalla, F.; Vande Bunte, E. K.; Chu, L.; Ondeyka,
D.; Kopka, I.; Li, B.; Ok, H.; Patel, M. J.; Xu, J.; Sisco, R.
WO2008030369 A1, 2008.
trapping (GSH and CH3ONH2), CYP reaction profiling
(CYP3A4 contribution), plasma protein binding (human, rat,
dog), hERG inhibition, inhibition of other cardiac ion channels
and AMES), in vivo PK experimental details and dose prediction
equations. This material is available free of charge via the Internet
at http://pubs.acs.org.
(10) Ogawa, A.; Ujjainwalla, F.; Vande Bunte, E. K.; Chu, L.; Ondeyka,
D.; Kopka, I. E.; Li, B.; Ok, H. O.; Patel, M.; Sisco, R. WO2008156721
A1, 2008.
(11) Ogawa, A.; Ujjainwalla, F.; Li, B.; Chu, L.; Patel, M.; Armstrong,
H.; Ok, H. O.; Sisco, R.; MacDonald, D.; Hutchinson, J.; Perrier, H.
WO2009048547 A1, 2009.
(12) Evans, J. F.; Ferguson, A. D.; Mosley, R. T.; Hutchinson, J. H.
What’s all the FLAP about?: 5-Lipoxygenase-activating protein
inhibitors for inflammatory diseases. Trends Pharmacol. Sci. 2008, 29,
72−78.
(13) Ferguson, A. D.; McKeever, B. M.; Xu, S.; Wisniewski, D.; Miller,
D. K.; Yamin, T.; Spencer, R. H.; Chu, L.; Ujjainwalla, F.; Cunningham,
B. R.; Evans, J. F.; Becker, J. W. Crystal structure of inhibitor-bound
human 5-lipoxygenase-activating protein. Science 2007, 317, 510−512.
(14) Leeson, P. D.; Springthorpe, B. The influence of drug-like
concepts on decision-making in medicinal chemistry. Nat. Rev. Drug
Discovery 2007, 6, 881−890.
(15) Gleeson, M. P.; Hersey, A.; Montanari, D.; Overington, J. Probing
the links between in vitro potency, ADMET and physicochemical
parameters. Nat. Rev. Drug Discovery 2011, 10, 197−208.
(16) Peters, J.; Schnider, P.; Mattei, P.; Kansy, M. Pharmacological
promiscuity: dependence on compound properties and target specifity
in a set of recent Roche Ccompounds. ChemMedChem 2009, 4, 680−
686.
(17) Chambers, R. J.; Marfat, A. Regiospecific carboalkoxylation of 2,5-
dibromopyridine. Synth. Commun. 1997, 27, 515−520.
(18) Marxer, A. Ringslutning mha NaH/butylbromide. Helv. Chim.
Acta 1969, 52, 262−270.
AUTHOR INFORMATION
Corresponding Author
*E-mail: malin.lemurell@astrazeneca.com. Phone: +46730667781.
Notes
■
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors thank Gabrielle Saarinen and David Blomberg for
synthetic chemistry support and the Separation Science
Laboratory and Structure Analysis Support at AstraZeneca
Molndal for help with compound purification and analysis. For
development of biological assays and generation of screening
results the authors acknowledge Bruno Becker, Alexander Vidal,
̈
Eva Carlsson, Johan Meuller, Magdalena Rhedin, Anita Dellsen
and Lisa Jinton.
́
,
ABBREVIATIONS USED
■
cysLT, cysteinyl leukotriene; COX, cyclooxygenase; DAST,
diethylaminosulfur trifluoride; EH, hepatic extraction; FLAP,
5-lipoxygenase activating protein; GSH, glutathione; hWB, human
whole blood; LTA4, leukotriene A4; LTB4, leukotriene B4; 5-LO,
5-lipoxygenase; NMM, N-methylmorpholine; PGE2, prostaglan-
din E2; rac, racemic; TBTU, O-(benzotriazol-1-yl)-N,N,N′,N′-
tetramethyluronium tetrafluoroborate; TXB2, thromboxane B2
(19) Gierczyk, B.; Nowak-Wydra, B.; Schroeder, G. Oxadiazole
synthesis. Magn. Reson. Chem. 2003, 41, 689−692.
(20) El-Azzounyl, A. A.; Maklad, Y. A.; Bartsch, H.; Zaghary, W. A.;
Ibrahim, W. M.; Mohamed, M. S. Hydroxyoxadiazoles. Sci. Pharm. 2003,
71, 331−356.
REFERENCES
■
(1) Peters-Golden, M. Cell biology of the 5-lipoxygenase pathway. Am.
J. Respir. Crit. Care Med. 1998, 157, 227−232, 247−248.
(2) Sala, A.; Folco, G.; Murphy, R. C. Transcellular biosynthesis of
eicosanoids. Pharmacol. Rep. 2010, 62, 503−510.
(21) Ogawa, A.; Ujjainwalla, F.; Vande Bunte, E. K.; Chu, L.; Ondeyka,
D.; Kopka, I. E.; Li, B.; Ok, H. O.; Patel, M.; Sisco, R. Diphenyl
Substituted Alkanes. WO2008/156721 A1, 2008; Merck and Co.
(22) Vozza, J. F. ringslutningar under sura betingelser av difenyl-
butandioler till difenyl-THF. J. Org. Chem. 1959, 24, 720−722.
(23) Gabrielsson, J.; Dolgos, H.; Gillberg, P.; Bredberg, U.; Benthem,
B.; Duker, G. Early integration of pharmacokinetic and dynamic
reasoning is essential for optimal development of lead compounds:
strategic considerations. Drug Discovery Today 2008, 14, 358−372.
(24) Green, A. R.; Gabrielsson, J.; Marsden, C. A.; Fone, K. C. F.
MDMA: on the translation from rodent to human dosing.
Psychopharmacology (Berlin, Ger.) 2009, 204, 375−378.
(25) Yang, C.; Hasselgren, C. H.; Boyer, S.; Arvidson, K.; Aveston, S.;
Dierkes, P.; Benigni, R.; Benz, R. D.; Contrera, J.; Kruhlak, N. L.;
Matthews, E. J.; Han, X.; Jaworska, J.; Kemper, R. A.; Rathman, J. F.;
Richard, A. M. Understanding genetic toxicity through data mining: the
process of building knowledge by integrating multiple genetic toxicity
databases. Toxicology Mech. Methods 2008, 18, 277−295.
(26) Kazius, J.; McGuire, R.; Bursi, R. Derivation and validation of
toxicophores for mutagenicity prediction. J. Med. Chem. 2005, 48, 312−
320.
(3) Folco, G.; Murphy, R. C. Eicosanoid transcellular biosynthesis:
from cell-cell interactions to in vivo tissue responses. Pharmacol. Rev.
2006, 58, 375−388.
(4) Matsuse, H.; Kohno, S. Leukotriene receptor antagonists
pranlukast and montelukast for treating asthma. Expert Opin.
Pharmacother. 2014, 15, 353−363.
(5) Whatling, C.; McPheat, W.; Hersloef, M. The potential link
between atherosclerosis and the 5-lipoxygenase pathway: investigational
agents with new implications for the cardiovascular field. Expert Opin.
Invest. Drugs 2007, 16, 1879−1893.
(6) Hakonarson, H.; Thorvaldsson, S.; Helgadottir, A.; Gudbjartsson,
D.; Zink, F.; Andresdottir, M.; Manolescu, A.; Arnar, D. O.; Andersen,
K.; Sigurdsson, A.; Thorgeirsson, G.; Jonsson, A.; Agnarsson, U.;
Bjornsdottir, H.; Gottskalksson, G.; Einarsson, A.; Gudmundsdottir, H.;
Adalsteinsdottir, A. E.; Gudmundsson, K.; Kristjansson, K.; Hardarson,
T.; Kristinsson, A.; Topol, E. J.; Glucher, J.; Kong, A.; Gurney, M.;
Thorgeirsson, G.; Stefansson, K. Effects of a 5-lipoxygenase-activating
protein inhibitor on biomarkers associated with risk of myocardial
infarction. A randomized trial. JAMA, J. Am. Med. Assoc. 2005, 293,
2245−2256.
(7) Tardif, J.; L’allier, P. L.; Ibrahim, R.; Gregoire, J. C.; Nozza, A.;
Cossette, M.; Kouz, S.; Lavoie, M.; Paquin, J.; Brotz, T. M.; Taub, R.;
Pressacco, J. Treatment with 5-lipoxygenase inhibitor VIA-2291
(atreleuton) in patients with recent acute coronary syndrome. Circ.
Cardiovasc. Imaging 2010, 3, 298−307.
(8) Lorrain, D. S.; Bain, G.; Correa, L. D.; Chapman, C.; Broadhead, A.
R.; Santini, A. M.; Prodanovich, P. P.; Darlington, J. V.; Stock, N. S.;
Zunic, J.; King, C. D.; Lee, C.; Baccei, C. S.; Stearns, B.; Roppe, J.;
(27) Maron, D. M.; Ames, B. N. Revised methods for the Salmonella
mutagenicity test. Mutat. Res., Environ. Mutagen. Relat. Subj. 1983, 113,
173−215.
(28) Zhang, L.; Zhang, Y. D.; Zhao, P.; Huang, S. M. Predicting drug-
drug interactions: an FDA perspective. AAPS J. 2009, 11, 300−306.
(29) Baillie, T. A. Metabolism and toxicity of drugs. Two decades of
progress in industrial drug metabolism. Chem. Res. Toxicol. 2008, 21,
129−137.
N
dx.doi.org/10.1021/jm501531v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
(30) Baillie, T. A. The use of stable isotopes in pharmacological
research. Pharmacol. Rev. 1981, 33, 81−132.
(31) McGinnity, D. F.; Collington, J.; Austin, R. P.; Riley, R. J.
Evaluation of human pharmacokinetics, therapeutic dose and exposure
predictions using marketed oral drugs. Curr. Drug Metab. 2007, 8, 463−
479.
(32) Uematsu, T.; Kanamaru, M.; Kosuge, K.; Hara, K.; Uchiyama, N.;
Takenaga, N.; Tanaka, W.; Friedman, B. S.; Nakashima, M.
Pharmacokinetic and pharmacodynamic analysis of a novel leukotriene
biosynthesis inhibitor, MK-0591, in healthy volunteers. Br. J. Clin.
Pharmacol. 1995, 40, 59−66.
(33) Bain, G.; King, C. D.; Schaab, K.; Rewolinski, M.; Norris, V.;
Ambery, C.; Bentley, J.; Yamada, M.; Santini, A. M.; van de Wetering de
Rooij, J.; Stock, N.; Zunic, J.; Hutchinson, J. H.; Evans, J. F.
Pharmacodynamics, pharmacokinetics and safety of GSK2190915, a
novel oral anti-inflammatory 5-lipoxygenase-activating protein inhibitor.
Br. J. Clin. Pharmacol. 2013, 75, 779−790.
(34) Obach, R. S.; Baxter, J. G.; Liston, T. E.; Silber, B. M.; Jones, B. C.;
Macintyre, F.; Rance, D. J.; Wastall, P. The prediction of human
pharmacokinetic parameters from preclinical and in vitro metabolism
data. J. Pharmacol. Exp. Ther. 1997, 283, 46−58.
O
dx.doi.org/10.1021/jm501531v | J. Med. Chem. XXXX, XXX, XXX−XXX