Structure- and property-based design of aminooxazoline xanthenes as selective, orally efficacious, and cns penetrable BACE inhibitors for the treatment of alzheimers disease

Article abstract of DOI:10.1021/jm300598e

A structure- and property-based drug design approach was employed to identify aminooxazoline xanthenes as potent and selective human β-secretase inhibitors. These compounds exhibited good isolated enzyme, cell potency, and selectivity against the structurally related aspartyl protease cathepsin D. Our efforts resulted in the identification of a potent, orally bioavailable CNS penetrant compound that exhibited in vivo efficacy. A single oral dose of compound 11a resulted in a significant reduction of CNS Aβ40 in naive rats.

Full text of DOI:10.1021/jm300598e

Article
pubs.acs.org/jmc
Structure- and Property-Based Design of Aminooxazoline Xanthenes
as Selective, Orally Efficacious, and CNS Penetrable BACE Inhibitors
for the Treatment of Alzheimer’s Disease
Hongbing Huang,*,†,∞ Daniel S. La,*,†,∞ Alan C. Cheng,‡,∞ Douglas A. Whittington,‡ Vinod F. Patel,†,△
Kui Chen,⊥ Thomas A. Dineen,† Oleg Epstein,† Russell Graceffa,† Dean Hickman,∥ Y.-H. Kiang,#
Steven Louie,∥ Yi Luo,§ Robert C. Wahl,⊥ Paul H. Wen,§ Stephen Wood,§ and Robert T. Fremeau, Jr.§
‡
†Department of Medicinal Chemistry and Department of Molecular Structure, Amgen Inc., 360 Binney Street, Cambridge,
Massachusetts 02142, United States
∥
⊥
§Department of Neuroscience, Department of Pharmacokinetics and Drug Metabolism, Department of HTS and Molecular
#
Pharmacology, and Department of Pharmaceutics, Amgen Inc., One Amgen Center Drive, Thousand Oaks, California 91320,
United States
S
* Supporting Information
ABSTRACT: A structure- and property-based drug design
approach was employed to identify aminooxazoline xanthenes
as potent and selective human β-secretase inhibitors. These
compounds exhibited good isolated enzyme, cell potency, and
selectivity against the structurally related aspartyl protease
cathepsin D. Our efforts resulted in the identification of a
potent, orally bioavailable CNS penetrant compound that
exhibited in vivo efficacy. A single oral dose of compound 11a
resulted in a significant reduction of CNS Aβ40 in naive rats.
attention as a potential disease-modifying treatment of AD.5 In
addition, the observation that BACE knockout mice lacking β-
INTRODUCTION
■
Alzheimer’s disease (AD) is the leading cause of dementia and
represents a substantial unmet medical need.1 AD is a chronic
secretase activity develop normally and display no profound
phenotype reinforces the strategy of BACE inhibition as a safe
neurodegenerative disorder characterized by progressive loss of
brain function, affecting memory, language skills, and bodily
functions, and inevitably leads to incapacitation and death. The
hallmarks of AD pathology include extracellular amyloid
plaques primarily containing aggregated β-amyloid (Aβ)
peptides derived from the β-amyloid precursor protein
approach to the treatment and prevention of AD.6
Despite the identification of BACE over a decade ago and the
relatively large number of preclinical programs within both
academia and industry, only a handful of compounds have
advanced into the clinic,7 underscoring the difficulties with
generating a safe, potent inhibitor with druglike properties
suitable for targeting the central nervous system (CNS). In the
arena of small molecule inhibitors of BACE, compounds from
Merck, Eisai, Eli Lilly, and CoMentis have entered clinical
trials.8 The localization of BACE protein in acidic intracellular
compartments poses a significant challenge for transition-state-
based BACE inhibitors, as many potent enzyme inhibitors fail
to show activity in cellular assays, presumably because of poor
permeability or lack of sufficient binding affinity under acidic
conditions.9 Furthermore, its site of action in the CNS makes
effective in vivo inhibition of BACE a pharmacokinetic
challenge.10 Several programs from the early 2000s were
based on hydroxyethylamine (HEA) and hydroxyethylene
(HE) structures11 which generally suffer from low brain
(APP), and intracellular neurofibrillary tangles containing
hyperphosphorylated tau proteins.2
The amyloid hypothesis states that production of Aβ
peptides from APP is responsible for the observed disease
pathology.3 The cleavage of APP by β-secretase (BACE)
generates a soluble version of APP (sAPPβ) and a resultant
membrane-bound C-terminal domain. Subsequent intramem-
brane proteolysis of the C-terminal domain by γ-secretase
produces amyloidogenic Aβ40 and Aβ42 peptides. Aβ42 is prone
to self-assembly into fibrils and is the major Aβ component in
amyloid plaques. The N-terminal domain of APP (N-APP),
produced by further cleavage of sAPPβ, has been shown to
trigger neuron degeneration via caspases.4 Therefore, an
increased production of Aβ caused by the dysregulation of
APP processing could induce cell death processes and disturb
axonal transport. These processes ultimately lead to defects in
cognition, synaptic plasticity, and the development of
subsequent tau pathology. Given the critical role of BACE in
APP metabolism, inhibition of BACE has gained considerable
Special Issue: Alzheimer's Disease
Received: April 30, 2012
© XXXX American Chemical Society
A
dx.doi.org/10.1021/jm300598e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 1. Calculated physiochemical properties of catalytic aspartate binding cores. PSA values were calculated in ChemDraw (version 9.0). pKa
values of structures shown were calculated using ACD LogD Suite (version 10, Advanced Chemistry Development Inc., Toronto, Canada).
Table 1. SAR of 2-Aminooxazoline BACE Inhibitors
a
IC50 (μM)
b
c
cd
,
compd
Ar
R
BACE
Cat D
cellular
HLM/RLM (μL min−1 mg−1
)
Papp (10−6 cm/s)
ER
0.6
1
2
3
4
5
6
7
5-Cl-2-F-Ph
3-Cl-Ph
Me
1.95 0.10
2.06 0.16
7.46 0.57
11.4 0.35
8.20 0.66
2.85 0.17
1.21 0.67
10.7 3.75
12.7 0.19
>100
12
54/>399
39/>399
110/>399
54/300
4.2
3.4
36
40
32
15
Me
18
0.7
12
pyridin-3-yl
pyrimidin-5-yl
6-F-pyr-3-yl
2-F-pyr-3-yl
(R)-2-F-pyr-3-yl
Me
5.5
13
Me
>100
8.8
5.7
1.6
1.4
Me
47.7 6.41
12.6 0.63
7.65 5.68
8.3
5.1
1.7
51/280
Me
165/>399
36/38
CHF2
12
a
b
Values represent the mean values of at least two experiments SD. Values without SD are for a single determination only. Compounds were
c
d
incubated with microsomes for 30 min at 1 μM. Parental LLC-PK1 cell line. B/A to A/B ratio.
penetration, poor oral bioavailability, and susceptibility to p-
glycoprotein (PGP) transport.5
toward rational design of aminooxazoline xanthenes as potent,
orally efficacious, and CNS penetrant inhibitors of BACE.
More recently, 2-aminoheterocyclic inhibitors of BACE have
emerged as promising leads.12 These classes of compounds are
reliant on a two-point contact between an “amidine” type
functional group and the catalytic aspartates (32 and 228) of
the active site (see boxed structure of Figure 1). Publicly
disclosed examples include compounds with aminopyridines,
aminohydantoins, and aminoquinolines as the Asp-binding
motif.13 While these compounds tend to have lower molecular
weight compared to HEA inhibitors, they often suffer from
PGP efflux, poor brain exposure, and undesired activity on the
hERG ion channel. Herein, we describe our initial efforts
RESULTS AND DISCUSSION
■
Our initial design approach took into consideration computed
properties that could improve CNS exposure: polar surface area
(PSA), pKa, and molecular weight.14,15 As shown in Figure 1,
computational analysis of core Asp binding moieties from both
our internal efforts and the literature revealed that PSA values
range from 38 to 67 Å2 and ACD calculated pKa values range
from 4.9 to 9.0.16 While the pKa values of the core structures
will be affected by substitution, the calculated values aided
prioritization of core structures relative to each other.17
B
dx.doi.org/10.1021/jm300598e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Presumably, the high pKa values exhibited by these 2-
aminoheterocycles were designed to complement the acidity
of the catalytic aspartates18 and to accommodate the acidic (pH
≈ 4.5) operating environment of BACE.19 Studies at Merck
have demonstrated that pKa plays a critical role in enzymatic
binding and cellular potency.20 They demonstrated that an
inhibitor containing the more basic 1-methyl-2-aminoimidazole
(H) exhibited improved potency relative to that bearing an
aminothiazole (G).20a The improvement was proposed to be a
result of increasing pKa which should be on the order of ∼3.6
log units. Design criteria, however, also targeted heterocycles
with pKa values of <10 in order to achieve high oral
bioavailability.21 We were aware that basic amidines often
required a prodrug strategy in order to achieve oral
bioavailability.22
Mindful of these considerations, the oxazoline Asp-binding
group (L, Figure 1) was chosen for further evaluation based on
2
its low molecular weight (86 amu), low PSA (48 Å ), and
relatively high pKa (8.7). In addition, the hydrogen bond donor
(HBD) count was limited to 2 in consideration of the
properties required for good CNS exposure. Furthermore, in
contrast to cores D−I, the sp3 hybridization of the center
adjacent to the “amidine” functional group (red in Figure 1) in
the oxazoline ring allowed for stereocontrolled projection of
substituents into S1 and S2′ (vide infra).
A series of racemic biaryl spiro-aminooxazolines was
prepared (Table 1). All compounds were evaluated for their
inhibitory activity against BACE in a homogeneous fluores-
cence resonance energy transfer (FRET) assay. We sought to
attain a high level of selectivity against other aspartyl proteases
in order to develop BACE inhibitors with a good safety profile.
As an initial test of selectivity, we evaluated compounds against
the closely related aspartyl protease cathepsin D (Cat D). For
advanced compounds, selectivity screening across a broader
panel of aspartyl proteases was undertaken. The cellular
potency of the compounds was assessed by measuring the
inhibition of Aβ40 production in HEK293 cells expressing
amyloid precursor protein. Initial SAR studies revealed that
only modest potency was achievable with various substituted
aryl rings. In general, substituted phenyl analogues exhibited no
cellular potency with a few exceptions. IC50 values in the single
digit micromolar range were observed with aryl rings containing
a 2-fluoro substituent (compounds 1 and 6). In addition, single
digit micromolar cell potencies were observed with the
introduction of a pyridyl ring, presumably because of
productive interactions with the water H-bond found in the
S3 pocket (shown in Figure 2). The data also indicated that Cat
D selectivity was potentially emerging. For example, greater
than 10-fold selectivity was observed with compound 3.
Most of the compounds within this series demonstrated high
clearance when incubated with rat liver microsomes and poor
permeability in LLC-PK1 cells as represented by compounds 1
and 2 (Table 1). By comparison, compound 6 showed
improved permeability while maintaining a low efflux ratio.
Unfortunately, the microsomal turnover of 6 remained high. A
vast improvement in RLM metabolic stability was observed
with compound 7. The stability imparted by the difluorome-
thoxy modification suggests that demethylation is the main
route of metabolism during RLM incubation for compounds
1−6. Encouragingly, 7 showed good permeability and PGP
efflux was low (ratio of 1.4). In addition 7 was chemically stable
under acidic conditions. For instance, the purity of 7 remained
unchanged even when stirred in 2.0 N HCl at room
Figure 2. X-ray cocrystal of aminooxazoline 6 in BACE.
temperature for 18 h. While increasing the potency of this
class remained an objective, the promising in vitro PK data
indicated that the oxazoline core had potential to be CNS
penetrant and metabolically stable, given its permeability, low
efflux ratio, and low microsomal turnover.
To better understand the origins of the moderate potency
and to gain insight to possible interactions that could be
exploited to improve potency, a cocrystal structure of
compound 6 bound to BACE was obtained (Figure 2). A
hydrogen bonding network and electrostatic interaction
between the catalytic diad (Asp228 and Asp32) and the
aminooxazoline core were observed. In addition, the phenyl-
pyridyl moiety occupied the S1−S3 pocket, with the pyridyl
nitrogen accepting a hydrogen bond from the backbone N−H
of Ser229 through a bridging water molecule. The dihedral
angle between the phenyl ring and the pyridyl ring was about
60°, suggesting that the 2-fluoride substitution not only served
to engage in van der Waals contacts but also may have
reinforced this conformation, thereby imparting the increased
potency found with 1 and 6. On the “prime” side of the
aminooxazoline, an edge-to-π face interaction was observed
between the phenyl ring and Tyr71 with a distance of about 3.8
Å. In addition, the ether of 6 was a hydrogen bond acceptor
with the side chain of Trp76.
From the structure−activity relationship (SAR) of HEA
inhibitors, we had previously discovered that hydrophobic
moieties placed in the S2′ pocket can provide a significant
potency increase.23 For instance, a cocrystal structure of 8a
bound to BACE showed that the neopentyl group extends into
the S2′ pocket (Figure 3A). An increase in van der Waals
contacts and hydrophobic interactions with the protein in this
region results in compounds 8b and 8c (R2 = neopentyl and
isobutyl, respectively) gaining 1000-fold in enzyme potency and
200-fold in cell potency compared to 8f (R2 = H) (Table 2).
C
dx.doi.org/10.1021/jm300598e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
meta position of the A ring would be appropriately positioned
to exploit the S1/S3 pockets of the enzyme.
The S1 pocket consists mostly of hydrophobic residues;
hydrophobic interactions with the enzyme in this region
contribute significantly to potency. Aromatic groups, such as a
phenyl ring, have been shown as good ligands for S1.12,25 This
pocket is adjacent to the S3 pocket which also consists of a
largely hydrophobic environment. The S3 site accommodates
large hydrophobic ligands through van der Waals contacts with
the enzyme.26 The hydrophobic environment stretching from
S1 to S3 presents an attractive opportunity to occupy the S3
binding pocket with a group extending from the S1 moiety.27
We envisioned the A ring of the xanthene core could occupy
the S1 hydrophobic pocket, and the meta position of the A ring
would project a substituent toward the nearby S3 site.
As a starting point, we decided to pursue a library of
meta,meta bis-substituted aminooxazoline xanthenes (AOX) 9
via parallel synthesis. To test our hypothesis, we first installed a
methoxy group at the 3 position of the B ring of the xanthene
core to partially occupy the S2′ pocket with the intention of
exploring the prime side pockets at a later stage. The methoxy
group could serve as an indicator of whether hydrophobic
interactions in the region could bring in better potency
compared to the biarylaminooxazoline 6. To maximize the
potential for optimal interactions with the protein, we first
investigated phenyl rings with a variety of substituents for
potential hydrophobic interactions with the S3 pocket.
Figure 3. Design of aminooxazoline xanthenes: (A) overlay of X-ray
BACE binding structure of HEA compound 8a and 2-aminooxazoline
6; (B) model of 6 where right side phenyl group is rotated by 90°; (C)
proposed binding model of aminooxazoline xanthene.
The SAR is summarized in Table 3. Monosubstitution at the
ortho- or para-position of the phenyl ring did not improve
potency; ortho-substitution alone proved to be detrimental to
the potency (9b, e). At para-positions of the phenyl ring, small
groups such as a chlorine atom or a methyl group provided
similar potency as an unsubstituted phenyl ring (9d,g). Larger
groups such as a trifluoromethoxy group (9h) decreased
potency. Meta-substitution proved to have a great impact on
potency; 9f (3-Cl-phenyl) is 10-fold more potent than 9c (3-
Me-phenyl) and 9i (3-CF3O-phenyl). The disubstituted phenyl
rings with a fluorine atom at the 2-position were beneficial to
potency. With a substituent at the 5-position of the 2-
fluorophenyl ring, 9l (2-F-5-Cl-phenyl) and 9n (2-F-5-MeO-
phenyl) demonstrated submicromolar enzymatic potency (IC50
= 0.28 and 0.74 μM, respectively). Moving the methoxy group
to the 3-position resulted in another inhibitor 9m (2-F-3-MeO-
phenyl) with good activity (IC50 = 0.9 μM). Substitution at the
4-position of the phenyl ring was not tolerated, as 9o (2-F-4-
Me-5-Cl-phenyl) lost significant potency compared to 9l.
To aid further design, a cocrystal structure of 9l bound to
BACE was solved at 1.95 Å resolution (Figure 4A). Although
racemic 9l was used for crystallography, only the (S)-
enantiomer crystallized with the protein. By analogy with
compound 6, 2-aminooxazoline 9l interacts with the catalytic
Asp32 and Asp228 through a hydrogen bond network. As
anticipated, the A ring of the xanthene core occupies the S1
pocket, from which the 2-fluoro-5-chlorophenyl group extends
into the S3 pocket where the chlorine atom makes hydrophobic
interactions with the backbone of residues Gly13, Gly230, and
Thr231. It seems likely that the same hydrophobic interactions
also accounted for the good potency of 9f. Notably a water
molecule, which is tightly bound to the carbonyl groups of
protein residues Ser229 and Thr231, is also close to the
chlorine atom in this region. In addition, the xanthene tricyclic
moiety of 9l makes a π-stacking interaction with Tyr71 from
the flap region of BACE. Furthermore, while the methoxy
Table 2. Examples of HEA SAR for P2′ Substituents
a
IC50 (μM)
compd
R1
R2
BACE
cellular
8a
8b
8c
8d
8e
8f
F
(CH3)3CCH2
(CH3)3CCH2
(CH3)2CHCH2
CH3CH2
Cl
0.005 0.002
0.006 0.006
0.005 0.002
0.091 0.056
0.957 0.043
5.28 1.18
0.01 0.006
0.003 0.004
0.004 0.002
0.039 0.023
1.88
H
H
H
H
H
H
0.804
a
Values represent the mean values of at least two experiments SD.
Values without SD are for a single determination only.
We envisioned that by exploiting similar interactions in the
S2′ pocket, we could enhance the potency of our aminooxazo-
line compounds. A key insight was the realization that rotation
of the right-most phenyl ring (ring B) by ∼90° would allow
direct access to S2′ by a meta substituent (Figure 3B). To lock
this conformation, we designed a xanthene based 2-amino-
oxazoline core (Figure 3C). By linkage of the adjacent phenyl
rings through an oxygen atom, the rigid tricycle structure
positions the substituent at the 3-position of the xanthene
toward the S2′ site; this rigid structure also decreases the
number of rotatable bonds, thereby potentially decreasing the
binding entropic penalty.24 In addition, aryl substitution at the
D
dx.doi.org/10.1021/jm300598e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 3. SAR of Xanthene BACE Inhibitors
a
IC50 (μM)
compd
Ar
BACE
cellular
Cat D
>40
9a
9b
9c
9d
9e
9f
phenyl
1.18 0.11
7.02 4.28
1.67 1.52
1.78 0.21
4.88 3.7
2.52 1.79
4.22
2-Me-phenyl
30.9 8.6
26.9 2.0
>40
3-Me-phenyl
4.54 2.43
3.63
4-Me-phenyl
2-Cl-phenyl
2.47
16.8 5.0
35.7 4.7
>40
3-Cl-phenyl
0.13 0.03
2.02 0.003
5.06 3.45
2.76 2.43
1.17 1.09
1.56 0.34
0.28 0.10
0.90 0.88
0.74 0.71
3.21 2.01
0.80 0.86
0.90 0.09
2.11 1.62
3.90 0.14
0.41 0.4
2.51 0.90
2.42 1.75
3.69 2.77
6.41 4.37
6.75 1.4
2.24 2.51
4.29 4.16
2.23 0.89
3.72 1.84
2.98
9g
9h
9i
4-Cl-phenyl
4-CF3O-phenyl
3-CF3O-phenyl
2-F-3-Cl-phenyl
2-F-4-Cl-phenyl
2-F-5-Cl-phenyl
2-F-3-MeO-phenyl
2-F-5-MeO-phenyl
2-F-4-Me-5-Cl-phenyl
2-F-pyridin-3-yl
4-F-pyridin-3-yl
2-Cl-pyridin-4-yl
pyridin-4-yl
>40
20.1 3.9
13.7 3.4
28.5 5.5
3.7 1.7
37.3 1.8
13.7 0.8
13.8 4.3
16.0 6.8
>40
9j
9k
9l
9m
9n
9o
9p
9q
9r
9s
9t
2.0 1.88
1.33 0.81
5.80 4.49
2.82
44.2 4.08
>40
(rac)-pyrimidin-5-yl
(S)-pyrimidin-5-yl
(R)-pyrimidin-5-yl
1.43 0.38
0.534 0.325
ND
37.0 7.4
29.7 1.48
>40
9u
9v
0.107 0.024
10.40 0.86
a
Values represent the mean values of at least two experiments SD. Values without SD are for a single determination only. ND = not determined.
group in 6 hydrogen-bonds to Trp76, the rigid tricyclic ring
projects the methoxy group of 9l to the upper portion of the
S2′ pocket. The structure confirmed our hypothesis and
indicated that potency could be improved by extending off the
methoxy group and deeper into the S2′ pocket. The presence
of water molecules in the S3 pocket prompted us to investigate
the possibility of gaining affinity by interacting with the enzyme
backbone residues through water bridged hydrogen bond
interactions. A variety of heteroaromatic groups were then
explored to fill the S3 pocket (Table 3, compounds 9p−v).
Heteroaromatic groups, such as substituted pyridyl and
pyrimidyl groups, displayed promising inhibitory activity. Not
surprisingly, the position of the heteroatom on the ring affects
potency. For instance, pyridine-3-yl is superior to pyridine-4-yl
(Table 3, compounds 9p−s). Notably, the enzyme potency of
9t (IC50 = 0.41 μM) represented a 28-fold increase compared
to its biarylaminooxazoline congener 4. The two enantiomers
of 9t were obtained in excellent optical purity (>99% ee) by
Figure 4. X-ray cocrystal structures: (A) 9l bound to BACE protein;
(B) 9l bound to BACE protein with superimposed HEA 8a in blue;
(C) 11a bound to BACE protein; (D) 11a bound to BACE protein
with superimposed HEA 8a in blue.
chiral HPLC separation. The (S)-enantiomer 9u was
significantly more potent (10-fold) than the (R)-isomer 9v
and is consistent with the cocrystallization of (S)-9l.
Table 4. Compound Profile of Selected Compounds
a
bc
,
c
compd
MW
PSA (Å2)
cLogP
HBD
rotatable bonds
BE
RLM/HLM (μL min−1 mg−1
)
hER
Papp (10−6 cm/s)
9f
9l
9t
393
411
360
66
5.07
5.21
1.90
2
2
2
2
2
2
0.33
0.31
0.32
200/390
170/130
52/80
1.0
0.5
3.0
5.4
5.7
17
66
91.9
a
b
c
Compounds were incubated with microsomes for 30 min at 1 μM. B/A to A/B ratio. Parental LLC-PK1 cell line.
E
dx.doi.org/10.1021/jm300598e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 5. SAR of Xanthene BACE Inhibitors
a
IC50 (μM)
b
bc
,
compd
R
BACE
cellular
Cat D
Papp (10−6 cm/s)
hER
9u
Me
Pr
0.107 0.024
0.016 0.003
0.008 0.0003
0.534 0.325
0.052 0.013
29.7 1.48
9.12 4.29
4.47 0.584
17
14
9.1
2.2
1.5
0.9
10a
11a
i-Bu
0.036 0.007
a
b
c
Values represent the mean values of at least two experiments SD. Parental LLC-PK1 cell line. B/A to A/B ratio.
Table 6. Pharmacokinetic Profile of Selected Compounds in Rats
a
a
b
b
b
c
c
compd
HLM (μL min−1 mg−1
)
RLM (μL min−1 mg−1
)
CL (L h−1 kg−1
)
Vss (L/kg)
T1/2 (h)
AUCinf (μM·h)
F
(%)
9u
97
55
19
22
21
20
2.86
4.92
2.76
9.37
3.49
3.7
3.2
3.2
20.9
5.66
7.29
260
10a
11a
110
31
1.08
a
b
c
Compounds were incubated with microsomes for 30 min at 1 μM. 2 mg/kg iv dose (solution in 100% DMSO). 10 mg/kg oral dose (solution in
(1% Tween 80)/(2% HPMC)).
Encouragingly, 9u also demonstrated 270-fold selectivity
against the closely related aspartyl protease Cat D. These
compounds also consistently demonstrated ≤5-fold enzyme to
cell shift. For instance, the most potent enzyme inhibitor 9u
recorded a cellular IC50 of 0.53 μM, a cell shift of 5-fold.
Selected compounds were further profiled to get an
understanding of their ability to penetrate the blood−brain
barrier (BBB) (Table 4). Compounds 9f, 9l, and 9t displayed
favorable calculated physiochemical properties required for
CNS drugs: a low molecular weight (<400 Da), few hydrogen
bond donors (2), limited rotatable bonds (2), and low polar
surface area.15,28 In addition, these compounds do not appear
to be substrates of the PGP transporter (efflux ratios of 1.3, 0.5,
and 3.0, respectively). With a cLogP of 1.9, compound 9t
demonstrated good apparent permeability (Papp of 17.2 × 10−6
cm/s). On the other hand, the more lipophilic compounds 9f
and 9l were less permeable (Papp of 5.4 × 10−6 and 5.7 × 10−6
cm/s, respectively), which might contribute to their high
enzyme to cell shift (Table 3). Furthermore, 9t demonstrated
better intrinsic stability in liver microsomes compared to 9f and
9l. Because of overall attractive physiochemical properties of 9t
for CNS penetration and the overall superior properties
brought by the pyrimidin-3-yl group, we chose to investigate
this series further.
To improve the potency of 9t, we pursued the second part of
our hypothesis, which was to gain affinity by filling the S2′
pocket through extension of the methoxy group. Specifically,
we investigated interactions with the S2′ pocket by exploring
substitution of the aryl ether in 9. As anticipated, filling the S2′
pocket with bulkier hydrophobic groups provided a significant
increase in enzymatic inhibition and cellular potency (Table 5).
Increasing the size from a methyl group to an n-propyl
substituent resulted in a more than 10-fold increase in both
enzyme and cell activity. A further increase in size to an isobutyl
group led to an additional ∼2-fold enhancement in potency.
These initial results confirmed our hypothesis that engaging
hydrophobic interactions in S2′ pocket could boost potency.
Encouragingly, compounds 10a and 11a maintained good
permeability and a low PGP-mediated efflux (Table 5). In
addition to the improved BACE potency, another appealing
outcome of extending deep into the S2′ pocket is the enhanced
selectivity against Cat D. For example, Cat D selectivity
increased from 270-fold for 9t to 560-fold for 11a.
A cocrystal structure of 11a bound to BACE protein was
obtained at 2.1 Å resolution (Figure 4C). Compound 11a
bound to the BACE protein similarly to previous structures.
The aminooxazoline interacts with the catalytic aspartate diad
in a fashion similar to 9l, with the xanthene moiety π-stacking
with Tyr71 and the pyrimidine ring of 11a occupying the S3
pocket. Unlike the hydrophobic interactions made by the 2-F-5-
Cl-phenyl group of 9l in the S3 region, the pyrimidine of 11a
interacts with BACE protein via two water-bridged hydrogen
bond networks. One N-atom of the pyrimidine ring interacts
with the side chain of Thr232 through a water-mediated
hydrogen bond. In the same region of the crystal structure of 9l
bound to BACE, a water molecule appears to tightly bind to
Thr232 and Ser229 via hydrogen bonds. In the structure of 11a
and BACE, the water molecule has been displaced outside
hydrogen bonding distance from the residue Ser229. The other
N-atom of the pyrimidine ring interacts with the backbone
carbonyl of Gly11 through a second water bridged H-bond
network. As expected, the isobutyl group of compound 11a
extends deeply into the upper region of the S2′ pocket. An
overlay of cocrystal structures of 11a/BACE with 8a/BACE
shows clearly that the isobutyl group of 11a occupies the same
space as the neopentyl group of 8a (Figure 4D).
With potent and selective BACE inhibitors 10a and 11a in
hand, we sought to understand the PKDM properties of these
fully elaborated aminooxazoline xanthenes (Table 6). Although
the hepatic intrinsic clearance was relatively low, in vivo
clearance (CL) was much higher than would be predicted.29
For instance, while compounds 9u, 10a, and 11a had similar
stability in rat liver microsomes, compounds 9u and 10a
displayed high in vivo clearance while 11a demonstrated
moderate clearance.30 The oral bioavailability of these
compounds was consistent with the moderate apparent
permeability.31 Furthermore, these compounds were poor
PGP substrates, with each having a low efflux ratio (<3).
Because of its good overall PK and potent cell activity,
F
dx.doi.org/10.1021/jm300598e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 5. Aβ40 reduction in wild type rats with 11a at 4 h (30 and 100 mg/kg, po): (∗) P < 0.0001.
Table 7. In Vivo Aβ40 Reduction in Wild Type Rats with 11a at 4 h (30 and 100 mg/kg, po)
a
dose (mg/kg) % inhibition in CSF [CSF] (μM) [CSF]/(cell IC50) % inhibition in brain [brain] (μM) [plasma]total (μM) [plasma]unbound (μM)
30
76
81
0.012
0.036
0.3
0.9
45
63
4.0
9.7
3.6
6.1
0.071
0.12
100
a
Dosed as a solution in mg/kg oral dose (solution in (1% Tween 80)/(2%HPMC)).
a
Scheme 1. Synthesis of Biaryl Aminooxazoline BACE Inhibitors
a
Reagents: (a) BBr3, CH2Cl2, 76%; (b) 2-chloro-2,2-difluoroacetic acid, K2CO3, H2O, DMF, 72%; (c) PPh3CH3Br, n-BuLi, THF, 99%; (d) AgNCO,
I2, THF; (e) NH4OH, acetone, 75% over two steps; (f) Ar(BOH)2, Pd(PPh3)4, DME.
compound 11a was chosen to study the effect of BACE
inhibition on levels of Aβ40 in the brain.
Aβ40 reduction in the CSF and brain improved with the
increased exposure in the corresponding compartments (Table
7, Figure 5).
The pharmacodynamic assay was conducted in naive male
Sprague−Dawley rats that were administered a single oral dose
of compound 11a at either 30 or 100 mg/kg. Aβ levels and
drug concentrations were measured at 4 h in the brain and
CSF. Compound 11a exhibited a significant reduction of Aβ40
in the central nervous system (Figure 5). The brain−blood
ratios of 11a, 1.1 at 30 mg/kg and 1.6 at 100 mg/kg, suggest
that the compound readily passes the blood−brain barrier
(Table 7). 11a significantly inhibited Aβ40 production in both
CSF (76%) and brain (45%) at 30 mg/kg relative to the
vehicle-treated animals. At 100 mg/kg, the inhibition of Aβ40
production was slightly improved in CSF (81%) and a more
robust Aβ40 reduction was seen in brain (63%). The degree of
CHEMISTRY
■
The synthesis of biaryl aminooxazolines is outlined in Scheme
1. Starting with commercially available ketone 12, demethyla-
tion mediated by BBr3 followed by difluoromethylation of the
resulting alcohol afforded diaryl ketone 13. Subsequent Wittig
olefination of 12/13 afforded 14. Addition of in situ generated
iodine isocyanate to 14 yielded iodide 15 which was cyclized to
the penultimate aminooxazoline bromide 16 in good yield. This
five-step protocol was reliable and scalable allowing for quick
access to precursor 16 which was elaborated to compounds 1−
7 through Suzuki coupling.
G
dx.doi.org/10.1021/jm300598e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Scheme 2. Synthesis of 2-Aminooxazoline Xanthenes
a
Reagents: (a) 4-bromophenol, Cu(OTf)2, Cs2CO3, cat. EtOAc, toluene, 110 °C, 68%; (b) H2SO4, 60 °C, 84%; (c) trimethylsilylmethyllithium,
THF, −78 °C; then AcCl, 23 °C; (d) I2, AgNCO, Et2O; then NH4OH, THF, 69%; (e) ArB(OH)2, Pd(PPh3)4, aq Na2CO3, dioxane, 90 °C.
a
Scheme 3. Synthesis of Xanthene BACE Inhibitors 10a and 11a
a
Reagents: (a) BBr3, CH2Cl2, 23 °C, 85%; (b) pyrimidin-5-ylboronic acid, Pd(PPh3)4, K2CO3, THF−H2O, 100 °C, 64%; (c) alkyl iodide (RI),
Cs2CO3, DMF; (d) chiral HPLC.
Aminooxazoline xanthenes were synthesized according to the
route outlined in Scheme 2. Diaryl ether 18 was prepared by
copper-catalyzed etherification of aryl bromide 17.32 Subse-
quent Friedel−Crafts acylation provided xanthenone 19 which
was converted to the terminal olefin 20 by a Peterson-type
olefination.33 The addition of in situ generated iodine
isocyanate to the olefin followed by treatment with ammonium
hydroxide led to 2-aminooxazoline aryl bromide 21.34 Suzuki
coupling of aryl bromide 21 with a variety of arylboronic acids
generated the desired 2-aminooxazoline xanthenes 9.
As shown in Scheme 3, the 2-aminooxazoline xanthene ether
analogues 10a and 11a were accessed in three steps from
intermediate 21. Boron tribromide mediated demethylation of
bromide 21 cleanly yielded phenol 22. A subsequent Suzuki
coupling provided 23. Selective alkylation of the phenol with
alkyl iodides then afforded the desired alkoxyxanthenes. It is
noteworthy that there were minimal N-alkylated products
observed under these conditions. Subsequent chiral separation
furnished optically pure enantiomers 10a and 11a.
cellular potency while maintaining properties necessary for
penetration across the blood−brain barrier. A single oral dose
of 11a produced a significant Aβ40 reduction in the CNS in
naive rats. Development and characterization of this series of
BACE inhibitors for the treatment of AD will be the subject of
future publications.
EXPERIMENTAL SECTION
■
BACE1 Enzymatic Assay. BACE1 enzymatic activity was
determined by the enhancement of fluorescence intensity upon
enzymatic cleavage of the fluorescence resonance energy transfer
substrate. The BACE1 recognition and cleavage sequence of the
substrate are derived from the reported literature,25a and the
fluorophore and quencher dyes are attached to the side chain of Lys
residues at the termini of the substrate peptide. The human
recombinant BACE135 assay was performed in 50 mM acetate, pH
4.5, 8% DMSO, 100 μM Genepol, 0.002% Brij-35. In dose-response
IC50 assays, 10 point 1:3 serial dilutions of compound in DMSO were
preincubated with the enzyme for 60 min at room temperature.
Subsequently, the substrate was added to initiate the reaction. After 60
min at room temperature, the reaction was stopped by addition of 0.1
M Tris base to raise the pH above the enzyme active range, and the
increase of fluorescence intensity was measured on Safire II microplate
CONCLUSION
■
A structure- and property-based approach led to the discovery
of aminooxazoline xanthenes as potent and selective BACE
inhibitors. Optimized compound 11a demonstrated good
reader (Tecan, Mannedorf, Switzerland).
̈
Cell-Based Assay. Human embryonic kidney cells (HEK293)
stably expressing APPSW were plated at a density of 100K cells/well in
H
dx.doi.org/10.1021/jm300598e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
96-well plates (Costar). The cells were cultivated for 6 h at 37 °C and
5% CO2 in DMEM supplemented with 10% FBS. Cells were incubated
overnight with test compounds at concentrations ranging from 0.0005
to 10 μM. Following incubation with the test compounds the
conditioned medium was collected and the Aβ40 levels were
determined using a sandwich ELISA. The IC50 was calculated from
the percent of control Aβ40 as a function of the concentration of the
test compound. The sandwich ELISA to detect Aβ40 was performed in
96-well microtiter plates, which were precoated with goat anti-rabbit
IgG (Pierce). The capture and detection antibody pair that was used to
detect Aβ40 from cell supernatants consists of affinity purified pAβ40
(Invitrogen) and biotinylated 6E10 (Covance), respectively. Con-
ditioned medium was incubated with capture antibody overnight at 4
°C followed by washing. The detecting antibody incubation was for 3
h at 4 °C, again followed by the wash steps as described previously.
The plate was developed using Delfia reagents (streptavidin−
europium and enhancement solution (Perkin-Elmer)), and time-
resolved fluorescence was measured on an EnVision multilabel plate
reader (Perkin-Elmer).
Permeability Assay. The wild type cell line LLC-PK1 (porcine
renal epithelial cells, WT-LLC-PK1) was purchased from American
Type Culture Collection (ATCC, Manassass, VA). Transfections of
WT-LLC-PK1 cells with human MDR1 gene (hMDR1-LLC-PK1) and
rat mdr1a gene (rMdr1a-LLC-PK1) were generated. Cells were grown
in medium 199 supplemented with 10% fetal bovine serum.36 Cells
were seeded onto Matrigel-coated Transwell filter membranes at a
density of 90 000 cells/well. Change of medium was performed on day
3. Compound incubations were performed 5−6 days after seeding. All
cultures were incubated at 37 °C in a humidified (95% relative
humidity) atmosphere of 5% CO2/95% air.
Inhibitor complexes were obtained by soaking apo BACE crystals in
mother liquor solutions supplemented with 1 mM compound for a
period of 6−14 h. Crystals were transferred briefly into a cryosolution
consisting of 25% (w/v) PEG 5000 MME, 0.1 M sodium citrate (pH
6.6), 0.2 M ammonium iodide, and 20% (v/v) glycerol prior to being
flash frozen in liquid nitrogen. Diffraction data were collected on a
FRE rotating anode X-ray source equipped with an RAXIS IV++
detector, and images were processed using the HKL suite of
programs.38 The structures were solved by molecular replacement
with AMORE using 1W50 as a search model, and they were refined
using REFMAC.39 Model building was performed with COOT.40 Data
collection and refinement statistics appear in the Supporting
Information.
Pharmacodynamic Assay. Male Sprague−Dawley rats (175−200
g) were purchased from Harlan and were maintained on a 12 h light/
dark cycle with unrestricted access to food and water until use. Rats
were administered compound by oral gavage at the appropriate dose.
Rats were euthanized with CO2 inhalation for 2 min, and cisterna
magna was quickly exposed by removing the skin and muscle above it.
CSF (50−100 μL) was collected with a 30 gauge needle through the
dura membrane covering the cisterna magna. Blood was withdrawn by
cardiac puncture and plasma obtained by centrifugation for drug
exposures. Brains were removed and, along with the CSF, immediately
frozen on dry ice and stored at −80 °C until use. The frozen brains
were subsequently homogenized in 10 volumes of (w/v) of 0.5%
Triton X-100 in TBS with protease inhibitors. The homogenates were
centrifuged at 100 000 rpm for 30 min at 4 °C. The supernatants were
analyzed for Aβ40 levels by immunoassay as follows: Meso Scale 96-
well avidin plates were coated with biotin-4G8 (Covance) and
detected with ruthenium-labeled Fab specific for Aβ40. Plates were
read in an MSD Sector6000 imager according to the manufacturer’s
recommended protocol (Meso Scale Discovery, Inc.). Aβ40
concentrations were plotted using Graphpad Prism and analyzed by
one-way ANOVA followed by Dunnett’s multiple comparison analysis
to compare drug-treated animals to vehicle-treated controls.
Prior to the transport experiment,37 culture medium was aspirated
from both apical and basolateral wells, and cells were rinsed with
warmed (37 °C) Hank’s balanced salt solution supplemented with 10
mM Hepes at pH 7.4 (HHBSS, Invitrogen, Grand Island, NY).
HHBSS was removed from wells prior to dosing with test drugs at 5
μM in transport buffer (HHBSS containing 0.1% bovine serum
albumin). An amount of 150 μL of transport buffer was added to
receiver chambers prior to dosing in triplicate to apical or basolateral
chambers. The dosed Transwell plates containing the cell monolayers
were incubated for 2 h at 37 °C on a shaking platform. At the end of
the incubation period, 100 μL samples were collected from receiver
reservoirs and analyzed by LC−MS/MS on an API4000 (Applied
Biosystem, Foster City, CA) triple quadruple mass spectrometer
interfaced with TurboIonSpray operated in positive mode using
Analyst 1.4.2 software.
Chemistry. Unless otherwise noted, all materials were obtained
from commercial suppliers and used without further purification.
Anhydrous solvents were obtained from Aldrich and used directly. All
reactions involving air- or moisture-sensitive reagents were performed
under a nitrogen or argon atmosphere. All final compounds were
purified to ≥95% purity as determined by high-performance liquid
chromatography (HPLC). Purity was measured using Agilent 1100
series HPLC systems with UV detection at 254 nm. System A had the
following parameters: Agilent Zorbax Eclipse XDB-C8 4.6 mm × 150
mm, 5 μm, 5−100% CH3CN in H2O with 0.1% TFA for 15 min at 1.5
mL/min. System B had the following parameters: Zorbax SB-C8, 4.6
mm × 75 mm, 10−90% CH3CN in H2O with 0.1% formic acid for 12
min at 1.0 mL/min. Silica gel columns were used with prepacked silica
The apparent permeability coefficient (Papp) of all tested agents was
estimated from the slope of a plot of cumulative amount of the agent
versus time based on the following equation:
1
gel cartridges (Biotage). H NMR spectra were recorded on a Bruker
P
= (dQ /dt)/(AC0)
app
AV-400 (400 MHz) spectrometer or on a Bruker AV-500 (500 MHz)
spectrometer at ambient temperature. NMR spectra were processed
using ACD SpecManager (version 12; ACD, Toronto, Canada). All
observed protons are reported as parts per million (ppm) downfield
from tetramethylsilane (TMS) or other internal reference in the
appropriate solvent indicated. Data are reported as follows: chemical
shift, multiplicity (s = singlet, d = doublet, t = triplet, q = quartet, br =
broad, m = multiplet), coupling constants, and number of protons.
Melting points were determined on a TA Instruments Q200 DSC
(New Castle, DE, U.S.). Low-resolution mass spectrometry (MS) data
were determined on an Agilent 1100 series LCMS with UV detection
at 254 nm and a low resonance electrospray mode (ESI). Exact mass
confirmation was performed on an Agilent 1200 series HPLC system
(Santa Clara, CA, U.S.) by flow injection analysis, and elution was with
binary solvent systems A and B (A, water with 0.1% FA; B, ACN with
0.1% FA) under gradient conditions (5−95% B over 3 min) at 0.3
mL/min with MS detection by an Agilent 6510-Q-TOF mass
spectrometer (Santa Clara, CA, U.S.).
where dQ/dt is the penetration rate of the agent (ng/s), A is the
surface area of the cell layer on the Transwell (0.11 cm2), and C0 is the
initial concentration of the test compound (ng/mL). Efflux ratio (ER)
was calculated from the basolateral-to-apical permeability divided by
the apical-to-basolateral permeability: ER = [Papp(B→A)]/[Papp(A→
B)].
Microsomal Stability Assay. Compounds (1 μM) were incubated
with liver microsomes (0.25 mg/mL in 67 mM phosphate buffer, pH
7.4) from human and rat at 37 °C for 30 min with or without 1 mM
NADPH in a total volume of 0.2 mL. The final concentration of
DMSO in the incubation was <0.1%. Incubations were stopped by
addition of 200 μL of ice-cold acetonitrile containing 0.5% formic acid
and an internal standard (500 ng/mL) followed by centrifugation at
3100 rpm for 20 min. The supernatants were analyzed directly
(without any further sample cleanup) by high performance liquid
chromatography (HPLC) and mass spectrometric detection.
X-ray Crystal Structure. The catalytic domain of human BACE1
(residues 14−453) was overexpressed in E. coli and refolded from
inclusion bodies following a procedure described previously.25b
Representative Synthetic Protocols of the Aminooxazolines
Shown in Schemes 1−3. Exemplified Analogues of Scheme 1.
Preparation of 4-(4-(Difluoromethoxy)phenyl)-4-(3-(2-fluoro-
I
dx.doi.org/10.1021/jm300598e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
pyridin-3-yl)phenyl)-4,5-dihydrooxazol-2-amine (7). Step a. To a
solution of (3-bromophenyl)(4-methoxyphenyl)methanone (4.00 g,
14 mmol) in DCM (69 mL) at 0 °C was added boron tribromide (21
mL, 21 mmol) dropwise. The resulting reaction mixture was warmed
to room temperature and stirred overnight. After the mixture was
stirred overnight, the reaction was not complete as indicated by
LCMS. An additional 21 mL of BBr3 (1 M) was added. The mixture
was stirred for an additional 24 h. Saturated aqueous NaHCO3 was
added slowly to quench the reaction mixture at room temperature.
The mixture was then transferred to a separatory funnel, and the
aqueous layer was extracted with EtOAc. The organic layers were
combined, dried over MgSO4, filtered, and concentrated in vacuo to an
oil. The product was purified by chromatography on silica gel (0−
100% 90:10:1 DCM/MeOH/NH4OH in DCM) to yield (3-
bromophenyl)(4-hydroxyphenyl)methanone (2.9 g, 76% yield). 1H
NMR (400 MHz, CDCl3) δ 8.19−8.23 (m, 1H), 8.03 (ddd, J = 1.03,
1.79, 7.95 Hz, 1H), 7.87−7.91 (m, 1H), 7.83 (ddd, J = 1.03, 1.96, 7.97
Hz, 1H), 7.75−7.80 (m, 1H), 7.69 (dddd, J = 1.08, 1.83, 7.86, 13.00
Hz, 1H), 7.39−7.45 (m, 1H), 7.33−7.39 (m, 1H), 6.91−6.96 (m, 1H).
MS m/z = 277.2 [M + H]+.
1.0 g of the (S)-4-(3-bromo-4-fluorophenyl)-4-(4-methoxyphenyl)-
4,5-dihydrooxazol-2-amine. The overall yield was 75% for two steps.
1H NMR (400 MHz, CDCl3) δ 7.45−7.53 (m, 2H), 7.31−7.38 (m,
2H), 7.20−7.30 (m, 3H), 7.08−7.16 (m, 2H), 6.75 (t, J = 73.75 Hz,
1H), 5.49 (d, J = 0.78 Hz, 1H), 5.48 (d, J = 0.88 Hz, 1H). MS m/z =
383.1 [M + H]+.
Step f. (R)-4-(3-Bromo-4-fluorophenyl)-4-(4-methoxyphenyl)-4,5-
dihydrooxazol-2-amine (0.070 g, 0.18 mmol) was added to a vial
followed by DME (1.0 mL). The vial was kept under a stream of
argon, and 2-fluoro-3-pyridineboronic acid (0.051 g, 0.37 mmol),
Pd(PPh3)4 (0.047 g, 0.037 mmol), and sodium carbonate (2.0 M, 0.40
mL) were added. The vial was capped and heated to 125 °C for 16 h.
The reaction mixture was cooled to room temperature and
concentrated in vacuo. The crude product was purified by RPHPLC
using 0.1% formic acid in acetonitrile and water as the mobile phase.
(R)-4-(4-(Difluoromethoxy)phenyl)-4-(3-(2-fluoropyridin-3-yl)-
phenyl)-4,5-dihydrooxazol-2-amine (0.046 g, 58% yield) was isolated
1
as a white solid. H NMR (400 MHz, DMSO-d6) δ 8.19−8.29 (m,
1H), 8.15 (s, 1H), 8.07 (ddd, J = 1.91, 7.46, 10.29 Hz, 1H), 7.69 (br s,
1H), 7.40−7.56 (m, 6H), 6.96−7.39 (m, 3H), 4.72−4.86 (m, 2H). MS
m/z = 400.2 [M + H]+.
Steps b and c. To a solution of (3-bromophenyl)(4-hydroxyphenyl)-
methanone (2.89 g, 10.4 mmol), potassium carbonate (6.49 g, 46.9
mmol), and water (2.10 mL) in DMF (20.9 mL) was added 2-chloro-
2,2-difluoroacetic acid (2.64 mL, 31.3 mmol) dropwise. After gas
ceased to evolve, the reaction vessel was sealed and heated to 100 °C.
This mixture was stirred overnight and then transferred to a separatory
funnel containing water. The aqueous layer was extracted with DCM.
The organic layers were combined, dried over MgSO4, filtered, and
concentrated in vacuo to an oil. The crude product was purified by
chromatography on silica gel (0−100% 90:10:1 DCM/MeOH/
NH4OH in DCM) to provide (3-bromophenyl)(4-(difluoromethoxy)-
phenyl)methanone (2.4 g, 70% yield). MS m/z 327.3 [M + H]+. To a
solution of methyltriphenylphosphonium bromide (3.7 g, 10 mmol) in
THF (25 mL) at 0 °C was added n-BuLi (5.5 mL, 8.8 mmol)
dropwise. After the mixture was stirred at 0 °C for 30 min, a solution
of (3-bromophenyl)(4-(difluoromethoxy)phenyl)methanone (2.4 g,
7.3 mmol) in THF (9.2 mL) was added dropwise. The solution was
allowed to warm to room temperature and stirred for 1 h and 15 min.
The reaction mixture was quenched with water, and the aqueous layer
was extracted with EtOAc. The combined organic layers were dried
over MgSO4, filtered, and concentrated in vacuo to an oil. The crude
product was purified by chromatography on silica gel (0−30% EtOAc
in hexanes) to provide 1-bromo-3-(1-(4-(difluoromethoxy)phenyl)-
Exemplified Analogues of Scheme 2. Preparation of 2′-
Methoxy-7′-(pyrimidin-5-yl)-5H-spiro[oxazole-4,9′-xanthen]-
2-amine 2,2,2-Trifluoroacetate (9t). Step a. 4-Bromophenol (8.7 g,
50 mmol), Cs2CO3 (16 g, 50 mmol), CuOTf−toluene complex (2:1)
(150 mg, 0.625 mmol), and EtOAc (0.25 mL) were added to a
solution of 2-bromo-5-methoxybenzoic acid (11.6 g, 50 mmol) in
toluene (40 mL) in a sealed tube. The reaction mixture was purged
with N2 and heated to 110 °C until the aryl halide was consumed as
determined by LCMS (48 h). After cooling to room temperature, the
mixture was filtered through a Celite plug. The Celite plug was washed
with EtOAc. The mixture was acidified with 1 N HCl and extracted
with EtOAc. The combined organic extracts were washed with brine,
dried over MgSO4, filtered, and concentrated in vacuo. The residue
was purified via column chromatography on silica gel (0−10% MeOH
in DCM) to afford 2-(4-bromophenoxy)-5-methoxybenzoic acid (11 g,
68% yield). 1H NMR (400 MHz, CDCl3) δ 7.67 (d, J = 3.23 Hz, 1H),
7.49 (d, J = 9.00 Hz, 1H), 7.49 (dd, J = 5.48, 15.45 Hz, 1H), 7.09 (dd,
J = 3.28, 9.05 Hz, 1H), 6.93 (d, J = 9.00 Hz, 1H), 6.93 (dd, J = 5.48,
15.36 Hz, 1H), 6.88 (d, J = 9.00 Hz, 1H), 3.87 (s, 3H). MS m/z =
324.9 [M + H]+.
Step b. Sulfuric acid (0.5M, 41 mL, 21 mmol) was added to 2-(4-
bromophenoxy)-5-methoxybenzoic acid (3.75 g, 12 mmol) at room
temperature. The reaction mixture was stirred at 60 °C for 1 h. LCMS
showed complete reaction. The reaction mixture was cooled to room
temperature and poured slowly over a stirred mixture of ice and water
(100 mL). The tan precipitate was filtered and washed with water (3 ×
30 mL), 0.5 N NaOH (2 × 30 mL), and water again. The residue was
recrystallized from 40 mL of THF to give 2-bromo-7-methoxy-9H-
1
vinyl)benzene (2.35 g, 99% yield). H NMR (400 MHz, CDCl3) δ
7.41 (t, J = 1.66 Hz, 1H), 7.33−7.37 (m, 1H), 7.14−7.19 (m, 3H),
7.11 (dd, J = 6.70, 7.80 Hz, 1H), 5.33 (d, J = 0.98 Hz, 1H), 5.26 (d, J =
1.08 Hz, 1H), 3.74 (s, 3H). MS m/z = 325.2 [M + H]+.
Steps d and e. To a solution of 1-bromo-3-(1-(4-(difluoromethoxy)-
phenyl)vinyl)benzene (2.4 g, 7.2 mmol) in Et2O (18 mL, 7.2 mmol)
was added cyanatosilver (3.2 g, 22 mmol) at room temperature. The
resulting mixture was cooled to −6 °C. Solid iodine (1.8 g, 7.2 mmol)
was added in one portion, and the solution was stirred for 4 h. The
solution was then filtered through a cotton/Celite plug and
concentrated in vacuo to an oil. The crude mixture was taken forward
to the following step without further purification. MS m/z = 493.5 [M
+ H]+. To a solution of the crude 1-bromo-3-(1-(4-(difluoromethoxy)-
phenyl)-2-iodo-1-isocyanatoethyl)benzene (3.6 g, 7.2 mmol) in
acetone (22 mL, 7.2 mmol) at room temperature was added
ammonium hydroxide (30% in water) (2.52 mL, 22 mmol) by
syringe. The resulting mixture was stirred overnight. The reaction
mixture was partitioned between water and DCM, and the aqueous
layer was extracted with DCM. The combined organic extracts were
dried over MgSO4, filtered, and concentrated in vacuo to an oil. The
product was purified by chromatography on silica gel (0−40% 90:10:1
DCM/MeOH/NH4OH in DCM). The enantiomers of the purified
racemic product were separated by chiral chromatography [5 μm
Chiralcel OJ (4.6 mm × 100 mm) under SFC conditions with a
mobile phase of 10% methanol with 0.2% diethylamine in CO2 and a
flow rate of 5 mL/min] to provide 1.1 g of the (R)-4-(3-bromo-4-
fluorophenyl)-4-(4-methoxyphenyl)-4,5-dihydrooxazol-2-amine and
1
xanthen-9-one (2.96 g, 84% yield). H NMR (400 MHz, CDCl3) δ
8.48 (dd, J = 0.29, 2.54 Hz, 1H), 7.80 (dd, J = 2.54, 8.90 Hz, 1H), 7.70
(d, J = 3.13 Hz, 1H), 7.46 (dd, J = 0.39, 9.10 Hz, 1H), 7.41 (dd, J =
0.29, 8.90 Hz, 1H), 7.37 (dd, J = 3.13, 9.10 Hz, 1H), 3.94 (s, 3H). MS
m/z = 305.2.
Step c. A solution of 2-bromo-7-methoxy-9H-xanthen-9-one (2.04 g,
6.7 mmol) in THF (67 mL) contained in a 250 mL round-bottom
flask was cooled in a dry ice/acetone bath for 10 min to give a milky
white mixture. Trimethylsilylmethyllithium (10 mL of a 1.0 M solution
in pentane, 10 mmol) was added dropwise over 5 min to give a clear
orange solution. The mixture was stirred for 15 min. Then acetyl
chloride (0.76 mL, 11 mmol) was added dropwise, resulting in the
formation of a clear, bright-yellow solution. The mixture was warmed
to room temperature for 3 h. Then an additional portion of acetyl
chloride (0.25 mL) was added. The mixture was stirred for an
additional 30 min before being diluted with saturated aqueous sodium
bicarbonate solution (100 mL). The biphasic mixture was extracted
with EtOAc (2 × 50 mL), and the combined organic extracts were
dried over MgSO4, filtered, and evaporated to give 2-bromo-7-
methoxy-9-methylene-9H-xanthene as a yellow solid that was used
without further purification. 1H NMR (400 MHz, CDCl3) δ 7.84 (d, J
J
dx.doi.org/10.1021/jm300598e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
= 2.25 Hz, 1H), 7.39 (dd, J = 2.35, 8.80 Hz, 1H), 7.16 (d, J = 2.93 Hz,
1H), 7.05−7.09 (m, 1H), 7.01 (d, J = 8.80 Hz, 1H), 6.92 (dd, J = 2.93,
8.90 Hz, 1H), 5.50 (s, 2H), 3.86 (s, 3H). MS m/z = 303.0 [M + H]+.
Step d. Crude 2-bromo-7-methoxy-9-methylene-9H-xanthene was
suspended in Et2O (33 mL). Silver cyanate (3.0 g, 20 mmol) and
iodine (1.7 g, 6.7 mmol) were added in sequence, resulting in a brown
mixture. After being stirred for 40 min at room temperature, the
reaction mixture was filtered through Celite with the aid of Et2O, and
the filtrate was evaporated. The residue was dissolved in a mixture of
THF (26 mL) and ammonium hydroxide (2.6 mL), and the mixture
was stirred for 15 h. The reaction mixture was partitioned between
water (100 mL) and DCM (70 mL). The layers were separated, and
the aqueous layer was extracted with DCM (2 × 70 mL). The
combined organic extracts were dried over MgSO4, filtered, and
evaporated. The residue was purified by chromatography on silica gel
(0−40% 90:10:1 DCM/MeOH/NH4OH in DCM) to provide 2′-
bromo-7′-methoxyspiro[1,3-oxazole-4,9′-xanthen]-2-amine (1.65 g,
solution, 12.1 mmol). The vessel was sealed and placed in a 100 °C oil
bath for 4 h. The reaction mixture was cooled to room temperature
and partitioned between EtOAc (50 mL) and water (50 mL). The
aqueous layer was extracted with EtOAc (2 × 50 mL), and the
combined organic extracts were dried over MgSO4, filtered, and
evaporated. The crude material was purified by chromatography on
silica gel (30−100% 90:10:1 DCM/MeOH/NH4OH in DCM) to give
2′-hydroxy-7′-(5-pyrimidinyl)spiro[1,3-oxazole-4,9′-xanthen]-2-amine
(538 mg, 64% yield) as an off-white solid. 1H NMR (400 MHz,
DMSO-d6) δ 9.35 (br s, 1H), 9.17 (s, 1H), 9.08 (s, 2H), 7.73 (dd, J =
2.30, 8.46 Hz, 1H), 7.65 (d, J = 2.25 Hz, 1H), 7.26 (d, J = 8.41 Hz,
1H), 7.00 (d, J = 8.71 Hz, 1H), 6.68−6.79 (m, 2H), 6.45 (br s, 2H),
4.22 (d, J = 8.31 Hz, 1H), 4.17 (d, J = 8.22 Hz, 1H). MS m/z = 347.2
[M + H]+.
Step c. A glass vial was charged with 2′-hydroxy-7′-(5-pyrimidinyl)-
spiro[1,3-oxazole-4,9′-xanthen]-2-amine (53.7 mg, 155 μmol), cesium
carbonate (75.7 mg, 232 μmol), DMF (0.62 mL), and 1-iodopropane
(16.6 μL, 170 μmol). The mixture was stirred at room temperature for
18 h, then poured into water (10 mL) and extracted with EtOAc (3 ×
7 mL). The combined organic extracts were dried over MgSO4,
filtered, and evaporated. The residue was purified by chromatography
on silica gel (0−80% 90:10:1 DCM/MeOH/NH4OH in DCM) to
give 2′-(1-propyloxy)-7′-(5-pyrimidinyl)spiro[1,3-oxazole-4,9′-xanth-
1
69% yield) as a pale yellow foam. H NMR (400 MHz, CDCl3) δ
7.49 (d, J = 2.25 Hz, 1H), 7.41 (dd, J = 2.35, 8.71 Hz, 1H), 7.05 (dd, J
= 0.34, 8.75 Hz, 1H), 6.99 (d, J = 8.71 Hz, 1H), 6.86−6.92 (m, 2H),
6.49 (br s, 2H), 4.48 (d, J = 8.80 Hz, 1H), 4.43 (d, J = 8.80 Hz, 1H),
3.82 (s, 3H). MS m/z = 361.2 [M + H]+.
Step e. A 10 mL resealable tube was charged with tetrakis-
(triphenylphosphine)palladium (122 mg, 105 μmol) and 5-pyrimidi-
nylboronic acid (98 mg, 789 μmol). A solution of 2′-bromo-7′-
methoxyspiro[1,3-oxazole-4,9′-xanthen]-2-amine (190 mg, 526 μmol)
in 3 mL of dioxane was added, and the resulting mixture was flushed
for 30 s with argon. Sodium carbonate (789 μL, 1578 μmol) (2 M
aqueous solution) was added, and the reaction mixture was stirred at
90 °C for 18 h. The resulting mixture was partitioned between water
and DCM. The aqueous layer was extracted with DCM. The organic
layers were combined, washed with brine, dried over MgSO4, and
concentrated in vacuo. The residue was purified by RPHPLC (column,
Xbridge 19 mm × 100 mm, 10 μm; flow rate 40 mL/min; mobile
phase 0.1% TFA in ACN and water) to afford 2′-methoxy-7′-
(pyrimidin-5-yl)-5H-spiro[oxazole-4,9′-xanthen]-2-amine 2,2,2-tri-
fluoroacetate (9t) (90 mg, 47% yield). Mp 220.3−222.9 °C. 1H
NMR (500 MHz, DMSO-d6) δ 11.25 (br s, 1H), 9.61 (br s, 1H), 9.43
(br s, 1H), 9.24 (s, 2H), 9.21 (s, 1H), 8.20 (d, J = 2.14 Hz, 1H), 7.95
(dd, J = 2.19, 8.60 Hz, 1H), 7.41 (d, J = 8.65 Hz, 1H), 7.29 (d, J = 8.98
Hz, 1H), 7.15 (dd, J = 2.94, 9.03 Hz, 1H), 6.98 (d, J = 2.88 Hz, 1H),
5.14 (s, 2H), 3.82 (s, 3H). HRMS m/z = 361.1293 [M + H]+. Calcd
for C20 H17N4O3: 361.1300.
1
en]-2-amine (55.0 mg, 91.4% yield) as a white solid. H NMR (400
MHz, DMSO-d6) δ 9.18 (s, 1H), 9.08 (s, 2H), 7.75 (dd, J = 2.30, 8.46
Hz, 1H), 7.67 (d, J = 2.25 Hz, 1H), 7.28 (d, J = 8.51 Hz, 1H), 7.12 (d,
J = 8.90 Hz, 1H), 6.93 (dd, J = 3.03, 8.90 Hz, 1H), 6.83 (d, J = 3.03
Hz, 1H), 6.45 (br s, 2H), 4.23 (ABq, J = 8.31 Hz, 2H), 3.84 − 3.98
(m, 2H), 1.68−1.79 (m, 2H), 0.99 (t, J = 7.43 Hz, 3H). HRMS m/z =
389.1618 [M + H]+. Calcd for C22 H21N4O3: 389.1613.
Step d. 2′-Propoxy-7′-(5-pyrimidinyl)spiro[1,3-oxazole-4,9′-xanth-
en]-2-amine (40 mg) was subjected to chromatography using
15:85:0.2 MeOH/CO2/DEA at 80 mL/min on a 20 mm × 250
mm, 5 μm ChiralPak AS-H column and 100 bar system pressure. The
first peak (tR =3.5 min) corresponded to (R)-2′-propoxy-7′-(5-
pyrimidinyl)spiro[1,3-oxazole-4,9′-xanthen]-2-amine (13.0 mg, >99%
ee), and the second peak (tR =4.3 min) corresponded to (S)-2′-
propoxy-7′-(5-pyrimidinyl)spiro[1,3-oxazole-4,9′-xanthen]-2-amine
(12.8 mg, >99% ee).
ASSOCIATED CONTENT
■
S
* Supporting Information
Exemplified Analogues of Scheme 3. 2′-Propoxy-7′-(5-
pyrimidinyl)spiro[1,3-oxazole-4,9′-xanthen]-2-amine (10a).
Step a. A solution of 2′-bromo-7′-methoxyspiro[1,3-oxazole-4,9′-
xanthen]-2-amine (1.034 g, 2863 μmol) in DCM (29 mL) contained
in a 100 mL round-bottom flask was cooled in an ice bath for 15 min.
A solution of boron tribromide (8.5 mL of a 1.0 M solution in DCM,
8.59 mmol) was added dropwise over 5 min, resulting in a dark brown
solution. The ice bath was removed, and the mixture was stirred for 1.5
h. The reaction mixture was carefully quenched with saturated aqueous
sodium bicarbonate solution (30 mL). The mixture was partitioned
between water (50 mL) and DCM (50 mL). The aqueous layer was
extracted with DCM (2 × 25 mL), and the combined organic extracts
were dried over MgSO4. The solution was filtered, and the filter cake
was washed with 10% MeOH/DCM. The combined filtrates were
concentrated in vacuo. The residue was purified by chromatography
on silica gel (0−70% 90:10:1 DCM/MeOH/NH4OH in DCM) to
give 2′-bromo-7′-hydroxyspiro[1,3-oxazole-4,9′-xanthen]-2-amine
Experimental details, analytical data, and X-ray crystallographic
information for cocrystals of BACE + 6, BACE + 9l, and BACE
+ 11a. This material is available free of charge via the Internet
at http://pubs.acs.org.
Accession Codes
The PDB accession codes for the X-ray cocrystal structures of
BACE + 6, BACE + 9l, and BACE + 11a are 4FRI, 4FRJ, and
4FRK, respectively.
AUTHOR INFORMATION
■
Corresponding Author
*For H.H.: phone, 617-444-5197; fax, 617-621-3908; e-mail,
hongbing.huang@amgen.com. For D.S.L.: phone, 617-444-
5187; fax, 617-621-3908; e-mail, daniel.la@amgen.com.
1
(845 mg, 85.1% yield). H NMR (400 MHz, DMSO-d6) δ 9.34 (br
Present Address
△Sanofi-Aventis, 270 Albany Street, Cambridge, Massachusetts
02139, United States.
s, 1H), 7.46 (dd, J = 2.54, 8.71 Hz, 1H), 7.37 (d, J = 2.35 Hz, 1H),
7.08 (d, J = 8.70 Hz, 1H), 6.97 (d, J = 9.00 Hz, 1H), 6.67−6.76 (m,
2H), 6.51 (br s, 2H), 4.10 (d, J = 8.41 Hz, 1H), 4.07 (d, J = 8.41 Hz,
1H). MS m/z = 347.0 [M + H]+.
Step b. A 150 mL pressure vessel was charged with 2′-bromo-7′-
hydroxyspiro[1,3-oxazole-4,9′-xanthen]-2-amine (845 mg, 0.243
mmol) in THF (24 mL), pyrimidin-5-ylboronic acid (754 mg, 6085
μmol), tetrakis(triphenylphosphine)palladium(0) (281 mg, 243
μmol), and potassium carbonate (10.1 mL of a 1.2 M aqueous
Author Contributions
∞These authors contributed equally to the rational design of
the aminooxazoline xanthene structure.
Notes
The authors declare no competing financial interest.
K
dx.doi.org/10.1021/jm300598e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
(6) (a) Luo, Y.; Bolon, B.; Kahn, S.; Bennett, B. D.; Babu-Khan, S.;
Denis, P.; Fan, W.; Kha, H.; Zhang, J.; Gong, Y.; Martin, L.; Louis, J.
C.; Yan, Q.; Richards, W. G.; Citron, M.; Vassar, R. Mice deficient in
BACE1, the Alzheimer’s beta-secretase, have normal phenotype and
abolished beta-amyloid generation. Nat. Neurosci. 2001, 4 (3), 231−
232. (b) Cai, H.; Wang, Y.; McCarthy, D.; Wen, H.; Borchelt, D. R.;
Price, D. L.; Wong, P. C. BACE1 is the major beta-secretase for
generation of Abeta peptides by neurons. Nat. Neurosci. 2001, 4 (3),
233−234. (c) Roberds, S. L.; Anderson, J.; Basi, G.; Bienkowski, M. J.;
Branstetter, D. G.; Chen, K. S.; Freedman, S. B.; Frigon, N. L.; Games,
D.; Hu, K.; Johnson-Wood, K.; Kappenman, K. E.; Kawabe, T. T.;
Kola, I.; Kuehn, R.; Lee, M.; Liu, W.; Motter, R.; Nichols, N. F.;
Power, M.; Robertson, D. W.; Schenk, D.; Schoor, M.; Shopp, G. M.;
Shuck, M. E.; Sinha, S.; Svensson, K. A.; Tatsuno, G.; Tintrup, H.;
Wijsman, J.; Wright, S.; McConlogue, L. BACE knockout mice are
healthy despite lacking the primary beta-secretase activity in brain:
implications for Alzheimer’s disease therapeutics. Hum. Mol. Genet.
2001, 10 (12), 1317−1324.
ACKNOWLEDGMENTS
■
We thank Grace Bi and Larry Miller for HTPP and chiral
separation. We thank Dhanashri Bagal for HRMS analysis.
ABBREVIATIONS USED
■
AD, Alzheimer’s disease; Aβ, β-amyloid peptide; APP, β-
amyloid precursor protein; BACE1, β-site β-amyloid precursor
protein cleaving enzyme; CNS, central nervous system; CSF,
cerebrospinal fluid; FRET, fluorescence resonance energy
transfer; HE, hydroxyethylene; HEA, hydroxyethylamine; Pgp,
p-glycoprotein
REFERENCES
■
(1) (a) Cummings, J. L. Biomarkers in Alzheimer’s disease drug
development. Alzheimer's Dementia 2011, 7 (3), e13−e44. (b) Cum-
mings, J. L. Alzheimer’s disease. N. Engl. J. Med. 2004, 351 (1), 56−67.
(c) Citron, M. Alzheimer’s disease: strategies for disease modification.
Nat. Rev. Drug Discovery 2010, 9 (5), 387−98. (d) Jakob-Roetne, R.;
Jacobsen, H. Alzheimer’s disease: from pathology to therapeutic
approaches. Angew Chem., Int. Ed. 2009, 48 (17), 3030−3059.
(2) (a) Medina, M. Recent developments in tau-based therapeutics
for neurodegenerative diseases. Recent Pat. CNS Drug Discovery 2011,
6 (1), 20−30. (b) Mattson, M. P. Pathways towards and away from
Alzheimer’s disease. Nature 2004, 430 (7000), 631−639. (c) Nuss-
baum, R. L.; Ellis, C. E. Alzheimer’s disease and Parkinson’s disease. N.
Engl. J. Med. 2003, 348 (14), 1356−1364. (d) Selkoe, D. J. Translating
cell biology into therapeutic advances in Alzheimer’s disease. Nature
1999, 399 (6738 Suppl.), A23−A31.
(3) (a) Checler, F.; Turner, A. J. Journal of Neurochemistry special
issue on Alzheimer’s disease: “amyloid cascade hypothesis20 years
on”. J. Neurochem. 2012, 120 (Suppl. 1), iii−iv. (b) Armstrong, R. A.
The pathogenesis of Alzheimer’s disease: a reevaluation of the
“amyloid cascade hypothesis”. Int. J. Alzheimer's Dis. 2011, 2011,
630865. (c) Cummings, J. Alzheimer’s disease: clinical trials and the
amyloid hypothesis. Ann. Acad. Med. Singapore 2011, 40 (7), 304−303.
(d) Sinha, S.; Lieberburg, I. Cellular mechanisms of beta-amyloid
production and secretion. Proc. Natl. Acad. Sci. U.S.A. 1999, 96 (20),
11049−11053.
(4) (a) Wang, Z.; Yang, L.; Zheng, H. Role of APP and Abeta in
synaptic physiology. Curr. Alzheimer Res. 2012, 9 (2), 217−226.
(b) Chasseigneaux, S.; Allinquant, B. Functions of Abeta, sAPPalpha
and sAPPbeta: similarities and differences. J. Neurochem. 2012, 120
(Suppl. 1), 99−108. (c) Guo, Q.; Li, H.; Gaddam, S. S.; Justice, N. J.;
Robertson, C. S.; Zheng, H. Amyloid precursor protein revisited:
neuron-specific expression and highly stable nature of soluble
derivatives. J. Biol. Chem. 2012, 287 (4), 2437−2445. (d) Farah, M.
H.; Pan, B. H.; Hoffman, P. N.; Ferraris, D.; Tsukamoto, T.; Nguyen,
T.; Wong, P. C.; Price, D. L.; Slusher, B. S.; Griffin, J. W. Reduced
BACE1 activity enhances clearance of myelin debris and regeneration
of axons in the injured peripheral nervous system. J. Neurosci. 2011, 31
(15), 5744−5754. (e) Nikolaev, A.; McLaughlin, T.; O’Leary, D. D.;
Tessier-Lavigne, M. APP binds DR6 to trigger axon pruning and
neuron death via distinct caspases. Nature 2009, 457 (7232), 981−
989.
(5) (a) Cho, J. E.; Kim, J. R. Recent approaches targeting beta-
amyloid for therapeutic intervention of Alzheimer’s disease. Recent Pat.
CNS Drug Discovery 2011, 6 (3), 222−233. (b) Ganjei, J. K. Targeting
amyloid precursor protein secretases: Alzheimer’s disease and beyond.
Drug News Perspect. 2010, 23 (9), 573−584. (c) Silvestri, R. Boom in
the development of non-peptidic beta-secretase (BACE1) inhibitors
for the treatment of Alzheimer’s disease. Med. Res. Rev. 2009, 29 (2),
295−338. (d) Albert, J. S. Progress in the development of beta-
secretase inhibitors for Alzheimer’s disease. Prog. Med. Chem. 2009, 48,
133−161. (e) Guo, T.; Hobbs, D. W. Development of BACE1
inhibitors for Alzheimer’s disease. Curr. Med. Chem. 2006, 13 (15),
1811−1829.
(7) Ghosh, A. K.; Brindisi, M.; Tang, J. Developing beta-secretase
inhibitors for treatment of Alzheimer’s disease. J. Neurochem. 2012,
120 (Suppl. 1), 71−83.
(8) (a) Safety Study of CTS21166 To Treat Alzheimer Disease.
http://clinicaltrials.gov/ct2/show/NCT00621010 (accessed April
2012). (b) Drahl, C. Beta Testing. Chem. Eng. News 2010, 88, 14.
(9) (a) Huang, X. P.; Chang, W. P.; Koelsch, G.; Turner, R. T., 3rd;
Lupu, F.; Tang, J. Internalization of exogenously added memapsin 2
(beta-secretase) ectodomain by cells is mediated by amyloid precursor
protein. J. Biol. Chem. 2004, 279 (36), 37886−37894. (b) He, X.;
Cooley, K.; Chung, C. H.; Dashti, N.; Tang, J. Apolipoprotein receptor
2 and X11 alpha/beta mediate apolipoprotein E-induced endocytosis
of amyloid-beta precursor protein and beta-secretase, leading to
amyloid-beta production. J. Neurosci. 2007, 27 (15), 4052−4060.
(c) Rajendran, L.; Schneider, A.; Schlechtingen, G.; Weidlich, S.; Ries,
J.; Braxmeier, T.; Schwille, P.; Schulz, J. B.; Schroeder, C.; Simons, M.;
Jennings, G.; Knolker, H. J.; Simons, K. Efficient inhibition of the
Alzheimer’s disease beta-secretase by membrane targeting. Science
2008, 320 (5875), 520−523.
(10) (a) Vassar, R.; Bennett, B. D.; Babu-Khan, S.; Kahn, S.;
Mendiaz, E. A.; Denis, P.; Teplow, D. B.; Ross, S.; Amarante, P.;
Loeloff, R.; Luo, Y.; Fisher, S.; Fuller, J.; Edenson, S.; Lile, J.;
Jarosinski, M. A.; Biere, A. L.; Curran, E.; Burgess, T.; Louis, J. C.;
Collins, F.; Treanor, J.; Rogers, G.; Citron, M. Beta-secretase cleavage
of Alzheimer’s amyloid precursor protein by the transmembrane
aspartic protease BACE. Science 1999, 286 (5440), 735−741. (b) Yan,
R.; Bienkowski, M. J.; Shuck, M. E.; Miao, H.; Tory, M. C.; Pauley, A.
M.; Brashier, J. R.; Stratman, N. C.; Mathews, W. R.; Buhl, A. E.;
Carter, D. B.; Tomasselli, A. G.; Parodi, L. A.; Heinrikson, R. L.;
Gurney, M. E. Membrane-anchored aspartyl protease with Alzheimer’s
disease beta-secretase activity. Nature 1999, 402 (6761), 533−537.
(c) Sinha, S.; Anderson, J. P.; Barbour, R.; Basi, G. S.; Caccavello, R.;
Davis, D.; Doan, M.; Dovey, H. F.; Frigon, N.; Hong, J.; Jacobson-
Croak, K.; Jewett, N.; Keim, P.; Knops, J.; Lieberburg, I.; Power, M.;
Tan, H.; Tatsuno, G.; Tung, J.; Schenk, D.; Seubert, P.; Suomensaari,
S. M.; Wang, S.; Walker, D.; Zhao, J.; McConlogue, L.; John, V.
Purification and cloning of amyloid precursor protein beta-secretase
from human brain. Nature 1999, 402 (6761), 537−540.
(11) Stachel, S. J. Progress toward the development of a viable BACE
inhibitor. Drug Dev. Res. 2009, 70, 101−110.
(12) (a) Cole, D. C.; Manas, E. S.; Stock, J. R.; Condon, J. S.;
Jennings, L. D.; Aulabaugh, A.; Chopra, R.; Cowling, R.; Ellingboe, J.
W.; Fan, K. Y.; Harrison, B. L.; Hu, Y.; Jacobsen, S.; Jin, G.; Lin, L.;
Lovering, F. E.; Malamas, M. S.; Stahl, M. L.; Strand, J.; Sukhdeo, M.
N.; Svenson, K.; Turner, M. J.; Wagner, E.; Wu, J.; Zhou, P.; Bard, J.
Acylguanidines as small-molecule beta-secretase inhibitors. J. Med.
Chem. 2006, 49 (21), 6158−6161. (b) Edwards, P. D.; Albert, J. S.;
Sylvester, M.; Aharony, D.; Andisik, D.; Callaghan, O.; Campbell, J. B.;
Carr, R. A.; Chessari, G.; Congreve, M.; Frederickson, M.; Folmer, R.
H.; Geschwindner, S.; Koether, G.; Kolmodin, K.; Krumrine, J.;
Mauger, R. C.; Murray, C. W.; Olsson, L. L.; Patel, S.; Spear, N.; Tian,
L
dx.doi.org/10.1021/jm300598e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
G. Application of fragment-based lead generation to the discovery of
novel, cyclic amidine beta-secretase inhibitors with nanomolar
potency, cellular activity, and high ligand efficiency. J. Med. Chem.
2007, 50 (24), 5912−5925. (c) Congreve, M.; Aharony, D.; Albert, J.;
Callaghan, O.; Campbell, J.; Carr, R. A.; Chessari, G.; Cowan, S.;
Edwards, P. D.; Frederickson, M.; McMenamin, R.; Murray, C. W.;
Patel, S.; Wallis, N. Application of fragment screening by X-ray
crystallography to the discovery of aminopyridines as inhibitors of
beta-secretase. J. Med. Chem. 2007, 50 (6), 1124−1132. (d) Baxter, E.
W.; Conway, K. A.; Kennis, L.; Bischoff, F.; Mercken, M. H.; Winter,
H. L.; Reynolds, C. H.; Tounge, B. A.; Luo, C.; Scott, M. K.; Huang,
Y.; Braeken, M.; Pieters, S. M.; Berthelot, D. J.; Masure, S.; Bruinzeel,
W. D.; Jordan, A. D.; Parker, M. H.; Boyd, R. E.; Qu, J.; Alexander, R.
S.; Brenneman, D. E.; Reitz, A. B. 2-Amino-3,4-dihydroquinazolines as
inhibitors of BACE-1 (beta-site APP cleaving enzyme): use of
structure based design to convert a micromolar hit into a nanomolar
lead. J. Med. Chem. 2007, 50 (18), 4261−4264. (e) Malamas, M. S.;
Erdei, J.; Gunawan, I.; Turner, J.; Hu, Y.; Wagner, E.; Fan, K.; Chopra,
R.; Olland, A.; Bard, J.; Jacobsen, S.; Magolda, R. L.; Pangalos, M.;
Robichaud, A. J. Design and synthesis of 5,5′-disubstituted amino-
hydantoins as potent and selective human beta-secretase (BACE1)
inhibitors. J. Med. Chem. 2010, 53 (3), 1146−1158.
Zega, A. Metabolism-directed optimisation of antithrombotics: the
prodrug principle. Curr. Pharm. Des 2006, 12 (1), 73−91.
(23) Kaller, M. R.; Harried, S. S.; Albercht, B.; Amarante, P.; Babu-
Khan, S.; Bartberger, M. D.; Brwon, J.; Brown, R.; Chen, K.; Cheng,
Y.; Citron, M.; Croghan, M. D.; Graceffa, R.; Hickman, D.; Judd, T.;
Kriemen, C.; La, D.; Li, V.; Lopez, P.; Luo, Y.; Masse, C.;
Monenschein, H.; Nguyen, T.; Pennington, L. D.; Miguel, T. S.;
Sickmier, E. A.; Wahl, R. C.; Weiss, M. M.; Wen, P. H.; Williamson,
T.; Wood, S.; Xue, M.; Yang, B.; Zhang, J.; Patel, V.; Zhong, W.;
Hitchcock, S. A potent and orally efficacious, hydroxyethylamine-based
inhibitor of β-secretase. ACS Med. Chem. Lett. [Online early access].
DOI: 10.1021/ml3000148. Published Online: April 22, 2012.
(24) (a) Ward, J. M.; Gorenstein, N. M.; Tian, J.; Martin, S. F.; Post,
C. B. Constraining binding hot spots: NMR and molecular dynamics
simulations provide a structural explanation for enthalpy-entropy
compensation in SH2-ligand binding. J. Am. Chem. Soc. 2010, 132
(32), 11058−11070. (b) Schneider, H.-J. In Protein−Ligand
Interactions: From Molecular Recognition to Drug Design; Bohm, H.-J.,
̈
Schneider, G., Eds; Wiley-VCH: Weinheim, Germany, 2003; pp 21−
50.
(25) (a) Turner, R. T., 3rd; Koelsch, G.; Hong, L.; Castanheira, P.;
Ermolieff, J.; Ghosh, A. K.; Tang, J. Subsite specificity of memapsin 2
(beta-secretase): implications for inhibitor design. Biochemistry 2001,
40 (34), 10001−10006. (b) Patel, S.; Vuillard, L.; Cleasby, A.; Murray,
C. W.; Yon, J. Apo and inhibitor complex structures of BACE (beta-
secretase). J. Mol. Biol. 2004, 343 (2), 407−416.
(13) Cole, D. C.; Bursavich, M. G. Nonpeptide BACE 1 Inhibitors:
Design and Synthesis. In Methods and Principles in Medicinal Chemistry;
Ghosh, A. K., Ed.; Wiley-VCH: Weinheim, Germany, 2010; Vol. 45,
pp 481−509.
(26) Hong, L.; Turner, R. T., 3rd; Koelsch, G.; Shin, D.; Ghosh, A.
K.; Tang, J. Crystal structure of memapsin 2 (beta-secretase) in
complex with an inhibitor OM00-3. Biochemistry 2002, 41 (36),
10963−10967.
(14) The Asp binding moiety would act as the pharmocaphore and
therefore be implemented into every compound henceforth. Thus, this
portion required the heaviest physiochemical design restrictions.
Wager, T. T.; Villalobos, A.; Verhoest, P. R.; Hou, X.; Shaffer, C. L.
Strategies to optimize the brain availability of central nervous system
drug candidates. Expert Opin. Drug Discovery 2011, 6, 371−381.
(15) Hitchcock, S. A.; Pennington, L. D. Structure−brain exposure
relationships. J. Med. Chem. 2006, 49 (26), 7559−7583.
(16) pKa values were calculated using the following: ACD LogD Suite,
version 10; Advanced Chemistry Development Inc.: Toronto, Canada.
(17) Benchmark compounds synthesized containing cores B and H
demonstrated efflux ratios of >40.
(18) Rajamani, R.; Reynolds, C. H. Modeling the protonation states
of the catalytic aspartates in beta-secretase. J. Med. Chem. 2004, 47
(21), 5159−5166.
(19) (a) Sidera, C.; Liu, C.; Austen, B. Pro-domain removal in ASP-2
and the cleavage of the amyloid precursor are influenced by pH. BMC
Biochem. 2002, 3, 25. (b) Dominguez, J. L.; Christopeit, T.; Villaverde,
M. C.; Gossas, T.; Otero, J. M.; Nystrom, S.; Baraznenok, V.;
Lindstrom, E.; Danielson, U. H.; Sussman, F. Effect of the protonation
state of the titratable residues on the inhibitor affinity to BACE-1.
Biochemistry 2010, 49 (34), 7255−7263.
(27) (a) Stachel, S. J.; Coburn, C. A.; Steele, T. G.; Jones, K. G.;
Loutzenhiser, E. F.; Gregro, A. R.; Rajapakse, H. A.; Lai, M. T.;
Crouthamel, M. C.; Xu, M.; Tugusheva, K.; Lineberger, J. E.; Pietrak,
B. L.; Espeseth, A. S.; Shi, X. P.; Chen-Dodson, E.; Holloway, M. K.;
Munshi, S.; Simon, A. J.; Kuo, L.; Vacca, J. P. Structure-based design of
potent and selective cell-permeable inhibitors of human beta-secretase
(BACE-1). J. Med. Chem. 2004, 47 (26), 6447−6450. (b) Coburn, C.
A.; Stachel, S. J.; Li, Y. M.; Rush, D. M.; Steele, T. G.; Chen-Dodson,
E.; Holloway, M. K.; Xu, M.; Huang, Q.; Lai, M. T.; DiMuzio, J.;
Crouthamel, M. C.; Shi, X. P.; Sardana, V.; Chen, Z.; Munshi, S.; Kuo,
L.; Makara, G. M.; Annis, D. A.; Tadikonda, P. K.; Nash, H. M.; Vacca,
J. P.; Wang, T. Identification of a small molecule nonpeptide active site
beta-secretase inhibitor that displays a nontraditional binding mode for
aspartyl proteases. J. Med. Chem. 2004, 47 (25), 6117−6119.
(28) van de Waterbeemd, H.; Camenisch, G.; Folkers, G.; Chretien,
J. R.; Raevsky, O. A. Estimation of blood−brain barrier crossing of
drugs using molecular size and shape, and H-bonding descriptors. J.
Drug Targeting 1998, 6 (2), 151−165.
(20) (a) Stachel, S. J.; Coburn, C. A.; Rush, D.; Jones, K. L.; Zhu, H.;
Rajapakse, H.; Graham, S. L.; Simon, A.; Katharine Holloway, M.;
Allison, T. J.; Munshi, S. K.; Espeseth, A. S.; Zuck, P.; Colussi, D.;
Wolfe, A.; Pietrak, B. L.; Lai, M. T.; Vacca, J. P. Discovery of
aminoheterocycles as a novel beta-secretase inhibitor class: pH
dependence on binding activity part 1. Bioorg. Med. Chem. Lett.
2009, 19 (11), 2977−2980. (b) Rajapakse, H. A.; Nantermet, P. G.;
Selnick, H. G.; Barrow, J. C.; McGaughey, G. B.; Munshi, S.; Lindsley,
S. R.; Young, M. B.; Ngo, P. L.; Holloway, M. K.; Lai, M. T.; Espeseth,
A. S.; Shi, X. P.; Colussi, D.; Pietrak, B.; Crouthamel, M. C.;
Tugusheva, K.; Huang, Q.; Xu, M.; Simon, A. J.; Kuo, L.; Hazuda, D.
J.; Graham, S.; Vacca, J. P. SAR of tertiary carbinamine derived BACE1
inhibitors: role of aspartate ligand amine pKa in enzyme inhibition.
Bioorg. Med. Chem. Lett. 2010, 20 (6), 1885−1889.
(29) (a) Obach, R. S. Nonspecific binding to microsomes: impact on
scale-up of in vitro intrinsic clearance to hepatic clearance as assessed
through examination of warfarin, imipramine, and propranolol. Drug
Metab. Dispos. 1997, 25 (12), 1359−1369. (b) Obach, R. S. Prediction
of human clearance of twenty-nine drugs from hepatic microsomal
intrinsic clearance data: an examination of in vitro half-life approach
and nonspecific binding to microsomes. Drug Metab. Dispos. 1999, 27
(11), 1350−1359. (c) Smith, D. A.; van de Waterbeemd, H.; Walker,
D. K. In Pharmacokinetics and Metabolishm in Drug Design; Mannhold,
R., Kubinyi, H., Folkers, G., Eds; Wiley-VCH: Weinheim, Germany,
2006; pp 165−177.
(30) The relatively low clearance of 11a is likely a result of high
plasma protein binding. The protein binding in rat plasma for 9u and
11a is 98% and 91.3%, respectively.
(21) Reynolds, D. P.; Lanevskij, K.; Japertas, P.; Didziapetris, R.;
Petrauskas, A. Ionization-specific analysis of human intestinal
absorption. J. Pharm. Sci. 2009, 98 (11), 4039−4054.
(31) Dineen, T. A.; Weiss, M. M.; Williamson, T.; Acton, P.; Babu-
Khan, S.; Bartberger, M. D.; Brown, J.; Chen, K.; Cheng, Y.; Citron,
M.; Croghan, M. D.; Dunn, R. T., 2nd; Esmay, J.; Graceffa, R. F.;
Harried, S. S.; Hickman, D.; Hitchcock, S. A.; Horne, D. B.; Huang,
H.; Imbeah-Ampiah, R.; Judd, T.; Kaller, M. R.; Kreiman, C. R.; La, D.
S.; Li, V.; Lopez, P.; Louie, S.; Monenschein, H.; Nguyen, T. T.;
Pennington, L. D.; San Miguel, T.; Sickmier, E. A.; Vargas, H. M.;
(22) (a) Kotthaus, J.; Hungeling, H.; Reeh, C.; Schade, D.; Wein, S.;
Wolffram, S.; Clement, B. Synthesis and biological evaluation of L-
valine-amidoximeesters as double prodrugs of amidines. Bioorg. Med.
Chem. 2011, 19 (6), 1907−1914. (b) Peterlin-Masic, L.; Cesar, J.;
M
dx.doi.org/10.1021/jm300598e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Wahl, R. C.; Wen, P. H.; Whittington, D. A.; Wood, S.; Xue, Q.; Yang,
B. H.; Patel, V. F.; Zhong, W. Design and synthesis of potent, orally
efficacious hydroxyethylamine derived beta-site amyloid precursor
protein cleaving enzyme (BACE1) inhibitors. J. Med. Chem. [Online
early access]. DOI: 10.1021/jm300118s. Published Online: April 18,
2012.
(32) Marcoux, J.; Doye, S; Buchwald, S. L. A general copper-
catalyzed syntesis of diaryl ethers. J. Am. Chem. Soc. 1997, 119, 10539−
10540.
(33) Peterson, D. J. Carbonyl olefination reaction using silyl-
substituted organometallic compounds. J. Org. Chem. 1968, 33, 780−
784.
(34) Fernandez, J. M. G.; Mellet, C. O.; Adrian, M. A. P.; Mota, J. F.
Syntheses and spectral properties of β-iodoureas and 2-amino-4,4-
diphenyl-2-oxazolines. J. Heterocycl. Chem. 1991, 28, 777−780.
(35) Haniu, M.; Denis, P.; Young, Y.; Mendiaz, E. A.; Fuller, J.; Hui,
J. O.; Bennett, B. D.; Kahn, S.; Ross, S.; Burgess, T.; Katta, V.; Rogers,
G.; Vassar, R.; Citron, M. Characterization of Alzheimer’s beta-
secretase protein BACE. A pepsin family member with unusual
properties. J. Biol. Chem. 2000, 275 (28), 21099−21106.
(36) Schinkel, A. H.; Wagenaar, E.; van Deemter, L.; Mol, C. A.;
Borst, P. Absence of the mdr1a P-glycoprotein in mice affects tissue
distribution and pharmacokinetics of dexamethasone, digoxin, and
cyclosporin A. J. Clin. Invest. 1995, 96 (4), 1698−1705.
(37) Booth-Genthe, C. L.; Louie, S. W.; Carlini, E. J.; Li, B.; Leake, B.
F.; Eisenhandler, R.; Hochman, J. H.; Mei, Q.; Kim, R. B.; Rushmore,
T. H.; Yamazaki, M. Development and characterization of LLC-PK1
cells containing Sprague-Dawley rat Abcb1a (Mdr1a): comparison of
rat P-glycoprotein transport to human and mouse. J. Pharmacol.
Toxicol. Methods 2006, 54 (1), 78−89.
(38) Otwinowski, Z.; Minor, W. Processing of X-ray diffraction data
collected in oscillation mode. Methods Enzymol. 1997, 276, 307−326.
(39) (a) Navaza, J. AMoRe: an automated package for molecular
replacement. Acta Crystallogr. A 1994, 50, 157−163. (b) Murshudov,
G. N.; Vagin, A. A.; Dodson, E. J. Refinement of macromolecular
structures by the maximum-likelihood method. Acta Crystallogr., Sect.
D: Biol. Crystallogr. 1997, 53 (Part 3), 240−255.
(40) Emsley, P.; Cowtan, K. Coot: model-building tools for
molecular graphics. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2004,
60, 2126−2132.
N
dx.doi.org/10.1021/jm300598e | J. Med. Chem. XXXX, XXX, XXX−XXX