Di-substituted pyridinyl aminohydantoins as potent and highly selective human β-secretase (BACE1) inhibitors

Article abstract of DOI:10.1016/j.bmc.2009.12.007

The identification of highly selective small molecule di-substituted pyridinyl aminohydantoins as β-secretase inhibitors is reported. The more potent and selective analogs demonstrate low nanomolar potency for the BACE1 enzyme as measured in a FRET assay, and exhibit comparable activity in a cell-based (ELISA) assay. In addition, these pyridine-aminohydantoins are highly selectivity (>500×) against the other structurally related aspartyl proteases BACE2, cathepsin D, pepsin and renin. Our design strategy followed a traditional SAR approach and was supported by molecular modeling studies based on the previously reported aminohydantoin 3a. We have taken advantage of the amino acid difference between the BACE1 and BACE2 at the S2′ pocket (BACE1 Pro₇₀ changed to BACE2 Lys₈₆) to build ligands with >500-fold selectivity against BACE2. The addition of large substituents on the targeted ligand at the vicinity of this aberration has generated a steric conflict between the ligand and these two proteins, thus impacting the ligand's affinity and selectivity. These ligands have also shown an exceptional selectivity against cathepsin D (>5000-fold) as well as the other aspartyl proteases mentioned. One of the more potent compounds (S)-39 displayed an IC₅₀ value for BACE1 of 10 nM, and exhibited cellular activity with an EC₅₀ value of 130 nM in the ELISA assay.

Full text of DOI:10.1016/j.bmc.2009.12.007

Bioorganic & Medicinal Chemistry 18 (2010) 630–639
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Di-substituted pyridinyl aminohydantoins as potent and highly selective human
b-secretase (BACE1) inhibitors
b
b
b
a
c
c
Michael S. Malamas a, , Keith Barnes , Matthew Johnson , Yu Hui , Ping Zhou , Jim Turner , Yun Hu ,
Erik Wagner c, Kristi Fan a, Rajiv Chopra d, Andrea Olland d, Jonathan Bard c, Menelas Pangalos c,
Peter Reinhart c, Albert J. Robichaud a
*
a Department of Chemical Sciences, Wyeth Research, CN 8000, Princeton, NJ 08543-8000, USA
b Albany Molecular Research, Albany, NY, USA
c Neuroscience, Wyeth Research, CN 8000, Princeton, NJ 08543-8000, USA
d Department of Chemical Sciences, Wyeth Research, 200 Cambridge Park Drive, Cambridge, MA 02140, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The identification of highly selective small molecule di-substituted pyridinyl aminohydantoins as b-
secretase inhibitors is reported. The more potent and selective analogs demonstrate low nanomolar
potency for the BACE1 enzyme as measured in a FRET assay, and exhibit comparable activity in a cell-
based (ELISA) assay. In addition, these pyridine-aminohydantoins are highly selectivity (>500ꢀ) against
the other structurally related aspartyl proteases BACE2, cathepsin D, pepsin and renin.
Our design strategy followed a traditional SAR approach and was supported by molecular modeling stud-
ies based on the previously reported aminohydantoin 3a. We have taken advantage of the amino acid dif-
ference between the BACE1 and BACE2 at the S20 pocket (BACE1 Pro70 changed to BACE2 Lys86) to build
ligands with >500-fold selectivity against BACE2. The addition of large substituents on the targeted ligand
at the vicinity of this aberration has generated a steric conflict between the ligand and these two proteins,
thus impacting the ligand’s affinity and selectivity. These ligands have also shown an exceptional selec-
tivity against cathepsin D (>5000-fold) as well as the other aspartyl proteases mentioned. One of the
more potent compounds (S)-39 displayed an IC50 value for BACE1 of 10 nM, and exhibited cellular activity
with an EC50 value of 130 nM in the ELISA assay.
Received 14 October 2009
Revised 30 November 2009
Accepted 2 December 2009
Available online 6 December 2009
Keywords:
Alzheimer’s disease (AD)
BACE1 inhibitors
b-Amyloid peptide (Ab)
Aminohydantoins
Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction
secretase complexes (intramembrane aspartyl proteases),7 as part
of the b-amyloidogenic pathway. During this process, two b-secre-
Alzheimer’s disease (AD) is a progressive, neurodegenerative
disease of the brain and is recognized as the leading cause of
dementia. At the early stage, AD is associated with gradual loss
of cognition that leads to complete deterioration of cognitive,
and behavioral functions and ultimately death. The pathological
hallmarks of AD include the extracellular deposition of b-amyloid
peptide (Ab), which leads to aggregation and plaque formation,
and the abnormal hyperphosphorylation of tau protein, which
leads to the intracellular formation of neurofibrillary tangles.1,2
b-Amyloid deposits are predominately composed of the Ab pep-
tides (Ab, 39–43 residues) resulting from the endoproteolysis of
the amyloid precursor protein (APP).3,4 Neurofibrillary tangles are
intracellular aggregates of the microtubule associated protein
tau.5 Ab peptides result from the sequential cleavage of APP, first
at the N-terminus by b-secretase enzyme (b-site APP cleaving en-
tase cleavage products are produced; a secreted ectodomain frag-
ment named APPsol, and the membrane bound C-terminal
fragment C99 of APP. Following b-secretase cleavage, a second pro-
tease, c-secretase, cleaves C99 to generate the toxic Ab peptides
(Ab, 39–43 residues) which are secreted from the cell. Although,
the cause of AD remains unknown, a large body of evidence is
beginning to accumulate that highlights the central role of Ab in
the pathogenesis of the disease.8–11 Thus, processes that limit Ab
production and deposition by preventing formation, inhibiting
aggregation, and/or enhancing clearance may offer effective treat-
ments for AD. Since b-secretase mediated cleavage of APP is the
first and rate-limiting step of the amyloidogenic possessing path-
way, BACE1 inhibition is considered a prominent therapeutic tar-
get for treating AD by diminishing Ab peptide formation in AD
patients.
zyme, BACE1),6,1 followed at the C-terminus by one or more
c-
Recently, we have disclosed the discovery of small molecule
aminohydantoins as potent BACE1 inhibitors.12 In early SAR inves-
tigations we quickly learned that the truncation of the tetrahydro-
pyrimidine portion of the high-throughput screening hit 1 (Fig. 1)
* Corresponding author. Tel.: +1 732 2744428.
E-mail address: malamam@wyeth.com (M.S. Malamas).
0968-0896/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2009.12.007

M. S. Malamas et al. / Bioorg. Med. Chem. 18 (2010) 630–639
631
H2N
H2N
Me
H2N
Me
N
N
N
N
R
N
N
N
O
O
N
truncate
SAR
OMe
Me
3a; R = F; 3b; R = H
IC50 = 0.02; 0.04 uM
IC50 = 0.78; 0.79 uM
IC50 = 2.60; 12.5 uM
1(HTS Hit)
BACE1: IC50 = 38 uM
BACE2: IC50 = 38.3 uM
Cathepsin D: IC50 >100 uM
2
IC50 = 3.4 uM
IC50 = 0.77 uM
IC50 >100 uM
Figure 1.
which led to the more compact aminohydantoin 2, possessed a 10-
fold enhancement in potency (IC50 = 3.4 M). Soon afterward we
ences between the two proteins at the extended ligand binding
pocket. In particular, we have considered that the residue differ-
ences (BACE1 Pro70 is replaced by BACE2 Lys86) among the two
proteins at the S20 pocket as a genuine opportunity for SAR inves-
tigation to potentially improve a ligand’s selectivity. This contin-
uing interest in this part of the extended BACE pocket is
noteworthy as we have demonstrated in earlier SAR studies with
aminohydantoin 3b that interaction between the ligand and the
enzyme at the S20 region might contribute to the ligand’s
selectivity.13
In this paper, we report our efforts to develop BACE1 inhibitors
with enhanced selectivity against BACE2, by methodologically
exploring the SAR around the S20 pocket of the enzyme. As de-
scribed, our intent was to capitalize on the residue differences be-
tween the BACE1 and BACE2 enzymes in this micro-region and
investigate the feasibility of rationally designing highly selective
ligands.
l
carried out detailed structure–activity relationship studies13
around compound 2 that ultimately led to compound 3a, which re-
sulted in 1900-fold improvement of ligand affinity (3a vs 1) for the
BACE1 enzyme (IC50 = 20 nM; Fig. 1). In addition, derivative 3a has
demonstrated about 40- and 100-fold selectivity against the close
related aspartyl proteases BACE2 and cathepsin D, respectively.
This modest selectivity over these two enzymes was a conse-
quence of the initial SAR development, but not one that was fo-
cused on as a primary endpoint.
The physiological functions of BACE2 have yet to be fully char-
acterized, and although BACE1/BACE2 double knockout mice have
been reported to be viable14 selective inhibitors against BACE2 are
highly desirable to avoid potential side effects in human clinical
trials. More importantly, selectivity against cathepsin D was
deemed to be essential to the clinical development of BACE1 inhib-
itors, since cathepsin D plays a critical role in cellular processes and
functions as a positive mediator in apoptosis.15
2. Chemistry
In an effort to further enhance the selectivity of the aminohyd-
antoin BACE1 inhibitors, we have closely examined the differences
of the amino acid sequences of BACE1 and BACE2 as it regards
binding of potential ligands to the active site. Not unexpectedly,
the catalytic domain of BACE1 is similar to that of the BACE2 with
79% sequence identity within the active site. However, an overlay
of the X-ray structure of BACE1 complexed with 1 and the homol-
ogy model of BACE2 (Fig. 2) has revealed several amino acid differ-
The compounds needed to delineate the SAR for this study were
prepared according to synthetic Scheme 1. In general, two routes
were used for the formation of biaryl acetylenes 5. In route a, Sono-
gashira coupling16 of disubstituted 4-ethynylpyridines 4 with 1-
bromo-3-iodobenzene afforded acetylenes 5 in 75% yield. Alterna-
tively, for enhanced SAR diversity, bromo-pyridines used in route
b, where palladium-catalyzed cross coupling reaction of 1-bro-
mo-3-iodobenzene with ethynyl(trimethyl)silane furnished adduct
6, which upon hydrolysis to the corresponding phenyl-acetylene 7,
and further Sonogashira coupling with bromo-pyridines 8 afforded
acetylenes 5, in 43% yield over the three steps. Oxidation of acety-
lenes 5 with potassium permanganate in the presence of magne-
sium sulfate and sodium bicarbonate furnished diketones 9, in
75% yield. These 1,2 di-substituted diketones were converted to
the aminohydantoins 10 upon treatment with 1-methylguanidine
in the presence of potassium carbonate in excellent yields.13 Palla-
dium-catalyzed cross coupling reaction of aminohydantoins 10
with any number of heteroaryl boronic acids 11 (Suzuki cou-
pling17) in the presence of suitable Pd(0) or Pd(II) catalysts pro-
duced the desired products 12.
The required key starting acetylenes, 4 and 4-bromopyridines 8
(Scheme 1) were either commercially available or prepared accord-
ing to Scheme 2, employing Methods A–E. In Method A, 2-alkyl
pyridines 14 were assembled according to the reaction conditions
first reported by Comin’s,18 where initial activation of the pyridines
13 with phenyl chloroformate followed by the addition of a Grig-
nard reagent and oxidation with o-chloranil to furnish the desired
adducts 14, in 53% yield. Sonogashira coupling of pyridines 14 with
ethynyl(trimethyl)silane and hydrolysis, as before, afforded acety-
lenes 15. In Method B, 2,6-dimethylpyridin-4-ol was treated with
phosphorus oxychloride to yield 4-chloropyridine 17, which upon
Figure 2. Crystal structure of BACE1 complexed with 1 (shown in green) and BACE2
homology (shown in magenta) are overlayed. Key amino acid differences between
BACE1 (yellow) and BACE2 (magenta) are highlighted.

632
M. S. Malamas et al. / Bioorg. Med. Chem. 18 (2010) 630–639
R6
N
Br
R6
N
Br
a
b
I
+
route a
R2
4
5
R2
R2, R6 = H, alkyl
SiMe3
a, c
R6
route b
Br
N
Br
a
8
R2
X
6; X = SiMe3
7; X = H
c
Me
H2N
R6
N
N
NH
O
N
O
CH3
Br
R6
H2N
N
Br
H
R2
N
d
O
9
R2
10
R2'
Me
H2N
R2'
X
Y
N
X
B(OH)2
11
N
O
Y
R6
X, Y = C, N
d
N
R2
12
Scheme 1. Reagents and conditions: (a) Pd(PPh3)2Cl2, CuI, Et3N, DMF; (b) K2CO3,MeOH; (c) KMnO4, NaHCO3, MgSO4, acetone; (d) methylguanididneꢂHCl, Na2CO3, EtOH,
water; (e) aryl or heteroaryl boronic acids Pd(PPh3)4, Na2CO3, 1,2-diethoxyethane, water.
palladium-catalyzed coupling with 2-methylbut-3-yn-2-ol and
subsequent hydrolysis with potassium hydroxide19 afforded acety-
lene 18, in 25% yield over the three steps. In Method C, 4-bromo-
2,6-dimethylpyridine 19 was alkylated with methyl iodide in the
presence of lithium di-isopropyl amide20 to furnish bromo-pyri-
dine 20, in 76% yield. In Method D, condensation of pentane-2,4-
dione with methyl 2-methylpropanoate in the presence of sodium
hydride, was followed by treatment with ammonium hydroxide to
afford pyridin-4(1H)-one 2221, in 53% yield over the two steps.
Compound 22 was converted to 23 upon treatment with phospho-
rus oxybromide. In Method E, self-condensation of 3-oxopentane-
dioic acid under acidic conditions furnished 2,6-diethyl-4H-
pyran-4-one22 25, which was then treated with ammonium
hydroxide23 and phosphorous oxybromide to afford the desired
product 27, in 38% yield over the three steps.
RE(EDANS)-IHPFHLVIHTK(DABCYL)-R for renin. Kinetic rates were
calculated and IC50 values were determined by fitting the % inhibi-
tion, as a function of compound concentration, to the Hill equation
(y = ((B ꢁ Kn) + (100 ꢁ xn))/(Kn + xn). In practice, we have routinely
screened all prepared compounds for BACE1, BACE2 and cathepsin
D inhibition and then selected compounds were assayed in the
pepsin and renin screens, based on their meeting the screening
protocol for affinity to the target. These data are reported in Tables
1 and 2.
Cellular potency of advancing compounds was done via a cell-
based Ab inhibition (Ab40 or Ab42) in an enzyme-linked immune
sandwich assay (ELISA) in Chinese Hamster Ovary (CHO) cells,
recombinantly expressing human wild-type APP (CHO-wt). The
concentration at which the cellular production of Ab40 or Ab42
was reduced by 50% (EC50) was determined and reported in the
data tables. Potential compound toxicity was assessed via mito-
chondrial function using a MTS readout (MTS kit from Promega)
and values are represented as LD50 or the dose of compound that
resulted in 50% of control signal.
3. Results and discussion
With the necessary tools needed to fully investigate the SAR, the
diverse array of compounds were profiled for their potency at the
target enzyme, BACE1, as well as the closely related enzyme sites
that were the focus of this investigation. The primary screening as-
says utilized for the program were homogenous, continuous fluo-
rescence resonance energy transfer (FRET) protocols, representing
competitive inhibition for BACE1, BACE2, cathepsin D, pepsin and
renin.24 The BACE1 and BACE2 activities were based on the cleav-
age of peptide substrate Abz-SEVNLDAEFR-Dpa (Swedish sub-
strate), while peptide substrate MOCAc-GKPILFFRLK (Dnp)-D-R-
NH2 was used for cathepsin D and pepsin, and peptide substrate
In this report, we have undertaken studies to methodically
investigate the SAR of the aminohydantoins at the S20 pocket of
the enzyme. As stated, our goal was to develop selective BACE1
inhibitors, by taking advantage of the amino acid difference be-
tween the BACE1 and BACE2 (BACE1 Pro70 is replaced by BACE2
Lys86) enzymes in this region. Towards this objective, we have used
X-ray structures of BACE1 co-crystallized with various ligands and
molecular modeling studies to help us in the design of such inhib-
itors. The initial unsubstituted baseline pyridine analog 28 showed
a modest threefold improvement in selectivity for BACE1 vs phenyl
derivative 2. As we have described in great detail in our previous

M. S. Malamas et al. / Bioorg. Med. Chem. 18 (2010) 630–639
633
c, d
SiMe3
a, b
N
N
Br
N
Br
A)
B)
R2MgCl
R2
R2
R2 = alkyl
13
14
17
15
18
Me
N
Me
N
Me
N
f
e, c
OH
OH
Me
Me
Me
16
Me
N
Me
g
C)
D)
Br
N
Br
Me
19
20
Me
Me
N
O
O
O
Br
h, i
j
N
H
Me
23
22
21
Me
Me
O
n, g
O
O
O
E)
k, l, m
N
X
O
HO
OH
25
24
Me
26; X = OH
27; X = Br
Scheme 2. Methods A–E. Reagents and conditions: (a) R2MgCl, PhOCOCl, THF; (b) o-chloranil, AcOH, toluene; (c) Pd(PPh3)2Cl2, CuI, Et3N, DMF; (d) K2CO3,MeOH; (e) POCl3; (f)
KOH, toluene; (g) LDA, THF, MeI; (h) NaH, methyl 2-methylpropanoate, DME; (i) NaOH; (j) POBr3; (k) H2SO4; (l) Na2CO3, AcOH; (m) HCl; (n) NH4OH.
Table 1
Pyridine-aminohydantoins
Me
H2N
N
N
O
R3'
R6
N
R2
a
a
R30
Compd
R2
R6
BACE1 IC50
(lM)
BACE2 IC50
(
l
M)
Cathepsin D % inh.b at 100
10
l
M
ELISA EC50 (lM)
28
29
30
31
32
33
34
35
36
37
38
S-38
R-38
S-39
H
H
H
H
H
H
H
H
H
H
Me
Me
Me
Et
Et
Et
Et
H
Ph
2.68
2.01 0.4
0.14 0.04
0.48 0.15
1.17
2.2 1.04
1.02
6.13
24.1
28% at 12.5 lM
5.5 0.9
0.9 0.4
0.3 0.05
3.8 1.3
0.13 0.05
0.06 0.01
0.1 0.002
0.08 0.001
0.03 0.01
0.07 0.01
0.14 0.01
0.15 0.03
0.3 0.14
0.04 0.009
0.03 0.003
IC50 = 30 7.4 lM
3-Pyridine
3-Pyridine
5-Pyrimidine
5-Pyrimidine
5-Pyrimidine
5-Pyrimidine
5-Pyrimidine
5-Pyrimidine
5-Pyrimidine
5-Pyrimidine
5-Pyrimidine
3-(2F-Pyridine)
45
44
45
49
47
14
17
29
32
22
31
Me
Me
Et
i-Pr
Me
Et
i-Pr
Et
Et
2.3
1
0.9 0.16
1.9 0.36
3.7 1.2
1.9 0.3
1.5 0.14
0.85 0.5
24% at 25
l
M
23.6
38% at 12.5
11 2.7
l
M
0.27 0.07
>2500
0.13 0.04
Et
Et
48% at 5
lM
0.01 0.005
6.7 0.8
IC50 = 54
2
lM
a
IC50 and EC50 values are the means of at least two experiments SD.
Experiments were performed in triplicate.
b
work,12 molecular modeling and X-ray crystallographic studies
have shown that building off the meta-position of the phenyl moi-
ety (occupying the S1 pocket) of potent biphenyl aminohydantoin
derivatives would allow for projection directly towards the unoc-

634
M. S. Malamas et al. / Bioorg. Med. Chem. 18 (2010) 630–639
Table 2
Pepsin and renin inhibition of selected compounds
Compd
Pepsin % inh.a at 200
lM
Renin % inh. at 200 lM
36
37
S-38
S-39
Inactive
Inactive
9
Inactive
Inactive
Inactive
26
13
a
Experiments were performed in triplicate.
cupied S3 region (see Fig. 3). This was proposed to have compara-
ble effects in the pyridine series, and to that end, we have intro-
duced a phenyl group at the meta-position of the phenyl moiety.
Not surprisingly, that led to a 20-fold improvement in potency
(29 vs 28), confirming this hypothesis. The analogous pyridine ana-
log 30 was approximately 45-fold better for BACE1 (30 vs 28),
again mirroring the previous SAR results. To confirm our modeling
calculations and our hypothesis, 30 was co-crystallized with
BACE1. As illustrated in Figure 4, the pyridine moiety projects deep
into S3 pocket and makes a water–bridge contact with Ser229
through the buried water found near the catalytic site of the en-
zyme. This water–bridge interaction together with the additional
van der Walls contacts between the ligand and the enzyme back-
bone at the S3 region are proposed to account for the ligand’s in-
creased potency. The X-ray structure also confirms, as before,
that the amino group of the ligand interacts with both aspartic
acids (Asp32 and Asp228) and the N3 nitrogen of the imidazole
ring interacts with Asp32 via a hydrogen-bonding network. Fur-
thermore, the BACE1:30 structure indicates that the pyridine nu-
cleus projecting into the S20 pocket makes a hydrogen-bonding
contact with residue Trp76, and may be partly responsible for
the increased potency. Compound 30 exhibited modest selectivity
(ꢃ8-fold) against BACE2, and it was poorly active against cathepsin
Figure 4. Crystal structure of BACE 1 complexed with 30. Key hydrogen-bonding
interactions between ligand 30 and protein at the catalytic aspartic acids Asp32 and
Asp228, Trp76 at the S20 region, and the water–bridge to Ser229 are highlighted
with yellow dashed lines.
offered an avenue to improve the BACE1 selectivity of 30 by target-
ing the BACE1 Pro70 to BACE2 Lys86 difference and generating a ste-
ric conflict between the ligand and the backbone of the BACE2
enzyme. To our satisfaction, introduction of a methyl group (entry
31; Table 1) at position 2 of the pyridine nucleus modestly affected
the ligand potency (two-fold decrease) for BACE1, while slightly
improving the ligand’s selectivity (about two-fold) against BACE2.
Replacement of the pyridine nucleus projecting into the S3 with a
pyrimidine moiety (32) resulted in an additional enhancement
(two-fold) of the ligand’s selectivity. With these initial results sup-
porting our hypothesis, we further increased the size of the ortho-
substituent of the pyridine moiety that projects into the S20 pocket.
The derivative bearing an ethyl group at position 2 was approxi-
mately 34-fold selective against BACE2 (entry 33), showing a
marked improvement in selectivity, while the even bulkier isopro-
pyl moiety (entry 34) demonstrated broader selectivity (87-fold vs
BACE2). The cathepsin D selectivity of these analogs tracked very
well with the previous findings, with all compounds exhibiting
D (45% inhibition at 100 lM). The importance of sustained activity
in the cell was not lost during this investigation. Pyridine 30
tracked well with its increased molecular binding with EC50 values
of 300 nM in the cell-based ELISA assay.
Close examination of the BACE1:30 X-ray structure revealed a
key opportunity that we believed would allow for selectivity opti-
mization between BACE1 and BACE2. We hypothesized that substi-
tution at position 2 of the pyridine group located in the S20 pocket,
weak affinity (IC50 ꢃ100
lM) for this site.
One cannot neglect the possibility that mere rotation of the pyr-
idine ring in this S20 pocket would move the alkyl group offering
the apparent selectivity from interacting with the backbone of
the BACE2 enzyme. This would effectively diminish the utility of
the 2-substituted pyridyl adducts, that being to cause an unfavor-
able interaction with the BACE2 backbone. Furthermore, molecular
modeling studies have indicated that this binding mode (rotation
of the alkyl group away from the Lys86 in BACE2) would be accom-
modated by the enzyme sites of both BACE1 and BACE2. In order to
maintain a sterically biased effect in this pocket, we have endeav-
ored to prepare di-substituted analogs at positions 2 and 6 of the
pyridine nucleus to abrogate the enzymes ability to remove the
offending interaction. To that end, introduction of methyl groups
at positions 2 and 6 of the pyridine nucleus was executed and this
derivative produced a marked increase in ligand selectivity (170-
fold; 35 vs 31). Having confirmed our hypothesis that the 2,6-pyr-
idine substitutions are optimal to effectively target the Pro70 to
Lys86 difference and improve BACE1 selectivity, we have systemat-
ically varied the size and steric encumbrance of the substituents at
the 2,6-positions of the pyridine nucleus to further optimize the li-
gand’s selectivity. This can be seen with both di-substituted ana-
logs, ethyl-methyl (36) and isopropyl-methyl (37), as they
Figure 3. Crystal structure of BACE1 complexed with 3b13 (shown in yellow) and
pyridine 28 (shown in green) are overlayed. Key hydrogen-bonding interactions
between ligand 3b and protein at the catalytic aspartic acids Asp32 and Asp228,
Trp76 at the S20 region, and the water–bridge to Ser229 are highlighted with yellow
dashed lines.

M. S. Malamas et al. / Bioorg. Med. Chem. 18 (2010) 630–639
635
possess only marginal affinity (ꢃ28% at 25
l
M) for BACE2, while
that projects towards the FLAP region interacts with Val69, while
the distal ethyl moiety interacts with Ser35. Furthermore, The
BACE1:(S)-38 structure revealed, similar to the previous X-ray
findings,12 that the aminohydantoin portion of the ligand directly
interacts with both catalytic-site aspartic acids (Asp32 and
Asp228) via a hydrogen-bonding network, and the pyrimidine moi-
ety of the ligand extends deep into the S3 pocket and makes a
water–bridge contact with Ser229. Lastly, the nitrogen of the pyr-
idine nucleus located at the S20 pocket makes a hydrogen-bond
interaction with residue Trp76.
Furthermore, an investigation of the chiral preference of the
racemic mixtures of this class of compounds, once separated,
showed a clear preference for the S-enantiomer [see (S)-38 vs
(R)-38]. This was apparent as the crystal structures and modeling
studies bore this out. For the less active enantiomer, one could
envision keeping the respective P1–P3 and P20 side-chains in their
designed pockets, but in so doing, the aminohydantoin would be
flipped into an unfavorable confirmation and pressed into the
backbone of the active site of the enzyme.
the BACE1 affinity was practically unaffected. The bulkier isopropyl
group has produced a slight ꢃ2-fold reduction of the BACE1 affinity
(36 vs 37). Furthermore, the ethyl–ethyl di-substituted analog 38
was the most selective ligand, showing 590-fold preference for
BACE1. Next, we have replaced the pyrimidine nucleus of 38 with
the previously optimized 2-F-pyridyl moiety,14 which was discov-
ered during the optimization of the SAR studies of compound 3a.
The 2,6-diethyl pyridyl analog 39 showed high potency for BACE1
(IC50 = 10 nM) and exquisite selectivity (670ꢀ) against BACE2.
Compound 39 was also highly selective >5000ꢀ against cathepsin
D. Furthermore, the cell-based activity of pyridine 39 tracked well
compared to its molecular binding with an EC50 value of 130 nM in
the ELISA assay.
In support of our modeling studies and design of selective
BACE1 ligands, compound (S)-38 was co-crystallized with the
BACE1 enzyme. Overlapping of the X-ray crystal structures of com-
pounds (S)-38 and 30 bound to the enzyme indicated a nearly per-
fect superimposition of these two ligands (Fig. 5). Close
examination of these overlapping structures, however, has re-
vealed a shift of the ‘flap’ loop region of the protein backbone of
the enzyme resulting in a more open orientation of the loop
(approximately 5 Å shift vs apo-position) for the more selective
compound (S)-38. Notice that with the unsubstituted ligand 30,
the loop is in a more closed orientation (apo-position), which has
obvious profound impact on the selectivity versus BACE2. The open
orientation of the ‘flap’ region in the BACE1:(S)-38 structure is per-
haps attributed to the steric influence exerted by the 2-ethyl group
of the S20-pyridine moiety. As discussed previously, the protein
amino acid sequences of the ‘flap’ loops of the BACE1 and BACE2
proteins differ in the S20 pocket (BACE1 Pro70 is replaced by BACE2
Lys86 and BACE1 Arg128 is replaced by Lys in BACE2) at close prox-
imity to the bound ligand. This minor amino acid difference to-
gether with the dynamic motion of the ‘flap’ loop upon ligand
engagement apparently influences the ligand/protein contacts
and impedes the ligand’s affinity for the BACE2 site, thus resulting
in increased selectivity. Additionally, both ethyl groups make
favorable contact with other residues of the S20 pocket backbone
of the BACE1 enzyme, thus contributing to the ligand’s affinity.
More specifically, the crystal structure shows that the ethyl group
To confirm the more general specificity of these ligands for the
BACE1 enzyme, a representative set of compounds were evaluated
for inhibition of renin and pepsin aspartyl proteases. As shown in
Table 2, all compounds demonstrated weak inhibition for these
closely related targets confirming their potential as selective
BACE1 ligands.
4. Conclusions
In this report, we have described a detailed and stepwise explo-
ration of substituted pyridine-aminohydantoins that led to the dis-
covery of highly potent and selective BACE1 inhibitors. The more
potent the selective analogs demonstrate low nanomolar potency
(IC50 = 10nM) for BACE1 in a FRET assay, and exhibit comparable
activity in a cell-based (ELISA) assay. In addition, these aminohyd-
antoins show >500-fold selectivity toward the other structurally
related aspartyl proteases BACE2, cathepsin D, pepsin, and renin.
Our design strategy followed a traditional SAR approach and
was supported by molecular modeling studies based on the previ-
ously reported aminohydantoin 3. We have taken advantage of the
amino acid difference between the BACE1 and BACE2 enzymes at
the S20 pocket (BACE1 Pro70 changed to BACE2 Lys86) to build li-
gands with >500-fold selectivity against BACE2. The addition of
large steric substituents at the vicinity of this mutation has gener-
ated an apparent conflict between the ligand and the BACE2 en-
zyme backbone, thus affecting the ligand’s affinity and
selectivity. These ligands have also shown a distinguished selectiv-
ity against cathepsin D (>5000). One of the more potent com-
pounds (S)-39 displayed an IC50 value for BACE1 of 10 nM, and
exhibited cellular activity with an EC50 value of 130 nM in the ELI-
SA assay. Ligand (S)-39 also showed >500-fold selectivity the other
related aspartyl proteases BACE2, cathepsin D, pepsin, and renin.
5. Experimental section
5.1. Chemistry
Melting points were determined in open capillary tubes on a
Mel-Temp-II apparatus, and reported uncorrected. 1H NMR spectra
were determined in the cited solvent on a Varian Unity or Varian
Inova (400 MHz) instrument, with tetramethylsilane as an internal
standard. Chemical shifts are given in ppm and coupling constants
are in hertz. Splitting patterns are designated as follows: s, singlet;
br s, broad singlet; d, doublet; t, triplet; q, quartet; m, multiplet.
The infrared spectra were recorded on a AVATAR 360 Nicolet spec-
trophotometer as KBr pellets or as solutions in chloroform. Mass
Figure 5. Crystal structures of BACE 1 complexed with (S)-38 (shown in yellow)
and 30 (shown in magenta) are overlayed. Movement of Tyr71 from the closed
position in complex 30 to the open orientation in complex with (S)-38 is shown. Key
hydrogen-bonding interactions between the ligand and residues Tyr71, Val69 and
Ser35 at the S20 region are shown in yellow dotted lines.

636
M. S. Malamas et al. / Bioorg. Med. Chem. 18 (2010) 630–639
spectra were recorded on a Micromass LCT, Waters spectrometer.
Elemental analyses (C, H, N) were performed on a Perkin–Elmer
240 analyzer and all compounds are within 0.4% of theory unless
otherwise indicated. HPLC techniques and high resolution mass
spectrometry were used to determine the purity of compounds
outside the range of the elemental analysis assessment. Purity of
all final products was >96% as determined by HPLC and/or combus-
tion analysis. Purity was determined by HPLC analysis using the
following protocols. Method A: Mobile Phase A: 10 mM ammo-
nium formate in water (pH 3.5); B: 50:50 ACN/MeOH; solvent gra-
dient 85:15 to 5:95 A:B in 2 min., hold 1.25 min, re-equilibrate
5.2.3. Preparation of 2-amino-5-(3-bromophenyl)-3-methyl-5-
(pyridin-4-yl)-3,5-dihydro-4H-imidazol-4-one (10, R2, R6 = H)
A
suspension of 1-phenyl-2-pyridin-4-ylethane-1,2-dione
(0.81 g, 3.8 mmol), N-methylguanidine hydrochloride (1.16 g,
4 mmol) and Na2CO3 (1.9 g, 18 mmol) in EtOH (80 mL), dioxane
(62 mL) and water (18 mL) was stirred at 85 °C for 1 h. The reaction
mixture was cooled to room temperature, poured into water and ex-
tracted with CHCl3 (3 ꢀ 60 mL). The combined organic extracts were
dried over MgSO4. Evaporation of the solvents and purification by
flash chromatography (silica gel, dichloromethane/MeOH, 10:1)
afforded 2-amino-5-(3-bromophenyl)-3-methyl-5-(pyridin-4-yl)-
3,5-dihydro-4H-imidazol-4-one as an off-white solid (0.73 g, 53%):
H NMR (300 MHz, DMSO-d6) d 2.95 (s, 3H, Me), 6.83 (br s, 2H, NH2),
7.34–7.3–7.35 (m, 1H, Ar-H), 7.36–7.38 (m, 2H, Ar-H), 7.43–7.45
(m, 2H, Ar-H), 7.6 (t, 1H, J = 1.83 Hz, Ar-H), 8.48 (dd , 2H, J = 4.53,
1.71 Hz, Ar-H); MS m/z 345 (M+H)+. Anal. Calcd for C15H13BrN4O: C,
52.19; H, 3.58; N, 16.23. Found: C, 52.15; H, 3.79; N, 16.35.
0.5 min; flow rate 1.1 mL/min; Column Agilent SB C18 1.8 lM,
3.0 ꢀ 50 mm; temperature 45 °C; detection at 210–370 nM. Meth-
od B: mobile phase water with A = 0.05% v/v trifluoroacetic acid,
B = acetonitrile with 0.05% v/v trifluoroacetic acid; solvent gradient
90:10 to 10:90 A:B in 20 min; flow rate 1.0 mL/min; Waters Sym-
metry C18 column (4.6 ꢀ 250 mm) with UV detection at 254 nm.
All products, unless otherwise noted, were purified by ‘flash chro-
matography’ with use of 220–400 mesh silica gel. Thin-layer chro-
matography was done on Silica Gel 60 F-254 (0.25 mm thickness)
plates. Visualization was accomplished with UV light and/or 10%
phosphomolybdic acid in ethanol. The hydration was determined
by the Karl Fischer titration, using a Mitsubishi moisture meter
Model CA-05. Unless otherwise noted, all materials were obtained
commercially and used without further purification. All reactions
were carried out under an atmosphere of dried argon or nitrogen.
5.2.4. Preparation of 2-amino-5-(2,6-diethylpyridin-4-yl)-3-
methyl-5-[3-(pyrimidin-5-yl)phenyl]-3,5-dihydro-4H-imidazol-
4-one (12, R2, R6 = Et, R0 = H, X, Y = N).
2
A mixture of 2-amino-5-(3-bromophenyl)-5-(2,6-diethylpyri-
din-4-yl)-3-methyl-3,5-dihydro-4H-imidazol-4-one
(0.12 g,
0.297 mmol), pyrimidine-5-boronic acid (0.044 g, 0.355 mmol),
tetrakis(triphenylphosphino)palladium(0) (0.018 g, 0.016 mmol)
and sodium carbonate (0.083 g, 0.783 mmol) in ethylene glycol di-
methyl ether (8 mL) and water (4 mL) was heated at reflux for 1 h.
The mixture was cooled to room temperature, concentrated, and
the residue partitioned between CH2Cl2 (50 mL) and water
(50 mL). The layers were separated, and the aqueous layer was ex-
tracted with methylene chloride (2 ꢀ 25 mL) and the combined or-
ganic extracts were dried over Na2SO4. Evaporation of the solvents
and purification by flash chromatography (silica gel, 96:4:0.5
methylene chloride/methanol/concentrated ammonium hydrox-
ide) afforded 2-amino-5-(2,6-diethylpyridin-4-yl)-3-methyl-5-(3-
pyrimidin-5-ylphenyl))-3,5-dihydro-4H-imidazol-4-one as an off-
white solid (0.059 g, 49% yield): mp 204.5 °C; 1H NMR (300 MHz,
CDCl3) d 1.26 (t, J = 7.6 Hz, 6H, 2 ꢀ CH2CH3), 2.76 (q, J = 7.6 Hz,
4H, 2 ꢀ CH2CH3), 3.14 (s, 3H, Me), 4.75 (br s, 2H, NH2), 7.15 (s,
2H, Ar-H), 7.49 (m, 2H, Ar-H), 7.62 (s, 1H, Ar-H), 7.74 (s, 1H, Ar-
H), 8.92 (s, 2H, Ar-H), 9.20 (s, 1H, Ar-H); MS m/z 401 (M+H)+. Anal.
Calcd for C23H24NO6 ꢀ 1H2O: C, 66.01; H, 6.26; N, 20.0: Found: C,
66.09; H, 6.20; N, 19.66.
5.2. Representative synthetic protocols of the aminohydantoins
shown in Scheme 1 are described below
5.2.1. Preparation of 4-[(3-bromophenyl)ethynyl]pyridine (5,
R2, R6 = H, route a)
To a solution of 1-bromo-3-iodobenzene (5.0 g, 17.6 mmol)
in DMF (50 mL) were added dichlorobis(triphenylphos-
phine)palladium (0.23 g, 0.29 mmol), copper iodide (0.03 g,
0.16 mmol), triethylamine (6.7 mL, 48.5 mmol) and 4-ethynyl-
pyridine (1.0 mL, 9.7 mmol). The reaction mixture was heated
at 65 °C for 3 h, and then cooled and quenched with water
(100 mL). The aqueous mixture was extracted with EtOAc
(3 ꢀ 50 mL) and the combined organic extracts were washed
with brine (50 mL), and dried over MgSO4. Evaporation of the
solvents and purification by flash chromatography (silica gel,
EtOAc/hexane 1:3) afforded 4-[(3-bromophenyl)ethynyl]pyri-
dine as a yellow solid (1.9 g, 76% yield): 1H NMR (300 MHz,
DMSO-d6) d 7.38–7.40 (m, 1H, Ar-H), 7.49–7.51(m, 2H, Ar-H),
7.6–7.61 (m, 1H, Ar-H), 7.65–7.65 (m, 1H, Ar-H), 7.81 (t, 1H,
J = 1.58 Hz, Ar-H), 8.61–8.62 (m, 2H, Ar-H); MS m/z 258
(M+H)+; Anal. Calcd for C13H8BrN ꢀ 0.1 H2O: C, 60.07; H, 3.18;
N, 5.38. Found: C, 60.02; H, 3.18; N, 5.1.
5.2.5. Preparation of [(3-bromophenyl)ethynyl](trimethyl)-
silane (6, X = SiMe3)
To a solution of 1-bromo-3-iodobenzene (4.1 g, 14.5 mmol) in
DMF (50 mL) were added dichlorobis(triphenylphosphine)palla-
dium (0.31 g, 0.43 mmol), copper iodide (0.055 g, 0.16 mmol), tri-
ethylamine (10 mL, 72.5 mmol) and ethynyl(trimethyl)silane
(2.05 mL, 14.5 mmol). The reaction mixture was heated at 65 °C
for 3 h, cooled at room temperature, and quenched with water
(100 mL). The aqueous mixture was extracted with EtOAc
(3 ꢀ 50 mL) and the combined organic extracts were washed with
brine (50 mL), and dried over MgSO4. Evaporation of the solvents
and purification by flash chromatography (silica gel, hexane) afford
[(3-bromophenyl)ethynyl](trimethyl)silane as a yellow oil (3.0 g,
82% yield): 1H NMR (300 MHZ, DMSO-d6) d 0.19 (s, 3H, SiMe3),
7.28–7.3 (m, 1H, Ar-H), 7.4(7.2-(m, 1H, Ar-H), 7.55–7.6 (m, 2H,
Ar-H); MS m/z 252 (M+H)+.
5.2.2. Preparation of 1-(3-bromophenyl)-2-pyridin-4-ylethane-
1,2-dione (9, R2, R6 = H)
A
solution of 4-[(3-bromophenyl)ethynyl]pyridine (1.81 g,
7.0 mmol) in acetone (63 mL) was added into a warm (40 °C) mix-
ture of NaHCO3 (0.35 g, 4.20 mmol) and MgSO4 (1.26 g,
10.50 mmol) in water (63 mL). Potassium permanganate (2.43 g,
15.40 mmol) was added in one portion and the reaction mixture
was stirred at room temperature for 4 min, poured into water
(300 mL) and extracted with 1:1 Et2O/hexane. The organic extracts
were dried over MgSO4. Evaporation of the solvents afforded 1-(3-
bromophenyl)-2-pyridin-4-ylethane-1,2-dione as a yellow solid
(1.52 g, 75% yield): mp 88–90 °C; 1H NMR (300 MHz, DMSO-d6) d
7.56–7.57 (m, 1H, Ar-H), 7.81–7.82 (m, 2H, Ar-H), 7.95–7.97 (m,
2H, Ar-H), 8.1 (t, 1H, J = 1.71 Hz, Ar-H), 8.86 (dd , 2H, J = 4.39,
1.59 Hz, Ar-H): MS m/z 289 (M+H)+. Anal. Calc for C13H8BrNO2: C,
53.82; H, 2.78; N, 4.83. Found: C, 53.49; H, 2.73; N, 4.63.
5.2.6. Preparation of 1-bromo-3-ethynylbenzene (7, X = H)
To
a solution of [(3-bromophenyl)ethynyl](trimethyl)silane
(10 g, 39.5 mmol) in dichloromethane (150 mL) and ethanol
(150 mL) was added Cs2CO3 (13 g, 39.9 mmol) at room tempera-
ture. After stirring for 4 h, the reaction mixture was diluted with

M. S. Malamas et al. / Bioorg. Med. Chem. 18 (2010) 630–639
637
Et2O (300 mL) and washed with water and brine. Evaporation of the
solvents and distillation (2 psi, 55 °C) afforded a clear oil (6.8 g,
93.9% yield): 1H NMR (300 MHZ, DMSO-d6) d 4.29 (s, 1H, acety-
lene–H), 7.28–7.30 (m, 1H, Ar-H), 7.43–7.46 (m,1H, Ar-H), 7.57–
7.59 (m, 1H, Ar-H), 7.63–7.64(m, 1H, Ar-H): MS m/z 180 M+. Anal.
Calcd for C8H5Br: C, 53.08;, H, 2.78: Found: C, 52.78; H, 2.65.
ride (2.63 g, 0.17 mol) under N2. The mixture was warmed to
reflux for 1 h then cooled to room temperature. Excess phospho-
rous oxychloride was removed under reduced pressure, and the
residue diluted with dichloroethane (50 mL). The mixture was
again concentrated to provide 4-chloro-2,6-dimethylpyridine
(2.0 g) as a red oil. This material was dissolved in DMF (65 mL) un-
der N2, and treated with bis(triphenylphosphino)palladium(II)
chloride (315 mg, 0.45 mmol), triphenylphosphine (243 mg,
0.93 mmol), copper(I) iodide (268 mg, 1.42 mmol), triethylamine
5.2.7. Preparation of 4-[(3-bromophenyl)ethynyl]pyridine (5,
R2, R6 = H, route b)
(11.1 g,
110 mmol),
and
2-methyl-3-butyn-2-ol
(2.43 g,
To a solution of 1-bromo-3-ethynylbenzene (0.8 g, 4.4 mmol) in
DMF (20 mL) were added dichlorobis(triphenylphosphine)palla-
dium (0.09 g, 0.13 mmol), copper iodide (0.09 g, 0.46 mmol), tri-
ethylamine (3.1 mL, 22 mmol) and 4-bromopyridine (1.05 g,
94.4 mmol). The reaction mixture was heated at 65 °C for 4 h, then
cooled to room temperature and quenched with water (100 mL).
The aqueous mixture was extracted with EtOAc (3 ꢀ 50 mL) and
the combined organic extracts were washed with brine (50 mL),
and dried over MgSO4. Evaporation and purification by flash chro-
matography (silica gel, EtOAc/hexane 1:4) afforded 4-[(3-bromo-
phenyl)ethynyl]pyridine as a yellow solid (0.6 g, 53%): 1H NMR
(300 MHz, DMSO-d6) d 7.38–7.40 (m, 1H, Ar-H), 7.49–7.51(m, 2H,
Ar-H), 7.6–7.61 (m, 1H, Ar-H), 7.65–7.65 (m, 1H, Ar-H), 7.81 (t,
1H, J = 1.58 Hz, Ar-H), 8.61–8.62 (m, 2H, Ar-H); MS m/z 258
(M+H)+. Anal. Calcd for C13H8BrN ꢀ 0.1 H2O: C, 60.07; H, 3.18; N,
5.38. Found: C, 59.92; H, 3.12; N, 5.23.
289 mmol). The mixture was heated at 100 °C overnight and
cooled to room temperature. Ethyl acetate (400 mL) and water
(200 mL) were added, the layers were separated, and the aqueous
layer was extracted with ethyl acetate (2 ꢀ 100 mL). The combined
organic layers were washed with 2% aq lithium chloride
(2 ꢀ 150 mL), and dried Na2SO4. Evaporation of the solvents and
purification by flash chromatography (silica gel, 100:0–98:2 meth-
ylene chloride/methanol) afforded 4-(2,6-dimethylpyridin-4-yl)-2-
methylbut-3-yn-2-ol as a brown solid (0.96 g, 36% yield): 1H NMR
(300 MHz, CDCl3) d 1.61 (s, 6H, 2 ꢀ Me); 2.14 (s, 1H, Me), 2.50 (s,
6H, 2 ꢀ Me), 6.97 (s, 2H, Ar-H); MS m/z 190 (M+H)+.
5.3.2.2. Step f. Preparation of 4-ethynyl-2,6-dimethylpyridine
(18). A solution of 4-(2,6-dimethylpyridin-4-yl)-2-methylbut-3-
yn-2-ol (102 mg, 0.54 mmol) in toluene (5 mL) under N2 was trea-
ted with powdered potassium hydroxide (51.0 mg, 0.91 mmol),
and the mixture was heated at reflux for 1 h. The mixture was
cooled to room temperature, diluted with ethyl acetate (100 mL)
and water (100 mL), and the layers were separated. The aqueous
layer was extracted with ethyl acetate (2 ꢀ 50 mL), and the com-
bined organic extracts were dried over Na2SO4, filtered, and con-
centrated to afford 4-ethynyl-2,6-dimethylpyridine as a light
orange solid (0.05 g, 70% yield): 1H NMR (300 MHz, CDCl3) d 2.51
(s, 6H, 2 ꢀ Me); 3.20 (s, 1H, acetylene–H), 7.04 (s, 2H, Ar-H): MS
m/z 132 (M+H)+.
5.3. The preparation of the substituted pyridines shown in
Scheme 2 were prepared according to the synthetic Methods A–E
5.3.1. Method A
5.3.1.1. Steps a and b. Preparation of 4-bromo-2-methylpyri-
dine (14, R2 = Me). To a cooled (ꢄ78 °C) suspension of 4-bromo-
pyridine hydrochloride (5.0 g, 25.7 mmol) in anhydrous THF
(90 mL) was added dropwise a solution of MeMgCl (3.0 M in THF,
21 mL, 63.0 mmol). After the addition, the reaction mixture was
stirred at ꢄ78 °C for 15 min. Phenyl chloroformate (3.8 mL,
30 mmol) in THF (10 mL) was added slowly and the mixture was al-
lowed to warm to room temperature. Afterwards, the reaction was
cooled at 0 °C, quenched with saturated NH4Cl and extracted with
Et2O. The combined organic extracts were successively washed
with water, HCl (1 N), and brine. The organic extracts were dried
over MgSO4 and the solvents were evaporated. The residue was dis-
solved in anhydrous toluene (100 mL) and a solution of o-chloranil
(7.8 g, 32 mmol) in AcOH (60 mL) was added dropwise and the mix-
ture was stirred for 22 h. A red suspension was formed and it was
then the mixture was made basic by using 10% NaOH until a black
emulsion was obtained. The mixture was filtered through a Celite
pad and washed with water. The organic layer was extracted three
times with HCl (1 N). The combined aqueous extracts were basified
with 50% NaOH and extracted with CH2Cl2. The combined organic
extracts were dried over MgSO4 and the solvents were removed un-
der vacuum to afford 4-bromo-2-methylpyridine as yellow oil
(2.35 g, 53% yield): 1H NMR (300 MHz, DMSO-d6) d 2.4 (s, 3H,
Me), 7.42 (1H, dd, J = 5.37, 1.95 Hz, Ar-H), 7.52 (d, 1H, J = 1.46 Hz,
Ar-H), 8.3 (d, 1H, J = 5.37, Ar-H); MS m/z 171 M+.
5.3.3. Method C
5.3.3.1. Preparation
of
4-bromo-2-ethyl-6-methylpyridine
(20). To a solution of diisopropylamine (1.49 g, 14.8 mmol) in tet-
rahydrofuran (55 mL) at ꢄ78 °C was added dropwise n-butyllith-
ium (10.2 mL of a 1.6 M solution in hexanes, 16.3 mmol). The
mixture was stirred for 15 min at ꢄ78 °C, warmed to 0 °C for
15 min and then cooled to ꢄ78 °C. The solution of lithium diisopro-
pylamide was then transferred via a cannula to a stirred solution of
commercially available 4-bromo-2,6-dimethyl pyridine (2.75 g,
14.8 mmol) in tetrahydrofuran (30 mL) at -78 °C and the mixture
stirred for 1 h. Methyl iodide (2.31 g, 16.3 mmol) was added and
the reaction stirred for an additional 30 min. Saturated aqueous
NH4Cl (20 mL) was then added, and the mixture was allowed to
warm to room temperature and then diluted with CH2Cl2
(200 mL) and water (100 mL). The organic layer was separated
and washed successively with water (100 mL) and brine
(100 mL), and dried over sodium sulfate. Evaporation of the sol-
vents and purification by flash chromatography (silica gel, 90:10
hexanes/ethyl acetate) afforded 4-bromo-2-ethyl-6-methylpyri-
dine as a yellow oil (2.26 g, 76% yield): 1H NMR (300 MHz, CDCl3)
d 1.28 (t, J = 7.6 Hz, 3H, CH2CH3), 2.50 (s, 3H, Me), 2.75 (q, J = 7.6 Hz,
2H, CH2CH3), 7.16 (s, 2H, Ar-H): MS m/z 200 (M+H)+.
5.3.1.2. Steps c and d. The preparation of 4-ethynyl-2-methyl-
pyridine 15 (R2 = Me; Scheme 2) was prepared from 4-bromo-2-
methylpyridine (14, R1 = Me) and ethynyl(trimethyl)silane accord-
ing to the synthetic protocols described in Scheme 1.
5.3.4. Method D
5.3.4.1. Steps h and i. Preparation of 2-methyl-6-(propan-2-
yl)pyridin-4(1H)-one (22). To a suspension of sodium hydride
(2.00 g of a 60% dispersion in oil, 50.0 mmol) and ethylene glycol
dimethyl ether (35 mL) at reflux was added dropwise a solution
of acetyl acetone (1.00 g, 10.0 mmol) and methyl isobutyrate
(1.53 g, 14.9 mmol) in ethylene glycol dimethyl ether (35 mL), at
5.3.2. Method B
5.3.2.1. Steps e and c. Preparation of 4-(2,6-dimethylpyridin-4-
yl)-2-methylbut-3-yn-2-ol
(17). 2,6-Dimethyl-4-hydroxypyri-
dine (1.74 g, 14.1 mmol) was treated with phosphorous oxychlo-

638
M. S. Malamas et al. / Bioorg. Med. Chem. 18 (2010) 630–639
a rate as to maintain gentle hydrogen evolution at a controllable
rate. The reaction was then refluxed for a further 4 h and cooled
to room temperature. The condenser was removed and replaced
with a distillation head, vacuum was applied and the solvent (ca.
40 mL) was removed. The resulting slurry was diluted with diethyl
ether (50 mL), cooled to 0 °C and quenched with water (50 mL).
After separating the layers, the organic layer was washed with
water (30 mL) and 1% NaOH (30 mL). Ice chips were added to the
combined aqueous layer, followed by concentrated hydrochloric
acid (10 mL). The mixture was extracted with ether (3 ꢀ 50 mL),
and the combined organic extracts were dried Na2SO4. The result-
ing product 7-methyloctane-2,4,6-trione was then dissolved in
concentrated NH4OH (45 mL) and heated at reflux for 3 h. The reac-
tion mixture was cooled and the solvents evaporated to afford 2-
an orange oil (3.86 g, 76% yield): 1H NMR (300 MHz, CDCl3) d
1.28 (t, J = 7.6 Hz, 6H, 2 ꢀ CH2CH3), 2.77 (q, J = 7.6 Hz, 4H,
2 ꢀ CH2CH3), 7.16 (s, 2H, Ar-H); MS m/z 214 (M+H)+.
5.4. Separation of enantiomers (5S)-2-amino-5-(2,6-
diethylpyridin-4-yl)-3-methyl-5-[3-(pyrimidin-5-yl)phenyl]-
3,5-dihydro-4H-imidazol-4-one and (5R)-2-amino-5-(2,6-
diethylpyridin-4-yl)-3-methyl-5-[3-(pyrimidin-5-yl)phenyl]-
3,5-dihydro-4H-imidazol-4-one
Racemic mixture of 2-amino-5-(2,6-diethylpyridin-4-yl)-3-
methyl-5-[3-(pyrimidin-5-yl)phenyl]-3,5-dihydro-4H-imidazol-4-
one was separated to its two enantiomers by HPLC on Chiralcel AD,
2 ꢀ 25 cm using mobile phase 15% EtOH in hexane and a flow rate
of 20 mL/min to afford (5S)-2-amino-5-(2,6-diethylpyridin-4-yl)-
3-methyl-5-[3-(pyrimidin-5-yl)phenyl]-3,5-dihydro-4H-imidazol-
methyl-6-(propan-2-yl)pyridin-4(1H)-one as
a
red viscous oil
1.27 (d,
(0.80 g, 53% yield): 1H NMR (300 MHz, CDCl3)
d
J = 6.9 Hz, 6H, CHMe2), 2.32 (s, 3H, Me), 2.84 (septet, J = 6.9 Hz,
1H, CHMe2), 6.21 (s, 1H, olefinic-H), 6.17 (s, 1H, olefinic-H); MS
m/z 152 (M+H)+.
4-one:
½
a 2D5
ꢅ
= ꢄ0.035 (c = 1% in CH3OH); 1H NMR (DMSO-d6
300 MHz) d 1.15 (t, J = 7.56 Hz, 6H, 2 ꢀ CH2CH3), 2.63 (q, J =
7.56 Hz, 4H, 2 ꢀ CH2CH3), 2.95 (s, 3H, Me), 6.72 (br s, 2H, NH2),
7.15 (s, 2H, Ar-H), 7.45 (t, J = 7.69 Hz, 1H, Ar-H), 7.51 (d, J = 8.05 Hz,
1H, Ar-H), 7.62 (dd, J = 7.68, 1.46 Hz), 1H, Ar-H), 7.75 (t, J = 1.59 Hz,
1H, Ar-H), 8.95 (s, 2H, Ar-H), 9.15 (s, 1H, Ar-H); MS m/z (M+H)+
401. Anal. Calcd for C23H24N6O ꢀ 0.5 H2O: C, 67.46; H, 6.15; N,
20.52. Found; C, 67.42; H, 5.96; N, 20.23; and (5R)-2-amino-5-(2,6-
diethylpyridin-4-yl)-3-methyl-5-[3-(pyrimidin-5-yl)phenyl]-3,5-
5.3.4.2. Step j. Preparation of 4-bromo-2-methyl-6-(propan-2-
yl)pyridine (23). A mixture of 2-methyl-6-(propan-2-yl)pyridin-
4(1H)-one (3.60 g, 23.8 mmol) and phosphorous oxybromide
(15.0 g, 52.3 mmol) in chloroform (10 mL) was heated at 130 °C
in a sealed tube for 45 min. The reaction mixture was then cooled,
quenched by the addition of ice, diluted with water (50 mL) and
neutralized with solid sodium carbonate. The mixture was ex-
tracted with methylene chloride (2 ꢀ 75 mL) and the combined or-
ganic extratcs were dried over Na2SO4. Evaporation of the solvents
and purification by flash chromatography (silica gel, 80:20–50:50
ethyl acetate/hexanes) afforded 4-bromo-2-methyl-6-(propan-2-
yl)pyridine as an orange oil (1.17 g, 23% yield): 1H NMR
(500 MHz, CDCl3) d 1.27 (d, J = 6.9 Hz, 6H, CHMe2); 2.50 (s, 3H,
Me), 2.99 (septet, J = 6.9 Hz, 1H, CHMe2), 7.14 (s, 2H, Ar-H); MS
m/z 214 (M+H)+.
dihydro-4H-imidazol-4-one: ½a D25
ꢅ
= +0.039 (c = 1% in CH3OH); 1H
NMR (DMSO-d6 300 MHz) d 1.15 (t, J = 7.56 Hz, 6H, 2 ꢀ CH2CH3),
2.63 (q, J = 7.56 Hz, 4H, 2 ꢀ CH2CH3), 2.95 (s, 3H, Me), 6.72 (br s,
2H, NH2), 7.15 (s, 2H, Ar-H), 7.45 (t, J = 7.69 Hz, 1H, Ar-H), 7.51 (d,
J = 8.05 Hz, 1H, Ar-H), 7.62 (dd, J = 7.68, 1.46 Hz), 1H, Ar-H), 7.75 (t,
J = 1.59 Hz, 1H, Ar-H), 8.95 (s, 2H, Ar-H), 9.15 (s, 1H, Ar-H); MS m/z
(M+H)+ 401. Anal. Calcd for C23H24N6O ꢀ 0.7 H2O: C, 66.87; H,
6.19; N, 20.3. Found; C, 66.89; H, 5.81; N, 20.0.
5.5. Biological methods
5.3.5. Method E
5.5.1. FRET-based peptide cleavage assays
5.3.5.1. Steps k–m. Preparation of 2,6-diethyl-4H-pyran-4-one
(25). The reaction of 1,3-acetonedicarboxylic acid (15.0 g,
103 mmol) with propionic anhydride (41 mL, 320 mmol) in the
presence of concentrated sulfuric acid (0.5 mL, 9 mmol) as de-
scribed in the literature22 afforded 2,6-diethyl-4H-pyran-4-one as
a pale yellow oil (6.03 g, 38% yield): 1H NMR (300 MHz, CDCl3) d
1.23 (t, J = 7.5 Hz, 6H, 2 ꢀ CH2CH3), 2.54 (q, J = 7.5 Hz, 2 ꢀ CH2CH3),
6.08 (s, 2H, 2 ꢀ olefinic-H); MS m/z 153 (M+H)+.
A homogenous, continuous fluorescence resonance energy trans-
fer(FRET)wasusedtoassesscompoundinhibitionforBACE1, BACE2,
cathepsin D, pepsin and renin activities.24 The BACE1 and BACE2
activities were based on the cleavage of peptide substrate Abz-SEV-
NLDAEFR-Dpa (Swedishsubstrate), whilepeptide substrateMOCAc-
GKPILFFRLK (Dnp)-D-R-NH2 was used for cathepsin D and Pepsin,
and peptide substrate RE(EDANS)-IHPFHLVIHTK(DABCYL)-R for
Renin. Kinetic rates were calculated and IC50 values were deter-
mined by fitting the % inhibition, as a function of compound concen-
tration, to the Hill equation (y = ((B ꢁ Kn) + (100 ꢁ xn))/(Kn + xn). (x is
the agonist concentration and y is response (binding). Bmax is the
maximal or asymptotic response as x increases without bound. K is
the EC50. The Hill equation has no obvious physical interpretation.
The role of the parameter n is to adjust the steepness of the curve).
5.3.5.2. Step n. 2,6-Diethylpyridin-4-ol (26). The reaction of 2,6-
diethyl-4H-pyran-4-one (3.61 g, 23.7 mmol) and concentrated
NH4OH (15 mL, approx. 210 mmol) following the literature proce-
dure gave 2,6-diethylpyridin-4-ol23 as a viscous red oil (4.0 g,
quantitative): 1H NMR (300 MHz, CDCl3) d 1.25 (t, J = 7.6 Hz, 6H,
2 ꢀ CH2CH3), 2.67 (q, J = 7.6 Hz, 4H, 2 ꢀ CH2CH3), 6.22 (s, 2H, Ar-
H); MS m/z 152 (M+H)+.
5.5.2. Cell-based Ab inhibition assay
CHO-K1 cells recombinantly expressing human wild-type APP
(CHO-WT) were grown to confluence and then treated with serum
free medium (Ultraculture) supplemented with test compound in
DMSO or DMSO alone (vehicle) at a final [DMSO] of 0.1% (v/v).
Conditioned medium was harvested at 24 h, and assayed using
streptavidin MSD plates, and an electrochemiluminescent immuno-
assay with biotinylated mouse monoclonal antibody 6E10 (Signet,
Dedham, MA) as capture, and rabbit anti-Ab40 or Ab42 antibodies
(Biosource, Camarillo, CA) as detection antibodies, with a secondary
of MSD ECL tagged Sheep anti Rabbit for electrochemiluminescent
amplification.
5.3.5.3. Step j. Preparation of 4-bromo-2,6-diethylpyridine
(27). A mixture of 2,6-diethylpyridin-4-ol (3.58 g, 23.7 mmol)
and phosphorous oxybromide (17.1 g, 59.6 mmol) was heated to
130 °C for 90 min, cooled to room temperature and the flask then
placed in an ice bath. Ice was added in small pieces, and vigorous
gas evolution was observed. The mixture was diluted with water
(100 mL) and carefully neutralized by adding solid sodium carbon-
ate in small portions. Once neutral, the aqueous layer was ex-
tracted with methylene chloride (2 ꢀ 100 mL), and the combined
organic extracts were dried over Na2SO4. Evaporation of the sol
vents and purification by flash chromatography (silica gel, 1:9
ethyl acetate/hexanes) afforded 4-bromo-2,6-diethylpyridine as
Data analysis of the MSD assay was performed by fitting the
percent inhibition, as a function of compound concentration, to a

M. S. Malamas et al. / Bioorg. Med. Chem. 18 (2010) 630–639
639
four-parameter logistic curve and the concentration at which the
cellular production of Ab40 or Ab42 was reduced by 50% (EC50
5.7. Molecular modeling. Docking calculations were performed
using the QXP software package29
)
was determined. Compound toxicity was assessed via mitochon-
drial function using an MTS readout (MTS kit from Promega); val-
ues are represented as LD50 or the dose of compound that resulted
in 50% of control signal (e.g., 50% of the highest MTS signal).
Once the X-ray ligands were minimized in the active site, con-
strained simulated annealing calculations were performed to re-
lieve any artifacts of the X-ray refinement perceived as
unfavorable interactions or strain by the modified AMBER force
field with QXP. The X-ray ligand was then redocked to confirm that
the lowest energy pose is in agreement with the X-ray structure.
Once the binding site model was generated, docking of analogues
was performed using the QXP Monte Carlo docking algorithm
mcdock in combination with combidock. Visualization of X-ray
structures and docking results was performed using the InsightII
software package (www.accelrys.com, Accelrys, Inc., San Diego,
CA). Conformational analysis was performed using Macromodel
(Macromodel 8.0; Schrodinger, LLC, Portland, OR), with the OPLS-
AA force fields, and a GBSA solvation model.
5.6. X-ray crystallography
5.6.1. Cloning/expression of BACE1
A human BACE1 secreting mammalian secreting cell line (CHO
cell line) was made by expressing a construct in which the prodo-
main plus ectodomain of human BACE1 (residues 22–454) was
fused to the honeybee melittin secretory leader sequence at the
50 end and the Fc region of IgG, separated from the BACE sequence
via an enterokinase cleavage site, was fused at the 30 end of BACE1.
The BACE1/Fc fusion protein was affinity purified by protein-A se-
pharose and removal of the Fc domain was achieved with entero-
kinase cleavage. Purified BACE1 protein was accomplished via
sequential size-exclusion chromatography. Escherichia coli-derived
expression material was used for co-crystallographic studies. A co-
don-derived BACE1 E. coli expression construct was fused to a car-
boxy-terminal 6X HIS tag. After scale up, inclusion bodies were
purified and protein refolded and dialyzed. The pro domain was
cleaved with furin and BACE1 was further purified by size exclu-
sion chromatography. The refolded E. coli-derived BACE1 showed
the same enzymatic activity as the CHO-derived material (data
not shown).
Supplementary data
X-ray crystallographic data (collection details, refinement sta-
tistics), 1HNMR and analytical data of the final compounds not
listed in the Experimental Section, and supplemental information
of biological assays are available. Supplementary data associated
with this article can be found, in the online version, at
doi:10.1016/j.bmc.2009.12.007.
References and notes
1. Dickson, D. W. J. Neuropathol. Exp. Neurol. 1997, 56, 321–339.
2. Selkoe, D. J. Science 1997, 275, 630–631.
3. Vassar, R.; Citron, M. Neuron 2000, 27, 416–422.
4. Selkoe, D. J. Physiol. Rev. 2001, 81, 741–766.
5. Selkoe, D. J. Ann. Intern. Med. 2004, 140, 627–638.
6. Arendt, T. Neuroscience 2001, 102, 723–765.
7. Hardy, J.; Selkoe, D. J. Science 2002, 297, 353–356.
8. Selkoe, D. Ann. N.Y. Acad. Sci. 2000, 924, 17–25.
9. Couglan, C.; Breen, K. C. Pharm. Ther. 2000, 86, 111–144.
10. Racchi, M.; Govone, S. Trends Pharmacol. Sci. 1999, 20, 418–423.
11. Checler, F. J. Neurochem. 1995, 65, 1431–1444.
12. Malamas, M. S.; Erdei, J.; Gunawan, I.; Barnes, K.; Johnson, M.; Hui, Y.; Turner,
J.; Hu, Y.; Erik Wagner, E.; Fan, K.; Olland, A.; Bard, J.; Robichaud, A. J. Med.
Chem 2009, 52, 6314–6323.
5.6.2. Crystallization and X-ray diffraction analysis of BACE1
Crystals were grown by hanging drop vapor diffusion at 18 °C in
drops containing 1.0
20 mM Tris–HCl, pH 7.5, 250 mM NaCl, and 240
interested diluted from a 100 mM stock in DMSO) mixed with
1.0 L well solution (6% PEG 3350, 0.1 M sodium acetate pH 5.4)
lL protein stock solution (200
lM protein,
lM compound of
l
end equilibrated against 0.5 mL well solution. Rod shaped crystals
grew in several days.
5.6.3. Data collection
13. Malamas, M. S.; Barnes, K.; Johnson, M.;Hui, Y.; Zhou, P.; Turner, J.; Hu, Y.;Erik
Wagner, E.; Fan, K.; Olland, A.; Bard, J.; Jacobsen, S.; Ronald L. Magolda, R.;
Pangalos, M.; Robichaud, A. J. Med. Chem., (online 7 December 2009).
14. Dominguez, D.; Tournoy, J.; Hartmann, D.; Huth, T.; Cryns, K.; Deforce, S.;
Serneels, L.; Camacho, I. E.; Marjaux, E.; Craessaerts, K.; Roebroek, A. J. M.;
Schwake, M.; D’Hooge, R.; Bach, P.; Kalinke, U.; Moechars, D.; Alzheimer, C.;
Reiss, K.; Paul Saftig, P.; De Strooper, B. J. Biol. Chem. 2005, 280, 30797–30806.
15. Kagedal, K.; Johansson, U.; Llinger, K. O. FASEB 2001, 1592–1594.
16. Sonogashira, K.; Tohda, Y.; Hagihara, N. Tetrahedron Lett. 1975, 50, 4467–4470.
17. Miyaura, N.; Suzuki, A. Chem. Rev. 1995, 95, 2457–2483.
18. Comins, D. L.; Mantlo, N. B. J. Org. Chem. 1985, 50, 4410–4411.
19. Ciana, L. D.; Haim, A. J. Heterocycl. Chem. 1984, 21, 607–608.
20. Dixon, J. WO 035507, 2005.
Crystals were drawn through a solution of 25% glycerol and 75%
well solution, and cooled rapidly in liquid nitrogen. Diffraction data
were recorded at the ALS beamline 5.0.1 on a q-210 ccd camera.
Intensities were integrated and scaled using the programs DENZO
25
and SCALEPACK
.
5.6.4. Phasing, model building and refinement
Structures were determined by molecular replacement using
AmorE26 and the apo structure of BACE1 (PDB ID 1W50) as the
search model. The final structures were obtained after several iter-
ative cycles of refinement using Phenix27 and model improvement
in Coot.28
21. Beak, P.; Covington, J.; Smith, S.; White, M.; Zeigler, J. J. Org. Chem. 1980, 45,
1354–1362.
22. Yates, P.; Hand, E. S.; Singh, P.; Roy, S. K.; Still, W. J. J. Org. Chem. 1969, 34, 4046.
23. Deshapande, S. J. Indian Chem. Soc. 1932, 303.
24. Wu, P.; Brand, L. Anal. Biochem. 1994, 218, 1–13.
25. Otwinowski, Z.; Minor, W. Methods Enzymol. 1997, 276, 307–326. Macromol.
Crystallogr., Part A.
26. CCP4 Acta Crystallogr., Sect. D Biol. Crystallogr. 1994, 50, 760–763.
27. Adams, P. D.; Grosse-Kunstleve, R. W.; Hung, L.-W.; Ioerger, T. R.; McCoy, A. J.;
Moriarty, N. W.; Read, R. J.; Sacchettini, J. C.; Sauter, N. K.; Terwillinger, T. C.
Acta Crystallogr., Sect. D 2002, 58, 1948–1954.
28. Emsley, P.; Cowtan, K. Acta Crystallogr., Sect. D Biol. Crystallogr. 2004, 60, 2126–
2132.
29. McMartin, C.; Bohacek, R. S. J. Comput. Aided Mol. Des. 1997, 11, 333–344.
5.6.5. Data deposition
The atomic coordinates of the BACE1 crystal structure for com-
pounds 3b (3INF)15, 30 (3IN3), 38 (3IN4) have been deposited in
the Protein Data Bank, Research Collaboratory for Structural Bioin-
formatics, Rutgers University, New Brunswick, New Jersey. The
atomic coordinates of the apo-BACE2 structure (2EWY) were ob-
tained from the Protein Data Bank.