Application of fragment-based NMR screening, X-ray crystallography, structure-based design, and focused chemical library design to identify novel μM leads for the development of nM BACE-1 (β-site APP cleaving enzyme 1) inhibitors

Article abstract of DOI:10.1021/jm901472u

Fragment-based NMR screening, X-ray crystallography, structure-based design, and focused chemical library design were used to identify novel inhibitors for BACE-1. A rapid optimization of an initial NMR hit was achieved by a combination of NMR and a funct

Full text of DOI:10.1021/jm901472u

942 J. Med. Chem. 2010, 53, 942–950
DOI: 10.1021/jm901472u
Application of Fragment-Based NMR Screening, X-ray Crystallography, Structure-Based Design, and
Focused Chemical Library Design to Identify Novel μM Leads for the Development of nM BACE-1
( β-Site APP Cleaving Enzyme 1) Inhibitors†
Yu-Sen Wang,*,‡ Corey Strickland,§ Johannes H. Voigt,§ Matthew E. Kennedy,§ Brian M. Beyer,§ Mary M. Senior,§
Elizabeth M. Smith,§ Terry L. Nechuta,§ Vincent S. Madison,§ Michael Czarniecki,§ Brian A. McKittrick,§
Andrew W. Stamford,§ Eric M. Parker,§ John C. Hunter,§ William J. Greenlee,§ and Daniel F. Wyss‡
‡Schering-Plough Research Institute, 320 Bent Street, Cambridge, Massachusetts 02141 and §Schering-Plough Research Institute, 2015 Galloping
Hill Road, Kenilworth, New Jersey 07033
Received October 5, 2009
Fragment-based NMR screening, X-ray crystallography, structure-based design, and focused chemical
library design were used to identify novel inhibitors for BACE-1. A rapid optimization of an initial
NMR hit was achieved by a combination of NMR and a functional assay, resulting in the identification
of an isothiourea hit with a Kd of 15 μM for BACE-1. NMR data and the crystal structure revealed that
this hit makes H-bond interactions with the two catalytic aspartates, occupies the nonprime side
region of the active site of BACE-1, and extends toward the S3 subpocket (S3sp). A focused NMR-
based search for heterocyclic isothiourea isosteres resulted in several distinct classes of BACE-1 active
site directed compounds with improved chemical stability and physicochemical properties. The strategy
for optimization of the 2-aminopyridine lead series to potent inhibitors of BACE-1 was demonstrated.
The structure-based design of a cyclic acylguanidine lead series and its optimization into nanomolar
BACE-1 inhibitors are the subject of the companion paper (J. Med. Chem. 2010, 53, DOI:10.1021/
jm901408p).
Introduction
periphery.8 BACE-1 is primarily responsible for the cleavage
of APP and the subsequent generation of Aβ peptides (e.g.,
BACE-1 knockout mice are normal, do not produce Aβ
peptides and have few overt phenotypes9-12), whereas the
function of BACE-2 is unknown at present. Therefore,
BACE-1 is believed to be the most clinically important iso-
form to the development of AD.13 The required selectivity for
a clinical compound for BACE-1 versus BACE-2 has yet to be
established, but the selectivity for BACE versus other aspartyl
proteases such as cathepsin D and E may be more impor-
tant.14 The essential role of BACE-1 in Aβ formation, the
observation that BACE-1 knockout mice are phenotypically
normal, and theavailabilityof a crystal structure ofBACE-115
to support structure-based drug design together suggested
that inhibition of BACE-1 is a very promising approach for
the treatment of AD.16,17 This theoretical enthusiasm was
tempered, however, by the known practical difficulties in
designing low molecular weight, orally active, brain penetrant
aspartic protease inhibitors. The BACE-1 crystal structure
also revealed that the BACE-1 active site would present
additional challenges to inhibitor design because of its relative
hydrophilicity and shallow, elongated topography.
Alzheimer’s disease (ADa) is a dementia-inducing illness
which has taken on significant importance as a result of the
aging populations worldwide.1,2 Currently marketed treat-
ments have limited efficacy, leaving a large unmet medical
need in terms of true disease-modifying therapies.3 Neuronal
cell death is symptomatic of AD and is a key contributor to
cognitive failure in AD patients and is a root cause of AD.3
Genetic and pathological evidence strongly supports the
amyloid cascade hypothesis of AD, which states that the
production of β-amyloid (Aβ) peptides in the brain, neuronal
cell death, and the subsequent symptoms associated with AD
are directly linked.4
Aβ is produced by the sequential catalytic cleavage of the
amyloid precursor protein (APP) by two proteases, β-secre-
tase (BACE, β-site APP cleaving enzyme) and γ-secretase,
respectively.5,6 BACE is a transmembrane aspartyl protease
that exists in two isoforms7 with BACE-1 located in the
central nervous system and BACE-2 expressed mainly in the
†Coordinates have been deposited in the RCSB Protein Data bank for
BACE-1 with compounds 3, 4, and 5 with accession codes 3KMX,
3KMY, and 3KNO, respectively.
*To whom correspondence should be addressed. Phone: (617)-499-
3645. Fax: (617) 499-3813. E-mail: allen.yu-sen.wang@spcorp.com.
a Abbreviations: Aβ, β-amyloid peptide; AD, Alzheimer’s disease;
APP, β-amyloid precursor protein; BACE-1, β-site APP cleaving
enzyme 1 (or β-secretase); HSQC, heteronuclear single-quantum corre-
lation; NMR, nuclear magnetic resonance; HTRF, homogeneous time-
resolved fluorescence; PDB, protein data bank.
Considerable efforts have been made to develop
BACE-1 inhibitors as a therapeutic strategy for the treatment
of AD.18-21 Such efforts have led to the development of
peptidomimetic BACE-1 inhibitors with nM affinity based
on, for example, the statine,22,23 hydroxyethylene,24,25 or
hydroxyethylamine26-28 scaffolds. However, a potential pro-
blem with peptidomimetics is that they tend to be relatively
r
pubs.acs.org/jmc
Published on Web 12/31/2009
2009 American Chemical Society

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 3 943
large and possess multiple hydrogen bond donors and accep-
tors, which may make them less attractive leads for a
therapeutic target that requires oral bioavailability and pene-
tration across the blood-brain barrier.29,30 Therefore, some
of the more recently discovered nonpeptidic small molecule
BACE inhibitors are of great interest.21 Heterocycle-based
BACE inhibitors are among the most intriguing inhibitors
which are likely to have appropriate pharmacokinetic proper-
ties.31-36
of small molecule compounds. Clusters which caused chemical
shift changes of active site residues of BACE-1 were deconvo-
luted to identify the active site binding compound(s) within
each active cluster. Using this approach, multiple novel hits
were identified which bind to the substrate-binding sites of
BACE-1 in the μM-mM range.
HSQC-based NMR screening is a sensitive and reliable
method for the detection of ligand binding over a virtually
unlimited affinity range. For weakly bound ligands (Kd >
∼10 μM), where the rate constant for the dissociation of the
complex (koff) is fast on the chemical shift time scale,
quantitative dissociation constants can be obtained by mon-
itoring the chemical shift changes of the target as a function
of ligand concentrations.54-56 If a small molecule binds to a
macromolecular target at a particular site, it causes chemical
shift changes of target resonances which in many cases can be
analyzed to locate the binding site of the small molecule
ligand. HSQC-based methods are not only very robust and
reliable to detect ligand binding but may also provide
invaluable structural information regarding the binding site
location and orientation of a small molecule ligand with
respect to the protein surface.57-59 Therefore, small mole-
cules that bind to the active site(s) of the protein (and nearby
sites) can be selected over those that bind to other regions of
the protein which are distant from the active site(s). Knowl-
edge of the 3D structure of the target protein (e.g., X-ray
crystal structure, NMR solution structure, or homology
model) and sequence-specific assignments of at least the
active site residues are, however, required to obtain such
structural information. At the start of the NMR screening
campaign, we did not have NMR resonance assignments of
active site residues, which, therefore, were mapped using a
known peptidomimetic BACE-1 inhibitor 1 (OM99-2)15 to
identify active-site binders. In the course of the NMR
screening campaign, we obtained almost complete backbone
NMR resonance assignments of the catalytic domain of
human BACE-1,60 which then allowed a more detailed
analysis of ligand binding.
In an effort to discover novel small molecule inhibitors for
this challenging drug target, we have applied fragment-based
NMR screening37-39 to identify small molecule hits which
bind to the active site of the protein and were used as starting
points for the development of validated lead inhibitors of
BACE-1. Fragment-based lead discovery is an emerging field
in which much lower molecular weight compounds are
screened relative to those in high throughput screening cam-
paigns.40-47 In theory, fragment-based methods offer the
possibility of identifying novel leads with improved affinity,
selectivity, and pharmaceutical properties, and the rationale
behind these fragment-based strategies makes intuitive sense.
However, fragment-based hits are typically weak inhibitors/
binders (IC50/Kd ∼ μM-mM range) and therefore need to be
screened at higher concentrations using very sensitive
detection techniques. Although fragment hits are simpler,
less functionalized compounds48 with correspondingly lower
potencies, they typically possess a high “ligand efficiency”
(LE)49-51 and therefore are highly suitable for optimization
into clinical candidates with good drug-like properties. Never-
theless, optimization of weak-binding fragments into high-
affinity binders can be challenging and fragment-based lead
discovery can be difficult in practice. Both the discovery
and advancement of fragment hits are areas of intense re-
search. Although there is still much work to be done to
fully exploit the potential of this approach, the increasing
number of successful applications that have appeared in the
literature,40-47 including the first examples of clinical drug
candidates44,47 originating from this approach, strongly sug-
gest its viability.
NMR-based fragment screening of the fragment library
by 15N-HSQC NMR resulted in the identification of nine
chemically distinct classes of BACE-1 active site hits. These
active site hits were identified by the chemical shift changes of
the active site residues of BACE-1. Representative structures
of the nine classes of BACE-1 NMR hits are shown in
Figure 1. The characterization and optimization of the
isothiourea hit 2 will be described in detail in this paper.
Some limited optimization of the other hit classes was
conducted by NMR screening of analogues. The number
of analogues tested and the dissociation constants (Kd) that
were measured for some of the hits by NMR titration
experiments are indicated in Figure 1. It is worth mentioning
that for the aminodiazobicycloheptene, the endo phenyl
stereoisomer induced large chemical shift changes of
BACE-1, whereas the exo phenyl stereoisomer showed only
small chemical shift changes. For example, a chemical shift
change of 0.15 ppm of Gly34 was induced by the endo phenyl
compound at 1 mM concentration. A much smaller chemical
shift change of 0.03 ppm was observed by the exo phenyl
compound under the same conditions. This observation
indicated that the stereochemistry of the compounds in that
series is important for binding to BACE-1.
Here we report the identification of novel small molecule
inhibitors of BACE-1 using fragment-based NMR screening
ofa highly customized fragmentlibrary. A jointapproach that
involved NMR, bioassay, and X-ray crystal structure deter-
mination for hit characterization and focused screens resulted
in the identification of more potent isothiourea BACE-1
inhibitor leads. Directed NMR screening of isothiourea iso-
steres to identify 2-aminopyridines as active site-directed
BACE-1 ligands and the structure-assisted synthesis of ex-
tended 2-aminopyridines to demonstrate a strategy for further
optimization of the 2-aminopyridine lead series to more
potent small molecule inhibitors of BACE-1 is also described.
Results and Discussion
Identification of Initial NMR Hits. Two-dimensional 15N-
HSQC NMR was used to screen a custom-built fragment
library of approximately 10000 compounds against 15N-
labeled BACE-1 catalytic domain. About 50% of the com-
pounds in this screening library followed the “Rule-of-
Three” concept48 or similar rules52,53 consisting of low
MW compounds that are highly pure and water-soluble, as
well as synthetically tractable, to allow a rapid structural
optimization of fragment hits. Hits were identified by com-
paring two NMR spectra obtained with and without mixtures
Characterization of the Isothiourea NMR Hit 2. One of the
most interesting hits from the initial NMR screen contained
an isothiourea (compound 2 in Figure 2). This small

944 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 3
Wang et al.
Figure 1. Fragment-based NMR screening hits for BACE-1. Nine classes of BACE-1 active site directed NMR hits were identified by
screening 10000 compounds from a customized NMR fragment library by 15N-HSQC NMR.
analogues were selected from the corporate compound re-
pository and submitted for a focused screen in a BACE-1
HTRF assay. Fifteen compounds showed significant activity
at 50 μg/mL in the HTRF functional assay and were further
characterized by NMR. Among those, compound 3
(Figure 2) dose-dependently inhibited BACE-1 with an
IC50 of about 200 μM in the BACE-1 HTRF assay
and was identified as the most potent analogue in this
series. A Kd value of 15 μM was determined for 3 from
an NMR-based titration experiment. The chemical shift
changes of Asp228, Gly34, and Ala39 as a function of
compound concentrations were used for the Kd deter-
mination. Thus, this combination approach rapidly
achieved a 36-fold increase in binding affinity over the
initial NMR hit without commitment of synthetic chem-
istry resources.
NMR was also used to obtain initial SAR regarding the
type and length of linker connecting the amidine and the
phenyl groups in 3. Seventeen additional analogues were
evaluated which contained different X atoms (X=S, C, O)
and linker lengths (n=1, 2). Careful inspection of the mag-
nitudes of chemical shift changes induced by binding of these
compounds to BACE-1 in 2D HSQC spectra of the enzyme
revealed that the isothiourea headgroup is preferred over the
isourea and amidine groups and a phenethyl extension is
preferred over shorter extensions (Figure 2).
Figure 2. Optimization of the initial isothiourea NMR hit 2 into
the improved isothiourea NMR hit 3 for BACE-1. Initial SAR of
the starting fragments was evaluated by NMR (see text for details).
fragment hit induced large chemical shift changes of selected
peaks of BACE-1 active site residues in the 2D HSQC NMR
spectrum of 15N-labeled BACE-1. This is consistent with the
observation that many of the same protein peaks were also
perturbed by the active site inhibitor 1 when added to the
protein sample. The S4 to S20 peptide substrate pockets of the
˚
BACE-1 active site are well-defined in the 1.9 A resolution
crystal structure of the complex between the aspartic pro-
tease and the peptidomimetic inhibitor 1, and the active site
of BACE-1 is more open and less hydrophobic than that of
other human aspartic proteases in this structure. The NMR
hit 2 caused chemical shift changes of the two active site
aspartates (Asp32 and Asp228) and chemical shifts changes
of Ile118 and Gly230 in the S1 region and Gly34 in the S10
region. Incremental addition of the NMR hit 2 to 15N-
labeled BACE-1 caused dose-dependent protein chemical
shift changes of active site residues, which is consistent with
weak, site-specific active-site directed binding of 2 to BACE-
1 in the micromolar range. An NMR-based titration yielded
a dissociation constant (Kd) of 550 μM for 2. The chemical
shift changes of Asp32 and Gly34 as a function of compound
concentrations were used for the Kd calculation.
Ligand efficiency, defined as the free binding energy
per non-hydrogen atom (kcal mol-1 /heavy-atom count),
3
is arguably a better parameter than potency alone in the
selection of a lead compound and in the lead optimization
process.49 Compounds with high ligand efficiency are pre-
dicted to have an improved probability of binding to the
target of interest. Although the NMR hits 2 and 3 have low
affinities, because of their low molecular weights they show
relatively high ligand efficiency values of 0.37 and 0.39,
respectively (Figure 2). These relative high ligand efficiencies
are a result of extensive interactions of the hits with the
protein active site residues as observed in the NMR binding
data and the X-ray crystal structures (see below). Therefore,
Identification of a More Potent Isothiourea BACE-1 In-
hibitor. To accelerate the discovery of higher affinity BACE-
1 inhibitors, a tandem approach using NMR and a
BACE-1 HTRF assay61 was employed. Over 200 isothiourea

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 3 945
Figure 4. Superposition of the X-ray crystal structures of 3 (white)
and 4 (yellow) in the BACE-1 active site (A). BACE-1 flap (light
blue) and Asp32, Asp228 (red) are highlighted. The structure of
BACE-1 is shown as a solvent-accessible surface in gray. The
hydrogen bond networks of 3 and 4 with Asp32 and Asp228 of
BACE-1 are shown (B).
Figure 3. Selected regions of 2D 15N-HSQC spectra measured with
a sample of 100 μM uniformly 15N-labeled BACE-1 alone (black), in
the presence of 27.5 (cyan), 55 (magenta), 110 (blue), 220 (green),
and 440 (red) μM of 3, suggesting its binding to the two active site
aspartates while extending toward S3 and leaving the flap un-
changed.
these NMR hits for BACE-1 were considered excellent starting
points for further chemical optimization to potent leads.
Structural Characterization of the Isothiourea Hit 3. The
more potent isothiourea NMR hit 3 induced significant
chemical shift changes in the 2D 15N-HSQC spectrum of
BACE-1 as shown in Figure 3. It induced large, dose-
dependent chemical shift changes of the two active side
aspartates (Asp32 and Asp228) (Figure 3A) and for the
residues in the S3sp pocket such as Gly13 (Figure 3B). In
addition, it caused significant chemical shift changes of
Gly34 in the S10 subsite, Gly230 in the S1/S2 subsites, and
Ile118 and Leu119 in the S1 subsite. In contrast, it induced
only minor chemical shift changes of residues within the flap,
as can be seen for example in Figure 3B for the flap residue
Gly74. The flap of BACE-1 is known to be flexible and plays
an important role in the binding of inhibitors.15 The NMR
peak of Gly74 falls into an isolated, well resolved region of
the 2D 15N-HSQC spectrum of BACE-1 and can be used as a
marker to reveal whether or not ligand binding induces
conformation changes in the flap. Overall, the chemical shift
perturbation data indicated that 3 interacts with the two
active site aspartates and binds to the nonprime side region
of the active site of BACE-1, likely extending through the S1
pocket toward S3sp.
Figure 5. Structure-guided design of the heterocyclic isothiourea
isosteres lead to the discovery of the 2-aminopyridine lead series and
the iminohydantoin lead series based on the hydrogen bond struc-
ture formed between the isothiourea moiety of the NMR hit 3 and
the two active site aspartates of BACE-1.
hit-to-lead optimization approach was pursued. Analysis of
the X-ray crystal structure of 3 suggested that optimization
of binding in the S1 and S3sp pockets was a logical approach
to improve affinity. However it was considered that hydro-
lytic instability of the isothiourea moiety of 3 may render this
functional group unsuitable for drug development.62-64 It
was therefore desirable to search for isothiourea isosteres
with improved chemical stability. Hence, novel heterocyclic
cores that preserved the amidine-like active site binding
motif were conceptualized as illustrated in Figure 5. A
focused NMR-based search for heterocyclic isothiourea
isosteres resulted in the identification of two distinct classes
of compounds. Directed NMR screening led to the discovery
of a 2-aminopyridine lead series. The characterization and
optimization of the 2-aminopyridines will be further illu-
strated in this paper. The structure-based design of an
iminohydantoin lead series, starting from the NMR iso-
thiourea hit 3, and its optimization into nanomolar BACE-
1 inhibitors are the subject of the companion paper.72
These observations from the NMR chemical shift pertur-
˚
bation data were largely confirmed by a 1.8 A resolution
X-ray crystal structure of the complex between the isothiour-
ea NMR hit 3 and BACE-1, revealing in detail the binding of
the isothiourea moiety to Asp32 and Asp228 at the active
site. As shown in Figure 4, an extensive hydrogen bonding
network was observed between the NH groups of the iso-
thiourea moiety and the two catalytic aspartates (Asp32 and
Asp228). The crystal structure further revealed that the
phenyl group occupies the S1 pocket and the butyl group
extends toward the S3sp region. The discovery of aspartyl
protease ligands possessing an amidine-type group partici-
pating in an H-bond donor-acceptor array with the cataly-
tic Asp residues of the protease, as depicted for 3 in Figure 4,
was unprecedented at the time of its discovery.
Characterization of a 2-Aminopyridine BACE-1 Ligand.
NMR played an important role in the exploration of the
structure-activity relationship in this lead series because the
BACE-1 potency of these 2-aminopyridines were initially too
Focused NMR-Based Search for Heterocyclic Isosteres.
With this X-ray crystal structure in hand, a structure-based

946 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 3
Wang et al.
Figure 6. Schematic drawing illustrating the strategies pursued in
the optimization of 2-aminopyridine 4 into potent inhibitors of
BACE-1.
Figure 7. BACE-1 cocrystal structures of
3 (yellow) and 5
weak to be reliably detected in the biological assay. Directed
NMR screening of 32 2-aminopyridines rapidly identified
compound 4, which binds to the active site of BACE-1. The
chemical structure of 4 is shown in Figure 6. Compound 4
induced chemical shift changes of the two active site aspar-
tates (Asp32 and Asp228), Gly34 in the S10 subsite, and
Gly230 and Thr231 in the S1/S2 substrate binding pockets. It
also perturbed residues in the flap region, such as Gly74.
However, chemical shifts of residues in the S3sp pocket, for
instance Gly13, were not perturbed. The 2-aminopyridine 4
bound to BACE-1 with a Kd value of 32 μM (LE = 0.38)
as derived from NMR-based titration experiments. Taken
together, these NMR data strongly suggested that the
2-aminopyridine 4 binds with high ligand efficiency to the
active site of BACE-1 in a binding mode very similar to that
of the isothiourea NMR hit 3. However, 4 did not show
significant inhibition in the BACE-1 HTRF assay up to a
concentration of 1 mM. Nevertheless, an X-ray crystal
structure of 4 complexed to BACE-1 could be determined
(magenta). The substitution at 30 position of 5 extends to the S3
binding pocket of BACE-1.
(Figure 6). Initial efforts were thus focused on exploiting
binding in the hydrophobic subsites (S1, S3sp, and S20) of
BACE-1 and on engaging additional hydrogen bonding
interactions with Gly34 (S10), Gly230 (S1/S2), and Thr232
(S3sp/S4). Focused small compound libraries were designed
around this concept and produced for experimental testing.
Analysis of ligand-BACE-1 binding by NMR played an
important role in this effort because many analogues bound
weakly to BACE-1, and their potencies could not be detected
accurately in the HTRF assay. NMR data were used to
derive accurate Kd values for weak inhibitors in the μM-mM
range and infer their binding site locations. X-ray crystal
structures were determined for the most promising hits to
assist SAR development for chemical optimization. Struc-
ture-based design and chemical synthesis resulted in a highly
focused and rapid optimization of the initial hits into leads
with low-μM potency. As an example of a proof-of-concept,
compound 5, as shown in Figure 7, was designed to interact
with S1 and S3sp pocket and synthesized with extension at
the 30 position of the phenyl ring. It showed large chemical
shift perturbations of active site residues including residues
in the S3sp pocket such as Gly13 and its Kd is about 100 uM
by NMR. It inhibited BACE-1 with an IC50 value of 250 μM
in the BACE-1 HTRF assay. NMR titration is a reliable
method to measure Kd’s of weakly bound compounds
(∼μM-mM range). The IC50 values of compounds 3 and 5
revealed that these compounds also showed weak dose-
dependent activity in a functional assay, as would be ex-
pected from their binding mode to BACE-1. The crystal
structure of 5 in complex with BACE-1 confirmed that the
compound made favorable interactions with the non-
prime (S1, S2, and S3sp) subsites as shown in Figure 7. The
2-aminopyridine moiety makes the same hydrogen bond
interaction with the two catalytic aspartic acids as the initial
hit 4. The 3-phenethyl group forms hydrophobic interaction
in the S1 pocket. The furan moiety indeed reaches the S3sp
binding pocket. This structure-assisted synthesis of extended
2-aminopyridines demonstrated a viable strategy for further
optimization of the 2-aminopyridine lead series. Given the
high ligand efficiency and improved physiochemical proper-
ties of 4, the 2-aminopyridine series promised high potential
for further optimization to potent small molecule inhibitors
for BACE-1.
˚
at a resolution of 2.0 A, which revealed that 4 bound as
suggested from the NMR data, occupying the same position
as 3, showing the same array of hydrogen bonds with the two
active site aspartates and placing the phenyl ring into the S1
pocket (Figure 4). However, the phenyl ring of 4 is turned
away from the S3sp pocket. The X-ray crystal structure is
consistent with the NMR chemical shift perturbation data.
The 2-aminopyridine core embodies a novel BACE-1 cata-
lytic aspartate binding motif with improved chemical stabi-
lity relative to the isothiourea 3 and its chemical accessibility
is suited for rapid SAR development. Furthermore the
aminopyridine moiety of 4 is weakly basic with a predicted
pKa of ∼7.2,65-67 which is considered to be favorable for
brain penetration, and 4 was considered to be an ideal
starting point for hit-to-lead optimization of potent
BACE-1 inhibitors.
Strategy for Hit-to-Lead Optimization of 2-Aminopyri-
dines. The X-ray crystal structure of 4 suggested possible
modifications of the 2-aminopyridine core to increase both
the potency against BACE-1 and its selectivity against other
aspartic proteases by making favorable interactions with the
prime and nonprime binding sites adjacent to the catalytic
aspartate residues. The X-ray crystal structure of 4 in com-
plex with BACE-1 suggested that a 3-phenethyl extension off
the 2-aminopyridine core could place the phenyl into the S1
pocket and extension at the 30 and/or 40 positions in the
phenyl could extend toward the S3sp pocket, whereas exten-
sions from the 6-position of the 2-aminopyridine core could
probe the prime subsites of the aspartic acid protease.
Established chemistry methods could be employed to allow
rapid SAR development by introducing substitutions at C6,
C30, and C40 positions in the 3-phenethyl framework, en-
abling access to the hydrophobic subsites near the active site
Recently, several publications and patents have emerged
which also report promising nonpeptidic small molecule
BACE inhibitors which are of great interest to the field.21
For example, Murray, et al. reported an application of
fragment-based screening by X-ray crystallography to

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 3 947
BACE-1 in which three distinct chemotypes were identified
as binding to the catalytic aspartates in the millimolar range
either via an aminoheterocycle, a piperidine, or an aliphatic
hydroxyl group.32 Virtual screening around the 2-aminoqui-
noline motif identified an aminopyridine with increased
potency, which was used as a starting point for the develop-
ment of sub-μM 3-substituted 2-aminopyridine BACE-1
inhibitors which show an extensive H-bonding network with
the two catalytic aspartates and extend into the adjacent S20
and S1-S3sp binding pockets.33 Another recent report by
Cole, et al., describes the structure-based optimization of an
initial low μM acylguanidine inhibitor, which was discovered
in a high-throughput screen using a FRET assay into sub-
μM acylguanidine inhibitors which interact with the two
catalytic aspartates and extend into the adjacent S1-S3sp,
S20, and S10 binding pockets.68
low molecular weight cores for BACE-1 as a result of
this highly focused fragment-based approach, which present
uniqueopportunitiesfor the development of sub-μM BACE-1
inhibitors.
Experimental Section
Protein Expression and Purification. The expression of the
human BACE-1 protein (residues 14-454) was described in
detail previously.70 Briefly, the human BACE-1 protein (residue
1-454) was expressed in E. coli BL21 (DE3) strain as a precursor
protein containing the Pre, Pro, and protease domains. The
precursor protein was insoluble and refolded from inclusion
bodies that were treated with denaturant and detergent and
subsequently subjected to a Superdex-200 gel filtration column.
The fractions with BACE catalytic activity were collected and
combined. The protein solution was first applied to a Resource-
Q column and subjected to a Superdex-200 column for further
purification. The Pre and Pro peptides of BACE-1 were auto-
catalytically removed to yield residues 46-454 of the catalytic
domain with a molecular weight of 45341 Da (BACE-1). Uni-
form 15N-labeling of BACE-1 was performed in 15N labeled
minimal medium using 15NH4Cl as a sole nitrogen source. The
efficiency of refolding BACE is roughly 40%, and the amount of
starting material recovered from E. coli prior to refolding was
100 mg/L of cell paste (grown in minimal media with 15N
labeling). Therefore, each liter of E. coli that was grown yielded
40 mgs of refolded material for NMR analysis after refolding
and protein purification. Uniformly 15N labeled BACE-1 pro-
tein was exchanged into a phosphate buffer containing 75 mM
KiPO4, pH 7.5, 150 mM NaCl, and 5% D2O for NMR studies.
Conclusions
Fragment-based drug discovery provides exciting new op-
portunities to tackle “difficult” drug targets such as BACE-1.
For this CNS aspartic acid protease, structurally novel, highly
brain penetrant inhibitors needed to be developed while
restricting the molecular weight and the number of H-bond
donors and acceptors to impart oral absorption and blood-
-brain barrier permeability.29,30 Fragment-based drug dis-
covery approaches might also be suitable for new target
classes such asprotein-proteininteractions which historically
have not been considered “druggable”.69
In this work, we employed a highly integrated approach of
NMR-based fragment screening, X-ray crystallography,
structure-based design, and focused chemical library design
to identify novel μM leads for the development of nM BACE-
1 inhibitors. HSQC-based NMR screening of the fragment
library yielded multiple, active-site directed fragment hits in
the μM-mM range, which could be categorized into nine
distinct chemical classes. A rapid optimization of an initial
NMR hit 2 was achieved by a combination of NMR and a
functional assay, leading to the identification of an isothiour-
ea hit 3 with a dissociation constant (Kd) of 15 μM for BACE-
1, reflecting a 36-fold increase in binding affinity over the
starting fragment 2. NMR structural data and the crystal
structure of the isothiourea 3 complexed with BACE-1 re-
vealed that the small molecule makes hydrogen bond interac-
tions with the two catalytic aspartic acid residues, occupies
the nonprime side region of the active site of BACE-1, and
extends toward the S3sp binding pocket. Structure-based
design and directed NMR screening led to the discovery of
the 2-aminopyridine lead series as a novel isostere for the
isothiourea NMR hits with improved stability, physiochem-
ical properties, and chemical accessibility. The crystal
structure along with NMR data suggested that the 2-amino-
pyridine 4 binds to the BACE-1 active site. The X-ray
structures of compound 4 suggested distinct opportunities
for the optimization of potency and selectivity. Focused
small compound libraries were synthesized to exploit binding
in the hydrophobic S1, S3sp, and S20 subsites. For example,
structure-guided design and chemical synthesis resulted
in compound 5, which extends through S1 toward the
S3sp binding pocket as designed, thereby validating the
2-aminopyridine core as a novel aspartyl protease active site
binding motif for the development of small molecule inhibi-
tors for BACE-1.
The BACE-1 inhibitor
Bioscience Inc. (King of Prussia, PA).
1 was purchased from Bachem
BACE-1 HTRF Assay. The BACE-1 homogeneous time-
resolved fluorescence (HTRF) assays were conducted according
to the procedures described previously.61 The HTRF assay
utilizes the fluorescence resonance energy transfer (FRET) pair
europium and allophycocyanin for measuring BACE-1 activity
in a high-throughput manner.
NMR Spectroscopy. NMR experiments were performed at
25 °C on a Varian INOVA 600 MHz NMR spectrometer
equipped with a 5 mM triple resonance probe head. The fast
1H-15N-HSQC experiment was used to improve sensitivity and
to avoid water saturation.71 All HSQC experiments were ac-
quired with 96 scans per t1 increment, and a recycle delay of 0.7 s.
1
The H and 15N spectral widths were 8000 Hz with 1024 (t2)
complex points and 2200 Hz with 48 indirect complex points,
respectively. Linear prediction to 96 complex points was per-
formed in the indirect dimension before Fourier transformation.
NMR data were processed and analyzed with the Felix program
(Accelrys, CA) on a Silicon Graphics workstation.
Compounds in the custom-built fragment library were
screened in clusters of 10 compounds each against BACE-1.
The 2D HSQC spectra were acquired on a sample of 500 μL of
100 μM uniformly 15N-labeled BACE-1 in the absence or
presence of compounds. In the initial screening, the compound
concentration was 500 μM. Clusters which showed chemical
shift changes in the 2D HSQC spectra of 15N-BACE-1 were
deconvoluted to identify those compounds that caused the
chemical shift changes. Active site directed hits were identified
by chemical shift changes of residues within the substrate
binding pockets and the two active site aspartates of BACE-1.
Identified hits were further confirmed by repeating the NMR
experiments with freshly prepared compound from solid sam-
ples. Using this procedure, multiple weak hits with affinities in
the tens of μM to low mM range were discovered. These novel
chemically distinct classes of hits bind to the active site region of
BACE-1 based on the chemical shift changes of 15N-BACE-1.
NMR titration experiments were performed to determine the
dissociation constants (Kd) of ligands. The chemical shift
The work presented here and in the companion paper72
describe several attractive, ligand efficient, active site directed

948 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 3
Wang et al.
Figure 8. Synthesis of the 2-aminopyridine 5.
changes were monitored as a function of the compound con-
centration and the dissociation constants were obtained by
nonlinear fits of the NMR titration data.
(4) Hardy, J.; Selkoe, D. J. The Amyloid Hypothesis of Alzheimer’s
Disease: Progress and Problems on the Road to Therapeutics.
Science 2002, 297, 353–356.
(5) 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, 735–741.
(6) Wolfe, M. S.; Xia, W.; Ostaszewski, B. L.; Diehl, T. S.; Kimberly,
W. T.; Selkoe, D. J. Two transmembrane aspartates in presenilin-1
required for presenilin endoproteolysis and γ-secretase activity.
Nature 1999, 398, 513–517.
(7) Farzan, M.; Schnitzler, C. E.; Vasilieva, N.; Leung, D.; Choe,
H. BACE2, a beta-secretase homolog, cleaves at the beta site
and within the amyloid-beta region of the amyloid-beta
precursor protein. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 9712–
9717.
(8) Saunders, A. J.; Kim, T.-W; Tanzi, R. E. BACE Maps to Chromo-
some 11 and a BACE Homolog, BACE2, Reside in the Obligate
Down Syndrome Region of Chromosome 21. Science 1999, 286,
1255a.
(9) 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, 231–232.
X-ray Crystallography. X-ray crystal structure of the complex
of BACE-1 and compound 3 was obtained by cocrystallization.
BACE-1 (16 mg/mL, 150 mM NaCl, and 20 mM Hepes at pH
7.5) was complexed with 1 mM of compound 3 and mixed 1:1
with reservoir (15% PEG3350 and 0.2 M Na/K Tartrate). The
hanging drops were incubated at 4 °C for several weeks until
diffraction quality crystals grow. Crystals were cryoprotected
(15% PEG3350, 0.2 M Na/K tartrate and 15% PEG400) and
flash frozen prior to data collection. Data was collected with an
Raxis IV detector using a Raxis RU-H2R generator. The final
˚
data set was 99% complete to 1.8 A resolution with an R merge
of 5.7%. Crystal structures of the complexes for 4 and 5 were
obtained by soaking the cocrystals of compound 3. The data for
˚
compound 4 extend to 1.9 A resolution and was collected similar
˚
to compound 3. The data for compound 5 extend to 1.85 A
resolution and was collected at IMCA-CAT 17-BM (Advanced
Photon Source).
Chemistry. LC-Electrospray-Mass spectroscopy with a C-18
column using a gradient of 5% to 95% MeCN in water as the
mobile phase was used to determine the molecular mass and
retention time. Compounds described in this work were g95%
purity, with the exception of compound 5, which was produced
from a parallel synthesis experiment and the purity was deter-
mined to be 74% by LCMS with no single contaminant present
in an amount greater than 5%. X-ray structure determination of
compound 5 confirmed the structural assignment.
(10) Luo, Y.; Bolon, B.; Damore, M. A.; Fitzpatrick, D.; Liu, H.;
Zhang, J.; Yan, Q.; Vassar, R.; Citron, M. BACE1 (beta-secretase)
knockout mice do not acquire compensatory gene expression
changes or develop neural lesions over time. Neurobiol. Dis. 2003,
14, 81–88.
(11) Willem, M.; Garratt, A. N.; Novak, B.; Citron, M.; Kaufmann, S.;
Rittger, A.; DeStrooper, B.; Saftig, P.; Birchmeier, C.; Haass, C.
Control of Peripheral Nerve Myelination by the β-Secretase
BACE1. Science 2006, 314, 664–666.
(12) Wang, H.; Song, L.; Laird, F.; Wong, P. C.; Lee, H.-K. BACE1
knock-outs display deficits in activity-dependent potentiation of
synaptic transmission at mossy fiber to CA3 synapses in the
hippocampus. J. Neurosci. 2008, 28, 8677–8681.
The amine 5 was prepared starting from 2-amino-3-methyl-
pyridine by reaction with BOC anhydride and subsequent
alkylation with 3-cyanobenzylbromide. Reduction of the nitrile
to the aldehyde with DIBAL, followed by reductive alkylation
with 2-N-methyl methylaminofuran and removal of the BOC
group provided the amine 5 (Figure 8).
(13) Beher, D.; Graham, S. L. Protease inhibitors as potential disease-
modifying therapeutics for Alzheimer’s disease. Expert Opin. In-
vest. Drugs 2005, 14, 1385–1409.
(14) Bennett, B. D.; Babu-Khan, S.; Loeloff, R.; Louis, J. -C.;
Curran, E.; Citron, M.; Vassar, R. Expression analysis of BACE2
in brain and peripheral tissues. J. Biol. Chem. 2000, 275, 20647–
20651.
(15) Hong, L.; Koelsh, G.; Lin, X.; Wu, S.; Terzyan, S.; Ghosh, A. K.;
Zhang, X. C.; Tang, J. Structure of the Protease Domain of
Memapsin 2 (β-Secretase) Complexed with Inhibitor. Science
2000, 290, 150–153.
Acknowledgment. Use of the IMCA-CAT beamline 17-
BM at the Advanced Photon Source was supported by the
companies of the Industrial Macromolecular Crystallography
Association. Use of the Advanced Photon Source was sup-
ported by the U.S. Department of Energy, Office of Science,
Office of Basic Energy Sciences, under contract no. W-31-109-
Eng-38.
(16) Vassar, R. Beta-secretase (BACE) as a drug target for Alzheimer’s
disease. Adv. Drug Delivery Rev. 2002, 54, 1589–1602.
(17) Citron, M. Beta-secretase inhibition for the treatment of
Alzheimer’s disease-promise and challenge. Trends Pharmacol.
Sci. 2004, 25, 92–97.
(18) Cumming, J. N.; Iserloh, U.; Kennedy, M. E. Design and devel-
opment of BACE-1 inhibitors. Curr. Opin. Drug Discovery Dev.
2004, 7, 536–556.
References
(1) Cummings, J. L. Alzheimer’s Disease. New Engl. J. Med. 2004, 351,
56–67.
(2) Nussbaum, R. L.; Ellis, C. E. Alzheimer’s Disease and Parkinson’s
Disease. New Engl. J. Med. 2003, 348, 1356–1364.
(3) Citron, M. Strategies for disease modification in Alzheimer’s
disease. Nat. Rev. Neurosci. 2004, 5, 677–685.

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 3 949
(19) Durham, T. B.; Shepherd, T. A. Progress toward the discovery and
development of efficacious BACE inhibitors. Curr. Opin. Drug
Discovery Dev. 2006, 9, 776–791.
(20) Varghese, J. Human β-secretase (BACE) and BACE Inhibitors:
Progress Report. Curr. Top. Med. Chem. 2006, 6, 569–578.
(21) Stachel, S. J. Progress toward the development of a viable BACE-1
inhibitor. Drug Dev. Res. 2009, 70, 101–110.
Congreve, M.; Frederickson, M.; Folmer, R. H. A.; Geschwindner,
S.; Koether, G.; Kolmodin, K.; Krumrine, J.; Mauger, R. C.;
Murray, C. W.; Olsson, L.-L.; Patel, S.; Spear, N.; Tian, G.
Application of Fragment-Based Lead Generation to the Discovery
of Novel, Cyclic Amidine β-Secretase Inhibitors with Nanomolar
Potency, Cellular Activity, and High Ligand Efficiency. J. Med.
Chem. 2007, 50, 5912–5925.
(36) Stachel, S. J.; Coburn, C. A.; Rush, D.; Jones, K. L. G.; Zhu, H.;
Rajapakse, H.; Graham, S. L.; Simon, A.; Holloway, M. K.;
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 β-secretase inhibitor class: pH de-
pendence on binding activity part 1. Bioorg. Med. Chem. Lett. 2009,
19, 2977–2980.
(37) Shuker, S. B.; Hajduk, P. J.; Meadows, R. P.; Fesik, S. W.
Discovering High-Affinity Ligands for Proteins: SAR by NMR.
Science 1996, 274, 1531–1534.
(38) Wyss, D. F.; McCoy, M. A.; Senior, M. M. NMR-based ap-
proaches for lead discovery. Curr. Opin. Drug Discovery Dev.
2002, 5, 630–647.
(39) Pellecchia, M.; Bertini, I.; Cowburn, D.; Dalvit, C.; Giralt, E.;
Jahnke, W.; James, T. L.; Homans, S. W.; Kessler, H.; Luchinat,
C.; Meyer, B.; Oschkinat, H.; Peng, J.; Schwalbe, H.; Siegal, G.
Perspectives on NMR in drug discovery: a technique comes of age.
Nat. Rev. Drug Discovery 2008, 7, 738–745.
(40) Jahnke, W., Erlanson, D. A., Eds. Fragment-Based Approaches in
Drug Discovery; Methods and Principles in Medicinal Chemistry,
Vol. 34; Mannhold, R., Kubinyi, H., Folkers, G., Eds.; Wiley-VCH:
Weinheim, Germany, 2006.
(41) Rees, D. C.; Congreve, M.; Murray, C. W.; Carr, R. Fragment-
based lead discovery. Nat. Rev. Drug Discovery 2004, 3, 660–672.
(42) Erlanson, D. A.; McDowell, R. S.; O’Brien, T. Fragment-Based
Drug Discovery. J. Med. Chem. 2004, 47, 3463–3482.
(43) Erlanson, D. A. Curr. Opin. Biotechnol. 2006, 17, 643–652.
(44) Hajduk, P. J.; Greer, J. A decade of fragment-based drug design:
strategic advances and lessons learned. Nat. Rev. Drug Discovery
2007, 6, 211–219.
(45) Wyss, D. F.; Eaton, H. L. Frontiers in Drug Design & Discovery;
Caldwell, G.W., Atta-ur-Rahman, D'Andrea, M. R., Choudhary, M.I.,
Eds.; Bentham Science Publishers: Oak Park, IL, 2007; Vol. 3, in press .
(46) Congreve, M.; Chessari, G.; Tisi, D.; Woodhead, A. J. Recent
Developments in Fragment-Based Drug Discovery. J. Med. Chem.
2008, 51, 3661–3680.
(22) Hom, R. K.; Fang, L. Y.; Mamo, S.; Tung, J. S.; Guinn, A. C.;
Walker, D. E.; Davis, D. L.; Gailunas, A. F.; Thorsett, E. D.;
Sinha, S.; Knops, J. E.; Jewett, N. E.; Anderson, J. P.; John, V.
Design and Synthesis of Statine-Based Cell-Permeable Peptidomi-
metic Inhibitors of Human β-Secretase. J. Med. Chem. 2003, 46,
1799–1802.
€
ꢀ
€
(23) Back, M.; Nhylen, J.; Kvarnstrom, I.; Apelgren, S.; Borkakoti, N.;
€
Jansson, K.; Lindberg, J.; Nystrom, S.; Hallberg, A.; Rosenquist,
ꢀ
˚
A; Samuelsson, B. Design, synthesis and SAR of potent statine-
based BACE-1 inhibitors: exploration of P1 phenoxy and benzy-
loxy residues. Bioorg. Med. Chem. 2008, 16, 9471–9486.
(24) Hom, R. K.; Gailunas, A. F.; Mamo, S.; Fang, L. Y.; Tung, J. S.;
Walker, D. E.; Davis, D.; Thorsett, E. D.; Jewett, N. E.; Moon,
J. B.; John, V. Design and Synthesis of Hydroxyethylene-Based
Peptidomimetic Inhibitors of Human β-Secretase. J. Med. Chem.
2004, 47, 158–164.
(25) Hanessian, S.; Yun, H.; Hou, Y.; Yang, G.; Bayrakdarian, M.;
Therrien, E.; Moitessier, N.; Roggo, S.; Veenstra, S.; Tintelnot-
Blomley, M.; Rondeau, J.-M.; Ostermeier, C.; Strauss, A.; Ramage,
P.;Paganetti, P.;Neumann, U.;Betschart, C. Structure-Based Design,
Synthesis, and Memapsin 2 (BACE) Inhibitory Activity of Carbo-
cyclic and Heterocyclic Peptidomimetics. J. Med. Chem. 2005, 48,
5175–5190.
(26) 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
β-Secretase (BACE-1). J. Med. Chem. 2004, 47, 6447–6450.
(27) Barrow, J. C.; Rittle, K. E.; Ngo, P. L.; Selnick, H. G.; Graham,
S. L.; Pitzenberger, S. M.; McGaughey, G. B.; Colussi, D.; Lai,
M.-T.; Huang, Q.; Tugusheva, K.; Espeseth, A. S.; Simon, A. J.;
Munshi, S. K.; Vacca, J. P. Design and Synthesis of 2,3,5-Sub-
stituted Imidazolidin-4-one Inhibitors of BACE-1. ChemMed-
Chem 2007, 2, 995–999.
(47) Chessari, G; Woodhead, A. J. From fragment to clinical
candidate-a historical perspective. Drug Discovery Today 2009,
14, 668–675.
(48) Congreve, M.; Carr, R.; Murray, C.; Jhoti, H. A “Rule of Three”
for fragment-based lead discovery? Drug Discovery Today 2003, 8,
876–877.
(49) Hopkins, A. L.; Groom, C. R.; Alex, A. Ligand efficiency: a useful
metric for lead selection. Drug Discovery Today 2004, 9, 430–431.
(50) Abad-Zapatero, C.; Metz, J. T. Ligand efficiency indices as guide-
posts for drug discovery. Drug Discovery Today 2005, 10, 464–469.
(51) Bembenek, S. D.; Tounge, B. A.; Reynolds, C. H. Ligand efficiency
and fragment-based discovery. Drug Discovery Today 2009, 14,
278–283.
(52) Hann, M. M.; Oprea, T. I. Pursuing the leadlikeness concept in
pharmaceutical research. Curr. Opin. Chem. Biol. 2004, 8, 255–263.
(53) Oprea, T. I.; Blaney, J. M. Cheminformatics Approaches to
Fragment-Based Lead Discovery. Methods Principles Med. Chem.
2006, 34, 89–111.
(28) Charrier, N.; Clarke, B.; Cutler, L.; Demont, E.; Dingwall, C.;
Dunsdon, R.; East, P.; Hawkins, J.; Howes, C.; Hussain, I.; Jeffrey,
P.; Maile, G.; Matico, R.; Mosley, J.; Naylor, A.; O’Brien, A.;
Redshaw, S.; Rowland, P.; Soleil, V.; Smith, K. J.; Sweitzer, S.;
Theobald, P.; Vesey, D.; Walter, D. S.; Wayne, G. Second Gen-
eration of Hydroxyethylamine BACE-1 Inhibitors: Optimizing
Potency and Oral Bioavailability. J. Med. Chem. 2008, 51, 3313–
3317.
(29) Clark, D. E. Rapid calculation of polar molecular surface area and
its application to the prediction of transport phenomena. 2. Pre-
diction of blood-brain barrier penetration. J. Pharm. Sci. 1999, 88,
815–821.
(30) Pardridge, W. M. Blood-brain delivery. Drug Discovery Today
2007, 12, 54–61.
(31) Baxter, E. W.; Conway, K. A.; Kennis, L.; Bischoff, F.; Mercken,
M. H.; De Winter, H. L.; Reynolds, C. H.; Tounge, B. A.; Luo, C.;
Scott, M. K.; Huang, Y.; Braeken, M.; Pieters, S. M. A.; Berthelot,
D. J. C.; 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 (β-Site APP Cleaving Enzyme): Use of Structure Based
Design to Convert a Micromolar Hit into a Nanomolar Lead.
J. Med. Chem. 2007, 50, 4261–4264.
(32) Murray, C. W.; Callaghan, O.; Chessari, G.; Cleasby, A.; Congreve,
M.; Frederickson, M.; Hartshorn, M. J.; McMenamin, R.; Patel, S.;
Wallis, N. Application of Fragment Screening by X-ray Crystal-
lography to β-Secretase. J. Med. Chem. 2007, 50, 1116–1123.
(33) Congreve, M.; Aharony, D.; Albert, J.; Callaghan, O.; Campbell, J;
Carr, R. A. E.;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 β-Secretase. J. Med. Chem. 2007, 50,
1124–1132.
€
(54) Pellecchia, M.; Sem, D. S.; Wuthrich, K. NMR in drug discovery.
Nat. Rev. Drug Discovery 2002, 1, 211–219.
(55) Hajduk, P. J.; Meadows, R. P.; Fesik, S. W. NMR-based screening
in drug discovery. Q. Rev. Biophys. 1999, 32, 211–240.
(56) Zuiderweg, E. R. P. Mapping Protein-Protein Interactions in
Solution by NMR Spectroscopy. Biochemistry 2002, 41, 1–7.
(57) McCoy, M. A.; Wyss, D. F. Alignment of weakly interacting
molecules to protein surfaces using simulations of chemical shift
perturbations. J. Biomol. NMR 2000, 18, 189–198.
(58) Medek, A.; Hajduk, P. J.; Mack, J.; Fesik, S. W. The Use of
Differential Chemical Shifts for Determining the Binding Site
Location and Orientation of Protein-Bound Ligands. J. Am. Chem.
Soc. 2000, 122, 1241–1242.
(59) McCoy, M. A.; Wyss, D. F. Spatial Localization of Ligand Binding
Sites from Electron Current Density Surfaces Calculated from
NMR Chemical Shift Perturbations. J. Am. Chem. Soc. 2002,
124, 11758–11763.
(60) Liu, D.; Wang, Y.-S.; Gesell, J. J.; Wilson, E.; Beyer, B. M.; Wyss,
D. F. Letter to the Editor: Backbone Resonance Assignments of
the 45.3 kDa Catalytic Domain of Human BACE1. J. Biomol.
NMR 2004, 29, 425–426.
(34) Geschwindner, S.; Olsson, L.-L.; Albert, J. S.; Deinum, J.;
Edwards, P. D.; de Beer, T.; Folmer, R. H. A. Discovery of a
Novel Warhead against β-Secretase through Fragment-Based
Lead Generation. J. Med. Chem. 2007, 50, 5903–5911.
(35) Edwards, P. D.; Albert, J. S.; Sylvester, M.; Aharony, D.; Andisik,
D.; Callaghan, O.; Campbell, J. B.; Carr, R. A.; Chessari, G.;

950 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 3
Wang et al.
(61) Kennedy, M. E.; Wang, W.; Song, L.; Lee, J.; Zhang, L.; Wong, G.;
Wang, L.; Parker, E. Measuring human β-secretase (BACE1)
activity using homogeneous time-resolved fluorescence. Anal. Bio-
chem. 2003, 319, 49–55.
(62) Hueper, W. C.; Ichniowsky, C. T. Toxicopathologic studies on
S-methylisothiourea. Virchows Arch. Pathol. Anat. Physiol. Klin.
Med. 1944, 37, 253–263.
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 β-Secretase Inhibitors.
J. Med. Chem. 2006, 49, 6158–6161.
(69) Arkin, M. R.; Wells, J. A. Small-molecule inhibitors of protein-
protein interactions: progressing towards the dream. Nat. Rev.
Drug Discovery 2004, 3, 301–317.
(63) Dieke, S. H.; Allen, G. S.; Richter, C. P. The acute toxicity of
thioureas and related compounds to wild and domestic Norway
rats. J. Pharmacol. Exp. Ther. 1947, 90, 260–270.
(64) Peterlin-Maiic, L.; Cesar, J.; Zega, A. Metabolism-Directed Opti-
misation of Antithrombotics: The Prodrug Principle. Curr. Pharm.
Design 2006, 12, 73–91.
(65) Fastier, F. N.; McDowall, M. A. Analgesic activity of 4-methyl-2-
amino-pyridine and of some related compounds. Austr. J. Exp.
Biol. Med. Sci. 1958, 36, 491–498.
(70) Wang, Y.-S; Beyer, B. M.; Senior, M. M.; Wyss, D. F. Character-
ization of Autocatalytic Conversion of Precursor BACE1 by
Heteronuclear NMR Spectroscopy. Biochemistry 2005, 44,
16594–16601.
(71) Mori, S.; Abeygunawardana, C.; Johnson, M. O.; Berg, J.; van Zijl,
P. C. Improved sensitivity of HSQC spectra of exchanging protons
at short interscan delays using a new fast HSQC (FHSQC) detec-
tion scheme that avoids water saturation. J. Magn. Reson., Ser. B
1995, 108, 94–98.
(66) Sun, M. S.; Brewer, D. G. Comparative formation constants for
complexes of copper, nickel, and silver with some substituted
pyridines. Can. J. Chem. 1967, 45, 2729–2739.
(67) Brignell,P.J.;Johnson,C.D.;Katritzky,A.R.;Shakir,N.;Tarhan,H.O.;
Walker, G. Acidity functions and the protonation of weak bases. Part V.
Second protonation of diacid bases. J. Chem. Soc., B 1967, 1233–1235.
(68) 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.;
(72) Zhu, Z.; Sun, Z.-Y.; Ye, Y.; Voigt, J.; Strickland, C.; Smith,
E. M.; Cumming, J.; Wang, L.; Wong, J.; Wang, Y.-S.; Wyss,
D. F.; Chen, X.; Kuvelkar, R.; Kennedy, M. E.; Favreau, L.;
Parker, E.; McKittrick, B. A.; Stamford, A.; Czarniecki,
M.; Greenlee, W.; Hunter, J. C. Discovery of Cyclic Acyl-
guanidines as Highly Potent and Selective beta-site Amyloid
Cleaving Enzyme (BACE) Inhibitors: Part I;Inhibitor Design
and Validation. J. Med. Chem. 201053, DOI: 10.1021/
jm901408p.