Acyl guanidine inhibitors of β-secretase (bACE-1): Optimization of a micromolar hit to a nanomolar lead via iterative solid- and solution-phase library synthesis

Article abstract of DOI:10.1021/jm300931y

This report describes the discovery and optimization of a BACE-1 inhibitor series containing an unusual acyl guanidine chemotype that was originally synthesized as part of a 6041-membered solid-phase library. The synthesis of multiple follow-up solid- and

Full text of DOI:10.1021/jm300931y

Article
pubs.acs.org/jmc
Acyl Guanidine Inhibitors of β‑Secretase (BACE-1): Optimization of a
Micromolar Hit to a Nanomolar Lead via Iterative Solid- and
Solution-Phase Library Synthesis
Samuel W. Gerritz,*,† Weixu Zhai,† Shuhao Shi,† Shirong Zhu,† Jeremy H. Toyn,† Jere E. Meredith, Jr.,†
Lawrence G. Iben,† Catherine R. Burton,† Charles F. Albright,† Andrew C. Good,† Andrew J. Tebben,‡
Jodi K. Muckelbauer,‡ Daniel M. Camac,‡ William Metzler,‡ Lynda S. Cook,† Ramesh Padmanabha,†
Kimberley A. Lentz,† Michael J. Sofia,† Michael A. Poss,‡ John E. Macor,† and Lorin A. Thompson, III†
†Bristol-Myers Squibb Research, 5 Research Parkway, Wallingford, Connecticut 06492, United States
‡Bristol-Myers Squibb Research, P.O. Box 4000, Princeton, New Jersey 08543-4000, United States
S
* Supporting Information
ABSTRACT: This report describes the discovery and
optimization of a BACE-1 inhibitor series containing an
unusual acyl guanidine chemotype that was originally
synthesized as part of a 6041-membered solid-phase library.
The synthesis of multiple follow-up solid- and solution-phase
libraries facilitated the optimization of the original micromolar
hit into a single-digit nanomolar BACE-1 inhibitor in both
radioligand binding and cell-based functional assay formats. The
X-ray structure of representative inhibitors bound to BACE-1 revealed a number of key ligand:protein interactions, including a
hydrogen bond between the side chain amide of flap residue Gln73 and the acyl guanidine carbonyl group, and a cation−π
interaction between Arg235 and the isothiazole 4-methoxyphenyl substituent. Following subcutaneous administration in rats, an
acyl guanidine inhibitor with single-digit nanomolar activity in cells afforded good plasma exposures and a dose-dependent
reduction in plasma Aβ levels, but poor brain exposure was observed (likely due to Pgp-mediated efflux), and significant
reductions in brain Aβ levels were not obtained.
cleaving enzyme 1 (BACE-1, also called β-secretase), an
aspartyl protease that performs a selective proteolysis of APP to
afford the N-terminus of Aβ, and (2) γ-secretase, a protein
complex with presenilin at its catalytic core that cleaves APP to
generate the C-terminus. Inhibiting either of these enzymes
blocks the production of Aβ, prevents the formation of β-
amyloid plaques, and reverses amyloid-induced cognitive
deficits in laboratory animals.5 In this paper, we report the
discovery and optimization of a novel series of acyl guanidine-
based BACE-1 inhibitors.
INTRODUCTION
■
Alzheimer’s disease (AD) is a progressive neurodegenerative
disorder whose effects are initially manifested in the loss of
short-term memory, followed by increasingly worsening senile
dementia and leading ultimately to death.1 In 2009, it was
estimated that AD afflicted more than 35 million people
worldwide.2 Current treatments include the acetylcholinester-
ase inhibitors and the NMDA blocker memantine; however,
these agents treat disease symptoms and do not halt or delay
disease progression. AD is defined by two characteristic brain
lesions, intraneuronal neurofibrillary tangles comprising a
hyperphosphorylated form of the microtubule-binding protein
tau and extracellular amyloid plaques comprising a complex
mixture of β-amyloid (Aβ) peptides spanning 36 to 43 amino
acids. Data derived from studies of genetic mutations in early
onset AD suggest that overproduction of Aβ1−42 is a
significant contributor to AD.3 Substantial progress in under-
standing the underlying etiology of AD has been made over the
past decade, and current work suggests that soluble forms of
Aβ1−42 are central to the disease process.3 Therapeutic agents
that reduce Aβ production in the brain thus represent a logical
approach to discovering a disease-modifying therapy.4 Aβ is
produced via sequential proteolytic cleavage of amyloid
precursor protein (APP) by two enzymes:1 (1) β -APP-
RESULTS AND DISCUSSION
■
Owing to their peptidomimetic origins, early BACE-1
inhibitors6 contained multiple amide bonds and polar func-
tional groups, which served to limit their brain exposure. In an
effort to discover novel nonpeptidic BACE-1 inhibitors, we
performed a structure-biased fragment screening campaign in
which fragments that scored high in a BACE-1 binding model
were tested in an NMR direct binding assay and a high
concentration (30 μM) HTS assay. These screening efforts
Special Issue: Alzheimer's Disease
Received: July 2, 2012
Published: September 4, 2012
© 2012 American Chemical Society
9208
dx.doi.org/10.1021/jm300931y | J. Med. Chem. 2012, 55, 9208−9223

Journal of Medicinal Chemistry
Article
culminated in the discovery of acyl guanidine 1 (Figure 1).
When tested in our radioligand binding assay,7 1 afforded a
were submitted without further purification, but for the BACE-
1 hit-to-lead efforts described herein, all analogues of
compound 8 were purified via reverse-phase HPLC prior to
submission for biological testing.
Over 5000 acyl guanidine analogues were assayed in the
BACE-1 HTS, and every active sample contained both the 5-
methyl-3-phenyl isoxazole (R1) found in 1 and a small group of
benzylamine (R2) substructures. The high degree of structural
similarity among the HTS hits prompted the synthesis of a
matrix library in which the benzylamine and isoxazole10,11 side
chains were varied simultaneously. Table 1 summarizes the
Figure 1. BACE-1 HTS hit acyl guanidine 1.
Table 1. Acyl Guanidine Matrix Library Summary
BACE-1 Ki = 3900 nM, several orders of magnitude less potent
than that of reported peptide-based inhibitors. The ligand
efficiency8,9 (LE) of compound 1 was 0.25 kcal/mol·heavy
atom, utilizing the following equation: LE = −RT ln(Ki)/heavy
atom count, and this value was relatively low for a nascent
screening hit. However, the unusual structure and ease of
synthesis of 1 prompted further exploration of the chemotype
in search of improved BACE-1 activity. Compound 1 was
initially synthesized10 on solid phase as part of a 6041-
membered, ion channel-targeted library comprising two points
of diversity: (1) an amine reagent set (e.g., the 1-amino-
methylnaphthyl side chain of 1) and (2) a carboxylic acid
reagent set (e.g., the 5-methyl-3-phenylisoxazole moiety of 1).
Although 1 was originally submitted as a crude library product
with an HPLC purity of 70% (based on UV absorbance, λ =
254 nm), the resynthesis and purification of 1 successfully
recapitulated the observed BACE-1 activity.
The library pedigree of 1 provided a robust solid-phase
synthesis of acyl guanidines as shown in Scheme 1. The
synthesis commenced with treatment of commercially available
p-nitrophenylcarbonate-modified polystyrene Wang (PS-
Wang) resin11 (2) with S-methylisothiourea under basic
conditions to provide the corresponding solid-supported
carbamate 3. By utilizing carefully optimized conditions
(PyAOP, DIEA), the primary nitrogen of 3 was acylated with
carboxylic acid 4 to provide the polymer-bound S-methyl-
thiourea 5 and, in the process, activating the thiomethyl group
for subsequent displacement by amine 6 to afford the solid-
supported carbamate-protected acyl guanidine 7. Treatment of
7 with TFA provided acyl guanidine 8. In the case of the
aforementioned ion channel library, compound 8 analogues
a
Mean radioligand displacement activity (n ≥ 2) according to ref 7.
Each Ki value is within 2.2-fold of the mean.
Scheme 1
9209
dx.doi.org/10.1021/jm300931y | J. Med. Chem. 2012, 55, 9208−9223

Journal of Medicinal Chemistry
Article
BACE-1 activity for a 4 amine (R1) × 3 acid (R2) matrix and
highlights how a matrix approach can afford greater SAR insight
than optimizing a single position at a time. For example, when
R1 = 1-naphthylmethyl, the R2 substituent does not appear to
have a significant effect on BACE-1 activity (cf. 1 vs 8a vs 8b).
However, when R1 = 3,4-dichlorobenzyl, the R2 substituent
strongly influences the BACE-1 activity as seen by the
progression from 8f (R2 = H) to 8g (R2 = F) to 8h (R2 =
OMe); with a Ki = 910 nM, compound 8h represents a nearly
10-fold improvement in BACE-1 activity over 8f (Ki = 8800
nM). The same trend is observed when R1 = 3,5-dichlorobenzyl
(analogues 8i−8k), and this side chain afforded the most
potent analogue synthesized to date (8k) with a BACE-1 Ki =
670 nM. Given that the number of heavy atoms did not change
from 1 to 8k, the LE improved from 0.25 kcal/mol·heavy atom
to 0.29 kcal/mol·heavy atom. The 3,5-dichlorobenzyl side chain
replaced the 1-naphthylmethyl side chain for all subsequent
SAR studies, and it is likely that this side chain would not have
been identified without synthesizing a matrix library, as the
significance of the R2 = OMe substituent was not apparent in
the 1-naphthylmethyl series (cf. 1 and 8b), and the activity
enhancement provided by the 3,5-dichlorobenzyl side chain
was not observed when R2 = H (cf. 1 and 8i); it was only when
R1 and R2 were varied simultaneously that a substantial
improvement in activity was realized.
Table 2. Heterocyclic Isoxazole Replacements
The discovery of compound 8k prompted us to hold the 3,5-
dichlorobenzyl side chain constant and explore the isoxazole
side chain more extensively. This effort did not afford a
significant improvement in BACE-1 activity, but it did provide
two key SAR findings (Supporting Information, Table SI-1):
(1) the 5-position methyl group was required for activity (i.e.,
the 5-H and 5-Et analogues were inactive), and (2) although
the SAR for the 3-phenyl side chain was fairly flat, we observed
a consistent preference for electron-donating substituents (i.e.,
4-MeO). This SAR was applied to the exploration of isoxazole
replacements, as shown in Table 2. The “reverse” isoxazole 8l
afforded similar BACE-1 activity as the parent (8k), but
pyrazole 8m and isothiazole12 8o both provided a modest but
reproducible increase in BACE-1 activity. The N-methyl
analogue (8n) of pyrazole 8m was 20-fold less potent in the
BACE-1 binding assay, suggesting that the nitrogen of 8m (and
8k) engaged in a direct interaction with the BACE-1 protein.
Even though pyrazole 8m and isothiazole 8o were equipotent,
the pyrazole introduced another H-bond donor and this
prompted us to focus on isoxazoles and isothiazoles for
subsequent analogue efforts. The discovery of isothiazole 8o
prompted a re-examination of benzylamine substitution (Table
2, entries 8p−8s). Although the benzylamine SAR remained
narrow, two close analogues of the 3,5-dichlorobenzyl group
were equipotent with 8k: the 3-bromo-5-chlorobenzyl analogue
8r was an important precursor for a subsequent Suzuki
coupling library and the 3,5-dimethylbenzyl analogue 8s
suggested that chloro and methyl substituents could be used
interchangeably.
Starting with compound 1, the iterative synthesis of solid-
phase libraries led to the identification of 8o and a 10-fold
improvement in BACE-1 potency. However, with a BACE-1 Ki
= 290 nM, the potency of 8o remained at least an order of
magnitude from presumed therapeutic utility, and ensuing
optimization efforts focused on exploiting protein−ligand
interactions established for other classes of BACE-1 ligands.
To elucidate the binding mode of acyl guanidines to BACE-1,
X-ray crystallography efforts were initiated with compound 8k.
a
Mean radioligand displacement activity (n ≥ 2) according to ref 7.
Each Ki value is within 2.2-fold of the mean.
Co-crystals of a Phe-statine peptide13,14 bound to BACE-1 were
soaked with 8k, X-ray diffraction data were collected, and the
structure was determined. Initial electron density maps were of
marginal quality due to data resolution (2.8 Å) and incomplete
displacement of the Phe-statine peptide, resulting in partial
9210
dx.doi.org/10.1021/jm300931y | J. Med. Chem. 2012, 55, 9208−9223

Journal of Medicinal Chemistry
Article
Figure 2. Computational model of 8k bound to BACE-1 derived from the X-ray crystal structure. (A) Initial density of compound 8k contoured at
1.0σ (white mesh). Contiguous electron density is present for the 3,5-dichlorobenzyl and guanidine moieties but is incomplete for the acyl-isoxazole
and 4-methoxyphenyl groups. The computational model derived from the crystal structure is shown for reference (yellow carbons). Hydrogen bonds
to the conserved catalytic aspartates, Asp32 and Asp228, are shown as hashed yellow lines. (B) Computational model of 8k (yellow carbons) bound
to BACE-1. Residues predicted to form key contacts with 8k are shown as thin green sticks. Subsites within the BACE1 active site are indicated with
blue text. F108 is shown as thick green sticks. The figures were produced using Pymol.16
density for compound 8k. Even so, the electron density map
provided sufficient information to unambiguously place the acyl
guanidine and the 3,5-dichlorobenzyl groups of 8k in the active
site of BACE-1 (Figure 2A). The 3,5-dichlorobenzyl side chain
occupies the S1 pocket, and the acyl guanidine forms hydrogen
bonds with Asp32 and Asp228. The electron density for the 3-
(4-methoxyphenyl)-5-methylisoxazole side chain of 8k was very
weak, making unambiguous placement of this group impossible.
This resulted in a partial model of 8k bound to BACE-1.
Computational modeling was then used to evaluate the two
potential positions of the 4-methoxyphenyl side chain, one
which placed it in the S2′ pocket and a second orienting it in
the S2 pocket. The model of the former configuration resulted
in steric clashes with the side chain of Val332, ruling it out as a
reasonable bound conformation. No steric clashes were
observed in the S2 model, and so it was utilized as the
structural basis for further design as shown in Figure 2B. The
information gleaned from the X-ray crystal structure and the
computational models of BACE-1 with 8k suggested two
potential strategies for improving BACE-1 binding affinity.
First, it was apparent that the S3 site was unoccupied. Knowing
that introduction of substituents into the S3 pocket is a well
precedented strategy for improving BACE-1 binding affin-
ity,13−15 chemistry efforts were initiated at the 3-position of the
benzylamine side chain to reach into the S3 pocket. The model
also revealed that the backbone carbonyl of Phe108, at the edge
of the S1 pocket, was exposed and less than 4 Å from the 4-
position of the benzylamine of 8k. The proximity of this
carbonyl to the benzylamine suggested that the addition of
hydrogen bond donors to the 4-position could form favorable
interactions, and chemistry efforts were directed toward a
second set of analogues to investigate this hypothesis.
SI-1). As described in Tables SI-2 and SI-3 in the Supporting
Information, these analogues did not afford improved BACE-1
activity, and we turned our attention to C-linked analogues as
described below.
The solid-phase synthesis of acyl guanidines (Scheme 1)
offered a number of advantages over a solution-phase synthesis,
including the ability to utilize a split-mix synthetic protocol for
the rapid synthesis of small matrices of analogues in which both
the R1 and R2 substituents were varied. In addition, the crude
acyl guanidine product was generally very pure and this greatly
facilitated the final purification step via preparatory HPLC.
However, solid-phase synthesis is generally not the most
efficient method for the synthesis of multigram quantities of
material, and so we developed a solution-phase route to acyl
guanidines as shown in Scheme 2. By analogy to the synthesis
of a CBZ-protected guanidinylating reagent,17 treatment of S-
methylisothiourea with BOC anhydride provided BOC-S-
methylisothiourea 12 as a crystalline, stable solid. Acylation of
12 under standard conditions with our preferred isoxazole
(13a) or isothiazole (13b) provided S-methylisothiourea 14a
(or 14b), which upon sequential treatment with amine 15 and
TFA afforded acyl guanidine 17. We routinely synthesized
multigram quantities of intermediates 14a and 14b and utilized
Our initial efforts to increase BACE-1 potency via occupation
of the S3 pocket focused on the parallel synthesis of ethers (i.e.,
9), carbamates (i.e., 10), and amides (i.e., 11) obtained via
Mitsunobu alkylation of solid-supported phenols, carbamoyla-
tion of solid-supported benzyl alcohols, and acylation of solid-
supported amines, respectively (Supporting Information, Figure
9211
dx.doi.org/10.1021/jm300931y | J. Med. Chem. 2012, 55, 9208−9223

Journal of Medicinal Chemistry
Article
summarized in Table 3, aryl side chains (i.e., 19a and 19b) were
significantly less active than alkenyl substituents, possibly
because the resulting biphenyl group was too rigid to effectively
“reach” into the S3 pocket. In contrast, small alkene
substituents (i.e., 19c) were equipotent with 8r, but we were
intrigued that introduction of an oxygen, either as an alcohol
(19d) or ether (19f), 3 or 4 atoms distal from the phenyl ring
afforded a significant improvement in activity, particularly in
comparison to the “all carbon” analogue 19e. Even though 19d
and 19f were among the most potent acyl guanidines identified
at the time, two concerns prompted us to abandon this SAR
vector: (1) the preference for oxygen-containing substituents
suggested displacement of an adjacent H2O molecule rather
than occupation of the highly lipophilic S3 pocket, (2) in other
chemotypes, introducing substituents which occupied the S3
pocket afforded improvements in BACE-1 activity on the order
of 10−100-fold13−15 rather than the modest 4-fold improve-
ment observed between 8r (Ki = 510 nM) and 19d (Ki = 140
nM).
Having failed to identify substituents that occupy the S3
pocket, we turned our attention to substituents at the 4-
position of the benzylamine side chain, as the binding model
shown in Figure 1 suggested that BACE-1 backbone amide
bonds lay in close proximity. The chemistry to access these 4-
position analogues proved challenging, as the 3,5-dichloro
substitution pattern imparted significant steric and electronic
barriers to substitution at the 4-position. This necessitated the
introduction of the entire 4-position side chain prior to the
reaction of benzylamine 15 with 14a (or 14b). As shown in
Table 4, the 4-OH analogue 17a did not afford any
Scheme 2
this chemistry to synthesize gram quantities of key
intermediates, including the 3-bromo-5-chlorobenzyl analogue
8r as the penultimate intermediate in the synthesis of C-linked
3-benzyl analogues via Suzuki coupling chemistry.
As shown in eq 1, 8r facilitated the introduction of aryl and
alkenyl side chains via straightforward Suzuki chemistry. As
Table 4. BACE-1 Activity of 4-Substituted-3,5-
Dichlorobenzylamine Analogues
Table 3. BACE-1 Activity of 3-Aryl and 3-Alkenyl Analogues
BACE-1 Ki
HEK-Sw IC50
a
b,c
compd
X
R
(nM)
(nM)
8k
O
S
H
670
290
480
>5000
120
40
1600
8o
H
>5000
nd
17a
17b
17c
17d
17e
17f
O
O
O
S
OH
O-nPr
NH2
NH2
nd
nd
630
790
310
O
S
NHC(O)CH3
10
NHC(O)CH3
5
a
Mean radioligand displacement activity (n ≥ 2) according to ref 7.
b
Each Ki value is within 2.2-fold of the mean. Mean Aβ-lowering
activity in HEK 293 cells (n ≥ 2) according to ref 18. Each IC50 value
c
is within 1.6-fold of the mean. nd = not determined.
improvement in BACE-1 activity, and the corresponding n-Pr
ether 17b was inactive. However, the 4-NH2 side chain
provided a dramatic increase in activity with both the isoxazole
(17c, Ki = 120 nM) and isothiazole (17d, Ki = 40 nM)
analogues. With an LE of 0.34 kcal/mol·heavy atom, 17d
provided a significant improvement in ligand efficiency.
Acetylation of the 4-NH2 group afforded two very potent
acyl guanidines: isoxazole 17e (Ki = 10 nM) and isothiazole 17f
(Ki = 5 nM), both of which were approximately 60-fold more
a
Mean radioligand displacement activity (n ≥ 2) according to ref 7.
Each Ki value is within 2.2-fold of the mean.
9212
dx.doi.org/10.1021/jm300931y | J. Med. Chem. 2012, 55, 9208−9223

Journal of Medicinal Chemistry
Article
Figure 3. Crystal structures of compounds 17d (A) and 17g (B) bound to BACE1. Ligands are rendered in yellow sticks. Refined ligand electron
density contoured at 1.0σ is shown as white mesh. Residues within 4.0 Å of the ligand are shown as thin green sticks. Protein−ligand hydrogen bonds
are indicated as green dashed lines and ligand intramolecular hydrogen bonds as yellow dashed lines. The figures were produced using Pymol.16
Figure 4. Interactions between BACE1 (green) and compound 17g (yellow) rendered as a surface view (A), in 3D (B), and 2D (C). Ligand:protein
bonding interactions are depicted as dashed lines: hydrogen bonds, yellow; π−π, red; π−cation, blue; hydrophobic, green.The figures were produced
using Pymol.16
Table 5. Binding and Cellular BACE-1 Activity of 4-Acylamino Analogues
a
b
compd
X
R1
R2
BACE-1 Ki (nM)
HEK-Sw IC50 (nM)
IC50/Ki ratio
17f
S
Cl
CH3C(O)
5
20
280
40
30
50
30
30
50
20
60
50
50
310
240
62
12
17g
17h
17i
S
Me
Cl
CH3C(O)
c
S
CH3S(=O)2
nd
O
S
Me
Me
Cl
CH3CH2C(O)
780
430
19
13
40
21
63
32
16
23
4
17j
PhC(O)
17k
17l
O
O
O
O
O
O
O
O
CH3SCH2C(O)
CH3SCH2CH2C(O)
CH3S(=O)2CH2C(O)
cyclopropylC(O)
cyclopropylC(O)
2-furylC(O)
2000
640
Me
Cl
17m
17n
17o
17p
17q
17r
1900
1600
330
Cl
Me
Me
Cl
1400
210
(CH3)2NCH2C(O)
(CH3)2NCH2C(O)
Me
90
2
a
b
Mean radioligand displacement activity (n ≥ 2) according to ref 7. Each Ki value is within 2.2-fold of the mean. Mean Aβ lowering activity in HEK
293 cells (n ≥ 2) according to ref 18. nd = not determined. Each IC50 value is within 1.6-fold of the mean.
c
9213
dx.doi.org/10.1021/jm300931y | J. Med. Chem. 2012, 55, 9208−9223

Journal of Medicinal Chemistry
Article
Scheme 3
Table 6. Binding and Cellular BACE-1 Activity of Aminoacetamide Analogues
a
b
Mean radioligand displacement activity (n ≥ 2) according to ref 7. Each Ki value is within 2.2-fold of the mean. Mean Aβ lowering activity in HEK
293 cells (n ≥ 2) according to ref 18. Each IC50 value is within 1.6-fold of the mean.
potent than the corresponding 4-H analogues (8k and 8o,
respectively). Compound 17f maintained an LE of 0.34 kcal/
mol·heavy atom. Along with the dramatic improvement in
BACE-1 binding activity, these potent analogues were also the
first acyl guanidine analogues to afford submicromolar activity
in our cell-based (HEK-Sw) assay18 in which HEK cells were
engineered to overexpress amyloid precursor protein (APP)
containing the Swedish mutation and the effect of BACE-1
inhibitors on the production of Aβ was measured via ELISA.
The consistent disparity between binding and cellular activity
will be discussed in greater detail.
density for the entire compound. This structure was largely
consistent with the partial X-ray structure and resulting
computational model of 8k bound to BACE-1 (Figure 2); in
particular, it confirmed the predicted orientation of the 3-(4-
methoxyphenyl)-5-methylisoxazole into the S2 pocket (Figure
3A). As anticipated from SAR and modeling, the amino group
at the 4-position of the benzylamine side chain of 17d formed a
hydrogen bond to the backbone carbonyl of Phe108, and we
observed an additional hydrogen bond to the backbone
carbonyl from Lys107. Collectively, these hydrogen bonds
provide a structural basis for the observed increase in potency.
The structure also revealed an unexpected conformation of the
side chain of flap residue Gln73, allowing the Gln73 side chain
to form a hydrogen bond to the isoxazole carbonyl oxygen of
At this point, we were gratified to obtain the X-ray crystal
structure of 17d bound to BACE-1 (Figure 3A). Initial electron
density maps were of good quality and revealed unambiguous
9214
dx.doi.org/10.1021/jm300931y | J. Med. Chem. 2012, 55, 9208−9223

Journal of Medicinal Chemistry
Article
17d. The unanticipated conformation of the Gln73 side chain is
additionally stabilized by van der Waals interactions with the 4-
methoxyphenyl group of the isoxazole. A crystal structure of the
N-acetyl analogue 17g was also obtained (Figure 3B). In this
structure, a hydrogen bond between the backbone carbonyl of
Phe108 and the acetamide NH is observed with the acetamide
moiety itself being oriented away from the S1 pocket, toward
solvent. The structure also reveals a total of 18 protein−ligand
interactions encompassing nearly every heavy atom of 17g and
highlighting the high ligand efficiency for the series (Figure 4).
The discovery of the 4-acetamido-3,5-dichlorobenzyl ana-
logues 17e and 17f prompted a thorough exploration of other
4-acylamino groups (and acylamino mimetics) as summarized
in Table 5. The SAR for 4-acylamino analogues was largely flat;
replacing the carbonyl group of 17f with a sulfonyl group (17h)
had a detrimental effect on the BACE-1 Ki value, but otherwise
all of the acyl analogues contained in Table 5 were equipotent.
However, while developing the chemistry to access these
analogues, we discovered an unusual phenomenon: when a 3-
methyl group was substituted for the 3-chloro group in
compound 17f, the ratio between the BACE-1 IC50 (measured
in the cellular/HEK-Sw assay) and the BACE-1 Ki (measured
in the radioligand binding assay) was consistently lower for the
methyl analogue (17g). This trend was also observed with
analogues 17n and 17o: the BACE-1 Ki values were not
affected (within experimental error) by the change from R1 =
Cl to R1 = Me, but the methyl-containing compound 17o was
5-fold more potent in the HEK-Sw assay. We have no
explanation for this “methyl effect” on cellular potency, and we
were further surprised by the 4-(2-(N,N-dimethylamino)-
acetamido) analogues 17q and 17r; in this case, the
introduction of a basic side chain dramatically improved the
cellular activity and provided IC50 to Ki ratios approaching
unity. A similar “pKa effect” on cellular potency has been
reported19,20 for two structurally unrelated BACE-1 chemo-
types, but our case is unique in several respects. In the first
report,19 the authors discovered that increasing the pKa of an
amino group that directly interacts with the active site
aspartates afforded an activity increase in both binding and
cellular assays. In contrast, the basic side chain of our series is
distal to the active site and the activity increase is observed only
in the cellular assay; the binding activity is unaffected. In the
second report,20 a 3-fold improvement in cellular activity (with
no effect on binding activity) was obtained when a pyridyl N
was substituted for a phenyl CH, although the ratio between
cellular and binding activities remained high at 19-fold. This is
consistent with our observation, although the magnitude of the
shift in our series is more pronounced: introduction of a basic
amine afforded an 8-fold improvement in cellular activity (cf.
17i and 17r), and a 2- to 4-fold IC50 to Ki ratio. The origin of
this “basic amine effect” is likely the result of altered drug
distribution within cells, and we hypothesize that the basic
amine modifies the physical properties of this series to afford a
higher effective concentration of drug within the cellular
compartment where BACE-1 processes Aβ.
both the BACE-1 binding and HEK-Sw cellular assays were
identified, and azetidine-containing analogue 22h emerged as
the most potent acyl guanidine identified to date, with single-
digit nanomolar activity in both assays. Compound 22h
exhibited no Herg activity in a thallium flux assay21 and was
inactive against a panel of aspartyl proteases7 (i.e., cathepsin D,
cathepsin E, and pepsin; Supporting Information Table SI-4).
As described previously, acetamide analogues bearing a basic
side chain (i.e., 17q−17r, 22a−22h) afforded a significant
increase in activity in the cellular assay relative to nonbasic
acetamides (i.e., 17f−17p), with little or no effect on activity in
the binding assay. We were thus not surprised to find that the
X-ray co-crystal structure of 17r and BACE-1 (unpublished
data) did not reveal any additional ligand−protein interactions,
and the pendant N,N-dimethylamino group appears to be
solvent-exposed. However, we did note the close proximity of
the N,N-dimethylamino group and the 4-methoxyphenyl group
of the isoxazole; in the crystal structure, the C−C distance
́
between these methyl groups is 4.9 Å. This observation
prompted the development of a straightforward ring-closing
metathesis route to macrocyclic isoxazolyl acyl guanidines, and
the synthesis and biological properties of these macrocycles will
be described in a future publication.
The cellular potency of compound 22h prompted us to
evaluate its effect on Aβ1−40 levels in the plasma, brain. and
cerebral spinal fluid (CSF) of rats.13,18 Three subcutaneous (sc)
doses were selected: 5, 15. and 40 mg/kg, and Aβ1−40 and
drug levels were measured 5 h after administration of 22h. The
40 mg/kg group was pretreated with a 50 mg/kg oral dose of
aminobenzotriazole (ABT), a pan-cytochrome P450 (CYP
P450) inhibitor, in order to improve drug levels by blocking
CYP-mediated metabolism. As shown in Figure 5, dramatic and
dose-dependent reductions in plasma Aβ1−40 were observed.
However, despite the robust peripheral effect, significant
reductions of brain or CSF Aβ1−40 were not observed at
any dose.
Additional amine-containing analogues were synthesized
from α-bromoacetamide intermediates 20a and 20b as shown
in Scheme 3, and the BACE-1 binding and cellular data are
summarized in Table 6. Comparing the IC50/Ki ratios for each
pair of analogues (i.e., 22a and 22b, 22c and 22d) provided
additional evidence for the aforementioned “methyl effect” on
cellular potency. Combined with the “basic amine effect” on
cellular potency, a number of samples with low nM activity in
Figure 5. Levels of Aβ1−40 in the plasma, brain, and CSF of
laboratory rats 5 h after sc administration of 5, 15, and 40 mg/kg doses
of 22h. All three doses of 22h produced statistically significant
reductions in plasma Aβ levels relative to vehicle controls.
9215
dx.doi.org/10.1021/jm300931y | J. Med. Chem. 2012, 55, 9208−9223

Journal of Medicinal Chemistry
Article
The source of the lack of Aβ reduction in the brain and CSF
is readily apparent in Figure 6: although high plasma
minor modification produced a 10-fold increase in BACE-1
binding affinity and a 10-fold decrease in functional activity in
KBV cells. All attempts to overcome this P-gp efflux issue while
maintaining low nM BACE-1 activity were unsuccessful.
CONCLUSIONS
■
This work establishes the viability of the acyl guanidine
functional group as an aspartyl protease active site isostere. By
taking advantage of a previously developed route10 for the
solid-phase synthesis of acyl guanidine libraries and the
information gleaned from X-ray crystal structures, we rapidly
optimized a micromolar HTS hit into a single digit nanomolar
lead that afforded statistically significant reductions in
peripheral Aβ levels in rats 5 h after administration of a
subcutaneous dose. Brain Aβ levels were not reduced at any
dose, most likely due to P-gp mediated efflux. X-ray
crystallography of a representative acyl guanidine:BACE-1
complex revealed the unusual binding mode of this series;
even though other BACE-1 inhibitors bearing acyl guani-
dine23−29 or acyl guanidine-like30−32 functional groups have
been reported, none of them bind in the same orientation or
exploit the same pattern of hydrogen binding interactions.
Additional studies that explore the BACE-1 activity of
macrocyclic isoxazolyl acyl guanidines will be reported in due
course.
Figure 6. Levels of 22h in the plasma, brain, and CSF of laboratory
rats 5 h after sc administration of 5, 15, and 40 mg/kg doses of 22h.
EXPERIMENTAL SECTION
■
General. Unless otherwise noted, all starting materials, reagents,
and solvents were purchased from commercial vendors (Aldrich,
VWR, Lancaster, etc.) and used as supplied without further
purification. Wang resin was purchased from Polymer Laboratories.
NMR spectra were recorded on a Bruker DPX-500, Bruker DPX-400,
or Bruker DPX-300 NMR spectrometer. Chemical shifts are expressed
in parts per million (δ) using a residual solvent proton as the internal
standard. The following abbreviations are used to designate the
multiplicities: s, singlet; d, doublet; t, triplet; q, quartet; m, multiplet.
LC/MS analysis to assess final purity was performed on a Shimadzu
HPLC equipped with a Phenomenex C18, 3.0 mm × 50 mm column,
a Micromass spectrometer, and eluting with a mixture of MeOH and
H2O containing 0.1% TFA with a gradient of 10−100% over 5 min. All
samples provided HPLC purity (area% at 254 nm) ≥95%. High
resolution mass spectrometry (HRMS) analyses were performed on a
Fourier Transform Orbitrap mass spectrometer (Exactive, Thermo
Fisher Scientific, San Jose, CA) in positive or negative ionization
electrospray mode operating at 25000 resolution (full width at half
height maximum, fwhm). The instrument was daily calibrated
according to manufacturer’s specifications resulting in mass accuracy
of or better than 5 ppm. The operating software Xcalibur was used to
calculate theoretical mass-to-charge values and to process the obtained
data.
concentrations of 22h were achieved, low drug levels were
observed in the brain and CSF, independent of dose. As has
been observed with many other BACE-1 ligands, the source of
the low brain and CSF levels can be ascribed to P-glycoprotein
(P-gp) efflux. To directly measure this effect, we utilized a
cellular assay in which KBV cells (which express high levels of
P-gp in comparison to HEK cells) were transfected with a
plasmid that expressed APP containing the Swedish mutation.18
The KBV-Sw cellular assay was conducted with a representative
group of acyl guanidines both in the presence and absence of
the P-gp inhibitor tariquidar,22 as shown in Table 7. The effect
Table 7. Effect of P-gp Inhibition on KBV-Sw IC50 Values for
Representative Acyl Guanidines
BACE-1 Ki KBV-Sw IC50 KBV-Sw + P-gp inhibitor
P-gp
d
a
b
c
compd
(nM)
(nM)
IC50 (nM)
shift
8k
670
60
5
3300
1700
>10000
180
3900
1030
310
8
0.8
1.7
>32
23
17d
17f
22g
22h
4
5
180
5
36
General Procedure for Solid Phase Synthesis of Acyl
Guanidines (Method A). Resin-Bound 4-Nitrophenylcarbonate
Intermediate (2). To a suspension of Wang resin (50 g, 1.3 mmol/g,
65.0 mmol, 1.0 equiv) in 800 mL of anhydrous DCM was added 4-
nitrophenyl chloroformate (39.3 g, 195.0 mmol, 3.0 equiv) and N-
methylmorpholine (29.0 mL, 260 mmol, 4.0 equiv) at 25 °C. The
resulting mixture was shaken at 25 °C overnight, then filtered and
washed with DCM (4 × 500 mL) and ether (2 × 500 mL), and dried
under house vacuum overnight to afford 74.5 g of 2. Then 30 mg of
the above resin was treated with 2 mL of 30% TFA in DCM for 1 h,
then filtered. The filtrate was concentrated in vacuo to afford 5.1 mg of
4-nitrophenol, corresponding to a calculated loading of 1.2 mmol/g.
Resin-Bound S-Methylisothiourea Intermediate (3). To resin 2
(14 g, 16.8 mmol) suspended in 400 mL of DMF was added S-
methylisothiourea sulfate (18.7 g, 67.2 mmol, 4.0 equiv) and cesium
carbonate (21.9 g, 67.2 mmol, 4.0 equiv). The resulting suspension
was shaken at room temperature for 2 d, then filtered, and the resin
a
Mean radioligand displacement activity (n ≥ 2) according to ref 7.
b
c
Aβ lowering activity in KBV cells (n ≥ 2) according to ref 18. Mean
Aβ lowering activity in the presence of 20 μM P-gp inhibitor in KBV
cells (n ≥ 2) according to ref 18. Calculated as the ratio of KBV-SW
IC50 values in the absence and presence of tariquidar (P-gp inhibitor).
d
of P-gp inhibition on the KBV-Sw IC50 values (expressed as a
ratio that we called the “P-gp shift”) provided strong evidence
that P-gp efflux was preventing the compound from inhibiting
intracellular BACE-1 activity. It is striking that the initial acyl
guanidine analogues (i.e., 8k and 17d) did not exhibit large P-
gp shifts, but as BACE-1 binding potency increased, so did the
magnitude of the P-gp shift. For example, the only difference
between 17d and 17f is the addition of an acetyl group; this
9216
dx.doi.org/10.1021/jm300931y | J. Med. Chem. 2012, 55, 9208−9223

Journal of Medicinal Chemistry
Article
was washed with DMF (3 × 200 mL), THF (3 × 200 mL), and
MeOH (3 × 200 mL) and then dried in vacuo for 1 d to afford 14 g of
3. Resin 3 was distributed into 400 Microkans (purchased from
IRORI, 35 mg resin/Microkan, 42 μmol) and stored under nitrogen
until needed.
Resin-Bound Acyl S-Methylisothiourea Intermediate (5). To a
reaction flask containing a Microkan loaded with resin 3 (35 mg, 42
μmol) was added NMP (2 mL), the appropriate isoxazole acid 4 (0.21
mmol, 5.0 equiv), PyAOP (0.21 mmol, 5.0 equiv), and DIEA (0.21
mmol, 5.0 equiv). The reaction flask was shaken at 25 °C for 24 h,
filtered, and washed with DMF (3 × 2 mL), MeOH (3 × 2 mL), DCM
(3 × 2 mL), and MeOH (3 × 2 mL) and then dried in vacuo overnight
to afford a Microkan containing resin-bound acyl S-isothiourea 5 (35
mg, 42 μmol).
N-(Amino(benzylamino)methylene)-5-methyl-3-(4-
methoxyphenyl)isoxazole-4-carboxamide (8e). Compound 8e was
prepared from 3 via method A using 3-(4-methoxyphenyl)-5-
methylisoxazole-4-carboxylic acid as acid 4 and benzylamine as
amine 6. Preparative HPLC purification afforded 2.5 mg (16% yield
1
from 3) of 8e as a white solid. H NMR (500 MHz, CD3OD) δ 7.63
(d, J = 7.4 Hz, 2H), 7.44−7.42 (m, 2H), 7.39 (d, J = 7.2 Hz, 2H), 4.58
(s, 2H), 3.87 (s, 3H), 2.69 (s, 3H). MS (ESI+) (m/z) 365.1 ([M +
H]+).
N-(Amino(3,4-dichlorobenzylamino)methylene)-5-methyl-3-phe-
nylisoxazole-4-carboxamide (8f). Compound 8f was prepared from 3
via method A using 5-methyl-3-phenylisoxazole-4-carboxylic acid as
acid 4 and 3,4-dichlorobenzylamine as amine 6. Preparative HPLC
1
purification afforded 15.1 mg (89% yield from 3) of 8f as an oil. H
NMR (500 MHz, CD3OD) δ 7.69−7.63 (m, 2H), 7.56−7.52 (m, 5H),
7.32 (s, 1H), 4.58 (s, 2H), 2.72 (s, 3H). MS (ESI+) (m/z) 403.1 ([M
+ H]+).
Resin-Bound Acyl Guanidine (7). To a flask containing a Microkan
loaded with the appropriate acyl S-methylthiourea intermediate 5 (35
mg, 42 μmol) was added DCM (2 mL), DIEA (0.21 mmol, 5.0 equiv),
and amine 6 (0.21 mmol, 5.0 equiv). The flask was shaken at 25 °C for
16 h, filtered, and washed with DCM (3 × 2 mL), MeOH (3 × 2 mL),
then dried in vacuo overnight to afford a Microkan containing resin-
bound acyl guanidine 7 (35 mg, 42 μmol).
Acyl Guanidines 1, 8a−8s. To a vial containing a Microkan loaded
with the appropriate resin-bound acyl guanidine 7 (35 mg, 42 μmol)
was added 50% TFA/DCM (2 mL), and the resulting suspension was
shaken at 25 °C for 1 h. The Microkan was filtered and washed with
50% TFA/DCM (2 × 1 mL), and the combined TFA/DCM fractions
were concentrated in vacuo and purified by preparative HPLC to
afford compounds 1 and 8a−8s.
N-(Amino(naphthalen-1-ylmethylamino)methylene)-5-methyl-3-
phenylisoxazole-4-carboxamide (1). Compound 1 was prepared
from 3 via method A using 5-methyl-3-phenylisoxazole-4-carboxylic
acid as acid 4 and 1-naphthylmethylamine as amine 6. Preparative
HPLC purification afforded 7.7 mg (48% yield from 3) of 1 as a white
solid. 1H NMR (500 MHz, CD3OD) δ 7.99−7.96 (m, 4H), 7.69−7.60
(m, 3H), 7.55−7.53 (m, 5H), 5.04 (s, 2H), 2.70 (s, 3H). HRMS (ESI
+): calculated for C23H21N4O2, [M + H]+ = 385.1659; observed, [M +
H]+ = 385.1650.
N-(Amino(3,4-dichlorobenzylamino)methylene)-5-methyl-3-(4-
fluorophenyl)isoxazole-4-carboxamide (8g). Compound 8g was
prepared from 3 via method A using 3-(4-fluorophenyl)-5-methyl-
isoxazole-4-carboxylic acid as acid 4 and 3,4-dichlorobenzylamine as
amine 6. Preparative HPLC purification afforded 3.4 mg (19% yield
1
from 3) of 8g as an oil. H NMR (300 MHz, CD3OD) δ 7.77−7.72
(m, 2H), 7.61−7.58 (m, 2H), 7.29−7.23 (m, 3H), 4.58 (s, 2H), 2.71
(s, 3H). MS (ESI+) (m/z) 421.1 ([M + H]+).
N-(Amino(3,4-dichlorobenzylamino)methylene)-5-methyl-3-(4-
methoxyphenyl)isoxazole-4-carboxamide (8h). Compound 8h was
prepared from 3 via method A using 3-(4-methoxyphenyl)-5-
methylisoxazole-4-carboxylic acid as acid 4 and 3,4-dichlorobenzyl-
amine as amine 6. Preparative HPLC purification afforded 3.1 mg
1
(17% yield from 3) of 8h as an oil. H NMR (500 MHz, CD3OD) δ
7.64−7.57 (m, 4H), 7.33−7.31 (m, 1H), 7.06 (d, J = 8.7 Hz, 2H), 4.58
(s,2 H), 3.86 (s, 3H), 2.71 (s, 3H). MS (ESI+) (m/z) 433.1 ([M +
H]+).
N-(Amino(3,5-dichlorobenzylamino)methylene)-5-methyl-3-phe-
nylisoxazole-4-carboxamide (8i). Compound 8i was prepared from 3
via method A using 5-methyl-3-phenylisoxazole-4-carboxylic acid as
acid 4 and 3,5-dichlorobenzylamine as amine 6. Preparative HPLC
purification afforded 9.6 mg (57% yield from 3) of 8i as a white solid.
1H NMR (300 MHz, CD3OD) δ 7.69 (d, J = 6.1 Hz, 2H), 7.55−7.53
N-(Amino(naphthalen-1-ylmethylamino)methylene)-3-(4-fluoro-
phenyl)-5-methylisoxazole-4-carboxamide (8a). Compound 8a was
prepared from 3 via method A using 3-(4-fluorophenyl)-5-methyl-
isoxazole-4-carboxylic acid as acid 4 and 1-naphthylmethylamine as
amine 6. Preparative HPLC purification afforded 9.7 mg (60% yield
(m, 3H), 7.48 (s, 1H), 7.37 (s, 2H). 4.59 (s, 2H), 2.72 (s, 3H). MS
(ESI+) (m/z) 403.1 ([M + H]+).
1
N-(Amino(3,5-dichlorobenzylamino)methylene)-5-methyl-3-(4-
fluorophenyl)isoxazole-4-carboxamide (8j). Compound 8j was
prepared from 3 via method A using 3-(4-fluorophenyl)-5-methyl-
isoxazole-4-carboxylic acid as acid 4 and 3,5-dichlorobenzylamine as
amine 6. Preparative HPLC purification afforded 4.6 mg (26% yield
from 3) of 8a as an oil. H NMR (500 MHz, CD3OD) δ 8.03−7.88
(m, 4H), 7.73−7.54 (m, 5H), 7.24−7.19 (m, 2H), 5.05 (s, 2H), 2.70
(s, 3H). MS (ESI+) (m/z) 403.2 ([M + H]+).
N-(Amino(naphthalen-1-ylmethylamino)methylene)-3-(4-me-
thoxyphenyl)-5-methylisoxazole-4-carboxamide (8b). Compound
8b was prepared from 3 via method A using 3-(4-methoxy)-5-
methylisoxazole-4-carboxylic acid as acid 4 and 1-naphthylmethyl-
amine as amine 6. Preparative HPLC purification afforded 5.6 mg
1
from 3) of 8j as a white solid. H NMR (500 MHz, CD3OD) δ 7.75
(dd, J = 8.5 Hz, J = 5.7 Hz, 2H), 7.48 (s, 1H), 7.38 (s, 2H), 7.27 (t, J =
8.1 Hz, 2H), 4.60 (s, 2H), 2.72 (s, 3H). MS (ESI+) (m/z) 421.1 ([M
+ H]+).
1
(33% yield from 3) of 8b as a white solid. H NMR (500 MHz,
N-(Amino(3,5-dichlorobenzylamino)methylene)-5-methyl-3-(4-
methoxyphenyl)isoxazole-4-carboxamide (8k). Compound 8k was
prepared from 3 via method A using 3-(4-methoxyphenyl)-5-
methylisoxazole-4-carboxylic acid as acid 4 and 3,4-dichlorobenzyl-
amine as amine 6. Preparative HPLC purification afforded 15 mg (83%
yield from 3) of 8k as a white solid. 1H NMR (500 MHz, CD3OD) δ
7.64 (d, J = 7.0 Hz, 2H), 7.48 (s, 1H), 7.37 (s, 2H), 7.07 (d, J = 7.0
Hz, 2H), 4.59 (s, 2H), 3.87 (s, 3H), 2.70 (s, 3H). HRMS (ESI+):
calculated for C20H19N4O3Cl2, [M + H]+ = 433.0829; observed, [M +
H]+ = 433.0818.
CD3OD) δ 8.01−7.95 (m, 4H), 7.65−7.51 (m, 5H), 7.05−7.03 (m,
2H), 5.04 (s, 2H), 3.85 (s, 3H), 2.70 (s, 3H). MS (ESI+) (m/z) 414.1
([M + H]+).
N-(Amino(benzylamino)methylene)-5-methyl-3-phenylisoxazole-
4-carboxamide (8c). Compound 8c was prepared from 3 via method
A using 5-methyl-3-phenylisoxazole-4-carboxylic acid as acid 4 and
benzylamine as amine 6. Preparative HPLC purification afforded 9.6
1
mg (68% yield from 3) of 8c as a white solid. H NMR (500 MHz,
CD3OD) δ 7.69 (d, J = 7.3 Hz, 2H), 7.53 (d, J = 7.3 Hz, 3H), 7.42−
7.38 (m, 5H), 4.58 (s, 2H), 2.71 (s, 3H). MS (ESI+) (m/z) 335.2 ([M
+ H]+).
N-(Amino(3,5-dichlorobenzylamino)methylene)-3-methyl-5-(4-
methoxyphenyl)isoxazole-4-carboxamide (8l). Compound 8l was
prepared from 3 via method A using 3-methyl-5-(4-methoxyphenyli-
soxazole-4-carboxylic acid as acid 4 and 3,5-dichlorobenzylamine as
amine 6. Preparative HPLC purification afforded 15.7 mg (86% yield
N-(Amino(benzylamino)methylene)-5-methyl-3-(4-fluorophenyl)-
isoxazole-4-carboxamide (8d). Compound 8d was prepared from 3
via method A using 3-(4-fluorophenyl)-5-methylisoxazole-4-carboxylic
acid as acid 4 and benzylamine as amine 6. Preparative HPLC
purification afforded 10.9 mg (74% yield from 3) of 8d as a white
solid. 1H NMR (500 MHz, CD3OD) δ 7.69 (d, J = 7.3 Hz, 2H), 7.53
(d, J = 7.3 Hz, 2H), 7.42−7.38 (m, 5H), 4.58 (s, 2H), 2.71 (s, 3H).
MS (ESI+) (m/z) 353.3 ([M + H]+).
1
from 3) of 8l as an off-white solid. H NMR (500 MHz, CD3OD) δ
7.81 (d, J = 8.6 Hz, 2H), 7.49 (s, 1H), 7.39 (s, 2H), 7.13 (d, J = 8.6
Hz, 2H), 4.61 (s, 2H), 3.90 (s, 3H), 2.46 (s, 3H). MS (ESI+) (m/z)
433.1 ([M + H]+).
9217
dx.doi.org/10.1021/jm300931y | J. Med. Chem. 2012, 55, 9208−9223

Journal of Medicinal Chemistry
Article
N-(Amino(3,5-dichlorobenzylamino)methylene)-3-(4-methoxy-
phenyl)-5-methyl-1H-pyrazole-4-carboxamide (8m). Compound 8m
was prepared from 3 via method A using 3-(4-methoxyphenyl)-5-
methyl-1H-pyrazole-4-carboxylic acid as acid 4 and 3,5-dichlorobenzyl-
amine as amine 6. Preparative HPLC purification afforded 3.8 mg
(21% yield from 3) of 8m as an oil. 1H NMR (500 MHz, CD3OD) δ
7.56 (d, J = 8.5 Hz, 2H), 7.48 (s, 1H), 7.35 (s, 2H), 7.06 (d, J = 8.5
Hz, 2H), 4.61 (s, 2H), 3.85 (s, 3H), 2.51 (s, 3H). MS (ESI+) (m/z)
432.1 ([M + H]+).
N-(Amino(3,5-dichlorobenzylamino)methylene)-5-(4-methoxy-
phenyl)-1,3-dimethyl-1H-pyrazole-4-carboxamide (8n). Compound
8n was prepared from 3 via method A using 5-(4-methoxyphenyl)-1,3-
dimethyl-1H-pyrazole-4-carboxylic acid as acid 4 and 3,5-dichlor-
obenzylamine as amine 6. Preparative HPLC purification afforded 3.5
mg (19% yield from 3) of 8n as an oil. 1H NMR (500 MHz, CD3OD)
δ 7.47−7.43 (m, 4H), 7.31 (s, 1H), 7.11 (d, J = 8.75 Hz, 2H), 4.61 (s,
2H), 3.72 (s, 3H), 2.84 (s, 3H), 2.46 (s, 3H). MS (ESI+) (m/z) 446.1
([M + H]+).
× 500 mL) and dried over Na2SO4. Filtration and concentration
provided tert-butyl methylthiocarbonoimidoylcarbamate (12) as a
white solid (35.5 g, 0.187 mol, 94% yield based on Boc2O). H NMR
1
(500 MHz, CDCl3) δ 2.47 (s, 3H), 1.52 (s, 9H).
tert-Butyl N-(3-(4-methoxyphenyl)-5-methylisoxazole-4-
carbonyl)(methylthio)carbonoimidoylcarbamate (14a). A mixture
of tert-butyl methylthiocarbonoimidoylcarbamate (0.98 g, 5.15 mmol),
3-(4-methoxyphenyl)-5-methyl-4-isoxazolecarboxylic acid (1.17 g, 5.02
mmol), EDCI (1.20 g, 6.26 mmol), and DMAP (1.22 g, 10.0 mmol) in
DCM (30 mL) was stirred for 3 h. The solvent was evaporated by a
rotary evaporator and residue was purified by flash chromatography
(DCM, Rf 0.38). The product 14a was obtained as a white solid (1.2 g,
60% yield). 1H NMR (500 MHz, CD3OD) δ 7.46 (d, J = 8.9 Hz, 2H),
6.99 (d, J = 8.5 Hz, 2H), 3.84 (s, 3H), 2.74 (s, 3H), 1.87 (s, 3H), 1.50
(s, 9H). MS (ESI+) (m/z) 406 ([M + H]+).
tert-Butyl N-(3-(4-Methoxyphenyl)-5-methylisothiazole-4-
carbonyl)(methylthio)carbonoimidoylcarbamate (14b). Compound
14b was synthesized by analogy to the preparation of 14a, substituting
3-(4-methoxyphenyl)-5-methyl-4-isothiazolecarboxylic acid for 3-(4-
methoxyphenyl)-5-methyl-4-isoxazolecarboxylic acid to afford 14b as a
white solid. 1H NMR (400 MHz, CDCl3) δ 12.36 (s, 1H), 7.46 (d, J =
8.8 Hz, 2H), 6.92 (d, J = 8.8 Hz, 2H), 3.84 (s, 3H), 2.75 (s, 3H), 1.72
(s, 3H), 1.53 (s, 9H). MS (ESI+) (m/z) 422 ([M + H]+).
General Procedure for the Preparation of BOC-Protected Acyl
Guanidine (16). The reaction mixture of 20 mg of tert-butyl N-(3-(4-
methoxyphenyl)-5-methylisothiazole-4-carbonyl)(methylthio)-
carbonoimidoylcarbamate (14b), 23 mg of amine 15, and 0.030 mL of
DIEA in 4 mL of DCM was stirred at room temperature overnight and
then concentrated in vacuo to afford 16 as a crude product.
General Procedure for the Preparation of Acyl Guanidine (17).
Intermediate 16 was treated with 0.5 mL of 50% TFA in DCM at
room temperature for 1 h. The reaction mixture was concentrated in
vacuo, and the product was purified by preparative HPLC (0.1% TFA
MeOH/H2O) to afford acyl guanidine 17.
N-(Amino(3,5-dichlorobenzylamino)methylene)-3-(4-methoxy-
phenyl)-5-methylisothiazole-4-carboxamide (8o). Compound 8o
was prepared from 3 via method A using 3-(4-methoxyphenyl)-5-
methylisothiazole-4-carboxylic acid as acid 4 and 3,5-dichlorobenzyl-
amine as amine 6. Preparative HPLC purification afforded 11.6 mg
1
(62% yield from 3) of 8o as a white solid. H NMR (500 MHz,
CD3OD) δ 7.60 (d, J = 8.78 Hz, 2H), 7.47−7.46 (m, 1H), 7.33 (s,
2H), 7.02 (d, J = 8.8 Hz, 2H), 4.59 (s, 2H), 3.84 (s, 3H), 2.73 (s, 3H).
MS (ESI+) (m/z) 449.0 ([M + H]+).
N-(Amino(naphthalen-1-ylmethylamino)methylene)-3-(4-me-
thoxyphenyl)-5-methylisothiazole-4-carboxamide (8p). Compound
8p was prepared from 3 via method A using 3-(4-methoxyphenyl)-5-
methylisothiazole-4-carboxylic acid as acid 4 and 1-naphthylmethyl-
amine as amine 6. Preparative HPLC purification afforded 14 mg (77%
yield from 3) of 8p as a white solid. 1H NMR (500 MHz, CD3OD) δ
8.0−7.93 (m, 3H), 7.67−7.48 (m, 6H), 7.02−6.99 (m, 2H), 5.05 (s,
2H), 3.86 (s, 3H), 2.73 (s, 3H). MS (ESI+) (m/z) 431.2 ([M + H]+).
N-(Amino(naphthalen-2-ylmethylamino)methylene)-3-(4-me-
thoxyphenyl)-5-methylisothiazole-4-carboxamide (8q). Compound
8q was prepared from 3 via method A using 3-(4-methoxyphenyl)-5-
methylisothiazole-4-carboxylic acid as acid 4 and 2-naphthylmethyl-
amine as amine 6. Preparative HPLC purification afforded 2.7 mg
N - ( ( 4 - A m i n o - 3 , 5 - d i c h l o r o b e n z y l a m i n o ) ( t e r t -
butoxycarbonylamino)methylene)-3-(4-methoxyphenyl)-5-methyli-
sothiazole-4-carboxamide (16a). A mixture of 14b (0.830 g, 1.97
mmol), 3,5-dichloro-4-aminobenzylamine (0.450 g, 2.36 mmol), and
DIPEA (0.61 g, 4.72 mmol) in DCM (20 mL) was stirred at room
temperature overnight. Water (100 mL) and DCM (80 mL) were
added to the reaction mixture, and the solution was transferred to a
separatory funnel. The two layers were separated, and the aqueous
solution was extracted with 100 mL of DCM. The combined extract
was dried over anhydrous sodium sulfate and then filtered and
concentrated in vacuo.The crude product was purified by flash
chromatography (DCM, Rf 0.21) to give 0.920 g (83% yield) of 16a as
1
(15% yield from 3) of 8q as a white solid. H NMR (500 MHz,
CD3OD) δ 8.0−7.93 (m, 3H), 7.67−7.48 (m, 6H), 7.02−6.99 (m,
2H), 5.05 (s, 2H), 3.86 (s, 3H), 2.73 (s, 3H). MS (ESI+) (m/z) 431.2
([M + H]+).
N-(Amino(3-chloro-5-bromobenzylamino)methylene)-3-(4-me-
thoxyphenyl)-5-methylisothiazole-4-carboxamide (8r). Compound
8r was prepared from 3 via method A using 3-(4-methoxyphenyl)-5-
methylisothiazole-4-carboxylic acid as acid 4 and 3-chloro-5-
bromobenzylamine as amine 6. Preparative HPLC purification
1
a white solid. H NMR (500 MHz, CDCl3) δ 12.09 (s, 1H), 8.58 (s,
1H), 7.50 (d, J = 8.7 Hz, 2H), 6.95 (s, 2H), 6.88 (d, J = 8.7 Hz, 2H),
4.43 (s, 2H), 3.90 (d, J = 5.5 Hz, 2H), 3.77 (s, 3H), 2.69 (s, 3H), 1.49
(s, 9H). MS (ESI+) (m/z) 564 ([M + H]+).
1
afforded 5.8 mg (20% yield from 3) of 8r as a white solid. H NMR
(500 MHz, MeOD) δ 7.3 (m,3H), 7.49 (s, 1H), 7.38 (s, 1H), 7.04 (d,
J = 8.5 Hz, 2H), 4.59 (s, 2H), 3.85 (s, 3H), 2.74 (s, 3H). MS (ESI+)
(m/z) 492.92, 494.92 ([M + H]+).
tert-Butyl N-(4-Amino-3-chloro-5-methylbenzyl)-N′-(3-(4-me-
t h o x y p h e n y l ) - 5 - m e t h y l i s o t h i a z o l e - 4 - c a r b o n y l ) -
carbamimidoylcarbamate (16b). Compound 16b (1.97 g) was
prepared by analogy to 16a, substituting 4-(aminomethyl)-2-chloro-6-
methylaniline for 3,5-dichloro-4-aminobenzylamine to afford 1.94 g
N-(Amino((3,5-dimethylbenzyl)amino)methylene)-3-(4-methoxy-
phenyl)-5-methylisothiazole-4-carboxamide (8s). Compound 8s was
prepared from 3 via method A using 3-(4-methoxyphenyl)-5-
methylisothiazole-4-carboxylic acid as acid 4 and 3,5-dimethylbenzyl
amine as amine 6. Preparative HPLC purification afforded 2.7 mg
(15% yield from 3) of 8s as a white solid. 1H NMR (500 MHz,
CD3OD) δ 7.60 (d, J = 8.5 Hz, 2H), 7.03−6.98 (m, 3H), 6.94 (br s,
2H), 4.47 (br. s., 2H), 3.83 (s, 3H), 2.71 (d, J = 0.9 Hz, 3H), 2.32 (s,
6H). MS (ESI+) (m/z) 409.2 ([M + H]+).
General Procedure for Solution Phase Synthesis of Acyl
Guanidines (Method B). tert-Butyl Methylthiocarbonoimidoylcar-
bamate (12). To a vigorously stirred suspension of S-methylisothiour-
ea hemisulfate (60.8 g, 0.437 mol) in CH2Cl2 (600 mL) was added 2N
NaOH (300 mL, 0.6 mol). This was cooled to 0 °C on an ice bath, and
a solution of di-tert-butyl dicarbonate (43.2 g, 0.198 mol) was added
dropwise over 6 h. Upon completion of the addition, the mixture was
stirred an additional 20 min and diluted with 1 L CH2Cl2, and the
phases were separated. The organic portion was washed with water (2
1
(79% yield) of 16b as white solid. H NMR (CDCl3, 500 MHz) δ
12.09 (s, 1H), 8.53 (s, 1H), 7.52 (d, J = 8.6 Hz, 2H), 6.93 (s, 1H), 6.89
(d, J = 8.6 Hz, 2H), 6.74 (s, 1H), 4.00 (s, 2H), 3.92 (d, J = 5.5 Hz,
2H), 3.77 (s, 3H), 2.70 (s, 3H), 2.16 (s, 3H), 1.48 (s, 9H). MS (ESI+)
(m/z) 544 ([M + H]+).
N-(Amino((3,5-dichloro-4-hydroxybenzyl)amino)methylene)-3-
(4-methoxyphenyl)-5-methylisothiazole-4-carboxamide (17a).
Compound 17a was prepared from 3 via method A using 3-(4-
methoxyphenyl)-5-methylisothiazole-4-carboxylic acid as acid 4 and
3,5-dichloro-4-hydroxybenzylamine as amine 6. Preparative HPLC
purification afforded 5.6 mg (10% yield from 3) of 17a as a white solid.
1H NMR (500 MHz, CD3OD) δ 7.62 (d, J = 7.8 Hz, 2H), 7.42 (s,
2H), 7.07 (d, J = 7.8 Hz, 2H), 4.46 (s, 2H), 3.86 (s, 3H), 2.68 (s, 3H).
MS (ESI+) (m/z) 449.1 ([M + H]+).
N-(Amino((3,5-dichloro-4-propoxybenzyl)amino)methylene)-3-
(4-methoxyphenyl)-5-methylisothiazole-4-carboxamide (17b).
9218
dx.doi.org/10.1021/jm300931y | J. Med. Chem. 2012, 55, 9208−9223

Journal of Medicinal Chemistry
Article
Compound 17b was prepared from 3 via method A using 3-(4-
methoxyphenyl)-5-methylisothiazole-4-carboxylic acid as acid 4 and
3,5-dichloro-4-propoxybenzylamine as amine 6. Preparative HPLC
purification afforded 20 mg (32% yield from 3) of 17b as a white solid.
1H NMR (500 MHz, CD3OD) δ 7.64 (d, J = 7.6 Hz, 2H), 7.42 (s,
2H), 7.07 (d, J = 7.9 Hz, 2H), 4.53 (s, 2H), 4.02 (t, J = 6.4 Hz, 2H),
3.87 (s, 3H), 2.69 (s, 3H), 1.84−1.90 (m, 2H), 1.12 (t, 3H, J = 7.5 Hz,
3H). MS (ESI+) (m/z) 491.1 ([M + H]+).
N-(Amino(4-amino-3,5-dichlorobenzylamino)methylene)-3-(4-
methoxyphenyl)-5-methylisoxazole-4-carboxamide (17c). Com-
pound 17c was prepared from 14a via method B using 3,5-dichloro-
4-aminobenzylamine as amine 15. Preparative HPLC purification
afforded 11 mg (47% yield from 14a) of 17c as a white solid. 1H NMR
(500 MHz, CD3OD) δ 7.60 (d, J = 8.8 Hz, 2H), 7.24 (s, 2H), 7.03 (d,
J = 8.6 Hz, 2H), 4.39 (s, 2H), 3.84 (s, 3H), 2.66 (s, 3H). MS (ESI+)
(m/z) 448 ([M + H]+).
N-(Amino(4-benzamido-3-chloro-5-methylbenzylamino)-
methylene)-3-(4-methoxyphenyl)-5-methylisothiazole-4-carboxa-
mide (17j). Compound 17j was prepared from 14b via method B
using 4-benzamido-3-chloro-5-methylbenzylamine as amine 15.
Preparative HPLC purification afforded 17.2 mg (67% yield from
14b) of 17j as a white powder. 1H NMR (400 MHz, CD3OD) δ 7.97−
8.03 (m, 2H), 7.59−7.67 (m, 3H), 7.51−7.58 (m, 2H), 7.36 (br s,
1H), 7.24 (br s, 1H), 7.00−7.07 (m, 2H), 4.58 (br s, 2H), 3.83 (s,
3H), 2.73 (s, 3H), 2.34 (s, 3H). MS (ESI+) (m/z) 548 ([M + H]+).
N-(Amino(3,5-dichloro-4-(2-(methylthio)acetamido)-
benzylamino)methylene)-3-(4-methoxyphenyl)-5-methylisoxazole-
4 -carb oxami de (1 7 k). S tep 1. N- ((4 -Amino -3 , 5 -
dichlorobenzylamino)(tert-butoxycarbonylamino)methylene)-3-(4-
methoxyphenyl)-5-methylisoxazole-4-carboxamide. A mixture of
14a (2.10 g, 5.18 mmol), 3,5-dichloro-4-aminobenzylamine (1.20 g,
6.28 mmol), and DIPEA (1.80 mL, 10.3 mmol) in DCM (40 mL) was
stirred at room temperature overnight. Water (200 mL) and DCM
(160 mL) were added to the reaction mixture, and the solution was
transferred to a separatory funnel. The two layers were separated, and
the aqueous solution was extracted with 150 mL of DCM. The
combined extract was dried over anhydrous sodium sulfate, then
filtered and concentrated in vacuo.The crude product was purified by
flash chromatography (DCM, Rf 0.25) to give 2.78 g (98% yield) of
N-(Amino(4-amino-3,5-dichlorobenzylamino)methylene)-3-(4-
methoxyphenyl)-5-methylisothiazole-4-carboxamide (17d). Com-
pound 17d was prepared from 14b via method B using 3,5-dichloro-4-
aminobenzylamine as amine 15. Preparative HPLC purification
1
afforded 13 mg (45% yield from 14b) of 17d as a white solid. H
NMR (500 MHz, CD3OD) δ 7.60 (d, J = 8.6 Hz, 2H), 7.22 (s, 2H),
7.00 (d, J = 8.8 Hz, 2H), 4.39 (s, 2H), 3.83 (s, 3H), 2.71 (s, 3H).
1
the title compound as a white solid. H NMR (500 MHz, CDCl3) δ
HRMS (ESI+): calculated for C20H20N5O2Cl2S, [M + H]+
=
12.18 (s, 1H), 8.67 (br s, 1H), 7.53 (d, J = 8.5 Hz, 2H), 6.96 (s, 2H),
6.90 (d, J = 8.2 Hz, 2H), 4.45 (br s, 2H), 4.03 (d, J = 5.8 Hz, 2H), 3.77
(s, 3H), 2.73 (s, 3H), 1.49 (s, 9H). MS (ESI+) (m/z) 548 ([M +
H]+).
464.0709; observed, [M + H]+ = 464.0697.
N-((4-Acetamido-3,5-dichlorobenzylamino)(amino)methylene)-
3-(4-methoxyphenyl)-5-methylisoxazole-4-carboxamide (17e).
Compound 17e was prepared from 14a via method B using 3,5-
dichloro-4-acetamidobenzylamine as amine 15. Preparative HPLC
purification afforded 12.1 mg (78% yield from 14a) of 17e as a white
solid. 1H NMR (400 MHz, CDCl3) δ 7.50 (d, J = 7.8 Hz, 2H), 7.22 (s,
2H), 6.92 (d, J = 8.3 Hz, 2H), 4.42 (s, 2H), 3.79 (s, 3H), 2.69 (s, 3H),
2.18 (s, 3H). MS (ESI+) (m/z) 491 ([M + H]+).
N-((4-Acetamido-3,5-dichlorobenzylamino)(amino)methylene)-
3-(4-methoxyphenyl)-5-methylisothiazole-4-carboxamide (17f).
Compound 17f was prepared from 14b via method B using 3,5-
dichloro-4-acetamidobenzylamine as amine 15. Preparative HPLC
purification afforded 64.5 mg (84% yield from 14b) of 17f as a white
solid. 1H NMR (500 MHz, CD3OD) δ 7.60 (d, J = 8.6 Hz, 2H), 7.43
(s, 2H), 7.02 (d, J = 8.8 Hz, 2H), 4.39 (s, 2H), 3.83 (s, 3H), 2.72 (s,
3H), 2.19 (s, 3H). HRMS (ESI+): calculated for C22H22N5O3Cl2S, [M
+ H]+ = 506.0815; observed, [M + H]+ = 506.0805.
N-((4-Acetamido-3-chloro-5-methylbenzylamino)(amino)-
methylene)-3-(4-methoxyphenyl)-5-methylisothiazole-4-carboxa-
mide (17g). Compound 17g was prepared from 14b via method B
using N-(4-(aminomethyl)-2-chloro-6-methylphenyl)acetamide as
amine 15. Preparative HPLC purification afforded 17.4 mg (72%
yield from 3) of 17g as a white solid. 1H NMR (500 MHz, CD3OD) δ
7.60 (d, J = 8.5 Hz, 2H), 7.30 (br s, 1H), 7.17 (br s, 1H), 7.01 (d, J =
8.9 Hz, 2H), 4.53 (br s, 2H), 3.83 (s, 3H), 2.71 (s, 3H), 2.27 (s, 3H),
2.18 (s, 3H). HRMS (ESI+): calculated for C23H25N5O3ClS, [M + H]+
= 486.1361; observed, [M + H]+ = 486.1349.
N-(Amino(3,5-dichloro-4-(methylsulfonamido)benzylamino)-
methylene)-3-(4-methoxyphenyl)-5-methylisothiazole-4-carboxa-
mide (17h). Compound 17h was prepared from 14b via method B
using 3,5-dichloro-4-(methylsulfonamido)benzylamine as amine 15.
Preparative HPLC purification afforded 11.8 mg (27% yield from 14b)
of 17h as a white powder. 1H NMR (500 MHz, CD3OD) δ 7.60 (d, J
= 8.5 Hz, 2H), 7.45 (br s, 2H), 7.02 (d, J = 8.5 Hz, 2H), 4.57 (br s,
2H), 3.83 (d, J = 3.4 Hz, 3H), 3.23 (s, 3H), 2.70−2.74 (m, 3H). MS
(ESI+) (m/z) 542 ([M + H]+).
Step 2. N-((4-(2-Bromoacetamido)-3,5-dichlorobenzylamino)-
(tert-butoxycarbonylamino)methylene)-3-(4-methoxyphenyl)-5-
methylisoxazole-4-carboxamide (isoxazole analogue of 20a). N-
((4-Amino-3,5-dichlorobenzylamino)(tert-butoxycarbonylamino)-
methylene)-3-(4-methoxyphenyl)-5-methylisoxazole-4-carboxamide
(1.23 g, 2.24 mmol) was dissolved in 20 mL of anhydrous DCM in a
250 mL round-bottom flask, followed by the addition of 0.343 mL
(2.46 mmol, 1.1 equiv) of triethylamine. The solution was cooled to 0
°C, and nitrogen was introduced. Bromoacetyl bromide (0.214 mL,
2.46 mmol, 1.1 equiv) was added dropwise, and the reaction mixture
was stirred at 0 °C for 1 h. More bromoacetyl bromide (0.100 mL)
was added, and the reaction mixture was stirred at 0 °C for 1 h. DCM
(80 mL) and saturated NaHCO3 aqueous solution (100 mL) were
added, and the layers were separated. The aqueous was extracted with
DCM (50 mL). The combined organic extracts were dried over
anhydrous sodium sulfate. The crude product was purified by flash
chromatography (EtOAc/DCM, 5:95, Rf 0.34) to give 1.43 g (96%
1
yield) of the title compound as a white solid. H NMR (500 MHz,
CDCl3) δ 12.20 (s, 1H), 8.84 (t, J = 5.6 Hz, 1H), 7.91 (br s, 1H),
7.53−7.46 (m, 2H), 7.11 (s, 2H), 6.88 (d, J = 8.9 Hz, 2H), 4.11 (d, J =
5.8 Hz, 2H), 4.09 (s, 2H), 3.79 (s, 3H), 2.70 (s, 3H), 1.51 (s, 9H). MS
(ESI+) (m/z) 668 ([M + H]+).
Step 3. N-((tert-butoxycarbonylamino)(3,5-dichloro-4-(2-
(methylthio)acetamido)benzylamino)methylene)-3-(4-methoxy-
phenyl)-5-methylisoxazole-4-carboxamide (BOC-17k). The mixture
of 147 mg (0.506 mmol) of 1,3-di-Boc-2-methylisothiourea, 24 mg
(0.406 mmol) of n-propylamine, and 0.088 mL (0.051 mmol) of
DIPEA in 1.5 mL of DCM was stirred at room temperature overnight
to generate methanethiolate in situ. N-((4-(2-Bromoacetamido)-3,5-
dichlorobenzylamino)(tert-butoxycarbonylamino)methylene)-3-(4-
methoxyphenyl)-5-methylisoxazole-4-carboxamide (100 mg, 0.149
mmol) was added to the reaction mixture, which was stirred at
room temperature for 16 h. The reaction mixture was washed with
water and solvent was evaporated in vacuo to give the crude product of
N-((tert-butoxycarbonylamino)(3,5-dichloro-4-(2-(methylthio)-
acetamido)benzylamino)methylene)-3-(4-methoxyphenyl)-5-methyli-
soxazole-4-carboxamide (BOC-17k). To half of the crude product was
added 1.0 mL of 50% TFA in DCM, and the resulting solution was
stirred at RT for 1 h. The solution was concentrated in vacuo, and the
product was purified by preparative HPLC (0.1% TFA, MeOH/water)
N-(Amino(3-chloro-5-methyl-4-propionamidobenzylamino)-
methylene)-3-(4-methoxyphenyl)-5-methylisoxazole-4-carboxa-
mide (17i). Compound 17i was prepared from 14a via method B using
3-chloro-5-methyl-4-propionamidobenzylamine as amine 15. Prepara-
tive HPLC purification afforded 11.1 mg (46% yield from 14a) of 17i
as a white solid. 1H NMR (400 MHz, CD3OD) δ 7.61 (d, J = 8.8 Hz,
2H), 7.33 (s, 1H), 7.21 (s, 1H), 7.04 (d, J = 8.8 Hz, 2H), 4.53 (s, 2H),
3.84 (s, 3H), 2.67 (s, 3H), 2.46 (q, J = 7.5 Hz, 2H), 2.26 (s, 3H), 1.26
(t, 3H). MS (ESI+) (m/z) 484 ([M + H]+).
1
to give 15 mg (36% yield) of title compound. H NMR (500 MHz,
CD3OD) δ 7.61 (d, J = 8.2 Hz, 2H), 7.48 (br s, 2H), 7.04 (d, J = 8.2
9219
dx.doi.org/10.1021/jm300931y | J. Med. Chem. 2012, 55, 9208−9223

Journal of Medicinal Chemistry
Article
Hz, 2H), 4.58 (br s, 2H), 3.84 (s, 3H), 3.38 (s, 2H), 2.67 (s, 3H), 2.28
(s, 3H). MS (ESI+) (m/z) 536 ([M + H]+).
N-(Amino(3-chloro-4-(2-(dimethylamino)acetamido)-5-
methylbenzylamino)methylene)-3-(4-methoxyphenyl)-5-methyli-
soxazole-4-carboxamide (17r). Compound 17r was prepared from
14a via method B using 3-chloro-4-(2-(dimethylamino)acetamido)-5-
methylbenzylamine as amine 15. Preparative HPLC purification
N-(Amino(3-chloro-5-methyl-4-(3-(methylthio)propanamido)-
benzylamino)methylene)-3-(4-methoxyphenyl)-5-methylisoxazole-
4-carboxamide (17l). Compound 17l was prepared from 14a via
method B using 3-chloro-5-methyl-4-(3-(methylthio)propanamido)-
benzylamine as amine 15. Preparative HPLC purification afforded 6.1
mg (24% yield from 14a) of 17l as a white solid. 1H NMR (400 MHz,
CD3OD) δ 7.62 (d, J = 8.8 Hz, 2H), 7.34 (br s, 1H), 7.21 (br s, 1H),
7.05 (d, J = 8.8 Hz, 2H), 4.53 (br s, 2H), 3.85 (s, 3H), 2.89−2.83 (m,
2H), 2.78−2.72 (m, 2H), 2.68 (s, 3H), 2.30 (s, 3H), 2.17 (s, 3H). MS
(ESI+) (m/z) 530 ([M + H]+).
1
afforded 16.8 mg (67% yield from 14a) of 17r as a white solid. H
NMR (400 MHz, CD3OD) δ 7.59 (d, J = 9.0 Hz, 2H), 7.38 (s, 1H),
7.26 (s, 1H), 7.02 (d, J = 8.8 Hz, 2H), 4.55 (s, 2H), 4.29 (s, 2H), 3.83
(s, 3H), 3.01 (s, 6H), 2.66 (s, 3H), 2.30 (s, 3H). MS (ESI+) (m/z)
513 ([M + H]+).
N-(Amino(3-chloro-5-(thiophen-3-yl)benzylamino)methylene)-3-
(4-methoxyphenyl)-5-methylisoxazole-4-carboxamide (19a). Com-
pound 8r (19 mg, 40 μmol) was heated with 3-thiophenylboronic acid
(15 mg), tetrakis(triphenylphosphine)palladium (5 mg), and aqueous
tripotassium phosphate (2 M, 0.1 mL) in DMF (1.5 mL, degas with
N2) at 85 °C for 15 h. The crude mixture was purified by preparative
N-(Amino(3,5-dichloro-4-(2-(methylsulfonyl)acetamido)-
benzylamino)methylene)-3-(4-methoxyphenyl)-5-methylisoxazole-
4-carboxamide (17m). N-((tert-Butoxycarbonylamino)(3,5-dichloro-
4-(2-(methylthio)acetamido)benzylamino)methylene)-3-(4-methoxy-
phenyl)-5-methylisoxazole-4-carboxamide (BOC-17k) (64 mg of
crude product) and 47 mg of mCPBA in 2 mL of DCM was stirred
at room temperature overnight. Solvent was evaporated, and the
residue was treated with 1.0 mL of 50% TFA in DCM at room
temperature for 1 h. Solvent and TFA were evaporated in vacuo. The
product was purified by preparative HPLC (0.1% TFA, MeOH/water)
1
HPLC to afford 7.5 mg of 19a (30% yield from 8r). H NMR (500
MHz, CD3OD) δ 7.75 (dd, J1 = 1.46 Hz, J2 = 2.93 Hz, 1H), 7.70 (dd,
J1 = 1.46 Hz, J2 = 1.83 Hz, 1H), 7.62 (m, 3H), 7.53 (dd, J1 = 2.93 Hz,
J2 = 5.12 Hz, 1H), 7.48 (dd, J1 = 1.46 Hz, J2 = 5.12 Hz, 1H), 7.31 (s,
1H), 7.02 (d, J = 8.8 Hz, 2H), 5.36 (d, J = 11 Hz, 1H), 4.61 (s, 2H),
3.82 (s, 3H), 2.67 (s, 3H). MS (ESI+) (m/z) 481 ([M + H]+).
N-(Amino(3-chloro-5-(pyridin-3-yl)benzylamino)methylene)-3-(4-
methoxyphenyl)-5-methylisoxazole-4-carboxamide (19b). Com-
pound 19b was prepared by analogy to 19a, substituting 3-
pyridylboronic acid for 3-thiophenylboronic acid to afford 12.4 mg
1
to give 6.4 mg (15% yield) of title compound. H NMR (500 MHz,
CD3OD) δ 7.62 (d, J = 8.9 Hz, 2H), 7.49 (br s, 2H), 7.05 (d, J = 8.9
Hz, 2H), 4.59 (br s, 2H), 4.31 (s, 2H), 3.84 (s, 3H), 3.19 (s, 3H), 2.68
(s, 3H). MS (ESI+) (m/z) 568 ([M + H]+).
1
N-(Amino(3,5-dichloro-4-(cyclopropanecarboxamido)-
benzylamino)methylene)-3-(4-methoxyphenyl)-5-methylisoxazole-
4-carboxamide (17n). Compound 17n was prepared from 14a via
method B using 3,5-dichloro-4-(cyclopropanecarboxamido)-
benzylamine as amine 15. Preparative HPLC purification afforded
23.6 mg (37% yield from 14a) of 17n as a white solid. 1H NMR (500
MHz, CD3OD) δ 7.61 (d, J = 8.9 Hz, 2H), 7.46 (br s, 2H), 7.04 (d, J =
8.9 Hz, 2H), 4.57 (br s, 2H), 3.84 (s, 3H), 2.67 (s, 3H), 1.93−1.83
(m,1H), 1.01−0.86 (m, 4H). MS (ESI+) (m/z) 516 ([M + H]+).
N-(Amino(3-chloro-4-(cyclopropanecarboxamido)-5-
methylbenzylamino)methylene)-3-(4-methoxyphenyl)-5-methyli-
soxazole-4-carboxamide (17o). Compound 17o was prepared from
14a via method B using 3-chloro-4-(cyclopropanecarboxamido)-5-
methylbenzylamine as amine 15. Preparative HPLC purification
(42% yield from 8r) of 19b as a white solid. H NMR (500 MHz,
CD3OD) δ 9.09 (s, 1H), 8.80 (d, J = 4.27 Hz, 1H), 8.65 (d, J = 8.24
Hz, 1H), 7.98 (dd, J1 = 5 Hz, J2 = 8.24 Hz, 1H), 7.84 (s, 1H), 7.71 (s,
1H), 7.58 (d, J = 8.8 Hz, 2H), 7.57 (s, 1H), 7.00 (d, J = 8.8 Hz, 2H),
5.36 (d, J = 11 Hz, 1H), 4.68 (s, 2H), 3.82 (s, 3H), 2.67 (s, 3H). MS
(ESI+) (m/z) 476 ([M + H]+).
N-(Amino(3-chloro-5-vinylbenzylamino)methylene)-3-(4-me-
thoxyphenyl)-5-methylisoxazole-4-carboxamide (19c). Compound
19c was prepared by analogy to 19a, substituting 4,4,5,5-tetramethyl-2-
vinyl-1,3,2-dioxaborolane for 3-thiophenylboronic acid to afford 7.5
mg (33% yield from 8r) of 19c as a white solid. 1H NMR (500 MHz,
CD3OD) δ 7.61 (d, J = 8.8 Hz, 2H), 7.46 (s, 1H), 7.36 (s, 1H), 7.29
(s, 1H), 7.04 (d, J = 8.8 Hz, 2H), 6.73 (dd, J1 = 17.4 Hz, J2 = 11 Hz,
1H), 5.86 (d, J = 17.4 Hz, 1H), 5.36 (d, J = 11 Hz, 1H), 4.56 (s, 2H),
3.84 (s, 3H), 2.67 (s, 3H). MS (ESI+) (m/z) 425 ([M + H]+).
N-(Amino(3-chloro-5-((E)-4-hydroxybut-1-enyl)benzylamino)-
methylene)-3-(4-methoxyphenyl)-5-methylisoxazole-4-carboxa-
mide (19d). Compound 19d was prepared by analogy to 19a,
substituting (E)-trimethyl((4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-
2-yl)but-3-en-1-yl)oxy)silane for 3-thiophenylboronic acid to afford 2
mg (8% yield from 8r) of 19d as a white solid. MS (ESI+) (m/z) 469
([M + H]+).
1
afforded 23.7 mg (97% yield from 14a) of 17o as a white solid. H
NMR (400 MHz, CD3OD) δ 7.61 (d, J = 9.0 Hz, 2H), 7.32 (br s, 1H),
7.19 (br s, 1H), 7.05 (d, J = 8.8 Hz, 2H), 4.53 (br s, 2H), 3.84 (s, 3H),
2.67 (s, 3H), 2.26 (s, 3H), 1.83−1.92 (m, 1H), 0.85−1.00 (m, 4H).
MS (ESI+) (m/z) 496 ([M + H]+).
N - ( A m i n o ( 3 - c h l o r o - 4 - ( f u r a n - 2 - c a r b o x a m i d o ) - 5 -
methylbenzylamino)methylene)-3-(4-methoxyphenyl)-5-methyli-
soxazole-4-carboxamide (17p). Compound 17p was prepared from
14a via method B using 3-chloro-4-(furan-2-carboxamido)-5-methyl-
benzylamine as amine 15. Preparative HPLC purification afforded 14.4
mg (57% yield from 14a) of 17p as a white solid. 1H NMR (400 MHz,
CD3OD) δ 7.76 (d, J = 1.0 Hz, 1H), 7.63 (d, J = 8.8 Hz, 2H), 7.38 (br
s, 1H), 7.22−7.30 (m, 2H), 7.03−7.09 (m, 2H), 6.67 (dd, J = 3.5, 1.8
Hz, 1H), 4.56 (br s, 2H), 3.85 (s, 3H), 2.68 (s, 3H), 2.32 (s, 3H). MS
(ESI+) (m/z) 522 ([M + H]+).
N-(Amino(3,5-dichloro-4-(2-(dimethylamino)acetamido)-
benzylamino)methylene)-3-(4-methoxyphenyl)-5-methylisoxazole-
4-carboxamide (17q). N-((4-(2-Bromoacetamido)-3,5-
dichlorobenzylamino)(tert-butoxycarbonylamino)methylene)-3-(4-
methoxyphenyl)-5-methylisoxazole-4-carboxamide (22 mg, 0.033
mmol) was dissolved in 2 mL of anhydrous DCM, followed by the
addition of dimethylamine (0.085 mL of 2.0 M solution in methanol,
0.17 mmol). The reaction mixture was stirred at room temperature
overnight, and then the reaction mixture was concentrated in vacuo.
To the residue was added 0.8 mL of 50% TFA in DCM, and the
resulting solution was stirred at room temperature for 1 h and then
concentrated in vacuo and purified by preparative HPLC (0.1% TFA,
MeOH/water) to afford 9.8 mg (39% yield) of 17q as a white solid.
1H NMR (500 MHz, CD3OD) δ 7.60 (d, J = 8.5 Hz, 2H), 7.52 (br s,
N-(Amino(3-chloro-5-((E)-pent-1-enyl)benzylamino)methylene)-
3-(4-methoxyphenyl)-5-methylisoxazole-4-carboxamide (19e).
Compound 19e was prepared by analogy to 19a, substituting (E)-
4,4,5,5-tetramethyl-2-(pent-1-en-1-yl)-1,3,2-dioxaborolane for 3-thio-
phenylboronic acid to afford 9 mg (37% yield from 8r) of 19e as a
1
white solid. H NMR (500 MHz, CD3OD) δ 7.62 (d, J = 8.24 Hz,
2H), 7.38 (s, 1H), 7.29 (s, 1H), 7.21 (s, 1H), 7.04 (d, J = 8.24 Hz,
2H), 6.39 (s, 2H), 4.54 (s, 2H), 3.84 (s, 3H), 2.67 (s, 3H), 2.22 (m,
2H), 1.51 (m, 2H), 0.97 (t, J = 7.32 Hz, 3H). MS (ESI+) (m/z) 467
([M + H]+).
N-(Amino(3-chloro-5-((E)-3-methoxyprop-1-enyl)benzylamino)-
methylene)-3-(4-methoxyphenyl)-5-methylisoxazole-4-carboxa-
mide (19f). Compound 19f was prepared by analogy to 19a,
substituting (E)-2-(3-methoxyprop-1-en-1-yl)-4,4,5,5-tetramethyl-
1,3,2-dioxaborolane for 3-thiophenylboronic acid to afford 8.1 mg
1
(33% yield from 8r) of 19f as a white solid. H NMR (500 MHz,
CD3OD) δ 7.62 (d, J = 8.5 Hz, 2H), 7.45 (s, 1H), 7.36 (s, 1H), 7.28
(s, 1H), 7.04 (d, J = 8.5 Hz, 2H), 6.64 (d, J = 14.7, 1H), 6.43 (m, 1H),
4.56 (s, 2H), 4.10 (d, J = 5.8 Hz, 2H), 3.85 (s, 3H), 3.39 (s, 3H), 2.68
(s, 3H). MS (ESI+) (m/z) 469 ([M + H]+).
N-((4-(2-Bromoacetamido)-3,5-dichlorobenzylamino)(tert-
butoxycarbonylamino)methylene)-3-(4-methoxyphenyl)-5-methyli-
sothiazole-4-carboxamide (20a). Intermediate 16a (1.00 g, 1.77
2H), 7.03 (d, J = 8.5 Hz, 2H), 4.60 (br s, 2H), 4.29 (s, 2H), 3.84 (s,
3H), 3.02 (s, 6H), 2.67 (s, 3H). MS (ESI+) (m/z) 533 ([M + H]+).
9220
dx.doi.org/10.1021/jm300931y | J. Med. Chem. 2012, 55, 9208−9223

Journal of Medicinal Chemistry
Article
mmol) was dissolved in 20 mL of anhydrous DCM in a 100 mL
round-bottom flask, followed by the addition of 0.272 mL (1.95 mmol,
1.1 equiv) of triethylamine. The solution was cooled to 0 °C, and
nitrogen was introduced. Bromoacetyl bromide (0.170 mL, 1.95
mmol, 1.1 equiv) was added dropwise, and the reaction mixture was
stirred at 0 °C for 1 h. More bromoacetyl bromide (another 0.120 mL)
was added and the reaction mixture was stirred at 0 °C for 1 h. DCM
(80 mL) and saturated NaHCO3 aqueous solution (100 mL) were
added and the layers were separated. The aqueous was extracted with
DCM (100 mL). The combined organic extracts were dried over
anhydrous sodium sulfate. The crude product was purified by flash
chromatography (EtOAc/DCM, 5:95) to give 1.18 g (98% yield) of
(br s, 2H), 7.00 (d, J = 8.9 Hz, 2H), 4.60 (br s, 2H), 4.12 (s, 2H), 3.82
(s, 3H), 3.18 (q, J = 7.3 Hz, 2H), 2.72 (s, 3H), 1.35 (t, J = 7.3 Hz,
3H). MS (ESI+) (m/z) 549 ([M + H]+).
N-(Amino(3-chloro-4-(2-(ethylamino)acetamido)-5-
methylbenzylamino)methylene)-3-(4-methoxyphenyl)-5-methyliso-
thiazole-4-carboxamide (22f). Compound 22f was prepared by
analogy to 22a, substituting 20b for 20a. Preparative HPLC
purification afforded 18.7 mg (65% yield from 20b) of 22d as a
white solid. 1H NMR (500 MHz, CD3OD) δ 7.58 (d, J = 8.9 Hz, 2H),
7.34 (br s, 1H), 7.22 (br s, 1H), 7.00 (d, J = 8.9 Hz, 2H), 4.55 (br s,
2H), 4.11 (s, 2H), 3.82 (s, 3H), 3.17 (q, J = 7.3 Hz, 2H), 2.71 (s, 3H),
2.31 (s, 3H), 1.36 (t, J = 7.3 Hz, 3H). MS (ESI+) (m/z) 529 ([M +
H]+).
1
20a. H NMR (500 MHz, CDCl3) δ 12.13 (s, 1H), 8.80 (br s, 1H),
N - ( A m i n o ( 4 - ( 2 - ( a z e t i d i n - 1 - y l ) a c e t a m i d o ) - 3 , 5 -
dichlorobenzylamino)methylene)-3-(4-methoxyphenyl)-5-methyli-
sothiazole-4-carboxamide (22g). Compound 22g was prepared by
analogy to 22a, substituting azetidine for dimethylamine. Preparative
HPLC purification afforded 6.4 mg (21% yield from 20a) of 22g as a
white solid. 1H NMR (500 MHz, CD3OD) δ 7.59 (d, J = 8.9 Hz, 2H),
7.47 (br s, 2H), 7.01 (d, J = 8.9 Hz, 2H), 4.59 (br s, 2H), 4.40 (s, 2H),
4.34 (br s, 2H), 4.19 (br s, 2H), 3.82 (s, 3H), 2.72 (s, 3H). MS (ESI+)
(m/z) 561 ([M + H]+).
7.96 (br s, 1H), 7.48 (d, J = 8.5 Hz, 2H), 7.12 (s, 2H), 6.86 (d, J = 8.9
Hz, 2H), 4.08 (s, 2H), 4.03 (d, J = 5.8 Hz, 2H), 3.80 (s, 3H), 2.68 (s,
3H), 1.53 (s, 9H). MS (ESI+) (m/z) 684 ([M + H]+).
N-((4-(2-Bromoacetamido)-3-chloro-5-methylbenzylamino)(tert-
butoxycarbonylamino)methylene)-3-(4-methoxyphenyl)-5-methyli-
sothiazole-4-carboxamide (20b). Compound 20b (1.15 g) was
prepared by analogy to compound 20a, substituting 16b for 16a to
afford 1.07 g (88% yield) of 20b as a white solid. 1H NMR (500 MHz,
CDCl3) δ 12.11 (s, 1H), 8.73 (s, 1H), 7.86 (s, 1H), 7.49 (d, J = 8.7,
2H), 7.05 (s, 1H), 6.91 (s, 1H), 6.87 (d, J = 8.7, 2H), 4.06 (s, 2H),
4.02 (d, J = 5.5 Hz, 2H), 3.77 (s, 3H), 2.67 (s, 3H), 2.24 (s, 3H), 1.50
(s, 9H). MS (ESI+) (m/z) 664 ([M + H]+).
N-(Amino(3,5-dichloro-4-(2-(dimethylamino)acetamido)-
benzylamino)methylene)-3-(4-methoxyphenyl)-5-methylisothia-
zole-4-carboxamide (22a). Compound 20a (26 mg, 0.038 mmol) was
dissolved in 2 mL of anhydrous DCM, followed by the addition of
dimethylamine (0.095 mL of 2.0 M methanol solution, 0.19 mmol)
and DIPEA (25 mg, 0.19 mmol). The reaction mixture was stirred at
room temperature overnight, and then the reaction mixture was
concentrated in vacuo. To the residue was added 0.8 mL of 50% TFA
in DCM, and the resulting solution was stirred at room temperature
for 1 h, then concentrated in vacuo and purified by preparative HPLC
(0.1% TFA, MeOH/water) to afford 14.5 mg (49% yield) of 22a as a
white solid. 1H NMR (500 MHz, CD3OD) δ 7.59 (d, J = 8.9 Hz, 2H),
7.49 (br s, 2H), 7.01 (d, J = 8.5 Hz, 2H), 4.60
N-(Amino(4-(2-(azetidin-1-yl)acetamido)-3-chloro-5-
methylbenzylamino)methylene)-3-(4-methoxyphenyl)-5-methyliso-
thiazole-4-carboxamide (22h). Step 1. Compound 20b (100 mg)
was added to 1.0 mL of 50% TFA in DCM. The resulting solution was
stirred at room temperature for 1 h. DCM and TFA were evaporated
in vacuo to give (E)-N-(amino(4-(2-bromoacetamido)-3-chloro-5-
methylbenzylamino)methylene)-3-(4-methoxyphenyl)-5-methyliso-
thiazole-4-carboxamide as a TFA salt, which was used without further
1
purification. H NMR (500 MHz, CD3OD) δ 7.58 (d, J = 8.6, 2H),
7.29 (s, 1H), 7.17 (s, 1H), 6.99 (d, J = 8.6, 2H), 4.52 (s, 2H), 4.04 (s,
2H), 3.81 (s, 3H), 2.70 (s, 3H), 2.28 (s, 3H). MS (ESI+) (m/z) 564
([M + H]+).
Step 2. To the crude product from Step 1 was added 4 mL of
anhydrous DCM, followed by the addition of azetidine (25 mg, 0.44
mmol). The resulted reaction mixture was stirred at room temperature
overnight. The residual solvent was evaporated in vacuo to give a
crude product, which was purified by preparative HPLC to yield 80.9
mg (72% from 20b) of 22h. 1H NMR (500 MHz, CDCl3) δ 8.79 (br s,
1H), 7.54 (d, J = 7.3 Hz, 2H), 6.97 (br s, 1H), 6.86 (d, J = 7.3 Hz,
2H), 6.82 (br s, 1H), 3.96 (br s, 2H), 3.80 (s, 3H), 3.48 (t, J = 7.0 Hz,
4H), 3.29 (s, 2H), 2.56 (s, 3H), 2.13−2.24 (m, 5H). HRMS (ESI+):
calculated for C26H30N6O3ClS, [M + H]+ = 541.1783; observed, [M +
H]+ = 541.1769.
N-(Amino(3-chloro-4-(2-(dimethylamino)acetamido)-5-
methylbenzylamino)methylene)-3-(4-methoxyphenyl)-5-methyliso-
thiazole-4-carboxamide (22b). Compound 22b was prepared by
analogy to 22a, substituting 20b for 20a. Preparative HPLC
purification afforded 47 mg (48% yield from 20b) of 22b as a white
1
powder. H NMR (500 MHz, CD3OD) δ 7.59 (d, J = 8.9 Hz, 2H),
7.35 (br s, 1H), 7.23 (br s, 1H), 7.00 (d, J = 8.9 Hz, 2H), 4.56 (br s,
2H), 4.28 (s, 2H), 3.82 (br s, 3H), 3.02 (s, 6H), 2.71 (s, 3H), 2.31 (s,
3H). MS (ESI+) (m/z) 529 ([M + H]+).
X-ray Crystallography. As previously reported,13 co-crystals of
BACE-1 were prepared using a Phe-statine peptide and used in the
following soaking experiments.
Compound 8k. The crystals were soaked with 2.5 mM 8k, 35%
PEG8K, 0.2 M ammonium sulfate, and 0.1 M sodium cacodylate at pH
6.2 for nine days. The resulting X-ray diffraction data indicated partial
displacement of the Phe-statine peptide, and only the 3,5-
dichlorobenzyl group of 8k could be observed.
Compound 17d. The crystals were soaked with 10 mM 17d, 18%
PEG5K MME, 0.2 M ammonium iodide, and 0.1 M sodium citrate at
pH 6.4 for 10 days. The resulting X-ray diffraction data indicated
complete displacement of the Phe-statine peptide, and the X-ray
structure was solved with a resolution of 2.65 Å. The coordinates for
the complex of 17d with BACE-1 have been deposited in the Protein
Data Bank (www.pdb.org) under the PDB code 4SFE.
Compound 17g. The crystals were soaked with 10 mM 17g, 18%
PEG5K MME, 0.2 M ammonium iodide, and 0.1 M sodium citrate at
pH 6.4 for 11 days. The resulting X-ray diffraction data indicated
complete displacement of the Phe-statine peptide, and the X-ray
structure was solved with a resolution of 2.5 Å. The coordinates for the
complex of 17g with BACE-1 have been deposited in the Protein Data
Bank (www.pdb.org) under the PDB code 4FSL.
N - ( A m i n o ( 4 - ( 2 - ( b e n z y l a m i n o ) a c e t a m i d o ) - 3 , 5 -
dichlorobenzylamino)methylene)-3-(4-methoxyphenyl)-5-methyli-
sothiazole-4-carboxamide (22c). Compound 22c was prepared by
analogy to 22a, substituting benzylamine for dimethylamine.
Preparative HPLC purification afforded 24.7 mg (72% yield from
20a) of 22b as a white solid. 1H NMR (500 MHz, CD3OD) δ 7.58 (d,
J = 8.9 Hz, 2H), 7.44−7.55 (m, 7H), 7.00 (d, J = 8.9 Hz, 2H), 4.60 (br
s, 2H), 4.32 (s, 2H), 4.11 (s, 2H), 3.82 (s, 3H), 2.72 (s, 3H). MS (ESI
+) (m/z) 611 ([M + H]+).
N-(Amino(4-(2-(benzylamino)acetamido)-3-chloro-5-
methylbenzylamino)methylene)-3-(4-methoxyphenyl)-5-methyliso-
thiazole-4-carboxamide (22d). Compound 22d was prepared by
analogy to 22a, substituting 20b for 20a. Preparative HPLC
purification afforded 22.0 mg (71% yield from 20b) of 22d as a
white solid. 1H NMR (500 MHz, CD3OD) δ 7.58 (d, J = 8.5 Hz, 2H),
7.46−7.55 (m, 5H), 7.34 (br s, 1H), 7.22 (br s, 1H), 7.00 (d, J = 8.9
Hz, 2H), 4.55 (br s, 2H), 4.32 (s, 2H), 4.10 (s, 2H), 3.82 (s, 3H), 2.71
(s, 3H), 2.29 (s, 3H). MS (ESI+) (m/z) 591 ([M + H]+).
N-(Amino(3,5-dichloro-4-(2-(ethylamino)acetamido)-
benzylamino)methylene)-3-(4-methoxyphenyl)-5-methylisothia-
zole-4-carboxamide (22e). Compound 22e was prepared by analogy
to 22a, substituting ethylamine for dimethylamine. Preparative HPLC
purification afforded 19.1 mg (68% yield from 20a) of 22e as a white
solid. 1H NMR (500 MHz, CD3OD) δ 7.59 (d, J = 8.5 Hz, 2H), 7.48
Biological Assays. Radioligand binding assays were conducted as
described in ref 7. HEK-Sw and KBV-Sw cellular assays were
conducted as described in ref 18.
9221
dx.doi.org/10.1021/jm300931y | J. Med. Chem. 2012, 55, 9208−9223

Journal of Medicinal Chemistry
Article
Rat Studies. Rat studies were performed as described in refs 13
and 18.
(9) Hopkins, A. L.; Groom, C. R.; Alex, A. Ligand efficiency: a useful
metric for lead selection. Drug Discovery Today 2004, 9, 430−431.
(10) Dodd, D. S.; Zhao, Y. Solid-phase synthesis of N,N′-substituted
acylguanidines. Tetrahedron Lett. 2001, 42, 1259−1262.
ASSOCIATED CONTENT
* Supporting Information
■
(11) Wang, S. S. p-Alkoxybenzyl alcohol resin and p-alkoxybenzylox-
ycarbonylhydrazide resin for solid phase synthesis of protected peptide
fragments. J. Am. Chem. Soc. 1973, 95, 1328−1333.
S
Additional SAR tables and experimental procedures for
noncommercial benzylamines and compounds in the tables.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(12) McGregor, D. N.; Corbin, U.; Swigor, J. E.; Cheney, L. C.
Synthesis of isothiazoles: the transformation of isoxazoles into
isothiazoles. Tetrahedron 1969, 25, 389−395.
(13) Thompson, L. A.; Shi, J.; Decicco, C. P.; Tebben, A. J.; Olson, R.
E.; Boy, K. M.; Guernon, J. M.; Good, A. C.; Liauw, A.; Zheng, C.;
Copeland, R. A.; Combs, A. P.; Trainor, G. L.; Camac, D. M.;
Muckelbauer, J. K.; Lentz, K. A.; Grace, J. E.; Burton, C. R.; Toyn, J.
H.; Barten, D. M.; Marcinkeviciene, J.; Meredith, J. E.; Albright, C. F.;
Macor, J. E. Synthesis and in vivo evaluation of cyclic diaminopropane
BACE-1 inhibitors. Bioorg. Med. Chem. Lett. 2011, 21, 6909−6915.
(14) Marcin, L. R.; Higgins, M. A.; Zusi, F. C.; Zhang, Y.; Dee, M. F.;
Parker, M. F.; Muckelbauer, J. K.; Camac, D. M.; Morin, P. E.;
Ramamurthy, V.; Tebben, A. J.; Lentz, K. A.; Grace, J. E.;
Marcinkeviciene, J. A.; Kopcho, L. M.; Burton, C. R.; Barten, D. M.;
Toyn, J. H.; Meredith, J. E.; Albright, C. F.; Bronson, J. J.; Macor, J. E.;
Thompson, L. A. Synthesis and SAR of indole- and 7-azaindole-1,3-
dicarboxamide hydroxyethylamine inhibitors of BACE-1. Bioorg. Med.
Chem. Lett. 2011, 21, 537−541.
Accession Codes
X-ray structures of 17d (4SFE) and 17g (4FSL) bound to
BACE-1 have been deposited in the PDB (www.pdb.org).
AUTHOR INFORMATION
Corresponding Author
*Phone: 203-677-6884. Fax: 203-677-7884. Email: samuel.
gerritz@bms.com.
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We acknowledge Hong Huang and Ivar MacDonald for their
synthetic contributions, the Bioanalytical Research Group for
the analysis of plasma, brain, and CSF drug levels, Rushith
Anumula at Biocon Bristol-Myers Squibb R&D Center
(Bangalore, India) for X-ray crystal structure refinements,
Dieter Drexler for performing HRMS analyses, and Ramkumar
Rajamani, Ken Boy, and Larry Marcin for helpful discussions.
(15) Stachel, S. J.; Coburn, C. A.; Steele, T. G.; Crouthamel, M. C.;
Pietrak, B. L.; Lai, M. T.; Holloway, M. K.; Munshi, S. K.; Graham, S.
L.; Vacca, J. P. Conformationally biased P3 amide replacements of
beta-secretase inhibitors. Bioorg. Med. Chem. Lett. 2006, 16, 641−644.
(16) The PyMOL Molecular Graphics System, version 1.3r1;
Schrodinger LLC, 2010.
(17) Chenguang, Y.; Williams, R. M. An efficient method for the
preparation of amidinoureas. Tetrahedron Lett. 1996, 37, 1945−1948.
(18) Meredith, J. E., Jr.; Thompson, L. A.; Toyn, J. H.; Marcin, L.;
Barten, D. M.; Marcinkeviciene, J.; Kopcho, L.; Kim, Y.; Lin, A.; Guss,
V.; Burton, C.; Iben, L.; Polson, C.; Cantone, J.; Ford, M.; Drexler, D.;
Fiedler, T.; Lentz, K. A.; Grace, J. E., Jr.; Kolb, J.; Corsa, J.;
Pierdomenico, M.; Jones, K.; Olson, R. E.; Macor, J. E.; Albright, C. F.
P-Glycoprotein efflux and other factors limit brain amyloid beta
reduction by beta-site amyloid precursor protein-cleaving enzyme 1
inhibitors in mice. J. Pharmacol. Exp. Ther. 2008, 326, 502−513.
(19) Stachel, S. J.; Coburn, C. A.; Rush, D.; Jones, K. L.; Zhu, H.;
Rajapakse, H.; Graham, S. L.; Simon, A.; Katharine Holloway, M.;
Allison, T. J.; Munshi, S. K.; Espeseth, A. S.; Zuck, P.; Colussi, D.;
Wolfe, A.; Pietrak, B. L.; Lai, M. T.; Vacca, J. P. Discovery of
aminoheterocycles as a novel beta-secretase inhibitor class: pH
dependence on binding activity part 1. Bioorg. Med. Chem. Lett.
2009, 19, 2977−2980.
(20) Stachel, S. J.; Steele, T. G.; Petrocchi, A.; Haugabook, S. J.;
McGaughey, G.; Katharine Holloway, M.; Allison, T.; Munshi, S.;
Zuck, P.; Colussi, D.; Tugasheva, K.; Wolfe, A.; Graham, S. L.; Vacca,
J. P. Discovery of pyrrolidine-based beta-secretase inhibitors: lead
advancement through conformational design for maintenance of ligand
binding efficiency. Bioorg. Med. Chem. Lett. 2012, 22, 240−244.
(21) Weaver, C. D.; Harden, D.; Dworetzky, S. I.; Robertson, B.;
Knox, R. J. A thallium-sensitive, fluorescence-based assay for detecting
and characterizing potassium channel modulators in mammalian cells.
J. Biomol. Screening 2004, 9, 671−677.
(22) Fox, E.; Bates, S. E. Tariquidar (XR9576): a P-glycoprotein drug
efflux pump inhibitor. Expert Rev. Anticancer Ther. 2007, 7, 447−459.
(23) Cole, D. C.; Bursavich, M. G. Nonpeptide BACE1 Inhibitors:
Design and Synthesis. In Aspartic Acid Proteases as Therapeutic Targets;
Wiley-VCH Verlag GmbH & Co. KGaA: New York, 2010; pp 481−
509.
(24) 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 acylguanidines as highly potent and selective beta-site amyloid
ABBREVIATIONS USED
■
ABT, aminobenzotriazole; CYP P450, cytochrome P450;
DIEA, diisopropylethylamine; DCM, dichloromethane; DIEA,
diisopropylethylamine; DMAP, 4-N,N-dimethylaminopyridine;
DMF, N,N-dimethyl formamide; EDC, 1-(3-dimethylamino-
propyl)-3-ethylcarbodiimide hydrochloride; EtOH, ethanol;
HOBT, 1-hydroxybenzotrizole; MeOH, methanol; NMP, N-
methylpyrrolidinone; PyAOP, (7-azabenzotriazol-1-yloxy)-
tripyrrolidinophosphonium hexafluorophosphate; PS, polystyr-
ene; TFA, trifluoroacetic acid; THF, tetrahydrofuran
REFERENCES
■
(1) Querfurth, H. W.; LaFerla, F. M. Alzheimer’s disease. N. Engl. J.
Med. 2010, 362, 329−344.
(2) World Alzheimer Report 2011; Alzheimer’s Disease International;
http://www.alz.co.uk (accessed June 28, 2012).
(3) Walsh, D. M.; Selkoe, D. J. Ab oligomersa decade of discovery.
J. Neurochem. 2007, 101, 1172−1184.
(4) Citron, M. Alzheimer’s disease: strategies for disease
modification. Nature Rev. Drug Discovery 2010, 9, 387−398.
(5) McConlogue, L.; Buttini, M.; Anderson, J. P.; Brigham, E. F.;
Chen, K. S.; Freedman, S. B.; Games, D.; Johnson-Wood, K.; Lee, M.;
Zeller, M.; Liu, W.; Motter, R.; Sinha, S. Partial reduction of BACE1
has dramatic effects on Alzheimer plaque and synaptic pathology in
APP Transgenic Mice. J. Biol. Chem. 2007, 282, 26326−26334.
(6) Ghosh, A. K.; Shin, D.; Downs, D.; Koelsch, G.; Lin, X.;
Ermolieff, J.; Tang, J. Design of Potent Inhibitors for Human Brain
Memapsin 2 (beta-Secretase). J. Am. Chem. Soc. 2000, 122, 3522−
3523.
(7) Iben, L. G.; Kopcho, L.; Marcinkeviciene, J.; Zheng, C.;
Thompson, L. A.; Albright, C. F.; Toyn, J. H. [3H]BMS-599240a
novel tritiated ligand for the characterization of BACE1 inhibitors. Eur.
J. Pharmacol. 2008, 593, 10−15.
(8) Kuntz, I. D.; Chen, K.; Sharp, K. A.; Kollman, P. A. The maximal
affinity of ligands. Proc. Natl. Acad. Sci. U. S. A. 1999, 96, 9997−10002.
9222
dx.doi.org/10.1021/jm300931y | J. Med. Chem. 2012, 55, 9208−9223

Journal of Medicinal Chemistry
Article
cleaving enzyme (BACE) inhibitors: Part Iinhibitor design and
validation. J. Med. Chem. 2010, 53, 951−965.
(25) Cole, D. C.; Manas, E. S.; Stock, J. R.; Condon, J. S.; Jennings,
L. D.; Aulabaugh, A.; Chopra, R.; Cowling, R.; Ellingboe, J. W.; Fan, K.
Y.; Harrison, B. L.; Hu, Y.; Jacobsen, S.; Jin, G.; Lin, L.; Lovering, F. E.;
Malamas, M. S.; Stahl, M. L.; Strand, J.; Sukhdeo, M. N.; Svenson, K.;
Turner, M. J.; Wagner, E.; Wu, J.; Zhou, P.; Bard, J. Acylguanidines as
small-molecule beta-secretase inhibitors. J. Med. Chem. 2006, 49,
6158−6161.
(26) Fobare, W. F.; Solvibile, W. R.; Robichaud, A. J.; Malamas, M.
S.; Manas, E.; Turner, J.; Hu, Y.; Wagner, E.; Chopra, R.; Cowling, R.;
Jin, G.; Bard, J. Thiophene substituted acylguanidines as BACE1
inhibitors. Bioorg. Med. Chem. Lett. 2007, 17, 5353−5356.
(27) Cole, D. C.; Stock, J. R.; Chopra, R.; Cowling, R.; Ellingboe, J.
W.; Fan, K. Y.; Harrison, B. L.; Hu, Y.; Jacobsen, S.; Jennings, L. D.;
Jin, G.; Lohse, P. A.; Malamas, M. S.; Manas, E. S.; Moore, W. J.;
O’Donnell, M. M.; Olland, A. M.; Robichaud, A. J.; Svenson, K.; Wu,
J.; Wagner, E.; Bard, J. Acylguanidine inhibitors of beta-secretase:
optimization of the pyrrole ring substituents extending into the S1 and
S3 substrate binding pockets. Bioorg. Med. Chem. Lett. 2008, 18, 1063−
1066.
(28) Malamas, M. S.; Barnes, K.; Johnson, M.; Hui, Y.; Zhou, P.;
Turner, J.; Hu, Y.; Wagner, E.; Fan, K.; Chopra, R.; Olland, A.; Bard,
J.; Pangalos, M.; Reinhart, P.; Robichaud, A. J. Di-substituted pyridinyl
aminohydantoins as potent and highly selective human beta-secretase
(BACE1) inhibitors. Bioorg. Med. Chem. 2010, 18, 630−639.
(29) Malamas, M. S.; Erdei, J.; Gunawan, I.; Turner, J.; Hu, Y.;
Wagner, E.; Fan, K.; Chopra, R.; Olland, A.; Bard, J.; Jacobsen, S.;
Magolda, R. L.; Pangalos, M.; Robichaud, A. J. Design and synthesis of
5,5′-disubstituted aminohydantoins as potent and selective human
beta-secretase (BACE1) inhibitors. J. Med. Chem. 2010, 53, 1146−
1158.
(30) Malamas, M. S.; Erdei, J.; Gunawan, I.; Barnes, K.; Johnson, M.;
Hui, Y.; Turner, J.; Hu, Y.; Wagner, E.; Fan, K.; Olland, A.; Bard, J.;
Robichaud, A. J. Aminoimidazoles as potent and selective human beta-
secretase (BACE1) inhibitors. J. Med. Chem. 2009, 52, 6314−6323.
(31) Baxter, E. W.; Conway, K. A.; Kennis, L.; Bischoff, F.; Mercken,
M. H.; Winter, H. L.; Reynolds, C. H.; Tounge, B. A.; Luo, C.; Scott,
M. K.; Huang, Y.; Braeken, M.; Pieters, S. M.; Berthelot, D. J.; Masure,
S.; Bruinzeel, W. D.; Jordan, A. D.; Parker, M. H.; Boyd, R. E.; Qu, J.;
Alexander, R. S.; Brenneman, D. E.; Reitz, A. B. 2-Amino-3,4-
dihydroquinazolines as inhibitors of BACE-1 (beta-site APP cleaving
enzyme): Use of structure based design to convert a micromolar hit
into a nanomolar lead. J. Med. Chem. 2007, 50, 4261−4264.
(32) Barrow, J. C.; Stauffer, S. R.; Rittle, K. E.; Ngo, P. L.; Yang, Z.;
Selnick, H. G.; Graham, S. L.; Munshi, S.; McGaughey, G. B.;
Holloway, M. K.; Simon, A. J.; Price, E. A.; Sankaranarayanan, S.;
Colussi, D.; Tugusheva, K.; Lai, M. T.; Espeseth, A. S.; Xu, M.; Huang,
Q.; Wolfe, A.; Pietrak, B.; Zuck, P.; Levorse, D. A.; Hazuda, D.; Vacca,
J. P. Discovery and X-ray crystallographic analysis of a spiropiperidine
iminohydantoin inhibitor of beta-secretase. J. Med. Chem. 2008, 51,
6259−6262.
9223
dx.doi.org/10.1021/jm300931y | J. Med. Chem. 2012, 55, 9208−9223