Discovery of 7-tetrahydropyran-2-yl chromans: β-site amyloid precursor protein cleaving enzyme 1 (BACE1) inhibitors that reduce amyloid β-protein (Aβ) in the central nervous system

Article abstract of DOI:10.1021/jm401635n

In an attempt to increase selectivity vs Cathepsin D (CatD) in our BACE1 program, a series of 1,3,4,4a,10,10a-hexahydropyrano[4,3-b]chromene analogues was developed. Three different Asp-binding moieties were examined: spirocyclic acyl guanidines, aminooxazolines, and aminothiazolines in order to modulate potency, selectivity, efflux, and permeability. Using structure-based design, substitutions to improve binding to both the S3 and S2′ sites of BACE1 were explored. An acyl guanidine moiety provided the most potent analogues. These compounds demonstrated 10-420 fold selectivity for BACE1 vs CatD, and were highly potent in a cell assay measuring Aβ_1-40 production (5-99 nM). They also suffered from high efflux. Despite this undesirable property, two of the acyl guanidines achieved free brain concentrations (C _free,brain) in a guinea pig PD model sufficient to cover their cell IC₅₀s. Moreover, a significant reduction of Aβ_1-40 in guinea pig, rat, and cyno CSF (58%, 53%, and 63%, respectively) was observed for compound 62.

Full text of DOI:10.1021/jm401635n

Article
pubs.acs.org/jmc
Discovery of 7‑Tetrahydropyran-2-yl Chromans: β‑Site Amyloid
Precursor Protein Cleaving Enzyme 1 (BACE1) Inhibitors That Reduce
Amyloid β‑Protein (Aβ) in the Central Nervous System
Allen A. Thomas,*,†,§ Kevin W. Hunt,† Matthew Volgraf,‡ Ryan J. Watts,‡ Xingrong Liu,‡ Guy Vigers,†
Darin Smith,† Douglas Sammond,†,∥ Tony P. Tang,† Susan P. Rhodes,† Andrew T. Metcalf,†
Karin D. Brown,† Jennifer N. Otten,† Michael Burkard,† April A. Cox,†,⊥ Mary K. Geck Do,†,○
Darrin Dutcher,†,∇ Sumeet Rana,†,¶ Robert K. DeLisle,† Kelly Regal,†,# Albion D. Wright,†
Robert Groneberg,†,□ Kimberly Scearce-Levie,‡ Michael Siu,‡ Hans E. Purkey,‡ Joseph P. Lyssikatos,‡
and Indrani W. Gunawardana†
†Array BioPharma, 3200 Walnut Street, Boulder, Colorado 80301, United States
‡Genentech, Inc., 1 DNA Way, South San Francisco, California 94080, United States
S
* Supporting Information
ABSTRACT: In an attempt to increase selectivity vs Cathepsin D (CatD) in our BACE1
program, a series of 1,3,4,4a,10,10a-hexahydropyrano[4,3-b]chromene analogues was
developed. Three different Asp-binding moieties were examined: spirocyclic acyl guanidines,
aminooxazolines, and aminothiazolines in order to modulate potency, selectivity, efflux, and
permeability. Using structure-based design, substitutions to improve binding to both the S3 and
S2′ sites of BACE1 were explored. An acyl guanidine moiety provided the most potent
analogues. These compounds demonstrated 10−420 fold selectivity for BACE1 vs CatD, and
were highly potent in a cell assay measuring Aβ1−40 production (5−99 nM). They also suffered
from high efflux. Despite this undesirable property, two of the acyl guanidines achieved free
brain concentrations (Cfree,brain) in a guinea pig PD model sufficient to cover their cell IC50s.
Moreover, a significant reduction of Aβ1−40 in guinea pig, rat, and cyno CSF (58%, 53%, and
63%, respectively) was observed for compound 62.
comes from clinical trial results from Eli Lilly9 and Merck10 in
which significant decreases in CSF Aβ were observed after
dosing with BACE1 inhibitors. In addition to APP, BACE1
INTRODUCTION
■
During the last few decades, breakthroughs in medicine have
resulted in a decline of deaths associated with many major
processes other substrates that may limit its usefulness as a
diseases. In contrast, deaths due to Alzheimer’s disease have
therapeutic target.11 Nonetheless, Merck’s BACE1 inhibitor
steadily increased, in part due to a growing elderly population.1
MK-8931 has been well tolerated12 and is currently undergoing
The worldwide prevalence of dementia-related diseases, of
evaluation in phase II/III trials.13
which Alzheimer’s is the most common, was estimated to be 36
One of the greatest challenges in developing BACE1
inhibitors for Alzheimer’s disease is the need to cross the
million people in 2010, and this number is expected to almost
double every 20 years.2 Currently there is a great need for
blood−brain barrier. In our discovery program, efforts were
disease-modifying treatments, as the current standards of care
focused on compounds with physicochemical properties14 that
(acetylcholinesterase inhibitors and NMDA antagonists) do not
delay disease progression.3
Targeting the formation of the amyloid peptide (Aβ) has
received considerable attention as a strategy for treating
might give adequate brain levels to elicit a pharmacodynamic
response. In conflict with this goal, our most potent BACE1
inhibitors15 contained either an acyl guanidine (1) or
Alzheimer’s disease.4 Much research has revolved around
aminooxazoline/aminothiazoline (2 and 3, respectively) war-
head (Figure 1), which was necessary for potent interactions
inhibiting the two major proteases responsible for processing
with the catalytic Asps in the active site (Figure 2). These
the amyloid-β precursor protein (APP), γ and β-site APP
cleaving enzyme 1 (BACE1).5 We chose to pursue BACE1
warheads added two hydrogen bond donors and increased the
inhibitors because of the preclinical6 and genetic7 evidence
total polar surface area (tPSA) of our compounds, both of
which increased efflux liability and made obtaining brain
supporting its role in production of Aβ in the brain. More
penetrant molecules more challenging.14 In addition to potency
recently, it was demonstrated that a rare mutation in the APP,
which reduces the efficiency of BACE1 cleavage, correlated
with a dramatic improvement in cognition among elderly
Received: October 21, 2013
Scandinavians.8 Compelling evidence for targeting BACE1
© XXXX American Chemical Society
A
dx.doi.org/10.1021/jm401635n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
hexanes. Compound 9 was then subjected to the Bucherer−
Bergs reaction, which gave diastereomer 10 as the major
isomer, which was enriched to >80% diastereomeric purity by
trituration with hot i-PrOH. Methylation of hydantoin 10
provided compound 11, which was subsequently converted to
the thiohydantoin 12 with Lawesson’s reagent. Oxidation/
aminolysis of thiohydantion 12 afforded intermediate 13.
Analogues in Table 1 (56−68) were prepared via Suzuki
couplings with 13. Racemic compounds 13 and 61 were
resolved by chiral SFC to provide compounds 55, 62, and 63
(Table 1). Compound 59 was prepared from enantiopure
intermediate 55. A cocrystal structure of 62 with BACE1
(Figure 2) was obtained, and by analogy, the stereochemistries
for compounds of Table 1 (55−68) were assigned.
Figure 1. Numbering system and ring designation to be used for
1,3,4,4a,10,10a-hexahydropyrano[4,3-b]chromene analogues (“7-THP
chromans”) discussed in this article. Spirocyclic warheads (WH): acyl
guanidine (1), aminooxazoline (2), and aminothiazoline (3).
Analogues in Table 2, containing aminooxazoline or
aminothiazoline warheads, were prepared according to Scheme
2. Chromanone 9 (from Scheme 1) was reacted with Tebbe’s
reagent to afford alkene 14. Using published conditions,17
alkene 14 was converted to aminooxazoline 15 (X = O) or
aminothiazoline 16 (X = S), obtained as a mixture of
stereoisomers. Diastereomers of 15 (X = O) were separated
by silica gel chromatography to provide 15a and 15b in nearly
equivalent amounts. Diastereomers 16 were carried forward as
a mixture and separated by silica chromatography at the final
product stage. Analogous to compounds of Table 1, Suzuki
coupling of intermediates 15 or 16 with various boronic acids
provided compounds 69−76. The stereochemical assignments
for 15a and 15b, and by analogy, compounds 69−74 (Table 2)
were made based on the cocrystal structure of compound 72
with BACE1.18 The alternate diastereomers of aminothiazolines
75 and 76 were not isolated. The stereochemistries for
compounds 75 and 76 (Table 2) were determined by cocrystal
structure of compound 76 with BACE1.18
Aminooxazolines 77−82 (Table 3), substituted with a
methyl group at the β-position, were synthesized according to
Scheme 3. Salicylaldehyde 18 was cyclized to hemiacetal 19 in
the presence of Et3N and 3-methylbut-2-enal 17. Diol 20 was
formed upon treatment with NaBH4 and could then be
converted to ketone 21 in the presence of MnO2. Treatment
with MOMCl provided MOM-protected ether 22. Cyclic ether
24 was synthesized by deprotonation of 22 with LHMDS and
trapping with TMSCl to provide intermediate 23 (not
isolated), which was then treated with TiCl4 to give 24 as
the cis ring junction diastereomer. This cis ketone was then
epimerized by deprotonation with LHMDS followed by kinetic
reprotonation at −78 °C with ethyl salicylate to provide ketone
25 as a 5:1 mixture of trans/cis diastereomers. Olefin 26 was
formed upon treatment with Tebbe’s reagent. The spirocyclic
aminooxazolines 27 were then formed by treatment with I2 and
AgOCN followed by the addition of aq. NH4OH, in an
analogous fashion to compounds 15 (Scheme 2). A mixture of
diastereomers was obtained and carried forward to final
product. Suzuki coupling then provided compounds 77−82,
which were isolated as pure diastereomers by silica gel
chromatography or as single enantiomers after chiral SFC
purification.
Figure 2. X-ray cocrystal structure of BACE1 and 62. The two key
active site aspartic acids, Asp32 and Asp228, are labeled. The distances
between water oxygens (depicted as red spheres) and 62 are given in
angstroms.
and brain penetration, we also desired compounds with >100-
fold selectivity against another aspartyl protease, Cathepsin D
(CatD), which has been implicated in multiple physiologically
important roles.16 In an effort to overcome these challenges, we
hypothesized that modification of our previously described
BACE1 inhibitors15 by incorporation of a 7-THP A-ring would
generate 1,3,4,4a,10,10a-hexahydropyrano[4,3-b]chromenes
(1−3; hereafter referred to as “7-THP chromans”), which
might be highly potent vs BACE1 and >100-fold selective
against CatD, due to differences in their active sites (see
Supporting Information). The SAR and PK for these 7-THP
chromans, which culminated with advancement of 62 to PD
studies and its effect on lowering Aβ in the CSF will be the
main focus of this article.
CHEMISTRY
■
Scheme 1 describes the synthesis of spirocyclic acyl guanidines
of Table 1. Enamine 5 was prepared via the condensation of
morpholine and pyranone 4. Subsequent addition of 5 to
aldehyde 6 followed by intramolecular aminal formation
provided compound 7. Swern oxidation of 7, followed by in
situ elimination afforded chromene-one 8. L-Selectride
mediated reduction then gave chromanone 9, which was
isolated as the pure trans diastereomer after trituration with
3-Fluoro chromans 92 and 93 (Table 4) were synthesized
from 28 in an analogous fashion to compounds 77−82
according to Scheme 3. Relative stereochemistries for
compounds 92 and 93 were determined by NMR comparison
with compounds 77−82, for which absolute stereochemistry
had been unambiguously determined by cocrystal structure of
79 with BACE1.
B
dx.doi.org/10.1021/jm401635n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Scheme 1
a
Reagents and conditions: (a) morpholine, toluene, reflux, Dean−Stark trap; (b) toluene, rt; (c) oxalyl chloride, DMSO, Et3N, DCM, −78 °C; (d)
HOAc, rt; (e) L-selectride, THF, −78 °C; (f) KCN, (NH4)2CO3, NaHSO3, EtOH, 130 °C; (g) MeI, K2CO3, DMF, rt; (h) Lawesson’s reagent,
toluene, 90 °C; (i) t-BuOOH, aq. NH4OH, MeOH, 50 °C; (j) RB(OH)2, Pd[P(Ph)3]4, aq. Na2CO3, dioxane, 90 °C. All chiral compounds depicted
above were racemic, except for compounds 59, 62, and 63 as discussed below.
a
Table 1. SAR of 7-THP Chromans Containing a Spirocyclic Acyl Guanidine
IC50 (nM)
b
c
d
e
f
compd
R
BACE1
CatD
CatD/BACE1
cell IC50 (nM)
efflux ratio
Papp AB
g
55
Br
670
28
23
31
19
31
27
9.5
490
33
27
28
24
39
24000
1800
780
36
1200
27
3.4
47
49
37
69
28
91
67
140
56
170
61
61
12
22
56
57
58
3-methoxyphenyl
64
25
3-(difluoromethoxy)phenyl
3-cyanophenyl
34
21
16
2700
960
87
14
18
g
59
3-cyanophenyl
51
7.3
44
14
h
60
61
3-chloro-5-fluorophenyl
2-fluoropyridin-3-yl
2-fluoropyridin-3-yl
2-fluoropyridin-3-yl
5-fluoropyridin-3-yl
5-chloropyridin-3-yl
5-methoxypyridin-3-yl
5-cyanopyridin-3-yl
pyrimidin-5-yl
2300
74
11
1100
840
41
9.9
5.4
550
16
9.8
9.0
8.2
6.6
16
g
62
i
88
63
14000
6200
3200
5400
6300
11000
29
64
65
66
67
68
190
120
190
260
280
11
8.3
21
3.7
2.4
1.2
39
a
Compounds were racemic, with relative stereochemistry as depicted, unless specified otherwise. All assay data was obtained at least in triplicate, and
b
c
the values listed are the average from multiple experiments. BACE1 enzyme assay performed at 20 nM substrate concentration. Cathepsin D
enzyme assay was performed using enzyme derived from liver cells, unless specified otherwise. Cell assay measuring Aβ1−40 production. Efflux ratio
d
e
is the (B to A)/(A to B) value derived from LLC-PK1 cells transfected with human MDR1 (P-gp). Compounds were tested at 1 μM concentration.
f
Control P-gp substrate, digoxin, had an average efflux ratio of 51 30. Papp AB was determined using the LLC-PK1 parental cell line, in units of
g
h
10−6 cm/s. Single enantiomer with R configuration at the spirocyclic center and S,S configuration at the ring fusion. Cathepsin D enzyme assay was
i
performed using enzyme derived from spleen cells. Single enantiomer with S configuration at the spirocyclic center and R,R configuration at the ring
fusion.
Scheme 4 depicts the synthesis of compounds 83−85 (Table
3), substituted with a methyl group at the 8-position. Acetyl-
substituted chromenone 40 was prepared using published
procedures for related hydroxy aryl ethanones.19 For the
synthesis of these analogues, as well as for compounds 90 and
91 (Scheme 6), the THP A-ring was built via a hetero Diels−
Alder reaction (e.g., 40 to 41). The stereochemistry of adduct
41 was not unambiguously determined, thus the depicted
stereochemistry is analogous to structurally related hetero
Diels−Alder products described previously.20 Compound 42
C
dx.doi.org/10.1021/jm401635n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Table 2. SAR of 7-THP Chromans Containing Spirocyclic Aminooxazolines or Aminothiazolines
IC50 (nM)
b
c
d
e
f
compd
X
diastereomer
R
BACE1
CatD
CatD/BACE1
cell IC50 (nM)
efflux ratio
Papp AB
69
70
71
72
73
74
75
76
O
O
O
O
O
O
S
A
B
A
B
A
B
B
B
5-chloropyridin-3-yl
5-chloropyridin-3-yl
2-fluoropyridin-3-yl
2-fluoropyridin-3-yl
pyrimidin-5-yl
760
12000
16
230
120
670
270
650
940
79
4.3
2.6
2.8
2.1
4.9
5.5
1.7
1.6
29
31
31
22
13
9.7
22
31
440
120000
270
12
g
1300
830
15000
g
170000
205
30
g
2500
2000
220
76000
g
pyrimidin-5-yl
>200000
57000
>100
260
290
5-chloropyridin-3-yl
2-fluoropyridin-3-yl
S
330
96000
180
a
Compounds were racemic, with relative stereochemistry as depicted, unless specified otherwise. All assay data was obtained at least in triplicate, and
b
c
the values listed are the average from multiple experiments. BACE1 enzyme assay performed at 20 nM substrate concentration. Cathepsin D
enzyme assay was performed using enzyme derived from liver cells, unless specified otherwise. Cell assay measuring Aβ1−40 production. Efflux ratio
is the (B to A)/(A to B) value derived from LLC-PK1 cells transfected with human MDR1 (P-gp). Compounds were tested at 1 μM concentration.
d
e
f
Control P-gp substrate, digoxin, had an average efflux ratio of 51 30. Papp AB was determined using the LLC-PK1 parental cell line, in units of
g
10−6 cm/s. Cathepsin D assay was performed using enzyme derived from spleen cells.
a
Scheme 2
a
Reagents and conditions: (a) Tebbe’s reagent, THF, rt; (b) I2, AgXCN (X = O or S), aq. NH4OH, CH3CN, THF, rt; for X = O, diastereomers were
separated; (c) RB(OH)2, Pd[P(Ph)3]4, aq. Na2CO3, dioxane, 90 °C, for X = S, single diastereomers were isolated. All chiral compounds depicted
above were racemic.
was generated as a single diastereomer by DIBALH reduction
of 41. Triethyl silane with boron trifluoride etherate effected
reduction of the acetal functionality of 42 to afford THP 43.
The relative stereochemistry for 43 as well as for the other
compounds in Scheme 4 were assigned from cocrystal
structures of final products 83−85 with BACE1.18 Methyl
hydantoin 44 was prepared from THP 43 in an analogous
manner to compound 11 (Scheme 1), and obtained as a 60:40
mixture of trans diastereomers at the ring fusion. The two trans
diastereomers were separated by silica gel chromatography to
provide a single diastereomer 44, which was converted to final
product 83 in an analogous fashion to compounds 56−68
(Scheme 1). Compounds 84 and 85 were similarly prepared
from THP 43 as described for 69−76 (Scheme 2), and the
diastereomers were separated at final product by silica gel
chromatography.
7-THP chromans 86−89 (Table 3) substituted with a methyl
group at the α-position were formed according to Scheme 5.
Chromanone 9, prepared as described in Scheme 1, was
methylated using LDA with HMPA to yield 45 in >90%
diastereomeric purity after silica gel chromatography. Con-
version of 45 to products 86−88 was executed analogous to
compounds 56−68 (Scheme 1). The Bucherer−Bergs reaction
of ketone 45 (Scheme 5, step b) gave >85% diastereomerically
D
dx.doi.org/10.1021/jm401635n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Table 3. SAR of 7-THP Chromans with Methyl Substitution at the α, β, and 8-Positions
a
Compounds were racemic, with relative stereochemistry as depicted, unless specified otherwise. All assay data was obtained at least in triplicate, and
b
c
the values listed are the average from multiple experiments. BACE1 enzyme assay performed at 20 nM substrate concentration. Cathepsin D
d
e
enzyme assay was performed using enzyme derived from liver cells. Cell assay measuring Aβ1−40 production. Efflux ratio is the (B to A)/(A to B)
value derived from LLC-PK1 cells transfected with human MDR1 (P-gp). Compounds were tested at 1 μM concentration. Control P-gp substrate,
f
g
digoxin, had an average efflux ratio of 51 30. Papp AB was determined using the LLC-PK1 parental cell line, in units of 10−6 cm/s. Single
enantiomer with R configuration at the spirocyclic center and S,S configuration at the ring fusion.
pure spirocyclic hydantoin (not depicted), which was
subsequently transformed (Scheme 5, steps c−f) to products
86−88. The stereochemistry of ketone 45 and products 86−88
was determined by cocrystal structure of analogue 87 with
BACE1.18 Product 89 was prepared analogous to compounds
from Scheme 2. Formation of the aminooxazoline moiety
(Scheme 5, step h) was >90% diastereoselective giving the
relative stereochemistry depicted for 89, which was confirmed
by cocrystal structure with BACE1.18
Spirocyclic acyl guanidines 90 and 91 (Table 4) with fluoro
substitution at the C-ring’s 3-position were prepared according
to Scheme 6. In order to avoid competing nucleophilic aromatic
substitution (SNAr) with a potentially reactive 3-fluoro 2-bromo
chromanone during the Bucherer−Bergs reaction (Scheme 6,
step h), we chose to utilize a methoxy group in place of the
bromo substituent previously used as a handle for Suzuki
couplings in Schemes 1−5. Baeyer−Villiger oxidation of
starting material 46 yielded 47, which was subjected to a
Fries rearrangement to give phenyl ethanone 48. Reaction of 48
with triphosgene afforded aldehyde 49, which was converted to
50 via a hetero Diels−Alder reaction analogous to compound
41 in Scheme 4. Hydrogenation of olefin 50 lead to partial
reduction of the ketone functionality, so the crude was stirred
with MnO2 in DCM to provide 51. Removal of the acetal
functionality was accomplished with Et3SiH/BF3-Et2O to give
ketone 52, which was transformed to 53 (mixture of
E
dx.doi.org/10.1021/jm401635n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Scheme 3
a
Reagents and conditions: (a) Et3N, dioxane, 55 °C; (b) NaBH4, MeOH, THF, 0 °C; (c) MnO2, DCM, rt; (d) MOMCl, i-Pr2EtN, DCM, rt; (e)
LHMDS, TMSCl, THF, −70 °C; (f) TiCl4, DCM, −40 °C; (g) LHMDS, THF, −78 °C; (h) ethyl 2-hydroxybenzoate, THF, −78 °C; (i) Tebbe’s
reagent, THF, −78 °C to rt; (j) I2, AgOCN, aq. NH4OH, CH3CN, EtOAc, 40 °C; (k) RB(OH)2, Pd[P(Ph)3]4, aq. Na2CO3, dioxane, 90 °C;
diastereomers were separated. All chiral compounds depicted above were racemic, except for compounds 77, 78, 92, and 93 as discussed below.
a
Table 4. SAR of 7-THP 3-Fluoro Chromans
a
Compounds were racemic, with relative stereochemistry as depicted, unless specified otherwise. All assay data was obtained at least in triplicate, and
b
c
the values listed are the average from multiple experiments. BACE1 enzyme assay performed at 20 nM substrate concentration. Cathepsin D
d
e
enzyme assay was performed using enzyme derived from liver cells. Cell assay measuring Aβ1−40 production. Efflux ratio is the (B to A)/(A to B)
value derived from LLC-PK1 cells transfected with human MDR1 (P-gp). Compounds were tested at 1 μM concentration. Control P-gp substrate,
f
g
digoxin, had an average efflux ratio of 51 30. Papp AB was determined using the LLC-PK1 parental cell line, in units of 10−6 cm/s. MLM: mouse
h
liver microsomes. Single enantiomer with R configuration at the spirocyclic center and S,S configuration at the ring fusion.
diastereomers) analogous to compound 13 in Scheme 1. Acyl
guanidine 53 was carried forward as a mixture to give triflate 54
after HBr-mediated deprotection of the methoxy group,
protection as an amidine, and conversion of the phenol to a
triflate. The amidine protecting group made the diastereomers
of 54 more amenable to separation by silica gel chromatog-
raphy. Final products 90 and 91 were generated by Suzuki
couplings with triflate 54, and the amidine protecting group was
simultaneously removed under the Suzuki reaction conditions.
The relative stereochemistry for racemates 90 and 91 was
assigned based on NMR comparison with compounds 56−68
(Scheme 1), for which the stereochemistry had been
unambiguously determined by cocrystal structure of compound
62 with BACE1 (Figure 2).
RESULTS AND DISCUSSION
■
Structure−Activity Relationships (SAR). 7-THP chro-
man analogues were evaluated for activity in BACE1 and CatD
enzyme assays, a cell assay for production of Aβ1−40, and for
efflux ratio and permeability in an MDR1-transfected LLC-PK1
cell line. Identification of compounds with <100 nM cell
potency, >100-fold selectivity for CatD (due to its role in
multiple physiological functions16), an efflux ratio <2 (to
minimize P-gp liability), and apparent cell permeability in the
apical to basolateral direction (Papp AB) > 10 × 10−6 cm/s was
the goal. Coverage ratio21 (Cfree,brain/cell IC50) was a key
assessment that incorporated PK, potency, and brain
penetration. Although we were unable to identify a compound
F
dx.doi.org/10.1021/jm401635n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Scheme 4
a
Reagents and conditions: (a) DMF-dimethyl acetal, toluene, reflux; (b) Ac2O, py, rt; (c) ethyl vinyl ether, 100 °C; (d) DIBALH, THF, −78 °C; (e)
Et3SiH, BF3-Et2O, DCM, rt; (f) KCN, (NH4)2CO3, NaHSO3, EtOH, 130 °C; (g) MeI, K2CO3, DMF, rt; diastereomers were separated; (h)
Lawesson’s reagent, toluene, 90 °C; (i) t-BuOOH, aq. NH4OH, MeOH, 50 °C; (j) (2-fluoropyridin-3-yl)boronic acid, Pd[P(Ph)3]4, aq. Na2CO3,
dioxane, 90 °C; (k) Tebbe’s reagent, THF, rt; (l) I2, AgXCN (X = O or S), aq. NH4OH, CH3CN, THF, rt. All chiral compounds depicted above
were racemic.
a
Scheme 5
a
Reagents and conditions: (a) LDA, HMPA, MeI, THF, −78 °C; (b) KCN, (NH4)2CO3, NaHSO3, EtOH, 130 °C; (c) MeI, K2CO3, DMF, rt; (d)
Lawesson’s reagent, toluene, 90 °C; (e) t-BuOOH, aq. NH4OH, MeOH, 50 °C; (f) RB(OH)2, Pd[P(Ph)3]4, aq. Na2CO3, dioxane, 90 °C; (g)
Tebbe’s reagent, THF, rt; (h) I2, AgOCN, aq. NH4OH, CH3CN, THF, rt. All chiral compounds depicted above were racemic.
meeting all of these criteria, we discovered that some of our
compounds were still able to attain sufficient coverage ratios to
decrease Aβ in the CSF (Table 7 and Figure 5, vide infra). We
utilized the SAR from Tables 1−4 to select compounds with
different profiles of cell potency, efflux, and permeability to
evaluate in mouse PK. The results from mouse PK (Table 5), in
particular the coverage ratio, then guided the optimization and
selection of compounds to progress to PD studies.
compound 55 (R = Br) was >17-fold less potent than
compounds in which R was a larger aromatic moiety. As a
result of this preferred binding orientation, enantiomers were
expected to have dramatically different potencies. Indeed, pure
enantiomer 63 was 52-fold less potent for BACE1 than its
antipode 62. The various aromatic R groups of Tables 1−4
were chosen based on our earlier SAR15 and supported by
crystallography, so as to maximize binding affinity to BACE1.
Thus, compounds 56−62 and 64−68 were all <40 nM in the
enzyme assay. These compounds were also highly potent in the
cell assay, wherein the two most potent racemic analogues were
61 and 66 (R = 2-fluoropyridin-3-yl and 5-methoxypyridin-3-yl;
IC50s = 9.9 and 8.3 nM, respectively).
From our previous work15 and others,22 we expected the
three different Asp-binding warheads to anchor inhibitors in the
BACE1 active site such that substitution off the aromatic C-ring
with various R groups would gain potency by projection into
the S3 pocket (Figure 2). This was found to be the case, as
G
dx.doi.org/10.1021/jm401635n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Scheme 6
a
Reagents and conditions: (a) m-CPBA, DCM, 40 °C; (b) TfOH, 60 °C; (c) triphosgene, DMF, DCE, rt; (d) ethyl vinyl ether, 100 °C; (e) Pd/C,
EtOH, H2 (50 psi), rt; (f) MnO2, DCM, rt; (g) Et3SiH, BF3−Et2O, DCM, rt; (h) KCN, (NH4)2CO3, NaHSO3, EtOH, 130 °C; (i) MeI, K2CO3,
DMF, rt; (j) Lawesson’s reagent, toluene, 90 °C; (k) t-BuOOH, aq. NH4OH, MeOH, 50 °C; (l) 48% aq. HBr, 100 °C; (m) DMF−dimethyl acetal,
DMF, rt; (n) Tf2NPh, Et3N, DCM, rt; diastereomers were separated; (o) RB(OH)2, Pd[P(Ph)3]4, aq. Na2CO3, dioxane, 90 °C. All chiral
compounds depicted above were racemic.
a
Table 5. Mouse PK: iv, po Dose
a
b
Racemic compounds (unless specified otherwise) were dosed at 3 mg/kg iv and 10 mg/kg po in male CD-1 mice: 3 animals/cohort. iv and po
c
d
formulation: 40% PEG 400/10% EtOH/50% H2O. iv and po formulation: 30% captisol in H2O. iv formulation: 50 mM citrate in saline (pH 4). po
formulation: 1% carboxymethyl cellulose (CMC)/0.5% Tween/5 mM citrate in H2O. iv formulation: 40% PEG 400/10% EtOH/50% H2O. po
formulation: 1% CMC/0.5% Tween/5 mM citrate in H2O. Single enantiomer with R configuration at the spirocyclic center and S,S configuration at
the ring fusion. Compounds (1 μM) incubated with mouse liver microsomes for 20 min at 37 °C in the presence of NADPH. CLhep,syst,u
CLhep,syst/Fu,plasma. Compound levels were BLQ. Coverage ratio = Cfree,brain/cell IC50. Control P-gp substrate, digoxin, had an average efflux ratio of
51 30. See ref 28.
e
f
g
h
=
i
j
k
l
H
dx.doi.org/10.1021/jm401635n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Consistent with our earlier SAR,15 selectivity against CatD
for compounds of Table 1 generally improved with more polar
R groups. The most selective were 64 and 66−68 (190−280-
fold), and these compounds had low cLogPs, which ranged
from −1.0 to 0.3. In addition to selectivity from the more polar
aromatic R groups, the THP A-ring also contributed to
selectivity for BACE1. For example, 7-THP chroman 56 was
both more potent against BACE1 and less potent against CatD
than the corresponding dimethyl chroman (Figure 3). This
result is consistent with data on other chromans from our
previous work.15b
69−74, the two diastereomers (type A and B) with opposite
trans stereochemistry at the ring fusion were synthetically
accessible (via Scheme 2), whereas only diastereomer B for
aminothiazolines 75 and 76 was generated in sufficient quantity
for isolation. Significant differences in selectivity were observed
between pairs of trans diastereomers 69−74. Diastereomers of
type B were found to be significantly more selective for BACE1
than type A, supporting our previous hypothesis that the THP
A-ring contributes to CatD selectivity. Changing from the acyl
guanidine in type A diastereomers 65, 61, and 68 to the
aminooxazoline in 69, 71, and 73, respectively, resulted in
decreased selectivity. Because these aminooxazolines and
aminothiazolines provided lower efflux ratios and improved
permeability relative to the acyl guanidines, we were prompted
to further explore the SAR around the THP A-ring with a goal
of increasing potency. On the basis of the cocrystal structure for
62 (Figure 2), we identified the α-, β-, and 8-positions on the
THP A-ring as sites for substitution.
Since the flexible loops of BACE1 and CatD are very
different (see Figure S1, Supporting Information), it was
suggested that addition of a methyl group to the β-position of
the A-ring might increase CatD selectivity. Aminooxazolines
with methyl substitution at the β-position (Table 3)24
demonstrated improved enzyme and cell potencies relative to
the corresponding diastereomers from Table 2. The β-methyl
group also provided a dramatic increase in selectivity for
BACE1 (e.g., 81, 190-fold, vs 71, 12-fold). Moreover,
compound 82 was inactive against CatD (40% inhibition at
200 μM).
Low efflux ratios and high permeability were maintained for
77−82. Methyl substitution at the 8-position25 also resulted in
improved enzyme and cell potencies for analogues 84 and 85
relative to the corresponding diastereomers 72 and 76.
However, incorporation of the 8-methyl group was not
sufficient to provide potency comparable to the corresponding
acyl guanidine 83. Selectivities for 8-methyl analogues 84 and
85 were similar to des-methyl analogues 72 and 76 and easily
met our criteria for selectivity (>100-fold). Methyl substitution
at the α-position for compound 8926 resulted in a modest
decrease in potency relative to comparator 72. The α-methyl
substituted acyl guanidines 86−88 had similar enzyme
potencies to the corresponding des-methyl analogues in
Table 1 (58, 61, and 68), but gained a ∼2−3-fold improvement
in cell potency. The α-methyl group offered no significant
benefit toward reducing efflux for acyl guanidines 86−88.
Although methyl substitution at the β- and 8-positions of the
THP A-ring generally improved potency relative to their des-
methyl counterparts, analogues prepared with aminooxazolines
or aminothiazolines still had inferior potency relative to the
corresponding acyl guanidines.
Compounds from Tables 1−3 were generally predicted to
have high hepatic clearance based on a mouse liver microsome
(MLM) assay, and several analogues selected for mouse PK
also suffered from high systemic clearance (Table 5). It is
known that replacement of hydrogen with fluorine can improve
metabolic stability.27 On the basis of observations made from
our previous work on chromans,15b,c we hypothesized that
substitution of the C-ring’s 3-position with a fluorine would
improve metabolic stability. To test this hypothesis, 3-fluoro
chromans of Table 4 were prepared as described in Schemes 3
and 6. Contrary to what we hoped, stability in MLMs was not
substantially different for the two acyl guanidines 90 and 91
(CLhep = 21 and 40 mL/min/kg, respectively) compared to
Figure 3. Comparison of 7-THP chroman 56 with a less selective
dimethyl chroman (both compounds were racemic).15b
From a cocrystal crystal structure of 62 with BACE1 (Figure
2), the THP A-ring projects into the S2′ pocket and the A-ring
oxygen interacts with a well-defined water network. Compar-
isons of this structure with the published structure of CatD
(pdb code: 1LYA) show that there are many differences
between the two active sites, both in the flexible loop and in the
S1 and S2′ pockets (see Figure S1, Supporting Information). It
is likely that several of these structural differences, including the
interaction of the A-ring with the water network, contribute to
the selectivity of these compounds for BACE1 over CatD.
None of the spirocyclic acyl guanidines of Tables 1, 3, and 4
met our goal of an efflux ratio <2. We partially attributed the
large efflux ratios to the compounds’ tPSAs (77−110) and the
presence of two hydrogen bond donors, both of which are
likely to increase P-gp mediated efflux.14 As one might expect,
there was a general trend for compounds containing more polar
R groups (e.g., compounds 66−68) to be less permeable.
However, we appreciate that recognition by transporters such
as P-gp is a whole molecule property since 60 (R = 3-chloro-5-
fluorophenyl) had an 8-fold greater efflux ratio than 55 (R =
Br), despite the fact that their tPSAs are both 77 and that they
both contain acyl guanidines.
We hypothesized that reducing tPSA by changing from the
acyl guanidine (e.g., 61, tPSA = 90) to the aminooxazoline (e.g.,
71, tPSA = 78) or aminothiazoline moieties (e.g., 76, tPSA =
69) would have a positive impact on both efflux and
permeability. To our satisfaction, compounds of Table 2 all
exhibited lower efflux ratios and greater permeability. It was
recently reported that efflux ratio was directly correlated with
pKa for a series of BACE1 inhibitors;23 however, we observed
the opposite trend in this work. For example, acylguanidine 62
(efflux ratio = 67) had a measured pKa of 5.8, whereas
aminooxazolines 78, 84, and 89 (efflux ratios < 2.5) had pKa
values of 6.7, 6.4, and 6.4, respectively. Apparently the lower
tPSA values for the aminooxazolines had a greater influence on
efflux ratio than pKa.
Unfortunately, changing to the alternate Asp binders resulted
in an 8−64-fold decrease in BACE1 potency with a
commensurate drop in cell potency. For aminooxazolines
I
dx.doi.org/10.1021/jm401635n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
des-fluoro analogues 57 and 61 (CLhep = 10 and 46 mL/min/
kg, respectively). In like manner, the two aminooxazolines 92
and 93 also demonstrated no improvement in microsome
stability by comparison with 79 and 81 (CLhep = 60 and 61
mL/min/kg, respectively). The enzyme and cell potencies for
the two acyl guanidines 90 and 91 were ∼2−5-fold less than
those for des-fluoro analogues 57 and 61. Aminooxazolines 92
and 93 were tested as eutomers, so their enzyme and cell
potencies were likewise not substantially different (∼1−3-fold)
from the analogous racemates 79 and 81, after correcting for
the difference in enantiomeric purity. A more marked effect of
fluorine substitution was on selectivity, as compounds 90−93
demonstrated a ∼3−4-fold decrease relative to their des-fluoro
versions. In light of these disappointing results, fluorine
substitution at the 3-position was not pursued on the other
7-THP chroman cores from Tables 2 and 3.
0.0018). In contrast, acyl guanidines 58, 61, and 87 were
moderate to high clearance compounds, but they had ∼10−50-
fold greater free fraction in the brain than 60. A large free
fraction contributed to their achieving coverage ratios of 0.31−
0.85 at the 4 h time point, in spite of moderate to high
clearance and high efflux. However, free brain to plasma (b/p)
ratios for 58, 61, and 87 were poor (0.025−0.15, 4 h time
point), likely due to their high efflux. Oral bioavailability (F%)
for 58 and 61 was very good (F%: 10328 and 86, respectively),
but 87 had F% of 15, even though its permeability was
comparable to 61. It might be that addition of the α-methyl
group in 87 lead to increased first pass metabolism, but this was
unverified. Though 83 had similar PK and brain tissue binding
to structurally related 61, its brain levels were below the level of
quantitation (BLQ) at the 4 h time point for reasons not
readily apparent.
As the acyl guanidines (Tables 1, 3, and 4) suffered from high
efflux, we wanted to further improve their cell potencies; and by
doing so, potentially increase coverage ratios. From our
previously published work,15b we had determined that replacing
the chroman-based scaffold (Figure 4, X = O) with a
Aminooxazolines 78 and 79 demonstrated moderate
systemic clearance that was better than predicted in vitro,
and both had low to moderate F%. Compound 79 achieved free
brain concentrations comparable to 61, but because of its being
∼2-fold less potent, the coverage ratio was inferior. Amino-
oxazoline 78, one of the least effluxed (efflux ratio = 1.5) of the
7-THP chromans in this work, also demonstrated a superior
free b/p ratio (0.68, 4 h time point). However, its moderate cell
potency (IC50 = 41 nM) and high brain tissue binding (Fu,brain
=
0.0031) contributed to a poor coverage ratio of 0.061 at the 4 h
time point.
Pharmacokinetics for 62 in Multiple Species. Com-
pound 61 demonstrated the best coverage ratio in the mouse
(Table 5: 1.1, 1 h; 0.85, 4 h), and its eutomer 62 passed
multiple in vitro toxicological hurdles (e.g., hERG patch clamp
>100 μM, IC50 >10 μM against CYP’s, clean in Ames and
micronucleus assays, and highly selective in receptor, kinase,
and protease panels). Thus, the PK for 62 in multiple species
(Table 6) was evaluated to enable in vivo toxicology as well as
PD studies in guinea pig, rat, and cyno. There was a good in
vitro−in vivo correlation (IVIVC) of microsome and
hepatocyte stability with in vivo clearances observed for
mouse and rat. For cyno, there was a good IVIVC for
microsomes; however, hepatocytes underpredicted clearance.
In contrast, guinea pig demonstrated very high in vivo clearance
with no apparent IVIVC. Oral bioavailability was good for
mouse and rat (F% = 61 and >100, respectively), but poor for
guinea pig and cyno (F% = 17 and 10, respectively). Unbound
levels of 62 in plasma were ∼10−50-fold greater than in brain
for mouse and rat, and ∼10−30-fold greater than CSF levels in
rat and cyno, likely a result of the compound’s high efflux.
Because of the exceptionally high clearance in guinea pig, it was
not surprising that plasma levels of 62 were >20-fold lower than
for rat and cyno. However, the difference between plasma and
brain/CSF levels (∼1−7-fold) was significantly smaller in
guinea pig than the other species. In spite of this disconnect,
CSF concentrations for 62 in rat, guinea pig, and cyno, the
three species chosen for PD studies (Table 7 and Figure 5)
were greater than cell IC50 at both time points.
Figure 4. Comparison of racemic compounds 61 and 94.
tetrahydronaphthalene (THN; Figure 4, X = C) generally
resulted in greater cell potency. Indeed, compound 94 was ∼6-
fold more potent in our cell assay than 61, even though their
enzyme potencies were within 2-fold of each other. Besides a
difference in cell potency, 94 and 61 had similar selectivity,
were both highly effluxed, and both had acceptable
permeability. Unfortunately, the THN-based 7-THP analogues
were synthetically less accessible than the corresponding
chromans, and they were not pursued further.
Pharmacokinetic Studies in Mice. From the 7-THP
chromans described above, we selected seven acyl guanidines
58, 60, 61, 64, 65, 83, and 87 and two aminooxazolines 78 and
79 to evaluate in mouse PK (Table 5). In spite of the efflux
liability for the acyl guanidines, we wanted to assess their
coverage ratios (Cfree,brain/cell IC50) to prioritize analogues for
progression to pharmacodynamic studies. Aminooxazolines 78
and 79 were selected to determine if their lower efflux ratios
(1.5 and 4.3, respectively) would translate to greater brain
levels.
Compounds were dosed iv and po (3 and 10 mg/kg,
respectively) in male CD-1 mice, with plasma samples
withdrawn and brains harvested (oral groups only) at 1 and
4 h time points (Table 5). Generally, MLM’s underpredicted
hepatic clearance for the acyl guanidines, with most having
moderate to high systemic clearance. Compounds 64 and 65
were cleared faster than hepatic blood flow, resulting in brain
levels that were below level of quantitation (BLQ) at the 4 h
time point. Despite having the lowest in vivo clearance, 60 had
free brain levels that were 91-fold lower than its cell IC50 at the
Pharmacodynamic Studies. Enantiopure compounds 59
and 62 were selected for PD studies in guinea pig (Table 7),
and 62 was also evaluated in rat and cyno PD models (Table 7
and Figure 5, respectively). The maximum reduction observed
in guinea pig CSF Aβ1−40 was 58% at the 3 h time point, for
both the 60 and 100 mg/kg doses of 62. For guinea pig, there
was a crude correlation between free brain concentration of 59
and 62 and PD response, with the maximum response (52%
4 h time point, due in part to a very low free fraction (Fu,brain
=
J
dx.doi.org/10.1021/jm401635n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
For cyno, the maximum observed reduction in CSF Aβ1−40
for 62 was 63% at 8 h, which lagged behind its maximum CSF
concentration of 120 nM at 2 h (Figure 5). As with rat, there
was a sustained PD response in cyno, with a 49% reduction in
CSF Aβ1−40 at the 24 h time point. However, the PD response
in cyno persisted even though the CSF concentration for 62
had decreased by 24-fold from its peak concentration after 24 h.
By the 48 h time point, cyno CSF Aβ1−40 levels had returned to
baseline. Full details of this cyno PD study are published
elsewhere.30
Table 6. PK for Eutomer 62 in Multiple Species: iv, po Dose
SUMMARY
■
Using structure-based design, we substituted a 1,3,4,4a,10,10a-
hexahydropyrano[4,3-b]chromene (“7-THP chroman”) core to
improve affinity for BACE1 and selectivity against CatD. Three
different warheads (spirocyclic acyl guanidine, aminooxazoline,
and aminothiazoline), which anchored the molecules in the
active site by binding to the catalytic Asp residues, were
examined for their effect on potency, selectivity, efflux, and
permeability. The acyl guanidine moiety provided the most
potent analogues, but at the expense of high efflux ratios and
lower permeability than the two alternative warheads. On the
basis of crystallography and modeling of the S2′ site, methyl
substitution was explored at three different positions (α-, β-,
and 8-positions) around the 7-THP A-ring in order to boost
potency and selectivity. Although we did identify BACE1
inhibitors with superior potency and selectivity relative to their
des-methyl counterparts, the improvements in potency were
unable to provide compounds (e.g., 78 and 79) with free brain
levels sufficient to cover their cell IC50 values in vivo.
Substitution with various aromatic R groups, which provided
potency and selectivity through binding to the S3 site, resulted
in the identification of lead molecule 62 (R = 2-fluoropyridin-3-
yl). In spite of high efflux (efflux ratio = 67), 62 demonstrated
the best coverage ratio of the various 7-THP chromans
evaluated in mouse PK. Furthermore, 62 significantly reduced
CSF Aβ1−40 in three different species (i.e., guinea pig, rat, and
cyno). Our initial criteria for selectivity against CatD had been
>100-fold, but to minimize risks associated with inhibition of
CatD in vivo,16 we chose to pursue compounds with even
greater selectivity. Guided by modeling and the SAR presented
in this work, we discovered a series of THP A-ring chromans
with much improved selectivity (>1000-fold against CatD),
which will be the subject of a future publication.
guinea
a
b
c
d
species
mouse
IV: 3
rat
IV: 3
pig
cyno
IV: 0.5
PO: 60 PO: 100
dose (mg/kg)
IV: 3
PO: 10 PO: 60
Fu,plasma
Fu,brain
0.20
0.14
36
0.22
0.12
35
0.37
0.077
9.7
0.39
0.13
5.8
g
CL
CLhep,int
h
(mL/min/kg)
CLhep,in vitro
CLhep,syst
38
23
17
25
41
30
110
17
22
F%
61
>100
3100
10
f
Cfree,plasma (nM)
1h
4h
1h
4h
1h
4h
1h
4h
300
78
140
3200
1700
e
e
2200
68
23
Cfree,brain (nM)
free b/p
24
19
e
e
15
66
34
0.080
0.19
0.022
0.030
290
0.14
e
e
e
1.5
f
CCSF (nM)
69
110
e
e
150
23
84
f
CSF/free plasma 1h
0.094
0.068
0.49
0.034
0.049
e
4h
1.0
a
iv formulation: 5 mM citrate in saline. po formulation: 1% CMC/
0.5% Tween/5 mM citrate in H2O. Male CD-1 mice: 3 animials/
cohort. iv formulation: 30% captisol/5 mM citrate in saline. po
formulation: 1% CMC/0.5% Tween/5 mM citrate in H2O. Male
Sprague−Dawley rats: 3 animals/cohort. iv formulation: 30%
captisol/5 mM citrate in saline. po formulation: 1% CMC/0.5%
Tween/5 mM citrate in H2O. Male Hartley guinea pigs: 4 animals/iv
route and 6 animals/po route. iv formulation: PEG400/EtOH/water
(40/10/50) solution. po formulation: 1% carboxymethylcellulose
sodium (CMC), 0.5% Tween 80, 5 mM citrate buffer, suspension.
Female cynos: 3 animals/dosing route. Measured at the 3 h time
point. Measured at the 2 h time point. Compounds (2 μM)
incubated with cryopreserved hepatocytes for 60 and 120 min at 37
b
c
d
e
f
g
h
°C. Compounds (1 μM) incubated with mouse liver microsomes for
EXPERIMENTAL SECTION
20 min at 37 °C in the presence of NADPH.
■
Chemistry. Column chromatography was done on a Biotage
system (Manufacturer: Dyax Corporation) having a silica gel column,
on a silica SepPak cartridge (Waters), or using silica gel 60 (EMD
brand, 230−400 mesh) (unless otherwise stated). Preparative TLC
purification was performed using PLC-type silica gel 60, F254, 20 × 20
cm plates (EMD brand; 0.5, 1.0, or 2.0 mm thickness). 1H NMR were
recorded on a Varian or Bruker instrument operating at 400 MHz. 1H
NMR spectra were obtained as CDCl3, CD3OD, D2O, (CD3)2SO,
(CD3)2CO, C6D6, and CD3CN solutions (reported in ppm), using
tetramethylsilane (0.00 ppm) or residual solvent (CDCl3, 7.26 ppm;
CD3OD, 3.31 ppm; D2O, 4.79; (CD3)2SO, 2.50 ppm; (CD3)2CO, 2.05
ppm; C6D6, 7.16 ppm; CD3CN, 1.94 ppm) as the reference standard.
When peak multiplicities are reported, the following abbreviations are
used: s (singlet), d (doublet), t (triplet), q (quartet), m (multiplet), br
(broadened), dd (doublet of doublets), and dt (doublet of triplets).
Coupling constants, when given, are reported in Hertz (Hz). All
temperatures are set forth in degrees Celsius. Low-resolution mass
spectral (MS) data were obtained on an Agilent Technologies 6120
and 58%, respectively) for both compounds corresponding with
their peak brain concentrations at the 3 h time point. In rat, 62
had a relatively constant free brain concentration between 46−
84 nM, which lead to a sustained PD effect out to the 12 h time
point. Even though 59 had ∼10-fold greater unbound plasma
levels than 62 in guinea pig, it had a lower free brain
concentration due in part to increased brain tissue binding
(Fu,brain = 0.021 for 59 vs 0.077 for 62). Despite a lower
coverage ratio, 59 demonstrated a comparable PD effect to 62
at both time points. For compound 62, coverage ratios
corresponding with maximum CSF Aβ1−40 reduction for guinea
pig and rat were significantly different (6.3 vs 15, respectively).
This discrepancy may be due to interspecies differences in the
APP sequence.29
K
dx.doi.org/10.1021/jm401635n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 7. PD in Guinea Pig and Rat
PO dose
(mg/kg)
time
point (h)
Cfree,plasma
(nM)
Cfree,brain
(nM)
b/p
free
CCSF
(nM)
coverage
a
CSF Aβ1−40
Reduction (%)
compd
formulation
species
ratio
c
62
1% CMC/0.5% Tween/5 mM
citrate in H2O
guinea
b
30
3
5.8
140
23
9.5
19
34
12
5.2
11
50
1.6
9.9
69
1.8
3.5
6.3
2.2
0.96
2.0
9.3
20
35
58
7.0
23
15
58
d
pig
60
1
0.14
1.5
3
23
5
7.0
1.7
6.9
1.6
2.6
56
f
8
<3.0
>1.7
>3.7
1.3
f
12
3
<3.0
c
100
38
e
c
rat
60
1
3100
2200
2700
1700
960
68
66
79
84
46
0.022
0.030
0.029
0.049
0.048
290
150
200
130
85
13
12
15
16
8.5
26
40
53
38
30
3
5
8
12
d
g
59
40/10/50 PEG/EtOH/H2O
guinea
b
60
3
260
79
12
0.046
0.085
90
40
1.6
52
15
c
g
pig
5
6.7
0.92
a
b
c
d
e
Coverage ratio = Cfree,brain/cell IC50. Male Hartley guinea pigs. Eight animals/cohort. Six animals/cohort. Male Sprague−Dawley rats.
f
g
Compound levels were BLQ. Three animals were used to determine brain concentration.
Figure 5. Cyno CSF concentration of Aβ1−40 (dotted line) and compound 62 (solid line). Po formulation: 1% carboxymethylcellulose sodium
(CMC), 0.5% Tween 80, 5 mM citrate buffer, suspension; 10 male cynomolgus monkeys, crossover design (n = 10 vehicle, n = 10 at 100 mg/kg).
series quadrupole LC−MS with UV detection at 220 and 254 nm and
multimode ionization using atmospheric-pressure chemical ionization
(APCI) or ESI. The reactions set forth below were done generally
under a positive pressure of nitrogen or argon or with a drying tube
(unless specified otherwise) in anhydrous solvents, and the reaction
flasks were typically fitted with rubber septa for the introduction of
substrates and reagents via syringe. Glassware was oven-dried and/or
heat dried using a heat gun while under vacuum. Reagents were
purchased from commercial suppliers such as Sigma-Aldrich, Alfa
Aesar, TCI, or Acros and were used without further purification unless
otherwise indicated. Anhydrous solvents (e.g., THF, DMF, DMA,
DMSO, MeOH, DCM, and toluene) were purchased from EMD
chemicals (DriSolv line) or from Sigma-Aldrich and used directly.
Unless indicated otherwise, final compounds were purified to ≥95%
purity as determined by high-performance liquid chromatography
(HPLC) or liquid chromatography mass spectrometry (LC−MS).
HPLC purity was measured using a Varian ProStar with UV detection
at 220 and 254 nm, a Waters YMC ODS-AQ C18 3.0 mm × 50 mm,
3.0 μm, 5−95% CH3CN in H2O with 0.1% NH4OAc for 5 min at 1.0
mL/min. LC−MS purity was measured using an Agilent Technologies
6120 series LC−MS with UV detection at 220 and 254 nm, a Halo
C18 2.1 mm × 30 mm, 2.7 μm, 5−95% CH3CN in H2O with 0.1%
NH4OAc for 3 min at 1.0 mL/min. pKa values were determined at
Analiza using a 24 point parallel capillary electrophoresis method with
quantitation by UV absorbance at 229 nm.
4-(3,6-Dihydro-2H-pyran-4-yl)morpholine (5). A mixture of
dihydro-2H-pyran-4(3H)-one 4 (100 g, 999 mmol) and morpholine
(131 mL, 1498 mmol) in toluene (333 mL) was refluxed under
Dean−Stark trap overnight. More than 1 equiv of water was collected.
This reaction mixture was then concentrated down to give crude 5
(169 g) as an oil, which was carried forward to the next step without
purification.
8-Bromo-4a-morpholino-1,3,4,4a,10,10a-hexahydropyrano-
[4,3-b]chromen-10-ol (7). A mixture of 5 (178.1 g, 1052 mmol) and
5-bromo-2-hydroxybenzaldehyde 6 (211.6 g, 1052 mmol) in toluene
(351 mL) was stirred overnight at rt. A white solid crashed out and
was filtered off. This was washed with toluene (50 mL). The solid
product was collected and dried to give 7 (306.8 g, 79% yield) as a 2:1
1
mixture of diastereomers, which were not separated. H NMR (400
MHz, CDCl3) isomer 1 (major diastereomer) δ 7.55 (m, 1H), 7.29
(dd, J = 3, 9 Hz, 1H), 6.71 (d, J = 9 Hz, 1H), 5.10 (d, J = 6 Hz, 1H),
4.08 (m, 1H), 3.96 (m, 1H), 3.61−3.85 (m, 2H), 3.58 (m, 4H), 3.31
(t, J = 11 Hz, 1H), 2.59−2.84 (m, 5H), 1.83 (m, 2H); isomer 2 (minor
diastereomer) δ 7.47 (m, 1H), 7.35 (dd, J = 3, 9 Hz, 1H), 6.79 (d, J =
L
dx.doi.org/10.1021/jm401635n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
9 Hz, 1H), 4.08 (m, 1H), 3.96 (m, 1H), 3.61−3.85 (m, 2H), 3.58 (m,
4H), 3.02 (t, J = 11 Hz, 1H), 2.59−2.84 (m, 5H), 2.00 (m, 2H), 1.89
(m, 1H).
8.72 (s, 1H), 7.42 (dd, J = 2, 9 Hz, 1H), 7.28 (d, J = 2 Hz, 1H), 6.89
(d, J = 9 Hz, 1H), 4.76 (td, J = 5, 11 Hz, 1H), 3.97 (m, 1H), 3.88 (m,
1H), 3.37 (m, 1H), 2.97 (t, J = 11 Hz, 1H), 2.88 (s, 3H), 2.13 (m,
2H), 1.68 (m, 1H).
8-Bromo-3,4-dihydropyrano[4,3-b]chromen-10(1H)-one (8).
DMSO (204 mL, 2878 mmol) was added dropwise to oxalyl chloride
(1726.9 mmol) in DCM (8 L) at −78 °C. This was added such that
the temperature did not rise above −65 °C. This was then stirred for
40 min at −78 °C. Compd 7 (533 g, 1439 mmol) was added as a solid
(temperature did not rise), and this was stirred for 2 h at −78 °C. The
solid did not fully go into solution. Triethylamine (602 mL, 4317
mmol) was added dropwise (some exotherm was seen; however, the
reaction temperature did not get above −65 °C). This was stirred for
30 min at −78 °C. During the entire course of the reaction, the
mixture was continually purged with N2, which exited the flask via a
line fed into a bleach trap. The mixture was then concentrated down.
Glacial acetic acid (1000 mL) was added to the mixture. The material
went into solution initially; however, after 5 min of stirring, product
began to crash out. The material was stirred overnight at rt. White
solid had crashed out and this was filtered. The white solid was washed
with glacial acetic acid (200 mL). This gave 8 (340.8 g, 84% yield) as a
(4R*,4a′S*,10a′S*)-8′-Bromo-1-methyl-2-thioxo-
3′,4′,4a′,10a′-tetrahydro-1′H-spiro[imidazolidine-4,10′-
pyrano[4,3-b]chromen]-5-one (12). A thick walled, glass pressure
vessel with 11 (15.9 g, 43.3 mmol), Lawesson’s reagent (10.5 g, 26.0
mmol), and toluene (150 mL) was charged. The mixture was degassed
with N2 for several minutes. The mixture was then heated to 90 °C for
15 h with stirring. Reaction had gone to approx 50% conversion by
HPLC and 1H NMR. Then more Lawesson’s reagent (3.5 g, 0.2 equiv)
was added and heated to 100 °C for an additional 22 h. HPLC
indicated >95% consumption of starting material. After cooling to rt,
much solid had formed. The suspension was cooled in a freezer at 5
°C for 2 h, then solids were filtered (14.9 g) and washed with toluene.
The mother liquor was saved. This solid (14.9 g) was mostly desired
1
product by H NMR, and it was partially purified by silica gel plug,
eluting with 10% Et2O in DCM. The mother liquor that had been
saved was partitioned between EtOAc (200 mL) and satd. aq.
NaHCO3 (200 mL). Phases were separated, and the aqueous phase
was re-extracted with EtOAc (150 mL). Combined organic phases
were washed with brine (200 mL), dried (MgSO4), filtered, and
1
white solid. H NMR (400 MHz, CDCl3) δ 8.31 (d, J = 3 Hz, 1H),
7.74 (dd, J = 2, 9 Hz, 1H), 7.32 (d, J = 9 Hz, 1H), 4.65 (m, 2H), 4.02
(m, 2H), 2.79 (m, 2H).
1
(4aS*,10aS*)-8-Bromo-1,4,4a,10a-tetrahydropyrano[4,3-b]-
chromen-10(3H)-one (9). L-Selectride (587 mL, 587 mmol, 1 M in
THF) was added to a mixture of 8 (150 g, 534 mmol) in DCM (2809
mL) at −78 °C. This was stirred for 45 min. TLC showed that the
reaction was complete. The mixture was placed in an ice bath.
Aqueous Rochelle’s salt (0.5 M) was added to the mixture as it was
warming to 0 °C. This was then worked up with EtOAc/water. The
organics were extracted twice, washed with brine, dried (Na2SO4), and
concentrated. The crude was then triturated with hexanes to give 9
concentrated to obtain 15.3 g of crude material. By H NMR, this
crude contained approx 20% desired product. To improve upon the
product yield, it was chromatographed on a Biotage Flash 75L system,
eluting with 5% Et2O/DCM isocratic. From this column recovered 2.1
g additional product, which was combined with the 13 g of product
purified by silica gel plug, to obtain 15.1 g (59% yield) of 12. 1H NMR
(400 MHz, (CD3)2SO) δ 10.76 (s, 1H), 7.47 (dd, J = 2, 9 Hz, 1H),
7.07 (d, J = 2 Hz, 1H), 6.93 (d, J = 9 Hz, 1H), 4.75 (td, J = 5, 11 Hz,
1H), 3.97 (m, 1H), 3.73 (m, 1H), 3.38 (m, 1H), 3.14 (s, 3H), 2.98 (t,
J = 11 Hz, 1H), 2.19 (m, 2H), 1.70 (m, 1H).
1
1
(100 g, 66% yield) in >95% diastereomeric purity by H NMR. H
NMR (400 MHz, CDCl3) δ 7.96 (d, J = 3 Hz, 1H), 7.57 (dd, J = 3, 9
Hz, 1H), 6.89 (d, J = 9 Hz, 1H), 4.51 (m, 1H), 4.32 (m, 1H), 4.12 (m,
1H), 3.45 (td, J = 3, 12 Hz, 1H), 3.38 (m, 1H), 2.83 (m, 1H), 2.05−
2.20 (m, 2H).
(4R*,4a′S*,10a′S*)-2-Amino-8′-bromo-1-methyl-
3′,4′,4a′,10a′-tetrahydro-1′H-spiro[imidazole-4,10′-pyrano-
[4,3-b]chromen]-5(1H)-one (13). A round bottomed flask plus stir
bar with 12 (15.1 g, 39.4 mmol), MeOH (200 mL), 70% aqueous t-
butyl hydroperoxide (38 mL, 276 mmol), and 30% aq. NH4OH (77
mL, 591 mmol) was charged. The mixture was heated to 50 °C for 2 h
with stirring. After cooling to rt, the mixture was diluted with water
(20 mL) and concentrated (but not to dryness) in vacuo. The residue
was diluted with EtOAc (150 mL), and phases were separated and re-
extracted with aqueous EtOAc (2 × 75 mL). Combined organic phases
were washed with brine (150 mL), dried (MgSO4), filtered, and
concentrated. Purified crude (15.6 g) by silica gel chromatography on
a Biotage Flash 75L system, eluting with 7−10% MeOH/DCM. A
total of 7.4 g (47%) of 13 as a white powde was obtained, in >95%
(4S*,4a′S*,10a′S*)-8′-Bromo-3′,4′,4a′,10a′-tetrahydro-1′H-
spiro[imidazolidine-4,10′-pyrano[4,3-b]chromene]-2,5-dione
(10). A mixture of 9 (75 g, 265 mmol), KCN (34.5 g, 530 mmol),
ammonium carbonate (204 g, 2119 mmol), and NaHSO3 (11.0 g, 106
mmol) in EtOH (265 mL) was heated to 130 °C overnight in a steel
bomb with stirring. The mixture was poured into an Erlenmeyer flask
with side arm in an ice bath. The flask was purged with N2, and the
outlet line was fed into a 2 N NaOH soln to quench HCN. To the
flask was carefully added concentrated HCl until the pH was about 1.
This was then stirred in an ice bath for 1 h while purging with N2. The
resulting white solid was filtered off and collected. This solid was dried
and then taken up in i-PrOH (500 mL) and heated at reflux for 30
min. This was then cooled to rt and then to 5 °C in an ice bath. The
white solid was filtered and washed with i-PrOH (50 mL) to give 10
1
1
diastereomeric purity by H NMR. H NMR (400 MHz, CDCl3) δ
7.27 (dd, J = 2, 9 Hz, 1H), 7.01 (d, J = 2 Hz, 1H), 6.76 (d, J = 9 Hz,
1H), 4.85 (td, J = 5, 11 Hz, 1H), 4.68 (br s, 2H), 4.05 (m, 1H), 3.94
(m, 1H), 3.45 (m, 1H), 3.05 (s, 3H), 3.00 (t, J = 11 Hz, 1H), 2.11 (m,
2H), 1.82 (m, 1H).
1
1
(44.8 g, 43% yield) in 84% diastereomeric purity by H NMR. H
NMR (400 MHz, (CD3)2SO) δ 11.14 (s, 1H), 8.45 (s, 1H), 7.44 (dd,
J = 2, 9 Hz, 1H), 7.22 (d, J = 2 Hz, 1H), 6.89 (d, J = 9 Hz, 1H), 4.74
(td, J = 5, 11 Hz, 1H), 3.97 (m, 2H), 3.40 (m, 1H), 3.06 (t, J = 11 Hz,
1H), 2.10 (m, 2H), 1.69 (m, 1H).
(4aS*,10aS*)-8-Bromo-10-methylene-1,3,4,4a,10,10a-
hexahydropyrano[4,3-b]chromene (14). A round bottomed flask
plus stir bar with 9 (544 mg, 1.92 mmol) and anhyd. THF (5 mL) was
charged and cooled to 0 °C under N2, and added Tebbe’s reagent (5.7
mL, 2.9 mmol, 0.5 M in toluene). This mixture was stirred for 2 h at rt.
Then, it was cooled in an ice bath and added MeOH (5 mL) very
carefully (vigorous exotherm and bubbling), then added aq. 2 N
NaOH (5 mL) dropwise. To enable stirring, DCM (10 mL) was
added. The biphasic suspension was stirred for 15 min at rt. The
biphase was filtered to remove solids through celite, rinsing with
DCM. Phases were separated and re-extracted with aqueous DCM (5
mL). Combined organics were washed with brine (20 mL), dried
(MgSO4), filtered, and concentrated. Purified crude by Biotage Flash
40 silica gel chromatography, eluting with 10−15% EtOAc/hexanes.
Yield of 14: 168 mg (30%). 1H NMR (400 MHz, CDCl3) δ 7.63 (d, J
= 2 Hz, 1H), 7.27 (dd, J = 2, 9 Hz, 1H), 6.74 (d, J = 9 Hz, 1H), 5.48
(d, J = 2 Hz, 1H), 4.71 (d, J = 2 Hz, 1H), 4.44 (m, 1H), 4.12 (m, 1H),
(4R*,4a′S*,10a′S*)-8′-Bromo-1-methyl-3′,4′,4a′,10a′-tetra-
hydro-1′H-spiro[imidazolidine-4,10′-pyrano[4,3-b]chromene]-
2,5-dione (11). A round bottomed flask plus stir bar was charged with
DMF (100 mL) and 10 (16.3 g, 46.2 mmol). The mixture was cooled
in an ice bath under N2, and K2CO3 (9.6 g, 69 mmol) and
iodomethane (2.9 mL, 46 mmol) were added sequentially. It was
stirred in the ice bath for 10 min, and then removed from the bath and
allowed to stir at rt for 2 h. The reaction mixture was diluted with
EtOAc (300 mL) and water (200 mL). The phases were separated and
re-extracted with aqueous EtOAc (150 mL). Combined organic phases
were washed with water (200 mL), brine (200 mL), dried (MgSO4),
filtered, and concentrated. Yield of 11: 15.9 g (94% yield). Product was
approximately 85% pure based on HPLC, and was carried forward to
the next step without purification. 1H NMR (400 MHz, (CD3)2SO) δ
M
dx.doi.org/10.1021/jm401635n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
3.79−3.89 (m, 1H), 3.48 (m, 1H), 3.35 (m, 1H), 2.51 (m, 1H), 2.09
(m, 1H), 1.96 (m, 1H).
chromatography on silica gel (2:1 hexanes/EtOAc) to give 21 (10.8
g, 65% yield) as a yellow solid. 1H NMR (400 MHz, CDCl3) δ 7.96 (d,
J = 3 Hz, 1H), 7.56 (dd, J = 3, 9 Hz, 1H), 6.84 (d, J = 9 Hz, 1H),
3.89−3.86 (m, 2H), 2.93−2.64 (m, 2H), 2.11−1.95 (m, 2H), 1.45 (s,
3H).
(4R*,4a′S*,10a′S*)-8′-Bromo-3′,4′,4a′,10a′-tetrahydro-
1′H,5H-spiro[oxazole-4,10′-pyrano[4,3-b]chromen]-2-amine
(15a) and (4R*,4a′R*,10a′R*)-8′-Bromo-3′,4′,4a′,10a′-tetrahy-
dro-1′H,5H-spiro[oxazole-4,10′-pyrano[4,3-b]chromen]-2-
amine (15b). This analogue was prepared from 14, as described for 8-
bromo-3,4,4a,10a-tetrahydro-1H,5′H-spiro[pyrano[4,3-b]chromene-
10,4′-thiazol]-2′-amine 16, replacing silver thiocyanate with silver
cyanate. Diastereomers were partially separated by preparative TLC (2
mm thickness), eluting with 10% MeOH/DCM. Mixed fractions were
repurified by Biotage Flash 40 column, eluting with a gradient of 1:1
EtOAc/hexanes, neat EtOAc, then 2.5%−5% MeOH/EtOAc. Yield of
15a: 62 mg (16%); 1H NMR (400 MHz, CDCl3) δ 7.42 (d, J = 2 Hz,
1H), 7.25 (dd, J = 2, 9 Hz, 1H), 6.67 (d, J = 9 Hz, 1H), 4.57 (br s,
2H), 4.39 (d, J = 9 Hz, 1H), 4.10 (m, 2H), 3.97 (d, J = 9 Hz, 1H), 3.46
(m, 1H), 3.23 (t, J = 11 Hz, 1H), 2.15 (m, 2H), 1.91 (m, 2H). Yield of
15b: 49 mg (15%); 1H NMR (400 MHz, CDCl3 + CD3OD) δ 7.31 (d,
J = 2 Hz, 1H), 7.26 (dd, J = 2, 9 Hz, 1H), 6.72 (d, J = 9 Hz, 1H), 4.54
(d, J = 9 Hz, 1H), 4.36 (d, J = 9 Hz, 1H), 4.19 (td, J = 5, 11 Hz, 1H),
4.05 (m, 2H), 3.52 (m, 1H), 3.37 (m, 1H), 2.16 (m, 1H), 1.88 (m,
2H).
8-Bromo-3,4,4a,10a-tetrahydro-1H,5′H-spiro[pyrano[4,3-b]-
chromene-10,4′-thiazol]-2′-amine (16). A stirred solution of 14
(168 mg, 0.598 mmol) in diethyl ether (2 mL) was cooled to 0 °C
under N2. In a separate flask, silver thiocyanate (397 mg, 2.39 mmol)
was suspended in CH3CN (1 mL), and to this suspension, iodine (303
mg, 1.20 mmol) in THF (1 mL) was added, and the resulting mixture
was shaken for 30 s. The suspension was then poured into the alkene
solution at 0 °C, and the vial was rinsed with CH3CN and added to
the reaction mixture. The reaction mixture was stirred at rt for 4 h.
Aqueous NH4OH (1 mL) was added directly to the reaction mixture,
and it was stirred at rt for 2 h, then left sitting for 18 h at rt. Filtered
salts through Celite, rinsing with EtOAc. The filtrate was washed with
aq. satd. Na2S2O3 (20 mL). After shaking and then separating the
phases, the aqueous layer was re-extracted with EtOAc (20 mL). The
combined organic layers were washed with brine (20 mL), dried
(MgSO4), filtered, and concentrated. Crude was partially purified by
preparative TLC (2 mm thickness) eluting with 10% MeOH/DCM
and then carried forward to the next step as a mixture of
diastereomers.
6-Bromo-2-(2-(methoxymethoxy)ethyl)-2-methylchroman-
4-one (22). To a solution of 21 (10.8 g, 38 mmol) and i-Pr2EtN (13.5
mL, 76 mmol) in DCM (30 mL) was added MOMCl (4.4 mL, 57
mmol) at 0 °C. The reaction mixture was stirred at rt for 12 h. The
reaction was then quenched with satd. aq. NH4Cl and extracted with
DCM. The organic layer was dried over Na2SO4, filtered, and
concentrated to give the crude product, which was purified by column
chromatography on silica gel (10:1 hexanes/EtOAc) to give 22 (10.1
1
g, 81%) as an oil. H NMR (400 MHz, CDCl3) δ 7.96 (d, J = 2 Hz,
1H), 7.55 (dd, J = 3, 9 Hz, 1H), 6.84 (d, J = 9 Hz, 1H), 4.59 (d, J = 0.8
Hz, 2H), 3.72−3.68 (m, 2H), 3.34 (s, 3H), 2.91−2.86 (m, 1H), 2.70−
2.66 (m, 1H), 2.12−1.99 (m, 2H), 1.44 (s, 3H).
((6-Bromo-2-(2-(methoxymethoxy)ethyl)-2-methyl-2H-chro-
men-4-yl)oxy)trimethylsilane (23). To a solution of 22 (20.0 g,
60.9 mmol) in THF (300 mL) was added dropwise LHMDS (66.9
mL, 66.9 mmol, 1 M in THF) at −70 °C. The mixture was stirred for
10 min. Then TMSCl (7.26 g, 66.8 mmol) was added to the mixture.
After 30 min, the reaction was quenched with 10% aq. NaHCO3 at
−70 °C, and extracted with DCM. The organic layer was dried over
Na2SO4, filtered, and concentrated to afford crude 23, which was
carried to the next step without further purification.
( 4 a S * , 1 0 a R * ) - 8 - B r o m o - 4 a - m e t h y l - 1 , 4 , 4 a , 1 0 a -
tetrahydropyrano[4,3-b]chromen-10(3H)-one (24). To a solu-
tion of TiCl4 (23.4 mL, 213 mmol) in DCM (200 mL) was added 23
in DCM (100 mL) at −40 °C. The mixture was stirred at −40 °C for
30 min. The mixture was then poured into satd. aq. NaHCO3,
extracted with DCM twice, washed with brine, and the organic layer
was then dried over Na2SO4, filtered, and concentrated to give the
crude product, which was purified by column chromatography on silica
gel (10:1 hexanes/EtOAc) to give 24 (12.0 g, 67% yield, 2 steps),
which was the pure cis stereoisomer at the ring fusion. 1H NMR (400
MHz, CDCl3) δ 7.96 (d, J = 2 Hz, 1H), 7.59 (dd, J = 2, 8 Hz, 1H),
6.90 (d, J = 9 Hz, 1H), 3.89−3.82 (m, 3H), 3.60 (t, J = 11 Hz, 1H),
2.71 (dd, J = 11, 5 Hz, 1H), 2.00−1.96 (dt, J = 14, 3 Hz, 1H), 1.78
(ddd, J = 14, 11, 6 Hz, 1H), 1.37 (s, 3H).
( 4 a S * , 1 0 a S * ) - 8 - B r o m o - 4 a - m e t h y l - 1 , 4 , 4 a , 1 0 a -
tetrahydropyrano[4,3-b]chromen-10(3H)-one (25). Compound
24 (7.0 g, 24 mmol) was dissolved in THF (150 mL) and cooled to
−78 °C. LHMDS (25 mL, 25 mmol, 1 M in THF) was then added
slowly via syringe. After stirring for 10 min at −78 °C, the enolate-
containing solution was added over 10 min via cannula to a solution of
ethyl 2-hydroxybenzoate (14 mL, 94 mmol) in THF (100 mL) also
cooled to −78 °C. After stirring for 10 min, the reaction mixture was
quenched with satd. aq. NH4Cl, partitioned between DCM and brine,
and the phases were then separated. The organic layer was dried over
Na2SO4, filtered, and concentrated. The crude material was then
purified by silica gel eluting with a linear gradient of 0−20% EtOAc/
heptane to provide a 5:1 mixture of trans/cis stereoisomers at the ring
fusion. Yield of 25: 6.3 g, 90%.
8-Bromo-2-methyl-2,3,4,6-tetrahydro-2,6-methanobenzo-
[b][1,5]dioxocin-4-ol (19). To a solution of 5-bromo-2-hydrox-
ybenzaldehyde 18 (30 g, 150 mmol) and Et3N (10.5 mL, 75 mmol) in
dioxane (56 mL) and H2O (19 mL) was added methylbut-2-enal 17
(12.6 g, 150 mmol). The mixture was stirred at 55 °C for 12 h.
Saturated aq. NaHCO3 was then added to the mixture and the
aqueous phase was extracted with EtOAc. The organic layer was dried
over Na2SO4, filtered, and concentrated to give the crude product,
which was purified by column chromatography on silica gel (eluting
with 1:1 EtOAc/hexanes) to give yellow solid 19 (26.5 g, 63%) as a
1
single unknown diastereomer. H NMR (400 MHz, CDCl3) δ 7.30−
7.26 (m, 2H), 6.69 (d, J = 10 Hz, 1H), 4.85−4.80 (m, 2H), 3.17 (d, J =
6 Hz, 1H), 2.20−2.10 (m, 2H), 1.81−1.77 (m, 1H), 1.60−1.55 (m,
1H), 1.45 (s, 3H).
(4aS*,10aS*)-8-Bromo-4a-methyl-10-methylene-
1,3,4,4a,10,10a-hexahydropyrano[4,3-b]chromene (26). Com-
pound 25 (2.45 g, 8.25 mmol) was dissolved in THF (25 mL) and
cooled to −78 °C. Tebbe’s reagent (24.7 mL, 12.4 mmol, 0.5 M in
toluene) was then added slowly via syringe. The reaction mixture was
allowed to warm to rt overnight. The next morning, the reaction
mixture was cooled to 0 °C and slowly quenched with MeOH (∼20
mL) followed by the addition of 1 N aq. NaOH (∼10 mL). After
stirring 15 min at rt, the mixture was diluted with DCM (200 mL) and
filtered through Celite. The filtrate was partitioned between additional
DCM and brine, phases separated, and the organic layer was then
dried over Na2SO4, filtered, and concentrated. The crude material was
purified by silica gel eluting with a linear gradient of 0−30% EtOAc/
heptane to provide 26 (1.47 g, 60% yield) as a 3:1 mixture of trans/cis
stereoisomers at the ring fusion.
6-Bromo-2-(2-hydroxyethyl)-2-methylchroman-4-ol (20). To
a solution of 19 (20.5 g, 71.8 mmol) in MeOH (60 mL) and THF
(240 mL) was added NaBH4 (1.4 g, 35.9 mmol). The mixture was
stirred at 0 °C for 2 h. The reaction was quenched with satd. aq.
NaHCO3 and extracted with DCM. The organic layer was dried over
Na2SO4, filtered, and concentrated to give white solid 20 (16.6 g, 81%)
1
as a single unknown diastereomer. H NMR (400 MHz, CDCl3) δ
7.58 (dd, J = 0.8, 2 Hz, 1H), 7.28−7.25 (m, 1H), 6.68 (d, J = 9 Hz,
1H), 4.84−4.81 (m, 1H), 3.90−3.82 (m, 2H), 2.16−1.90 (m, 6H),
1.35 (s, 3H).
6-Bromo-2-(2-hydroxyethyl)-2-methylchroman-4-one (21).
To a solution of 20 (16.6 g, 58 mmol) in DCM (300 mL) was
added MnO2 (25.2 g, 290 mmol). The reaction mixture was stirred at
rt for 12 h. The reaction was filtered, and the filtrate was concentrated
to give the crude product, which was purified by column
N
dx.doi.org/10.1021/jm401635n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
8′-Bromo-4a′-methyl-3′,4′,4a′,10a′-tetrahydro-1′H,5H-
spiro[oxazole-4,10′-pyrano[4,3-b]chromen]-2-amine (27). Io-
dine (1.38 g, 5.44 mmol) dissolved in EtOAc (39 mL) was added
over 5 min to a suspension of 26 (1.46 g, 4.95 mmol) and silver
cyanate (899 mg, 5.93 mmol) in EtOAc (6.0 mL) and CH3CN (13.2
mL) at 0 °C. The reaction mixture was then warmed to rt and allowed
to stir for 30 min and then heated at 40 °C for an additional 30 min.
The heterogeneous reaction mixture was then filtered through celite
and concentrated. The crude material was redissolved in THF (30
mL) and NH4OH solution (20 mL). The mixture was then heated at
40 °C for 1 h. The reaction mixture was then partitioned between
DCM and brine, phases separated, and the organic layer dried over
Na2SO4, filtered, and concentrated. The crude material was then
purified by silica gel eluting with a linear gradient of 0−6% MeOH/
DCM + 1% NH4OH solution to provide 27 (585 mg, 33% yield) as a
mixture of diastereomers, which was carried on to the next reaction
without additional purification.
A round bottomed flask plus stir bar with 41 (43 g, 127 mmol) and
THF (500 mL) was charged and cooled to −78 °C in a dry ice/
acetone bath. To the mixture was added DIBALH (101 mL, 152
mmol, 1.5 M in toluene) dropwise, and it was stirred at −78 °C for 1
h. Reaction remained a suspension the entire time. The reaction
mixture was quenched by inverse addition (via cannula) to Rochelle’s
salt (500 mL) and was stirred at rt. The mixture was worked up by
extraction with EtOAc (2 × 500 mL). The combined organics were
washed with brine (500 mL), dried (MgSO4), filtered, and
concentrated. The crude was purified by Biotage Flash 75 silica gel
chromatography, eluting with 5−10% EtOAc/hexanes. Yield of 42:
22.4 g (36%), which contained approximately 30% unreacted starting
1
material, carried forward without further purification. H NMR (400
MHz, CDCl3) δ 7.91 (d, J = 2 Hz, 1H), 7.56 (dd, J = 2, 9 Hz, 1H),
6.87 (d, J = 9 Hz, 1H), 4.47 (dd, J = 2, 10 Hz, 1H), 4.36 (m, 1H), 3.98
(m, 1H), 3.69 (m, 1H), 3.57 (m, 1H), 2.49 (m, 1H), 2.43 (m, 1H),
2.00 (m, 1H), 1.62 (d, J = 6 Hz, 3H), 1.26 (t, J = 7 Hz, 3H).
(1R*,4aR*,10aR*)-8-Bromo-1-methyl-1,4,4a,10a-
tetrahydropyrano[4,3-b]chromen-10(3H)-one (43). A round
bottomed flask plus stir bar with 42 (22.2 g, 65.1 mmol), DCM
(200 mL), and Et3SiH (51.8 mL, 325 mmol) was charged and cooled
in an ice bath under N2. Then BF3-Et2O (24.7 mL, 195 mmol) was
added dropwise and stirred overnight at rt. The mixture was carefully
quenched with satd. aq. NaHCO3 (200 mL) and stirred for 1 h. Phases
were separated and re-extracted with aqueous DCM (2 × 75 mL).
Combined organic phases were washed with brine (200 mL), dried
(MgSO4), filtered, and concentrated. The crude was purified by
Biotage Flash 65 silica gel chromatography, eluting with 10−20%
EtOAc/hexanes. Yield of 43: 13.6 g (60%). 1H NMR (400 MHz,
CDCl3) δ 7.91 (d, J = 2 Hz, 1H), 7.55 (dd, J = 2, 9 Hz, 1H), 6.86 (d, J
= 9 Hz, 1H), 4.35 (m, 1H), 4.07 (m, 1H), 3.64 (m, 1H), 3.49 (m, 1H),
2.52 (dd, J = 9, 13 Hz, 1H), 2.13 (m, 2H), 1.57 (d, J = 6 Hz, 3H).
(1′R*,4R*,4a′R*,10a′R*)-8′-Bromo-1,1′-dimethyl-
3′,4′,4a′,10a′-tetrahydro-1′H-spiro[imidazolidine-4,10′-
pyrano[4,3-b]chromene]-2,5-dione (44). A stainless steel bomb
plus stir bar with EtOH (10 mL) and 43 (3 g, 10 mmol) was charged.
Next ammonium carbonate (4.9 g, 50 mmol), KCN (1.3 g, 20 mmol)
and sodium hydrogensulfite (0.26 g, 2.5 mmol) were added. The
reaction was heated to 130 °C for 16 h with stirring in an oil bath.
After cooling to rt, reaction contents were transferred to an
Erlenmeyer flask using EtOAc (20 mL) and water (10 mL) to aid
in transfer. The mixture was chilled in an ice bath and carefully
acidified with conc. HCl, then bubbled N2 through mixture for 30 min
to sparge HCN (near back of hood). Phases were separated, and the
aqueous phase was re-extracted with EtOAc (2 × 10 mL). The
combined organic phases were washed with brine (50 mL), dried
(MgSO4), filtered, and concentrated. The crude was subjected to a
Biotage Flash 65 silica gel chromatography column, eluting with 5−
10% MeOH/DCM. A 60:40 mixture of diastereomers was obtained
and carried forward without separation as follows. A round bottomed
flask plus stir bar with DMF (10 mL) and the two diastereomers (1.6
g, 4.36 mmol) was charged and cooled in an ice bath under N2, then
K2CO3 (0.903 g, 6.54 mmol) and MeI (0.217 mL, 3.49 mmol) were
added sequentially. The mixture was stirred at rt overnight. The
mixture was diluted with EtOAc (20 mL) and water (20 mL). Phases
were separated and re-extracted with aqueous EtOAc (20 mL).
Combined organic phases were washed with water (20 mL), brine (20
mL), dried (MgSO4), filtered, and concentrated. Diastereomers were
separated by Biotage Flash 40 silica gel chromatography, eluting with a
gradient of 20% EtOAc/hexanes to 1:1 EtOAc/hexanes. Yield of 44:
554 mg (15%, 2 steps). 1H NMR (400 MHz, CDCl3) δ 7.35 (m, 1H),
6.99 (m, 1H), 6.79 (d, J = 9 Hz, 1H), 6.24 (br s, 1H), 4.06 (m, 1H),
3.85 (m, 1H), 3.52 (m, 1H), 3.16 (s, 3H), 2.36 (m, 1H), 2.21 (m, 1H),
2.00 (m, 1H), 1.27 (m, 1H), 1.11 (d, J = 6 Hz, 3H).
5-Bromo-4-fluoro-2-hydroxybenzaldehyde (28). To a solution
of 3-fluorophenol (100.0 g, 893 mmol) in anhydrous CH3CN (1.0 L)
was added MgCl2 (254.7 g, 2.7 mol) portionwise at 0 °C. Et3N (494
mL, 3.5 mol) was added dropwise to the mixture over 25 min,
followed by portionwise addition of paraformaldehyde (160.7 g, 5.3
mol). After complete addition, the mixture was heated at reflux for 3 h.
The mixture was cooled and quenched by the addition of cold conc.
HCl (347 mL) and extracted with EtOAc. The combined EtOAc
layers were washed with water and brine, dried over Na2SO4, filtered,
and concentrated to give 4-fluoro-2-hydroxybenzaldehyde (100.0 g,
80% yield) as a yellow oil, which was carried forward without
purification. To a solution of 4-fluoro-2-hydroxy-benzaldehyde (100.0
g, 714 mmol) in acetic acid (1.0 L) was added bromine (118.5 g, 750
mmol) dropwise at 10 °C. The mixture was stirred at rt overnight. To
the reaction mixture was added dropwise satd. aq. Na2SO3 slowly at 0
°C until the brown color disappeared. The reaction was then poured
into ice−water, the solid was collected by filtration and dried to give
28 (50.0 g, 26% yield, 2 steps) as a yellow solid. It was used in the next
1
step without purification. H NMR (400 MHz, CDCl3) δ 11.34 (br,
1H), 9.77 (s, 1 H), 7.74 (dd, J = 3, 7 Hz, 1H), 6.79 (d, J = 10 Hz, 1H).
(E)-1-(5-Bromo-2-hydroxyphenyl)-3-(dimethylamino)prop-
2-en-1-one (39). Similar to a published procedure for related
compounds,19 a mixture of 1-(5-bromo-2-hydroxyphenyl)ethanone 38
(50 g, 233 mmol) and DMF-dimethylacetal (42 g, 349 mmol) in dry
toluene (250 mL) was refluxed for 3 h. After cooling to rt, the mixture
was concentrated to half volume, and the resulting suspension was
cooled in an ice bath, then the solids were filtered and washed with
minimum amounts of toluene. Yield of 39: 56 g (87%). 1H NMR (400
MHz, CDCl3) δ 13.95 (s, 1H), 7.88 (d, J = 12 Hz, 1H), 7.76 (d, J = 2
Hz, 1H), 7.42 (dd, J = 2, 9 Hz, 1H), 6.83 (d, J = 9 Hz, 1H), 5.67 (d, J
= 12 Hz, 1H), 3.20 (s, 3H), 2.99 (s, 3H).
3-Acetyl-6-bromo-4H-chromen-4-one (40). Similar to a pub-
lished procedure for related compounds,19 to a solution of 39 (56 g,
207 mmol) in dry pyridine (84 mL) was added acetic anhydride (196
mL), and the mixture was stirred at rt for 18 h. The mixture was
concentrated on the rotovap to one-half volume at 80 °C. Cooled the
resulting suspension to rt, and then filtered the solids. Washed solids
1
with hexanes. Dried under high vacuum. Yield of 40: 48 g (85%). H
NMR (400 MHz, CDCl3) δ 8.59 (s, 1H), 8.39 (d, J = 2 Hz, 1H), 7.81
(dd, J = 2, 9 Hz, 1H), 7.42 (d, J = 9 Hz, 1H), 2.73 (s, 3H).
( 3 R * , 4 a R * ) - 8 - B r o m o - 3 - e t h o x y - 1 - m e t h y l - 4 , 4 a -
dihydropyrano[4,3-b]chromen-10(3H)-one (41). A stainless steel
bomb plus stir bar with ethyl vinyl ether (169 mL, 1760 mmol) and 40
(47 g, 176 mmol) was charged and heated to 100 °C for 15 h. After
cooling to rt, the reaction mixture was filtered, and the solids were
washed with a minimum amount of EtOAc. Yield of 41: 44 g (72%),
isolated as a single diastereomer.20 1H NMR (400 MHz, CDCl3) δ
8.05 (d, J = 2 Hz, 1H), 7.49 (dd, J = 2, 9 Hz, 1H), 6.83 (d, J = 9 Hz,
1H), 5.14 (dd, J = 2, 9 Hz, 1H), 5.07 (m, 1H), 3.99 (m, 1H), 3.67 (m,
1H), 2.53 (m, 1H), 2.38 (d, J = 2 Hz, 3H), 2.32 (m, 1H), 1.28 (t, J = 7
Hz, 3H).
(4aS*, 10aS*)-8-Bromo-10a-methyl-1, 4, 4a, 10a-
tetrahydropyrano[4,3-b]chromen-10(3H)-one (45). A round
bottomed flask plus stir bar with THF (25 mL) and i-Pr2NH (2.40
mL, 17.1 mmol) was charged. The reaction vessel was cooled to −78
°C (dry ice/acetone) under N2 and added n-BuLi (6.22 mL, 15.5
mmol, 2.5 M in hexanes). Next a solution of 9 (2.2 g, 7.77 mmol) in
(1R*,3R*,4aR*,10aR*)-8-Bromo-3-ethoxy-1-methyl-
1,4,4a,10a-tetrahydropyrano[4,3-b]chromen-10(3H)-one (42).
O
dx.doi.org/10.1021/jm401635n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
THF (10 mL) was added. The reaction vessel was transferred to a −10
°C ice/NaCl bath and stirred for 20 min in the bath. The vessel was
recooled to −78 °C and added HMPA (9.46 mL, 54.4 mmol) followed
by MeI (1.45 mL, 23.3 mmol) and stirred for 1 h at −78 °C. It was
quenched by the addition of satd. aq. NH4Cl (20 mL) at −78 °C. The
mixture was removed from dry ice bath and stirred for 15 min as
reaction mixture warmed. Phases were partitioned between EtOAc (50
mL) and water (50 mL). Phases were separated and re-extracted with
aqueous EtOAc (30 mL). Combined organic phases were washed with
water (50 mL), brine (50 mL), dried (MgSO4), filtered, and
concentrated. The crude was purified by Biotage Flash 40, eluting
with 10% EtOAc/hexanes. Yield of 45: 850 mg (31%). Isolated
sion of 50 (13.3 g, 45.3 mmol) in EtOH (100 mL) was treated with
Pd/C (10 wt %, 0.8 g) and shaken in a Parr shaker under H2 (50 psi)
for 3 h. The reaction was filtered through GF/F paper, and the filtrate
was concentrated. The solid was suspended in DCM (100 mL) and
stirred with MnO2 (7.9 g, 90.5 mmol) overnight. The crude was
filtered and concentrated to provide 51 (12.1 g) as a white solid, which
was carried forward crude without purification at this step.
( 4 a S *, 10 aS * ) - 7 - F l u o r o - 8 - m e t h o x y - 1 , 4 , 4 a , 1 0 a -
tetrahydropyrano[4,3-b]chromen-10(3H)-one (52). A solution of
51 (11.1 g, 37 mmol) in DCM (3 mL) was cooled to 0 °C and treated
with Et3SiH (18 mL, 112 mmol), then BF3−Et2O (9.2 mL, 75 mmol).
The reaction was allowed to stir at rt overnight. The reaction was
incomplete so an additional 3 equiv of Et3SiH and 2 equiv of BF3−
Et2O were added and the reaction continued to stir at rt. After 40 h,
the reaction mixture was dissolved in EtOAc (30 mL) and MeOH (5
mL), then quenched with satd. aq. NaHCO3 and stirred for 4 h. The
organic phase was separated and washed with brine. The organic phase
was dried (Na2SO4) and concentrated in vacuo to provide 52 (8.58 g),
which was carried forward crude without purification at this step.
2-Amino-7′-fluoro-8′-methoxy-1-methyl-3′,4′,4a′,10a′-tet-
rahydro-1′H-spiro[imidazole-4,10′-pyrano[4,3-b]chromen]-
5(1H)-one (53). Prepared from 52 using reaction conditions
described for the synthesis of compounds 10−13, crude 53 was
purified by silica gel chromatography (gradient: 0−10% (80% DCM/
19% MeOH/1% NH4OH)/DCM) to obtain a 60:40 mixture of trans/
cis stereoisomers at the ring fusion, which was carried forward without
1
1
product was >90% diastereomerically pure by H NMR. H NMR
(400 MHz, CDCl3) δ 7.96 (d, J = 2 Hz, 1H), 7.57 (dd, J = 2, 9 Hz,
1H), 6.89 (d, J = 9 Hz, 1H), 4.30 (dd, J = 5, 12 Hz, 1H), 4.09 (m, 2H),
3.43 (m, 2H), 2.22 (m, 1H), 1.93 (m, 1H), 1.26 (s, 3H).
3-Fluoro-4-methoxyphenyl acetate (47). A stirred solution of
1-(3-fluoro-4-methoxyphenyl)ethanone 46 (50 g, 297 mmol) in DCM
(1.2 L) was treated with m-CPBA (83.3 g, 372 mmol). The suspension
was heated to 40 °C with stirring, and the suspension became a yellow
solution. The reaction was stirred for 72 h at 40 °C, and TLC
suggested only partial conversion. The reaction was cooled to rt, and
an additional 80 g of m-CPBA was added in a single portion. The
reaction was returned to 40 °C, and the reaction stirred for an
additional 48 h. TLC confirmed conversion of starting material, and
the reaction was cooled to rt and washed with aq. 1 N NaOH,
repeating until organic phase was clear. The organic phase was then
washed with brine, dried (Na2SO4), and concentrated to a yellow oil.
Yield of 47: 47.7 g (87%). 1H NMR (400 MHz, CDCl3) δ 6.92 (d, J =
9 Hz, 1H), 6.83 (dd, J = 3, 11 Hz, 1H), 6.83 (m, 1H), 3.88 (s, 3H),
2.27 (s, 3H).
1-(4-Fluoro-2-hydroxy-5-methoxyphenyl)ethanone (48).
Neat TfOH (194 g, 1295 mmol) was added dropwise by addition
funnel into 47 (47.7 g, 259 mmol) with stirring at 0 °C. The reaction
was heated to 60 °C for 1 h and cooled to rt. The reaction was poured
carefully into an ice slurry (1 L). The resulting suspension was filtered,
and the solid was partitioned between Et2O and satd. aq. NaHCO3.
The organic phase was washed with brine, dried (Na2SO4), and
concentrated under vacuum. Yield of 48: 44.2 g, (93%) as a brown oil.
1H NMR (400 MHz, CDCl3) δ 12.31 (s, 1H), 7.23 (d, J = 9 Hz, 1H),
1
separation at this step. Yield: 1.70 g (12%, 4 steps). H NMR (400
MHz, CDCl3) trans isomer, δ 6.65 (d, J = 12 Hz, 1H), 6.46 (m, 1H),
5.32 (br, 2H), 4.82 (td, J = 5, 11 Hz, 1H), 4.02 (m, 1H), 3.92 (m, 1H),
3.76 (s, 3H), 3.44 (m, 1H), 3.07 (s, 3H), 3.00 (t, J = 11 Hz, 1H),
1.73−2.16 (m, 3H); cis isomer δ 6.71 (d, J = 12 Hz, 1H), 6.46 (m,
1H), 5.32 (br, 2H), 5.16 (m, 1H), 3.92 (m, 1H), 3.78 (m, 1H), 3.76 (s,
3H), 3.75 (m, 1H), 3.35 (t, J = 12 Hz, 1H), 3.08 (s, 3H), 1.73−2.16
(m, 3H).
(4R*,4a′S*,10a′S*)-2-((E)-((Dimethylamino)methylene)-
amino)-7′-fluoro-1-methyl-5-oxo-1,3′,4′,4a′,5,10a′-hexahydro-
1′H-spiro[imidazole-4,10′-pyrano[4,3-b]chromen]-8′-yl Tri-
fluoromethanesulfonate (54). Compound 53 (500 mg, 1.49
mmol) was treated with 48% aq. HBr (7.5 mL, 1.49 mmol) and
heated to 100 °C in a sealed vessel for 3 h. The reaction was cooled to
rt and treated with DCM (50 mL) and satd. aq. NaHCO3 until slightly
basic. The pH was brought to ∼5 by careful addition of aq. 1 N HCl.
The biphasic mixture was then concentrated in vacuo, and the residue
was triturated with 10% MeOH/DCM. The resulting suspension was
filtered, and the filtrate was concentrated to provide a crude oil, which
was then stirred with DMF-dimethylacetal (0.90 mL, 7.45 mmol) in
DMF (10 mL) at rt for 15 h. The reaction mixture was then
concentrated in vacuo to an oil. This oil was subsequently stirred with
Et3N (0.42 mL, 3.0 mmol) and DCM (7 mL) and treated with 1,1,1-
trifluoro-N-phenyl-N-(trifluoromethylsulfonyl)methanesulfonamide
(0.799 g, 2.24 mmol). The reaction was stirred for 15 h at rt. The
reaction mixture was washed with water and brine and dried over
Na2SO4. The crude was purified by silica gel chromatography to
provide 54 (90 mg, 12% yield, 3 steps). 1H NMR (400 MHz, CDCl3)
δ 8.59 (s, 1H), 7.17 (m, 1H), 6.74 (m, 1H), 4.98 (td, J = 5, 11 Hz,
1H), 4.05 (m, 1H), 3.85 (m, 1H), 3.46 (m, 1H), 3.18 (s, 3H), 3.17 (s,
3H), 3.07 (s, 3H), 3.01 (t, J = 11 Hz, 1H), 2.26 (m, 1H), 2.15 (m,
1H), 1.82 (m, 1H).
6.72 (d, J = 12 Hz, 1H), 3.89 (s, 3H), 2.60 (s, 3H).
7-Fluoro-6-methoxy-4-oxo-4H-chromene-3-carbaldehyde
(49). A solution of triphosgene (71.2 g, 240 mmol) in DCE (160 mL)
was added dropwise to a flask containing a mixture of DMF (254 mL,
2880 mmol) and DCE (300 mL) that was stirred in an ice bath. The
reaction temperature was maintained below 25 °C. After addition, the
reaction was cooled to 0 °C and treated with a solution of 48 (44.2 g,
240 mmol) in DCE (160 mL). The ice bath was removed and the
reaction was allowed to warm to rt while monitoring by HPLC. After 5
h of stirring at rt, the reaction was poured into a 2 L ice slurry and
stirred for an additional 2 h. The aqueous was extracted multiple times
with DCE, and the combined organic extracts were washed with satd.
aq. NaHCO3 and brine, and then the organic phase was concentrated
in vacuo. The resulting solid was placed in a vacuum oven and heated
1
to 60 °C overnight. Yield of 49: 28 g (53%) as a brown powder. H
NMR (400 MHz, CDCl3) δ 10.39 (s, 1H), 8.52 (s, 1H), 7.76 (d, J = 9
Hz, 1H), 7.29 (d, J = 10 Hz, 1H), 4.02 (s, 3H).
(3S*, 4aS*)-3-Ethoxy-7-fluoro-8-methoxy-4, 4a-
dihydropyrano[4,3-b]chromen-10(3H)-one (50). A stirred sus-
pension of 49 (31.4 g, 141 mmol) and ethyl vinyl ether (67.9 mL, 707
mmol) was heated at 100 °C in a Teflon-lined stainless steel bomb for
8 h with stirring. The heat was removed, and the reaction continued to
stir an additional 7 h at rt. The resulting residue was crystallized from
hot EtOH, and the solids were filtered. Yield of 50: 16.6 g (40%) as a
single diastereomer.20 1H NMR (400 MHz, (CD3)2SO) δ 7.57 (d, J =
1 Hz, 1H), 7.41 (d, J = 10 Hz, 1H), 7.04 (d, J = 12 Hz, 1H), 5.39 (dd,
J = 2, 10 Hz, 1H), 5.31 (m, 1H), 3.90 (m, 1H), 3.85 (s, 3H), 3.68 (m,
1H), 2.55 (m, 1H), 2.11 (m, 1H), 1.19 (t, J = 7 Hz, 3H).
(4R,4a′S,10a′S)-2-Amino-8′-bromo-1-methyl-3′,4′,4a′,10a′-
tetrahydro-1′H-spiro[imidazole-4,10′-pyrano[4,3-b]chromen]-
5(1H)-one (55). The title compound was obtained by resolution of
the racemate 13 using chiral SFC, performed as follows. System, Thar
350; column, Chiralpak IB, 5 cm × 25 cm, 5 μm; mobile phase A,
CO2; mobile phase B, 80% MeOH with 0.1% NH4OH, 20% DCM;
isocratic conditions, 65% A, 35% B; flow rate, 300 mL/min;
backpressure, 100 bar; temperature, 40 °C; UV detection, 220 nm;
racemic 13 was dissolved in a 50:35:15 mixture of MeOH/DCM/
formic acid, which was loaded on the column in 3 mL injections every
5.5 min. Compound 55 was >98% pure by LC−MS (254 nm) and
1
>98% ee by chiral SFC analysis. H NMR (400 MHz, CDCl3) δ 7.27
(3S*,4aS*,10aS*)-3-Ethoxy-7-fluoro-8-methoxy-1,4,4a,10a-
tetrahydropyrano[4,3-b]chromen-10(3H)-one (51). A suspen-
(dd, J = 2, 9 Hz, 1H), 7.01 (d, J = 2 Hz, 1H), 6.76 (d, J = 9 Hz, 1H),
P
dx.doi.org/10.1021/jm401635n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
4.84 (m, 1H), 4.68 (br s, 2H), 4.03 (m, 1H), 3.94 (m, 1H), 3.45 (m,
1H), 3.05 (s, 3H), 3.00 (t, J = 11 Hz, 1H), 2.11 (m, 2H), 1.82 (m,
1H); m/z (APCI-pos) M + 1 = 366, 368.
(0.17 g, 0.15 mmol), and 2 N aq. Na2CO3 (9.0 mL, 18 mmol) was
charged. The mixture was sparged with N2 for 15 min, then heated to
90 °C for 1 h with stirring. The reaction mixture was partitioned
between EtOAc (50 mL) and water (50 mL). Phases were separated
and re-extracted with aqueous EtOAc (2 × 30 mL). The combined
organic phases were washed with brine (50 mL), dried (MgSO4),
filtered, and concentrated. The crude was taken up in DCM (30 mL)
and concentrated again to remove residual solvents from the workup.
Then the crude solid was triturated in DCM (10 mL) at rt. The solids
were filtered, rinsing with DCM (3 × 5 mL). Compound 61 was
obtained as an off-white solid (1.7 g, 74% yield) in 96% purity (HPLC,
220 nm). 1H NMR (400 MHz, CDCl3) δ 8.05 (m, 1H), 7.75 (m, 1H),
7.36 (m, 1H), 7.18 (m, 1H), 7.14 (m, 1H), 6.98 (d, J = 9 Hz, 1H),
5.73 (br s, 2H), 4.95 (td, J = 5, 11 Hz, 1H), 4.05 (dd, J = 5, 12 Hz,
1H), 3.98 (dd, J = 4, 11 Hz, 1H), 3.47 (m, 1H), 3.04 (s, 3H), 3.03 (m,
1H), 2.18 (m, 2H), 1.83 (m, 1H); m/z (APCI-pos) M + 1 = 383.
(4R,4a′S,10a′S)-2-Amino-8′-(2-fluoropyridin-3-yl)-1-methyl-
3′,4′,4a′,10a′-tetrahydro-1′H-spiro[imidazole-4,10′-pyrano-
[4,3-b]chromen]-5(1H)-one (62) and (4S,4a′R,10a′R)-2-Amino-
8′-(2-fluoropyridin-3-yl)-1-methyl-3′,4′,4a′,10a′-tetrahydro-
1′H-spiro[imidazole-4,10′-pyrano[4,3-b]chromen]-5(1H)-one
(63). The title compounds were obtained by resolution of the
racemate 61 using chiral SFC, performed as follows. System, Thar 350;
column, Phenomenex Lux Cellulose-2, 3 cm × 25 cm, 5 μm; mobile
phase A, CO2; mobile phase B, MeOH with 0.1% NH4OH; isocratic
conditions, 50% A, 50% B; flow rate, 200 mL/min; backpressure, 100
bar; temperature, 40 °C; UV detection, 220 nm; racemic 61 was
dissolved in a 3:1 mixture of MeOH/formic acid, which was loaded on
the column in 3 mL injections every 4 min. Compounds 62 and 63
were 98% and 99% chemically pure, respectively (HPLC, 220 nm).
Both compounds were >98% ee by chiral SFC analysis. 1H NMR (400
MHz, CDCl3) δ 8.05 (m, 1H), 7.75 (m, 1H), 7.36 (m, 1H), 7.18 (m,
1H), 7.14 (m, 1H), 6.98 (d, J = 9 Hz, 1H), 5.73 (br s, 2H), 4.95 (td, J
= 5, 11 Hz, 1H), 4.05 (dd, J = 5, 12 Hz, 1H), 3.98 (dd, J = 4, 11 Hz,
1H), 3.47 (m, 1H), 3.04 (s, 3H), 3.03 (m, 1H), 2.18 (m, 2H), 1.83 (m,
1H); m/z (APCI-pos) M + 1 = 383.
(4R*,4a′S*,10a′S*)-2-Amino-8′-(3-methoxyphenyl)-1-meth-
yl-3′,4′,4a′,10a′-tetrahydro-1′H-spiro[imidazole-4,10′-pyrano-
[4,3-b]chromen]-5(1H)-one (56). Compound 56 was prepared from
13, as described for 61, replacing 2-fluoropyridin-3-ylboronic acid with
3-methoxyphenylboronic acid. Crude product was purified by
preparative TLC (0.5 mm thickness, Rf = 0.49) eluting with 10%
MeOH (containing 7N NH3)/DCM. Yield: 9 mg (41%). 99% pure by
1
LC−MS (220 nm). H NMR (400 MHz, CDCl3 + CD3OD) δ 7.42
(m, 1H), 7.31 (t, J = 8 Hz, 1H), 7.08 (m, 1H), 7.06 (m, 1H), 6.99 (m,
1H), 6.96 (d, J = 9 Hz, 1H), 6.85 (m, 1H), 4.94 (td, J = 5, 11 Hz, 1H),
4.08 (m, 1H), 3.95 (m, 1H), 3.84 (s, 3H), 3.51 (m, 1H), 3.08 (s, 3H),
3.06 (t, J = 11 Hz, 1H), 2.25 (m, 2H), 1.89 (m, 1H); m/z (APCI-pos)
M + 1 = 394.
(4R*,4a′S*,10a′S*)-2-Amino-8′-(3-(difluoromethoxy)-
phenyl)-1-methyl-3′,4′,4a′,10a′-tetrahydro-1′H-spiro-
[imidazole-4,10′-pyrano[4,3-b]chromen]-5(1H)-one (57). Com-
pound 57 was prepared from 13, as described for 61, replacing 2-
fluoropyridin-3-ylboronic acid with 3-(difluoromethoxy)phenylboronic
acid. Crude product was purified by preparative TLC (0.5 mm
thickness, Rf = 0.49) eluting with 10% MeOH (containing 7 N NH3)/
1
DCM. Yield: 10 mg (42%), >98% pure by LC−MS (220 nm). H
NMR (400 MHz, CDCl3 + CD3OD) δ 7.39 (dd, J = 2, 8 Hz, 1H), 7.37
(t, J = 8 Hz, 1H), 7.31 (m, 1H), 7.20 (m, 1H), 7.06 (d, J = 2 Hz, 1H),
7.05 (m, 1H), 6.97 (d, J = 9 Hz, 1H), 6.56 (t, J = 74 Hz, 1H), 4.94 (td,
J = 5, 11 Hz, 1H), 4.06 (m, 1H), 3.96 (m, 1H), 3.50 (m, 1H), 3.08 (s,
3H), 3.05 (t, J = 12 Hz, 1H), 2.27 (m, 2H), 1.89 (m, 1H); m/z (APCI-
pos) M + 1 = 430.
3-((4R*,4a′S*,10a′S*)-2-Amino-1-methyl-5-oxo-
1,3′,4′,4a′,5,10a′-hexahydro-1′H-spiro[imidazole-4,10′-
pyrano[4,3-b]chromen]-8′-yl)benzonitrile (58). Compound 58
was prepared from 13, as described for 61, replacing 2-fluoropyridin-3-
ylboronic acid with 3-cyanophenylboronic acid. Crude product was
purified by preparative TLC (0.5 mm thickness, Rf = 0.51) eluting with
10% MeOH (containing 7 N NH3)/DCM. Yield: 11 mg (50%), 97%
pure by LC−MS (220 nm). 1H NMR (400 MHz, CDCl3 + CD3OD) δ
7.74 (s, 1H), 7.70 (m, 1H), 7.57 (m, 1H), 7.51 (m, 1H), 7.41 (m, 1H),
7.06 (m, 1H), 7.00 (d, J = 9 Hz, 1H), 4.95 (td, J = 5, 11 Hz, 1H), 4.07
(m, 1H), 3.96 (m, 1H), 3.51 (m, 1H), 3.10 (s, 3H), 3.05 (t, J = 11 Hz,
1H), 2.24 (m, 2H), 1.91 (m, 1H); m/z (APCI-pos) M + 1 = 389.
3 - ( ( 4 R , 4 a ′ S , 1 0 a ′ S ) - 2 - A m i n o - 1 - m e t h y l - 5 - o x o -
1,3′,4′,4a′,5,10a′-hexahydro-1′H-spiro[imidazole-4,10′-
pyrano[4,3-b]chromen]-8′-yl)benzonitrile (59). Compound 59
was prepared from 55 (700 mg, 1.91 mmol), as described for 61,
replacing 2-fluoropyridin-3-ylboronic acid with 3-cyanophenylboronic
acid (421 mg, 2.87 mmol). Yield: 474 mg (63%), >98% pure by LC−
(4R*,4a′S*,10a′S*)-2-Amino-8′-(5-fluoropyridin-3-yl)-1-
methyl-3′,4′,4a′,10a′-tetrahydro-1′H-spiro[imidazole-4,10′-
pyrano[4,3-b]chromen]-5(1H)-one (64). Compound 64 was
prepared from 13, as described for 61, replacing 2-fluoropyridin-3-
ylboronic acid with 5-fluoropyridin-3-ylboronic acid. Crude product
was purified by preparative TLC (0.5 mm thickness, Rf = 0.29) eluting
with 10% MeOH (containing 7 N NH3)/DCM. Yield: 12 mg (55%),
96% pure by LC−MS (220 nm). 1H NMR (400 MHz, CDCl3
+
CD3OD) δ 8.50 (m, 1H), 8.34 (d, J = 3 Hz, 1H), 7.56 (m, 1H), 7.43
(dd, J = 2, 9 Hz, 1H), 7.09 (d, J = 2 Hz, 1H), 7.02 (d, J = 9 Hz, 1H),
4.97 (td, J = td, 1H), 4.09 (m, 1H), 3.96 (m, 1H), 3.51 (m, 1H), 3.10
(s, 3H), 3.05 (t, J = 11 Hz, 1H), 2.27 (m, 2H), 1.90 (m, 1H); m/z
(APCI-pos) M + 1 = 383.
1
MS (220 nm). H NMR (400 MHz, CDCl3 + CD3OD) δ 7.74 (s,
1H), 7.70 (m, 1H), 7.57 (m, 1H), 7.51 (m, 1H), 7.41 (m, 1H), 7.06
(m, 1H), 7.00 (d, J = 9 Hz, 1H), 4.95 (td, J = 5, 11 Hz, 1H), 4.07 (m,
1H), 3.96 (m, 1H), 3.51 (m, 1H), 3.10 (s, 3H), 3.05 (t, J = 11 Hz,
1H), 2.24 (m, 2H), 1.91 (m, 1H); m/z (APCI-pos) M + 1 = 389.
(4R*,4a′S*,10a′S*)-2-Amino-8′-(3-chloro-5-fluorophenyl)-1-
methyl-3′,4′,4a′,10a′-tetrahydro-1′H-spiro[imidazole-4,10′-
pyrano[4,3-b]chromen]-5(1H)-one (60). Compound 60 was
repared from 13, as described for 61, replacing 2-fluoropyridin-3-
ylboronic acid with 3-chloro-5-fluorophenylboronic acid. Crude
product was purified by preparative TLC (1 mm thickness, Rf =
0.29) eluting with 5% MeOH/EtOAc. Yield: 19 mg (16%), >98% pure
by LC−MS (220 nm). 1H NMR (400 MHz, CDCl3) δ 7.37 (dd, J = 2,
8 Hz, 1H), 7.22 (s, 1H), 7.05 (m, 2H), 7.00 (m, 1H), 6.97 (d, J = 9
Hz, 1H), 4.93 (td, J = 5, 11 Hz, 1H), 4.07 (dd, J = 5, 12 Hz, 1H), 3.99
(dd, J = 4, 11 Hz, 1H), 3.48 (td, J = 2, 13 Hz, 1H), 3.13 (s, 3H), 3.04
(t, J = 11 Hz, 1H), 3.03 (br s, 2H), 2.27 (td, J = 4, 11 Hz, 1H), 2.18
(m, 1H), 1.87 (m, 1H); m/z (APCI-pos) M + 1 = 416.
(4R*,4a′S*,10a′S*)-2-Amino-8′-(5-chloropyridin-3-yl)-1-
methyl-3′,4′,4a′,10a′-tetrahydro-1′H-spiro[imidazole-4,10′-
pyrano[4,3-b]chromen]-5(1H)-one (65). Compound 65 was
prepared from 13, as described for 61, replacing 2-fluoropyridin-3-
ylboronic acid with 5-chloropyridin-3-ylboronic acid. Crude product
was purified by preparative TLC (0.5 mm thickness, Rf = 0.50) eluting
with 10% MeOH (containing 7 N NH3)/DCM. Yield: 14 mg (50%),
97% pure by LC−MS (220 nm). 1H NMR (400 MHz, CDCl3) δ 8.56
(d, J = 2 Hz, 1H), 8.45 (d, J = 2 Hz, 1H), 7.70 (t, J = 2 Hz, 1H), 7.37
(dd, J = 2, 9H, 1H), 7.06 (d, J = 2 Hz, 1H), 6.99 (d, J = 9 Hz, 1H),
4.91 (td, J = 5, 11 Hz, 1H), 4.77 (br s, 2H), 4.03 (dd, J = 5, 11 Hz,
1H), 3.94 (dd, J = 4, 11 Hz, 1H), 3.46 (m, 1H), 3.08 (s, 3H), 3.02 (m,
1H), 2.13 (m, 2H), 1.82 (m, 1H); m/z (APCI-pos) M + 1 = 399.
(4R*,4a′S*,10a′S*)-2-Amino-8′-(5-methoxypyridin-3-yl)-1-
methyl-3′,4′,4a′,10a′-tetrahydro-1′H-spiro[imidazole-4,10′-
pyrano[4,3-b]chromen]-5(1H)-one (66). Compound 66 was
prepared from 13, as described for 61, replacing 2-fluoropyridin-3-
ylboronic acid with 5-methoxypyridin-3-ylboronic acid. Crude product
was purified by preparative TLC (0.5 mm thickness, Rf = 0.24) eluting
with 10% MeOH (containing 7 N NH3)/DCM. Yield: 10 mg (45%),
(4R*,4a′S*,10a′S*)-2-Amino-8′-(2-fluoropyridin-3-yl)-1-
methyl-3′,4′,4a′,10a′-tetrahydro-1′H-spiro[imidazole-4,10′-
pyrano[4,3-b]chromen]-5(1H)-one (61). A thick walled, glass
pressure vessel plus stir bar with 13 (2.2 g, 6.0 mmol), dioxane (30
mL), 2-fluoropyridin-3-ylboronic acid (1.3 g, 9.0 mmol), Pd[P(Ph)3]4
1
>98% pure by LC−MS (220 nm). H NMR (400 MHz, CDCl3 +
CD3OD) δ 8.25 (d, J = 2 Hz, 1H), 8.16 (d, J = 3 Hz, 1H), 7.41 (m,
Q
dx.doi.org/10.1021/jm401635n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
1H), 7.33 (m, 1H), 7.08 (d, J = 2 Hz, 1H), 7.01 (d, J = 9 Hz, 1H), 4.96
(td, J = 5, 11 Hz, 1H), 4.09 (m, 1H), 3.94 (m, 1H), 3.92 (s, 3H), 3.51
(m, 1H), 3.10 (s, 3H), 3.05 (t, J = 11 Hz, 1H), 2.27 (m, 2H), 1.88 (m,
1H); m/z (APCI-pos) M + 1 = 395.
5-((4R*,4a′S*,10a′S*)-2-Amino-1-methyl-5-oxo-
1,3′,4′,4a′,5,10a′-hexahydro-1′H-spiro[imidazole-4,10′-
pyrano[4,3-b]chromen]-8′-yl)nicotinonitrile (67). Compound 67
was prepared from 13, as described for 61, replacing 2-fluoropyridin-3-
ylboronic acid with 5-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)-
nicotinonitrile. Crude product was purified by preparative TLC (2
mm thickness, Rf = 0.56) eluting with 10% MeOH (containing 7 N
NH3)/DCM. Yield: 64 mg (59%), 98% pure by LC−MS (220 nm).
1H NMR (CDCl3, 400 MHz) δ 8.90 (m, 1H), 8.76 (m, 1H), 8.06 (m,
structures with BACE1 using the same conditions as described for
compound 62. Yield: 10 mg (67%), >98% purity by LC−MS (220
1
nm). H NMR (400 MHz, CDCl3 + CD3OD) δ 8.05 (m, 1H), 7.77
(m, 1H), 7.34 (m, 2H), 7.22 (m, 1H), 6.86 (d, J = 8 Hz, 1H), 4.55 (d,
J = 9 Hz, 1H), 4.34 (d, J = 9 Hz, 1H), 4.20 (m, 1H), 3.99 (m, 2H),
3.47 (m, 1H), 3.32 (m, 1H), 2.11 (m, 1H), 1.91 (m, 1H), 1.81 (m,
1H); m/z (APCI-pos) M + 1 = 356.
(4R*,4a′S*,10a′S*)-8′-(Pyrimidin-5-yl)-3′,4′,4a′,10a′-tetrahy-
dro-1′H,5H-spiro[oxazole-4,10′-pyrano[4,3-b]chromen]-2-
amine (73). Compound 73 was prepared from 15a, as described for
61, replacing 2-fluoropyridin-3-ylboronic acid with pyrimidin-5-
ylboronic acid. Crude product was purified by preparative TLC (0.5
mm thickness, Rf = 0.48) eluting with 10% MeOH (containing 7 N
NH3)/DCM. Yield: 7 mg (38%), >98% purity by LC−MS (220 nm).
1H NMR (400 MHz, CDCl3 + CD3OD) δ 9.03 (s, 1H), 8.84 (s, 2H),
7.47 (m, 1H), 7.34 (m, 1H), 6.88 (d, J = 9 Hz, 1H), 4.42 (d, J = 9 Hz,
1H), 4.03 (m, 3H), 3.97 (d, J = 9 Hz, 1H), 3.44 (m, 1H), 3.23 (m,
1H), 2.12 (m, 2H), 1.88 (m, 1H); m/z (APCI-pos) M + 1 = 339.
(4R*,4a′R*,10a′R*)-8′-(Pyrimidin-5-yl)-3′,4′,4a′,10a′-tetrahy-
dro-1′H,5H-spiro[oxazole-4,10′-pyrano[4,3-b]chromen]-2-
amine (74). Compound 74 was prepared from 15b, as described for
61, replacing 2-fluoropyridin-3-ylboronic acid with pyrimidin-5-
ylboronic acid. Crude product was purified by preparative TLC (0.5
mm thickness, Rf = 0.51) eluting with 10% MeOH (containing 7 N
NH3)/DCM. Yield: 5 mg (35%), >98% purity by LC−MS (220 nm).
1H NMR (400 MHz, CDCl3 + CD3OD) δ 9.12 (s, 1H), 8.90 (s, 2H),
1H), 7.42 (m, 1H), 7.08 (m, 1H), 7.03 (d, J = 9 Hz, 1H), 4.97 (m,
1H), 4.08 (m, 1H), 3.95 (m, 1H), 3.47 (d, J = 12 Hz, 1H), 3.09 (s,
3H), 3.03 (t, J = 12 Hz, 1H), 2.20 (m, 2H), 1.88 (m, 1H); m/z (APCI-
pos) M + 1 = 390.
(4R*,4a′S*,10a′S*)-2-Amino-1-methyl-8′-(pyrimidin-5-yl)-
3′,4′,4a′,10a′-tetrahydro-1′H-spiro[imidazole-4,10′-pyrano-
[4,3-b]chromen]-5(1H)-one (68). Compound 68 was prepared from
13, as described for 61, replacing 2-fluoropyridin-3-ylboronic acid with
pyrimidin-5-ylboronic acid. Crude product was purified by preparative
TLC (0.5 mm thickness, Rf = 0.33) eluting with 10% MeOH
(containing 7 N NH3)/DCM. Yield: 9 mg (44%), >98% pure by LC−
1
MS (220 nm). H NMR (400 MHz, CDCl3 + CD3OD) δ 9.10 (s,
1H), 8.87 (s, 2H), 7.46 (m, 1H), 7.12 (s, 1H), 7.06 (d, J = 9 Hz, 1H),
4.96 (td, J = 5, 11 Hz, 1H), 4.10 (m, 1H), 3.97 (m, 1H), 3.51 (m, 1H),
3.12 (s, 3H), 3.05 (t, J = 11 Hz, 1H), 2.27 (m, 2H), 1.91 (m, 1H); m/z
(APCI-pos) M + 1 = 366.
7.43 (m, 2H), 7.00 (d, J = 9 Hz, 1H), 4.64 (d, J = 9 Hz, 1H), 4.43 (d, J
= 9 Hz, 1H), 4.28 (m, 1H), 4.08 (m, 2H), 3.56 (m, 1H), 3.41 (m, 1H),
2.21 (m, 1H), 2.00 (m, 1H), 1.90 (m, 1H); m/z (APCI-pos) M + 1 =
339.
(4R*,4a′S*,10a′S*)-8′-(5-Chloropyridin-3-yl)-3′,4′,4a′,10a′-
tetrahydro-1′H,5H-spiro[oxazole-4,10′-pyrano[4,3-b]-
chromen]-2-amine (69). Compound 69 was prepared from 15a, as
described for 61, replacing 2-fluoropyridin-3-ylboronic acid with 5-
chloropyridin-3-ylboronic acid. Crude product was purified by
preparative TLC (1 mm thickness, Rf = 0.57) eluting with 10%
(4aR*,4′R*,10aR*)-8-(5-Chloropyridin-3-yl)-3,4,4a,10a-tetra-
hydro-1H,5′H-spiro[pyrano[4,3-b]chromene-10,4′-thiazol]-2′-
amine (75). Compound 75 was prepared from 16, as described for
61, replacing 2-fluoropyridin-3-ylboronic acid with 5-chloropyridin-3-
ylboronic acid. Crude product was purified by preparative TLC (1 mm
thickness, Rf = 0.66) eluting with 10% MeOH (containing 7 N NH3)/
DCM. Diastereomer 75 was repurified by a second preparative TLC
(0.5 mm thickness, Rf = 0.45) eluting with 10% MeOH/DCM. The
1
MeOH (containing 7 N NH3)/DCM. Yield: 12 mg (41%). H NMR
(400 MHz, CDCl3 + CD3OD) δ 8.47 (d, J = 2 Hz, 1H), 8.30 (d, J = 2
Hz, 1H), 7.75 (m, 1H), 7.38 (d, J = 2 Hz, 1H), 7.25 (m, 1H), 6.77 (d,
J = 8 Hz, 1H), 4.36 (d, J = 9 Hz, 1H), 3.95 (m, 3H), 3.90 (d, J = 9.0
Hz, 1H), 3.38 (m, 1H), 3.16 (d, J = 11 Hz, 1H), 2.04 (m, 2H), 1.78
(m, 1H); m/z (APCI-pos) M + 1 = 372; 93% pure by LC−MS (220
nm).
(4R*,4a′R*,10a′R*)-8′-(5-Chloropyridin-3-yl)-3′,4′,4a′,10a′-
tetrahydro-1′H,5H-spiro[oxazole-4,10′-pyrano[4,3-b]-
chromen]-2-amine (70). Compound 70 was prepared from 15b, as
described for 61, replacing 2-fluoropyridin-3-ylboronic acid with 5-
chloropyridin-3-ylboronic acid. Crude product was purified by
preparative TLC (1 mm thickness, Rf = 0.57) eluting with 10%
MeOH (containing 7 N NH3)/DCM. Yield: 9 mg (40%), >98% purity
by LC−MS (220 nm). 1H NMR (400 MHz, CDCl3 + CD3OD) δ 8.49
(d, J = 2 Hz, 1H), 8.36 (d, J = 2 Hz, 1H), 7.74 (t, J = 2 Hz, 1H), 7.29
(m, 2H), 6.85 (d, J = 9 Hz, 1H), 4.54 (d, J = 9 Hz, 1H), 4.32 (d, J = 9
Hz, 1H), 4.17 (td, J = 5, 11 Hz, 1H), 3.97 (m, 2H), 3.44 (m, 1H), 3.30
(d, J = 12 Hz, 1H), 2.09 (m, 1H), 1.85 (m, 2H); m/z (APCI-pos) M +
1 = 372.
1
minor diastereomer was not isolated. Yield of 75: 10 mg (21%). H
NMR (400 MHz, CDCl3) δ 8.65 (d, J = 2 Hz, 1H), 8.50 (d, J = 2 Hz,
1H), 7.77 (t, J = 2 Hz, 1H), 7.73 (d, J = 2 Hz, 1H), 7.38 (dd, J = 2, 9
Hz, 1H), 6.95 (d, J = 9 Hz, 1H), 4.63 (br s, 2H), 4.36 (m, 1H), 4.14
(m, 1H), 4.09 (m, 1H), 3.87 (d, J = 12 Hz, 1H), 3.76 (d, J = 12 Hz,
1H), 3.53 (m, 1H), 3.39 (m, 1H), 2.17 (m, 1H), 2.05 (m, 1H), 1.90
(m, 1H); m/z (APCI-pos) M + 1 = 388; 88% purity by LC−MS (220
nm).
(4aR*,4′R*,10aR*)-8-(2-Fluoropyridin-3-yl)-3,4,4a,10a-tetra-
hydro-1H,5′H-spiro[pyrano[4,3-b]chromene-10,4′-thiazol]-2′-
amine (76). Compound 76 was prepared from 16, as described for
61. Crude product was purified by preparative TLC (1 mm thickness,
Rf = 0.68) eluting with 10% MeOH (containing 7 N NH3)/DCM. The
alternate diastereomer was not isolated. Relative stereochemistry was
confirmed from cocrystal structures with BACE1 using the same
1
conditions as described for compound 62. Yield: 9 mg (21%). H
NMR (400 MHz, CDCl3) δ 8.14 (d, J = 5 Hz, 1H), 7.80 (m, 1H), 7.75
(m, 1H), 7.36 (m, 1H), 7.24 (m, 1H), 6.93 (d, J = 9 Hz, 1H), 4.56 (br
s, 2H), 4.37 (m, 1H), 4.16 (m, 1H), 4.08 (m, 1H), 3.90 (d, J = 12 Hz,
1H), 3.75 (d, J = 12 Hz, 1H), 3.53 (m, 1H), 3.39 (t, J = 11 Hz, 1H),
2.19 (m, 1H), 2.05 (m, 1H), 1.92 (m, 1H); m/z (APCI-pos) M + 1 =
372; 92% purity by LC−MS (220 nm).
3-((4R,4a′S,10a′S)-2-Amino-4a′-methyl-3′,4′,4a′,10a′-tetra-
hydro-1′H,5H-spiro[oxazole-4,10′-pyrano[4,3-b]chromene]-8′-
yl)benzonitrile (77). Compound 77 was prepared from 27 as
described for 79, replacing (5-chloro-3-pyridyl)boronic acid with (3-
cyano-phenyl)boronic acid. Purification by silica gel provided a
mixture of product diastereomers. The mixture was further purified
by chiral SFC on a Chiralpak AD (2 × 15 cm) column eluting with
25% MeOH (0.1% NH4OH)/CO2 at 100 bar at a flow rate of 70 mL/
min. The peaks isolated were analyzed on a Chiralpak AD (50 × 0.46
cm) column eluting with 25% MeOH (0.1% NH4OH)/CO2, at 120
bar (flow rate 5 mL/min, 220 nm). From this purification, 77 (peak-3,
(4R*,4a′S*,10a′S*)-8′-(2-Fluoropyridin-3-yl)-3′,4′,4a′,10a′-
tetrahydro-1′H,5H-spiro[oxazole-4,10′-pyrano[4,3-b]-
chromen]-2-amine (71). Compound 71 was prepared from 15a, as
described for 61. Crude product was purified by preparative TLC (0.5
mm thickness, Rf = 0.65) eluting with 10% MeOH (containing 7 N
1
NH3)/DCM. Yield: 7 mg (36%), 97% pure by LC-MS (220 nm). H
NMR (400 MHz, CDCl3 + CD3OD) δ 8.12 (m, 1H), 7.88 (m, 1H),
7.51 (m, 1H), 7.41 (m, 1H), 7.29 (m, 1H), 6.90 (d, J = 9 Hz, 1H),
4.48 (d, J = 9 Hz, 1H), 4.02−4.14 (m, 4H), 3.52 (m, 1H), 3.30 (m,
1H), 2.20 (m, 2H), 1.97 (m, 1H); m/z (APCI-pos) M + 1 = 356.
(4R*,4a′R*,10a′R*)-8′-(2-Fluoropyridin-3-yl)-3′,4′,4a′,10a′-
tetrahydro-1′H,5H-spiro[oxazole-4,10′-pyrano[4,3-b]-
chromen]-2-amine (72). Compound 72 was prepared from 15b, as
described for 61. Crude product was purified by preparative TLC (0.5
mm thickness, Rf = 0.62) eluting with 10% MeOH (containing 7 N
NH3)/DCM. Relative stereochemistry was confirmed from cocrystal
R
dx.doi.org/10.1021/jm401635n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
1
8.0 mg, 8% yield, chemical purity >99%, > 99% ee) was isolated. H
NMR (400 MHz, DMSO-d6) δ 8.05−8.01 (m, 1H), 7.88 (d, J = 8.0
Hz, 1H), 7.76 (d, J = 7.6 Hz, 1H), 7.64 (t, J = 7.8 Hz, 1H), 7.51 (dd, J
= 8.4, 2.4 Hz, 1H), 7.39 (d, J = 2.4 Hz, 1H), 6.84 (d, J = 8.5 Hz, 1H),
6.13 (s, 2H), 4.46 (d, J = 8.9 Hz, 1H), 4.11 (d, J = 9.0 Hz, 1H), 3.92
(dd, J = 11.9, 4.5 Hz, 1H), 3.71 (dd, J = 11.2, 4.3 Hz, 1H), 3.55 (td, J =
11.9, 2.7 Hz, 2H), 2.19 (dd, J = 11.3, 4.2 Hz, 1H), 1.89 (td, J = 12.8,
5.0 Hz, 1H), 1.80 (d, J = 12.5 Hz, 1H), 1.32 (s, 3H); m/z (ESI-pos) M
+ 1 = 376.2.
(m, 1H), 8.02 (ddd, J = 10.0, 7.5, 1.9 Hz, 1H), 7.44 (ddd, J = 7.0, 4.7,
1.9 Hz, 1H), 7.39 (dt, J = 8.5, 2.0 Hz, 1H), 7.34−7.27 (m, 1H), 6.85
(d, 1H), 6.13 (s, 2H), 4.44 (d, J = 8.9 Hz, 1H), 4.04 (d, J = 8.9 Hz,
1H), 3.92 (dd, J = 11.8, 4.5 Hz, 1H), 3.71 (dd, J = 11.3, 4.3 Hz, 1H),
3.61−3.49 (m, 2H), 2.22 (dd, J = 11.3, 4.2 Hz, 1H), 1.90 (td, J = 12.8,
5.2 Hz, 1H), 1.84−1.73 (m, 1H), 1.32 (s, 3H). m/z (ESI-pos) M + 1 =
370.1.
(4R*,4a′S*,10a′S*)-4a′-Methyl-8′-(pyrimidin-5-yl)-
3′,4′,4a′,10a′-tetrahydro-1′H,5H-spiro[oxazole-4,10′-pyrano-
[4,3-b]chromen]-2-amine (82). Compound 82 was prepared from
27 as described for 79, replacing (5-chloro-3-pyridyl)boronic acid with
pyrimidin-5-ylboronic acid. Yield: 18.3 mg (18%). 1H NMR (400
MHz, DMSO-d6) δ 9.13 (s, 1H), 9.02 (s, 2H), 7.57 (dd, J = 8.4, 2.4
Hz, 1H), 7.44 (d, J = 2.3 Hz, 1H), 6.89 (d, J = 8.4 Hz, 1H), 6.13 (s,
2H), 4.46 (d, J = 8.9 Hz, 1H), 4.12 (d, J = 8.9 Hz, 1H), 3.92 (dd, J =
11.7, 4.7 Hz, 1H), 3.71 (dd, J = 11.3, 4.3 Hz, 1H), 3.55 (td, J = 10.8,
10.1, 2.4 Hz, 2H), 2.20 (dd, J = 11.3, 4.3 Hz, 1H), 1.90 (td, J = 12.7,
5.1 Hz, 1H), 1.81 (d, J = 12.4 Hz, 1H), 1.33 (s, 3H). m/z (ESI-pos) M
+ 1 = 353.1. HPLC purity: 87% (254 nm).
(4R,4a′S,10a′S)-8′-(3-Chloro-5-fluorophenyl)-4a′-methyl-
3′,4′,4a′,10a′-tetrahydro-1′H,5H-spiro[oxazole-4,10′-pyrano-
[4,3-b]chromen]-2-amine (78). Compound 78 was prepared from
8′-bromo-4a′-methyl-3′,4′,4a′,10a′-tetrahydro-1′H,5H-spiro[oxazole-
4,10′-pyrano[4,3-b]chromen]-2-amine 27, as described for 79,
replacing (5-chloro-3-pyridyl)boronic acid with (3-chloro-5-fluoro-
phenyl)boronic acid. Purification by silica gel provided a mixture of
diastereomers. The mixture was further purified by chiral SFC on a
Chiralpak AD (2 × 15 cm) column eluting with 25% MeOH (0.1%
NH4OH)/CO2 at 100 bar at a flow rate of 70 mL/min. The peaks
isolated were analyzed on Chiralpak AD (50 × 0.46 cm) column
eluting with 25% MeOH (0.1% NH4OH)/CO2, at 120 bar (flow rate 5
mL/min, 220 nm). From the chiral SFC separation, 78 (peak-4, 43
(1′R*,4R*,4a′R*,10a′R*)-2-Amino-8′-(2-fluoropyridin-3-yl)-
1,1′-dimethyl-3′,4′,4a′,10a′-tetrahydro-1′H-spiro[imidazole-
4,10′-pyrano[4,3-b]chromen]-5(1H)-one (83). Compound 83 was
prepared from 44 in three steps using reaction conditions described
for the synthesis of compounds 12, 13, and 61. Crude 83 was purified
by preparative TLC (2 mm thickness, Rf = 0.44) eluting with 10%
MeOH (containing 7 N NH3)/DCM. Then the resulting product was
triturated with a minimum amount of DCM and filtered. Relative
stereochemistry was confirmed from cocrystal structures with BACE1
using the same conditions as described for compound 62. Yield: 23 mg
1
mg, 15% yield, chemical purity >99%, > 99% ee) was isolated. H
NMR (400 MHz, (CD3)2SO) δ 7.51 (dd, J = 7, 3 Hz, 1H), 7.42 (ddd,
J = 9, 4, 3 Hz, 1H), 7.34 (dd, J = 13, 6 Hz, 2H), 7.26 (s, 1H), 6.83 (d, J
= 9 Hz, 1H), 6.13 (s, 2H), 4.44 (d, J = 9 Hz, 1H), 4.06 (d, J = 9 Hz,
1H), 3.92 (dd, J = 12, 4 Hz, 1H), 3.71 (dd, J = 11, 4 Hz, 1H), 3.54 (t, J
= 11 Hz, 2H), 2.21 (dd, J = 11, 4 Hz, 1H), 1.96 − 1.83 (m, 1H), 1.80
(d, J = 13 Hz, 1H), 1.32 (s, 3H); m/z (ESI-pos) M + 1 = 403.0.
(4R*,4a′S*,10a′S*)-8′-(5-Chloropyridin-3-yl)-4a′-methyl-
3′,4′,4a′,10a′-tetrahydro-1′H,5H-spiro[oxazole-4,10′-pyrano-
[4,3-b]chromen]-2-amine (79). Diastereomers 27 (95 mg, 0.269
mmol), (5-chloro-3-pyridyl)boronic acid (55 mg, 0.350 mmol),
Na2CO3 (86 mg, 0.807 mmol), and Pd[P(Ph)3]4 (31.1 mg, 0.0269
mmol) were added as solids to a vial, followed by the simultaneous
addition of dioxane (4.6 mL) and degassed water (0.5 mL). The vial
was sealed and heated at 90 °C for 4 h. The reaction mixture was then
cooled and diluted with DCM and a mixture of brine and NH4OH
solution. The organic layer was dried over Na2SO4, filtered, and
concentrated. The crude material was purified by silica gel eluting with
a linear gradient of 0−6% MeOH/DCM + 1% NH4OH solution to
provide 79 (17.5 mg, 17% yield) as the second eluting UV active
fraction. Relative stereochemistry was confirmed from cocrystal
structures with BACE1 using the same conditions as described for
1
(7%, 3 steps). H NMR (400 MHz, CDCl3 + CD3OD) δ 8.11 (m,
1H), 7.78 (m, 1H), 7.37 (m, 1H), 7.25 (m, 1H), 6.97 (d, J = 9 Hz,
1H), 6.93 (m, 1H), 4.18 (m, 1H), 4.02 (m, 1H), 3.55 (m, 1H), 3.46
(m, 1H), 3.18 (s, 3H), 2.25 (m, 2H), 2.00 (m, 1H), 1.03 (d, J = 6 Hz,
3H); m/z (APCI-pos) M + 1 = 397; 94% purity by LC−MS (220
nm).
(1′R*,4R*,4a′R*,10a′R*)-8′-(2-Fluoropyridin-3-yl)-1′-methyl-
3′,4′,4a′,10a′-tetrahydro-1′H,5H-spiro[oxazole-4,10′-pyrano-
[4,3-b]chromen]-2-amine (84). Compound 84 was prepared from
43 in three steps using reaction conditions described for the synthesis
of compounds 14, 15, and 61. Diastereomers were partially separated
by preparative TLC (1 mm thickness) eluting with 10% MeOH/
DCM. Then 84 was repurified by preparative TLC (1 mm thickness)
eluting with 7.5% MeOH (containing 7 N NH3)/DCM. Relative
stereochemistry was confirmed from cocrystal structures with BACE1
using the same conditions as described for compound 62. Yield: 13 mg
1
compound 62. HPLC purity: 100% (254 nm). H NMR (400 MHz,
1
(6%, 3 steps). H NMR (400 MHz, CDCl3 + CD3OD) δ 8.14 (m,
(CD3)2SO) δ 8.74 (d, J = 2 Hz, 1H), 8.56 (d, J = 2 Hz, 1H), 8.12 (t, J
= 2 Hz, 1H), 7.55 (dd, J = 9, 2 Hz, 1H), 7.42 (d, J = 2 Hz, 1H), 6.85
(d, J = 9 Hz, 1H), 6.13 (s, 2H), 4.46 (d, J = 9 Hz, 1H), 4.13 (d, J = 9
Hz, 1H), 3.92 (dd, J = 11, 4 Hz, 1H), 3.71 (dd, J = 11, 4 Hz, 1H), 3.55
(t, J = 11 Hz, 2H), 2.19 (dd, J = 11, 4 Hz, 1H), 1.88 (dd, J = 13, 5 Hz,
1H), 1.80 (d, J = 13 Hz, 1H), 1.32 (s, 3H); m/z (ESI-pos) M + 1 =
386.1.
1H), 8.13 (m, 1H), 7.86 (m, 1H), 7.38 (m, 1H), 7.31 (m, 1H), 6.92
(d, J = 8 Hz, 1H), 4.56 (d, J = 9 Hz, 1H), 4.51 (d, J = 9 Hz, 1H), 4.07
(m, 2H), 3.59 (m, 2H), 2.19 (m, 1H), 1.95 (m, 1H), 1.76 (t, J = 10
Hz, 1H), 1.35 (d, J = 6 Hz, 3H); m/z (APCI-pos) M + 1 = 370; 91%
purity by LC−MS (220 nm).
(1R*,4aR*,4′R*,10aR*)-8-(2-Fluoropyridin-3-yl)-1-methyl-
3,4,4a,10a-tetrahydro-1H,5′H-spiro[pyrano[4,3-b]chromene-
10,4′-thiazol]-2′-amine (85). Compound 85 was prepared from 43
in three steps using reaction conditions described for the synthesis of
compounds 14, 16, and 61. Diastereomers were separated by
preparative TLC (2 mm thickness) eluting with 10% MeOH/DCM
+ 1% HOAc. Compound 85 was then triturated with 1:1 DCM/Et2O
and filtered. Relative stereochemistry was confirmed from cocrystal
structures with BACE1 using the same conditions as described for
compound 62. Yield: 53 mg (8%, 3 steps), >98% purity by LC−MS
(4R*,4a′S*,10a′S*)-8′-(5-Fluoropyridin-3-yl)-4a′-methyl-
3′,4′,4a′,10a′-tetrahydro-1′H,5H-spiro[oxazole-4,10′-pyrano-
[4,3-b]chromen]-2-amine (80). Compound 80 was prepared from
27 as described for 79, replacing (5-chloro-3-pyridyl)boronic acid with
1
(5-fluoropyridin-3-yl)boronic acid. Yield: 29.4 mg (30%). H NMR
(400 MHz, DMSO-d6) δ 8.69−8.65 (m, 1H), 8.51 (d, J = 2.7 Hz, 1H),
7.93 (dt, J = 10.5, 2.3 Hz, 1H), 7.54 (dd, J = 8.5, 2.4 Hz, 1H), 7.42 (d,
J = 2.4 Hz, 1H), 6.86 (d, J = 8.5 Hz, 1H), 6.13 (s, 2H), 4.46 (d, J = 9.0
Hz, 1H), 4.12 (d, J = 8.9 Hz, 1H), 3.92 (dd, J = 11.7, 4.6 Hz, 1H), 3.71
(dd, J = 11.3, 4.3 Hz, 1H), 3.55 (td, J = 11.6, 3.0 Hz, 2H), 2.19 (dd, J =
11.3, 4.3 Hz, 1H), 1.90 (td, J = 12.6, 5.0 Hz, 1H), 1.80 (d, J = 12.4 Hz,
1H), 1.32 (s, 3H). m/z (ESI-pos) M + 1 = 370.2. HPLC purity: 88%
(254 nm).
1
(220 nm). H NMR (CDCl3 + CD3OD) δ 8.13 (m, 1H), 7.87 (m,
1H), 7.81 (s, 1H), 7.34 (m, 2H), 6.93 (d, J = 8 Hz, 1H), 4.14 (m, 1H),
4.04 (m, 1H), 3.75 (m, 2H), 3.61 (m, 1H), 3.53 (m, 1H), 2.17 (m,
1H), 1.95 (m, 1H), 1.82 (m, 1H), 1.43 (d, J = 5 Hz, 3H); m/z (APCI-
pos) M + 1 = 386.
3-((4R*,4a′S*,10a′S*)-2-Amino-1,10a′-dimethyl-5-oxo-
1,3′,4′,4a′,5,10a′-hexahydro-1′H-spiro[imidazole-4,10′-
pyrano[4,3-b]chromen]-8′-yl)benzonitrile (86). Compound 86
was prepared from 45 using reaction conditions described for the
synthesis of compounds 10−13 and 61, and replacing 2-fluoropyridin-
(4R*,4a′S*,10a′S*)-8′-(2-Fluoropyridin-3-yl)-4a′-methyl-
3′,4′,4a′,10a′-tetrahydro-1′H,5H-spiro[oxazole-4,10′-pyrano-
[4,3-b]chromen]-2-amine (81). Compound 81 was prepared from
27 as described for 79, replacing (5-chloro-3-pyridyl)boronic acid with
(2-fluoropyridin-3-yl)boronic acid. Yield: 23.0 mg (20%). HPLC
1
purity: 99% (254 nm). H NMR (400 MHz, DMSO-d6) δ 8.20−8.15
S
dx.doi.org/10.1021/jm401635n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
3-ylboronic acid with (3-cyanophenyl)boronic acid. Crude 86 was
purified by preparative TLC (2 mm thickness, Rf = 0.47), eluting with
10% MeOH (containing 7 N NH3)/DCM. Yield: 39 mg (12%, 5
(m, 1H), 3.14−3.02 (m, 1H), 3.07 (s, 3H), 2.28−2.10 (m, 2H), 1.92−
1.77 (m, 1H); m/z (APCI-pos) M + 1 = 401; 90% purity by LC−MS
(254 nm).
1
steps). H NMR (400 MHz, CDCl3 + CD3OD) δ 7.75 (s, 1H), 7.72
(4R,4a′S,10a′S)-8′-(5-Chloropyridin-3-yl)-7′-fluoro-4a′-meth-
yl-3′,4′,4a′,10a′-tetrahydro-1′H,5H-spiro[oxazole-4,10′-
pyrano[4,3-b]chromen]-2-amine (92). Compound 92 was pre-
pared from 28 using reaction conditions described for the synthesis of
compounds 19−27 and 79. Crude 92 (mixture of diastereomers) was
purified by silica gel eluting with a linear gradient of 0−6% DCM/
MeOH + 1% NH4OH to provide 8′-(5-chloropyridin-3-yl)-7′-fluoro-
4a′-methyl-3′,4′,4a′,10a′-tetrahydro-1′H,5H-spiro[oxazole-4,10′-
pyrano[4,3-b]chromen]-2-amine (42 mg) as a 2:1 mixture of
diastereomers. This material was further purified by chiral SFC on a
Chiralpak AD (2 × 15 cm) column eluting with 25% MeOH (0.1%
NH4OH)/CO2 at 100 bar at a flow rate of 70 mL/min. The peaks
isolated were analyzed on Chiralpak AD (50 × 0.46 cm) column
eluting with 25% MeOH (0.1% NH4OH)/CO2, at 120 bar (flow rate 5
mL/min, 220 nm). Compound 92 (peak-3, 16 mg, 0.1% yield for 10
(d, J = 8 Hz, 1H), 7.56 (m, 1H), 7.50 (t, J = 8 Hz, 1H), 7.42 (m, 1H),
7.08 (m, 1H), 7.00 (d, J = 9 Hz, 1H), 5.29 (m, 1H), 4.07 (m, 1H),
3.60 (d, J = 11 Hz, 1H), 3.48 (t, J = 12 Hz, 1H), 3.13 (m, 1H), 3.09 (s,
3H), 2.10 (m, 1H), 1.94 (m, 1H), 1.19 (s, 3H); m/z (APCI-pos) M +
1 = 403; 93% purity by LC−MS (220 nm).
(4R*,4a′S*,10a′S*)-2-Amino-8′-(2-fluoropyridin-3-yl)-1,10a′-
dimethyl-3′,4′,4a′,10a′-tetrahydro-1′H-spiro[imidazole-4,10′-
pyrano[4,3-b]chromen]-5(1H)-one (87). Compound 87 was
prepared from 45 using reaction conditions described for the synthesis
of compounds 10−13 and 61. Crude 87 was purified by preparative
TLC (2 mm thickness, Rf = 0.46), eluting with 10% MeOH
(containing 7 N NH3)/DCM. Relative stereochemistry was confirmed
from cocrystal structures with BACE1 using the same conditions as
1
described for compound 62. Yield: 46 mg (14%, 5 steps). H NMR
1
(400 MHz, CDCl3 + CD3OD) δ 8.11 (m, 1H), 7.84 (t, J = 8 Hz, 1H),
7.43 (m, 1H), 7.29 (m, 1H), 7.11 (m, 1H), 7.00 (d, J = 9 Hz, 1H),
5.29 (m, 1H), 4.10 (m, 1H), 3.60 (d, J = 11 Hz, 1H), 3.49 (t, J = 12
Hz, 1H), 3.12 (d, J = 11 Hz, 1H), 3.07 (s, 3H), 2.14 (m, 1H), 1.98 (m,
1H), 1.19 (s, 3H); m/z (APCI-pos) M + 1 = 397; 91% purity by LC−
MS (220 nm).
steps, chemical purity >99%, >99% ee) was isolated. H NMR (400
MHz, DMSO-d6) δ 8.69−8.59 (m, 2H), 8.05 (s, 1H), 7.26 (d, J = 8.8
Hz, 1H), 6.77 (d, J = 12.0 Hz, 1H), 6.13 (s, 2H), 4.45 (d, J = 8.9 Hz,
1H), 4.12 (d, J = 8.9 Hz, 1H), 3.92 (dd, J = 12.2, 4.4 Hz, 1H), 3.70
(dd, J = 11.4, 4.4 Hz, 1H), 3.60−3.48 (m, 2H), 2.18 (dd, J = 11.3, 4.3
Hz, 1H), 1.90 (td, J = 12.5, 4.9 Hz, 1H), 1.81 (d, J = 12.5 Hz, 1H),
1.34 (s, 3H). m/z (ESI-pos) M + 1 = 404.1. HPLC purity = 100%
(254 nm).
(4R*,4a′S*,10a′S*)-2-Amino-1,10a′-dimethyl-8′-(pyrimidin-
5-yl)-3′,4′,4a′,10a′-tetrahydro-1′H-spiro[imidazole-4,10′-
pyrano[4,3-b]chromen]-5(1H)-one (88). Compound 88 was
prepared from 45 using reaction conditions described for the synthesis
of compounds 10−13 and 61, and replacing 2-fluoropyridin-3-
ylboronic acid with pyrimidin-5-ylboronic acid. Crude 88 was purified
by preparative TLC (2 mm thickness, Rf = 0.29), eluting with 10%
MeOH (containing 7N NH3)/DCM. Yield: 31 mg (12%, 5 steps),
(4R,4a′S,10a′S)-7′-Fluoro-8′-(2-fluoropyridin-3-yl)-4a′-meth-
yl-3′,4′,4a′,10a′-tetrahydro-1′H,5H-spiro[oxazole-4,10′-
pyrano[4,3-b]chromen]-2-amine (93). Compound 93 was pre-
pared from 28 using reaction conditions described for the synthesis of
compounds 19−27 and 79, and replacing (5-chloropyridin-3-
yl)boronic acid with (2-fluoropyridin-3-yl)boronic acid. Crude 93
(mixture of diastereomers) was purified as described for compound
92. Compound 93 (17 mg, 0.1% yield for 10 steps, chemical purity
>99%, >99% ee) was isolated. 1H NMR (400 MHz, DMSO-d6) δ 8.27
(d, J = 5.1 Hz, 1H), 8.06−7.96 (m, 1H), 7.46 (ddd, J = 7.1, 4.8, 1.8 Hz,
1H), 7.17 (d, J = 8.4 Hz, 1H), 6.77 (d, J = 11.3 Hz, 1H), 6.08 (s, 2H),
4.30 (s, 2H), 3.92 (dd, J = 11.9, 4.9 Hz, 1H), 3.85 (dd, J = 11.7, 4.0
Hz, 1H), 3.58 (td, J = 12.4, 2.6 Hz, 1H), 3.32 (dd, J = 14.1, 8.4 Hz,
1H), 2.11 (dd, J = 10.9, 4.0 Hz, 1H), 1.87 (dd, J = 12.7, 5.1 Hz, 1H),
1.79 (d, J = 12.7 Hz, 1H), 1.49 (s, 3H); m/z (ESI-pos) M + 1 = 388.1.
X-ray Crystal Structure. Human BACE1 (residues 57−453) was
expressed in E. coli with a noncleavable 6-His tag at the C-terminus.
The protein was refolded and purified in a manner similar to that from
Tomasselli et al.31 Briefly, cells were lysed using mechanical disruption
and inclusion bodies recovered by differential centrifugation. Inclusion
body proteins were solubilized in urea and refolded by dilution into
water. Following a 3 week refold period, BACE protein was recovered
by Q-Sepharose anion-exchange chromatography. Fractions containing
active protein were pooled and further purified over Source-Q anion
exchange and Sephadex 200 size-exclusion columns. Purified BACE
protein was concentrated to 10 mg/mL and frozen for crystallographic
studies. Crystals were grown by hanging drop vapor diffusion with a
precipitant of 20% PEG200, 0.1 M sodium acetate, pH 4.5−5.6, and
0−2 mM cobalt chloride. Crystals grew in space group C2221 in
approximately 1 week. Crystals were then transferred to a soaking
solution (27−30% PEG200, 0.1 M NaOAc, pH 5.0−5.6, 0−2 mM
CoCl2, and 10% DMSO) with compound at a final concentration of 1
mM. Crystals were soaked for 18−24 h and either flash-frozen in liquid
nitrogen or mounted directly for data collection. The soaking solution
also served as a cryprotectant for data collection at 100 K.
1
89% purity by LC−MS (220 nm). H NMR (400 MHz, CDCl3 +
CD3OD) δ 9.00 (s, 1H), 8.78 (s, 2H), 7.38 (m, 1H), 7.03 (m, 1H),
6.98 (m, 1H), 5.25 (m, 1H), 3.0−4.2 (m, 4H), 3.01 (s, 3H), 2.07 (m,
1H), 1.90 (m, 1H), 1.10 (s, 3H); m/z (APCI-pos) M + 1 = 380.
(4R*,4a′R*,10a′R*)-8′-(2-Fluoropyridin-3-yl)-10a′-methyl-
3′,4′,4a′,10a′-tetrahydro-1′H,5H-spiro[oxazole-4,10′-pyrano-
[4,3-b]chromen]-2-amine (89). Compound 89 was prepared from
45 using reaction conditions described for the synthesis of compounds
14, 15, and 61. Crude 89 was purified by preparative TLC (2 mm
thickness, Rf = 0.71), eluting with 10% MeOH (containing 7 N NH3)/
DCM. Relative stereochemistry was confirmed from cocrystal
structures with BACE1 using the same conditions as described for
1
compound 62. Yield: 37 mg (20%, 3 steps). H NMR (400 MHz,
CDCl3 + CD3OD) δ 8.13 (m, 1H), 7.87 (m, 1H), 7.46 (m, 1H), 7.39
(m, 1H), 7.30 (m, 1H), 6.94 (m, 1H), 4.64 (m, 1H), 4.51 (m, 1H),
4.43 (m, 1H), 3.67 (m, 1H), 3.3−3.6 (m, 3H), 2.06 (m, 1H), 1.97 (m,
1H), 1.00 (s, 3H); m/z (APCI-pos) M + 1 = 370; 94% purity by LC−
MS (220 nm).
(4R*,4a′S*,10a′S*)-2-Amino-8′-(3-(difluoromethoxy)-
phenyl)-7′-fluoro-1-methyl-3′,4′,4a′,10a′-tetrahydro-1′H-
spiro[imidazole-4,10′-pyrano[4,3-b]chromen]-5(1H)-one (90).
Compound 90 was prepared from 54, as described for compound
61, replacing 2-fluoropyridin-3-ylboronic acid with (3-
(difluoromethoxy)phenyl)boronic acid. The crude was purified by
silica gel chromatography (gradient: 0−100% (80% EtOAc/19%
MeOH/1% Et3N)/EtOAc). Yield: 2.3 mg (9%), 95% purity by LC−
1
MS (254 nm). H NMR (400 MHz, CDCl3) δ 7.38 (m, 1H), 7.31−
7.25 (m, 1H), 7.19 (m, 1H), 7.09 (m, 1H), 6.94 (m, 1H), 6.72 (m,
1H), 6.54 (t, J = 74 Hz, 1H), 4.96 (m, 1H), 4.10−3.91 (m, 2H), 3.48
(m, 1H), 3.08 (s, 3H), 3.04 (m, 1H), 2.23−2.11 (m, 2H), 1.91−1.78
(m, 1H); m/z (APCI-pos) M + 1 = 448.
Diffraction data were collected on a Rigaku FR-E generator and
Raxis IV++ detector (Rigaku Inc., Woodlands, TX). Data were
processed in Moslfm,32 and the initial structure was solved by
molecular replacement from pdb code 1FNK using Molrep. All crystal
structures were refined in Refmac, all three of these programs being
part of the CCP4 suite (Winn et al., 2011).33 Model building was
performed in Coot.34
(4R*,4a′S*,10a′S*)-2-Amino-7′-fluoro-8′-(2-fluoropyridin-3-
yl)-1-methyl-3′,4′,4a′,10a′-tetrahydro-1′H-spiro[imidazole-
4,10′-pyrano[4,3-b]chromen]-5(1H)-one (91). Compound 91 was
prepared from 54, as described for compound 61. The crude was
purified by silica gel chromatography (gradient: 0−100% (80%
EtOAc/19% MeOH/1% Et3N)/EtOAc). Yield: 2.5 mg (11%). 1H
NMR (400 MHz, CDCl3) δ 8.16 (s, 1H), 7.77 (m, 1H), 7.22 (m, 1H),
6.97 (m, 1H), 6.73 (m, 1H), 4.98 (m, 1H), 4.10−3.92 (m, 2H), 3.48
BACE1 Enzyme Assay. The activity of purified recombinant
human BACE1, expressed in CHO cells, was determined using a
T
dx.doi.org/10.1021/jm401635n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
custom synthesized biotinylated peptide based upon sAPP. Uncleaved
substrate was detected via HTRF whereby europium-labeled anti-Aβ
(amino acids 1−17, clone 6E10) was combined with D2-labeled
streptavidin. Specifically, 20 nM enzyme was incubated with 150 nM
peptide (Biotin-Lys-Thr-Glu-Glu-Ile-Ser-Glu-Val-Asn-Leu-Asp-Ala-
Glu-Phe-Arg-His-Asp-Ser-Gly-Tyr-Glu-Val-His-His-Gln-Lys-Leu, rep-
resenting P10−P17′) for 6 h at 22 °C in an assay buffer consisting of
50 mM sodium acetate (pH 4.4) and 0.1% CHAPS. In IC50 studies,
compounds were serially diluted (1:2.5) in DMSO (11 points);
compound and enzyme were preincubated for 15 min at room
temperature and the enzymatic reaction was initiated with peptide.
The detection solution consisted of 43 nM streptavidin-D2 and 1.1
nM antibody in 200 mM Tris (pH 8.0), 20 mM EDTA (pH 8.0), 0.1%
BSA, and 0.8 M KF. Following an incubation for 2 h at 22 °C, the
samples were excited at 320 nm and the emission ratio of 665 nm/615
nm was determined.
Cathepsin D Enzyme Assay. The catalytic activity of Cathepsin
D, purified from human spleen (Calbiochem), was assessed in a FRET
peptide substrate assay using 50 mM sodium acetate (pH 4.4) and
0.1% CHAPS. In IC50 studies, compounds were serially diluted (1:2.5)
in DMSO (11 points). The substrate (Anaspec catalog # 72097)
contains a 5-FAM/QXL 520 pair whereby cleavage results in an
increase of 5-FAM fluorescence that is measured using an excitation of
485 nm and an emission of 535 nm. The fluorescence was monitored
continuously for one hour at room temperature and initial rates were
calculated from the fluorescence increase. The final concentrations
were 10 nM enzyme and 2 μM substrate.
Cell-Based Assay. Human embryonic kidney cells (HEK293)
stably expressing APPwt were plated at a density of 35K cells/well in
96-well collagen coated plates (BD-Biosciences). The cells were
cultivated overnight at 37 °C and 5% CO2 in DMEM supplemented
with 10% FBS. The following day the media was changed, and the cells
were incubated for 48 h with test compounds at concentrations
ranging from 0.0008 to 16 μM. Following incubation with the test
compounds, the conditioned medium was collected, and the Aβ1−40
levels were determined using an HTRF assay (CisBio). The IC50 was
calculated from the percent of control Aβ40 as a function of the
concentration of the test compound. The HTRF to detect Aβ1−40 was
performed in 384-well microtiter plates (Corning). The antibody pair
that was used to detect Aβ1−40 from cell supernatants was a pair of
monoclonal antibodies, one labeled with Cryptate and one labeled
with XL655 as supplied by the manufacturer (CisBio). Conditioned
medium was incubated with antibody pair overnight at 4 °C. The
plates were measured for timeresolved fluorescence on a Victor2 plate
reader (Wallac).
Permeability and Efflux Assays. Both LLC-PK1 and MDR1
transfected LLC-PK1 cells were cultured and plated according to
manufacturer’s recommendations with the exception that the passage
media contained only 2% fetal bovine serum so as to extend passage
time out to seven days. Both positive and negative controls were used
to assess functionality of P-gp efflux in the assay. Stock solutions for
assay controls and the test article were prepared in DMSO for final test
concentrations of 10 and 1 μM, respectively. Final organic
concentration in the assay was 1%. All dosing solutions contained
10 μM lucifer yellow to monitor LLC-PK1 cell monolayer integrity.
For the apical to basolateral determination (A to B), 75 μL of the test
article in transport buffer were added to the apical side of the
individual transwells and 250 μL of basolateral media, without
compound or lucifer yellow, were added to each well. For the
basolateral to apical determination (B to A), 250 μL of test article in
transport buffer were added to each well and 75 μL transport buffer,
without compound or lucifer yellow, were added to each transwell. All
tests were performed in triplicate, and each compound was tested for
both apical to basolateral and basolateral to apical transport.
Propranolol, a high permeability compound (Papp AB > 8 × 10−6
cm/s), sulfasalazine, a medium permeability compound (Papp > 2 ×
10−6 cm/s; Papp < 8 × 10−6 cm/s), and cefuroximine, a low
permeability compound (Papp AB < 2 × 10−6 cm/s), were used as
negative controls for P-gp efflux (ER < 2). Digoxin, a medium
permeability compound, and quinidine, a high permeability com-
pound, were used as positive controls for P-gp efflux (ER > 2). The
plates were incubated for 2 h on a Lab-line Instruments Titer Orbital
Shaker (VWR, West Chester, PA) at 50 rpm and 37 °C with 5% CO2.
All culture plates were removed from the incubator, and 50 μL of
media were removed from the apical and basolateral portion of each
well and added to 150 μL of 1 μM labetalol in 2:1 CH3CN: H2O, v/v.
The plates were read using a Molecular Devices (Sunnyvale, CA)
Gemini Fluorometer to evaluate the lucifer yellow concentrations at
excitation/emission wavelengths of 425/535 nm. These values were
accepted when found to be below 5% for apical to basolateral and
basolateral to apical flux across the MDR1 transfected LLC-PK1 cell
monolayers. The plates were sealed, and the contents of each well
analyzed by LC−MS/MS. The compound concentrations were
determined from the ratio of the peak areas of the compound to the
internal standard (labetalol) in comparison to the dosing solution. The
LC−MS/MS system comprised an HTS-PAL autosampler (Leap
Technologies, Carrboro, NC), an HP1200 HPLC (Agilent, Palo Alto,
CA), and an MDS Sciex 4000 Q Trap system (Applied Biosystems,
Foster City, CA). Chromatographic separation of the analyte and
internal standard was achieved at rt using a C18 column (Kinetics, 50
× 300 mm, 2.6 μm particle size, Phenomenex, Torrance, CA) in
conjunction with gradient conditions using mobile phases A (water
containing 1% i-PrOH and 0.1% formic acid) and B (0.1% formic acid
in CH3CN). The total run time, including re-equilibration time, for a
single injection was 1.2 min. Mass spectrometric detection of the
analytes was accomplished using the ion spray positive mode. Analyte
responses were measured by multiple reaction monitoring (MRM) of
transitions unique to each compound (the protonated precursor ion
and selected product ions for each test article and m/z 329 to m/z 162
for labetalol, the internal standard).
Plasma Protein and Brain Tissue Binding. Aliquots of
undiluted plasma or diluted brain samples (three volumes of 0.1 M
sodium phosphate buffer, pH 7.4) from each animal species were
spiked with test compound in DMSO to give final concentrations of 5
μM (0.1% final DMSO) test article. Rapid equilibrium dialysis (RED)
inserts were prepared in triplicate by adding 500 μL of PBS for plasma
protein binding or 0.1 M sodium phosphate for brain tissue binding to
one chamber and 300 μL of spiked plasma or brain homogenate to the
opposing chamber. The RED units (inserts and bases) were sealed and
incubated at 37 °C on a plate shaker at 300 rpm for 6 h to equilibrate
the samples. One hundred microliters of each spiked matrix sample
and respective buffer were transferred to a 96-deep well plate for
sample extraction. The matrix samples were diluted with 100 μL of
either PBS or 0.1 M sodium phosphate buffer, and the buffer samples
were diluted with 100 μL of either brain tissue homogenate or plasma
to yield identical matrixes between buffer and nonbuffer samples.
Plasma or brain proteins were precipitated with 800 μL of CH3CN
spiked with 0.2 μM final concentration (0.02% DMSO) of labetalol,
the internal standard. The plates were spun in a centrifuge for 10 min
at 3000g. One hundred microliters of the supernatant were transferred
into a shallow 96-well plate and diluted 1:1 (v/v) with HPLC grade
water and sealed for LC−MS/MS analysis (method described below).
Liver Microsomal Incubations. A 100 mM potassium phosphate
assay buffer solution (KPB) was prepared as follows. Both KH2PO4
and K2HPO4 were dissolved separately in reagent grade water resulting
in a final concentration of 100 mM. A 75:25 mixture v/v of K2HPO4/
KH2PO4 was prepared, and the pH of the solution was adjusted to 7.4
using diluted HCl or diluted NaOH solutions. A stock solution of the
test article(s) was prepared at 10 mM (active compound) in DMSO.
The stock solution was diluted immediately before use to 2.5 μM using
the KPB solution to create the working standard. All test compounds
were completely soluble by visual inspection at rt. The NADPH-
regenerating solution (NRS) was prepared on the day of analysis by
diluting one volume of 17 mg/mL NADP+ with one volume of 78
mg/mL glucose-6-phosphate (both prepared in KPB, pH 7.4) and 7.9
volumes of 20 mM MgCl2. The final concentrations of NADP+ and
glucose-6-phosphate were 1.7 and 7.8 mg/mL, respectively. Immedi-
ately prior to use, the NRS was activated by the addition of 10 μL of
glucose-6-phosphate dehydrogenase (150 units/mL in KPB, pH 7.4)
per mL of NRS stock solution. Liver microsomes were diluted to 2.5
U
dx.doi.org/10.1021/jm401635n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
mg protein/mL using KPB. For each test article or positive control
(i.e., dextromethorphan, diazepam, diltiazem, phenacetin, tolbutamide,
and verapamil), 20 μL of 2.5 μM working standard solution of test
compound and 20 μL of microsomes (2.5 mg protein/mL) were
added to each well of a 96-well polypropylene plate (Costar, VWR,
West Chester, PA) in duplicate. The plates were placed in an incubator
at 37 °C for 5 min before adding the start solution. A 10 μL aliquot of
the NRS solution was added to each original well to initiate
metabolism. The concentration of the test compound during
incubation was 1 μM. One incubation plate was prepared for each
time point (i.e., 0 and 20 min). Incubations were conducted at 37 °C
and 100% relative humidity. At each time point, the appropriate
incubation plate was removed from the incubator and a solution
containing internal standard (150 μL, 0.25 μM labetalol in 60%
CH3CN) was added to each well. The plate was immediately spun in a
centrifuge at 2095g for 7 min at rt using an Allegra benchtop centrifuge
(Beckman Coulter, Fullerton, CA). A 200 μL aliquot of the
supernatant was transferred from each well to a 96-well shallow
plate (Costar). The plates were sealed using reusable plate mats.
Analysis was performed as described below.
Hepatocyte Incubations. A stock solution of the test article(s)
was prepared at 10 mM (active compound) in DMSO. The in vitro
stability of each test article or positive control was assessed in the
presence of hepatocytes as follows. Cryopreserved hepatocytes were
thawed, isolated from shipping media, and diluted to a density of 1 ×
106 viable cells/mL, according to the supplier’s guidelines, using
Dulbecco’s Modified Eagle Medium, 1×, high glucose (DMEM,
Invitrogen, Carlsbad, CA). Viability was determined by trypan blue
exclusion using a hemocytometer (3500 Hausser, VWR, West Chester,
PA). The 10 mM stock solution of test article(s) or control compound
was diluted to 2 μM using supplemented DMEM to create the
working standard. A 20 μL aliquot of test compound or control
(antipyrine, diazepam, diltiazem, lorazepam, propranolol, verapamil,
and 7-ethyl-10-hydroxycamptothecin [SN-38]) was added to each test
well of a 96-well polypropylene plate (Costar, VWR, West Chester,
PA) immediately followed by the addition of 20 μL of the hepatocyte
suspension. One incubation plate was prepared for each time point
(i.e., 0, 60, and 120 min) with samples being prepared in duplicate.
Incubations were conducted at 37 °C and 100% relative humidity. At
each time point, the appropriate incubation plate was removed from
the incubator and a solution containing internal standard (200 μL,
0.25 μM labetalol in 60% CH3CN) was added to each well. The plate
was mixed at 700 rpm for 1 min on a plate shaker (IKA MTS 2/4
Digital Microtiter Shaker, VWR) and immediately spun in a centrifuge
at 2095g for 10 min at rt using an Allegra benchtop centrifuge
(Beckman Coulter, Fullerton, CA). A 200 μL aliquot of the
supernatant was transferred from each well to a 96-well shallow
plate (Costar). The plates were sealed using reusable plate mats.
Analytical Quantitation for Microsomal/Hepatocyte Stabil-
ity and Tissue Binding Assays. The LC−MS/MS system
comprised an HTS-PAL autosampler (Leap Technologies, Carrboro,
NC), an HP1200 HPLC (Agilent, Palo Alto, CA), and an API4000
triple quadrupole mass spectrometer (PE Sciex, a division of Applied
Biosystems, Foster City, CA). Chromatographic separation of the
analyte and internal standard was achieved at rt using a C18 column
(Kinetex, 30 × 3.0 mm, 2.6 μm particle size, Phenomenex, Torrance,
CA) in conjunction with gradient conditions using mobile phases A
(aqueous 0.1% formic acid with 1% i-PrOH) and B (0.1% formic acid
in CH3CN). The total run time, including re-equilibration, for a single
injection was 2 min. Mass spectrometric detection of the analytes was
accomplished using the ESI+ ionization mode. Ion current was
optimized during infusion of a stock solution of each test article.
Analyte responses were measured by multiple reaction monitoring
(MRM) of transitions unique to each compound. Data were acquired,
and peak areas were calculated for test compounds and the internal
standard using Analyst 1.5.1 software (Sciex). For the liver microsomal
and hepatocyte stability assessments, peak area tables were exported to
BioAssay Enterprise (CambridgeSoft, Cambridge, MA), where the
average analyte to internal standard peak area ratios were used to
calculate percent remaining (%REM), half-life (t1/2), predicted hepatic
clearance (CLhep), and predicted hepatic extraction ratio (ER).
Pharmacokinetic Studies. Animals and Drug Treatment: Male
CD-1 mice (4−6 weeks old, 25−35 g) and Male Hartley guinea pigs
(surgically implanted jugular cannulas, 6−8 weeks old, 250−300 g)
were obtained from Charles River (Portage, MI, and Kingston, NY,
respectively). The cannulae were exteriorized at the back of the neck,
and each animal was placed singly in a metabolic cage. Male Sprague−
Dawley rats (6−8 weeks old, 250−300 g) were obtained from Harlan
(Indianapolis, IN). All the animals were housed at controlled
temperature and humidity in an alternating 12 h light and dark
cycle with free access to food and water, in accordance with Array
BioPharma Institutional Animal Care and Use Committee guidelines
and in harmony with the Guide for Laboratory Animal Care and Use.
The guinea pigs were housed individually according to NIH guidelines.
Mice (3 per group, per time-point) received a 3 mg/kg dose of drug (5
mL/kg) intravenously or 10 mg/kg dose (10 mL/kg) orally. Rats (3
per group) received a 3 mg/kg dose of compound intravenously or 60
mg/kg dose orally. Guinea pigs (4 per group) received a 3 mg/kg dose
(1 mL/kg) of compound intravenously via cannula or 60 mg/kg dose
(5 mL/kg) orally. Plasma collection: Mice were euthanized by CO2
inhalation, rats and guinea pigs were anesthetized by isoflurane at
designated times between 1 min and 24 h postdose. The whole blood
samples were drawn via cardiac puncture (mice), tail vein (rats), and
jugular vein cannula (guinea pigs) in tubes containing 1.5% EDTA.
The blood samples were centrifuged at 14000 rpm for 10 min; plasma
was decanted and stored at −20 °C before drug concentration analysis.
In guinea pigs, after each blood draw, the cannulas were flushed with
heparinized saline (volume equal to the cannula volume). At
termination of the study, guinea pigs were euthanized by euthasol
(250 mg/kg) via intraperitoneal injection. Brain collection: The orally
dosed mice, postmortem at 30 min to 8 h postdose, were perfused
with 5 mL sterile saline. Brains were harvested, and the tissue weights
were recorded. The brains were snap frozen immediately in liquid
nitrogen and stored at −80 °C until determination of drug
concentration in brain samples.
Rat and Guinea Pig Pharmacodynamic Studies. Animals and
Drug Treatment: Male Sprague−Dawley rats and male Hartley guinea
pigs (6−8 weeks old, 250−350 g), were obtained from Harlan
(Indianapolis, IN) and Charles River Laboratories (Kingston, NY),
respectively. The animals were housed in groups of 3, at controlled
temperature and humidity in an alternating 12 h light and dark cycle
with free access to food and water, in accordance with Array
BioPharma Institutional Animal Care and Use Committee guidelines
and in harmony with the Guide for Laboratory Animal Care and Use.
Rats (8 per group) received orally either vehicle (1% carboxymethyl
cellulose/0.5% Tween 80) or 62 (60 mg/kg) at a volume of 4 mL/kg.
Guinea pigs (6−8 per group) received orally either vehicle (a) 1%
carboxymethyl cellulose/0.5% Tween 80 or (b) 40% PEG400/10%
ethanol/50% water or drugs: 62 (30, 10, and 60 mg/kg) or 59 (60
mg/kg) at a volume of 4−5 mL/kg. Plasma, CSF, and brain collection:
At 1, 3, 5, 8, and 12 h after receiving the vehicle or drug dose, the rats
were euthanized by CO2 inhalation, and guinea pigs were euthanized
by euthasol (250 mg/kg) via intraperitoneal injection. The whole
blood samples were collected postmortem via cardiac puncture, in
tubes containing 1.5% EDTA, centrifuged at 14000 rpm for 10 min.
The plasma was stored at −20 °C until drug concentration analysis.
The CSF samples were drawn postmortem via cisterna magna
puncture through a 25 gauge needle (Becton Dickinson Franklin
Lakes, NJ), into tubes containing 50 μL of 10% BSA in water. The
CSF samples were snap frozen in liquid nitrogen and stored at −80 °C
until Aβ 40 assay. Sample weights were recorded to determine the
actual volume of CSF collected. The postmortem animals were
perfused with 20 mL of sterile saline and brain hemispheres were
dissected (brain stem and cerebellum were not collected), halved
longitudinally, weighed, and placed only left half of the brain in a 15
mL conical tube and snap frozen in liquid nitrogen. All tissues were
stored at −80 °C until drug concentration determination. CSF Aβ 40
assay: CSF Aβ 40 concentration in rat and guinea pig CSF samples was
analyzed by MSD 96-Well MULTI-ARRAY Human/Rodent (4G8)
V
dx.doi.org/10.1021/jm401635n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Aβ 40 Ultra-Sensitive Assay (rat) or Human (6E10) Aβ 40 Ultra-
Sensitive Kit (Meso Scale Discovery, Gaithersburg, MD). The 96- well
MULTI-SPOT Aβ 40 peptide plate was incubated with 1% BSA
solution at room temperature with shaking for 1 h and washed as per
manufacturer’s instructions with Tris wash buffer (Tris wash buffer in
deionized water). The detection antibody solution (25 μL/well, sulfo-
tag 4G8 in 1% BSA) and CSF samples or standards (25 μL/well) were
added to the plate for incubation at room temperature for 2 h. The
plate was washed three times before the addition of MSD read buffer
and was read on Sector Imager 2400 (Meso Scale Discovery,
Gaithersburg, MD) immediately after. The Aβ 40 levels in neat CSF
samples were then determined using a standard curve, values were
obtained as pg/mL, and the results were expressed as mean Aβ 40
reduction percentage (%) of control samples from vehicle-treated
animals.
Cyno Pharmacodynamic Studies. A single oral dose of
compound at 100 mg/kg was given to 10 male cynomolgus monkeys
(3−5 year, 2−3 kg) at Maccine. The study utilized a crossover design:
n = 10 vehicle, n = 10 at 100 mg/kg, with an 18−20 days washout
period between vehicle and treatment. Baseline was established 216−
24 h before dosing. PK was determined for parent drug concentration
in plasma and CSF, and PD was determined by measuring Aβ1−40
levels in plasma and CSF. Full details are published elsewhere.30
work; and Jaimie Rogan for LLC-PK1 cell permeability/efflux
assays. We thank the following people from Genentech: Chris
Hamman and Mengling Wong for chiral separation; and Young
Shin and Melis Coraggio for bioanalytical analysis for cyno PK/
PD.
ABBREVIATIONS USED
■
BLQ, below the level of quantitation; IVIVC, in vitro−in vivo
correlation; Papp AB, apparent cell permeability in the apical to
basolateral direction; APCI, atmospheric-pressure chemical
ionization; CHO, Chinese hamster ovary; HTRF, homoge-
neous time-resolved fluorescence; 5-FAM, 5-carboxyfluores-
cein; DMEM, Dulbecco’s modified eagle medium; LLC-PK1,
pig kidney epithelial cells; RED, rapid equilibrium dialysis; ER,
extraction ratio; KPB, potassium phosphate buffer; NRS,
NADPH-regenerating solution; PDB, protein data bank;
tPSA, total polar surface area; SNAr, nucleophilic aromatic
substitution
REFERENCES
■
(1) 2011 Alzheimer’s Disease Facts and Figures. http://www.alz.org/
downloads/Facts_Figures_2011.pdf.
ASSOCIATED CONTENT
* Supporting Information
■
(2) World Alzheimer Report 2009. http://www.alz.co.uk/research/
files/WorldAlzheimerReport-ExecutiveSummary.pdf.
(3) Citron, M. Alzheimer’s disease: strategies for disease
S
Figure S1 and experimental details for compound 94. This
material is available free of charge via the Internet at http://
pubs.acs.org.
modification. Nat. Rev. Drug Discovery 2010, 9, 387−398.
(4) (a) 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. (b) Younkin, S. G. The role of A beta 42 in
Alzheimer’s disease. J. Physiol. 1998, 92, 289−292. (c) Querfurth, H.
W.; LaFeral, F. M. Mechanisms of Disease: Alzheimer’s disease. N.
Engl. J. Med. 2010, 362, 329−344.
Accession Codes
The PDB accession number for the coordinates of the cocrystal
structure of BACE1 + compound 62 is 4N00.
AUTHOR INFORMATION
Corresponding Author
*(A.A.T.) Phone: 970-214-0574. E-mail: allenathomas@
comcast.net.
(5) (a) De Strooper, B.; Vasar, R.; Golde, T. The secretases: enzymes
with therapeutic potential in Alzheimer’s disease. Nat. Rev. Neurol.
2010, 6, 99−107. (b) 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. Beta-
secretase cleavage of Alzheimer’s amyloid precursor protein by the
transmembrane aspartic protease BACE. Science 1999, 286, 735−741.
(c) Vassar, R.; Kovacs, D. M.; Yan, R.; Wong, P. C. The β-secretase
enzyme in health and Alzheimer’s disease: regulation, cell biology,
function, and therapeutic potential. J. Neurosci. 2009, 29, 12787−
12794. (d) Imbimbo, B. P. Alzheimer’s disease: γ-secretase inhibitors.
Drug Discovery Today 2008, 5, 169−175. (e) Wu, W.-L.; Zhang, L. γ-
Secretase inhibitors for the treatment of Alzheimer’s disease. Drug Dev.
Res. 2009, 70, 94−100.
(6) (a) Luo, Y.; Bolon, B.; Kahn, S.; Bennett, B. D.; Babu-Khan, S.;
Denis, P.; Fan, W.; Kha, H.; Zhang, J.; Gong, Y.; Martin, L.; Louis, J.-
C.; Yan, Q.; Richards, W. G.; Citron, M.; Vassar, R. Mice deficient in
BACE1, the Alzheimer’s beta-secretase, have normal phenotype and
abolished beta-amyloid generation. Nat. Neurosci. 2001, 4, 231−232.
(b) Roberds, S. L.; Anderson, J.; Basi, G.; Bienkowski, M. J.;
Branstetter, D. G.; Chen, K. S.; Freedman, S. B.; Frigon, N. L.; Games,
D.; Hu, K.; Johnson-Wood, K.; Kappenman, K. E.; Kawabe, T. T.;
Kola, I.; Kuehn, R.; Lee, M.; Liu, W.; Motter, R.; Nichols, N. F.;
Power, M.; Robertson, D. W.; Schenk, D.; Schoor, M.; Shopp, G. M.;
Shuck, M. E.; Sinha, S.; Svensson, K. A.; Tatsuno, G.; Tintrup, H.;
Wijsman, J.; Wright, S.; McConlogue, L. BACE knockout mice are
healthy despite lacking the primary beta secretase activity in brain:
implications for Alzheimer’s disease therapeutics. Hum. Mol. Genet.
2001, 10, 1317−1324. (c) Ohno, M.; Sametsky, E. A.; Younkin, L. H.;
Oakley, H.; Younkin, S. G.; Citron, M.; Vassar, R.; Disterhoft, J. F.
BACE1 deficiency rescues memory deficits and cholinergic dysfunc-
tion in a mouse model of Alzheimer’s disease. Neuron 2004, 41, 27−
33. (d) McConlogue, L.; Buttini, M.; Anderson, J. P.; Brigham, E. F.;
Chen, K. S.; Freedman, S. B.; Games, D.; Johnson Wood, K.; Lee, M.;
■
Present Addresses
§(A.A.T.) Community College of Denver, Science Building, SI
1006, 1151 Arapahoe St., Denver, Colorado, 80204, United
States.
∥(D.S.) 7484 Old Post Road, Boulder, Colorado 80301, United
States.
⊥(A.A.C.) Clinical Sciences Building, Room 309, Medical
University of South Carolina, 96 Jonathan Lucas Street,
Charleston, South Carolina 29403, United States.
○(M.K.G.D.) Institute for Applied Cancer Science, The
University of Texas MD Anderson Cancer Center, 1515
Holcombe Boulevard, Unit 1956, Houston, Texas 77030,
United States.
∇(D.D.) Boehringer Ingelheim Pharmaceuticals, 900 Ridgebury
Road, Ridgefield, Connecticut 06877, United States.
¶(S.R.) E-mail: sumeetrana1@gmail.com
#(K.R.) ProPharma Services, 3195 East Yarrow Circle, Superior,
Colorado 80027, United States.
□(R.G.) Agilent Technologies, Inc., 5555 Airport Boulevard.
Suite 100, Boulder, Colorado 80301, United States.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank the following people from Array BioPharma: Andrew
Allen for multidimensional NMR spectroscopy support and
chiral HPLC methods development; Julie Harris for PK/PD
W
dx.doi.org/10.1021/jm401635n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
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.
(7) Haass, C.; Lemere, C. A.; Capell, A.; Citron, M.; Seubert, P.;
Schenk, D.; Lannfelt, L.; Selkoe, D. J. The Swedish mutation causes
early-onset Alzheimer’s disease by β-secretase cleavage within the
secretory pathway. Nat. Med. 1995, 1, 1291−1296.
Levie, K.; Siu, M.; Purkey, H. E.; Lyssikatos, J. P. Unpublished results.
(b) Hunt, K. W.; Cook, A. W.; Watts, R. J.; Clark, C. T.; Vigers, G.;
Smith, D.; Metcalf, A. T.; Gunawardana, I. W.; Burkard, M.; Cox, A.
A.; Geck Do, M. K.; Dutcher, D.; Thomas, A. A.; Rana, S.; Kallan, N.
C.; DeLisle, R. K.; Rizzi, J. P.; Regal, K.; Sammond, D.; Groneberg, R.;
Siu, M.; Purkey, H.; Lyssikatos, J. P.; Marlow, A.; Liu, X.; Tang, T. P.
Spirocyclic β-site amyloid precursor protein cleaving enzyme 1
(BACE1) inhibitors: from hit to lowering of cerebrospinal fluid
(CSF) amyloid β in a higher species. J. Med. Chem. 2013, 56, 3379−
3403. (c) Hunt, K. H.; Cook, A. W.; Watts, R. J.; Clark, C. T.; Vigers,
G.; Smith, D.; Metcalf, A. T.; Gunawardana, I. W.; Burkard, M.; Cox,
A. A.; Geck Do, M. K.; Dutcher, D.; Thomas, A. A.; Rana, S.; Kallan,
N. C.; DeLisle, R. K.; Rizzi, J. P.; Regal, K.; Sammond, D.; Groneberg,
R.; Siu, M.; Purkey, H.; Lyssikatos, J. P.; Marlow, A.; Liu, X.; Tang, T.
P. Unpublished results.
(8) Jonsson, T.; Atwal, J. K.; Steinberg, S.; Snaedal, J.; Jonsson, P. V.;
Bjornsson, S.; Stefansson, H.; Sulem, P.; Gudbjartsson, D.; Maloney,
J.; Hoyte, K.; Gustafson, A.; Liu, Y.; Lu, Y.; Bhangale, T.; Graham, R.
R.; Huttenlocher, J.; Bjornsdottir, G.; Andreassen, O. A.; Jonsson, E.
̈
G.; Palotie, A.; Behrens, T. W.; Magnusson, O. T.; Kong, A.;
Thorsteinsdottir, U.; Watts, R. J.; Stefansson, K. A mutation in APP
protects against Alzheimer’s disease and age-related cognitive decline.
Nature 2012, 488, 96−99.
(9) May, P. C.; Dean, R. A.; Lowe, S. L.; Martenyi, F.; Sheehan, S.
M.; Boggs, L. N.; Monk, S. A.; Mathes, B. M.; Mergott, D. J.; Watson,
B. M.; Stout, S. L.; Timm, D. E.; Smith Labell, E.; Gonzales, C. R.;
Nakano, M.; Jhee, S. S.; Yen, M.; Ereshefsky, L.; Lindstrom, T. D.;
Calligaro, D. O.; Cocke, P. J.; Greg Hall, D.; Friedrich, S.; Citron, M.;
Audia, J. E. Robust central reduction of amyloid-β in humans with an
orally available, non-peptidic β-secretase inhibitor. J. Neurosci. 2011,
31, 16507−16516.
(10) Stone, J.; Kleijn, H. J.; Dockendorf, M.; Ma, L.; Palcza, J.; Tseng,
J.; Tanen, M.; Forman, M. Consistency of Bace inhibitor-mediated
brain amyloid production inhibition by MK-8931 in Alzheimer’s
disease patients and healthy young adults. Alzheimer's Dementia 2013,
9 (4), P690−P691.
(11) (a) Cheret, C.; Willem, M.; Fricker, F. R.; Wende, H.; Wulf-
Goldenberg, A.; Tahirovic, S.; Nave, K.-A.; Saftig, P.; Haass, C.;
Garratt, A. N.; Bennett, D. L.; Birchmeier, C. Bace1 and Neuregulin-1
cooperate to control formation and maintenance of muscle spindles.
EMBO J. 2013, 32, 2015−2028. (b) 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 beta-secretase BACE1. Science 2006, 314, 664−666. (c) Kandale-
pas, P. C.; Vassar, R. Identification and biology of β-secretase. J.
Neurochem. 2012, 120 (Suppl 1), 55−61. (d) Kuhn, P. H.; Koroniak,
K.; Hogl, S.; Colombo, A.; Zeitschel, U.; Willem, M.; Volbracht, C.;
Schepers, U.; Imhof, A.; Hoffmeister, A.; Haass, C.; Rossner, S.; Brase,
S.; Lichtenthaler, S. F. Secretome protein enrichment identifies
physiological BACE1 protease substrates in neurons. EMBO J. 2012,
31, 3157−3168. (e) Zhou, L.; Barao, S.; Laga, M.; Bockstael, K.;
Borgers, M.; Gijsen, H.; Annaert, W.; Moechars, D.; Mercken, M.;
Gevaert, K.; De Strooper, B. The neural cell adhesion molecules L1
and CHL1 are cleaved by BACE1 protease in vivo. J. Biol. Chem. 2012,
287, 25927−25940.
(12) Tseng, J.; Dockendorf, M.; Krishna, G.; Ma, L.; Palcza, J.;
Leempoels, J.; Ramael, S.; Han, D.; Jhee, S.; Ereshefsky, L.; Wagner, J.;
Troyer, M.; Forman, M. Safety and pharmacokinetics of the novel
BACE inhibitor MK-8931 in healthy subjects following single- and
multiple-dose administration. Alzheimer's Dementia 2012, 8 (4),
P184−P185.
(13) An Efficacy and Safety Trial of MK-8931 in Mild to Moderate
Alzheimer's Disease (P07738) (EPOCH). http://clinicaltrials.gov/
show/NCT01739348.
(14) (a) Wager, T. T.; Hou, X.; Verhoest, P. R.; Villalobos, A.
Moving beyond rules: the development of a central nervous system
multiparameter optimization (CNS MPO) approach to enable
alignment of druglike properties. ACS Chem. Neurosci. 2010, 1,
435−449. (b) Clark, D. E. Rapid calculation of polar molecular surface
area and its application to the prediction of transport phenomena. 2.
Prediction of blood−brain barrier penetration. J. Pharm. Sci. 1999, 88,
815−821. (c) Hitchcock, S. A.; Pennington, L. D. Structure−brain
exposure relationships. J. Med. Chem. 2006, 49, 7559−7583.
(15) (a) Tang, T. P.; Thomas, A. A.; Metcalf, A. T.; Hunt, K. W.;
Vigers, G.; Rizzi, J. P.; Watts, R. J.; Liu, X.; Smith, D.; Rhodes, S. P.;
Brown, K. D.; Otten, J. N.; Burkard, M.; Cox, A. A.; Geck Do, M. K.;
Dutcher, D.; Rana, S.; DeLisle, R. K.; Regal, K.; Wright, A. D.; Scearce-
(16) (a) Im, E.; Kazlauskas, A. The role of cathepsins in ocular
physiology and pathology. Exp. Eye Res. 2007, 84, 383−388.
(b) Benes, P.; Vetvicka, V.; Fusek, M. Cathepsin D–many functions
of one aspartic protease. Crit. Rev. Oncol. Hematol. 2008, 68, 12−28.
(17) (a) Hassner, A.; Lorber, M. E.; Heathcock, C. Addition of iodine
isocyanate to olefins. Scope and synthetic utility. J. Org. Chem. 1967,
32, 540−549. (b) Wittekind, R. R.; Rosenau, J. D.; Poos, G. I. Ring
cleavage reactions of trans-2-amino-3a,4,5,6,7,7a-hexahydrobenzoxa-
zole. J. Org. Chem. 1961, 26, 444−446. (c) Gebelein, C. G.; Rosen, S.;
Swern, D. Observations on the mechanism of addition of iodine
isocyanate to unsaturated compounds. J. Org. Chem. 1969, 34, 1677−
1679.
(18) Final products 72, 76, 79, 83−85, 87, and 89 were each
cocrystallized with BACE1 using the same conditions as for compound
62, in order to unambiguously assign relative stereochemistry.
(19) (a) Badawy, D. S.; Kandeel, E. M.; Awad, N. M.; Abdel-Rahman,
A.-R. H. Synthesis of some new naphthopyran, pyrazole, pyridine, and
thienobenzochromene derivatives using 1-(1-hydroxy-2-naphthyl)
ethanone as a versatile starting material. Phosphorus, Sulfur Silicon
Relat. Elem 2009, 184, 179−196. (b) Compound 40 has reportedly
been synthesized by an alternative procedure: Shaw, G. A novel high
yield γ-chromone synthesis. J. Chem. Soc., Chem. Comm. 1987, 1735−
1736.
(20) Stereochemistry for compounds 41 and 50 assigned by analogy
to related compounds from the literature. (a) Saengchantara, S. T.;
Wallace, T. W. Heterodiene cycloadditions of 3-acylchromones with
enol ethers. J. Chem. Soc., Perkin Trans. I 1986, 789−794. (b) Coutts,
S. J.; Wallace, T. W. Heterodiene cycloadditions: Preparation and
transformations of some substituted pyrano[4,3-b][1]benzopyrans.
Tetrahedron 1994, 50, 11755−11780.
(21) We found that using the 1 and 4 h time points for evaluating
coverage ratio gave a comparable result to using an AUC-based
coverage ratio for rank ordering compounds.
(22) (a) Edwards, P. D.; Albert, J. S.; Sylvester, M.; Aharony, D.;
Andisik, D.; Callaghan, O.; Campbell, J. B.; Carr, R. A.; Chessari, G.;
Congreve, M.; Frederickson, M.; Folmer, R. H.; Geschwindner, S.;
Koether, G.; Kolmodin, K.; Krumrine, J.; Mauger, R. C.; Murray, C.
W.; Olsson, L. L.; Patel, S.; Spear, N.; Tian, 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. (b) 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 β-
secretase (BACE1) inhibitors. J. Med. Chem. 2009, 52, 6314−6323.
(c) Swahn, B.-M.; Holenz, J.; Kihlstrom, J.; Kolmodin, K.; Lindstrom,
̈
̈
J.; Plobeck, N.; Rotticci, D.; Sehgelmeble, F.; Sundstrom, M.; von
̈
Berg, S.; Falting, J.; Georgievska, B.; Gustavsson, S.; Neelissen, J.; Ek,
̈
M.; Olsson, L.-L.; Berg, S. Aminoimidazoles as BACE-1 inhibitors: the
challenge to achieve in vivo brain efficacy. Bioorg. Med. Chem. Lett.
2012, 22, 1854−1859. (d) Huang, H.; La, D. S.; Cheng, A. C.;
Whittington, D. A.; Patel, V. F.; Chen, K.; Dineen, T. A.; Epstein, O.;
Graceffa, R.; Hickman, D.; Kiang, Y.-H.; Louie, S.; Luo, Y.; Wahl, R.
C.; Wen, P. H.; Wood, S.; Fremeau, R. T., Jr. Structure- and property-
X
dx.doi.org/10.1021/jm401635n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
based design of aminooxazoline xanthenes as selective, orally
efficacious, and CNS penetrable BACE inhibitors for the treatment
of Alzheimer’s disease. J. Med. Chem. 2012, 55, 9156−9169.
(e) Cheng, Y.; Judd, T. C.; Bartberger, M. D.; Brown, J.; Chen, K.;
Fremeau, R. T., Jr.; Hickman, D.; Hitchcock, S. A.; Jordan, B.; Li, V.;
Lopez, P.; Louie, S. W.; Luo, Y.; Michelsen, K.; Nixey, T.; Powers, T.
S.; Rattan, C.; Sickmier, E. A.; St. Jean, D. J., Jr.; Wahl, R. C.; Wen, P.
H.; Wood, S. From fragment screening to in vivo efficacy: optimization
of a series of 2-aminoquinolines as potent inhibitors of beta-site
amyloid precursor protein cleaving enzyme 1 (BACE1). J. Med. Chem.
2011, 54, 5836−5857. (f) Gravenfors, Y.; Viklund, J.; Blid, J.; Ginman,
Schauer, S. P.; Regal, K.; Hunt, K. W.; Thomas, A. A.; Siu, M.;
Lyssikatos, J.; Deshmukh, G.; Hop, C. E. C. A. Mechanistic
pharmacokinetic−pharmacodynamic modeling of BACE1 inhibition
in monkeys: development of a predictive model for amyloid precursor
protein processing. Drug Metab. Dispos. 2013, 41, 1319−1328.
(31) Tomasselli, A. G.; Paddock, D. J.; Emmons, T. L.; Mildner, A.
M.; Leone, J. W.; Lull, J. M.; Cialdella, J. I.; Prince, D. B.; Fischer, H.
D.; Heinrikson, R. L.; Benson, T. E. High yield expression of human
BACE constructs in Eschericia coli for refolding, purification, and high
resolution diffracting crystal forms. Protein Pept. Lett. 2008, 15, 131−
143.
(32) Leslie, A. G. W.; Powell, H. R. Processing Diffraction Data with
Mosfim. In Evolving Methods for Macromolecular Crystallography; Read,
R. J., Sussman, J. L., Eds.; NATO Science Series II: Mathematics,
Physics, and Chemistry 245; Springer: New York, 2007; pp 41−51.
(33) Winn, M. D.; Ballard, C. C.; Cowtan, K. D.; Dodson, E. J.;
Emsley, P.; Evans, P. R.; Keegan, R. M.; Krissinel, E. B.; Leslie, A. G.;
McCoy, A.; McNicholas, S. J.; Murshudov, G. N.; Pannu, N. S.;
Potterton, E. A.; Powell, H. R.; Read, R. J.; Vagin, A.; Wilson, K. S.
Overview of the CCP4 suite and current developments. Acta
Crystallogr., Sect. D: Biol. Crystallogr. 2011, D67, 235−242.
(34) Emsley, P.; Cowtan, K. Coot: model-building tools for
molecular graphics. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2004,
60, 2126−2132.
T.; Karlstrom, S.; Kihlstrom, J.; Kolmodin, K.; Lindstrom, J.; von Berg,
̈
̈
̈
S.; von Kieseritzky, F.; Slivo, C.; Swahn, B.-M.; Olsson, L.-L.;
Johansson, P.; Eketjall, S.; Falting, J.; Jeppsson, F.; Stromberg, K.;
̈
̈
̈
Janson, J.; Rahm, F. New aminoimidazoles as β-secretase (BACE-1)
inhibitors showing amyloid-β (aβ) lowering in brain. J. Med. Chem.
2012, 55, 9297−9311. (g) Mandal, M.; Zhu, Z.; Cumming, J. N.; Liu,
X.; Strickland, C.; Mazzola, R. D.; Caldwell, J. P.; Leach, P.; Grzelak,
M.; Hyde, L.; Zhang, Q.; Terracina, G.; Zhang, L.; Chen, X.; Kuvelkar,
R.; Kennedy, M. E.; Favreau, L.; Cox, K.; Orth, P.; Buevich, A.; Voigt,
J.; Wang, H.; Kazakevich, I.; McKittrick, B. A.; Greenlee, W.; Parker, E.
M.; Stamford, A. W. Design and validation of bicyclic iminopyr-
imidinones as beta amyloid cleaving enzyme-1 (BACE1) inhibitors:
conformational constraint to favor a bioactive conformation. J. Med.
Chem. 2012, 55, 9331−9345. (h) Swahn, B.-M.; Kolmodin, K.;
Karlstrom, S.; von Berg, S.; Soderman, P.; Holenz, J.; Berg, S.;
̈
̈
Lindstrom, J.; Sundstrom, M.; Turek, D.; Kihlstrom, J.; Slivo, C.;
̈
̈
̈
Andersson, L.; Pyring, D.; Rotticci, D.; Ohberg, L.; Kers, A.; Bogar, K.;
von Kieseritzky, F.; Bergh, M.; Olsson, L.-L.; Janson, J.; Eketjall, S.;
Georgievska, B.; Jeppsson, F.; Falting, J. Design and synthesis of β-site
̈
̈
amyloid precursor protein cleaving enzyme (BACE1) inhibitors with
in vivo brain reduction of β-amyloid peptides. J. Med. Chem. 2012, 55,
9346−9361. (i) Cumming, J. N.; Smith, E. M.; Wang, L.; Misiaszek, J.;
Durkin, J.; Pan, J.; Iserloh, U.; Wu, Y.; Zhu, Z.; Strickland, C.; Voigt, J.;
Chen, X.; Kennedy, M. E.; Kuvelkar, R.; Hyde, L. A.; Cox, K.; Favreau,
L.; Czarniecki, M. F.; Greenlee, W. J.; McKittrick, B. A.; Parker, E. M.;
Stamford, A. W. Structure based design of iminohydantoin BACE1
inhibitors: identification of an orally available, centrally active BACE1
inhibitor. Bioorg. Med. Chem. Lett. 2012, 22, 2444−2449. (j) 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 β-secretase (BACE1)
inhibitors. J. Med. Chem. 2010, 53, 1146−1158.
(23) Hilpert, H.; Guba, W.; Woltering, T. J.; Wostl, W.; Pinard, E.;
Mauser, H.; Mayweg, A. V.; Rogers-Evans, M.; Humm, R.;
Krummenacher, D.; Muser, T.; Schnider, C.; Jacobsen, H.; Ozmen,
L.; Bergadano, A.; Banner, D. W.; Hochstrasser, R.; Kuglstatter, A.;
David-Pierson, P.; Fischer, H.; Polara, A.; Narquizian, R. β-Secretase
(BACE1) inhibitors with high in vivo efficacy suitable for clinical
evaluation in Alzheimer’s disease. J. Med. Chem. 2013, 56, 3980−3995.
(24) The alternate trans diastereomers (not depicted), analogous to
diastereomers B in Table 2, were significantly less potent.
(25) The alternate trans diastereomers (not depicted), analogous to
diastereomers A in Table 2, were significantly less potent.
(26) The alternate trans diastereomer (not depicted), analogous to
diastereomers A in Table 2, was not accessible using the synthetic
route described by Scheme 5.
(27) Meanwell, N. A. Synopsis of some recent tactical application of
bioisosteres in drug design. J. Med. Chem. 2011, 54, 2529−2591.
(28) F% of 103% was within experimental error of 100%.
(29) Du, P.; Wood, K. M.; Rosner, M. H.; Cunningham, D.; Tate, B.;
Geoghegan, K. F. Dominance of amyloid precursor protein sequence
over host cell secretases in determining β-amyloid profiles studies of
interspecies variation and drug action by internally standardized
immunoprecipitation/mass spectrometry. J. Pharm. Exp. Ther. 2007,
320, 1144−1152.
(30) Liu, X.; Wong, H.; Scearce-Levie, K.; Watts, R. J.; Coraggio, M.;
Shin, Y. G.; Peng, K.; Wildsmith, K. R.; Atwal, J. K.; Mango, J.;
Y
dx.doi.org/10.1021/jm401635n | J. Med. Chem. XXXX, XXX, XXX−XXX