Preparation and biological evaluation of conformationally constrained BACE1 inhibitors

Article abstract of DOI:10.1016/j.bmc.2015.04.062

Abstract The BACE1 enzyme is a key target for Alzheimer's disease. During our BACE1 research efforts, fragment screening revealed that bicyclic thiazine 3 had low millimolar activity against BACE1. Analysis of the co-crystal structure of 3 suggested that potency could be increased through extension toward the S3 pocket and through conformational constraint of the thiazine core. Pursuit of S3-binding groups produced low micromolar inhibitor 6, which informed the S3-design for constrained analogs 7 and 8, themselves prepared via independent, multi-step synthetic routes. Biological characterization of BACE inhibitors 6-8 is described.

Full text of DOI:10.1016/j.bmc.2015.04.062

Bioorganic & Medicinal Chemistry xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Preparation and biological evaluation of conformationally
constrained BACE1 inhibitors
Leonard L. Winneroski, Matthew A. Schiffler, Jon A. Erickson, Patrick C. May à, Scott A. Monk, David E. Timm,
James E. Audia , James P. Beck, Leonard N. Boggs, Anthony R. Borders, Robert D. Boyer, Richard A. Brier,
Kevin J. Hudziak, Valentine J. Klimkowski à, Pablo Garcia Losada, Brian M. Mathes, Stephanie L. Stout,
⇑
Brian M. Watson, Dustin J. Mergott
Lilly Research Laboratories, A Division of Eli Lilly & Co., Lilly Corporate Center, Indianapolis, IN 46285, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The BACE1 enzyme is a key target for Alzheimer’s disease. During our BACE1 research efforts, fragment
screening revealed that bicyclic thiazine 3 had low millimolar activity against BACE1. Analysis of the
co-crystal structure of 3 suggested that potency could be increased through extension toward the S3
pocket and through conformational constraint of the thiazine core. Pursuit of S3-binding groups
produced low micromolar inhibitor 6, which informed the S3-design for constrained analogs 7 and 8,
themselves prepared via independent, multi-step synthetic routes. Biological characterization of BACE
inhibitors 6–8 is described.
Received 5 February 2015
Revised 10 April 2015
Accepted 20 April 2015
Available online xxxx
Keywords:
BACE inhibitor
Complex organic synthesis
Conformational constraint
Stereoselective
Ó 2015 Elsevier Ltd. All rights reserved.
Structure-based drug design
1. Introduction
conformational constraint and S3-pocket binding hypotheses. The
successful testing of these medicinal chemistry hypotheses through
Alzheimer’s disease is a neurodegenerative disorder character-
ized by progressive memory loss ultimately leading to dementia.
A key hallmark of the disease is the abnormal extracellular accumu-
lation of Ab in the brain as soluble oligomers or insoluble plaques.1–5
Owing to its critical role in the production of Ab, BACE1 has been
aggressively pursued as a key target for development of a potential
Alzheimer’s disease therapy.6–8 We have previously reported on the
fragment-based discovery of LY2811376 (1, Fig. 1), the first mole-
cule reported to demonstrate robust reduction of human CSF Ab
in a Phase I clinical trial.9 More recently, we reported the discovery
of LY2886721 (2), a potent BACE1 inhibitor that reached Phase 2
clinical trials in early Alzheimer’s disease.10 As part of our
Alzheimer’s disease program, we also targeted topologically com-
plex, polycyclic aminothiazines as BACE1 inhibitors to test
multi-step organic synthesis and the biological characterization of
the resulting BACE inhibitors is described herein.
2. Results and discussion
2.1. General design considerations
During our fragment-based discovery efforts toward BACE1
inhibitors, we identified bicyclic aminothiazine fragment ( )-3
(Fig. 1), which weakly inhibited BACE1 in a high concentration
recombinant BACE1 enzyme assay. Crystallization of ( )-3 with
BACE1 revealed a binding mode in which the aminothiazine
engaged in a hydrogen bonding network with the catalytic aspartic
acids, with the fused cyclohexyl ring projecting into the S1 pocket
(Fig. 2). Based on the X-ray structural observations and
computational docking studies, we hypothesized that the potency
and efficiency of this fragment could be increased through two key
structural changes: conformational constraint to reinforce the
observed binding mode, and extension of suitable functionality
toward the S3 pocket of the enzyme.
Abbreviations: MBP, maltose binding protein; mcaFRET, (7-methoxycoumarin-
4-yl)acetyl modified fluorescence resonance energy transfer; MDCK, Madin-Darby
canine epithelial cells; NT, not tested; PDAPP, transgenic mice overexpressing
mutant human amyloid precursor protein V717F.
⇑
Corresponding author. Tel.: +1 317 277 2261; fax: +1 317 433 0552.
E-mail address: mergottdu@lilly.com (D.J. Mergott).
Thiazine 4 was initially targeted to test these hypotheses.
Although a conformational analysis revealed that the bicyclic
Current address: Constellation Pharmaceuticals, Cambridge, MA 02140, USA.
Retired.
à
http://dx.doi.org/10.1016/j.bmc.2015.04.062
0968-0896/Ó 2015 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Winneroski, L. L.; et al. Bioorg. Med. Chem. (2015), http://dx.doi.org/10.1016/j.bmc.2015.04.062

2
L. L. Winneroski et al. / Bioorg. Med. Chem. xxx (2015) xxx–xxx
syntheses of more complex constrained scaffolds were pursued.
This molecule exhibited an mcaFRET BACE1 IC50 of 37.5 M, a
ꢀ100-fold increase in potency compared to fragment 3.
Surprisingly, however, unsubstituted benzamide displayed
l
5
similar potency and improved efficiency compared to 4, prompting
re-evaluation of the substitution of the aromatic amide ring. This
led to the discovery of 6, which displayed a BACE1 IC50 value of
1.83 l
M (LEAN = 0.27).12 This represented a >1500-fold increase
in potency compared to fragment 3, a significant increase in ligand
efficiency, and resulted in BACE1 potency within ꢀ10-fold of
LY2811376 (1), which had already demonstrated robust reduction
of human CSF Ab.9 The BACE1 potency of 6 bolstered our interest in
the constrained scaffolds to increase potency.
An X-ray crystal structure of 6 bound in the BACE1 active site
indicated that the binding mode of fragment 3 was conserved, with
the amide extending toward the S3 pocket (Fig. 2). The binding
mode of amide 6 indicated that the core aromatic group preferred
to bind at the entrance to the S3 pocket with the chloro substituent
extending into the pocket. Space appeared to be available in the
BACE1 active site to constrain the bicyclic aminothiazine via
three-atom fused or bridged ring systems. Furan and pyran rings
were selected as constraints over the corresponding all-carbon
rings owing to the presumed metabolic and physicochemical prop-
erty benefits afforded by the oxygen-containing constraints. Based
on these X-ray observations and the SAR data for 4, 5, and 6, we
targeted meta-chloro analogs 7 and 8 (Fig. 1).
2.3. Chemistry
Figure 1. Aminothiazine BACE1 inhibitors.
2.3.1. Model tricyclic system
In parallel with the aforementioned SAR, the chemistry toward
tricycle 7 was piloted on model thiourea 9,13 lacking the amide
substituent. The two key steps in the proposed reaction sequence
were an iodocyclization reaction (conversion of 9–10) and a reduc-
tive etherification (conversion of 10–11) (Scheme 1). In order to
generate 11 via this sequence, the diastereoselectivity of the
iodocyclization would need to favor 10 over 12. Thiourea 9 was
prepared to explore the iodocyclization/etherification sequence.14
Upon treatment of 9 with I2 in CH2Cl2, iodocyclization proceeded
smoothly, affording a mixture of isomers in 76% yield. However,
undesired isomer 12 was the major product with a diastereomeric
ratio of roughly 90:10.15 Given the undesired selectivity exhibited
by 9, we elected to target a different cyclization substrate to access
the key tricyclic framework present in 7.
We reasoned that the preference for isomer 12 in the cycliza-
tion of 9 may have been driven by the development of a steric
interaction between the ester and the iodomethyl substituent in
the transition state leading to 10. In our revised synthetic
approach, we posited that constraining the ester via bridged lac-
tone 13 (Scheme 2) would reverse the stereochemical preference
of the cyclization in favor of 14. This was supported by computa-
tional modeling, which suggested that conformer 15a (leading to
Figure 2. X-ray crystal structures of BACE1 with 3 (A), 6 (B), 7 (C), and 8 (D).
Oxygen, nitrogen, chlorine and sulfur atoms are colored red, blue, green and yellow,
respectively. Hydrogen bond distances (in Å) are indicated in yellow.
thiazines 3 and 4 appeared to bind in their low energy conforma-
tions, it was hypothesized that additional constraints could essen-
tially lock the compounds into their binding conformations,
providing meaningful increases in potency.11
2.2. Initial SAR
Compound 4, which, like 2, contains a fluoropyridine S3 binding
group, was targeted to test the S3 binding hypothesis while
Scheme 1. Initial model system for iodocyclization.
Please cite this article in press as: Winneroski, L. L.; et al. Bioorg. Med. Chem. (2015), http://dx.doi.org/10.1016/j.bmc.2015.04.062

L. L. Winneroski et al. / Bioorg. Med. Chem. xxx (2015) xxx–xxx
3
afforded ( )-14 as the sole product in 75% yield, supporting our
hypothesis that constraining the ester via a bicyclic lactone would
alter the stereochemical outcome of the iodocyclization.15 After
extensive experimentation, it was found that one-pot reductive
ring opening of lactone ( )-14 and in situ cyclization of the result-
ing putative alkoxide ( )-19 afforded ( )-20 in 19% yield. Alcohol
( )-20 was then oxidized to ketone ( )-21, which was converted
to sulfinyl imine ( )-22 via Ti(IV)-promoted condensation with
t-butyl sulfinamide. In situ reduction of ( )-22 with sodium
borohydride proceeded exclusively through axial approach on the
conformationally locked system to furnish ( )-23 as the sole
observed diastereomer. Finally, deprotection of ( )-23 followed
by acylation to install the 3-chlorobenzoyl group, chiral separation,
and then removal of the benzoyl protecting group produced
aminothiazine 7. The synthesis of 7 included 18 steps and pro-
ceeded in 0.1% overall yield from commercially available materials.
2.3.3. Synthesis of 8
Constrained tricycle ( )-8 was prepared via a 21-step linear
sequence for study alongside 7 (Scheme 4).14 Acid chloride 2417
was reduced to the corresponding diol and mono-silylated to yield
( )-25. Two-step oxidation furnished acid ( )-26, which was sub-
jected to a Curtius rearrangement and hydroboration/oxidation.
Oxidation of the resulting crude alcohol, desilylation, and separa-
tion of regioisomers then provided ketone ( )-27. Hydrogenolysis
of the benzyl carbamate was followed by treatment with benzyl
isocyanate to provide urea ( )-28. Activation of the alcohol with
Ghosez’s reagent18 triggered cyclization to yield the thiazine
( )-29. To convert the ketone to the amine, a similar protocol
was employed to that used for the conversion of ( )-21 to ( )-23.
However, this reduction produced only a 3:2 mixture of diastere-
omers. Deprotection of the resulting sulfonamide yielded ( )-30,
which was acylated and deprotected to give the target compound
( )-8.
2.4. In vitro and in vivo evaluation
The successful delivery of aminothiazines 7 and ( )-8 enabled
us to test the hypothesis that conformational constraint would
increase potency. Tricycle 7 displayed a BACE1 IC50 of 2.25
similar to that of the unconstrained molecule 6, while ( )-8
demonstrated a BACE1 IC50 of 36.1 M, which was considerably
lM,
l
less potent than 6 (Table 1). The crystal structure of 7 complexed
to BACE1 suggested that the conformational constraint imposed
by the bridged pyran system had indeed reinforced the bound con-
formation of 6 (compare Fig. 2, panels B and C). However, this con-
formational constraint did not translate to greater in vitro BACE1
potency. Gas phase conformational analysis of 6 estimated that
the alternate boat conformation of the cyclohexyl ring to be
approximately 1.3 kcal/mol above the favored chair, which would
result in a small but significant population; however, the antici-
pated entropic gain in binding free energy was not realized when
this conformation was eliminated by the pyran ring constraint in
7. On the other hand, the X-ray structure of 8 suggests that substi-
tution on the bridgehead position to install the constraint, is not
well tolerated in the binding pocket as it evidentially projects too
far into the mobile flap region, resulting in decreased potency.
Further biochemical characterization revealed that 6 and 7 did
not display significant cathepsin D inhibition up to concentrations
Scheme 2. Putative conformations leading to 14 (15a) and epi-14 (15b). Note the
strain shown in green disfavoring 15b.
14) was significantly lower in energy than conformer 15b (leading
to epi-14) owing to a steric interaction (shown in green) in 15b
between the iodonium species and the carbon bridge.16 We
expected that this energy difference would be reflected in the tran-
sition states arising from 15a and 15b, thereby favoring 14.
Furthermore, this constrained system would contain functionality
at C(6), enabling installation of the amide side chain late in the
synthetic sequence, which would facilitate additional SAR.
2.3.2. Synthesis of 7
Toward the synthesis of 7 (Scheme 3), allyl carbonate 1614 was
treated with n-BuLi followed by Pd(PPh3)4 to induce an allylic sub-
stitution, furnishing ( )-17 in 31% isolated yield, along with 17% of
the undesired epimer ( )-18. Cleavage of the t-butyl ester and sub-
sequent Curtius rearrangement afforded the corresponding benzyl
carbamate. Removal of the Cbz group and acylation with benzoyl
isothiocyanate then afforded thiourea ( )-13 as the key iodocy-
clization substrate. Treatment of ( )-13 with I2 in dichloromethane
of 100 and 300 lM, respectively. However, neither compound dis-
played selectivity against BACE2.19 Compounds 6 and 7 also pos-
sessed low mouse microsomal metabolism. Given the single-digit
micromolar potency, low microsomal mouse metabolism, and the
level of synthetic investment in the initial SAR, we elected to use
6 and 7 as tool compounds for in vivo assessment. Both molecules
were administered subcutaneously at 100 mg/kg to young female
Please cite this article in press as: Winneroski, L. L.; et al. Bioorg. Med. Chem. (2015), http://dx.doi.org/10.1016/j.bmc.2015.04.062

4
L. L. Winneroski et al. / Bioorg. Med. Chem. xxx (2015) xxx–xxx
Scheme 3. Synthesis of 7. Reagents and conditions: (a) n-BuLi, toluene, À78 °C, then Pd(PPh3)4, À78 °C to rt, 31% ( )-17 and 17% ( )-18; (b) ( )-17, TFA, CH2Cl2, 96%; (c) DPPA,
BnOH, Et3N, toluene, 79%; (d) Pd(OAc)2, Et3SiH, Et3N, CH2Cl2, rt, 77%; (e) benzoyl isothiocyanate, THF, rt, 91%; (f) I2, CH2Cl2, 0 °C–rt, 75%; (g) LiBH4, THF, H2O, 19%; (h) TPAP,
NMO, CH2Cl2, 85%; (i) ( )-tert-butyl sulfinamide, Ti(OEt)4, THF, 65 °C, then NaBH4, MeOH, À30 °C, 99%; (j) HCl, Et2O, 91%; (k) 3-chlorobenzoic acid, HOBt, EDCI, DIEA, THF, 91%;
(l) chiral SFC, 43%; (m) MeONH2–HCl, pyridine, EtOH, 50 °C, 93%.
Table 1
Physicochemical and selectivity data for BACE1
6ÁHCl
7ÁHCl
( )-8ÁHCl
31ÁTFA
MW
324
2.6
68
366
2.2
77
366
2.2
77
325
3.5
65
clogP
PSA (Å2)
pKa
9.7
9.0
8.2
—
BACE1 IC50
LEAN
BACE2 IC50
(
(
l
M)
M)
1.83
0.27
0.497
>100
10
2.25
0.24
1.34
>300
8.5
36.1
0.19
22.7
277
NT
0.761
0.29
0.584
>100
2.7
l
Cathepsin D IC50
Pgp efflux ratio
(lM)
Permeability (10À6 cm/s)
2.9
2.2
NT
7.2
700
600
500
400
300
200
100
0
57%
p < 0.01
Scheme 4. Synthesis of 8. Reagents and conditions: (a) LiAlH4, THF, 80 °C, 92%; (b)
NaH, THF, then TBSCl, rt, 83%; (c) PCC, CH2Cl2, rt, 78%; (d) NaClO2, 2-methyl-2-
butene, t-BuOH, THF, NaH2PO4(aq), 0 °C–rt, 100%; (e) BnOH, DIEA, toluene; DPPA,
110 °C, 77%; (f) BH3–S(CH3)2, THF, then NaOH(aq), H2O2 (aq); (g) DMP, CH2Cl2, rt,
88% (2 steps); (h) TBAF, THF, 50%; (i) H2, Pd/C, EtOH, EtOAc, then benzoyl
isothiocyanate, 76%; (j) 1-chloro-N,N,2-trimethylpropenylamine, CH2Cl2, 91%; (k)
( )-tert-butyl sulfinamide, Ti(OEt)4, THF, 65 °C, then NaBH4, MeOH, 0 °C–rt, 52%, 3:2
dr; (l) HCl, Et2O, CH2Cl2, 82%; (m) 3-chlorobenzoic acid, HOBt, EDCI, DIEA, THF, 24%;
(n) MeONH2–HCl, pyridine, EtOH, 50 °C, 94%.
Veh.
Cmpd 6 Cmpd 7 Pos. Cntrl.
Figure 3. Brain Ab concentrations in PDAPP mice after subcutaneous injection of 6
and 7.
PDAPP (V717F) transgenic mice20,21 and total Ab was measured in
brain cortex samples by sandwich ELISAs.9,22 Unlike other mole-
cules with low micromolar BACE1 potency that we had tested in
PDAPP mice, neither 6 nor 7 produced significant changes in Ab
concentrations in brain (Fig. 3). Concentrations of compounds 6
and 7 were measured at 9.3 and 6.7 lM in plasma, but 0.75 and
Please cite this article in press as: Winneroski, L. L.; et al. Bioorg. Med. Chem. (2015), http://dx.doi.org/10.1016/j.bmc.2015.04.062

L. L. Winneroski et al. / Bioorg. Med. Chem. xxx (2015) xxx–xxx
5
0.61
l
M in brain, respectively. The lack of reduction in brain Ab
(3Â). The combined organic layers were dried (Na2SO4) and the
solvent was removed to afford ( )-tert-butyl 3-oxo-5-exo-vinyl-2-
oxabicyclo[2.2.2]octane-4-carboxylate (1.36 g, 96% yield). 1H
NMR (400 MHz, CDCl3): d 5.87–5.81 (m, 1H), 5.28 (s, 1H), 5.26–
5.24 (m, 1H), 4.79–4.77 (m, 1H), 3.24–3.20 (m, 1H), 2.40–2.34
(m, 2H), 2.16–2.11 (m, 2H), 1.83–1.75 (m, 1H), 1.68 (ddd, J = 14.4,
4.4, 1.3 Hz, 1H). LCMS (ESIÀ) (m/z) 195 [MÀH]À.
levels was consistent with the low brain exposure, with total brain
concentrations of each compound representing less than 0.3-fold
the respective in vitro IC50 value. The low in vivo brain penetration
for these compounds is consistent with the low permeability and
high P-gp mediated efflux subsequently observed in MDCK cells
(Table 1).23 Additional SAR within the bicyclic scaffold demon-
strated that replacement of the amide with an ester linkage (31,
Fig. 1) eliminated P-gp efflux and increased permeability
(Table 1). These data suggest that the amide linkage contributes
to the P-gp efflux and low permeability observed with 6 and 7.
A room temperature solution of the product from the previous
step (1.36 g, 6.93 mmol), benzyl alcohol (3.00 g, 27.7 mmol) and
triethylamine (2.81 g, 27.7 mmol) in toluene (60 mL) was treated
with diphenylphosphonic azide (2.29 g, 8.32 mmol), then the reac-
tion was heated to reflux for 4 h with stirring under nitrogen. The
reaction was cooled to room temperature, and then diluted with
water and EtOAc. The organic layer was washed with 1 N aqueous
HCl and brine. The aqueous layer was extracted two more times
with EtOAc. The combined organic layers were dried (Na2SO4)
and the solvent was removed to afford crude product that was
purified on silica gel with a 0–100% gradient of EtOAc in hexanes
to afford ( )-4-amino-5-exo-vinyl-2-oxabicyclo[2.2.2]octan-3-one
(1.66 g, 79% yield). Rf = 0.30 (1:1 hexanes/EtOAc, KMnO4 stain to
visualize). 1H NMR (400 MHz, CDCl3): d 7.36–7.27 (m, 5H), 5.98–
5.95 (m, 1H), 5.21–5.12 (m, 3H), 5.08 (s, 2H), 4.71–4.68 (m, 1H),
3.06–2.96 (m, 1H), 2.60–2.48 (m, 1H), 2.45–2.36 (m, 1H), 2.20–
1.96 (m, 2H), 1.89–1.81 (m, 1H), 1.69–1.64 (m, 1H). LCMS (ESI+)
(m/z) 302 [M+H]+.
3. Conclusions
After aminothiazine 3 was identified as a promising fragment
starting point through high concentration in vitro testing and
X-ray fragment soaking experiments, a structure-based design
approach was pursued to optimize this fragment. This strategy
was enabled by the execution of challenging synthetic chemistry
which, in the case of constrained molecule 7, required a second
generation approach to produce the final molecule. Although com-
pounds 6 and 7 did not achieve significant Ab reduction in PDAPP
mice owing to low brain exposure, the successful synthesis of the
compounds enabled key SAR hypotheses to be tested, revealing
that the potency of this series of BACE inhibitors was increased
by >1500-fold through S3 binding, but was not further enhanced
through conformational constraint. This work exemplifies the con-
tinued importance of synthetic organic chemistry in the pursuit of
increasingly difficult drug discovery targets.24
A
room temperature mixture of triethylsilane (1.60 g,
13.77 mmol) and triethylamine (84 mg, 826 mol) in dry CH2Cl2
(40 mL) was treated with Pd(OAc)2 (93 mg, 413 mol) and stirred
l
l
under nitrogen for 5 min and the dark mixture was then treated
dropwise over 5 min with a solution of the product from the pre-
vious step (1.66 g, 5.51 mmol) in CH2Cl2 (40 mL) and the reaction
was stirred at room temperature for 7 h under nitrogen. The sol-
vent was removed, and the crude product was purified on silica
gel with a 1–10% (80% MeOH/20% isopropylamine) in CH2Cl2 gradi-
ent, and the mixed fractions were purified with a 10 g SCX column
using 2:1 CH2Cl2/MeOH and then 2:1 (CH2Cl2/7 N NH3) in MeOH to
4. Experimental section
4.1. Chemistry
4.1.1. ( )-tert-Butyl 3-oxo-5-exo-vinyl-2-oxabicyclo[2.2.2]-
octane-4-carboxylate (( )-17)
afford
( )-4-amino-5-exo-vinyl-2-oxabicyclo[2.2.2]octan-3-one
A À78 °C solution of 16 (7.457 g, 22.7 mmol) in dry toluene
(42 mL) was treated dropwise with a solution of n-butyl lithium
in hexanes (1.6 M, 14.9 mL, 23.9 mmol) and the reaction was stir-
red at À78 °C for 2 min under nitrogen and was then treated with
tetrakis(triphenylphosphine)palladium (1.31 g, 1.14 mmol), and
the reaction was warmed to room temperature and stirred over-
night. The reaction was quenched with 1.0 N aqueous HCl, diluted
with water, and extracted three times with EtOAc. The combined
organic layers were dried (Na2SO4) and the solvent was removed
to afford crude product that was purified on silica gel with a step
gradient of 10% then 20% EtOAc in hexanes to afford ( )-17
(1.778 g, 31% yield) and ( )-18 (0.969 g, 17% yield).
(0.714 g, 77% yield). 1H NMR (400 MHz, CDCl3): d 5.97–5.88 (m,
1H), 5.27 (dt, J = 10.0, 1.1 Hz, 1H), 5.19 (dt, J = 17.0, 1.1 Hz, 1H),
4.67–4.64 (m, 1H), 2.47–2.46 (m, 1H), 2.38–2.34 (m, 1H), 2.12–
2.06 (m, 2H), 1.86–1.80 (m, 1H), 1.66–1.61 (m, 2H). LCMS (ESI+)
(m/z) 168 [M+H]+.
A room temperature mixture of the product from the previous
step (0.711 g, 4.25 mmol) in CH2Cl2 (30 mL) was treated with ben-
zoyl isothiocyanate (833 mg, 5.10 mmol). The reaction was stirred
at room temperature for 45 min under nitrogen. The solvent was
removed to afford crude product that was purified on silica gel
with
a 0–100% EtOAc in hexanes gradient to afford ( )-13
(1.273 g, 91% yield). Rf = 0.36 (1:1 hexanes/EtOAc, I2 stain). 1H
NMR (400 MHz, CDCl3): d 10.90 (br s, 1H), 8.9 (br s, 1H),
7.81–7.78 (m, 2H), 7.61–7.59 (m, 1H), 7.50–7.48 (m, 2H),
5.97–5.93 (m, 1H), 5.45–5.34 (m, 2H), 4.73–4.70 (m, 1H), 2.53–
2.51 (m, 1H), 2.34–2.33 (m, 3H), 1.94–1.93 (m, 1H), 1.76–1.75
(m, 2H). LCMS (ESI+) (m/z) 331 [M+H]+.
( )-17: Rf = 0.25 (3:1 hexanes/EtOAc, KMnO4 stain) 1H NMR
(400 MHz, CDCl3): d 5.78–5.72 (m, 1H), 5.21 (dt, J = 5.7, 1.3 Hz,
1H), 5.18–5.17 (m, 1H), 4.69–4.66 (m, 1H), 3.21–3.19 (m, 1H),
2.32–2.29 (m, 1H), 2.16–2.09 (m, 3H), 1.74–1.68 (m, 1H), 1.61–
1.58 (m, 1H), 1.45 (s, 9H). LCMS (ESI+) (m/z) 197 [M+2HÀt-Bu]+.
( )-18: Rf = 0.15 (3:1 hexanes/EtOAc, KMnO4 stain) 1H NMR
(400 MHz, CDCl3): d 5.79–5.73 (m, 1H), 5.03–5.01 (m, 1H), 4.99–
4.98 (m, 1H), 4.67–4.64 (m, 1H), 2.95–2.90 (m, 1H), 2.25–2.17
(m, 1H), 2.08–2.04 (m, 3H), 1.89–1.84 (m, 1H), 1.79–1.76 (m,
1H), 1.43 (s, 9H). LCMS (ESI+) (m/z) 197 [M+2HÀt-Bu]+.
4.1.3. N-((4RS,4aRS,6SR,8aSR)-4-(Iodomethyl)-9-oxo-4,4a,5,6,7,
8-hexahydro-6,8a-(epoxymethano)benzo[d][1,3]thiazin-2-yl)-
benzamide (( )-14)
A 0 °C solution of ( )-13 (1.28 g, 3.87 mmol) in CH2Cl2 (70 mL)
was treated with I2 (1.97 g, 7.75 mmol) and the reaction was
warmed to room temperature and stirred for 30 min under nitro-
gen. The red-brown reaction slurry was diluted with aqueous
sodium bisulfite and extracted three times with EtOAc. The com-
bined organic layers were dried (Na2SO4) and the solvent was
removed to afford crude product that was purified with silica gel
with a 5–100% EtOAc in hexanes gradient to afford the titled
4.1.2. ( )-N-((3-Oxo-5-exo-vinyl-2-oxabicyclo[2.2.2]octan-4-
yl)carbamothioyl)benzamide (( )-13)
A solution of ( )-17 (1.82 g, 7.21 mmol) in CH2Cl2 (40 mL) was
treated with TFA (20 mL) and stirred at room temperature for 1
h. The solvent was removed, and the residue was diluted with
water and EtOAc. The aqueous layer was extracted with EtOAc
Please cite this article in press as: Winneroski, L. L.; et al. Bioorg. Med. Chem. (2015), http://dx.doi.org/10.1016/j.bmc.2015.04.062

6
L. L. Winneroski et al. / Bioorg. Med. Chem. xxx (2015) xxx–xxx
product (1.323 g, 75% yield). Rf = 0.09 (1:1 hexanes/EtOAc). The rel-
ative stereochemistry was assigned based on the measured cou-
pling constant of 11.1 Hz between the adjacent methane protons
(H4 and H4a), which agreed with predicted coupling constants of
the minimized structure of the isomer shown above. The predicted
methine coupling constants is 11.0 Hz for the isomer shown and
3.5 Hz for the isomer with the opposite iodomethylene configura-
tion. 1H NMR (400 MHz, CDCl3): d 8.10–8.09 (m, 2H), 7.50–7.47 (m,
1H), 7.41–7.37 (m, 2H), 4.80–4.78 (m, 1H), 3.53–3.50 (m, 1H),
3.46–3.42 (m, 2H), 2.66–2.62 (m, 1H), 2.43–2.40 (m, 3H), 2.02–
2.01 (m, 2H), 1.64–1.63 (m, 2H). LCMS (ESI+) (m/z) 457 [M+H]+.
three times with EtOAc. The combined organic layers were dried
(Na2SO4) and the solvent was removed to afford crude product
that was purified on silica gel with a 15 min 0.5–10% MeOH in
CH2Cl2 gradient to afford the title product (149 mg, 99% yield).
Rf = 0.30 (9:1 CH2Cl2/MeOH). 1H NMR (400 MHz, CDCl3): d 8.20–
8.18 (m, 2H), 7.51–7.47 (m, 1H), 7.42 (t, J = 7.4 Hz, 2H), 4.19–
4.12 (m, 1H), 3.89–3.86 (m, 1H), 3.76 (d, J = 11.4 Hz, 2H), 3.39–
3.30 (m, 1H), 3.22 (d, J = 5.7 Hz, 1H), 3.08 (d, J = 1.5 Hz, 1H),
2.11–2.03 (m, 3H), 1.77–1.67 (m, 4H), 1.20 (s, 9H). 13C
(100 MHz, CDCl3): d 131.9, 129.3, 128.1, 74.44, 74.40, 53.6, 53.0,
52.2, 52.1, 42.7, 42.6, 36.6, 36.3, 35.3, 34.1, 31.4, 31.2, 29.4,
28.6, 22.6, 22.1. HRMS (ESI+) (m/z) calcd for C21H30N3O3S2
(M+H): 436.1723, found 436.1725.
4.1.4. N-((4RS,4aRS,6SR,8aSR)-6-Hydroxy-4,4a,5,6,7,8-hexahy-
dro-1H,3H-4,8a-(epithiomethenoazeno)isochromen-10-yl)-
benzamide (( )-20)
4.1.7. (+)-N-((4S,4aS,6S,8aR)-10-Amino-4,4a,5,6,7,8-hexahydro-
1H,3H-4,8a-(epithiomethenoazeno)isochromen-6-yl)-3-
chlorobenzamide (7)
A room temperature solution of ( )-14 (1.341 g, 2.94 mmol) in
98:2 THF/water (97 mL) was treated quickly with a 2 M solution
of LiBH4 in THF (14.7 mL, 29.39 mmol), and the reaction was stir-
red at room temperature for 10 min under nitrogen. The reaction
was quenched with MeOH (30 mL) and was stirred at room tem-
perature for 2 min until a clear solution developed. The solvent
was removed to afford crude product that was purified with silica
gel with a step gradient of 75% EtOAc in hexanes to 100% EtOAc,
then 10% MeOH in CH2Cl2 to afford the titled product (185 mg,
19% yield). Rf = 0.29 (9:1 CH2Cl2/MeOH). 1H NMR (400 MHz,
CD3OD): d 8.07–8.04 (m, 2H), 7.48–7.46 (m, 1H), 7.39–7.35 (m,
2H), 4.11–4.09 (m, 2H), 3.96–3.92 (m, 1H), 3.62–3.59 (m, 1H),
3.44–3.40 (m, 1H), 3.09–3.07 (m, 1H), 2.61–2.58 (m, 1H), 1.88–
1.87 (m, 5H), 1.48–1.45 (m, 1H). 13C (100 MHz, CD3OD): d 131.4,
128.8, 127.7, 76.6, 74.0, 64.4, 53.1, 42.8, 32.8, 26.6, 25.3. HRMS
(ESI+) (m/z) calcd for C17H21N2O3S (M+H): 333.1273, found
333.1273.
A 0 °C solution of ( )-23 (211 mg, 484
lmol) in CH2Cl2 (6 mL)
was treated with a 1 N Et2O solution of HCl (3.39 mL, 3.39 mmol).
A white slurry developed that was warmed to room temperature
and stirred under nitrogen for 30 min. The solvent was removed,
and the crude product was split and purified in parallel on two
10 g SCX columns, each using 2:1 CH2Cl2 then 2:1 (CH2Cl2/7 N
NH3) in MeOH to elute the product. The solvent was removed from
the basic washes to afford N-((4RS,4aRS,6RS,8aSR)-6-amino-
4,4a,5,6,7,8-hexahydro-1H,3H-4,8a-(epithiomethenoazeno)iso-
chromen-10-yl)benzamide (146 mg, 91% yield). 1H NMR (400 MHz,
CDCl3): d 8.19 (d, J = 7.2 Hz, 2H), 7.50–7.47 (m, 1H), 7.42 (t,
J = 7.3 Hz, 2H), 4.16 (d, J = 11.4 Hz, 1H), 3.85 (d, J = 11.3 Hz, 1H),
3.72 (d, J = 11.4 Hz, 1H), 3.28 (d, J = 11.4 Hz, 1H), 3.07 (s, 1H),
2.91–2.76 (m, 1H), 2.12–2.10 (m, 1H), 1.94–1.91 (m, 1H), 1.78 (d,
J = 12.4 Hz, 1H), 1.73–1.63 (m, 1H), 1.58–1.41 (m, 3H). 13C
(100 MHz, CDCl3): d 137.0, 131.7, 129.3, 128.0, 74.5, 52.2, 42.9,
36.7, 31.5. HRMS (ESI+) (m/z) calcd for C17H22N3O2S (M+H):
332.1433, found 332.1423.
4.1.5. N-((4RS,4aRS,8aSR)-6-Oxo-4,4a,5,6,7,8-hexahydro-1H,3H-
4,8a-(epithiomethenoazeno)isochromen-10-yl)benzamide (( )-
21)
A solution of the product from the previous step (146 mg, 441
A room temperature solution of ( )-20 (183 mg, 551
lmol) and
lmol), 3-chlorobenzoic acid (83 mg; 529
lmol), and 1-hydroxy-
N-methylmorpholine-N-oxide (129 mg, 1.10 mmol) in dry CH2Cl2
benzotriazole hydrate (88 mg; 573 mol) in THF (4 mL) was
l
(16 mL) was treated with tetrapropylammonium perruthenate
treated with diisopropylethylamine (114 mg, 881
l
mol). Then,
(19 mg, 550
lmol) and the reaction was stirred under nitrogen
1-(3-dimethylaminopropyl)-3-ethylcarbodiimide
hydrochloride
for 1 h. The solvent was removed, and the residue was purified
on silica gel with a 0.5–10% MeOH in CH2Cl2 gradient to afford
the titled product (154 mg, 85% yield). Rf = 0.37 (9:1
CH2Cl2/MeOH). 1H NMR (400 MHz, CDCl3): d 8.13–8.12 (m, 2H),
7.51–7.50 (m, 1H), 7.43–7.39 (m, 2H), 4.22–4.19 (m, 1H), 3.85–
3.83 (m, 2H), 3.36–3.32 (m, 1H), 3.09–3.07 (m, 1H), 2.82–2.79
(m, 2H), 2.44–2.42 (m, 3H), 2.02–2.00 (m, 1H), 1.81–1.78 (m,
1H). 13C (100 MHz, CDCl3): d 207.4, 132.2, 128.9, 128.3, 73.9,
68.1, 63.7, 51.9, 48.4, 42.6, 42.1, 37.1, 36.4, 33.9, 32.5, 22.7, 14.1.
HRMS (ESI+) (m/z) calcd for C17H19N2O3S (M+H): 331.1116, found
331.1111.
(110 mg, 573 lmol) was added, and the reaction was stirred at
room temperature for 1.5 h under nitrogen. The reaction mixture
was diluted with EtOAc, water, and saturated aqueous NaHCO3,
and then the combined aqueous layers were extracted three times
with EtOAc. The combined organic layers were dried (Na2SO4) to
give crude product that was purified on silica gel with a 0.5–10%
MeOH in CH2Cl2 gradient to afford (À)-N-((4S,4aS,6S,8aR)-10-ben-
zamido-4,4a,5,6,7,8-hexahydro-1H,3H-4,8a (epithiomethenoaze-
no)isochromen-6-yl)-3-chlorobenzamide (188 mg, 91% yield).
Rf = 0.47 (9:1 CH2Cl2/MeOH). This material was subjected to chiral
SFC (Column: Chiralpak AD-H, 2.1 cm  15 cm; eluent: 40% iso-
propanol and 60% CO2; flow 70 mL/min at UV 290 nm). Analysis
of the first eluting isomer (Column: Chiralpak AD-H, 0.46 cm Â
15 cm; 40% isopropanol and 60% CO2; flow 5 mL/min at UV
225 nm) indicated that the product was isolated in >99% ee (tR
(major) = 1.3 min; tR (minor) = 2.0 min), (69 mg, 43% yield). 1H
NMR (CDCl3, 400 MHz): d 11.6 (br s, 1H), 8.18–8.16 (m, 2H), 7.75
(t, J = 1.8 Hz, 1H), 7.60 (dt, J = 7.8, 1.2 Hz, 1H), 7.51–7.45 (m, 5H),
6.25 (d, J = 8.0 Hz, 1H), 4.20–4.13 (m, 2H), 3.89 (dd, J = 1.3,
11.7 Hz, 1H), 3.77 (d, J = 11.4 Hz, 1H), 3.35 (d, J = 11.5 Hz, 1H),
3.08–3.07 (m, 1H), 2.16 (dt, J = 12.3, 3.4 Hz, 1H), 2.06–1.97 (m,
2H), 1.79–1.67 (m, 4H). 13C NMR (100 MHz, CDCl3): d 165.5,
136.0, 134.8, 131.9, 131.6, 129.9, 129.3, 128.1, 127.4, 124.9, 74.4,
4.1.6. N-((4RS,4aRS,6RS,8aSR)-6-((tert-Butylsulfinyl)amino)-
4,4a,5,6,7,8-hexahydro-1H,3H-4,8a-(epithiomethenoazeno)-
isochromen-10-yl)benzamide (( )-23)
A solution of ( )-21 (114 mg, 345
propane-2-sulfinamide (150 mg, 414.03
was treated with Ti(OEt)4 (197 mg, 863
l
mol) and ( )-2-methyl-
mol) in THF (6 mL)
mol) and then heated
l
l
to 65 °C under nitrogen for 4 h. The reaction was cooled to
À30 °C and treated portion-wise with NaBH4 (25 mg, 690 mol)
and the reaction mixture was stirred for min, then was
l
5
quenched with MeOH (1 mL) and warmed to room temperature
and stirred under nitrogen for 20 min. The reaction was treated
with water, diluted with EtOAc, and the milky mixture was fil-
tered through a small amount of diatomaceous earth using
EtOAc to rinse. The filtrate was diluted with water and washed
52.1, 47.3, 42.7, 36.4, 32.7, 31.3, 27.0. HRMS (ESI+) (m/z) calcd for
20
C
24H25ClN3O3S (M+H): 470.1305, found 470.1302. [
a]
À181.70
D
(c 10, CH2Cl2).
Please cite this article in press as: Winneroski, L. L.; et al. Bioorg. Med. Chem. (2015), http://dx.doi.org/10.1016/j.bmc.2015.04.062

L. L. Winneroski et al. / Bioorg. Med. Chem. xxx (2015) xxx–xxx
7
A solution of the product from the previous step (69 mg,
147 mol), O-methylhydroxylamine hydrochloride (123 mg,
A 0 °C solution of the product from the previous step (7.66 g,
25.84 mmol) in THF (50 mL), t-butyl alcohol (75 mL) and 2-
methyl-2-butene (50 mL) was treated with a freshly prepared solu-
tion of sodium chlorite (29.21 g, 258.36 mmol), and then a solution
of monosodium phosphate monohydrate (35.65 g, 258.36 mmol)
in water (100 mL) was added slowly to the solution. The solution
was stirred at 0 °C for 30 min and was then warmed to room tem-
perature and stirred for 16 h. The reaction was quenched with
aqueous saturated ammonium chloride. Dichloromethane and
water were added, and the organic layer was separated and the
aqueous layer was extracted with more CH2Cl2. The combined
organic extracts were dried over MgSO4 and the solvent was
removed to afford crude product that was purified over silica gel
with a 0–50% EtOAc in hexanes gradient to afford the titled product
(8.58 g, 100% yield). 1H NMR (400 MHz, CDCl3): d 5.73–5.72 (m,
2H), 4.27 (d, J = 8.6 Hz, 1H), 3.90 (d, J = 8.4 Hz, 1H), 3.70–3.55 (m,
4H), 2.53–2.51 (m, 1H), 2.34–2.29 (m, 1H), 2.03–1.98 (m, 1H),
1.27–1.18 (m, 1H), 0.85 (s, 9H), 0.01 (s, 3H), 0.00 (s, 3H). LCMS
(ESI+) (m/z) 313 [M+H]+.
l
1.5 mmol), and pyridine (116 mg; 1.5 mmol) in ethanol (4 mL)
was heated to 50 °C overnight. The solvent was removed to afford
a crude residue that was purified on silica gel with a 1–10% metha-
nol (w/20% isopropylamine) in CH2Cl2 gradient to furnish 7 (50 mg,
93% yield). 1H NMR (400 MHz, CDCl3): d 7.75 (t, J = 1.8 Hz, 1H), 7.63
(d, J = 7.7 Hz, 1H), 7.47 (ddd, J = 8.0, 2.0, 1.0 Hz, 1H), 7.37 (t,
J = 7.9 Hz, 1H), 6.11 (d, J = 7.8 Hz, 1H), 4.15–4.07 (m, 2H), 3.81
(dd, J = 1.2, 11.5 Hz, 1H), 3.64 (d, J = 11.0 Hz, 1H), 3.22 (d,
J = 10.9 Hz, 1H), 3.07 (d, J = 1.1 Hz, 1H), 1.93–1.86 (m, 3H), 1.68–
1.58 (m, 4H). 13C NMR (100 MHz, CDCl3): d 165.4, 152.0, 136.4,
134.7, 131.5, 129.9, 127.3, 125.0, 74.7, 51.5, 48.1, 42.3, 34.0, 33.9,
33.2, 27.6, 21.1. HRMS (ESI+) (m/z) calcd for C17H21ClN3O2S
20
(M+H): 366.1043, found 366.1042. [
a]
+3.70 (c 10, CH2Cl2).
D
4.1.8. (3aRS,7aRS)-7a-(((tert-Butyldimethylsilyl)oxy)methyl)-
1,4,7,7a-tetrahydroisobenzofuran-3a(3H)-yl)methanol (( )-25)
A 80 °C solution of lithium aluminum hydride (1.0 M in THF,
65.04 mL, 65.04 mmol) was treated dropwise with a solution of
2416 (8.1 g, 32.52 mmol) in THF (50 mL). The reaction stirred at
80 °C for 2 h and was then cooled to room temperature and
quenched with Na2SO4Á10H2O. Ethyl acetate and water were
added, and the organic layer was separated and the aqueous layer
was extracted with EtOAc. The combined organic extracts were
washed with brine and then dried over Na2SO4 and the solvent
4.1.10. Benzyl ((3aRS,7aRS)-7a-(hydroxymethyl)-6-oxohexahy-
droisobenzofuran-3a(3H)-yl)carbamate (( )-27)
( )-26 (7.61 g, 24.35 mmol), benzyl alcohol (3.95 g,
36.53 mmol), and diisopropylethylamine (12.59 g, 97.41 mmol)
were dissolved in toluene (100 mL). Diphenylphosphonic azide
(10.05 g, 36.53 mmol) was added, and the reaction was heated to
110 °C for 18 h. The reaction was cooled to room temperature.
Ethyl acetate and water were added. The organic layer was sepa-
rated, and the aqueous layer was extracted with EtOAc. The com-
bined organic extracts were dried over Na2SO4, and the solvent
was removed to afford crude product that was purified on silica
gel with a 0–30% EtOAc in hexanes gradient to afford benzyl
((3aRS,7aSR)-7a-(((tert-butyldimethylsilyl)oxy)methyl)-1,4,7,7a-
tetrahydroisobenzofuran-3a(3H)-yl)carbamate (7.82 g, 77% yield).
1H NMR (400 MHz, CDCl3): d 7.32–7.22 (m, 5H), 5.60 (s, 2H), 5.02
(s, 2H), 4.28–4.26 (m, 1H), 4.06–4.02 (m, 1H), 3.85–3.81 (m, 2H),
3.66–3.59 (m, 2H), 2.75–2.58 (m, 2H), 2.19–2.02 (m, 2H), 0.84 (s,
9H), 0.057 (s, 3H), 0.046 (s, 3H). LCMS (ESI+) (m/z) 418 [M+H]+.
A 0 °C solution of the product from the previous step (3.36 g,
8.05 mmol]) in THF (50 mL) was treated with borane–methyl sul-
fide complex (3.06 g, 40.23 mmol) and the reaction was stirred at
0 °C for 1 h and was then warmed to room temperature and stirred
for 2 h. Aqueous 1 N NaOH (50 mL) was added slowly to the reac-
tion and the reaction was stirred for 30 min. Hydrogen peroxide
(30% aqueous solution, 30 mL, 295.81 mmol) was then added,
and the reaction was stirred at room temperature for 2 h. The reac-
tion was diluted with (3:1 CH2Cl2/isopropanol) and water. The
organic layer was separated, and the aqueous layer was extracted
with CH2Cl2. The combined organic extracts were dried over
Na2SO4 and the solvent was removed to afford crude product that
was purified on silica gel with a 0–50% EtOAc in hexanes gradient
to afford 2.52 g (72%) of an intermediate regioisomeric mixture of
alcohols that was dissolved in CH2Cl2 (20 mL) and treated with
3,3,3-triacetoxy-3-iodophthalide (5.06 g, 11.57 mmol) and the
was
removed
to
afford
(cis-4,7-dihydroisobenzofuran-
3a,7a(1H,3H)-diyl)dimethanol (5.49 g, 92% yield). 1H NMR
(400 MHz, CDCl3): d 5.68 (t, J = 1.8 Hz, 2H), 3.81 (s, 2H), 3.79 (d,
J = 2.5 Hz, 2H), 3.65–3.63 (m, 2H), 3.58 (d, J = 11.5 Hz, 2H), 2.56–
2.53 (m, 2H), 2.29–2.22 (m, 2H), 2.09–2.03 (m, 2H).
Sodium hydride (60% dispersion in mineral oil, 1.60 g,
39.89 mmol) was added to THF (50 mL). This suspension was
treated dropwise with a solution of the product from the previous
step (7.35 g, 39.9 mmol) in THF (50 mL). The reaction stirred at
room temperature for 30 min and then t-butyldimethylchlorosi-
lane (6.07 g, 39.89 mmol) was added to the reaction and it was
stirred at room temperature for an additional 18 h. Ethyl acetate
and water were added, and the organic layer was separated and
the aqueous layer was extracted with EtOAc. The combined
organic extracts were washed with brine and then dried over
Na2SO4 and the solvent was removed to afford crude product that
was purified on silica gel with a 0–50% EtOAc in hexanes gradient
to afford the titled product (9.87 g, 83% yield). 1H NMR (400 MHz,
CDCl3): d 5.70–5.67 (m, 2H), 3.81–3.72 (m, 4H), 3.62 (dd, J = 8.2,
5.3 Hz, 2H), 3.56 (d, J = 10.6 Hz, 1H), 3.51–3.47 (m, 1H), 3.33–
3.29 (m, 1H), 2.32–2.27 (m, 2H), 2.09–2.03 (m, 2H), 0.89 (s, 9H),
0.07 (s, 6H).
4.1.9. (3aRS,7aSR)-7a-(((tert-Butyldimethylsilyl)oxy)methyl)-
1,4,7,7a-tetrahydroisobenzofuran-3a(3H)-carboxylic acid (( )-
26)
A solution of ( )-25 (9.87 g, 33.07 mmol) in CH2Cl2 (100 mL)
was treated with pyridinium chlorochromate (14.55 g,
66.13 mmol) and the reaction was stirred at room temperature
for two days. The reaction slurry was filtered through a pad of
diatomaceous earth, and the filter cake was washed multiple times
with CH2Cl2. The solvent was removed from the filtrate to afford
crude product that was purified on silica gel with a 0–20%
EtOAc in hexanes gradient to afford (3aRS,7aSR)-7a-(((tert-
butyldimethylsilyl)oxy)methyl)-1,4,7,7a-tetrahydroisobenzofu-
ran-3a(3H)-carbaldehyde (7.66 g, 78% yield). 1H NMR (400 MHz,
CDCl3): d 9.70 (s, 1H), 5.71 (t, J = 1.8 Hz, 2H), 4.17 (d, J = 8.9 Hz,
1H), 3.86 (d, J = 8.4 Hz, 1H), 3.68–3.61 (m, 3H), 3.46 (d,
J = 10.3 Hz, 1H), 2.45–2.39 (m, 1H), 2.30–2.23 (m, 1H), 2.07–2.04
(m, 1H), 2.03–1.99 (m, 1H), 0.83 (s, 9H), À0.01 (s, 6H).
reaction was stirred at room temperature for
4
h.
Dichloromethane and water were added and the two-phased mix-
ture was filtered through a pad of diatomaceous earth to remove
insoluble materials. From the filtrate, the organic layer was sepa-
rated, and the aqueous layer was extracted with CH2Cl2. The com-
bined organic extracts were dried over Na2SO4, and the solvent was
removed to afford crude product that was purified on silica gel
with a 0–50% EtOAc in hexanes gradient to afford a 1:1 mixture
of regioisomeric ketones that was carried on as is (2.20 g, 88%
yield). 1H NMR (400 MHz, CDCl3): d 7.31–7.25 (m, 5H), 5.04–4.98
(m, 2H), 4.39–4.36 (m, 0.5H), 4.23–4.22 (m, 0.5H), 3.99–3.83 (m,
2H), 3.77 (d, J = 9.5 Hz, 0.5H), 3.60 (d, J = 9.4 Hz, 0.5H), 3.54–3.49
Please cite this article in press as: Winneroski, L. L.; et al. Bioorg. Med. Chem. (2015), http://dx.doi.org/10.1016/j.bmc.2015.04.062

8
L. L. Winneroski et al. / Bioorg. Med. Chem. xxx (2015) xxx–xxx
(m, 1H), 3.38–3.35 (m, 0.5H), 3.29 (d, J = 10.6 Hz, 0.5H), 3.02–3.01
(m, 0.5H), 2.80–2.76 (m, 0.5H), 2.56–2.50 (m, 4H), 2.10–2.01 (m,
0.5H), 1.85–1.83 (m, 0.5H), 0.83 (s, 9H), 0.05 (s, 3H), 0.03 (s, 3H).
LCMS (ESI+) (m/z) 434 [M+H]+.
for 1 h. Ethyl acetate and water were added. The solution was then
filtered through diatomaceous earth, and the filter cake was
washed with EtOAc and water. The filtrate was collected, and the
organic layer was separated and aqueous layer was extracted with
EtOAc (3Â). The organic extracts were combined, dried over
Na2SO4, filtered, and the solvent was removed to afford crude
product that was purified on silica gel with a 25–100% EtOAc in
hexanes gradient to afford N-((4aRS,8aSR)-6-((1,1-dimethylethyl)
sulfonamido)-5,6,7,8-tetrahydro-4H-4a,8a-(methanooxymethano)
The product from the previous step (2.198 g, 5.07 mmol) was
dissolved in THF (10 mL) and treated with a 1 M solution of tetra-
butylammonium fluoride in THF (10.14 mL, 10.14 mmol), and the
reaction was stirred under nitrogen for 4 h. Dichloromethane and
water were added, and the organic layer was separated, and the
aqueous layer was extracted with more CH2Cl2. The combined
organic extracts were dried over Na2SO4, and the solvent was
removed to afford crude product that was purified on silica gel
with a 25–100% EtOAc in hexanes gradient to afford ( )-27 as a sin-
gle isomer (813 mg, 50% yield). 1H NMR (400 MHz, CDCl3): d 7.36–
7.34 (m, 5H), 6.85–6.81 (m, 1H), 5.06 (s, 2H), 4.44–4.43 (m, 1H),
3.99 (dd, J = 3.1, 10.5 Hz, 2H), 3.63 (d, J = 9.3 Hz, 1H), 3.47–3.40
(m, 2H), 2.62 (d, J = 15.3 Hz, 1H), 2.57–2.55 (m, 5H), 2.13–2.08
(m, 1H). LCMS (ESI+) (m/z) 320 [M+H]+.
benzo[d][1,3]thiazin-2-yl)benzamide as
a ca. 3:2 mixture of
C(6) epimers (346 mg, 52% yield). 1H NMR (400 MHz, CDCl3): d
8.14–8.08 (m, 2H), 7.51–7.50 (m, 1H), 7.43–7.38 (m, 2H), 4.24–
4.21 (m, 1H), 4.18–4.14 (m, 0.4H), 3.91–3.84 (m, 2H), 3.75–3.72
(m, 0.6H), 3.57–3.55 (m, 1.6H), 3.29–3.24 (m, 0.4H), 3.11–3.07
(m, 1H), 2.87–2.83 (m, 0.4H), 2.51 (dd, J = 13.6, 3.6 Hz, 0.6H),
2.05–2.00 (m, 6H), 1.22–1.21 (m, 3H), 1.19 (s, 3H), 1.18 (s, 3H).
LCMS (ESI+) (m/z) 436 [M+H]+.
A solution of the product from the previous step (346 mg, 794
lmol) in diethyl ether (5 mL) and CH2Cl2 (ꢀ1 mL) was treated with
4.1.11. N-(((3aRS,7aRS)-7a-(Hydroxymethyl)-6-oxohexahydro-
isobenzofuran-3a(3H)-yl)carbamothioyl)benzamide (( )-28)
A Parr bottle was charged with ( )-27 (1.064 g, 3.33 mmol), 10%
Pd/C (0.5520 g), EtOAc (80 mL), and EtOH (80 mL) and was purged
with nitrogen (4Â), purged with hydrogen (4Â) and then pressur-
ized with hydrogen (60 psi g) and was shaken at room temperature
for 10 min. The mixture was vented, and the catalyst was removed
by filtration, using EtOH to rinse. The filtrate was treated with ben-
zoyl isothiocyanate (1.09 g, 6.66 mmol), and the reaction was stir-
red at room temperature for 2 h. The solvent was removed, and the
crude product was purified over silica gel with a 0–50% EtOAc in
hexanes gradient to afford the titled product (885 mg, 76% yield).
1H NMR (400 MHz, CDCl3): d 11.49–11.45 (m, 1H), 8.92–8.91 (m,
1H), 7.82–7.80 (m, 2H), 7.65–7.61 (m, 1H), 7.53–7.50 (m, 2H),
4.79 (d, J = 10.1 Hz, 1H), 4.28 (d, J = 10.1 Hz, 1H), 4.14–4.07 (m,
1H), 3.68–3.61 (m, 2H), 3.48 (d, J = 9.3 Hz, 1H), 3.24–3.18 (m,
1H), 2.70 (d, J = 15.6 Hz, 1H), 2.56 (d, J = 15.6 Hz, 1H), 2.49–2.41
(m, 4H). LCMS (ESI+) (m/z) 349 [M+H]+.
a 1 N Et2O solution of HCl (4 mL, 4 mmol). The reaction was stirred
at room temperature for 3 h. The solvent was removed, and the
crude product was purified sequentially with a SCX column using
MeOH then (7 N NH3 in MeOH) to flush the product off the SCX col-
umn. The basic extract was concentrated, and the product was
purified again on silica with a 0–10% (7 N NH3 MeOH) in CH2Cl2
gradient to furnish ( )-30 as a 1:1 mixture of C(6) epimers
(215 mg, 82% yield). 1H NMR (400 MHz, CDCl3): d 8.14–8.07 (m,
2H), 7.49–7.44 (m, 1H), 7.41–7.36 (m, 2H), 4.29–4.25 (m, 0.5H),
4.18 (dd, J = 9.2, 6.5 Hz, 1H), 3.90–3.79 (m, 2H), 3.69 (d, J = 9.1 Hz,
0.5H), 3.53 (d, J = 13.5 Hz, 0.5H), 3.31 (d, J = 13.6 Hz, 0.5H), 3.06–
2.99 (m, 0.5H), 2.96–2.89 (m, 0.5H), 2.77–2.72 (m, 0.5H), 2.49 (d,
J = 13.6 Hz, 0.5H), 2.17–2.09 (m, 0.5H), 1.99–1.92 (m, 4.5H), 1.41–
1.20 (m, 2H), 0.88–0.79 (m, 2H). LCMS (ESI+) (m/z) 332 [M+H]+.
4.1.14. N-((4aRS,6RS,8aSR)-2-Amino-5,6,7,8-tetrahydro-4H-
4a,8a-(methanooxymethano)benzo[d][1,3]thiazin-6-yl)-3-
chlorobenzamide (( )-8)
A 0 °C mixture of ( )-30 (107 mg, 323
acid (104 mg, 646 mol), 1-hydroxybenzotriazole (65 mg, 384
mol), and N,N-diisopropylethylamine (83 mg, 113 L, 646 mol)
in THF (5 mL) was treated with 1-(3-dimethylaminopropyl)-3-
ethylcarbodiimide hydrochloride (124 mg, 646 mol), and the
lmol), 3-chlorobenzoic
4.1.12. N-((4aRS,8aSR)-6-Oxo-5,6,7,8-tetrahydro-4H-4a,8a-
(methanooxymethano)benzo[d][1,3]thiazin-2-yl)benzamide
(( )-29)
l
l
l
l
A 0 °C solution of ( )-28 (880 mg, 2.53 mmol) in CH2Cl2 (10 mL)
l
was
treated
with
1-chloro-N,N,2-trimethylpropenylamine
reaction was warmed to room temperature and stirred for 18 h.
The reaction was diluted with water and extracted multiple times
with CH2Cl2. The combined organics layers were dried (MgSO4),
filtered, and the solvent was removed to afford crude product
that was purified on silica gel with a 0–100% EtOAc in hexanes
gradient to afford N-((4aRS,6RS,8aRS)-2-benzamido-5,6,7,8-tetrahy-
dro-4H-4a,8a-(methanooxymethano)benzo[d][1,3]thiazin-6-yl)-3-
chlorobenzamide (37 mg, 24% yield). 1H NMR (400 MHz, CDCl3): d
8.10 (d, J = 7.2 Hz, 2H), 7.70 (t, J = 1.9 Hz, 1H), 7.57 (ddd, J = 7.7, 1.6,
1.2 Hz, 1H), 7.49–7.45 (m, 5H), 6.09–6.04 (m, 1H), 4.34 (d,
J = 9.3 Hz, 1H), 4.31–4.30 (m, 1H), 3.91–3.84 (m, 2H), 3.75
(d, J = 9.3 Hz, 1H), 3.56 (d, J = 13.7 Hz, 1H), 2.48 (d, J = 13.6 Hz,
1H), 2.13–2.11 (m, 1H), 2.02–1.97 (m, 4H), 1.55–1.54 (m, 1H).
LCMS (ESI+) (m/z) (35Cl/37Cl) 470/472 [M+H]+.
(506 mg, 3.79 mmol) and then the reaction was warmed to
room temperature and stirred under nitrogen for 18 h.
Dichloromethane and water were added, and the organic layer
was separated and the aqueous layer was extracted with CH2Cl2.
The combined organic extracts dried over Na2SO4, and the solvent
was removed to afford crude product that was purified over silica
gel with a 0–100% EtOAc in hexanes gradient to afford the titled
product (758 mg, 91% yield). 1H NMR (400 MHz, CDCl3): d 8.12–
8.06 (m, 2H), 7.51 (t, J = 7.3 Hz, 1H), 7.42 (t, J = 7.5 Hz, 2H), 4.03–
3.97 (m, 2H), 3.90 (t, J = 9.0 Hz, 2H), 3.15 (d, J = 13.6 Hz, 1H),
2.91–2.87 (m, 1H), 2.83–2.77 (m, 2H), 2.70–2.66 (m, 1H), 2.56–
2.50 (m, 2H), 2.33–2.22 (m, 2H). LCMS (ESI+) (m/z) 331 [M+H]+.
4.1.13. N-((4aRS,8aSR)-6-Amino-5,6,7,8-tetrahydro-4H-4a,8a-
(methanooxymethano)benzo[d][1,3]thiazin-2-yl)benzamide
(( )-30)
A mixture of the product from the previous step (37 mg, 78.7
l
mol), pyridine (63 mg, 64
lL, 787
l
mol), and O-methylhydroxy-
mol) in ethanol (3 mL) was
lamine hydrochloride (67 mg, 787
l
A solution of ( )-29 (500 mg, 1.51 mmol) was dissolved in THF
(5 mL) and then treated sequentially with ( )-2-methyl-2-
propanesulfinamide (220 mg, 1.82 mmol) and Ti(OEt)4 (863 mg,
3.78 mmol), and the reaction was heated to 65 °C for 5 h. The reac-
tion was cooled to room temperature and then to 0 °C. Sodium
borohydride (115 mg, 3.03 mmol) was added, followed by MeOH
(4 mL). The reaction was warmed to room temperature and stirred
heated to 50 °C for 17 h. The reaction was cooled to room temper-
ature, and the solvent was removed to afford crude product that
was purified over silica gel with a 0–10% (7 N NH3 methanol) in
CH2Cl2 gradient to afford ( )-8 (27 mg, 94% yield). 1H NMR
(400 MHz, DMSO-d6): d 8.46 (d, J = 7.9 Hz, 1H), 7.91 (t, J = 1.9 Hz,
1H), 7.81 (dt, J = 7.8, 1.3 Hz, 1H), 7.57 (ddd, J = 8.0, 2.1, 1.1 Hz,
1H), 7.47 (t, J = 7.8 Hz, 1H), 5.89–5.85 (br s, 2H), 4.15 (d,
Please cite this article in press as: Winneroski, L. L.; et al. Bioorg. Med. Chem. (2015), http://dx.doi.org/10.1016/j.bmc.2015.04.062

L. L. Winneroski et al. / Bioorg. Med. Chem. xxx (2015) xxx–xxx
9
J = 8.8 Hz), 4.01 (m, 1H), 3.59 (d, J = 8.8 Hz, 1H), 3.57 (d, J = 7.7 Hz,
References and notes
1H), 3.45 (d, J = 7.7 Hz, 1H), 3.26 (d, J = 13.2 Hz, 1H), 2.54 (d,
J = 13.2 Hz, 1H), 2.00 (t, J = 12.9 Hz, 1H), 1.81 (dt, J = 13.5, 2.5 Hz,
1H), 1.74 (dt, J = 13.5, 3.3 Hz, 1H), 1.60 (m, 2H), 1.31 (dq, J = 12.4,
3.2 Hz, 1H). 13C NMR (100 MHz, DMSO-d6) d 163.8, 147.2, 136.5,
133.0, 130.9, 130.2, 127.0, 126.1, 78.9, 75.4, 60.5, 44.7, 35.3, 34.1,
33.2, 31.4, 27.7. HRMS (ESI+) (m/z) calcd for C17H20ClN3O2S
(M+H): 365.0965, found 365.0968.
1. Koh, J.-Y.; Yang, L.-L.; Cotman, C. W. Brain Res. 1990, 533, 315.
2. Haass, C.; Selkoe, D. J. Nat. Rev. Mol. Cell Biol. 2007, 8, 101. and references
therein.
3. Kennedy, M. E.; Rajendran, L.; Wong, P. C.; Vassar, R.; Kuhn, P.-H.; Haass, C.;
Lichtenthaler, S. F. J. Neurochem. 2014, 130, 4.
4. Masters, C. L.; Selkoe, D. J. Cold Spring Harb. Perspect. Med. 2012, 2, a006262.
5. De Strooper, B.; Vassar, R.; Golde, T. Nat. Rev. Neurol. 2010, 6, 99.
6. Oehlrich, D.; Prokopcova, H.; Gijsen, H. J. M. Bioorg. Med. Chem. Lett. 2014, 24,
2033.
7. Stamford, A.; Strickland, C. Curr. Opin. Chem. Biol. 2013, 17, 320.
8. Yan, R.; Vassar, R. Lancet Neurol. 2014, 13, 319.
4.2. Bioanalytical methods
9. May, P. C.; Dean, R. A.; Lowe, S. L.; Martènyi, 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.
J. Neurosci. 2011, 31, 16507.
10. May, P. C.; Willis, B. A.; Lowe, S. L.; Dean, R. A.; Monk, S. A.; Cocke, P. J.; Audia, J.
E.; Boggs, L. N.; Borders, A. R.; Brier, R. A.; Calligaro, D. O.; Day, T. A.; Ereshefsky,
L.; Erickson, J. A.; Gevorkyan, H.; Gonzales, C. R.; James, D. E.; Jhee, S. S.;
Komjathy, S. F.; Li, L.; Lindstrom, T. D.; Mathes, B. M.; Martènyi, F.; Sheehan, S.
M.; Stout, S. L.; Timm, D. E.; Vaught, G. M.; Watson, B. M.; Winneroski, L. L.;
Yang, Z.; Mergott, D. J. J. Neurosci. 2015, 35, 1199.
Aliquots of homogenized brain or plasma samples and appro-
priate calibration standards were mixed with methanol/acetoni-
trile (1:1, v/v) containing internal standard to precipitate
proteins, then centrifuged to pellet the precipitant. A volume of
each sample’s supernatant was transferred to a new 96-well plate
and diluted 4-fold with water/methanol (1:1, v/v) then analyzed by
LC–MS/MS using a Betasil C18 (Thermo) 20 mm  2.1 mm  5
lm
11. Full conformational analysis performed with MacroModel using the OPLS2.1
force field in the gas phase. All conformers within 10 kcal/mol of the minima
were examined. Only the bound conformation was found for 7, and an
Javelin column and a gradient mobile phase system with A:
water/trifluoroacetic acid/1 M ammonium bicarbonate (1000:4:1,
v/v) and B: acetonitrile/trifluoroacetic acid/1 M ammonium bicar-
bonate (1000:4:1, v/v) delivered at 1.5 mL/min. Selected Reaction
Monitoring (SRM) transitions (positive ion mode) with for 6 (m/z
324.1 > 139.1), 7 (m/z 366.1 > 139.1), and internal standard (m/z
249.1 > 148.1) were acquired with an Applied Biosystems/MDS
Sciex API4000 tuned to achieve unit resolution (0.7 Da at 50%
FWHM) using Analyst software (version 1.4.2).
alternate ring conformation for
8 was 2.2 kcal/mol above the minima
Schrödinger Release 2013-2: MacroModel, version 10.1; Schrödinger, LLC: New
York, NY, 2013.
12. LEAN = (ÀlogIC50)/N, where N is the number of heavy (non-hydrogen) atoms
and IC50 is in M.
13. 9 was prepared as a single isomer. The absolute stereochemistry is unknown;
relative stereochemistry is as depicted.
14. See Supporting information for details.
15. See Supporting information for stereochemical analysis.
16. To estimate the transition state energy differences for the two cyclization
pathways, a reaction coordinate scan was carried out using the program Jaguar
(Jaguar, version 8.3, Schrödinger, LLC: New York, NY, 2014). Beginning with the
UDFT (m06-2Â)/3-21G⁄ geometry optimized iodonium ion intermediates, a
relaxed coordinate scan of the S–C distance from 1.8 to 3.6 Å was carried out.
The UDFT (m06-2Â) hybrid functional developed by Truhlar (Zhao, Y.; Truhlar,
D. G. Theor. Chem. Acc. 2008, 120, 215) was shown previously to work well with
iodonium ions (George, L.; Kalume, A.; Reid, S. A. Chem. Phys. Lett. 2012, 554,
86). Maxima were found at 3.0 Å for the favored attack; that is, 15a and 3.2 Å
for 15b. The energy difference for these two maxima was 5.7 kcal/mol in favor
of the cyclization leading to the desired isomer, 14.
17. Paquette, L. A.; Russell, R. K.; Burson, R. L. J. Am. Chem. Soc. 1975, 97, 6124.
18. Marchand-Brynaert, J.; Ghosez, L. J. Am. Chem. Soc. 1972, 94, 2869.
19. For a discussion of the challenges associated with BACE1/BACE2 selectivity, see
Ref. 10.
20. Johnson-Wood, K.; Lee, M.; Motter, R.; Hu, R.; Gordon, G.; Barbour, R.; Khan, K.;
Gordon, M.; Tan, H.; Games, D.; Lieberburg, I.; Schenk, D.; Seubert, P.;
McConlogue, L. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 1550.
21. Games, D.; Adams, D.; Alessandrini, R.; Barbour, R.; Berthelette, P.; Blackwell,
C.; Carr, T.; Clemens, J.; Donaldson, T.; Gillespie, F. Nature 1995, 373, 523.
22. Dovey, H. F.; John, V.; Anderson, J. P.; Chen, L. Z.; de Saint Andrieu, P.; Fang, L.
Y.; Freedman, S. B.; Folmer, B.; Goldbach, E.; Holsztynska, E. J.; Hu, K. L.;
Johnson-Wood, K. L.; Kennedy, S. L.; Kholodenko, D.; Knops, J. E.; Latimer, L. H.;
Lee, M.; Liao, Z.; Lieberburg, I. M.; Motter, R. N.; Nietz, J.; Quinn, K. P.; Sacchi, K.
L.; Seubert, P. A.; Shopp, G. M.; Thorsett, E. D.; Tung, J. S.; Wu, J.; Yang, S.; Yin, C.
T.; Schenk, D. B.; May, P. C.; Altstiel, L. D.; Bender, M. H.; Boggs, L. N.; Britton, T.
C.; Clemens, J. C.; Czilli, D. L.; Dieckman-McGinty, D. K.; Droste, J. J.; Fuson, K.
S.; Gitter, B. D.; Hyslop, P. A.; Johnstone, E. M.; Li, W. Y.; Little, S. P.; Mabry, T. E.;
Miller, F. D.; Audia, J. E. J. Neurochem. 2001, 76, 173.
Acknowledgments
Use of the Advanced Photon Source, an Office of Science User
Facility operated for the U.S. Department of Energy (DOE) Office
of Science by Argonne National Laboratory, was supported by the
U.S. DOE under Contract No. DE-AC02-06CH11357. Use of the
Lilly Research Laboratories Collaborative Access Team (LRL-CAT)
beamline at Sector 31 of the Advanced Photon Source was provided
by Eli Lilly Company, which operates the facility. The authors thank
Steven J. Green for assistance with the preparation of compounds 6
and 31, Christopher T. Reutter for obtaining optical rotation and
HRMS data; Jacyln A. Barrett for obtaining measured pKa data;
Eric P. Seest and Thomas J. Perun for chiral analysis and separation;
Linglin Li for preparation of PDAPP primary neuronal cultures; and
Zhixiang Yang and Theresa Day for technical assistance in conduct
and analysis of the in vivo study.
Supplementary data
Supplementary data (experimental procedures for 4ÁHCl, 5ÁHCl,
6ÁHCl, 10/12, 16, and 31ÁTFA, analytical data, and in vitro selectivity
and physicochemical data for compounds 4, 5–8, and 31) associ-
ated with this article can be found, in the online version, at
http://dx.doi.org/10.1016/j.bmc.2015.04.062.
23. Desai, P.; Sawada, G.; Watson, I.; Raub, T. Mol. Pharm. 2013, 10, 1249.
24. Tsukamoto, T. ACS Med. Chem. Lett. 2013, 4, 369.
Please cite this article in press as: Winneroski, L. L.; et al. Bioorg. Med. Chem. (2015), http://dx.doi.org/10.1016/j.bmc.2015.04.062