Discovery and Early Clinical Development of LY3202626, a Low-Dose, CNS-Penetrant BACE Inhibitor

Article abstract of DOI:10.1021/acs.jmedchem.1c00489

The beta-site APP cleaving enzyme 1, known as BACE1, has been a widely pursued Alzheimer's disease drug target owing to its critical role in the production of amyloid-beta. We have previously reported the clinical development of LY2811376 and LY2886721. L

Full text of DOI:10.1021/acs.jmedchem.1c00489

pubs.acs.org/jmc
Drug Annotation
Discovery and Early Clinical Development of LY3202626, a Low-
Dose, CNS-Penetrant BACE Inhibitor
David L. McKinzie, Leonard L. Winneroski, Steven J. Green, Erik J. Hembre, Jon A. Erickson,
Brian A. Willis, Scott A. Monk, Christopher D. Aluise, Thomas K. Baker, Jose E. Lopez, Jörg Hendle,
James P. Beck, Richard A. Brier, Leonard N. Boggs, Anthony R. Borders, Patrick J. Cocke,
Pablo Garcia-Losada, Stephen L. Lowe, Brian M. Mathes, Patrick C. May, Warren J. Porter,
Stephanie L. Stout, David E. Timm, Brian M. Watson, Zhixiang Yang, and Dustin J. Mergott*
Cite This: J. Med. Chem. 2021, 64, 8076−8100
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: The beta-site APP cleaving enzyme 1, known as
BACE1, has been a widely pursued Alzheimer’s disease drug target
owing to its critical role in the production of amyloid-beta. We have
previously reported the clinical development of LY2811376 and
LY2886721. LY2811376 advanced to Phase I before development
was terminated due to nonclinical retinal toxicity. LY2886721
advanced to Phase II, but development was halted due to
abnormally elevated liver enzymes. Herein, we report the discovery
and clinical development of LY3202626, a highly potent, CNS-
penetrant, and low-dose BACE inhibitor, which successfully
addressed these key development challenges.
INTRODUCTION
■
It is estimated that over 50 million people worldwide are
currently living with dementia. This number is projected to
increase to more than 82 million by 2030 and triple to 152
million by 2050.1 Alzheimer’s disease is the most common
cause of dementia, and this devastating, progressive neuro-
degenerative disease, for which no disease modifying therapy is
available, ultimately leads to death. An abnormal extracellular
plaque in the brain, principally composed of amyloid-beta
(Aβ), is a pathological hallmark of the disease. Beta-site APP
cleaving enzyme 1 (BACE1) is critical to the production of Aβ,
and consequently, it has been aggressively targeted for the
development of potential Alzheimer’s disease therapies.2−7 We
have previously reported the discovery and clinical develop-
ment of LY2811376 and LY2886721 (Figure 1). LY2811376
was terminated in Phase I due to adverse effects observed in
the retinal epithelium of rats upon longer term dosing, believed
to be connected to off-target protease activity.8 LY2886721
advanced to Phase II, but development was stopped because
liver enzymes were abnormally elevated.9 In order to reduce
the risk of drug-induced liver injury, we sought to develop a
more potent BACE inhibitor capable of achieving high BACE
inhibition at lower clinical doses.10 Herein, we report the
discovery and Phase I clinical development of LY3202626, a
highly potent, CNS-penetrant, and low-dose BACE inhibitor.
Figure 1. Previously reported Lilly BACE inhibitors LY2811376 and
LY2886721.
RESULTS AND DISCUSSION
■
We recently described a series of cyclopropyl BACE inhibitors
that leveraged a cyclopropyl group as a conformational
constraint.11 Initial efforts in this series focused on molecules
Received: March 17, 2021
Published: June 3, 2021
https://doi.org/10.1021/acs.jmedchem.1c00489
© 2021 American Chemical Society
J. Med. Chem. 2021, 64, 8076−8100
8076

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
Table 1. Data for Cyclopropyl BACE Inhibitors
a
b
c
Versus BACE1 IC50. MDCK cell bidirectional transport assay. m = mouse, r = rat, d = dog, h = human.
Figure 2. X-ray crystal structure of compound 5 (green) bound to the BACE1 binding site (amino acid residues labeled in yellow, protease pocket
labels in red, hydrogen bonds in dashed yellow lines).11 PDB access code 6UWV, final model derived by molecular replacement and structure
refinement from PDB 6UVV.
such as cis-fused THF-aminothiazine 3 (Table 1). While 3
demonstrated high microsomal stability across species, we
struggled to increase potency in this series beyond the ∼1 μM
range. Furthermore, 3 was a weak Pgp substrate and possessed
low passive permeability. We subsequently discovered that
potency could be increased ∼100× by employing a cis-fused
[4,5]-pyrrolidine ring system substituted with a 2-pyrimidine
group in compounds 4 and 5 (Table 1).12,13
Compounds 4 and 5 both demonstrated significant
liabilities. Amide 4 was a Pgp substrate with an efflux ratio
of 15. Compound 5, in which the amide was replaced by an
ether group, showed lower Pgp efflux and higher passive
permeability. However, microsomal stability was poor, and a
proof-of-concept study in PDAPP mice resulted in only 37%
hippocampal Aβ reduction at a dose of 100 mg/kg. The
cyclopropyl side chain also carried additional synthetic
complexity that rendered the SAR quite challenging. Thus,
8077
https://doi.org/10.1021/acs.jmedchem.1c00489
J. Med. Chem. 2021, 64, 8076−8100

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
a
Scheme 1. Synthesis of Compound 6
a
Reagents and conditions: (a) Bu4N−Br3, MeOH, CH2Cl2; (b) diallylamine, K2CO3, ACN, rt; (c) NH2OH−HCl, NaOAc, EtOH, 70 °C; (d)
xylenes, 150 °C, 35% yield (4 steps); (e) Zn, AcOH, 40 °C; (f) PhCONSC, bis(trimethylsilyl)trifluoroacetamide, THF, rt, 97% (2 steps); (g) 1-
chloro-N,N,2-trimethylpropenylamine, CH2Cl2, 0 °C for 2 h then warmed to rt, 39%; (h) chiral HPLC to separate mix of cis-fused enantiomers; (i)
N,N-dimethylbarbituric acid, Pd(Ph3)4, CHCl3, rt; (j) 2-chloropyrimidine, DIEA, 1,4-dioxane, 110 °C, 69% (2 steps); (k) 10% Pd/C, ammonium
formate, EtOH, 40 °C, 87%; (l) methoxylamine hydrochloride, pyridine, EtOH, 55 °C, 40%; (m) 1 M HCl in ether, rt, 100%.
Table 2. Key Data for Compounds 1, 2, 6, 15, 16, 17 and 18
1
2
6
15
16
17
18
MW
log P
PSA
pKa
BACE1 IC50
LEAN
cell IC50
cathepsin D IC50; ratio
Pgp efflux ratio
permeability (10−6 cm/s)
320
2.9
85
8.8
239 nM
0.29
126 nM
15 μM; 63×
2.2
390
2.1
115
7.7
317
1.9
121
7.6
121 nM
0.33
275 nM
40.9 μM; 338×
1.3
59
335
2.4
121
7.5
83.1 nM
0.32
324 nM
19 μM; 229×
1.1
331
0.8
118
7.1
470 nM
0.28
1700 nM
83 μM; 177×
1.3
360
2.3
145
6.6
110 nM
0.29
1210 nM
14 μM; 127×
ND
336
1.3
80
6.4
172 nM
0.31
2460 nM
29 μM; 169×
1.1
a
20.3 nM
0.29
10.7 nM
>300 μM; >15,000×
4.0
ab
,
a
c
60
55
60
54
ND
59
LysoTracker Rmax
3364
7863
NT
4090
59
91
20
a
b
c
Data are the means of at least 2 independent experiments. Cultured primary cortical neuronal cells.8 MDCK cell bidirectional transport.
Figure 3. X-ray crystal structure of compound 6 (green) bound to the BACE1 binding site (amino acid residues labeled in yellow, protease pocket
labels in red, hydrogen bonds in dashed yellow lines). PDB access code 7MYI, final model derived by molecular replacement and structure
refinement from PDB 6UVV.
8078
https://doi.org/10.1021/acs.jmedchem.1c00489
J. Med. Chem. 2021, 64, 8076−8100

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
a
Scheme 2. Synthesis of Compounds 15−18
a
Reagents and conditions: (a) K2CO3, NaI, Et3N, DMF; (b) (Boc)2O, Et3N, EtOH; (c) NaOH, MeOH (d) Piv-Cl, HN(Me)OMe-HCl, Et3N,
THF; (e) ArLi, THF, −40 °C; (f) Ti(OEt)4, DIPEA, toluene, reflux; (g) TFA; CH2Cl2, rt; (h) 2-chloro-4-fluoropyrimidine, DIEA, NMP, 100 °C;
(i) TFA, 60 °C; (j) Zn, AcOH; (k) PhCONSC, THF then CDI rt to reflux; (l) LiOH, MeOH; (m) HCl, CH2Cl2/Et2O; (n) BF3-Et2O, ArMgBr,
CH2Cl2/THF, −50 °C; (o) Zn, AcOH; (p) PhCONSC, THF; (q) 1-chloro-N,N,2-trimethylpropenylamine, CH2Cl2, 0 °C to rt; (r) TMS-I, ACN;
(s) 2-chloro-4-fluoropyrimidine, DIEA, NMP, 100 °C; (t) LiOH, MeOH; (u) HCl, CH2Cl2/Et2O.
we sought to move away from this series due to the
aforementioned liabilities.
these compounds induced accumulation of autofluorescent
granules in the retinal pigment epithelium of rats that was
associated with nonclinical retinal photoreceptor degenera-
tion.8,15 Subsequent research has provided compelling
evidence that this ocular toxicity effect was not attributable
to BACE1 inhibition, including a study in which LY2811376
induced such effects in BACE1 knockout mice.8 Additionally, a
comprehensive chemoproteomics study identified cathepsin D
(CatD) as the primary enzyme mediating the ocular toxicity.16
CatD is a lysosomal enzyme involved in phagocytosis, thereby
facilitating clearance of shed retinal material. Importantly,
CatD deficiency in animals and humans is associated with
retinal atrophy and visual deficits.17−19 In another study by
Ellis and colleagues, the conformational dynamics of
LY2811376 binding to BACE1 and CatD were evaluated
across a pH range reflective of endosomal and lysosomal
environments.20 They found that within the pH range that
these enzymes were active, LY2811376 interacted more
robustly with CatD versus BACE1, suggesting that the pH of
the cellular environment may influence the BACE1/CatD
selectivity of BACE1 inhibitor compounds.
Prior to the aforementioned chemoproteomics study, we
hypothesized that increasing CatD selectivity would reduce the
risk of retinal toxicity. This hypothesis was initially supported
by the improved nonclinical safety profile of LY2886721,
which helped enable its progression to a Phase II study. We
also reasoned that reducing the pKa of the aminothiazine
moiety would reduce risk of this off-target retinal epithelial
toxicity by decreasing lysosomal accumulation of the
corresponding BACE inhibitors. We used a LysoTracker
assay as a surrogate for accumulation of inhibitors in
lysosomes, and these data can be found in Table 2.21 Smaller
fluorescence response max (Rmax) values are indicative of lower
accumulation, whereas larger values suggest extensive
lysosomal accumulation. The comparison of Rmax values for
LY2811376 and LY2886721 indicated that both molecules
accumulate in lysosomes, which in turn suggested that this
accumulation must be counterbalanced by high CatD
selectivity in the case of LY2886721, thereby reducing the
The X-ray crystal structure of compound 5 is shown in
Figure 2. The binding mode highlights the cyclopropyl group
in the S1 pocket with the ether chain projecting the methyl
cyclopropyl moiety into the S3 pocket, while the pyrrolidine
ring system places the pyrimidine ring in the S2′ pocket. As
part of our SAR exploration, we evaluated the impact of
truncating the cyclopropyl side chain, thereby eliminating the
S3-binding group. We prepared compound 6, which contains
an S1-binding thiophene group in place of the cyclopropyl side
chain, to explore the impact of this structural change (Scheme
1). α-Bromination of acetylthiophene 7, followed by displace-
ment of the bromine by diallyl amine, afforded α-amino ketone
8. Treatment of 8 with hydroxylamine hydrochloride afforded
the oxime, which when heated in xylenes at 150 °C underwent
an intramolecular [3 + 2]-cycloaddition to afford the 5,5-cis-
fused pyrrolidine 9 as a mixture of diastereomers. N−O bond
cleavage using Zn powder, followed by treatment with benzoyl
isothiocyanate, afforded hydroxythiourea 10. Activation of the
alcohol using Ghosez’s reagent promoted ring closure to afford
the thiazine as a mixture of diastereomers which were
separated by chiral HPLC to afford the cyclized product 11.
Next, deprotection of the N-allyl pyrrolidine followed by
reaction with 2-chloropyrimidine afforded compound 12.
Debromination, followed by hydrolysis of the benzoyl group,
afforded compound 6 after treatment with HCl.
Interestingly, 6 demonstrated BACE1 IC50 = 121 nM along
with an exceptionally high LEAN14 value of 0.33, rendering it
more potent than LY2811376 and more LEAN efficient than
both LY2811376 and LY2886721 (Table 2). The X-ray crystal
structure of compound 6 (Figure 3) shows the thiophene in
the S1 pocket and the pyrimidine in the S2′ pocket. Thiophene
6 possessed a low Pgp efflux ratio (1.3) and high passive
permeability; however, cathepsin D selectivity was significantly
decreased compared to that of LY2886721.
Previously, development of early BACE1 inhibitors, such as
LY2811376 and AMG-8718, was halted due to findings of
ocular toxicity in dedicated animal toxicology studies. Both of
8079
https://doi.org/10.1021/acs.jmedchem.1c00489
J. Med. Chem. 2021, 64, 8076−8100

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
Figure 4. X-ray crystal structure of compound 18 (green) bound to the BACE1 binding site. (amino acid residues labeled in yellow, protease
pocket labels in red, hydrogen bonds in dashed yellow lines). PDB access code 7MYR, final model derived by molecular replacement and structure
refinement from PDB 6UVV.
Table 3. Key Data for LY2886721 (2) and Isothiazole 18
risk of retinal epithelial toxicity.16 We sought to achieve a
similar risk reduction by employing a less basic aminothiazine
group.
afforded 13. Addition of the desired aryl lithium afforded the
corresponding ketone, which when treated with paramethox-
ybenzyl hydroxylamine in the presence of Ti(OEt)4 and DIEA,
underwent a nitrone cycloaddition to afford the fused 5,5 ring
system. Following removal of the paramethoxybenzyl group, an
analogous sequence, as shown in Scheme 1, was used to
convert the tetrahydroisoxazole to the desired aminothiazine.
We also explored route 2, which moved the diversification of
the aryl S1 SAR closer to the end of the route compared to
route 1. Route 2 started with isooxazoline 14, which was
prepared from the cyclization of the known benzyl N-[2-
(hydroxyimino)ethyl]-N-2-propen-1-ylcarbamate. Addition of
We evaluated whether the pKa of the basic aminothiazine
group could be reduced by employing more electron-
withdrawing S1-binding heterocycles.22−24 Compounds 15−
18 were prepared via the routes shown in Scheme 2 to explore
the impact of pKa on lysosomal accumulation. We initially
explored route 1 for the synthesis of these compounds. In
route 1, 2-bromoethylacetate was alkylated with allylamine
followed by BOC protection. Hydrolysis of the ester and
conversion to the Weinreb amide via the mixed anhydride
8080
https://doi.org/10.1021/acs.jmedchem.1c00489
J. Med. Chem. 2021, 64, 8076−8100

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
a
Scheme 3. Synthesis of Compounds 20−22
a
Reagents and conditions: (a) Ar−COCl, ACN/EtOH, 50 °C.
aryl Grignard reagents to 14 in the presence of BF3-etherate
afforded the tetrahydroisoxazole, which then underwent ring
opening to the amino alcohol followed by conversion to the
thiourea. Ring closure and CBZ deprotection afforded the
fused NH-pyrrolidine. Treatment of the fused NH-pyrrolidine
with 2-chloro-4-fluoropyrimidine, followed by hydrolysis of the
benzoyl group, completed the synthesis. Compounds 16 and
18 were prepared via route 1, whereas compounds 15 and 17
were prepared by route 2.
enable time-averaged CSF Aβ reduction of 70% at a steady
state. Thus, isothiazole 18 did not meet our primary goal for
this effort, which was to deliver a BACE inhibitor with a
significantly lower dose than LY2886721.
As part of our chemistry efforts, we also sought to explore
amides such as 20−22 which, in addition to engaging the S1
and S2′ pockets, could also bind within the S3 pocket and
potentially increase BACE1 potency relative to LY2886721.
We also wanted to explore whether this S3 pocket binding
could increase selectivity against cathepsin D for amides 20−
22 compared to that with compound 18.27 However, we
recognized the high molecular weight and polar surface area of
these compounds would likely decrease brain penetration.
Nevertheless, we elected to explore compounds 20−22 and
make data-driven decisions regarding their strengths and
weaknesses. Compounds 20−22 were prepared in one step
from diamine 19 and the corresponding acid chloride as
described in Scheme 3.28
Amides 20−22 displayed potency at least 25-fold greater
than that of LY2886721 with high LEAN values, indicating
that the addition of heavy atoms efficiently translated to higher
BACE1 potency (Table 4). The X-ray structure of compound
22 (Figure 5) showed that the methoxypyrazine group filled
the S3 pocket of the BACE enzyme. Compounds 20−22 were
not selective over the BACE2 enzyme but were highly selective
over CatD, with ratios ranging from 105,000 to 220,000.
Compound 20 displayed similar potency between the BACE1
enzyme and the PDAPP neuronal cell culture assay, whereas
compounds 21−22 showed a slight increase in cellular
potency. We deprioritized compound 20 as it had the lowest
selectivity against the hERG channel (299-fold) compared to
21 and 22. We initially favored difluoropyridine carboxamide
21 due to its modest CatD potency (IC50 = 169 μM) and
exceptional selectivity (220,000-fold) over CatD. However,
compound 21 was also a weak Pgp substrate and demonstrated
low CNS penetration in mice and dogs (mouse brain kpu,u
and dog CSF/free plasma values of 0.05 and 0.2, respectively).
Upon further study of methoxypyrazine carboxamide 22, we
were surprised to find that 22 was freely CNS-penetrant in
both mouse and dog models, with mouse brain kpu,u and dog
CSF/free plasma values of 1.0 and 0.9, respectively. Given the
high MW and polar surface area of compound 22, it is not
clear why 22 displays free CNS penetration. The MW and
polar surface area are well outside the range that is often cited
for CNS-penetrant molecules.31 No single variable, such as
intramolecular hydrogen bonding or lipophilicity, is sufficient
to explain this observation.
Compared to S1-thiophene 15, introduction of polar atoms
or substituents in the S1-binding group resulted in loss of
potency but selectivity similar to that of cathepsin D (Table 2).
These structural changes did lower the log P and pKa of
compounds 16−18. Interestingly, similar to observations
involving the association of pKa and log P with phospholipi-
dosis, we discovered that both pKa and log P appeared to play
an important role in lysosomal accumulation as judged by the
LysoTracker Rmax values (Table 2).25 Compound 16, which
showed a modest pKa reduction (−0.4) but a significant log P
reduction (−1.6) relative to that of compound 15,
demonstrated a substantial decrease in LysoTracker Rmax
.
Compound 18, in which the S1-isothiazole group imparted
significant reduction of both pKa and log P relative to that of
15, demonstrated the lowest Rmax value. An X-ray crystal
structure (Figure 4) of compound 18 shows the thiazole group
bound in the S1 pocket of the BACE1 enzyme.
Beyond its low LysoTracker Rmax value, isothiazole 18
possessed many other favorable properties, including low Pgp
efflux, high passive permeability, and free equilibration across
the plasma/CNS barrier (Table 2). Compound 18 also
possessed suitable PK properties in rats and dogs (Table 3).
However, isothiazole 18 was about 14-fold less potent in
cultured primary cortical neuronal cells (IC50 = 2460 nM) than
in the BACE1 biochemical assay (IC50 = 172 nM) (Tables 2
and 3). As low pKa can reduce cellular potency by limiting
compound accumulation within the endosomal environment
to interact with BACE1,26 it is notable that 18 had the lowest
pKa (6.4) of the compounds in Table 2. We sought to
investigate this apparent enzyme−cell potency disconnect
using a cannulated dog CSF model to measure reduction of
CSF Aβ. After a single oral dose of 3 mg/kg of compound 18,
CSF Aβ 1−x levels were reduced ∼40−45% relative to
baseline at 6 and 9 h. By 24 h, CSF Aβ 1−x levels were back to
near baseline values. Data from this study indicated that the in
vivo potency in dogs (IC50 = 276 nM) was closer to the
BACE1 biochemical IC50 than the cell IC50. These data also
suggested that isothiazole 18 freely crossed the blood−brain
barrier, consistent with the absence of Pgp efflux and high
passive permeability (Table 2). Unfortunately, a human dose
projection indicated a QD dose >300 mg would be required to
The aforementioned CNS/free plasma ratio for compound
22 was obtained as part of a central pharmacology study
conducted in intrathecally cannulated dogs. CSF and blood
8081
https://doi.org/10.1021/acs.jmedchem.1c00489
J. Med. Chem. 2021, 64, 8076−8100

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
Table 4. Key Data for Compounds 20, 21, and LY3202626 (22)
8082
https://doi.org/10.1021/acs.jmedchem.1c00489
J. Med. Chem. 2021, 64, 8076−8100

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
Table 4. continued
a
b
c
Data are the means of at least 2 independent experiments. Cultured primary cortical neuronal cells.8 Percent turnover in hepatic microsomes in
d
30 min; h = human; r = rat; m = mouse; d = dog. MDCK cell bidirectional transport.
Figure 5. X-ray crystal structure of compound 22 (green) bound to the BACE1 binding site. (amino acid residues labeled in yellow, protease
pocket labels in red, hydrogen bonds in dashed yellow lines). PDB access code 7MYU, final model derived by molecular replacement and structure
refinement from PDB 6UVV.
were collected at baseline and select time points, after orally
dosing 1.5 mg/kg compound 22, to assess compound exposure
as well as pharmacodynamic responses. Robust decreases in
plasma and CSF levels of Aβ 1−x (approximately an 80%
reduction at nadir) were observed (Figure 6). Parallel changes
of similar magnitude were observed in CSF Aβ 1−40 and Aβ
1−42 (data not shown). The plasma free concentrations of
compound 22 exceeded the BACE1 IC50 by ∼30 to 1.7-fold
from 0.5 to 12 h postdose, respectively. As indicated above,
CSF concentrations of 22 were similar to free compound in
plasma, indicating good permeability across the blood−brain
barrier. The dog in vivo IC50 was 0.097 nM, which was about
3-fold more potent than the cellular BACE1 IC50. This shift
was observed for compounds 20 and 21 but not observed with
LY2886721. The reason for this difference is not clear.
Data from this dog study were used to develop a PK/PD
model describing plasma and CSF PK as well as CSF PD.
Changes in CSF Aβ were described using an indirect response
model driven by drug concentration in the CSF, and this
model was used to estimate an in vivo IC50 value. Human
clearance was predicted with simple allometry, utilizing
clearance from rats and dogs (Table 4). It was assumed that
the CSF/free plasma ratio and in vivo IC50 values in dogs
would reflect the CNS penetration and potency in humans.
8083
https://doi.org/10.1021/acs.jmedchem.1c00489
J. Med. Chem. 2021, 64, 8076−8100

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
Phase I single and multiple ascending dose study to explore
how its nonclinical properties would translate to clinical
performance.
The pharmacokinetics and pharmacodynamics of
LY3202626 were evaluated in
a clinical trial
(NCT02323334).32 The clinical trial was a randomized,
placebo-controlled, subject- and investigator-blind study
conducted at a single site following ethical review board
approval. The trial was conducted in compliance with the
Declaration of Helsinki and the International Conference on
Harmonisation/Good Clinical Practice Guidelines. The trial
was composed of a single ascending dose phase exploring
doses from 0.1 to 45 mg, a phase where single doses of 1.6, 10,
and 26 mg were followed by 36 h of serial CSF collection, and
a multiple dose phase where daily doses of 1, 6, or 26 mg were
administered for 14 days. Based upon serial CSF sampling
following single dose administration into healthy participants
(n = 5/cohort, except for the 10 mg dose group, where 2
additional subjects were included due to sample collection
issues), plasma/CSF LY3202626 concentration ratios ranged
between 0.0278 and 0.0428, similar to the measured in vitro
plasma free fraction value of 0.041. This suggests that unbound
drug concentrations were equivalent in CSF and plasma,
indicating free penetration across the blood−brain barrier.
Data from the human serial CSF cohort were used to fit a
PK/PD model. Pharmacokinetics were described using a three-
compartment model, with one of the compartments represent-
ing CSF concentrations. An indirect response model was used
to estimate the in vivo potency of LY3202626 against BACE1.
The model estimated IC50 for LY3202626 was 0.0712 ng/mL
(0.13 nM), generally consistent with the in vivo value of 0.097
nM estimated using data from the dog study. The model
projected doses to achieve 70 and 90% CSF Aβ reduction at a
Figure 6. Plasma and CSF Aβ 1−x response in beagle dog to 1.5 mg/
kg oral dose of LY3202626 (22). Value of **p < 0.01 vs baseline for
CSF 6, 12, and 24 h time points [ANOVA; F(5,28) = 9.40, p < 0.001;
Dunnett posthoc comparisons] and #p < 0.01 vs baseline for all
plasma time points [ANOVA; F(9,49) =43.93, p < 0.0001; Dunnett
posthoc comparison]. Data are means SE from 5 to 6 dogs.
These assumptions were used to project the daily doses that
would reduce CSF Aβ at a steady state (Figure 7). Compound
22 was projected to reduce CSF Aβ by 70% at a clinical daily
dose of 1.3 mg (90% CI: 0.46, 4.7 mg).
The combination of a low projected dose, low projected
steady-state exposure for 70% time-averaged Aβ reduction, and
free CNS penetration led us to prioritize LY3202626 (22) for
clinical development. After successfully completing nonclinical
safety testing, LY3202626 was advanced into a combined
Figure 7. Relationship between total daily dose and change in CSF Aβ for LY3202626 (22). Light blue region represents the 90% confidence
interval of the relationship. Vertical bars represent point estimate for doses projected to produce 50, 70, and 90% CSF Aβ reduction. Colored
horizontal bars represent the 90% confidence interval around these point estimates.
8084
https://doi.org/10.1021/acs.jmedchem.1c00489
J. Med. Chem. 2021, 64, 8076−8100

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
steady state were within the confidence interval of the values
projected from dog data, demonstrating the reliability of the
dog model to predict the degree of blood−brain barrier
penetration as well as the in vivo IC50 (Table 5).
projected human dose of >300 mg did not meet our low-dose
criteria for mitigating the risk of elevated liver enzymes
observed with LY2886721. We also felt that high CatD
selectivity was a more optimal approach to reducing retinal
toxicity risk. The other series, exemplified by compounds 20−
22, demonstrated pM potency and >100,000-fold selectivity
against CatD. Whereas compounds 20 and 21 displayed low
CNS penetration, LY3202626 displayed free CNS penetration
in mice and dogs, thus allowing its high BACE1 enzymatic and
cellular potency (IC50 = 615 and 275 pM, respectively) to
translate to a low projected human dose of 1.3 mg for steady-
state CSF levels targeting an Aβ 1−40 reduction of 70%.
Human exposure and dose projections proved to be accurate,
with Phase I data indicating that a dose of 2.4 mg would
reduce Aβ 1−40 by 70% at a steady state.32 Remarkably, a dose
of 9.2 mg was projected to reduce Aβ 1−40 by >90% at a
steady state. LY3202626 advanced to a Phase II clinical study
(NCT02791191) in patients with mild Alzheimer’s disease,
evaluating doses of 3 and 12 mg of LY3202626 administered
daily. While LY3202626 was well tolerated, with similar
numbers of treatment emergent adverse reported in the
placebo and 12 mg arms, an interim analysis of efficacy data
did not demonstrate a meaningful benefit to participants.
Based on this analysis, as well as reports of other BACE
inhibitors failing to slow disease progression (including
cognitive worsening in some studies) despite similar levels of
Aβ reduction, a decision was made to stop the Phase II
study.33
Table 5. Projected and Observed Pharmacokinetic and
Pharmacodynamic Parameters for LY3202626 (22)
human PK/PD projections
based on nonclinical data
observed
clinical value
CSF/free plasma ratio
human oral clearance
in vivo IC50
projected human dose for
70% CSF Aβ reduction
1.0
1.0
19 L/h (90% CI: 7.1, 51.2)
0.097 nM
1.3 mg (90% CI: 0.46, 4.7)
26 L/h
0.130 nM
2.4 mg
projected human dose for
90% CSF Aβ reduction
4.9 mg (90% CI: 1.5, 17)
9.2 mg
Based on the information obtained from the serial CSF
collections following single dose administration, doses of 1, 6,
and 26 mg were administered daily for 14 days in healthy
subjects (n = 9/cohort). These doses led to Aβ 1−40
reductions of 50.1, 75.7, and 93.7, and 46.7, 73.1, and 94.6% in
Aβ 1−42 at each of the three doses, respectively (Figure 8).
This effect was consistent with predictions made using the
model-estimated in vivo IC50 value. Overall, the preclinical
data enabled accurate human PK/PD projections.
Substantial progress has been made against this challenging
enzyme target since the initial cloning of the BACE1 enzyme
and disclosure of the first BACE1 enzyme crystal structure over
20 years ago.34−38 However, owing to multiple clinical trial
failures, interest in the BACE inhibitor field for the treatment
of AD has waned considerably. Recent data from antiamyloid
monoclonal antibodies have shown that targeting amyloid
remains a potentially viable therapeutic approach.39,40 It has
been hypothesized that BACE inhibitors may prove more
effective if evaluated years prior to Alzheimer’s disease onset, at
a time when Aβ levels are gradually accruing.40,41 Relatedly, all
recent clinical trials with BACE inhibitors have targeted soluble
Aβ reductions of >50%, with most clinical doses approaching
>80%.42 As BACE1 interacts with multiple proteins important
for synaptic plasticity and neuronal function, high levels of
BACE1 inhibition may affect other proteins beyond APP.43,44
All advanced clinical BACE inhibitors do not exhibit significant
selectivity of BACE1 over BACE2. While BACE2 inhibition
has been most associated with potential hypopigmentation
effects via interactions with pigment-cell-specific melanocyte
protein (PMEL), numerous other BACE2 protein substrates
have been identified.45 Development of selective BACE1
inhibitors without BACE2 activity could reduce off-target
pharmacology. Finally, it is known that individuals with the
A673T mutation in the APP gene results in less efficient
BACE1 cleavage of APP, resulting in approximately a 28%
plasma Aβ and a 40% reduction in Aβ deposition in elderly
individuals.46,47 These findings have led some researchers to
propose that future clinical trials targeting lower levels of
BACE inhibition may uncover therapeutic benefit; however, it
remains unclear if interest in BACE1 inhibitors will be
revived.48
Figure 8. Change in human CSF Aβ concentrations following daily
LY3202626 (22) Administration. Concentrations of CSF Aβ1−40 ( )
○
and Aβ1−42 (Δ) were measured at baseline and after 14 days of
LY3202626 administration. If the post-treatment Aβ concentration
was below detectable limits, a value of one-half the lower limit of
quantification was used in the calculation, and the value was plotted in
red to denote this. Nine healthy subjects were included in each
cohort.
CONCLUSIONS
■
We have described two new series of BACE inhibitors. One
series, exemplified by compound 18, achieved nM potency and
free CNS penetration. While this series appeared to have
reduced risk for retinal toxicity owing to its lower pKa and log
P values relative to those of other BACE inhibitors, its
8085
https://doi.org/10.1021/acs.jmedchem.1c00489
J. Med. Chem. 2021, 64, 8076−8100

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
concentrated in vacuo. The residue was dissolved in acetonitrile (200
mL), then potassium carbonate was added (20.22 g, 146.29 mmol),
followed by diallylamine (12.03 mL, 97.52 mmol). Stirring occurred
at rt for 2 h. Water was added to the reaction mixture and then
extracted with ethyl acetate. The combined organics were dried over
sodium sulfate and filtered and concentrated in vacuo. Crude product
was obtained (14.64 g, 100% yield). 1H NMR (400 MHz, CDCl3): δ
7.80 (d, J = 1.50 Hz, 1H), 7.46 (d, J = 1.76 Hz, 1H), 5.93−5.81 (m,
2H), 5.24−5.12 (m, 4H), 3.64 (s, 2H), 3.20 (d, J = 5.95 Hz, 4H).
ESI+ m/z: 300, 302 [M + H]+.
Table 6. Table of Crystallographic Statistics
BACE +
compound 6
PDB 7MYI
BACE +
compound 18
PDB 7MYR
BACE +
compound 22
PDB 7MYU
data collection
space group
P2(1)2(1)2(1)
P2(1)2(1)2(1)
P2(1)2(1)2(1)
cell dimensions
a, b, c (Å)
86.55, 89.49,
131.47
86.48, 89.74,
131.70
86.52, 90.54,
131.08
cell dimensions
α, β, γ (deg)
resolution (Å)
90, 90, 90
90, 90, 90
90, 90, 90
1-(4-Bromo-2-thienyl)-2-(diallylamino)ethanone oxime (8a).
20−1.25
31−1.72
(1.81−1.72)
0.133 (0.697)
36−1.94
(2.04−1.94)
0.155 (0.708)
a
a
a
(1.32−1.25)
0.101 (0.726)
8.9 (2.5)
R-merge
I/sigma(I)
9.3 (2.8)
7.8 (2.5)
completeness (%)
redundancy
98.6 (94.0)
7.2 (5.9)
99.9 (99.9)
7.4 (7.3)
99.7 (99.5)
7.2 (7.1)
refinement
A mixture of 1-(4-bromo-2-thienyl)-2-(diallylamino)ethanone (14.63
g, 48.748 mmol), sodium ethanoate (5.99 g), 73.122 mmol),
hydroxylamine hydrochloride (5.081g, 73.122 mmol), and ethanol
(195 mL) in a flask was heated to 70 °C and stirred overnight.
Reaction mixture was concentrated in vacuo and then diluted with
dichloromethane. Saturated sodium bicarbonate was added. The two
phases were separated, and the aqueous phase was extracted with
dichloromethane. The combined organics were dried over sodium
sulfate and filtered, and concentrated in vacuo. Crude product was
obtained (15.00 g, 100% yield). 1H NMR (400 MHz, CDCl3): δ 7.70
(d, J = 1.21 Hz, 0.6 H), 7.40 (d, J = 1.43 Hz, 0.4 H), 7.14 (d, J = 1.32
Hz, 0.6 H), 7.12 (d, J = 1.43 Hz, 0.4 H), 5.95−5.83 (m, 2H), 5.28 (s,
1H), 5.27−5.18 (m, 4H), 3.74 (s, 1.20 H), 3.59 (s, 0.80 H), 3.24−
318 (m, 4H).
resolution (Å)
no. of reflections
20−1.25
275060
31−1.72
109159
34−1.94
76594
R-work (%)/R-free
(%)
14.9/17.2
17.0/19.6
18.7/22.8
no. of non-hydrogen
atoms
6360/64/1188
6277/71/907
6187/75/730
protein/ligand/water
B-factors
13.9/13.2/29.5
0.007/1.47
15.6/19.1/30.0
0.006/1.25
23.3/22.5/33.3
0.005/1.18
protein/ligand/water
root-mean-squared
deviations bond
length (Å)/bond
angle (deg)
Ramachandran
distribution φ−ψ
favored (%)/φ−ψ
allowed (%)
99.0/100
98.8/100
98.6/100
( )-5-Allyl-6a-(4-bromo-2-thienyl)-3,3a,4,6-tetrahydro-1H-
pyrrolo[3,4-c]isoxazole (9).
a
Values in parentheses denote highest resolution shell. Please note
that protein coordinates and structure factors will be deposited with
the Protein Data Bank.
EXPERIMENTAL SECTION
■
A mixture of 1-(4-bromo-2-thienyl)-2-(diallylamino)ethanone oxime
(15.37 g, 48.748 mmol) and p-xylene (244 mL) was heated to 150 °C
for 4 h. Reaction mixture was cooled and the solvent removed under
reduced pressure. The crude mixture was purified over 120 g of silica
gel, with a 20 min gradient of 0 to 50% EtOAc/hexanes to afford the
General. All reagents and anhydrous solvents were used as
obtained from commercial sources without further purification. Air-
sensitive reactions were performed under a nitrogen atmosphere.
Reactions were monitored by thin layer chromatography using silica
gel 60 F254 plates from EMD and staining with ninhydrin or by
LCMS using an Agilent 1100 series HPLC on a Gemini-NX C18
110A, 50 mm × 2.00 mm column. Normal phase purifications were
performed using RediSep prepacked columns from Teledyne Isco.
Reversed-phase purifications were performed using RediSep Rf Gold
prepacked columns from Teledyne Isco. Analytical HPLC and LCMS
data were obtained using an Agilent 1100 series HPLC on a Gemini-
NX C18 110A, 50 mm × 2.00 mm column. The purity of the samples
was assessed using a UV detector at 254 nm. Unless otherwise
1
product (5.40 g, 35% yield) (over 4 steps). H NMR (400 MHz,
CDCl3): δ 7.07 (d, J = 1.4 Hz, 1H), 6.89 (d, J = 1.4 Hz, 1H), 5.88−
5.80 (m, 1H), 5.53−5.43 (m, 1H), 5.18 (dq, J = 17.1, 1.6 Hz, 1H),
5.13−5.10 (m, 1H), 4.59−4.38 (m, 1H), 3.65−3.48 (m, 1H), 3.20−
3.03 (m, 4H), 3.02−2.76 (m, 1H), 2.65−2.25 (m, 2H). LCMS
(ESI+): m/z 315, 317 [M + H]+.
( )-[1-Allyl-4-amino-4-(4-bromo-2-thienyl)pyrrolidin-3-yl]-
methanol (9a).
1
noted,all compounds were >95% pure as judged by HPLC. H NMR
spectra were obtained on a Varian Unity INOVA 400 at 400 MHz.
Synthesis Methods. Synthesis of Compound 6. 1-(4-Bromo-2-
thienyl)-2-(diallylamino)ethanone (8).
( )-5-Allyl-6a-(4-bromo-2-thienyl)-3,3a,4,6-tetrahydro-1H-pyrrolo-
[3,4-c]isoxazole (4.25 g, 13.482 mmol) was dissolved in acetic acid
(67 mL). Zinc powder (8.82 g, 134.820 mmol) was added, and the
reaction mixture was heated to 40 °C, with vigorous stirring. After 4 h,
the reaction was cooled to ambient temperature, diluted with ethanol,
filtered through Celite, and then washed with ethanol. The filtrate was
dried in vacuo and then dissolved in water. The solution was made
basic with 1 N NaOH, filtered through Celite, and then extracted with
3:1 chloroform/isopropyl alcohol. The organic phase was dried over
1-(4-Bromothiophen-2-yl)ethanone (10.00 g, 48.76 mmol) was
dissolved in a mixture of methanol (97 mL) and dichloromethane
(162 mL). Tetra-n-butylammonium tribromide (28.79 g, 58.51
mmol) was added. The reaction mixture was then concentrated in
the rotavap at 50 °C, diluted with minimal dichloromethane, loaded
to a 330 g silica gel cartridge, and eluted with dichloromethane. The
fractions containing the bromide intermediate were combined and
8086
https://doi.org/10.1021/acs.jmedchem.1c00489
J. Med. Chem. 2021, 64, 8076−8100

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
magnesium sulfate and filtered. The solvent was removed in vacuo to
give the crude product (4.00 g, 94% yield). H NMR (400 MHz,
N-[(4aR,7aR)-7a-(4-Bromo-2-thienyl)-4a,5,6,7-tetrahydro-4H-
pyrrolo[3,4-d][1,3]thiazin-2-yl]benzamide (11a).
1
CDCl3): δ 7.09 (d, J = 1.4 Hz, 1H), 6.95 (d, J = 1.3 Hz, 1H), 5.92−
5.87 (m, 1H), 5.24−5.18 (m, 1H), 5.14−5.10 (m, 1H), 3.85−3.81
(m, 1H), 3.73−3.68 (m, 1H), 3.19−3.15 (m, 2H), 2.88−2.85 (m,
1H), 2.79−2.76 (m, 2H), 2.71 (t, J = 9.3 Hz, 1H), 2.53−2.45 (m,
1H), 1.20 (d, J = 6.2 Hz, 1H). ESI+: m/z 317, 319 [M + H]+.
(
)-N-[[1-Allyl-3-(4-bromo-2-thienyl)-4-(hydroxymethyl)-
pyrrolidin-3-yl]carbamothioyl]benzamide (10).
A solution of N-[(4aR,7aR)-6-allyl-7a-(4-bromo-2-thienyl)-4,4a,5,7-
tetrahydropyrrolo[3,4-d][1,3]thiazin-2-yl]benzamide (1.05 g, 2.271
mol) in chloroform (23 mL) was degassed with a stream of nitrogen
for 10 min and then were added N,N-dimethylbarbituric acid (2.10 g,
13.44 mmol) and tetrakis(triphenylphosphine)palladium (1.05 g;
0.909 mmol). After 30 min, the reaction was complete. It was diluted
with dichloromethane and extracted with 1 N HCl.
The acidic aqueous was made basic with 5N NaOH, and extracted
with dichloromethane. The organic phase was dried over magnesium
sulfate and filtered. The solvent was removed in vacuo. The crude
product was obtained as a solid (960 mg, 100% yield). 1H NMR (400
MHz, CDCl3): δ 8.31−7.92 (m, 2H), 7.55−7.46 (m, 1H), 7.42 (t, J =
7.61 Hz, 2H), 6.95−6.91 (m, 1H), 3.44−3.23 (m, 5H), 2.94−2.84
(m, 1H), 2.84−2.75 (m, 1H). LCMS (ESI+): m/z 422, 424 [M + H]+.
N-[(4aR,7aR)-7a-(4-Bromo-2-thienyl)-6-pyrimidin-2-yl-4,4a,5,7-
tetrahydropyrrolo[3,4-d][1,3]thiazin-2-yl]benzamide (12).
To a solution of ( )-[1-allyl-4-amino-4-(4-bromo-2-thienyl)-
pyrrolidin-3-yl]methanol (4.00 g, 12.608 mmol) in tetrahydrofuran
(180 mL) was added bis(trimethylsilyl)trifluoroacetamide (3.25g,
12.608 mmol). The reaction was stirred for 1 h at ambient
temperature. Benzoyl isothiocyanate (2.04 mL, 15.130 mmol) was
added in one portion, and the reaction mixture was stirred for another
hour, quenched with 1 N HCl (20 mL), and stirred for one additional
hour. The reaction was made basic with 50 mL of saturated sodium
bicarbonate. It was then diluted with water (100 mL) and extracted
with dichloromethane. The organic phase was dried over sodium
sulfate and filtered. The solvent was removed in vacuo. The crude
product was triturated with 25 mL of 4:1 hexanes/ethyl acetate. The
precipitate was collected by filtration and dried under reduced
1
pressure. The product was obtained (5.90g, 98% yield). H NMR
A solution of N-[(4aR,7aR)-7a-(4-bromo-2-thienyl)-4a,5,6,7-tetrahy-
dro-4H-pyrrolo[3,4-d][1,3]thiazin-2-yl]benzamide (950 mg, 2.246
mmol) in 1,4-dioxane (25 mL) was treated with 2-chloro-pyrimidine
and (515 mg, 4.498 mmol) and diisopropylethylamine (1.96 mL,
11.246 mmol). The reaction was heated to 110 °C for 4 h. The
solvent was removed in vacuo. The crude product was purified over
40 g of silica gel, with a 20 min gradient of 0 to 50% EtOAc/hexanes
(400 MHz, CDCl3): δ 11.73 (s, 1H), 8.92−8.88 (m, 1H), 7.84 (d, J =
7.28 Hz, 2H), 7.64−7.60 (m, 1H), 7.51 (t, J = 7.8 Hz, 2H), 7.09 (dd,
J = 0.44, 1.43 Hz, 1H), 7.00 (dd, J = 0.44, 1.43 Hz, 1H), 5.98−5.85
(m, 1H), 5.26 (d, J = 16.98 Hz, 1H), 5.16 (d, J = 10.25 Hz, 1H),
3.97−3.94 (m, 2H), 3.65−3.57 (m, 1H), 3.53−3.41 (m, 1H), 3.28−
3.14 (m, 2H), 3.10−2.99 (m, 1H), 2.91−2.80 (m, 1H), 2.73−2.61
(m, 1H). LCMS (ESI+): m/z 480, 482 [M + H]+.
1
to afford the product (780 mg, 69% yield). H NMR (400 MHz,
N-[(4aR,7aR)-6-Allyl-7a-(4-bromo-2-thienyl)-4,4a,5,7-
tetrahydropyrrolo[3,4-d][1,3]thiazin-2-yl]benzamide (11).
CDCl3): δ 8.32 (d, J = 4.9 Hz, 2H), 8.07−7.93 (m, 1H), 7.53−7.46
(m, 1H), 7.41 (t, J = 7.61 Hz, 2H), 7.21−7.19 (m, 1H), 7.01−6.99
(m, 1H), 6.54 (t, J = 4.85 Hz, 1H), 4.23−4.16 (m, 1H), 4.07−4.02
(m, 1H), 4.01−3.95 (m, 2H), 3.36−3.26 (m, 1H), 3.06−2.98 (m,
1H), 2.98−2.91 (m, 1H). LCMS (ESI+): m/z 500, 502 [M + H]+.
N-[(4aR,7aR)-6-Pyrimidin-2-yl-7a-(2-thienyl)-4,4a,5,7-
tetrahydropyrrolo[3,4-d][1,3]thiazin-2-yl]benzamide (12a).
( )-N-[[1-Allyl-3-(4-bromo-2-thienyl)-4-(hydroxymethyl)pyrrolidin-
3-yl]carbamothioyl]benzamide (5.90g, 12.280 mmol) was dissolved
in dichloromethane (307 mL) and cooled to 0 °C. Chloro-N,N,2-
trimethylpropenylamine (3.25 mL, 24.560 mmol) was added slowly.
The mixture was stirred for 2 h and then warmed to ambient
temperature and stirred overnight. The reaction mixture was washed
with saturated sodium bicarbonate. The organic phase was dried over
magnesium sulfate and filtered. The solvent was removed in vacuo.
The crude product was purified over 120 g of silica gel, with a 20 min
gradient of 0 to 100% EtOAc/hexanes to afford the product (2.24g,
40% yield). This mixture of cis-fused enantiomers was chirally purified
using a 4.6 × 150 mm Chiralpak AD-H column and eluting with 0.2%
DMEA in EtOH at a flow rate of 0.60 mL/min. This gave the chirally
pure desired product (1.05 g, 99% ee). 1H NMR (400 MHz, CDCl3):
δ 8.19−8.06 (m, 2H), 7.52−7.47 (m, 1H), 7.42 (t, J = 7.72 Hz, 2H),
7.14−7.11 (m, 1H), 6.98−6.94 (m, 1H), 5.93−5.77 (m, 1H), 5.21 (d,
J = 17.31 Hz, 1H), 5.11 (d, J = 10.14 Hz, 1H), 3.29−3.11 (m, 4H),
3.10−3.01 (m, 1H), 3.01−2.90 (m, 3H), 2.77 (dd, J = 13.89, J = 4.08
Hz, 1H). MS (ESI+): m/z 462, 464 [M + H]+.
A slurry of N-[(4aR,7aR)-7a-(4-bromo-2-thienyl)-6-pyrimidin-2-yl-
4,4a,5,7-tetrahydropyrrolo[3,4-d][1,3]thiazin-2-yl]benzamide (775
mg, 1.549 mmol), palladium (1.648g, 1.549 mmol), formic acid,
ammonium salt, and ethanol (31 mL) was stirred in a sealed vessel at
40 °C for 1 h. To the reaction mixture were added Celite and
saturated sodium bicarbonate (1 mL).
The reaction was diluted with ethyl acetate and filtered through
Celite. The filtrate was dried over sodium sulfate, filtered and then the
solvent was removed in vacuo. The crude product was obtained as a
solid (570 mg, 87% yield).). 1H NMR (400 MHz, CDCl3): δ 8.32 (d,
J = 4.7 Hz, 2H), 8.18−7.98 (m, 1H), 7.50−7.44 (m, 1H), 7.42−7.38
(m, 2H), 7.32−7.27 (m, 1H), 7.10−7.08 (m, 1H), 7.02−7.00 (m,
1H)), 6.54 (t, J = 4.85 Hz, 1H), 4.30−4.21 (m, 1H), 4.17−4.10 (m,
1H), 4.04−3.97 (m, 2H), 3.40−3.34 (m, 1H), 3.11−3.02 (m, 1H),
2.95 (dd, J = 13.34, 4.08 Hz, 1H). LCMS (ESI+): m/z 422 [M + H]+.
8087
https://doi.org/10.1021/acs.jmedchem.1c00489
J. Med. Chem. 2021, 64, 8076−8100

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
(4aR,7aR)-6-Pyrimidin-2-yl-7a-(2-thienyl)-4,4a,5,7-
temperature exotherm to −48 °C noted. The resulting clear light
amber solution was stirred at −52 °C for 10 min. Next, thiophen-2-
ylmagnesium bromide (1 M in THF, 100 mL, 100 mmol) was added
slowly via cannula over a period of 25 min, while maintaining the
internal temperature below −45 °C. The solution was stirred at −50
°C for 10 min, and then the cooling bath was removed, and the
reaction was warmed to 0 °C over about 25 min at which point it was
placed in a 0 °C ice bath and continued stirring at 0 °C for 1 h. The
reaction was then quenched by the addition of saturated aqueous
NH4Cl (400 mL) and stirred for 10 min. The ice bath was removed,
and the mixture was stirred for 17 h. The mixtures were transferred to
a separatory funnel (combining the two side-by-side reactions at this
time) while being diluted with DCM (200 mL). The phases were
separated, and the aqueous phase was further extracted with DCM (2
× 300 mL). The combined organic extracts were dried over Na2SO4,
filtered, and concentrated to give the crude product as an amber oil.
The sample was purified via silica gel chromatography eluting with a
25−50% gradient of EtOAc in hexanes to give the desired product
(14.33g, 43.37 mmol, 87% yield). 1H NMR (400.15 MHz, CDCl3): δ
7.38−7.31 (m, 6H), 7.05−7.04 (m, 1H), 6.99 (dd, J = 3.6, 5.1 Hz,
1H), 5.13 (s, 2H), 4.27−4.20 (m, 1H), 3.96−3.94 (m, 4H), 3.71−
3.69 (m, 1H), 3.38−3.32 (m, 1H). LCMS (ESI+): m/z 331[M + H]+.
( )-Benzyl (3R,4R)-3-amino-4-(hydroxymethyl)-3-(2-thienyl)-
pyrrolidine-1-carboxylate (15b).
tetrahydropyrrolo[3,4-d][1,3]thiazin-2-amine dihydrochloride (6).
A solution of N-[(4aR,7aR)-6-pyrimidin-2-yl-7a-(2-thienyl)-4,4a,5,7-
tetrahydropyrrolo[3,4-d][1,3]thiazin-2-yl]benzamide (123 mg, 0.292
mmol) in ethanol (2 mL) was treated with pyridine (236 μL, 2.918
mmol) and o- methylhydroxylamine hydrochloride (244 mg, 2.918
mmol). The reaction was heated to 50 °C for 4 h in a sealed vessel.
The reaction was cooled to room temperature, and then the solvent
was removed in vacuo. The crude product was purified over 12 g of
silica gel, with a 12 min gradient of 0 to 5% 7 N NH3MeOH in
CH2Cl2 to afford the free base as a white solid (40 mg, 43% yield).
The 2 M HCl salt was prepared by dissolving the free base in
dichloromethane, then adding HCl (1.184 mmol, 1.18 mL, 1 M in
ether). The mixture was stirred at room temperature for 15 min. The
solvent was concentrated in vacuo and then the residue was dried in
the vacuum pump overnight. The 2 M HCl salt product was a white
1
solid (47 mg, 41% yield). H NMR (400 MHz, CD3OD): δ 8.61−
8.57 (m, 2H), 7.54 (dd, J = 1.2, 5.2 Hz, 1H), 7.22−7.20 (m, 1H), 7.12
(dd, J = 3.7, 5.1 Hz, 1H), 7.01−6.98 (m, 1H), 4.44 (d, J = 12.3 Hz,
1H), 4.30 (d, J = 12.4 Hz, 1H), 4.13−4.08 (m, 1H), 3.94−3.89 (m,
1H), 3.49−3.44 (m, 1H), 3.44−3.38 (m, 1H). HRMS (ESI+): m/z
calcd forC14H15N5S2 (M + H) 317.0769, found 317.0777.
Synthesis of Compound 15. Benzyl 3,3a,4,6-Tetrahydropyrrolo-
[3,4-c]isoxazole-5-carboxylate (14).
A rapidly stirring solution of ( )-benzyl (3aR,6aR)-6a-(2-thienyl)-
3,3a,4,6-tetrahydro-1H-pyrrolo[3,4-c]isoxazole-5-carboxylate (14.29g,
43.25 mmol) in acetic acid (144 mL) was treated with zinc powder
(28.28g, 432.50 mmol). The resulting mixture was stirred for 4.5 h,
while the reaction progress was monitored by LC/MS. Additional zinc
(14.14g, 216.25 mmol) was added, and stirring was continued for
another 22 h. The mixture was vacuum filtered through a plug of
Celite, while being rinsed with EtOAc (∼600 mL). The filtrate was
concentrated and placed under high vacuum to afford a pale-yellow
oil. This material was transferred to a separatory funnel while being
diluted with EtOAc (500 mL). Next, 1 M NaOH (aqueous, 300 mL)
was added to adjust the aqueous pH to ∼10. The phases were
separated, and the aqueous phase was further extracted with EtOAc
(200 mL). The combined organic extracts were washed with brine,
dried over Na2SO4, and filtered through a plug of Celite, while being
rinsed with EtOAc. The filtrate was concentrated to give the desired
To
a 50 mL RBF were added benzyl N-allyl-N-[2-
hydroxyiminoethyl]carbamate (24 g, 96.66 mmol) and dichloro-
methane (338 mL). Then, a solution of 5% w/w aqueous sodium
hypochlorite (143 mL) was added dropwise over 10 min. The
reaction mixture was then stirred at room temperature for 18 h. The
reaction was then quenched by adding a 40% aqueous solution of
sodium bisulfite (7.04g, 67.66 mmol). The mixture was transferred to
a separatory funnel. The organic layer was separated, dried over
magnesium sulfate, and filtered. The filtrate was evaporated to afford
the crude product as an oil. The sample was purified via silica gel plug
filtration using as eluent CH2Cl2/EtOAc (95/5). Fractions containing
pure product were combined and solvent evaporated to afford the
1
product (13.47 g, 40.53 mmol, 94% yield). H NMR (400.15 MHz,
1
desired product (18 g, 73.09 mmol, 76% yield). H NMR (300.16
CDCl3): δ 7.40−7.33 (m, 6H), 7.12−7.10 (m, 2H), 5.20−5.12 (m,
3H), 3.92−3.89 (m, 6H), 2.65−2.64 (m, 2H). LCMS (ESI+): m/z
333 [M + H]+.
( )-Benzyl (3R,4R)-3-(Benzoylcarbamothioylamino)-4-(hydroxy-
methyl)-3-(2-thienyl)pyrrolidine-1-carboxylate (15c).
MHz, CDCl3): δ 7.36 (s, 5H), 5.16 (s, 2H), 4.66−4.57 (m, 1H),
4.27−4.20 (m, 2H), 4.14−3.96 (m, 3H), 3.27−3.17 (m, 1H). LCMS
(ESI+): m/z 247 [M + H]+.
( )-Benzyl (3aR,6aR)-6a-(2-Thienyl)-3,3a,4,6-tetrahydro-1H-
pyrrolo[3,4-c]isoxazole-5-carboxylate (15a).
A stirred light amber solution of ( )-benzyl (3R,4R)-3-amino-4-
(hydroxymethyl)-3-(2-thienyl)pyrrolidine-1-carboxylate (13.47 g,
40.52 mmol) in tetrahydrofuran (135 mL) was chilled in a 0 °C ice
bath under nitrogen. Benzoyl isothiocyanate (5.74 mL, 42.55 mmol)
was added, and then the cooling bath was removed. The reaction was
stirred for 1.5 h while being monitored by LC/MS. The reaction was
then concentrated to give the crude product, which was carried on
without purification (21.59g, 43.56 mmol, 110% crude yield). 1H
NMR (400.15 MHz, CDCl3): δ 11.83−11.80 (m, 1H), 8.97−8.94 (m,
Two side-by-side reactions were run simultaneously. The reagent
amounts below represent the scale run for each of the two separately.
The amount of final product and yield is for the combined total. A
solution of benzyl 3,3a,4,6-tetrahydropyrrolo[3,4-c]isoxazole-5-car-
boxylate (6.16g, 25.01 mmol) in dichloromethane (250 mL, 3.90
mol) was cooled to −50 °C in a bath of dry ice and acetonitrile. The
reaction was treated with boron trifluoride etherate (6.33 mL, 50.03
mmol) dropwise over a period of 3 min, with a slight internal
8088
https://doi.org/10.1021/acs.jmedchem.1c00489
J. Med. Chem. 2021, 64, 8076−8100

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
1H), 7.84−7.80 (m, 2H), 7.69−7.64 (m, 1H), 7.50 (t, J = 7.6 Hz,
2H), 7.40−7.34 (m, 5H), 7.26−7.23 (m, 1H), 7.03−7.00 (m, 2H),
5.22−5.16 (m, 2H), 4.91−4.87 (m, 1H), 4.29−4.25 (m, 1H), 4.01−
3.97 (m, 2H), 3.84−3.81 (m, 3H). LCMS (ESI+): m/z 554 [M + H]+.
( )-Benzyl (4aR,7aR)-2-Benzamido-7a-(2-thienyl)-4,4a,5,7-
tetrahydropyrrolo[3,4-d][1,3]thiazine-6-carboxylate (15d).
to afford 33.3 g (89%) of product that was utilized as is in the next
step. 1H NMR (400.13 MHz, DMSO-d6): δ 8.01 (d, J = 7.3 Hz, 2H),
7.60−7.55 (m, 2H), 7.49 (t, J = 7.6 Hz, 2H), 7.19 (d, J = 3.1 Hz, 1H),
7.12−7.09 (m, 1H), 3.62−3.54 (m, 3H), 3.39−3.38 (m, 1H), 3.12−
3.06 (m, 1H), 3.02 (d, J = 3.1 Hz, 2H). LCMS (ESI+): m/z 344 [M +
H]+.
N-[(4aR,7aR)-6-(5-Fluoropyrimidin-2-yl)-7a-(2-thienyl)-4,4a,5,7-
tetrahydropyrrolo[3,4-d][1,3]thiazin-2-yl]benzamide (15g).
A solution of ( )-benzyl (3R,4R)-3-(benzoylcarbamothioylamino)-4-
(hydroxymethyl)-3-(2-thienyl)pyrrolidine-1-carboxylate (20.08 g,
40.52 mmol) in dichloromethane (202 mL) was chilled in a 0 °C
ice bath under an atmosphere of nitrogen. Then, 1-chloro-N,N,2-
trimethylpropenylamine (6.43 mL, 48.625 mmol) was added
dropwise over a period of 5 min. The resulting solution was stirred
at 0 °C for 10 min, and then the cooling bath was removed, and the
solution allowed to warm to rt with stirring for 1.5 h. The reaction was
quenched by the addition of 100 mL of saturated aqueous NaHCO3
and the mixture stirred for 10 min. The mixture was transferred to a
separatory funnel, and the layers were separated. The organic phase
was further washed with brine (1 × 100 mL), and then the organic
phase was dried over Na2SO4, filtered, and concentrated. The crude
material was then purified via silica gel chromatography using a 2−7%
EtOAc/DCM gradient to give the desired product (9.99g, 20.92
mmol, 51% yield). 1H NMR (400.15 MHz, CDCl3): δ 8.11−8.10 (m,
2H), 7.54−7.51 (m, 1H), 7.46−7.42 (m, 2H), 7.39−7.32 (m, 6H),
7.02−6.99 (m, 2H), 5.21−5.16 (m, 1H), 5.12−5.08 (m, 1H), 4.02−
3.96 (m, 2H), 3.87−3.82 (m, 2H), 3.30−3.27 (m, 1H), 2.97−2.94
(m, 2H). LCMS (ESI+): m/z 478 [M + H]+.
To a stirred solution of N-[(4aR,7aR)-7a-(2-thienyl)-4a,5,6,7-tetrahy-
dro-4H-pyrrolo[3,4-d][1,3]thiazin-2-yl]benzamide (35.9 g, 95.1
mmol) in N-methyl pyrrolidone (359 mL) at rt was added diisopro-
pylethylamine (28.2 mL, 161.7 mmol) followed by 5-fluoro-2-
chloropyrimidine (16.38 g, 11.81 mL, 123.6 mmol). The reaction
was heated to 85 °C for 6 h and monitored by LC/MS until complete.
The reaction was cooled and was poured into water (1.8 L) and solids
precipitated. The mixture was diluted with EtOAc (500 mL); the
organic layer was removed, and the aqueous layer was extracted with
EtOAc (500 mL then 300 mL). The combined organ-
ic phases were dried over Na2SO4, filtered, and concentrated to afford
64.7 g of crude product. The material was purified by silica gel
chromatography eluting with 5% EtOAc in CH2Cl2 to give the desired
product. 1H NMR (300.11 MHz, CDCl3): δ 8.23 (d, J = 0.6 Hz, 2H),
8.08 (d, J = 6.8 Hz, 2H), 7.52−7.41 (m, 3H), 7.32 (dd, J = 1.1, 5.1
Hz, 1H), 7.09 (d, J = 2.8 Hz, 1H), 7.03 (dd, J = 3.6, 5.1 Hz, 1H),
4.24−4.08 (m, 2H), 3.98 (d, J = 9.0 Hz, 2H), 3.39−3.34 (m, 1H),
3.13−3.04 (m, 1H), 2.96 (dd, J = 4.1, 13.5 Hz, 1H). LCMS (ESI+):
m/z 440 [M + H]+.
Benzyl (4aR,7aR)-2-Benzamido-7a-(2-thienyl)-4,4a,5,7-
tetrahydropyrrolo[3,4-d][1,3]thiazine-6-carboxylate (Chiral Purifi-
cation) (15e).
(4aR,7aR)-6-(5-Fluoropyrimidin-2-yl)-7a-(2-thienyl)-4,4a,5,7-
tetrahydropyrrolo[3,4-d][1,3]thiazin-2-amine (15h).
A solution of N-[(4aR,7aR)-6-(5-fluoropyrimidin-2-yl)-7a-(2-thien-
yl)-4,4a,5,7-tetrahydropyrrolo[3,4-d][1,3]thiazin-2-yl]benzamide
(24.1 g, 51.0 mmol) in methanol (482 mL) was treated with lithium
hydroxide (1.346 g, 56.2 mmol) and was heated to 76 °C for 2 h. The
reaction was cooled and then concentrated. Aqueous 1 N HCl (51
mL) was added, and the mixture was extracted with EtOAc (50 mL).
The organic layer was washed with 1 N HCl (51 mL). The combined
aqueous layers were basified to pH 10 with 50% aqueous NaOH, and
the mixture was extracted with EtOAc (3 × 100 mL). The combined
EtOAc extracts were dried over Na2SO4, filtered, and concentrated to
Benzyl rel-(4aR,7aR)-2-benzamido-7a-(2-thienyl)-4,4a,5,7-
tetrahydropyrrolo[3,4-d][1,3]thiazine-6-carboxylate (13.11g, 27.44
mmol) as a mixture of the cis-fused diastereomers was chirally
separated over multiple injections via SFC using a Chiralpak AS-H 50
× 150 mm column. The conditions used were 30% MeOH/70% CO2,
flow rate = 300 g/min, 100 bar, 40 °C, 7 mL injection size (∼525
1
mg). This resulted in the desired product (5.73g, 12.0 mmol). H
NMR (400.15 MHz, CDCl3): δ 8.10.
N-[(4aR,7aR)-7a-(2-Thienyl)-4a,5,6,7-tetrahydro-4H-pyrrolo[3,4-
d][1,3]thiazin-2-yl]benzamide (15f).
1
afford 17.2 g (99%) of product. H NMR (300.11 MHz, CDCl3): δ
8.23 (d, J = 0.8 Hz, 2H), 7.28−7.23 (m, 1H), 7.02−6.94 (m, 2H),
4.70−4.65 (m, 2H), 4.11−4.01 (m, 2H), 3.90−3.78 (m, 2H), 3.28
(dd, J = 3.9, 12.9 Hz, 1H), 2.95 (dd, J = 4.3, 13.0 Hz, 1H), 2.89−2.80
20
(m, 1H). LCMS (ESI+): m/z 336 [M + H]+. OR [α]589 = −121.4
(1.0, CH2Cl2).
(4aR,7aR)-6-(5-Fluoropyrimidin-2-yl)-7a-(2-thienyl)-4,4a,5,7-
tetrahydropyrrolo[3,4-d][1,3]thiazin-2-amine hydrochloride (15).
To a stirred solution of benzyl (4aR,7aR)-2-benzamido-7a-(2-thien-
yl)-4,4a,5,7-tetrahydropyrrolo[3,4-d][1,3]thiazine-6-carboxylate (47.1
g, 98.6 mmol) in acetonitrile (707 mL) was added iodotrimethylsi-
lane (26.1 mL, 177.5 mmol). The resulting yellow solution was
stirred at room temperature overnight. The reaction was the
concentrated to remove the acetonitrile, and aqueous 2 M HCl (1
L) was added followed by EtOAc (500 mL). The organic layer was
separated and washed with aqueous 2 M HCl (150 mL). The
combined aqueous layers were washed with isohexane (200 mL) and
then basified to pH 10 with 50% NaOH. The basified aqueous phase
was extracted with EtOAc (3 × 400 mL); the combined organic layers
were dried over sodium sulfate, and the solvent was removed in vacuo
A mixture of (4aR,7aR)-6-(5-fluoropyrimidin-2-yl)-7a-(2-thienyl)-
4,4a,5,7-tetrahydropyrrolo[3,4-d][1,3]thiazin-2-amine (603 mg, 1.18
mmol) in MTBE (7 mL) and EtOAc (4.5 mL) was heated to 65 °C
for 5 min to dissolve the solids. The resultant solution was then
treated with aqueous 1 N HCl (2.15 mL, 2.15 mmol), and a white
solid precipitated. The mixture was cooled to room temperature and
stirred and was then filtered, and the solids were washed with 1.6:1
8089
https://doi.org/10.1021/acs.jmedchem.1c00489
J. Med. Chem. 2021, 64, 8076−8100

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
MTBE/EtOAc. The solids were dried in a vacuum oven at 40 °C
overnight to afford 0.67 g (98%) of product. 1H NMR (400.13 MHz,
DMSO): δ 11.46 (s, 1H), 8.52 (s, 2H), 7.63 (dd, J = 1.0, 5.1 Hz, 1H),
7.23 (dd, J = 1.0, 3.7 Hz, 1H), 7.14 (dd, J = 3.7, 5.1 Hz, 1H), 4.25 (d,
J = 11.7 Hz, 1H), 4.05−4.02 (m, 1H), 3.91 (dd, J = 8.4, 11.1 Hz, 1H),
3.64−3.57 (m, 1H), 3.42 (dd, J = 4.0, 13.6 Hz, 1H), 3.36−3.30 (m,
1H), 3.21 (dd, J = 3.7, 13.4 Hz, 1H). HRMS (ESI+): m/z calcd for
C14H14FN5S2 (M + H) 335.0675, found 335.0684.
reddish-brown oil (38 g). The crude material was filtered through
silica gel using an eluent of 4:1 DCM/EtOAc. Fractions containing
desired product were combined and evaporated to give the desired
1
product (12.5 g, 45.1 mmol, 36% yield). H NMR (300.11 MHz,
CDCl3): δ 9.42−9.22 (m, 1H), 8.82−8.73 (m, 1H), 8.66−8.63 (m,
1H), 5.92−5.81 (m, 1H), 5.19−5.13 (m, 2H), 4.87 (s, 1H), 4.77 (s,
1H), 4.04−3.95 (m, 2H), 1.37 (s, 9H).). LCMS (ESI+): m/z 222 [M
+ H]+.
( )-tert-Butyl (3aR,6aR)-1-[(4-Methoxyphenyl)methyl]-6a-pyra-
zin-2-yl-3,3a,4,6-tetrahydropyrrolo[3,4-c]isoxazole-5-carboxylate
(16b).
Synthesis of Compound 16. tert-Butyl N-Allyl-N-[2-[methoxy-
(methyl)amino]-2-oxoethyl]carbamate (13).
A solution of Boc-N-(allyl)-glycine (85.5 g, 397.3 mmol) in
tetrahydrofuran (1.30 L) was chilled in a 0 °C ice bath under an
atmosphere of nitrogen. Triethylamine (83.1 mL, 596.0 mmol) was
added followed by the dropwise addition of pivaloyl chloride (53.5
mL, 437.0 mmol) over a period of 15 min, while maintaining the
internal temperature below 10 °C. The cooling bath was removed,
and the white mixture was stirred for 1 h while being warmed to rt.
The reaction was then returned to 0 °C and treated with additional
triethylamine (83.1 mL, 596.0 mmol) and N,O-dimethylhydroxyl-
amine hydrochloride (46.5 g, 476.8 mmol), after which the reaction
was moved to rt and stirred for 15 h. The mixture was concentrated,
and the resulting solid was transferred to a separatory funnel
containing EtOAc (700 mL), hexanes (100 mL), and 0.5 M aqueous
HCl (400 mL). The phases were partitioned, and the organic phase
further washed with dilute aqueous NaHCO3 (1 × 300 mL), water (1
× 200 mL), and saturated aqueous brine (1 × 200 mL). The organic
phase was dried over Na2SO4, filtered, and concentrated to give the
crude product as a pale-yellow liquid. The sample was purified via
silica gel chromatography eluting with hexanes/EtOAc (2-step
gradient from 70:30 to 60:40) to afford the desired product (80.26
g, 310.7 mmol, 78% yield). 1H NMR (400.15 MHz, CDCl3): δ 5.83−
5.77 (m, 1H), 5.16−5.08 (m, 2H), 4.12−4.09 (m, 1H), 4.01 (s, 1H),
3.96−3.92 (m, 1H), 3.92−3.89 (m, 1H), 3.68 (d, J = 10.7 Hz, 3H),
3.16 (d, J = 2.6 Hz, 3H), 1.43 (d, J = 12.1 Hz, 9H). LCMS (ESI+): m/
z 159 [M + H]+.
To a stirred solution of tert-butyl N-allyl-N-(2-oxo-2-pyrazin-2-
ylethyl)carbamate (0.14 g, 0.50 mmol), in anhydrous toluene (5
mL) were added N-[(4-methoxyphenyl)methyl]hydroxylamine hy-
drochloride (123 mg, 0.65 mmol) and triethylamine (104 μL, 0.75
mmol). The resulting solution was then treated with Ti(OEt)4 (416
uμL, 1.99 mmol) via syringe, and the mixture was heated at 75 °C
with stirring for 5 h. The mixture was cooled to rt followed by the
addition of water (5 mL) and EtOAc (50 mL). The mixture was
stirred for 10 min then filtered via Celite. The organic phase was dried
over MgSO4 and then concentrated. The sample was purified on a
column of SiO2 (24 g) using a gradient of 0−40% acetone in hexanes
to give the desired product as a mixture of cis-fused diastereomers
1
(126 mg; 0.290 mmol, 58% yield). H NMR (300.11 MHz, CDCl3):
δ 8.91 (s, 1H), 8.57−8.52 (m, 2H), 7.26−7.21 (m, 2H), 6.84 (d, J =
8.5 Hz, 2H), 4.33 (t, J = 8.2 Hz, 1H), 4.18−4.03 (m, 2H), 3.79 (s,
5H), 3.73−3.55 (m, 2H), 3.47−3.42 (m, 2H), 1.52 (s, 9H). LCMS
(ESI+): m/z 413 [M + H]+.
( )-(3aR,6aR)-1-[(4-Methoxyphenyl)methyl]-6a-pyrazin-2-yl-
3a,4,5,6-tetrahydro-3H-pyrrolo[3,4-c]isoxazole (16c).
tert-Butyl N-Allyl-N-(2-oxo-2-pyrazin-2-ylethyl)carbamate (16a).
Trifluoroacetic acid (21.45 mL, 283.64 mmol) was added to a stirred
solution of ( )-tert-butyl (3aR,6aR)-1-[(4-methoxyphenyl)methyl]-
6a-pyrazin-2-yl-3,3a,4,6-tetrahydropyrrolo[3,4-c]isoxazole-5-carboxy-
late (3.90 g, 9.45 mmol) in dichloromethane (100 mL) at rt. The
mixture was stirred at rt for 1.5 h and then concentrated. The residue
was dissolved in MeOH (15 mL). This solution was loaded onto an
ion-exchange resin (SCX) column (50 g). The column was first
washed with MeOH (80 mL) and then ammoniated MeOH (100
mL). The basic fraction contained product by LC/MS and was
concentrated to afford the semipure product. This material was
further purified on a column of SiO2 (120g) using a gradient 0−10%
ammoniated MeOH in DCM to give the desired product (1.98 g, 6.34
A solution of 2,2,6,6-tetramethylpiperidine (25.4 mL, 150 mmol) in
tetrahydrofuran (250 mL) was cooled to −70 °C. Butyl lithium (89.7
mL, 144 mmol) was added over 15 min. The solution was stirred at
−70 °C for 10 min, and then a solution of pyrazine (10 g, 124.9
mmol) in tetrahydrofuran (25 mL) was added over 30 min while the
temperature was kept below −65 °C. This resulted in a color change
in the solution from yellow to a dark red. After 5 min of stirring, a
solution of tert-butyl N-allyl-N-[2-[methoxy(methyl)amino]-2-
oxoethyl]carbamate (19.4 g, 74.9 mmol) in tetrahydrofuran (25
mL) was added dropwise, again keeping the temperature of the
reaction at −65 °C. The reaction was then stirred for an additional 30
min at −70 °C. Then reaction mixture was warmed to −30 °C,
checked by HPLC, and then diluted with MTBE (250 mL) and water
(100 mL). Some NaCl solution was added to improve the phase
separation and then transferred to a separatory funnel. The organic
layer was isolated and washed twice with aqueous MTBE (2 × 125
mL). The combined organic layers were dried over sodium sulfate and
filtered. Filtrates were evaporated to afford the crude product as a
1
mmol, 67% yield). H NMR (300.11 MHz, CDCl3): δ 8.86 (s, 1H),
8.55 (dd, J = 1.5, 2.4 Hz, 1H), 8.51 (d, J = 2.4 Hz, 1H), 7.20 (d, J =
8.7 Hz, 2H), 6.85−6.82 (m, 2H), 4.32−4.02 (m, 1H), 3.78 (s, 3H),
3.65 (dd, J = 4.4, 8.6 Hz, 1H), 3.43 (s, 2H), 3.39−3.21 (m, 3H), 2.99
(dd, J = 2.6, 11.9 Hz, 1H). LCMS (ESI+): m/z 313 [M + H]+.
8090
https://doi.org/10.1021/acs.jmedchem.1c00489
J. Med. Chem. 2021, 64, 8076−8100

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
( )-(3aR,6aR)-5-(5-Fluoropyrimidin-2-yl)-1-[(4-methoxyphenyl)-
methyl]-6a-pyrazin-2-yl-3,3a,4,6-tetrahydropyrrolo[3,4-c]isoxazole
(16d).
( )-N-[[(3R,4R)-1-(5-Fluoropyrimidin-2-yl)-4-(hydroxymethyl)-3-
pyrazin-2-yl-pyrrolidin-3-yl]carbamothioyl]benzamide (16g).
To a stirred suspension of [( )-(3R,4R)-4-amino-1-(5-fluoropyrimi-
din-2-yl)-4-pyrazin-2-yl-pyrrolidin-3-yl]methanol (960 mg, 3.31
mmol) in tetrahydrofuran (35 mL) under N2 was added benzoyl
isothiocyanate (546 μL, 3.97 mmol). The reaction was continued at rt
for 40 min. The solvent was evaporated to give the crude product.
The crude sample was purified on a column of SiO2 (40g) using a
gradient 0−10% MeOH(NH3) in DCM to give the desired product
A solution of ( )-(3aR,6aR)-1-[(4-methoxyphenyl)methyl]-6a-pyr-
azin-2-yl-3a,4,5,6-tetrahydro-3H-pyrrolo[3,4-c]isoxazole (1.98 g, 6.34
mmol) in 1,4-dioxane (44 mL) was treated with 2-chloro-5-
fluoropyrimidine (3.78 g, 28.52 mmol) and diisopropylethylamine
(5.53 mL, 31.69 mmol). The reaction was heated at 110 °C for 3 h.
The solvent was evaporated to afford the crude product. The crude
product was purified on a column of SiO2 (120g) using a gradient 0−
5% MeOH(NH3) in DCM to give the desired product (2.08 g; 4.58
mmol, 72% yield). 1H NMR (300.11 MHz, CDCl3): δ 9.01 (d, J = 1.3
Hz, 1H), 8.56−8.54 (m, 1H), 8.52 (d, J = 2.4 Hz, 1H), 8.28 (d, J =
0.8 Hz, 2H), 7.25−7.21 (m, 2H), 6.86−6.82 (m, 2H), 4.41−4.31 (m,
3H), 3.93−3.84 (m, 3H), 3.80−3.79 (m, 3H), 3.72 (s, 1H), 3.61−
3.50 (m, 2H). LCMS (ESI+): m/z 409 [M + H]+.
1
(1.57g, 3.46 mmol, 99%). H NMR (300.11 MHz, CDCl3): δ 11.97
(s, 1H), 9.02−8.96 (m, 1H), 8.84−8.81 (m, 1H), 8.57−8.55 (m, 1H),
8.50−8.47 (m, 1H), 8.26 (d, J = 0.8 Hz, 2H), 7.94−7.82 (m, 2H),
7.67−7.62 (m, 1H), 7.55−7.49 (m, 2H), 4.97 (d, J = 12.1 Hz, 1H),
4.76 (d, J = 12.2 Hz, 1H), 4.17−4.07 (m, 1H), 4.06−3.98 (m, 2H),
3.86−3.76 (m, 2H), 3.51 (s, 1H), 3.11−3.06 (m, 1H), 2.97−2.86 (m,
1H). LCMS (ESI+): m/z 454 [M + H]+.
( )-N-[(4aR,7aR)-6-(5-Fluoropyrimidin-2-yl)-7a-pyrazin-2-yl-
4,4a,5,7-tetrahydropyrrolo[3,4-d][1,3]thiazin-2-yl]benzamide
(16h).
( )-(3aR,6aR)-5-(5-Fluoropyrimidin-2-yl)-6a-pyrazin-2-yl-
3,3a,4,6-tetrahydro-1H-pyrrolo[3,4-c]isoxazole (16e).
Compound ( )-(3aR,6aR)-5-(5-fluoropyrimidin-2-yl)-1-[(4-
methoxyphenyl)methyl]-6a-pyrazin-2-yl-3,3a,4,6-tetrahydropyrrolo-
[3,4-c]isoxazole (2.08 g, 5.09 mmol) was mixed with trifluoroacetic
acid (20.4 mL, 269.9 mmol) and stirred at 60 °C for 1 h. The solvent
was evaporated, and the residue was dissolved in a mixture of MeOH
and DMSO. The mixture was loaded onto a 25 g ion-exchange (SCX)
column. The column was washed with MeOH (70 mL) and then with
ammoniated methanol (70 mL). The basic fraction was concentrated
to give the semipure product, which was then further purified via silica
gel chromatography using a 40 g SiO2 column and eluting with a
gradient of 0−10% MeOH(NH3) in DCM. This resulted in the
A solution of ( )-N-[[(3R,4R)-1-(5-fluoropyrimidin-2-yl)-4-(hydrox-
ymethyl)-3-pyrazin-2-yl-pyrrolidin-3-yl]carbamothioyl]benzamide
(1.57g, 3.46 mmol) in dichloromethane (35 mL) was cooled to 0 °C.
1-Chloro-N,N,2-trimethylpropenylamine (480 μL, 3.64 mmol) was
added. The solution was warmed to rt and stirred for 2 h. An
additional aliquot of 1-chloro-N,N,2-trimethylpropenylamine (192
μL, 1.45 mmol) was added, and the reaction was continued for 45
min, at which time the reaction was observed by LC/MS to be
complete. Saturated Na2CO3 (aqueous, 30 mL) was added and the
mixture diluted with DCM (150 mL). Aqueous phase was separated
and extracted additionally with DCM (150 mL). Combined organic
phases were dried, and solvent was evaporated to afford the crude
product. The crude product was dissolved in MeOH (10 mL) and
loaded on a SCX ion-exchange column (25 g). The column was first
washed with MeOH (50 mL) and then MeOH(NH3) (50 mL). The
basic fraction was concentrated to give the crude product, which was
purified via silica gel chromatography using a 120 g column and
eluted with a gradient of 0−5% ammoniated methanol in DCM to
give the desired product as a mixture of the cis-fused diastereomers
(900 mg, 2.07 mmol, 60% yield). 1H NMR (300.11 MHz, CDCl3): δ
13.14−13.11 (m, 0H), 8.86−8.80 (m, 1H), 8.62−8.60 (m, 2H), 8.24
(s, 2H), 8.15−8.11 (m, 2H), 7.57−7.46 (m, 3H), 7.28 (s, 3H), 5.32
(s, 1H), 4.25−4.13 (m, 2H), 4.02 (d, J = 9.0 Hz, 2H), 3.68−3.57 (m,
1H), 3.24−3.19 (m, 1H), 3.00−2.94 (m, 1H). LCMS (ESI+): m/z
436 [M + H]+.
1
desired product (1.1 g, 3.24 mmol, 64% yield). H NMR (300.11
MHz, CDCl3): δ 9.07 (d, J = 0.6 Hz, 1H), 8.56−8.52 (m, 2H), 8.25
(s, 2H), 4.55−4.42 (m, 1H), 4.27−4.07 (m, 1H), 4.04−3.86 (m, 2H),
3.84−3.77 (m, 3H). LCMS (ESI+): m/z 289 [M + H]+.
[( )-(3R,4R)-4-Amino-1-(5-fluoropyrimidin-2-yl)-4-pyrazin-2-yl-
pyrrolidin-3-yl]methanol (16f).
A solution of ( )-(3aR,6aR)-5-(5-fluoropyrimidin-2-yl)-6a-pyrazin-2-
yl-3,3a,4,6-tetrahydro-1H-pyrrolo[3,4-c]isoxazole (1.10g, 3.82 mmol)
in ethanol (30 mL) was hydrogenated at 50 psi in the presence of
Raney nickel (slurry in water) (1.4 g, 23.61 mmol). The reaction was
stirred for 4 h at rt. The mixture was diluted with MeOH and the
catalyst removed by decantation. Organic phase was additionally
filtered through Celite. The solvent was evaporated to afford the
(4aR,7aR)-6-(5-Fluoropyrimidin-2-yl)-7a-pyrazin-2-yl-4,4a,5,7-
tetrahydropyrrolo[3,4-d][1,3]thiazin-2-amine (16).
1
desired product (0.96 g, 3.14 mmol, 82% yield). H NMR (300.11
MHz, CDCl3): δ 8.89−8.88 (m, 1H), 8.57−8.56 (m, 2H), 8.24 (d, J =
0.6 Hz, 2H), 4.08−3.89 (m, 5H), 3.64 (dd, J = 9.4, 11.1 Hz, 1H),
3.08−3.00 (m, 1H). LCMS (ESI+): m/z 291 [M + H]+.
A mixture of ( )-N-[(4aR,7aR)-6-(5-fluoropyrimidin-2-yl)-7a-pyra-
zin-2-yl-4,4a,5,7-tetrahydropyrrolo[3,4-d][1,3]thiazin-2-yl]benzamide
8091
https://doi.org/10.1021/acs.jmedchem.1c00489
J. Med. Chem. 2021, 64, 8076−8100

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
(900 mg, 2.07 mmol), pyridine (2.01 mL, 24.80 mmol) and o-
methylhydroxylamine hydrochloride (1.59 g, 18.60 mmol) in ethanol
(65 mL) was heated to 80 °C with stirring for 3 h. The solvent was
evaporated, and the crude product was supported on SiO2 and
purified on a column of SiO2 (40 g) using a gradient 0−10%
MeOH(NH3) in DCM over 25 column volume, which resulted in a
semipure product. The product was dissolved in MeOH (20 mL) and
loaded on a SCX ion-exchange column (25g). The column was first
washed with MeOH (100 mL) then MeOH(NH3) (100 mL). The
basic fraction was concentrated to give the desired product as a mix of
cis-fused diastereomers. The sample was chirally purified via SFC
using a 150 × 21.2 mm i.d., 60A, 5u ODH column (Princeton
Chromatography) run at 100 bar, 35 °C, and eluted with a 0−50%
(0.2% diethylamine in MeOH)/CO2 gradient to give the desired
× 30 mL) and brine (1 × 30 mL). The organic phase was dried over
Na2SO4, filtered, and concentrated. The sample was purified via silica
gel chromatography eluting with a 20−100% EtOAc/hexanes gradient
to give the desired product (2.59 g, 5.15 mmol, 74% yield). 1H NMR
(399.83 MHz, CDCl3): δ 7.98−7.97 (m, 2H), 7.60−7.47 (m, 4H),
7.36−7.29 (m, 5H), 6.99−6.97 (m, 1H), 5.19−5.09 (m, 2H), 3.97−
3.91 (m, 3H), 3.17−3.15 (m, 1H), 2.94−2.91 (m, 2H). LCMS
(ESI+): m/z 503 [M + H]+.
N-[(4aR,7aR)-7a-(5-Cyano-2-thienyl)-4a,5,6,7-tetrahydro-4H-
pyrrolo[3,4-d][1,3]thiazin-2-yl]benzamide (17c).
1
chirally pure material (233 mg, 0.70 mmol, 34% yield). H NMR
(300.11 MHz, CDCl3): δ 8.81 (d, J = 1.3 Hz, 1H), 8.53−8.48 (m,
2H), 8.21 (s, 2H), 4.16 (d, J = 10.9 Hz, 1H), 4.00 (d, J = 10.9 Hz,
1H), 3.88 (d, J = 8.9 Hz, 2H), 3.47 (s, 2H), 3.39−3.31 (m, 1H), 3.14
(dd, J = 4.0, 13.2 Hz, 1H), 2.95 (dd, J = 4.4, 13.1 Hz, 1H). LCMS
(ESI+): m/z 332 [M + H]+.
Synthesis of Compound 17. Benzyl (4aR,7aR)-2-Benzamido-7a-
(5-bromo-2-thienyl)-4,4a,5,7-tetrahydropyrrolo[3,4-d][1,3]thiazine-
6-carboxylate (17a).
To a stirred clear and colorless solution of benzyl (4aR,7aR)-2-
benzamido-7a-(5-cyano-2-thienyl)-4,4a,5,7-tetrahydropyrrolo[3,4-d]-
[1,3]thiazine-6-carboxylate (641 mg, 1.28 mmol) in acetonitrile (26
mL) was added iodotrimethylsilane (0.55 mL, 3.83 mmol). The
resulting yellow solution was stirred at room temperature for 70 min.
Methanol (0.62 mL, 15.30 mmol) was added, and the reaction was
stirred for 5 min. The reaction was then concentrated to give an
amber solid. This material was purified via silica gel chromatography
eluting with a 0−10% MeOH/DCM gradient to give the desired
product as a white solid (590 mg, 1.60 mmol, 125% yield). Note,
some residual DCM was present, which accounts for the extra mass,
but the sample was taken on and used as is in the next step. 1H NMR
(400.15 MHz, CDCl3): δ 7.85−7.83 (m, 2H), 7.59−7.54 (m, 1H),
7.50 (d, J = 4.0 Hz, 1H), 7.46−7.41 (m, 2H), 7.10 (d, J = 4.0 Hz,
1H), 4.10−4.08 (m, 1H), 3.89−3.88 (m, 1H), 3.67−3.62 (m, 2H),
3.27−3.26 (m, 2H), 3.15−3.13 (m, 1H). LCMS (ESI+): m/z 369 [M
+ H]+.
To a stirred clear and colorless solution of benzyl (4aR,7aR)-2-
benzamido-7a-(2-thienyl)-4,4a,5,7-tetrahydropyrrolo[3,4-d][1,3]-
thiazine-6-carboxylate (3.71 g, 7.77 mmol) in dimethylformamide (78
mL) at rt and under an atmosphere of nitrogen was added N-
bromosuccinimide (1.66g, 9.32 mmol). The resulting solution was
stirred for 65 min. The solution was diluted with EtOAc (250 mL)
and hexanes (50 mL) and then transferred to a separatory funnel,
washed with water (2 × 50 mL) and with brine (1 × 50 mL). The
organic phase was dried over Na2SO4, filtered, and concentrated. The
residue was purified via silica gel chromatography eluting with
hexanes/EtOAc, gradient from 80:20 to 0:100 to afford the desired
N-[(4aR,7aR)-7a-(5-Cyano-2-thienyl)-6-(5-fluoropyrimidin-2-yl)-
4,4a,5,7-tetrahydropyrrolo[3,4-d][1,3]thiazin-2-yl]benzamide
(17d).
1
product as a white foam (3.88 g, 6.97 mmol, 90% yield). H NMR
(400.15 MHz, CDCl3): δ 8.11−8.08 (m, 2H), 7.57−7.56 (m, 1H),
7.49−7.44 (m, 2H), 7.38−7.33 (m, 5H), 6.97−6.95 (m, 1H), 6.79−
6.77 (m, 1H), 5.21−5.19 (m, 2H), 3.98−3.96 (m, 4H), 3.30−3.29
(m, 1H), 2.92−2.90 (m, 2H). LCMS (ESI+): m/z 556/558 [M + H]+.
Benzyl (4aR,7aR)-2-Benzamido-7a-(5-cyano-2-thienyl)-4,4a,5,7-
tetrahydropyrrolo[3,4-d][1,3]thiazine-6-carboxylate (17b).
To a stirred clear and colorless solution of N-[(4aR,7aR)-7a-(5-
cyano-2-thienyl)-4a,5,6,7-tetrahydro-4H-pyrrolo[3,4-d][1,3]thiazin-2-
yl]benzamide (470 mg, 1.28 mmol) in 1,4-dioxane (25 mL) at rt was
added diisopropylethylamine (0.667 mL, 3.825 mmol) followed by 5-
fluoro-2-chloropyrimidine (1.22 mL, 12.75 mmol). The reaction was
heated to reflux (heating block temp = 110 °C) for 5 h with
monitoring by LC/MS. The clear pale-yellow solution was cooled for
∼15 min. The solution was diluted with EtOAc (130 mL) and
hexanes (30 mL) and then washed with dilute aqueous NaHCO3 (1 ×
50 mL), water (2 × 40 mL), and brine (1 × 40 mL). The organic
phase was dried over Na2SO4, filtered, and concentrated. The residue
was purified by silica gel chromatography eluting with a gradient of
50−100% hexanes/Et2O to give the desired product (453 mg, 0.98
mmol, 76% yield). 1H NMR (400.15 MHz, CDCl3): δ 8.22 (d, J = 0.7
Hz, 2H), 8.00−7.99 (m, 2H), 7.56−7.53 (m, 2H), 7.47−7.44 (m,
2H), 7.04−7.03 (m, 1H), 4.18−4.17 (m, 1H), 4.03−4.01 (m, 3H),
3.29−3.23 (m, 1H), 3.07−3.06 (m, 2H). LCMS (ESI+): m/z 465 [M
+ H]+.
A mixture of benzyl (4aR,7aR)-2-benzamido-7a-(5-bromo-2-thienyl)-
4,4a,5,7-tetrahydropyrrolo[3,4-d][1,3]thiazine-6-carboxylate (3.88 g,
6.97 mmol), potassium hexacyanoferrate(II) (2.57g, 6.97 mmol),
copper(I) iodide (398 mg, 2.09 mmol), and 1-butylimidazole (2.75
mL, 20.92 mmol) in toluene (23 mL) was stirred together in a
pressure tube at rt. A stream of nitrogen was passed over the mixture
for a few minutes, and then the tube was sealed with a pressure gauge.
The stirred mixture was placed in a 140 °C oil bath for 18 h. The
reaction was then cooled to rt. The mixture was diluted with EtOAc
(100 mL) and water. The mixture was filtered through a pad of Celite
while being rinsed with additional EtOAc (100 mL) and water (15
mL). The filtrate was transferred to a separatory funnel, and the layers
were partitioned. The organic phase was further washed with water (2
8092
https://doi.org/10.1021/acs.jmedchem.1c00489
J. Med. Chem. 2021, 64, 8076−8100

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
5-[(4aR,7aR)-2-Amino-6-(5-fluoropyrimidin-2-yl)-4,4a,5,7-
tetrahydropyrrolo[3,4-d][1,3]thiazin-7a-yl]thiophene-2-carbonitrile
hydrochloride (17).
EtOAc/hexanes to afford the product as a yellow oil (80.26 g, 78%
1
yield). H NMR (400 MHz, CDCl3): δ 5.84−5.68 (m, 1H), 5.22−
5.09 (m, 2H), 4.04−3.84 (m, 4H), 1.95 (d, J = 8.79 Hz, 9 H). ESI+:
m/z 214 [M − H]−.
tert-Butyl N-Allyl-N-(2-isothiazol-5-yl-2-oxoethyl)carbamate
(18a).
A solution of N-[(4aR,7aR)-7a-(5-cyano-2-thienyl)-6-(5-fluoropyr-
imidin-2-yl)-4,4a,5,7-tetrahydropyrrolo[3,4-d][1,3]thiazin-2-yl]-
benzamide (442 mg, 0.951 mmol) in ethanol (19.0 mL) was treated
with pyridine (752 mg, 9.5 mmol) and o-methylhydroxylamine
hydrochloride (795 mg, 9.5 mmol). The stirred solution was heated
to 55 °C under an atmosphere of nitrogen for 12 h. The clear and
colorless solution was concentrated, diluted with DCM (30 mL), and
then reconcentrated. The residue was transferred to a separatory
funnel, while being diluted with EtOAc (50 mL), water (25 mL), and
saturated aqueous NaHCO3 (10 mL). The phases were separated, and
the aqueous phase extracted with EtOAc (1 × 30 mL). The combined
EtOAc extracts were dried over Na2SO4, filtered, and concentrated.
The residue was purified via silica gel chromatography using a 30−
100% EtOAc/hexanes gradient giving semipure product. The residue
was further purified by silica gel chromatography using a 30−100%
EtOAc/DCM gradient to afford the free base of the desired product.
The free base was dissolved in Et2O (20 mL) and treated with 1 M
HCl in Et2O (0.86 mL, 0.86 mmol). A white solid precipitated, and
the mixture was stirred for a few minutes, then the white precipitate
was collected by vacuum filtration through filter paper, rinsing with
Et2O to afford the desired product, 5-[(4aR,7aR)-2-amino-6-(5-
fluoropyrimidin-2-yl)-4,4a,5,7-tetrahydropyrrolo[3,4-d][1,3]thiazin-
7a-yl]thiophene-2-carbonitrile hydrochloride (255 mg, 0.643 mmol,
68% yield). 1H NMR (400.15 MHz, dmso): δ 11.58−11.57 (m, 1H),
8.51 (d, J = 0.8 Hz, 2H), 8.00 (d, J = 4.1 Hz, 1H), 7.42 (d, J = 4.1 Hz,
1H), 4.30−4.26 (m, 1H), 4.11−4.10 (m, 4H), 3.63−3.61 (m, 1H),
3.44−3.41 (m, 2H), 3.24−3.21 (m, 1H). HRMS (ESI+): m/z calcd
A solution of diisopropylamine (36.24 mL, 257.47 mmol) in
tetrahydrofuran (300 mL) was cooled to −78 °C. n-Butyl lithium
(124 mL, 123.78 mmol, 1.6 M in hexanes) was added dropwise over
20 min. The resulting clear solution was stirred at −78 °C for an
additional 45 min. A solution of isothiazole (16.86 g, 198.05 mmol) in
tetrahydrofuran (50 mL) was added dropwise over 15 min. The white
suspension was stirred at −78 °C for 45 min. This suspension was
treated dropwise with a solution of tert-butyl N-allyl-N-[2-(methoxy-
(methyl)amino)-2-oxoethyl]carbamate (25.58 g, 99.02 mmol) in
tetrahydrofuran (150 mL). Reaction mixture was allowed to gradually
reach room temperature and then stirred overnight. Reaction was
quenched with saturated ammonium chloride (500 mL) and extracted
with ethyl acetate. The two phases were separated, and the aqueous
phase was extracted with ethyl acetate (3 × 100 mL). The combined
organics were dried over magnesium sulfate and filtered. The solvent
was removed under reduced pressure. The crude mixture was purified
over 330 g of silica gel, with a 35 min gradient of 0 to 100% EtOAc/
hexanes to afford the product as a yellow solid (21.62 g, 77% yield).
1H NMR (400 MHz, CDCl3): δ 8.59−8.54 (m, 1H), 7.77−7.70 (m,
1H), 5.87−5.72 (m, 1H), 5.20−5.09 (m, 2H), 4.47 (s, 1H), 4.38−
4.32 (m, 1H), 4.04−3.98 (m, 1H), 3.96−3.89 (m, 1H), 1.47 (s,
4.5H), 1.34 (s, 4.5H). ESI+: m/z 281 [M − H]−.
( )-tert-Butyl-6a-isothiazol-5-yl-1-[(4-methoxyphenyl)methyl]-
3,3a,4,6-tetrahydropyrrolo[3,4-c]isoxazole-5-carboxylate (18b).
20
for C15H13FN6S2 (M + H) 360.0627, found 360.0632. OR [α]D
+19.986 (c = 0.1, MeOH).
Synthesis of Compound 18. tert-Butyl N-Allyl-N-[2-[methoxy-
To a stirred suspension of the tert-butyl N-allyl-N-(2-isothiazol-5-yl-2-
oxoethyl)carbamate (21.22 g, 75.15 mmol) in toluene (500 mL) were
added N-[(4-methoxyphenyl)methyl]hydroxylamine (18.53 g, 97.70
mmol), diisopropylethylamine (17.03 mL, 97.70 mmol), and titanium
tetraethoxide (31.43 mL, 150.30 mmol) under an atmosphere of
nitrogen. The reaction mixture became a clear solution and was
heated to 70 °C and stirred overnight, cooled to room temperature,
diluted with ethyl acetate and water, and then filtered through Celite.
The solid was washed with ethyl acetate; then the filtrate was placed
in a separatory funnel, and the phases were separated. The organic
phase was dried over magnesium sulfate and filtered. The solvent was
removed under reduced pressure. The crude mixture was purified over
330 g of silica gel, with a 30 min gradient of 0 to 50% EtOAc/hexanes
to afford the product as a yellow solid (31.00 g, 99% yield). 1H NMR
(400 MHz, CDCl3): δ 8.42 (d, J = 1.65 Hz, 1H), 7.26 (d, J = 9.02 Hz,
2H), 7.16−7.12 (m, 1H), 6.84 (d, J = 8.36 Hz, 2H), 4.13 (t, J = 8.25
Hz, 1H), 3.75 (s, 3H), 3.75−3.64 (m, 4H), 3.57−3.38 (m, 1H),
3.15−3.06 (m, 1H), 1.48 (s, 9H). ESI+: m/z 418 [M + H]+.
( )-tert-Butyl-6a-isothiazol-5-yl-3,3a,4,6-tetrahydro-1H-pyrrolo-
[3,4-c]isoxazole-5-carboxylate (18c).
(methyl)amino]-2-oxoethyl]carbamate (13).
To a three-neck round flask equipped with an overhead mechanical
stir paddle, were added ethyl 2-[allyl(tert-butoxycarbonyl)amino]-
acetate (85.52 g, 397.311 mmol) and tetrahydrofuran (1.32 L) and
then cooled to 0 °C in an ice bath and under an atmosphere of
nitrogen. Triethylamine (83.07 mL, 595.966 mmol) was added,
followed by pivaloyl chloride (53.53 mL, 437.04 mmol), which was
added dropwise over a period of 15 min. The internal reaction
temperature was maintained below 10 °C. Mixture was allowed to
gradually reach ambient temperature for 1 h. The white suspension
was cooled again to 0 °C, and then triethylamine (83.067 mL,
595.966 mmol) and N,O-dimethylhydroxylamine hydrochloride
(46.51, 476.773 mmol) were added. Reaction mixture was stirred at
ambient temperature for 16 h. The solvent was removed in vacuo, and
the white residue was diluted with ethyl acetate (700 mL), hexanes
(100 mL), and aqueous HCl (0.5 M, 400 mL). The phases were
separated, and the organic phase was further washed with saturated
NaHCO3 (300 mL), water (200 mL), and brine (200 mL). The
organic phase was dried over sodium sulfate and filtered. The solvent
was removed under reduced pressure. The crude mixture was purified
by silica gel chromatography, with a 35 min gradient of 30 to 95%
To
a solution of ( )-tert-butyl 6a-isothiazol-5-yl-1-[(4-
methoxyphenyl)methyl]-3,3a,4,6-tetrahydropyrrolo[3,4-c]isoxazole-5-
carboxylate (27.10 g, 64.91 mmol) in trifluoroacetic acid (250 mL)
8093
https://doi.org/10.1021/acs.jmedchem.1c00489
J. Med. Chem. 2021, 64, 8076−8100

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
was added triethylsilane (20.80 mL, 129.81 mmol). Reaction was
heated to 80 °C for 3 h, concentrated in vacuo to a minimum amount
of TFA, and then MeOH was added until a white precipitate formed.
The solid was collected by filtration and then was washed several
times with ethyl acetate. After being dissolved in tetrahydrofuran/
water (1:1, 150 mL), NaOH (5 N, 13 mL) was added. Reaction
mixture was stirred vigorously, and di-tert-butyl-dicarbonate (9.03 g,
41.3 mmol) was added as a solution in tetrahydrofuran (10 mL) and
stirred at room temperature for 10 min. Ethyl acetate (200 mL) and
brine (150 mL) were added. The phases were separated, and the
aqueous phase was extracted with ethyl acetate. The combined
organics were dried over magnesium sulfate and filtered. The solvent
was removed under reduced pressure to afford the product as an
amber color foam (12.4 g, 64.24% yield). 1H NMR (400 MHz,
CDCl3): δ 8.38 (d, J = 1.60 Hz, 1H), 7.12 (d, J = 1.60 Hz, 1H), 4.40
(t, J = 7.37 Hz 1H), 3.98−3.88 (m, 1H), 3.75−3.64 (m, 2H), 3.63−
3.56 (m, 1H), 3.53 (d, J = 13.19 Hz, 1H), 3.26−3.19 (m, 1H), 1.43
(s, 9H). LCMS (ESI+): m/z 298 [M + H]+.
( )-tert-Butyl-2-benzamido-7a-isothiazol-5-yl-4,4a,5,7-
tetrahydropyrrolo[3,4-d][1,3]thiazine-6-carboxylate (18f).
( )-tert-Butyl-3-(benzoylcarbamothioylamino)-4-(hydroxymethyl)-3-
isothiazol-5-yl-pyrrolidine-1-carboxylate (21.60 g, 46.70 mmol) was
dissolved in dichloromethane (500 mL) and cooled to 0 °C. Chloro-
N,N,2-trimethylpropenylamine (12.35 mL, 93.385 mmol) was added
slowly. The mixture was stirred for 2 h and then warmed to ambient
temperature and stirred for 3 h. The reaction mixture was washed
with water and then saturated sodium bicarbonate. The aqueous
phase was extracted with ethyl acetate (3×). The combined organic
phase was dried over magnesium sulfate and filtered. The solvent was
removed in vacuo. The crude residue was purified over 330 g of silica
gel, with a 20 min gradient of 0 to 50% EtOAc/dichloromethane, to
afford the product (19.54 g, 94% yield). 1H NMR (400 MHz,
CDCl3): δ 8.41 (s, 1H), 7.99−7.87 (m, 2H), 7.57−7.49 (m, 1H),
7.49−7.41 (m, 2H), 7.13−7.05 (m, 1H), 3.84−3.65 (m, 4H), 3.21−
3.04 (m, 1H), 2.87−2.72 (m, 2H), 1.45−1.41 (m, 9H). MS (ESI+):
m/z 445 [M + H]+.
( )-tert-Butyl-1-(benzoylcarbamothioyl)-6a-isothiazol-5-yl-
3,3a,4,6-tetrahydropyrrolo[3,4-c]isoxazole-5-carboxylate (18d).
tert-Butyl (4aR,7aR)-2-Benzamido-7a-isothiazol-5-yl-4,4a,5,7-
tetrahydropyrrolo[3,4-d][1,3]thiazine-6-carboxylate (18g, Chiral
Purification).
( )-tert-Butyl-6a-isothiazol-5-yl-3,3a,4,6-tetrahydro-1H-pyrrolo[3,4-
c]isoxazole-5-carboxylate (13.75 g, 46.24 mmol) was dissolved in
tetrahydrofuran (200 mL). Benzoyl isothiocyanate (9.98 mL, 73.98
mmol) was added under a nitrogen atmosphere and stirred at ambient
temperature for 18 h. The solvent was removed in vacuo. The crude
product was purified over 330 g of silica gel, with a 30 min gradient of
0 to 50% ethyl acetate/dichloromethane to afford the product (21.65
(
)-tert-Butyl-2-benzamido-7a-isothiazol-5-yl-4,4a,5,7-
tetrahydropyrrolo[3,4-d][1,3]thiazine-6-carboxylate (19.57 g, 44.019
mmol) as a mixture of the cis-fused diastereomers was chirally
separated over multiple injections via SFC using a Chiralpak IC, 50 ×
150 mm column. The conditions used were 40% MeOH/60% CO2,
flow rate = 300 g/min, 100 bar, 40 °C, 8 mL injection size (800 mg).
This resulted in the desired product (4.65 g, 10.473 mmol, DE =
1
g, 100% yield). H NMR (400 MHz, CDCl3): δ 8.82−8.65 (m, 1H),
8.37 (t, J = 1.4 Hz, 1H), 7.83−7.81 (m, 2H), 7.59−7.55 (m, 1H),
7.47 (t, J = 7.15 Hz, 2H), 7.23−7.23 (m, 1H), 4.86 (m, 1H), 4.51−
4.48 (m, 1H), 4.31−4.17 (m, 1H), 4.11−4.04 (m, 1H), 3.89−3.81
(m, 1H), 3.64−3.48 (m, 1H), 3.39−3.31 (m, 1H), 1.47 (d, J = 1.0 Hz,
9H). ESI+: m/z 461 [M + H]+.
1
98%). H NMR (400 MHz, CDCl3): δ 8.41 (d, J = 1.2 Hz, 1H),
8.04−7.87 (m, 2H), 7.53 (t, J = 7.04 Hz, 1H), 7.53 (t, J = 7.04 Hz,
2H), 7.13−7.05 (m, 1H), 3.92−3.66 (m, 4H), 3.17−3.07 (m, 1H),
2.93−2.78 (m, 2H), 1.43 (d, J = 4.95 Hz, 9H). MS (ESI+): m/z 445
[M + H]+.
( )-tert-Butyl-3-(benzoylcarbamothioylamino)-4-(hydroxymeth-
yl)-3-isothiazol-5-yl-pyrrolidine-1-carboxylate (18e).
N-[(4aR,7aR)-7a-Isothiazol-5-yl-4a,5,6,7-tetrahydro-4H-pyrrolo-
[3,4-d][1,3]thiazin-2-yl]benzamide (18h).
To a solution of ( )-tert-butyl-1-(benzoylcarbamothioyl)-6a-isothia-
zol-5-yl-3,3a,4,6-tetrahydropyrrolo[3,4-c]isoxazole-5-carboxylate
(21.65 g, 47.01 mmol) in acetic acid (235 mL) was added zinc (dust,
30.70 g, 470.06 mmol), stirred vigorously at ambient temperature for
45 min, and filtered through Celite; then the solvent was removed in
vacuo. The residue was dissolved in ethyl acetate and washed with
water and saturated sodium bicarbonate. The aqueous phase was
extracted with ethyl acetate. The combined organics were dried over
sodium sulfate and filtered. The solvent was removed under reduced
pressure to afford the product as a white foam (21.60 g, 99% yield).
1H NMR (400 MHz, CDCl3): δ 11.83 (m, 1H), (m, 1H), 9.00−8.93
(s, 1H), 7.85−7.81 (m, 2H), 7.65−7.61 (m, 1H), 7.54−7.50 (m, 2H),
7.17 (d, J = 1.8 Hz, 1H), 4.73 (d, J = 12.1 Hz, 1H), 4.25−4.09 (m,
1H), 4.01−3.95 (m, 2H), 3.80−3.68 (m, 1H), 3.67−3.57 (m, 1H),
2.74−2.61 (m, 1H), 1.48 (m, 9H). LCMS (ESI+): m/z 463 [M + H]+.
tert-Butyl (4aR,7aR)-2-benzamido-7a-isothiazol-5-yl-4,4a,5,7-
tetrahydropyrrolo[3,4-d][1,3]thiazine-6-carboxylate (4.65 g, 10.46
mmol) was dissolved in dichloromethane (100 mL). Trifluoroacetic
acid (15.82 mL) was added dropwise and stirred at room temperature
overnight. Reaction mixture was concentrated to a minimal amount of
solvent in vacuo. To the residue were added ethyl acetate and then
saturated sodium bicarbonate. Product was extracted with ethyl
acetate (3 × 50 mL), dried over magnesium sulfate, and filtered, and
the solvent was removed under reduced pressure to afford the product
as a white foam (3.06 g, 85% yield). 1H NMR (400 MHz, CD3OD): δ
8.47 (s, 1H), 7.89 (d, J = 8.36 Hz, 2H), 7.57 (t, J = 6.49 Hz, 1H), 7.48
(m, 2H), 7.35 (s, 1H), 3.62−3.45 (m, 4H), 3.23−3.07 (m, 2H),
3.00−2.93 (m, 1H). LCMS (ESI+): m/z 345 [M + H]+.
8094
https://doi.org/10.1021/acs.jmedchem.1c00489
J. Med. Chem. 2021, 64, 8076−8100

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
N-[(4aR,7aR)-6-(5-Fluoropyrimidin-2-yl)-7a-isothiazol-5-yl-
room temperature. The reaction was diluted with EtOAc and
saturated aqueous NaHCO3 and transferred to a separatory funnel.
The layers were separated, and the aqueous layer was further
extracted three times with EtOAc. The combined organic layers were
dried over Na2SO4, filtered, and concentrated to give the crude
product that was purified via silica gel chromatography using a
0.5 to 10% MeOH in CH2Cl2 gradient to give 119 mg (66%) of the
desired product. 1H NMR (400 MHz, CDCl3): δ 9.78−9.75 (m, 1H),
8.41 (d, J = 2.7 Hz, 1H), 8.28 (dd, J = 4.6, 8.7 Hz, 1H), 8.19 (s, 2H),
7.95−7.91 (m, 1H), 7.58−7.53 (m, 1H), 7.46−7.44 (m, 1H), 7.07
(dd, J = 8.9, 11.9 Hz, 1H), 4.29 (d, J = 11.1 Hz, 1H), 3.89−3.85 (m,
1H), 3.82−3.79 (m, 2H), 3.34−3.32 (m, 1H), 3.10−3.06 (m, 1H),
2.94−2.90 (m, 1H). LCMS (ESI+): m/z 486 [M + H]+.
4,4a,5,7-tetrahydropyrrolo[3,4-d][1,3]thiazin-2-yl]benzamide (18i).
A slurry of N-[(4aR,7aR)-7a-isothiazol-5-yl-4a,5,6,7-tetrahydro-4H-
pyrrolo[3,4-d][1,3]thiazin-2-yl]benzamide (1.21 g, 3.513 mmol) and
5-fluoro-2-chloropyrimidine (2.33 g, 17.564 mmol) in 1,4-dioxane
(80 mL) was treated with diisopropylethylamine (1.84 mL, 10.138
mmol). The resulting suspension was heated to reflux under an
atmosphere of nitrogen for 3 h to afford a yellow solution, cooled to
ambient temperature, and diluted with water and ethyl acetate. The
aqueous phase with extracted with ethyl acetate. Combined organics
were dried over magnesium sulfate and filtered, and then the solvent
was removed in vacuo. The crude product was purified over 120 g of
silica gel, with a 20 min gradient of 0 to 50% EtOAc in
dichloromethane to afford the product (713 mg, 46% yield). 1H
NMR (400 MHz, CDCl3): δ 8.43 (m, 1H), 8.20 (s, 2H), 8.03−7.84
(m, 2H), 7.51 (t, J = 6.71 Hz, 1H), 7.42 (t, J = 7.59 Hz, 2H), 7.20−
7.10 (m, 1H), 4.22−4.12 (m, 1H), 4.03−3.88 (m, 3H), 3.25−3.17
(m, 1H), 3.06−2.90 (m, 2H). LCMS (ESI+): m/z 441 [M + H]+.
(4aR,7aR)-6-(5-Fluoropyrimidin-2-yl)-7a-isothiazol-5-yl-4,4a,5,7-
tetrahydropyrrolo[3,4-d][1,3]thiazin-2-amine hydrochloride (18).
5-Fluoro-N-[4-fluoro-3-[rel-(4aR,7aS)-2-amino-6-(5-fluoropyri-
midin-2-yl)-4,4a,5,7-tetrahydropyrrolo[3,4-d][1,3]thiazin-7a-yl]-
phenyl]pyridine-2-carboxamide hydrochloride (20).
N-[3-[(4aR,7aS)-2-Amino-6-(5-fluoropyrimidin-2-yl)-4,4a,5,7-
tetrahydropyrrolo[3,4-d][1,3]thiazin-7a-yl]-4-fluorophenyl]-3,5-di-
fluoropyridine-2-carboxamide (175 mg, 0.36 mmol) was dissolved in
dichloromethane (1 mL) and MeOH (0.2 mL). The solution was
treated with HCl (1 M in ether, 0.36 mL, 0.36 mmol), which
immediately formed a white precipitate. The mixture was stirred for 5
min and then concentrated to give the desired product (183 mg, 97%
To a solution of N-[(4aR,7aR)-6-(5-fluoropyrimidin-2-yl)-7a-isothia-
zol-5-yl-4,4a,5,7-tetrahydropyrrolo[3,4-d][1,3]thiazin-2-yl]benzamide
(713 mg, 1.619 mmol) in methanol (30 mL) was added lithium
hydroxide (340 mg, 8.094 mmol). The reaction was heated to reflux
for 5 h. The solvent was removed in vacuo, and the residue was
diluted with water and extracted 3× with ethyl acetate. The combined
organic layers were dried over sodium sulfate and filtered, and then
the solvent was removed in vacuo. The crude product was purified
over 40 g of silica gel with a 15 min gradient of 0 to 3% 7 N
NH3MeOH in CH2Cl2 gradient. The free base was obtained as a
white solid (471 mg, 86% yield). The HCl salt was prepared by
dissolving the free base in dichloromethane (10 mL), then 1 N HCl in
Et2O (1.5 mL) was added. Mixture was stirred for 5 min, and then the
solvent was removed under vacuum. The product was a white solid
(565 mg, 94% yield). 1H NMR (400 MHz, CD3OD): δ 8.52 (s, 1H),
8.34 (s, 2H), 7.43 (s, 1H), 5.45 (s, 1H), 4.36 (d, J = 11.88 Hz, 1H),
4.16 (d, J = 12.42 Hz, 1H), 4.04−3.96 (m, 1H), 3.83−3.76 (m, 1H),
3.43−3.30 (m, 2H). LCMS (ESI+): m/z 337 [M + H]+.
1
yield). H NMR (400 MHz, CD3OD): δ 8.55 (dd, J = 0.4, 2.9 Hz,
1H), 8.33−8.33 (m, 2H), 8.21−8.18 (m, 1H), 8.01−7.98 (m, 1H),
7.81−7.78 (m, 2H), 7.26 (dd, J = 8.9, 12.1 Hz, 1H), 4.43−4.40 (m,
1H), 4.15−4.10 (m, 1H), 3.88−3.85 (m, 1H), 3.77−3.74 (m, 1H),
3.54−3.52 (m, 1H), 3.44−3.40 (m, 2H). HRMS (ESI+): m/z calcd
20
for C22H18F3N7OS (M + H) 485.1246, found 485.1251. OR [α]D
−9.898 (c = 0.1, MeOH).
Synthesis of Compound 21. N-[3-[(4aR,7aS)-2-Amino-6-(5-
fluoropyrimidin-2-yl)-4,4a,5,7-tetrahydropyrrolo[3,4-d][1,3]thiazin-
7a-yl]-4-fluorophenyl]-3,5-difluoropyridine-2-carboxamide(21a).
Synthesis of Compound 20. N-[3-[(4aR,7aS)-2-Amino-6-(5-
fluoropyrimidin-2-yl)-4,4a,5,7-tetrahydropyrrolo[3,4-d][1,3]thiazin-
7a-yl]-4-fluorophenyl]-5-fluoropyridine-2-carboxamide (20a).
A solution of acetonitrile (5 mL) and DMF (80.7 μL) was treated
with oxalyl chloride (90 μL, 1.04 mmol), which caused bubbling.
After 1−2 min, 3,5-difluoropyridine-2-carboxylic acid (165 mg, 1.04
mmol) was added in a single portion to prepare the acid chloride. A
separate solution of (4aR,7aS)-7a-(5-amino-2-fluorophenyl)-6-(5-
fluoropyrimidin-2-yl)-4,4a,5,7-tetrahydropyrrolo[3,4-d][1,3]thiazin-2-
amine (200 mg, 497 μmol) in ethanol (2.5 mL) was heated to 50 °C.
To this solution while at 50 °C was added dropwise the acid chloride
solution. The reaction was stirred at 50 °C for 1 h and then cooled to
room temperature. The reaction was diluted with EtOAc and
saturated aqueous NaHCO3 and transferred to a separatory funnel.
The layers were separated, and the aqueous layer was further extracted
three times with EtOAc. The combined organic layers were dried over
Na2SO4, filtered, and concentrated to give the crude product that was
purified via silica gel chromatography using a 0.5 to 10% MeOH in
A solution of acetonitrile (3.7 mL) and dimethylformamide (60.5 μL,
782 μmol) was treated with oxalyl chloride (68 μL, 782 μmol) which
caused bubbling. After 1−2 min, 5-difluoropyridine-2-carbox-
ylic acid (110 mg, 782 μmol) was added in a single portion to
prepare the acid chloride. A separate solution of (4aR,7aS)-7a-(5-
amino-2-fluorophenyl)-6-(5-fluoropyrimidin-2-yl)-4,4a,5,7-
tetrahydropyrrolo[3,4-d][1,3]thiazin-2-amine (150 mg, 372 μmol) in
ethanol (1.9 mL) and water (1.9 mL) was heated to 50 °C.28 To this
solution while at 50 °C was added dropwise the acid chloride
solution. The reaction was stirred at 50 °C for 1 h and then cooled to
1
CH2Cl2 gradient to give the desired product. H NMR (400 MHz,
CDCl3): δ 9.59−9.56 (m, 1H), 8.32 (d, J = 2.2 Hz, 1H), 8.19 (d, J =
0.8 Hz, 2H), 7.99−7.96 (m, 1H), 7.40−7.35 (m, 2H), 7.09−7.04 (m,
1H), 4.28 (d, J = 11.1 Hz, 1H), 3.89−3.86 (m, 3H), 3.36−3.34 (m,
8095
https://doi.org/10.1021/acs.jmedchem.1c00489
J. Med. Chem. 2021, 64, 8076−8100

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
1H), 3.11−3.07 (m, 1H), 2.95−2.92 (m, 1H). LCMS (ESI+): m/z
Recombinant BACE1 and BACE2 Assays. Human BACE1 and
BACE2 enzyme assays were conducted as described previously,8,49
using purified recombinant human BACE1 or BACE2 as an enzyme
source and a synthetic FRET peptide (sequence: methylcourmarine
(MCA)-S-E-V-N-L-D-A-E-F-R-K(dinitrophenol)-R-R-R-R-NH2) as a
substrate.
504 [M + H]+.
N-[3-[(4aR,7aS)-2-Amino-6-(5-fluoropyrimidin-2-yl)-4,4a,5,7-
tetrahydropyrrolo[3,4-d][1,3]thiazin-7a-yl]-4-fluorophenyl]-3,5-di-
fluoropyridine-2-carboxamide hydrochloride (21).
Cathepsin D Selectivity Enzyme Assay. The Cathepsin D assay
was conducted with commercially available reagents using the
protocol described by May and colleagues.8,9 Human liver cathepsin
D was purchased from Calbiochem (catalog #219041). Porcine
gastric mucosa pepsin was purchased from Sigma (catalog #P-6887).
For all enzyme assays, a 10-point inhibition curve was plotted and
fitted with a four-parameter logistic equation to obtain the IC50 values.
Lysotracker Assay. The LysoTracker Green DND-26 assay was
conducted with a commercially available reagent purchased from
ThermoFisher Scientific (catalog #L7526). HepG2 cells were
purchased from ATCC and cultured according to recommended
procedures (catalog #HB-8065). HepG2 cells were treated with test
article over a 10-point concentration response curve for 24 h, and
fluorescent response max (Rmax) was measured using an Opera, high
content imaging platform (PerkinElmer).
Mouse Primary Neuronal Cultures. Primary cortical neurons were
prepared from PDAPP embryos (gestational day 16) and cultured in
96-well plates (15 × 104 cells/well in DMEM/F12 (1:1) containing
B27 supplement. Neurons were incubated at 37 °C for 24 h in the
presence/absence of inhibitors at the desired concentration. Assays
were run in triplicate. At the end of the incubation period,
conditioned media were analyzed for Aβ peptides using a sandwich
ELISA assay using monoclonal 2G3 as a capture antibody for Aβ1−40
and monoclonal 21F12 as a capture antibody for Aβ1−42. Both ELISAs
used biotinylated 3D6 as the reporting antibody. Treated PDAPP
neuronal cultures from the conditioned medium was diluted 1:5 for
Aβ1−42 and 1:10 for Aβ1−40 in buffer (PBS, 0.25% casein, 0.05%
Tween20, pH 7.4) and 100 μL aliquots assayed in triplicate.
Beagle Dog In Vivo Pharmacology. Young beagle dogs were used
to determine PK/PD effects of BACE inhibitors. In this model
established at MPI research (Mattawan, MI), beagle dogs were
implanted with a cannula in the lumbar spine region and threaded up
toward the cervical spine allowing for multiple CSF collections
throughout a single 48 h study period. Following an oral dose of
compound in 1% HEC, plasma samples from 5 to 6 male dogs were
taken at 0.25, 0.5, 1, 2, 3, 6, 9, 12, 24, and 48 h later. To assess central
pharmacology, CSF was also collected 0.5 h prior to dosing to
establish a baseline and then again at 3, 6, 9, 24, and 48 h postdosing.
Aβ1−x levels were measured in CSF and plasma using a standard
sandwich ELISA protocol (m266.2 and biotinylated 3D6 as the
capture and reporter antibodies, respectively), averaged across all
dogs, and then plotted as a function of time relative to baseline.
Duplicate plasma and CSF samples were used for bioanalytical
determination of compound concentrations. The use of animals in
these pharmacology studies conformed to the Institutional Animal
Care and Use guidelines for Eli Lilly and Company.
N-[3-[(4aR,7aS)-2-Amino-6-(5-fluoropyrimidin-2-yl)-4,4a,5,7-
tetrahydropyrrolo[3,4-d][1,3]thiazin-7a-yl]-4-fluorophenyl]-3,5-di-
fluoropyridine-2-carboxamide (100 mg, 0.20 mmol) was dissolved in
dichloromethane (5 mL). The solution was treated with HCl (1 M in
Et2O, 0.20 mL, 0.20 mmol), which immediately formed a white
precipitate. The mixture was stirred for 5 min and then concentrated
1
to give the desired product (101 mg, 0.19 mmol, 95% yield). H
NMR (400 MHz, DMSO-d6): δ 10.84 (d, J = 4.5 Hz, 2H), 8.64 (d, J
= 2.3 Hz, 1H), 8.51 (s, 2H), 8.18−8.13 (m, 1H), 7.94 (ddd, J = 8.9,
4.1, 2.7 Hz, 1H), 7.86−7.82 (m, 1H), 7.37−7.32 (m, 1H), 4.28−4.25
(m, 1H), 4.12−4.09 (m, 1H), 3.80−3.75 (m, 1H), 3.63−3.57 (m,
1H), 3.48−3.47 (m, 5H). HRMS (ESI+): m/z calcd for
C22H17F4N7OS (M + H) 503.1151, found 503.1158.
Synthesis of LY3202626 (Compound 22). N-[3-[(4aR,7aS)-2-
Amino-6-(5-fluoropyrimidin-2-yl)-4,4a,5,7-tetrahydropyrrolo[3,4-
d][1,3]thiazin-7a-yl]-4-fluorophenyl]-5-methoxypyrazine-2-carbox-
amide hydrochloride (22).
To a slurry of 5-methoxypyrazine-2-carboxylic acid (102 mg, 0.66
mmol) in acetonitrile (7 mL) was added dimethylformamide (51 μL,
0.66 mmol). The reaction was treated with oxalyl chloride (57 μL,
0.66 mmol), which resulted in gas evolution and the slurry becoming
a solution. The reaction was stirred at room temperature for 15 min to
form the acid chloride. In a separate flask, (4aR,7aS)-7a-(5-amino-2-
fluorophenyl)-6-(5-fluoropyrimidin-2-yl)-4,4a,5,7-tetrahydropyrrolo-
[3,4-d][1,3]thiazin-2-amine (150 mg, 0.41 mmol) was dissolved in
ethanol (4 mL) and water (4 mL). This solution was warmed to 50
°C and then treated with the acid chloride solution while remaining at
50 °C. The reaction was stirred at 50 °C for 1.5 h and then cooled to
room temperature. The reaction was poured into saturated aqueous
NaHCO3 (200 mL) and extracted with CH2Cl2 (3 × 100 mL). The
combined organic layers were washed with brine, dried over MgSO4,
filtered, and concentrated. The sample was purified via silica gel
chromatography using a 24 g silica gel column and eluting with a 0−
10% (7 N NH3 in MeOH)/CH2Cl2 gradient to give the free base of
the desired product (110 mg, 53% yield). The free base was dissolved
in CH2Cl2 (8 mL) and treated with HCl (4 M in dioxanes, 0.75 mL,
3.000 mmol), which caused a white precipitate to form. The solvent
was removed to give the desired product (105 mg, 0.196 mmol, 47%
yield). 1H NMR (400 MHz, DMSO-d6): δ 10.77 (s, 1H), 8.84 (d, J =
1.3 Hz, 1H), 8.51 (d, J = 1.0 Hz, 2H), 8.38 (d, J = 1.3 Hz, 1H), 8.02−
7.99 (m, 1H), 7.96−7.93 (m, 1H), 7.33 (dd, J = 9.0, 12.0 Hz, 1H),
5.38−5.37 (m, 2H), 4.23 (d, J = 12.1 Hz, 1H), 4.13−4.09 (m, 1H),
3.99 (s, 3H), 3.80−3.75 (m, 1H), 3.62−3.57 (m, 1H), 3.42−3.34 (m,
3H), 3.25−3.22 (m, 1H). HRMS (ESI+): m/z calcd for
Nonclinical Bioanalytical Methods for Compound Detection.
Brain samples were weighed, and a 3-fold volume of water/methanol
(4:1, v/v) was added prior to homogenization with an ultrasonic
tissue disrupter. Appropriate unlabeled internal standard was added to
each brain, CSF, or plasma sample, and supernatants were subjected
to LC-MS/MS analysis using an Applied Biosystems/MDS Sciex API
4000 (Foster City, CA) equipped with a TurboIonSpray interface and
operated in positive ion mode.
Plasma and Brain Protein Binding. Plasma (mouse, dog, human)
and brain (mouse) protein binding was determined in vitro using
equilibrium dialysis, as described previously elsewhere.50 Unbound
brain and plasma concentrations were derived by multiplying the total
measured concentrations by the fraction unbound (free) determined
in vitro.
In Vitro Inhibition of CYP2D6, CYP3A4, and CYP2C9. Test
compounds were solubilized in DMSO at 10 mM and then diluted
with acetonitrile to 1 mM, 10% DMSO. A 2 μL aliquot of the
compound was added to 148 μL human liver microsomes (Xenotech
20
C22H20F2N8O2S (M + H) 498.1398, found 498.1413. OR [α]D
−12.248 (c = 0.1, CH2Cl2).
8096
https://doi.org/10.1021/acs.jmedchem.1c00489
J. Med. Chem. 2021, 64, 8076−8100

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
Xtreme Pool 200;100 mM NaPO4, pH 7.4) and preincubated for 10
min at 37 °C. The reaction was then initiated with addition of 50 μL
of a solution containing NADPH, bufuralol, midazolam, and
diclofenac (100 mM NaPO4, pH 7.4). Final concentrations in the
incubation were 10 μM test compound, 10 μM bufuralol, 5 μM
midazolam, 10 μM diclofenac, 1 mM NADPH, and 0.05 mg/mL
human liver microsomes. Substrate concentrations were approx-
imately Km concentration for the specific enzyme reaction. Following
a 3 min incubation, a 50 μL aliquot was quenched with 50 μL stop
solution containing internal standards (acetonitrile with 0.12 μM 1-
hydroxybufuralol maleate-d9, 0.17 μM α-hydroxymidazolam-d4, and
0.19 μM [13C6]-4′-hydroxydiclofenac). The quenched reaction
mixture was centrifuged at approximately 4000 rpm for 15 min at
room temperature. Following centrifugation, the samples are
submitted for LC/MS/MS analysis. The concentrations of 1-
hydroxybufuralol, α-hydroxymidazolam, and 4′-hydroxydiclofenac
were then compared to those from a vehicle-only (no test compound)
control to assess inhibition of CYP2D6, CYP3A4, and CYP2C9,
respectively.
ASSOCIATED CONTENT
■
sı
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.jmedchem.1c00489.
LCMS data for compounds 6, 15−18, and 20−22
(PDF)
LY3202626 molecular string formulas (CSV)
AUTHOR INFORMATION
■
Corresponding Author
Dustin J. Mergott − Lilly Research Laboratories, A Division of
Eli Lilly and Company, Indianapolis, Indiana 46285, United
States; orcid.org/0000-0003-1802-0758; Phone: 317-
277-2261; Email: mergottdu@lilly.com
Authors
In Vitro Liver Microsome Stability. Test compounds were
solubilized in DMSO to 10 μM, diluted into 50/50 water/acetonitrile
solution to 200 μM, then diluted 1:10 into 100 mM NaPO4, pH 7.4,
to a final concentration of 20 μM. A 100 μL aliquot of liver
microsome solution (human, rat, dog, or mouse) was added to a PCR
plate at 37 °C The reaction was initiated by adding prewarmed test
compound and NADPH (100 mM NaPO4, pH 7.4). Final
concentrations in the incubation were 2 μM test compound, 0.5
μM NADPH, and 0.5 mg/mL liver microsomes. Immediately, a 20 μL
aliquot was removed and the reaction was quenched by adding (1) 90
μL of water/acetonitrile/methanol (50/25/25) with 0.5% (v/v)
acetic acid containing 100 nM labetalol, 20 nM imipramine, and 500
nM diclofenac as internal standards and (2) 40 μL of 100 mM
NaPO4, pH 7.4 buffer. This sample was used as time 0. Following a 30
min incubation, another 20 μL was removed and quenched in the
same way as the time 0 sample. The quenched reaction mixtures were
centrifuged at approximately 4000 rpm for 15 min at ambient
temperature. Following centrifugation, the concentrations of test
compound at time 0 and 30 min were determined by LC-MS/MS
analysis. The percent metabolism was calculated as the percent
decrease in test compound from time 0 to 30 min.
Bidirectional Transport across MDCK Cells. MDCKII cells stably
expressing human wild-type ABCB1 (P-gp) were obtained from The
Netherlands Cancer Institute (Amsterdam, The Netherlands).
MDCK cells were maintained, and the assay was conducted as
described previously.51,52 Transport was measured in both directions
across uninhibited and inhibited cell monolayers using a substrate
concentration of 5 μM diluted from a 10 mM DMSO stock solution
(final DMSO concentration of 0.05%) and a single 60 min time
interval. A 2.5 μM concentration of LSN335984 was used to
David L. McKinzie − Lilly Research Laboratories, A Division
of Eli Lilly and Company, Indianapolis, Indiana 46285,
United States; Present Address: Stark Neurosciences
Research Institute, Indiana University School of Medicine,
Indianapolis, IN 46202.
Leonard L. Winneroski − Lilly Research Laboratories, A
Division of Eli Lilly and Company, Indianapolis, Indiana
46285, United States
Steven J. Green − Lilly Research Laboratories, A Division of
Eli Lilly and Company, Indianapolis, Indiana 46285, United
States
Erik J. Hembre − Lilly Research Laboratories, A Division of
Eli Lilly and Company, Indianapolis, Indiana 46285, United
States
Jon A. Erickson − Lilly Research Laboratories, A Division of
Eli Lilly and Company, Indianapolis, Indiana 46285, United
States
Brian A. Willis − Lilly Research Laboratories, A Division of Eli
Lilly and Company, Indianapolis, Indiana 46285, United
States
Scott A. Monk − Lilly Research Laboratories, A Division of Eli
Lilly and Company, Indianapolis, Indiana 46285, United
States
Christopher D. Aluise − Lilly Research Laboratories, A
Division of Eli Lilly and Company, Indianapolis, Indiana
46285, United States
Thomas K. Baker − Lilly Research Laboratories, A Division of
Eli Lilly and Company, Indianapolis, Indiana 46285, United
States
Jose E. Lopez − Lilly Research Laboratories, A Division of Eli
Lilly and Company, Indianapolis, Indiana 46285, United
States
Jörg Hendle − Lilly Research Laboratories, A Division of Eli
Lilly and Company, San Diego, California 92121, United
States
James P. Beck − Lilly Research Laboratories, A Division of Eli
Lilly and Company, San Diego, California 92121, United
States
selectively inhibit Pgp. The apparent permeability coefficients (Papp
)
were estimated as the slope of the mass transported per 60 min
relative to the total recovered mass.53 The basal-to-apical (B−A)/
apical-to-basal (A−B) Papp ratios were calculated in the absence or
presence of inhibitor for net efflux ratio (NER). If NER is greater than
3.0, then the compound is designated as a Pgp substrate.
In Vivo Pharmacokinetic Studies. To determine clearance and
volume of distribution, compounds were administered intravenously
in Captisol 20% w/v NaPO4 buffer 25 mM, pH 2, at 1 mg/kg (1 mL/
kg) in dogs and 2 mg/kg (2 mL/kg) in rats. Serial blood samples were
collected following administration. The samples were treated with K3-
EDTA, and plasma was obtained by centrifugation at 1600g for 10
min and stored at −70 °C until analysis by LC-MS/MS.
Richard A. Brier − Lilly Research Laboratories, A Division of
Eli Lilly and Company, Indianapolis, Indiana 46285, United
States
∥Leonard N. Boggs − Lilly Research Laboratories, A Division
of Eli Lilly and Company, Indianapolis, Indiana 46285,
United States; Department of Biochemistry, University of
Utah, Salt Lake City, Utah 84112, United States; Lilly
Crystallographic Methods. Crystal structures of BACE bound to
compounds were solved using minor modifications of previously
described methods and diffraction data collected at Sector 31 of the
Advanced Photon Source.8,34
8097
https://doi.org/10.1021/acs.jmedchem.1c00489
J. Med. Chem. 2021, 64, 8076−8100

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
Research Laboratories, A Division of Eli Lilly and Company,
San Diego, California 92121, United States
by Argonne National Laboratory under Contract No. DE-
AC02-06CH11357. https://www.aps.anl.gov/Science/
Publications/Acknowledgment-Statement-for-Publications
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 and Company, which
operates the facility. http://lrlcat.lilly.com/
Anthony R. Borders − Lilly Research Laboratories, A Division
of Eli Lilly and Company, Indianapolis, Indiana 46285,
United States
∥Patrick J. Cocke − Lilly Research Laboratories, A Division of
Eli Lilly and Company, Indianapolis, Indiana 46285, United
States; Department of Biochemistry, University of Utah, Salt
Lake City, Utah 84112, United States; Lilly Research
Laboratories, A Division of Eli Lilly and Company, San
Diego, California 92121, United States
ABBREVIATIONS
■
Aβ, amyloid-β; BACE1, beta-site APP cleaving enzyme 1;
ACN, acetonitrile; AD, Alzheimer’s Disease; APP, amyloid
precursor protein; CatD, cathepsin D; Cl, clearance; CSF,
cerebral spinal fluid; CNS, central nervous system; CYP,
cytochrome P450 enzymes; DCM, dichloromethane; DIEA,
N,N-diisopropylethylamine; EtOAc, ethyl acetate; EtOH,
ethanol; Ghosez’s reagent, 1-chloro-N,N,2-trimethyl-1-prope-
nylamine; HCl, hydrochloric acid; hERG, human ether-a-go-go
related gene; Hep, hepatic; HPLC, high-performance liquid
chromatography; MeOH, methanol; MW, molecular weight;
NA, not available, meaning data were not collected for that
compound in that assay; NMP, N-methyl-2-pyrrolidone; PD,
pharmacodynamic; PDAPP mice, transgenic mice overexpress-
ing mutant human amyloid precursor protein V717F; Pgp, P-
glycoprotein; Piv-Cl, pivaloyl chloride; rt, room temperature;
PK, pharmacokinetic; PSA, polar surface area; QD dose, one a
day dosing; SAR, structure−activity relationship; TFA,
trifluoroacetic acid; THF, tetrahydrofuran
Pablo Garcia-Losada − Lilly Research Laboratories, A
Division of Eli Lilly and Company, Indianapolis, Indiana
46285, United States
Stephen L. Lowe − Lilly Research Laboratories, A Division of
Eli Lilly and Company, Indianapolis, Indiana 46285, United
States
Brian M. Mathes − Lilly Research Laboratories, A Division of
Eli Lilly and Company, Indianapolis, Indiana 46285, United
States
∥Patrick C. May − Lilly Research Laboratories, A Division of
Eli Lilly and Company, Indianapolis, Indiana 46285, United
States; Department of Biochemistry, University of Utah, Salt
Lake City, Utah 84112, United States; Lilly Research
Laboratories, A Division of Eli Lilly and Company, San
Diego, California 92121, United States
∥Warren J. Porter − Lilly Research Laboratories, A Division of
Eli Lilly and Company, Indianapolis, Indiana 46285, United
States; Department of Biochemistry, University of Utah, Salt
Lake City, Utah 84112, United States; Lilly Research
Laboratories, A Division of Eli Lilly and Company, San
Diego, California 92121, United States
REFERENCES
■
(1) Patterson, C. World Alzheimer Report 2018, https://www.alzint.
org/resource/world-alzheimer-report-2018/ (accessed 02-07-2021).
(2) Vassar, R.; Kuhn, P. H.; Haass, C.; Kennedy, M. E.; Rajendran,
L.; Wong, P. C.; Lichtenthaler, S. F. Function, therapeutic potential
and cell biology of BACE proteases: current status and future
prospects. J. Neurochem. 2014, 130 (1), 4−28.
(3) Masters, C. L.; Selkoe, D. J. Biochemistry of amyloid beta-
protein and amyloid deposits in Alzheimer disease. Cold Spring
Harbor Perspect. Med. 2012, 2 (6), a006262.
Stephanie L. Stout − Lilly Research Laboratories, A Division
of Eli Lilly and Company, Indianapolis, Indiana 46285,
United States
David E. Timm − Lilly Research Laboratories, A Division of
Eli Lilly and Company, Indianapolis, Indiana 46285, United
States; Present Address: Department of Biochemistry,
University of 1644 Utah, Salt Lake City, Utah 84112,
United States
Brian M. Watson − Lilly Research Laboratories, A Division of
Eli Lilly and Company, Indianapolis, Indiana 46285, United
States
Zhixiang Yang − Lilly Research Laboratories, A Division of Eli
Lilly and Company, Indianapolis, Indiana 46285, United
States
(4) De Strooper, B.; Vassar, R.; Golde, T. The secretases: enzymes
with therapeutic potential in Alzheimer disease. Nat. Rev. Neurol.
2010, 6 (2), 99−107.
(5) Hsiao, C. C.; Rombouts, F.; Gijsen, H. J. M. New evolutions in
the BACE1 inhibitor field from 2014 to 2018. Bioorg. Med. Chem. Lett.
2019, 29 (6), 761−777.
(6) Rafii, M. S.; Aisen, P. S. Alzheimer’s Disease Clinical Trials:
Moving Toward Successful Prevention. CNS Drugs 2019, 33 (2), 99−
106.
(7) Imbimbo, B. P.; Watling, M. Investigational BACE inhibitors for
the treatment of Alzheimer’s disease. Expert Opin. Invest. Drugs 2019,
28 (11), 967−975.
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.jmedchem.1c00489
(8) 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-beta in humans with
an orally available, non-peptidic beta-secretase inhibitor. J. Neurosci.
2011, 31 (46), 16507−16.
(9) 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.; Martenyi,
F.; Sheehan, S. M.; Stout, S. L.; Timm, D. E.; Vaught, G. M.; Watson,
B. M.; Winneroski, L. L.; Yang, Z.; Mergott, D. J. The potent BACE1
inhibitor LY2886721 elicits robust central Abeta pharmacodynamic
Author Contributions
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
Notes
The authors declare no competing financial interest.
∥Retired.
ACKNOWLEDGMENTS
■
This research was funded by Eli Lilly and Company. The
crystallographic research used resources of the Advanced
Photon Source, a U.S. Department of Energy (DOE) Office of
Science User Facility operated for the DOE Office of Science
8098
https://doi.org/10.1021/acs.jmedchem.1c00489
J. Med. Chem. 2021, 64, 8076−8100

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
responses in mice, dogs, and humans. J. Neurosci. 2015, 35 (3), 1199−
210.
(10) Lammert, C.; Einarsson, S.; Saha, C.; Niklasson, A.; Bjornsson,
E.; Chalasani, N. Relationship between daily dose of oral medications
and idiosyncratic drug-induced liver injury: search for signals.
Hepatology 2008, 47 (6), 2003−9.
(20) Ellis, C. R.; Tsai, C. C.; Lin, F. Y.; Shen, J. Conformational
dynamics of cathepsin D and binding to a small-molecule BACE1
inhibitor. J. Comput. Chem. 2017, 38 (15), 1260−1269.
(21) Gonzalez-Polo, R. A.; Boya, P.; Pauleau, A. L.; Jalil, A.;
Larochette, N.; Souquere, S.; Eskelinen, E. L.; Pierron, G.; Saftig, P.;
Kroemer, G. The apoptosis/autophagy paradox: autophagic vacuoli-
zation before apoptotic death. J. Cell Sci. 2005, 118, 3091−102.
(22) O’Neill, B. T.; Beck, E. M.; Butler, C. R.; Nolan, C. E.;
Gonzales, C.; Zhang, L.; Doran, S. D.; Lapham, K.; Buzon, L. M.;
Dutra, J. K.; Barreiro, G.; Hou, X.; Martinez-Alsina, L. A.; Rogers, B.
N.; Villalobos, A.; Murray, J. C.; Ogilvie, K.; LaChapelle, E. A.;
Chang, C.; Lanyon, L. F.; Steppan, C. M.; Robshaw, A.; Hales, K.;
Boucher, G. G.; Pandher, K.; Houle, C.; Ambroise, C. W.; Karanian,
D.; Riddell, D.; Bales, K. R.; Brodney, M. A. Design and Synthesis of
Clinical Candidate PF-06751979: A Potent, Brain Penetrant, beta-Site
Amyloid Precursor Protein Cleaving Enzyme 1 (BACE1) Inhibitor
Lacking Hypopigmentation. J. Med. Chem. 2018, 61 (10), 4476−
4504.
(11) Winneroski, L. L.; Erickson, J. A.; Green, S. J.; Lopez, J. E.;
Stout, S. L.; Porter, W. J.; Timm, D. E.; Audia, J. E.; Barberis, M.;
Beck, J. P.; Boggs, L. N.; Borders, A. R.; Boyer, R. D.; Brier, R. A.;
Hembre, E. J.; Hendle, J.; Garcia-Losada, P.; Minguez, J. M.; Mathes,
B. M.; May, P. C.; Monk, S. A.; Rankovic, Z.; Shi, Y.; Watson, B. M.;
Yang, Z.; Mergott, D. J. Preparation and biological evaluation of
BACE1 inhibitors: Leveraging trans-cyclopropyl moieties as ligand
efficient conformational constraints. Bioorg. Med. Chem. 2020, 28 (1),
115194.
(12) For an example of different SAR using a bicyclic
iminopyrimidinone which appeared after this work was completed,
see: 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 iminopyrimidinones
as beta amyloid cleaving enzyme-1 (BACE1) inhibitors: conforma-
tional constraint to favor a bioactive conformation. J. Med. Chem.
2012, 55 (21), 9331−45.
(13) For an example of different SAR using a bicyclic
iminopyrimidinone which appeared after this work was completed,
see: Mandal, M.; Wu, Y.; Misiaszek, J.; Li, G.; Buevich, A.; Caldwell, J.
P.; Liu, X.; Mazzola, R. D.; Orth, P.; Strickland, C.; Voigt, J.; Wang,
H.; Zhu, Z.; Chen, X.; Grzelak, M.; Hyde, L. A.; Kuvelkar, R.; Leach,
P. T.; Terracina, G.; Zhang, L.; Zhang, Q.; Michener, M. S.; Smith, B.;
Cox, K.; Grotz, D.; Favreau, L.; Mitra, K.; Kazakevich, I.; McKittrick,
B. A.; Greenlee, W.; Kennedy, M. E.; Parker, E. M.; Cumming, J. N.;
Stamford, A. W. Structure-Based Design of an Iminoheterocyclic beta-
Site Amyloid Precursor Protein Cleaving Enzyme (BACE) Inhibitor
that Lowers Central Abeta in Nonhuman Primates. J. Med. Chem.
2016, 59 (7), 3231−48.
(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. beta-
Secretase (BACE1) inhibitors with high in vivo efficacy suitable for
clinical evaluation in Alzheimer’s disease. J. Med. Chem. 2013, 56
(10), 3980−95.
(24) Rombouts, F. J.; Tresadern, G.; Delgado, O.; Martinez-
Lamenca, C.; Van Gool, M.; Garcia-Molina, A.; Alonso de Diego, S.
A.; Oehlrich, D.; Prokopcova, H.; Alonso, J. M.; Austin, N.; Borghys,
H.; Van Brandt, S.; Surkyn, M.; De Cleyn, M.; Vos, A.; Alexander, R.;
Macdonald, G.; Moechars, D.; Gijsen, H.; Trabanco, A. A. 1,4-
Oxazine beta-Secretase 1 (BACE1) Inhibitors: From Hit Generation
to Orally Bioavailable Brain Penetrant Leads. J. Med. Chem. 2015, 58
(20), 8216−35.
(25) Ploemen, J.-P. H.T.M.; Kelder, J.; Hafmans, T.; van de Sandt,
H.; van Burgsteden, J. A.; Salemink, P. J.M.; van Esch, E. Use of
physicochemical calculation of pKa and CLogP to predict
phospholipidosis-inducing potential: a case study with structurally
related piperazines. Exp. Toxicol. Pathol. 2004, 55 (5), 347−355.
(26) Dominguez, J. L.; Christopeit, T.; Villaverde, M. C.; Gossas, T.;
Otero, J. M.; Nystrom, S.; Baraznenok, V.; Lindstrom, E.; Danielson,
U. H.; Sussman, F. Effect of the protonation state of the titratable
residues on the inhibitor affinity to BACE-1. Biochemistry 2010, 49
(34), 7255−63.
(14) LEAN (ligand efficiency by atom number) = −log(IC50)/N,
where N is the number of heavy (non-hydrogen) atoms and IC50 is in
M.
(15) Fielden, M. R.; Werner, J.; Jamison, J. A.; Coppi, A.; Hickman,
D.; Dunn, R. T., 2nd; Trueblood, E.; Zhou, L.; Afshari, C. A.;
Lightfoot-Dunn, R. Retinal Toxicity Induced by a Novel beta-
secretase Inhibitor in the Sprague-Dawley Rat. Toxicol. Pathol. 2015,
43 (4), 581−92.
(27) Rombouts, F.; Kusakabe, K. I.; Hsiao, C. C.; Gijsen, H. J. M.
Small-molecule BACE1 inhibitors: a patent literature review (2011 to
2020). Expert Opin. Ther. Pat. 2021, 31 (1), 25−52.
(28) Garcia-Losada, P.; DeBaillie, A. C.; Eugenio de Diego, J.;
Green, S. J.; Hansen, M. M.; Jaramillo, C.; Johnson, M.; Kaoudi, T.;
Li, J.; Lindsay-Scott, P. J.; Mateos, C.; Mergott, D. J.; Rincon, J. A.;
Rothhaar, R. R.; Seibert, K. D.; Watson, B. M.; Winneroski, L. L.;
Gangula, S.; Jing, D.; Sun, H.; Zhang, L.; Frederick, M. O. Synthesis,
Optimization, and Large-Scale Preparation of the Low-Dose Central
Nervous System-Penetrant BACE1 Inhibitor LY3202626 via a [3 + 2]
Nitrone Cycloaddition. Org. Process Res. Dev. 2020, 24 (2), 306−314.
(29) Dubin, A. E.; Nasser, N.; Rohrbacher, J.; Hermans, A. N.;
Marrannes, R.; Grantham, C.; Van Rossem, K.; Cik, M.; Chaplan, S.
R.; Gallacher, D.; Xu, J.; Guia, A.; Byrne, N. G.; Mathes, C.
Identifying modulators of hERG channel activity using the
PatchXpress planar patch clamp. J. Biomol. Screening 2005, 10 (2),
168−81.
(30) Ly, J. Q.; Shyy, G.; Misner, D. L. Assessing hERG Channel
Inhibition Using PatchXpress. JALA 2005, 10 (4), 225−230.
(31) Patel, N. C. Methods to optimize CNS exposure of drug
candidates. Bioorg. Med. Chem. Lett. 2020, 30 (23), 127503.
(32) Willis, B. A.; Lowe, S. L.; Daugherty, L. L.; Dean, R. A.; English,
B.; Ereshefsky, L.; Gevorkyan, H.; James, D. E.; Jhee, S. S.; Lin, Q.;
Lo, A. C.; Mergott, D. J.; Monk, S. A.; Nakano, M.; Zimmer, J. A.;
Irizarry, M. C. Pharmacokinetics, pharmacodynamics, safety, and
(16) Zuhl, A. M.; Nolan, C. E.; Brodney, M. A.; Niessen, S.;
Atchison, K.; Houle, C.; Karanian, D. A.; Ambroise, C.; Brulet, J. W.;
Beck, E. M.; Doran, S. D.; O’Neill, B. T.; Am Ende, C. W.; Chang, C.;
Geoghegan, K. F.; West, G. M.; Judkins, J. C.; Hou, X.; Riddell, D. R.;
Johnson, D. S. Chemoproteomic profiling reveals that cathepsin D off-
target activity drives ocular toxicity of beta-secretase inhibitors. Nat.
Commun. 2016, 7, 13042.
(17) Koike, M.; Nakanishi, H.; Saftig, P.; Ezaki, J.; Isahara, K.;
Ohsawa, Y.; Schulz-Schaeffer, W.; Watanabe, T.; Waguri, S.;
Kametaka, S.; Shibata, M.; Yamamoto, K.; Kominami, E.; Peters,
C.; von Figura, K.; Uchiyama, Y. Cathepsin D deficiency induces
lysosomal storage with ceroid lipofuscin in mouse CNS neurons. J.
Neurosci. 2000, 20 (18), 6898−906.
(18) Steinfeld, R.; Reinhardt, K.; Schreiber, K.; Hillebrand, M.;
Kraetzner, R.; Bruck, W.; Saftig, P.; Gartner, J. Cathepsin D deficiency
is associated with a human neurodegenerative disorder. Am. J. Hum.
Genet. 2006, 78 (6), 988−98.
(19) Tyynela, J.; Sohar, I.; Sleat, D. E.; Gin, R. M.; Donnelly, R. J.;
Baumann, M.; Haltia, M.; Lobel, P. A mutation in the ovine cathepsin
D gene causes a congenital lysosomal storage disease with profound
neurodegeneration. EMBO J. 2000, 19 (12), 2786−92.
8099
https://doi.org/10.1021/acs.jmedchem.1c00489
J. Med. Chem. 2021, 64, 8076−8100

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Drug Annotation
tolerability of LY3202626, a novel Bace1 inhibitor, in healthy subjects
and patients with Alzheimer’s disease. Alzheimer's Dementia 2016, 12,
P418.
The AMARANTH and DAYBREAK-ALZ Randomized Clinical
Trials. JAMA Neurol 2020, 77 (2), 199−209.
(45) Yan, R. Physiological Functions of the beta-Site Amyloid
Precursor Protein Cleaving Enzyme 1 and 2. Front. Mol. Neurosci.
2017, 10, 97.
(33) Lo, A. C.; Evans, C. D.; Mancini, M.; Wang, H.; Shcherbinin,
S.; Lu, M.; Natanegara, F.; Willis, B. A. Phase 2 (NAVIGATE-AD
study) Results of LY3202626 Effects on Patients with Mild
Alzheimer’s Disease Dementia. J. Alzheimers Dis Rep 2021, 5, 321.
(34) Hong, L.; Koelsch, G.; Lin, X.; Wu, S.; Terzyan, S.; Ghosh, A.
K.; Zhang, X. C.; Tang, J. Structure of the Protease Domain of
Memapsin 2 (β-Secretase) Complexed with Inhibitor. Science 2000,
290 (5489), 150−153.
(35) Hussain, I.; Powell, D.; Howlett, D. R.; Tew, D. G.; Meek, T.
D.; Chapman, C.; Gloger, I. S.; Murphy, K. E.; Southan, C. D.; Ryan,
D. M.; Smith, T. S.; Simmons, D. L.; Walsh, F. S.; Dingwall, C.;
Christie, G. Identification of a novel aspartic protease (Asp 2) as beta-
secretase. Mol. Cell. Neurosci. 1999, 14 (6), 419−27.
(36) Sinha, S.; Anderson, J. P.; Barbour, R.; Basi, G. S.; Caccavello,
R.; Davis, D.; Doan, M.; Dovey, H. F.; Frigon, N.; Hong, J.; Jacobson-
Croak, K.; Jewett, N.; Keim, P.; Knops, J.; Lieberburg, I.; Power, M.;
Tan, H.; Tatsuno, G.; Tung, J.; Schenk, D.; Seubert, P.; Suomensaari,
S. M.; Wang, S.; Walker, D.; Zhao, J.; McConlogue, L.; John, V.
Purification and cloning of amyloid precursor protein beta-secretase
from human brain. Nature 1999, 402 (6761), 537−40.
(37) Vassar, R.; Bennett, B. D.; Babu-Khan, S.; Kahn, S.; Mendiaz, E.
A.; Denis, P.; Teplow, D. B.; Ross, S.; Amarante, P.; Loeloff, R.; Luo,
Y.; Fisher, S.; Fuller, J.; Edenson, S.; Lile, J.; Jarosinski, M. A.; Biere,
A. L.; Curran, E.; Burgess, T.; Louis, J. C.; Collins, F.; Treanor, J.;
Rogers, G.; Citron, M. Beta-secretase cleavage of Alzheimer’s amyloid
precursor protein by the transmembrane aspartic protease BACE.
Science 1999, 286 (5440), 735−41.
(46) 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 (7409), 96−9.
(47) Martiskainen, H.; Herukka, S. K.; Stancakova, A.; Paananen, J.;
Soininen, H.; Kuusisto, J.; Laakso, M.; Hiltunen, M. Decreased plasma
beta-amyloid in the Alzheimer’s disease APP A673T variant carriers.
Ann. Neurol. 2017, 82 (1), 128−132.
(48) Vassar, R. Editorial: Implications for BACE1 Inhibitor Clinical
Trials: Adult Conditional BACE1 Knockout Mice Exhibit Axonal
Organization Defects in the Hippocampus. J. Prev Alzheimers Dis
2019, 6 (2), 78−84.
(49) Yang, H. C.; Chai, X.; Mosior, M.; Kohn, W.; Boggs, L. N.;
Erickson, J. A.; McClure, D. B.; Yeh, W. K.; Zhang, L.; Gonzalez-
DeWhitt, P.; Mayer, J. P.; Martin, J. A.; Hu, J.; Chen, S. H.; Bueno, A.
B.; Little, S. P.; McCarthy, J. R.; May, P. C. Biochemical and kinetic
characterization of BACE1: investigation into the putative species-
specificity for beta- and beta’-cleavage sites by human and murine
BACE1. J. Neurochem. 2004, 91 (6), 1249−59.
(50) Zamek-Gliszczynski, M. J.; Ruterbories, K. J.; Ajamie, R. T.;
Wickremsinhe, E. R.; Pothuri, L.; Rao, M. V.; Basavanakatti, V. N.;
Pinjari, J.; Ramanathan, V. K.; Chaudhary, A. K. Validation of 96-well
equilibrium dialysis with non-radiolabeled drug for definitive
measurement of protein binding and application to clinical develop-
ment of highly-bound drugs. J. Pharm. Sci. 2011, 100 (6), 2498−507.
(51) Desai, P. V.; Sawada, G. A.; Watson, I. A.; Raub, T. J.
Integration of in silico and in vitro tools for scaffold optimization
during drug discovery: predicting P-glycoprotein efflux. Mol.
Pharmaceutics 2013, 10 (4), 1249−61.
(52) Zhao, R.; Raub, T. J.; Sawada, G. A.; Kasper, S. C.; Bacon, J. A.;
Bridges, A. S.; Pollack, G. M. Breast cancer resistance protein interacts
with various compounds in vitro, but plays a minor role in substrate
efflux at the blood-brain barrier. Drug Metab. Dispos. 2009, 37 (6),
1251−8.
(38) Yan, R.; Bienkowski, M. J.; Shuck, M. E.; Miao, H.; Tory, M.
C.; Pauley, A. M.; Brashler, J. R.; Stratman, N. C.; Mathews, W. R.;
Buhl, A. E.; Carter, D. B.; Tomasselli, A. G.; Parodi, L. A.; Heinrikson,
R. L.; Gurney, M. E. Membrane-anchored aspartyl protease with
Alzheimer’s disease beta-secretase activity. Nature 1999, 402 (6761),
533−537.
(39) Schneider, L. A resurrection of aducanumab for Alzheimer’s
disease. Lancet Neurol. 2020, 19 (2), 111−112.
(40) Hampel, H.; Vassar, R.; De Strooper, B.; Hardy, J.; Willem, M.;
̀
Singh, N.; Zhou, J.; Yan, R.; Vanmechelen, E.; De Vos, A.; Nistico, R.;
Corbo, M.; Imbimbo, B. P.; Streffer, J.; Voytyuk, I.; Timmers, M.;
Tahami Monfared, A. A.; Irizarry, M.; Albala, B.; Koyama, A.;
Watanabe, N.; Kimura, T.; Yarenis, L.; Lista, S.; Kramer, L.; Vergallo,
A. The β-Secretase BACE1 in Alzheimer’s Disease. Biol. Psychiatry
2021, 89, 745−756.
(53) Ho, N. F. H.; Raub, T. J.; Burton, P. S.; Barsuhn, C. L.
Quantitative Approaches to Delineate Passive Transport Mechanisms
in Cell Culture Monolayers. In Transport Processes in Pharmaceutical
Systems; Gordon, G. L., Lee, P. I., Topp, E. M., Eds.; Marcel Dekker:
New York, 2000; pp 219−316.
(41) Lopez Lopez, C.; Tariot, P. N.; Caputo, A.; Langbaum, J. B.;
Liu, F.; Riviere, M. E.; Langlois, C.; Rouzade-Dominguez, M. L.;
Zalesak, M.; Hendrix, S.; Thomas, R. G.; Viglietta, V.; Lenz, R.; Ryan,
J. M.; Graf, A.; Reiman, E. M. The Alzheimer’s Prevention Initiative
Generation Program: Study design of two randomized controlled
trials for individuals at risk for clinical onset of Alzheimer’s disease.
Alzheimers Dement (N Y) 2019, 5, 216−227.
(42) Egan, M. F.; Kost, J.; Voss, T.; Mukai, Y.; Aisen, P. S.;
Cummings, J. L.; Tariot, P. N.; Vellas, B.; van Dyck, C. H.; Boada, M.;
Zhang, Y.; Li, W.; Furtek, C.; Mahoney, E.; Harper Mozley, L.; Mo,
Y.; Sur, C.; Michelson, D. Randomized Trial of Verubecestat for
Prodromal Alzheimer’s Disease. N. Engl. J. Med. 2019, 380 (15),
1408−1420.
(43) Barao, S.; Moechars, D.; Lichtenthaler, S. F.; De Strooper, B.
BACE1 Physiological Functions May Limit Its Use as Therapeutic
Target for Alzheimer’s Disease. Trends Neurosci. 2016, 39 (3), 158−
169.
(44) Wessels, A. M.; Tariot, P. N.; Zimmer, J. A.; Selzler, K. J.;
Bragg, S. M.; Andersen, S. W.; Landry, J.; Krull, J. H.; Downing, A.
M.; Willis, B. A.; Shcherbinin, S.; Mullen, J.; Barker, P.; Schumi, J.;
Shering, C.; Matthews, B. R.; Stern, R. A.; Vellas, B.; Cohen, S.;
MacSweeney, E.; Boada, M.; Sims, J. R. Efficacy and Safety of
Lanabecestat for Treatment of Early and Mild Alzheimer Disease:
8100
https://doi.org/10.1021/acs.jmedchem.1c00489
J. Med. Chem. 2021, 64, 8076−8100