Development of an Efficient Manufacturing Process for E2212 toward Rapid Clinical Introduction

Article abstract of DOI:10.1021/acs.oprd.8b00444

Process studies of E2212 (1) toward rapid clinical introduction are described. Through comprehensive route-finding studies and optimization of key condensation and cyclization steps, a racemate-based manufacturing route was established and successfully scaled-up to the hundred kilogram scale. For the rapid delivery of a drug substance containing the Z isomer for preclinical safety studies, the successful scale-up of the photoisomerization of an olefin in a flow system is also presented.

Full text of DOI:10.1021/acs.oprd.8b00444

Article
pubs.acs.org/OPRD
Cite This: Org. Process Res. Dev. XXXX, XXX, XXX−XXX
Development of an Efficient Manufacturing Process for E2212
toward Rapid Clinical Introduction
Minetaka Isomura,*,‡ Taiju Nakamura,‡ Atsushi Kamada,† Takeo Sasaki,§ Toshiyuki Uemura,§
Yorihisa Hoshino,† Masaaki Matsuda,† Yongbo Hu,∥ Daiju Hasegawa,§ Kazato Inanaga,‡ Nobuaki Sato,§
Kazuhiro Yoshizawa,‡ George A. Moniz,∥ Gordon D. Wilkie,∥ Francis G. Fang,∥ Yoshihiro Nishikawa,‡
and Katsuya Tagami*,‡
†API Research Japan, Pharmaceutical Science & Technology, CFU, Medicine Development Center, Eisai Co. Ltd., 5-1-3-Tokodai,
Tsukuba-shi, Ibaraki 300-2635, Japan
‡API Research Japan, Pharmaceutical Science & Technology, CFU, Medicine Development Center, Eisai Co. Ltd., 22-Sunayama,
Kamisu-shi, Ibaraki 314-0255, Japan
§Neurology Tsukuba Research Department, Discovery, Medicine Creation, NBG, Eisai Co. Ltd., 5-1-3-Tokodai, Tsukuba-shi,
Ibaraki 300-2635, Japan
∥Integrated Chemistry, Eisai AiM Institute, 4 Corporate Drive, Andover, Massachusetts 01810, United States
S
* Supporting Information
ABSTRACT: Process studies of E2212 (1) toward rapid clinical introduction are described. Through comprehensive route-
finding studies and optimization of key condensation and cyclization steps, a racemate-based manufacturing route was
established and successfully scaled-up to the hundred kilogram scale. For the rapid delivery of a drug substance containing the Z
isomer for preclinical safety studies, the successful scale-up of the photoisomerization of an olefin in a flow system is also
presented.
KEYWORDS: E2212, Alzheimer’s disease, γ-secretase, photoisomerization, flow
An initial synthetic route applied by the medicinal chemistry
group is shown in Scheme 1.4 1 was prepared from its racemic
precursor ( )-5 by chiral chromatography followed by salt
formation with D-tartaric acid. Precursor ( )-5 was obtained
from bisamide ( )-4 via chlorination in POCl3 followed by
cyclization in the presence of NH4OAc. Bisamide ( )-4 was
synthesized by condensation between cyclic hydrazide ( )-2
and carboxylic acid 3·TFA.4
By means of this route, the initial batches of drug substance for
preclinical studies were supplied. However, there were several
drawbacks in terms of scalability, especially in the chlorination
and cyclization steps: (1) heating (100 °C) in a solvent volume
of hazardous POCl3, (2) cyclization at high temperature (140
°C), (3) the use of silica gel chromatography, and (4) resolution
by chiral HPLC. Faced with a limited timeline toward initial
clinical introduction, process studies of 1 were initiated.
INTRODUCTION
■
Alzheimer’s disease (AD) is the most common cause of
dementia, constituting a serious social problem that affects the
quality of life of the afflicted.1 It has been hypothesized that
senile plaque formed by aggregation of amyloid-β protein (Aβ)
in patients’ brains is a primary cause of AD.2 In turn, Aβ is
produced from amyloid precursor protein (APP) by the
function of β-secretase and γ-secretase. Therefore, intensive
R&D studies have been conducted worldwide to develop agents
aimed at lowering Aβ by modulation of β- and γ-secretase.
E2212 (1) (Figure 1) is an orally bioavailable γ-secretase
modulator that was developed as an agent for the treatment of
AD.3 To this end, a reliable and scalable manufacturing process
of 1 was required to support drug development activities,
including pharmacological and toxicological evaluation as well as
formulation development and clinical studies.
RESULTS AND DISCUSSION
■
Selection of the Manufacturing Route. To identify the
best manufacturing route for future development, comprehen-
sive route-finding studies were attempted. First, a chiral
synthetic route was pursued because of its obvious mass
throughput advantage over a racemic process requiring chiral
Special Issue: Japanese Society for Process Chemistry
Received: December 15, 2018
Figure 1. Structure of E2212 (1).
© XXXX American Chemical Society
A
DOI: 10.1021/acs.oprd.8b00444
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Scheme 1. Medicinal Route
Scheme 2. Summary of Chiral Approaches
resolution. Scheme 2 summarizes three routes (A−C)5,6 that
were deemed feasible through laboratory evaluation.
(S)-5 was found to be low in Routes B and C even though chiral
precursors with high optical purity were used (see the table in
Scheme 2). Racemization of the cyclic amidine intermediate
(S)-9 was strongly indicated, considering that racemization of
(S)-5 did not proceed under these conditions.8 On the other
hand, the Heck coupling between olefin 6 and bromide (S)-7
proceeded without racemization in moderate yield. Thus, of the
In each route, coupling precursors (6/3/10 as pyridine
moieties and (S)-7/(S)-8/(S)-11 as chiral moieties) could be
prepared from common intermediates, bromide 137 and
commercially available carboxylic acid 14, as shown in Scheme
3. Key coupling reactions between the two moieties proceeded
in each route. However, the optical purity of the isolated product
B
DOI: 10.1021/acs.oprd.8b00444
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Scheme 3. Syntheses of Coupling Precursors
Scheme 4. Chiral Resolution and Recovery Process
chiral approaches that were attempted, Route A was essentially
the only viable candidate.
could be obtained in a practical yield with excellent optical
purity.9 The salt of (S)-5 could be successfully exchanged to D-
tartaric acid to give 1 without racemization. Thus, a chiral-salt-
based classical resolution process could be established,
eliminating the need for HPLC.
Next, recovery of the undesired enantiomer from the mother
liquor was attempted. At the outset, it was anticipated that the
pKa of the chiral center would be sufficiently acidic for
racemization to occur under basic conditions because of the
In parallel with the chiral syntheses, a racemic approach via
intermediate ( )-5 was also explored (Scheme 4). First,
diastereomeric salt formation of ( )-5 was attempted to replace
chiral HPLC, which is not suitable for further scale-up. After
screening of various chiral acids, dipivaloyl-D-tartaric acid (D-
DPTA) was identified as the resolving agent, and the
corresponding salt of the desired enantiomer (S)-5·D-DPTA
C
DOI: 10.1021/acs.oprd.8b00444
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Scheme 5. Syntheses of ( )-5
adjacent triazole ring and trifluoromethylphenyl moiety. In the
event, racemization of (R)-5 easily proceeded in alkaline
solution with heating. During scale-up production, multiple
mother liquors containing undesired (R)-5·D-DPTA and 1 were
mixed and racemized by treatment with 5 N NaOH. After
extractive workup and recrystallization, ( )-5 could be
recovered in good yield and purity.
Encouraged by the promising performance of the chiral
resolution and recovery processes, further synthetic studies to
obtain ( )-5 were continued. Because of the scalability concerns
associated with the medicinal chemistry route, Routes D and E
were explored (Scheme 5).6,9
In both routes, coupling precursors (10/20 as pyridine
moieties and ( )-19/( )-11 as racemic moieties) were easily
prepared in one or two steps without silica gel chromatography
starting from bromide 13 and commercially available nitrile 22,
respectively (Scheme 5). In both routes, coupling and
cyclization via cyclic amidine intermediate ( )-9 proceeded
under mild, scalable conditions to give ( )-5 in moderate
overall yield.
Compared with the chiral approach (Route A), the coupling
precursors in the racemic routes can be prepared in fewer steps
without silica gel chromatography. In addition, considering that
the recovery process of the undesired enantiomer was
established (Scheme 4), the overall yield of the racemic
approach was expected to be comparable to that of chiral
approach. Thus, the racemic approach was selected for further
development. In a comparison of Routes D and E, the cost of
TMSOTf in Route E was relatively high, and there was also a
quality concern associated with the instability of the hydrazone
intermediate ( )-21. Therefore, Route D was selected as a
future development route.
Optimization of the Condensation and Cyclization
Phases. After Route D was selected as the manufacturing route
for further development, efforts next focused on understanding
and optimizing the key cyclocondensation step between
hydrazide 10 and imidate ( )-23.10 The reaction consists of
two phases: the condensation phase and the cyclization phase
(Scheme 6). Major side products that were isolated during
optimization studies are also shown in Scheme 6 along with their
estimated formation pathways.
Through significant experimentation focused on controlling
the reaction pathway to improve the yield, it was recognized that
the reaction pathway is highly dependent on the pH of the
reaction mixture. Thus, an appropriate choice of base was
essential to the success of this reaction, as was stepwise control
of the reaction temperature. Selected experimental results
showing effects of the base and the temperature are shown in
Table 1.
Oxadiazole ( )-29 is an impurity derived from the
condensation phase. This impurity is formed via intermediate
( )-26, which in turn is generated from ( )-24, the tetrahedral
adduct between hydrazide 10 and imidate ( )-23. The
formation of ( )-29 and its precursor ( )-26 increased
significantly when no base or less basic pyridine was employed
(entries 1 and 4, Table 1), whereas it was almost completely
suppressed when a more basic amine (imidazole or Et3N) was
employed (entries 2, 3, 5, and 6). This is presumed to be due to
elimination of the amino group of ( )-24 promoted by
protonation in the relatively acidic medium. In addition, the
conversion rate of the condensation phase was also strongly
D
DOI: 10.1021/acs.oprd.8b00444
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Scheme 6. Reaction Pathway and Major Impurities
Table 1. Effects of Base and Temperature on the Reaction Profiles
triazole
oxadiazole
desired
temp.
(°C)
time hydrazide
condensation
condensation
oxadiazole
a
precursor
desired
a
regioisomer
a
a
a
a
a
entry
base (equiv)
(h)
10
( )-25
( )-26
( )-29
( )-9
( )-5
( )-28
1
2
3
4
5
6
pyridine (6)
imidazole (6)
Et3N (6)
none
imidazole (3)
Et3N (3)
rt
rt
rt
60
60
60
27
27
27
37
37
36
17
3
91
20
45
82
67
61
6
13
9
8
0
0
8
0
0
3
0
0
42
2
1
5
24
2
12
0
1
12
1
2
29
11
0
0
0
2
16
4
2
0
a
Each ratio was determined by HPLC area % at 254 nm.
influenced by the amine base added. As shown in Table 1, the
condensation was quite slow when 3−6 equiv of Et3N was used
both at room temperature and 60 °C (entries 3 and 6), whereas
it proceeded most efficiently when the mixture was buffered with
the same amount of imidazole (entry 2).
Another major impurity observed in this reaction is
regioisomer ( )-28. This impurity is derived from the
cyclization phase through competing minor alkylation of
( )-25, presumably due to the difference in the inherent
reactivities of the two nitrogen atoms in ( )-25. The cyclization
phase is a slow reaction, thus requiring higher temperature to
E
DOI: 10.1021/acs.oprd.8b00444
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Table 2. Selected Results of Optimization Studies
condensation phase
cyclization phase
a
b
entry
equiv of ( )-23
base 1 (equiv)
temp 1 (°C)
base 2 (equiv)
temp 2 (°C)
HPLC area % at 254 nm/271 nm
1
2
3
4
5
6
1.0
1.0
1.0
1.0
1.2
1.2
imidazole (6)
imidazole (12)
imidazole (12)
imidazole (12)
NaOAc (12)
NaOAc (6)
30
30
50
0
rt
rt
none
none
none
none
Et3N (4)
Et3N (4)
30
30
35
35
35
35
73/−
85/−
80/−
86/90
−/94
−/94
a
b
Each reaction was carried out on a 0.2−1.3 mmol scale. HPLC area % at the end of the reaction. The detection wavelength was changed from
254 to 271 nm through the development studies.
Table 3. Results of Pilot-Plant Production
a
generation mass of 10·HCl (kg) equiv of imidate ( )-23 base 1 (equiv) temp 1 (°C) base 2 (equiv) temp 2 (°C) HPLC area % isolated yield (%)
first
second
21
132
1.0
1.2
imidazole (12)
NaOAc (7)
5−13
10−15
none
Et3N (4)
45
40−45
90
96
58
78
a
HPLC area % (271 nm) at the end of the reaction.
complete compared with the condensation phase. The
regioisomer significantly increased at higher temperature
regardless of the basicity of the amine used (entries 4−6),
whereas the formation was negligible at room temperature
(entries 1−3).
operationally possible during the condensation phase. Entry 5
shows that imidazole could be replaced with cheaper NaOAc as
a neutralizing base. In addition, slightly increasing the amount of
imidate ( )-23 was effective for complete consumption of
hydrazide 10·HCl. Finally, addition of Et3N was also effective for
accelerating the cyclization phase, thus minimizing the
formation of the regioisomer and other unknown impurities to
give the highest HPLC purity. This is in contrast to the results in
Table 1, entries 3 and 6, where the conversion in the
condensation phase was strongly inhibited in the presence of
Et3N. Finally, as shown in entry 6, a comparable result could be
obtained by reducing the amount of NaOAc to 6 equiv.
Along with these optimization studies, two scale-up
production runs were performed in the pilot plant (Table 3).
In the initial pilot batch, the conditions based on entry 4 in Table
2 were employed on a 21 kg scale to afford a reproducible HPLC
profile and improved isolated yield (58%) compared with the
original conditions (45%; Scheme 5). Subsequently, the
optimized conditions based on entry 6 in Table 2 were applied
to the second pilot batch on a 132 kg scale to give ( )-5 in an
excellent isolated yield (78%). These scale-up batches
On the basis of these findings, further optimization was
performed to identify the best balance among (1) suppression of
the oxadiazole formation, (2) adequate condensation rate, and
(3) suppression of the regioisomer. Table 2 shows selected
experimental results obtained from this optimization effort.
With an equimolar mixture of 10·HCl and ( )-23,10opti-
optimization studies were carried out, focusing on the HPLC
area % at the end of the reaction since it represented the reaction
yield well. As shown in entries 1 and 2 in Table 2, increasing the
amount of imidazole was effective in keeping the mixture neutral
enough to prevent the formation of side products as well as
securing good conversion during the condensation phase,
increasing the HPLC purity. As shown in entries 2−4, a
tendency for decreasing HPLC area % was observed at higher
temperature during the condensation phase. To minimize the
formation of impurities, the temperature was set as low as
F
DOI: 10.1021/acs.oprd.8b00444
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Scheme 7. Isomerization of the Olefin
Table 4. Batch Isomerization Experiments
a
entry
volumes
time
E:Z:others
1
2
3
4
1000
1000
10
10 min
2 h
10 min
2 h
54:40:6
38:23:39
87:2:11
27:7:66
10
a
Determined by HPLC area (254 nm).
Figure 2. Flow system for photoisomerization.
demonstrated that the process is highly reproducible and robust
under mild conditions without requiring special measures in the
pilot plant and can easily be applied to manufacturing on a >100
kg scale.
light during operations (Scheme 7). Therefore, supplying the
drug substance containing a small amount (target: 1%) of the Z
isomer 30 to preclinical safety studies was essential to qualify the
Z isomer before human studies. Faced with a compressed
timeline, preparative studies to obtain a sufficient quantity of the
Z isomer were initiated.
To obtain a sufficient amount of Z isomer for preclinical
evaluation, a photomediated isomerization of 1 was considered
since sufficient material was available. First, preliminary
isomerization conditions were explored on a small scale using
a commercially available photoreactor (Table 4). These studies
revealed that the efficiency of the isomerization was greatly
influenced by the solvent volume. Highly diluted (1000 volume)
conditions resulted in fast and clean conversion, in contrast to
concentrated (10 volume) conditions (entry 1 vs 3). In addition,
it was also found that degradation gradually proceeded with
extended irradiation time (entry 1 vs 2). To minimize toxicity
Thus, the rapid establishment of the racemate-based
manufacturing process along with the HPLC-free chiral-salt-
based resolution, the recovery process, and the optimization of
the key condensation and cyclization phases contributed to the
timely delivery of a good manufacturing practice (GMP) drug
substance that enabled rapid clinical introduction of E2212.
Photoisomerization of the Olefin Using a Flow
System. For the fastest introduction into clinical studies, a
rapid supply of the drug substance for preclinical safety studies
can be as important as establishing a GMP manufacturing
process for clinical studies. For E2212 (1), it had been confirmed
that E/Z isomerization of the internal olefin proceeded during
the formulation process, presumably as a result of exposure to
G
DOI: 10.1021/acs.oprd.8b00444
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
risks during the safety studies, it was important that
contamination of unknown degradation impurities in the Z
isomer be minimized. Therefore, a short reaction time under
high-dilution conditions was considered necessary.
On the basis of the required quantity of the drug substance for
the preclinical safety studies and the target level of the Z isomer
(1% at 271 nm), isomerization of 185 g of the E isomer was
required in 185 L (1000 volumes) of solution, which far
exceeded the size of the photoreactor available. In addition, on
such a large scale it was considered to be impossible to realize a
short reaction time since the overall light exposure decreases
with increasing volume.
Thus, attention was turned to a flow system. As has been
reported in the literature, a photoreaction that is difficult to scale
using a traditional batch reactor can be successfully scaled using
a flow system.11 Thus, a single-path flow system consisting of a
PTFE tube wound around a traditional photoreactor was
installed in the laboratory, as shown in Figure 2. From the inlet
of the tube, a solution of 1 in 1000 volumes of MeCN/water was
fed using an appropriate pump, and an isomerized E/Z mixture
was eluted from the outlet. Since the lamp generates significant
heat during the long irradiation time, a cooling medium was
circulated via a circulator to keep the temperature inside the
PTFE tube constant throughout the irradiation. By variation of
the flow rate and the tube length exposed to light, conversions
under various residence times were monitored (Table 5).
safety studies on time, which would have been impossible to
achieve using traditional batch conditions.
CONCLUSION
■
Process studies of E2212 (1) enabling rapid clinical introduction
have been presented. Comprehensive route-finding studies
coupled with the chiral-salt-based resolution and enantiomer
recovery processes led to the identification of an optimal
manufacturing process proceeding via racemate ( )-5. Through
careful selection of the base and the temperature for a key
cyclocondensation process, the various reaction pathways
leading to side products were suppressed, and the highly
selective formation of the desired racemate ( )-5 was achieved
in excellent isolated yield and quality control. The overall
process is mild, operationally simple, and robust and was
successfully scaled up to hundred kilogram scale production,
contributing to the rapid delivery of GMP drug substance. In
addition, the preparation of drug substance containing the Z
isomer was achieved for the preclinical safety studies. The
photomediated isomerization of the internal olefin was also
readily scaled up to a multihundred liter scale with excellent
reproducibility by applying a rapid (within 1 min) single-path
flow irradiation system. A sufficient quantity of the Z isomer was
rapidly prepared, which would have been difficult to achieve
using a traditional batch reaction system.
EXPERIMENTAL SECTION
■
Table 5. Photoisomerization Using a Flow System
General. All of the reagents and solvents were obtained from
commercial sources, except for the made-to-order bromide 13.
The lamp was a commercially available 450 W high-pressure UV
lamp (UM-452, Ushio) fitted with a cooling condenser (Pyrex).
The PFA tubing was obtained from commercial sources, and the
length and dimensions used for each experiment were as
indicated in the text. The experimental procedures and the
NMR data for each intermediate in the route-finding studies
(Routes A−E) can be found in refs 4, 5, 6, and 9. All of the
positive/negative-ion mode electrospray ionization mass spectra
were recorded on an LTQ Orbitrap XL mass spectrometer
(Thermo Fisher Scientific). With regard to the analytical data for
E2212 (1), the IR absorption spectrum was measured using an
FT/IR-620 spectrometer (JASCO) by the potassium bromide
disk method (KBr, IR grade, WAKO, Japan), and the 1H and 13C
NMR spectra were recorded on a 600 MHz NMR spectrometer
(Bruker AVANCE). For other compounds, the IR absorption
spectra were measured using a Spectrum Two FT-IR
spectrometer (PerkinElmer), and the NMR spectra were
recorded on a 600 MHz NMR spectrometer (Bruker AVANCE
III).
Condensation and Cyclization of 10·HCl and ( )-23
(Table 3, First-Generation Process). A reactor was charged
with imidate HCl salt ( )-23 (21.0 kg, 61.0 mol), imidazole
(49.6 kg, 728.6 mol), hydrazide dihydrochloride hydrate 10·
HCl (21.1 kg, 60.9 mol), and MeOH (83.2 kg). The mixture was
stirred at 5−14 °C for 24 h and then warmed to 45 °C and stirred
for an additional 17 h. Upon reaction completion, the mixture
was cooled to room temperature. Half of the reaction mixture
was drained and kept in another reactor while the remaining half
of the mixture was diluted with EtOAc (94.4 kg), 15% aqueous
NaCl (63.0 kg), and 5 N HCl (34.1 kg) while the internal
temperature was kept below 35 °C. The pH was confirmed to be
6.5−7, and the layers were separated. To the organic phase was
added 30% aqueous NaHSO3 (63.1 kg), and the solution was
length
(m)
diameter
(mm)
flow rate
(mL/min)
residence
time (s)
a
entry
E:Z:others
1
2
3
4
1
1
1
2.5
2.5
2.5
3.0
9
5
33
60
600
22
54:43:3
48:47:5
42:45:13
56:39:5
0.5
184
10
a
Determined by HPLC area % (254 nm).
With the small-scale flow system using a 1 m length of 2.5 mm
diameter PTFE, the isomerization proceeded completely in 10
min (entry 3, Table 5). Since other impurities were detected at
higher levels compared with the corresponding batch result
(entry 1, Table 4), the residence time was further shortened to
assess whether the other impurities could be suppressed while
still affording a sufficient E/Z ratio. As a result, much shorter
times (33−60 s) were found to be enough to obtain an
acceptable E/Z ratio with improved HPLC purities (entries 1
and 2). Having identified good conditions for the isomerization
with the small-scale flow system, the isomerization was next
scaled-up using a longer (10 m) PTFE tube with a wider
diameter (3.0 mm) to increase the volume that can be treated
within the same working time. As shown in entry 4, comparable
results were obtained on the larger scale with almost the same
residence time (22 s). Encouraged by the good reproducibility in
entry 4, we applied this scale-up setting to the isomerization of
∼185 L of solution of the E isomer (185 g) (see the
Experimental Section). The feeding and irradiation were
completed in 17 h (two working days), and a sufficient quantity
of the required Z isomer was obtained with the same quality as in
the small-scale flow experiment. It should be noted that the
continuous cooling of the reaction mixture was important and
could be achieved by using a cooling circulator to dissipate the
heat generated. Thus, drug substance containing the desired
level of the Z isomer was successfully delivered for preclinical
H
DOI: 10.1021/acs.oprd.8b00444
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
warmed to 60 °C for 0.5 h. The solution was cooled to 30 °C,
and the aqueous layer was separated. To the organic phase was
added cooled 5 N HCl (56.9 kg), and after the pH was
confirmed to be 0, the organic phase was separated. The aqueous
phase was added back into the reactor, and EtOAc (94.6 kg) and
MeOH (4.2 kg) were added. The resulting mixture was cooled
to 7 °C followed by the addition of cooled 5 N NaOH (75.6 kg)
while the internal temperature was maintained below 20 °C. The
pH was confirmed to be 6.0−6.5. The aqueous layer was
separated, and the organic layer was washed with 5% aqueous
NaHCO3 (53.6 kg) and MeOH (4.2 kg). The pH was confirmed
to be 8.5. The layers were separated, and the organic layer was
washed with 5% aqueous NaCl (32.8 kg) and MeOH (4.2 kg).
The aqueous layer was separated and washed with water (31.5
kg). The aqueous layer was separated, and MeOH (4.2 kg) was
added to give organic solution 1. The remaining half of the
reaction mixture kept in another reactor was processed with the
same process as described above to give organic solution 2.
Organic solutions 1 and 2 were combined, and the organic layer
was filtered through a Zeta Carbon 35SP CUNO filter. The filter
was rinsed with EtOAc (57.2 kg). The organic layer was
concentrated under reduced pressure at 35−40 °C. The residue
was azeotroped with i-PrOH (33.0 kg × 3). The residue was
redissolved in i-PrOH (66.2 kg) at 70−80 °C. A seed crystal
(50.4 g) was added at 57 °C. The mixture was cooled to −5 to 0
°C at a rate of 10 °C/h and held at −5 to 0 °C. The crystals were
filtered, rinsed with 1:1 i-PrOH/n-heptane (25.0 kg), and dried
under reduced pressure at 60 °C to give crystals of the racemate
( )-5 (19.0 kg, including 10.4% solvents and water, 58.3% yield
from 10·HCl) as a white powder. FTIR (cm−1, KBr) 3435, 2952,
1.99 (m, 1H), 1.77 (m, 1H); 13C NMR (150 MHz, CDCl3) δ
166.9, 164.6, 156.0, 148.3, 138.8, 137.3, 136.6, 136.1, 132.7,
132.0, 129.2, 128.5 (q, JC−F = 29.7 Hz), 127.9, 127.0, 126.3 (q,
JC−F = 5.6 Hz), 124.3 (q, JC−F = 273.9 Hz), 122.5, 121.5, 118.1,
115.7, 113.9, 54.1, 44.2, 38.0, 32.3, 30.3, 13.7; HRMS (ESI+)
calcd for C25H24ClF3N5O2([M + H]+) 518.1565, found
518.1573.
Characterization of Regioisomer ( )-28. FTIR (cm−1,
KBr) 3412, 2955, 1581, 1499, 1450, 1403, 1310, 1158, 1118,
1034, 961, 784; 1H NMR (600 MHz, CDCl3) δ 7.82 (d, J = 1.3
Hz, 1H), 7.72 (d, J = 7.8 Hz, 1H), 7.68 (d, J = 15.6 Hz, 1H), 7.54
(d, J = 7.6 Hz, 1H), 7.53 (d, J = 15.6 Hz, 1H), 7.47 (dd, J = 7.8,
7.6 Hz, 1H), 7.38 (dd, J = 7.8, 7.6 Hz, 1H), 7.10 (d, J = 7.8 Hz,
1H), 7.04 (d, J = 7.7 Hz, 1H), 6.98 (dd, J = 1.3, 1.0 Hz, 1H), 4.73
(dd, J = 9.9, 5.8 Hz, 1H), 4.29 (ddd, J = 11.9, 5.5, 4.4 Hz, 1H),
4.14 (ddd, J = 11.9, 9.9, 5.1 Hz, 1H), 4.10 (s, 3H), 2.44 (dddd, J
= 13.7, 6.0, 6.0, 2.6 Hz, 1H), 2.31 (d, J = 1.0 Hz, 3H), 2.29 (m,
1H), 2.13 (m, 1H), 1.91 (dddd, J = 13.6, 11.7, 8.7, 2.5 Hz, 1H);
13C NMR (150 MHz, CDCl3) δ 156.0, 154.0, 151.2, 149.6,
140.4, 138.7, 136.7, 132.5, 132.4, 132.1, 130.2, 128.5 (q, JC−F
=
29.7 Hz), 127.3, 126.2 (q, JC−F = 5.5 Hz), 124.5 (q, JC−F = 274.1
Hz), 121.7, 117.7, 115.8, 114.7, 55.8, 42.9, 37.0, 30.6, 21.6, 13.7;
HRMS (ESI+) calcd for C25H24F3N6O ([M + H]+) 481.1958,
found 481.1963.
Condensation and Cyclization of 10·HCl and ( )-23
(Table 3, Second-Generation Process). A reactor was
charged with hydrazide 10·HCl (132.0 kg, 362.4 mol) under a
nitrogen atmosphere, and precooled MeOH (941.7 kg) at 10 °C
was added. With stirring, NaOAc (208.1 kg, 2536.9 mol) and
imidate HCl salt ( )-23 (149.7 kg, 434.9 mol) were successively
added, and the reactor wall was washed with MeOH (104.7 kg).
After the mixture was stirred at 10−15 °C for 25 h, Et3N (146.7
kg, 1449.7 mol) was added, and the mixture was warmed to 45
°C. After 60 h of stirring at 40−45 °C, the jacket temperature
was set to 35 °C, and the mixture was diluted with i-PrOAc
(1151.9 kg) followed by 5% aqueous NaCl (1320.0 kg) at 30−
44 °C. After separation of the phases, MeOH (209.8 kg) was
added to the organic layer, followed by washing with 5%
aqueous NaCl (1319.8 kg) at 34−35 °C. To this mixture was
added activated carbon (6.6 kg) in i-PrOAc (57.3 kg), and the
mixture was stirred for 1 h at 32 °C. Celite (6.6 kg) in i-PrOAc
(57.9 kg) was added, and the mixture was filtered through Celite
(3.6 kg) followed by a rinse with i-PrOAc (95.5 kg). The filtered
solution was concentrated under reduced pressure and then
azeotroped with i-PrOH (207.2 kg × 3) at 16−45 °C (internal
temperature). To the resulting solid was added i-PrOH (286.5
kg), and the mixture was heated to 61 °C to completely dissolve
the solid. Then the mixture was cooled, and a seed crystal (30.1
g) was added at 55 °C. The mixture was cooled to 8 °C at a rate
of 5 °C/h and held at 8 °C. The crystals were filtered and rinsed
with an i-PrOH/n-heptane mixture (155.5 and 135.4 kg,
respectively). The crystals were dried under reduced pressure
at 60 °C to give crystals of the racemate ( )-5 (136.6 kg, 78.5%
yield from 10·HCl) as a white powder.
1
2867, 1571, 1505, 1454, 1315, 1163, 1112, 973, 814, 773; H
NMR (600 MHz, CDCl3) δ 7.99 (d, J = 1.2 Hz, 1H), 7.75 (d, J =
7.6 Hz, 1H), 7.67 (d, J = 15.8 Hz, 1H), 7.51 (dd, J = 7.6, 7.5 Hz,
1H), 7.49 (d, J = 7.8 Hz, 1H), 7.46 (d, J = 15.8 Hz, 1H), 7.42
(dd, J = 7.8, 7.5 Hz, 1H), 7.04 (d, J = 7.8 Hz, 1H), 6.97 (dd, J =
1.2, 0.8 Hz, 1H), 6.96 (d, J = 7.8 Hz, 1H), 4.74 (dd, J = 8.4, 6.0
Hz, 1H), 4.39 (br ddd, J = 13.0, 7.4, 5.5 Hz, 1H), 4.35 (ddd, J =
13.0, 8.9, 5.0 Hz, 1H), 4.06 (s, 3H), 2.45 (dddd, J = 13.2, 6.7, 6.7,
2.5 Hz, 1H), 2.32 (d, J = 0.8, 3H), 2.30 (m, 1H), 2.19 (m, 1H),
1.98 (dddd, J = 13.6, 11.0, 8.3, 2.8 Hz, 1H); 13C NMR (150
MHz, CDCl3) δ 160.7, 155.9, 155.0, 150.6, 140.3, 138.5, 136.7,
132.3, 132.1, 131.1, 130.1, 128.4 (q, JC−F = 29.8 Hz), 127.3,
126.3 (q, JC−F = 5.6 Hz), 124.5 (q, JC−F = 274.0 Hz), 122.4,
120.9, 116.6, 115.9, 53.8, 47.4, 37.8, 31.2, 21.3, 13.7; HRMS
(ESI+) calcd for C25H24F3N6O ([M + H]+) 481.1958, found
481.1959. HPLC conditions to monitor the reaction: XBridge-
Shield-RP18 (5 μm, 4.6 mm × 250 mm), 1.0 mL/min, oven
temperature = 40 °C, UV detection at 271 nm, mobile phase A =
900:100:1 v/v/w H2O/MeCN/AcONH4, mobile phase B =
100:900:1 v/v/w H2O/MeCN/AcONH4, gradient (time
(min)/B conc (%)) = 0/5 → 5/45 → 30/45 → 35/100 →
40/100 → 40.01/5 → 47/stop, relative retention time (RRT) of
hydrazide (10) = 0.29, RRT of intermediate ( )-25 = 0.39, RRT
of intermediate ( )-9 = 0.72.
Characterization of Oxadiazole ( )-29. FTIR (cm−1,
KBr) 3419, 2952, 1588, 1499, 1403, 1313, 1120, 1036, 964, 821,
770, 736; 1H NMR (600 MHz, CDCl3) δ 7.82 (d, J = 1.3 Hz,
1H), 7.74 (d, J = 7.9 Hz, 1H), 7.58 (d, J = 15.7 Hz, 1H), 7.58 (m,
1H), 7.58 (m, 1H), 7.53 (d, J = 7.7 Hz, 1H), 7.42 (br dd, J = 7.9,
7.9 Hz, 1H), 7.36 (d, J = 15.7 Hz, 1H), 7.03 (d, J = 7.7 Hz, 1H),
6.98 (dd, J = 1.2, 0.9 Hz, 1H), 4.67 (dd, J = 7.6, 7.6 Hz, 1H), 4.07
(s, 3H), 3.57 (ddd, J = 6.6, 6.3, 2.2 Hz, 2H), 2.54 (dddd, J = 13.6,
10.7, 8.3, 5.1 Hz, 1H), 2.30 (d, J = 0.9 Hz, 3H), 2.29 (m, 1H),
Formation of the Diastereomeric Salt (S)-5·D-DPTA
Using D-DPTA (Scheme 4). A reactor was charged D-DPTA
(14.0 kg, 44.1 mol), MeOH (27.4 kg), and EtOAc (72.9 kg)
under a nitrogen atmosphere, and the mixture was stirred and
heated to 50 °C. To this mixture was added a solution of
racemate ( )-5 (38.5 kg, 80.1 mol, dissolved in MeOH (31.9
kg) and EtOAc (85.1 kg) at 55−60 °C in advance) over 1.5 h at
51−53 °C. (After addition of the first 20% of solution, seed
crystals (5.1 g) were added, and then the remaining ∼80% of the
I
DOI: 10.1021/acs.oprd.8b00444
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
solution was added). After completion of the addition, the
reactor for racemate ( )-5 was rinsed with MeOH (4.6 kg) and
EtOAc (12.2 kg), and the rinses were added to the mixture.
Then the mixture was stirred at 47−51 °C for 2 h, cooled to 8 °C
at a rate of 10 °C/h, and held at 8 °C. The crystals were filtered
and rinsed with a mixture of EtOAc and MeOH (125.0 and 12.2
kg) to give wet crystals (55.1 kg) and mother liquor 1 (327.6 kg).
The wet crystals were charged to a reactor, and then MeOH (8.1
kg) and EtOAc (81.9 kg) were added. The mixture was then
stirring and heated to 50 °C. After 3.5 h of stirring at 50−51 °C,
the mixture was cooled to 8 °C at a rate of 5 °C/h and held at 8
°C. The crystals were filtered and rinsed with a mixture of EtOAc
and MeOH (70.1 and 6.9 kg, respectively) to give wet crystals
(45.8 kg) and mother liquor 2 (160.8 kg). The wet crystals were
dried under reduced pressure at 60 °C for 18 h to give the D-
DPTA salt (S)-5·D-DPTA (24.2 kg, 37.8% yield, 98.1% ee) as a
white solid. Mother liquors 1 and 2 containing the undesired
enantiomer (R)-5·D-DPTA were recycled according to the
procedure described below. FTIR (cm−1, KBr) 3676, 3426,
3191, 2975, 2484, 1725, 1476, 1311, 1160, 1062, 974, 779; 1H
NMR (600 MHz, CDCl3) δ 8.34 (br s, 1H), 7.73 (d, J = 7.4 Hz,
1H), 7.65 (d, J = 15.7 Hz, 1H), 7.57 (d, J = 7.9 Hz, 1H), 7.48
(dd, J = 7.6, 7.4 Hz, 1H), 7.43 (d, J = 15.7 Hz, 1H), 7.40 (dd, J =
7.8, 7.6 Hz, 1H), 7.00 (d, J = 7.8 Hz, 1H), 6.99 (dd, J = 1.3, 1.1
Hz, 1H), 6.98 (d, J = 7.9 Hz, 1H), 5.91 (br s, 2H, OH), 5.55 (s,
2H), 4.70 (dd, J = 8.7, 5.8 Hz, 1H), 4.35 (ddd, J = 13.0, 5.8, 5.4
Hz, 1H), 4.35 (ddd, J = 13.0, 8.7, 4.9 Hz, 1H), 4.03 (s, 3H), 2.46
(dddd, J = 13.4, 6.6, 6.6, 2.0 Hz, 1H), 2.32 (d, J = 1.1 Hz, 3H),
2.27 (m, 1H), 2.16 (m, 1H), 1.95 (dddd, J = 13.5, 11.1, 8.5, 2.7
Hz, 1H), 1.20 (s, 9H); 13C NMR (150 MHz, CDCl3) δ 177.3,
169.7, 160.3, 155.7, 155.1, 152.3, 140.0, 135.8, 135.2, 133.4,
132.2, 131.0, 130.0, 128.4 (q, JC−F = 29.9 Hz), 127.4, 126.4 (q,
container containing D-tartaric acid was rinsed with n-PrOH (6.4
kg), and the rinse was added to the mixture. Seed crystals (5.0 g)
were added to the mixture, and then the mixture was stirred for 1
h at 70−71 °C, cooled to 8 °C at a rate of 10 °C/h, and held at 8
°C. The crystals were filtered and rinsed with a mixture of n-
PrOH and i-PrOAc (10.3 and 22.3 kg) to give wet crystals (48.4
kg) and mother liquor 3 (148.0 kg). The wet crystals were dried
under reduced pressure at 60 °C for 19.5 h to give E2212 (1)
(18.0 kg, 92.5% yield) as a white solid. Mother liquor 3 were
recycled according to the procedure described below. FTIR
(cm−1, KBr) 3461, 3173, 2956, 1734, 1584, 1536, 1476, 1309,
1130, 835, 765, 752; 1H NMR (600 MHz, DMSO-d6) δ 7.91 (s,
1H), 7.78 (d, J = 8.4 Hz, 1H), 7.77 (br d, J = 8.4 Hz, 1H), 7.61
(br dd, J = 7.8, 7.8 Hz, 1H), 7.49 (br dd, J = 7.8, 7.8 Hz, 1H),
7.46 (d, J = 15.6 Hz, 1H), 7.32 (d, J = 15.6 Hz, 1H), 7.27 (d, J =
7.8 Hz, 1H), 7.25 (br d, J = 7.8 Hz, 1H), 7.22 (s, 1H), 4.51 (dd, J
= 9.0, 6.0 Hz, 1H), 4.29 (s, 3H), 4.28 (m, 2H), 3.98 (s, 3H), 2.29
(m, 1H), 2.14 (s, 3H), 2.16 (m, 2H), 1.96 (m, 1H); 13C NMR
(150 MHz, DMSO-d6) δ 173.3, 159.3, 155.4, 155.0, 150.1,
141.1, 137.1, 136.9, 133.6, 132.9, 131.0, 130.5, 127.6, 127.1 (q,
JC−F = 30 Hz), 125.8 (q, JC−F = 5.6 Hz), 124.7 (q, JC−F = 270
Hz), 122.2, 120.7, 117.2, 116.5, 72.3, 53.7, 47.0, 37.6, 30.7, 21.3,
13.6; HRMS (ESI+) calcd for C25H23F3N6O ([M + H]+)
481.1958, found 481.1953.
Recovery of Racemate ( )-5 (Scheme 4). A reactor was
charged with mother liquors 1 and 2 (assumed to be 49.9 mol in
total) and mother liquor 3 (assumed to be 2.1 mol) described
above under a nitrogen atmosphere, and the mixture was
concentrated and azeotroped with n-PrOH three times (38.3 kg,
38.2 kg, 38.3 kg) under reduced pressure at 8−36 °C (internal
temperature). To this mixture were added n-PrOH (38.3 kg)
and 5 N NaOH (140.8 kg) followed by heating to 90 °C. After
49 h of stirring at 85−87 °C, the mixture was cooled to 45 °C,
and the aqueous layer was discarded. To the organic layer were
added i-PrOAc (41.5 kg) and water (95.3 kg). After the aqueous
layer was discarded, the organic layer was washed with water
(95.3 kg), and the aqueous layer was discarded. The organic
layer was concentrated and azeotroped with i-PrOH (three
times with 37.4 kg and then four times with 18.7 kg) under
reduced pressure at 30−45 °C to give crude racemate ( )-5. To
this mixture was added i-PrOH (37.9 kg), followed by heating to
70 °C to dissolve the solid. The mixture was cooled to 8 °C at a
rate of 5 °C/h and held at 8 °C (seed crystals (16.6 g) were
added when the internal temperature fell below 55 °C). The
crystals were filtered and rinsed with a mixture of i-PrOH and n-
heptane (28.0 and 24.4 kg, respectively) precooled at 0 °C to
give wet crystals (20.7 kg). The wet crystals were dried under
reduced pressure at 60 °C for 24 h to give racemate ( )-5 (19.4
kg, 77.8% yield) as a pale-yellow solid. The reaction progress was
monitored using the chiral HPLC conditions described above.
Photoisomerization of Olefin 1 in a Flow System. A
solution of 1 (185.1 g, 262.3 mol) in MeCN (73.1 kg) and water
(92.0 kg) was fed through PFE tubing (3.0 mm i.d., 0.5 mm
thickness, 10 m length, wound around the photoreactor (Pyrex)
equipped with a UV lamp) at a flow rate of 182 mL/min. The
flow and irradiation were continued for 17 h with occasional
monitoring of the flow rate and E/Z ratio. The temperature of
the cooling medium was kept at −5 to −1 °C during flow (after 9
h, feeding of the solution of 1 was stopped, and 50% aqueous
MeCN was fed at 182 mL/min for 3 min to wash the flow system
and then turned off; the flow was resumed the following day).
The isomerized product eluted from outlet of the PFE tubing
was concentrated under reduced temperature below 42 °C
JC−F = 5.6 Hz), 124.5 (q, JC−F = 273.9 Hz), 123.0, 119.2, 117.1,
116.8, 71.4, 54.1, 50.7, 47.4, 38.7, 37.8, 31.1, 27.0, 21.3, 11.4;
HRMS (ESI+) calcd for C25H24F3N6O ([M + H]+) 481.1958,
found 481.1961, calcd for C14H21O8 ([M − H]−) 317.1242,
found 317.1227; Chiral HPLC (CHIRALPAK IB 5 μm, 4.6 mm
× 150 mm, 1.0 mL/min, mobile phase = 3:2 v/v n-hexane/
EtOH, oven temperature = 25 °C, RRT of the undesired
enantiomer (R)-5 = 0.69.
E2212 (1). A reactor was charged with D-DPTA salt (S)-5·D-
DPTA (22.0 kg, 27.5 mol), and EtOAc (198.4 kg) was added
under a nitrogen atmosphere, after which stirring was started. To
this mixture was added 5 N HCl (110.0 kg) at 16−17 °C with
cooling. After the phases were separated, i-PrOAc (172.6 kg)
and MeOH (17.4 kg) were added to the aqueous layer, which
was cooled to 14 °C. To this mixture, 5 N NaOH (160.1 kg) was
added at 14−26 °C to adjust the pH to 9.8 (the acceptable range
is 7−10) and the aqueous layer was discarded. To the organic
layer were added purified water (220.0 kg) and MeOH (8.7 kg),
and the aqueous layer was discarded. To the organic layer were
added purified water (110.0 kg) and MeOH (8.7 kg), and the
aqueous layer was discarded. After MeOH (17.4 kg) was added
to the organic layer, polish filtration of the resulting solution was
performed, followed by rinsing with i-PrOAc (19.1 kg). The
filtered solution was concentrated and azeotroped three times
with i-PrOAc (38.4 kg, 57.5 kg, 57.5 kg) under reduced pressure
at 23−53 °C (internal temperature) to give the free base (S)-5.
To this were added i-PrOAc (87.7 kg) and n-PrOH (6.4 kg), and
the mixture was heated to 73 °C to completely dissolve the solid.
To this mixture was added a solution of D-tartaric acid (7.3 kg,
48.6 mol, dissolved in n-PrOH (47.9 kg) at 52 °C in advance) at
68−73 °C over 1 h. After completion of the addition, the
J
DOI: 10.1021/acs.oprd.8b00444
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
and pharmacodynamics of the novel γ-secretase modulator, E2212, in
healthy human subjects. J. Clin. Pharmacol. 2014, 54, 528.
(4) Kimura, T.; Kitazawa, N.; Kaneko, T.; Sato, N.; Kawano, K.; Ito,
K.; Takaishi, M.; Sasaki, T.; Yoshida, Y.; Uemura, T.; Doko, T.;
Shinmyo, D.; Hasegawa, D.; Miyagawa, T.; Hagiwara, H. Polycyclic
Compound. WO2009028588, 2009.
(5) Kamada, A.; Sasaki, T.; Uemura, T. Method for Producing
Polycyclic Compound and Intermediate for Same. WO2010098332,
2010.
(6) Uemura, T.; Sasaki, T.; Hoshino, Y.; Isomura, M. Process for
Producing Tetrahydrotriazolopyridine Derivative. WO2010098496,
2010.
(7) Burnett, D. A.; Bursavich, M. G.; McRiner, A. J. Fused
Morpholinopyrimidines and Methods of Use Thereof.
WO2015109109, 2015.
(8) It was difficult to obtain optical purity data for cyclic amidine (S)-
8. Therefore, racemization from (S)-18 to (S)-8 cannot be excluded in
route B.
(9) Nakamura, T.; Matsuda, M.; Hu, Y.; Hasegawa, D.; Hoshino, Y.;
Inanaga, K.; Isomura, M.; Sato, N.; Yoshizawa, K.; Moniz, G. A.; Wilkie,
G. D.; Fang, F. G.; Nishikawa, Y. Process for Preparing Certain
Cinnamide Compounds. WO2010025197, 2010.
(10) During the process studies, the structure of the imidate was
changed from Me-imidate ( )-19 to Et-imidate ( )-23 because of its
improved stability. The salt form of hydrazide 10 was also changed from
free to dihydrochloride hydrate.
(11) Hook, B. D. A.; Dohle, W.; Hirst, P. R.; Pickworth, M.; Berry, M.
B.; Booker-Milburn, K. I. A Practical Flow Reactor for Continuous
Organic Photochemistry. J. Org. Chem. 2005, 70, 7558.
(internal temperature), and the residue was azeotroped with n-
PrOH twice (500 mL, 300 mL) to give an E/Z mixture of E2212
(196.0 g (containing residual n-PrOH), E:Z = 61.8:37.4 by UV
1
(271 nm), 1.3:1.0 by H NMR) as an orange oil. HPLC
conditions to monitor the isomerization conversion and E/Z
ratio: XBridge-Shield-RP18 (5 μm, 4.6 mm × 250 mm), 1.0 mL/
min, oven temperature = 40 °C, mobile phase A = 900:100:1 v/
v/w H2O/MeCN/AcONH4, mobile phase B = 100:900:1 v/v/
w H2O/MeCN/AcONH4, gradient (time (min)/B conc (%)) =
0/5 → 5/45 → 35/45 → 50/100 → 55/100 → 55.01/5 → 65/5
→ 65.01/stop, RRT of Z form = 0.73.
From this mixture, a small portion was purified by silica gel
column chromatography to give the Z isomer in free form. FTIR
(cm−1, KBr) 3416, 2952, 1586, 1500, 1487, 1313, 1161, 1114,
1036, 966, 858, 769; 1H NMR (600 MHz, CDCl3) δ 7.90 (d, J =
7.9 Hz, 1H), 7.72 (d, J = 1.2 Hz, 1H), 7.70 (d, J = 7.9 Hz, 1H),
7.45 (dd, J = 7.6, 7.4 Hz, 1H), 7.37 (dd, J = 7.7, 7.6 Hz, 1H), 7.30
(d, J = 7.9 Hz, 1H), 7.02 (d, J = 7.9 Hz, 1H), 6.93 (dd, J = 1.2, 1.0
Hz, 1H), 6.73 (d, J = 13.3 Hz, 1H), 6.64 (d, J = 13.3 Hz, 1H),
4.63 (dd, J = 9.3, 5.9 Hz, 1H), 4.34 (br ddd, J = 13.0, 5.6, 4.1 Hz,
1H), 4.28 (ddd, J = 13.0, 9.9, 4.9 Hz, 1H), 3.92 (s, 3H), 2.45
(dddd, J = 13.2, 6.5, 6.5, 2.6 Hz, 1H), 2.29 (d, J = 1.0 Hz, 3H),
2.28 (m, 1H), 2.15 (m, 1H), 1.94 (dddd, J = 12.9, 11.4, 8.3, 2.6
Hz, 1H); 13C NMR (150 MHz, CDCl3) δ 159.4, 155.2, 154.4,
150.8, 140.2, 138.3, 136.6, 133.2, 131.9 131.6, 129.9, 128.5 (q,
JC−F = 29.8 Hz), 127.2, 126.2 (q, JC−F = 5.6 Hz), 124.4 (q, JC−F =
274.0 Hz), 121.8, 120.1, 118.4, 116.0, 53.6, 47.3, 37.9, 31.0, 21.7,
13.6; HRMS (ESI+) calcd for C25H24F3N6O ([M + H]+)
481.1958, found 481.1960.
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.oprd.8b00444.
1H and 13C NMR, FT-IR, and HRMS characterization
data and representative HPLC data of key compounds
(( )-5, (S)-5·D-DPTA, 1, and the (Z)-olefin) and
impurities (regioisomer ( )-28, oxadiazole ( )-29)
(PDF)
AUTHOR INFORMATION
Corresponding Authors
*E-mail: m-isomura@hhc.eisai.co.jp.
*E-mail: k-tagami@hhc.eisai.co.jp.
■
ORCID
Minetaka Isomura: 0000-0002-7344-5129
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We are grateful to Hiroki Matsunaga for analytical support.
REFERENCES
■
(1) For a recent report on Alzheimer’s disease, see: Alzheimer’s
Association. 2018 Alzheimer’s disease facts and figures. Alzheimer's
Dementia 2018, 14, 367.
(2) For a recent review of the amyloid hypothesis, see: Selkoe, D. J.;
Hardy, J. The amyloid hypothesis of Alzheimer’s disease at 25 years.
EMBO Mol. Med. 2016, 8, 595.
(3) Yu, Y.; Logovinsky, V.; Schuck, E.; Kaplow, J.; Chang, M.-K.;
Miyagawa, T.; Wong, N.; Ferry, J. Safety, tolerability, pharmacokinetics,
K
DOI: 10.1021/acs.oprd.8b00444
Org. Process Res. Dev. XXXX, XXX, XXX−XXX