Optimization of 1,4-Oxazine β-Secretase 1 (BACE1) Inhibitors Toward a Clinical Candidate

Article abstract of DOI:10.1021/acs.jmedchem.8b00304

In previous studies, the introduction of electron withdrawing groups to 1,4-oxazine BACE1 inhibitors reduced the pK_a of the amidine group, resulting in compound 2 that showed excellent in vivo efficacy, lowering Aβ levels in brain and CSF. However, a suboptimal cardiovascular safety margin, based on QTc prolongation, prevented further progression. Further optimization resulted in the replacement of the 2-fluoro substituent by a CF₃-group, which reduced hERG inhibition. This has led to compound 3, with an improved cardiovascular safety margin and sufficiently safe in GLP toxicity studies to progress into clinical trials.

Full text of DOI:10.1021/acs.jmedchem.8b00304

Subscriber access provided by UNIVERSITY OF CONNECTICUT
Article
Optimization of 1,4-Oxazine #-Secretase 1
(BACE1) Inhibitors Towards a Clinical Candidate
Harrie J.M. Gijsen, Sergio A Alonso de Diego, Michel De Cleyn, Aránzazu García-Molina, Gregor
J. Macdonald, Carolina Martinez-Lamenca, Daniel Oehlrich, Hana Prokopcova, Frederik J.R.
Rombouts, Michel Surkyn, Andrés A. Trabanco, Sven Van Brandt, Dries Van den Bossche,
Michiel Van Gool, Nigel Austin, Herman Borghys, Deborah Dhuyvetter, and Diederik Moechars
J. Med. Chem., Just Accepted Manuscript • Publication Date (Web): 29 May 2018
Downloaded from http://pubs.acs.org on May 29, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 38
1
2
3
4
5
6
7
8
Optimization of 1,4ꢀOxazine βꢀSecretase 1 (BACE1) Inhibitors Towards a
Clinical Candidate
9
Harrie J. M. Gijsen,1,* Sergio A. Alonso de Diego,2 Michel De Cleyn,1 Aránzazu GarcíaꢀMolina,2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Gregor J. Macdonald,1 Carolina MartínezꢀLamenca,1 Daniel Oehlrich,1 Hana Prokopcova,1 Frederik J. R.
Rombouts,1 Michel Surkyn,1 Andrés A. Trabanco,2 Sven Van Brandt,1 Dries Van den Bossche,1 Michiel
Van Gool,2 Nigel Austin,3 Herman Borghys,3 Deborah Dhuyvetter,3 and Diederik Moechars.4
1Neuroscience Medicinal Chemistry, Janssen Research & Development, Janssen Pharmaceutica NV,
Turnhoutseweg 30, Bꢀ2340, Beerse, Belgium
2Neuroscience Medicinal Chemistry, Janssen Research & Development, Janssen−Cilag SA, C/Jarama
75A, 45007 Toledo, Spain
3Discovery Sciences, Janssen Research & Development, Janssen Pharmaceutica NV, Turnhoutseweg 30,
Bꢀ2340 Beerse, Belgium
4Neuroscience Biology, Janssen Research & Development, Janssen Pharmaceutica NV, Turnhoutseweg
30, Bꢀ2340 Beerse, Belgium
ABSTRACT: In previous studies, the introduction of electron withdrawing groups to 1,4ꢀoxazine
BACE1 inhibitors reduced the pKa of the amidine group, resulting in compound 2 that showed excellent
in vivo efficacy, lowering Aβ levels in brain and CSF. However, a suboptimal cardiovascular safety
margin, based on QTc prolongation, prevented further progression. Further optimization resulted in the
replacement of the 2ꢀfluoro substituent by a CF3ꢀgroup, which reduced hERG inhibition. This has led to
compound 3, with an improved cardiovascular safety margin and sufficiently safe in GLP toxicity
studies to progress into clinical trials.
INTRODUCTION
ACS Paragon Plus Environment

Page 3 of 38
Journal of Medicinal Chemistry
Alzheimer’s disease (AD) is the most common form of dementia, causing progressive decline of
cognitive function, behavioral changes, and eventually death, due to neuronal loss. It is affecting over
44 million people worldwide, and numbers are predicted to increase even further over the coming years,
creating a huge physical, social and economic burden for patients and their families. 1 The main
pathological hallmarks are intracellular neurofibrillary tangle and extracellular plaque deposition. These
are composed of hyperphosphorylated tau protein and amyloidꢀß (Aß) peptides, respectively. 2
Supported by genetic and histopathological data, the amyloidꢀß cascade hypothesis states that
disturbance of the equilibrium between the production and clearance of toxic Aß peptides causes the
onset of the disease.3 The first and rateꢀlimiting step in the generation of Aß peptides is the cleavage of
amyloid precursor protein (APP) by the ßꢀsecretase aspartyl protease 1 (BACE1). Therefore, inhibition
of BACE1 is one of the most compelling approaches in the treatment of AD,4 resulting in a highly
competitive landscape with several compounds progressing in clinical trials.5 In these trials, BACE1
inhibitors have been shown to efficiently lower Aβ levels, but so far have failed to slow the disease
progression in mildꢀtoꢀmoderate or even prodromal AD patients, highlighted by the recent termination
of the EPOCH and APECS trials with verubecestat.6 To achieve therapeutic efficacy with a BACE1
inhibitor, treatment may need to start well before the disease becomes symptomatic, in a preventive
approach.7 This will require highly safe compounds, hence our continued interest in novel chemotypes.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In a recent publication, we have described the design and optimization of a series of 1,4ꢀoxazine
BACE1 inhibitors.8 The decoration of the 1,4ꢀoxazine ring with electron withdrawing groups to reduce
the amidine pKa resulted in orally bioavailable, centrally active BACE1 inhibitors, capable of lowering
brain and cerebrospinal fluid (CSF) Aβ levels in mouse and dog, respectively.
From these efforts, the 2ꢀfluoroꢀ1,4ꢀoxazine 2 emerged as a very effective agent in lowering CSF Aβ
levels in dogs (Figure 1). The introduction of the 2ꢀfluoro group in 1 reduced the pKa of the amidine
basic moiety by over 1 log unit, resulting in reduced Pꢀgp efflux and a slight reduction in hERG affinity
(8.3 ꢀM binding for 2 vs 5 ꢀM binding for 1). This however did not result in an improvement in the
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 38
functional hERG channel patchꢀclamp assay, where both compounds showed a similar level of
1
2
3
inhibition of ~50% at 3 ꢀM concentration.
4
5
6
7
8
In a subsequent in vivo model using anesthetized guineaꢀpigs, 9 compound 2 doseꢀdependently
prolonged the QTcB interval starting at a plasma concentration (Cmax) of 475 ng/mL. The no observed
adverse effect level (NOAEL) was determined at 170 ng/mL, giving an insufficient safety margin over
the EC50 concentration of 20 ng/mL, as determined in dog.10 Although it is currently unknown how
much reduction in Aβ levels will be therapeutically beneficial, a 50% reduction in Aβ levels is the
minimum targeted in most ongoing clinical trials,5 and a relevant BACE clinical candidate should be
able to achieve this with sufficient safety margin.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. Different generations of 1,4ꢀoxazine BACE1 inhibitors, improving Pꢀgp efflux and the
cardiovascular safety profile.
In this paper, we describe the optimization of 2 toward molecules combining good in vivo efficacy with
a sufficient cardiovascular safety margin. This included further investigation of other electronꢀ
withdrawing substituents on the 1,4ꢀoxazine ring and more extensive variations on the distal
ACS Paragon Plus Environment

Page 5 of 38
Journal of Medicinal Chemistry
heteroaromatic ring (B). From this exploration, 3 has emerged as an oxazineꢀcontaining BACE1
1
2
3
inhibitor combining all the desired properties.
4
5
6
RESULTS AND DISCUSSION
7
8
9
Our initial studies had indicated an influence of the Bꢀring substitution on hERG binding and inhibition.
To further investigate this finding, a small set of close analogs of 2 containing the 2ꢀfluoroꢀ1,4ꢀoxazine
amidine warhead and the original cyanopyridyl Bꢀring containing orthoꢀsubstituents was prepared.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Analogs 4 and 5 indeed showed an improvement in hERG binding and in the more relevant patchclamp
functional assays. However, when dosed to mice, no brain levels were detected at 2 and 4 h for either of
the two compounds after an oral administration of a 30 mg/kg (4) or 10 mg/kg (5) dose. This can be
attributed to an increased efflux for 4 and poor bioavailability for 5. While the latter compound was
sufficiently stable when tested in mouse liver microsomes (mLM), a significant instability in mouse
hepatocytes was observed (55% turnover after 60 min of incubation at 1 µM), explaining the low
plasma levels in mice (38 ng/mL after 4 h).
Table 1. Profile of 2ꢀfluoroꢀ1,4ꢀoxazine analogs.a
2
4
5
BACE1 IC50 nM
hAβ42 cell IC50 nM
hERG IC50 µM
7.2
8.3
8
18.4
12.2
>10
10.7
15.5
>10
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 38
hERG PC (% inh. @ 3 µM)
hLM (% metabolized @ 15 min)
mLM (% metabolized @ 15 min)
mHepatocytes (% metabolized @ 60 min)
Papp +Elacridar (nm/s)
56%
0
24%
11
19%
1
1
2
3
4
5
6
7
8
2
19
0
0
n.d.
69.5
10.5
6.6
55
9
183
51
3.6
94.8
30.6
3.1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Papp ꢀElacridar (nm/s)
Papp ratio
a.
See supplementary information for assay details. The IC50 values for the enzymatic BACE1
and hAß42 cellular assays represent the mean values of at least two independent experiments.
n.d.: not determined
The challenges in finding an optimal profile via a suitable combination of the 2ꢀfluoroꢀ1,4ꢀoxazine
amidine warhead and a substituted Bꢀring, made us revisit alternatively substituted oxazine warheads
described in our previous paper.8 Matched pair analysis of a set of 2ꢀfluoroꢀ and 2ꢀCF3ꢀ1,4ꢀoxazine
analogs bearing a standard set of substituted Bꢀrings, indicated a reduced hERG inhibition potential for
the 2ꢀCF3ꢀ1,4ꢀoxazine analogs, despite a moderate increase in lipophilicity of 0.3 log units for the
trifluoromethylꢀsubstituted analogs (Table 2).11
Table 2: hERG Patch Clamp inhibition: Matched pair analysis of 2ꢀfluoroꢀ and 2ꢀCF3ꢀ1,4ꢀoxazine
analogs.
ACS Paragon Plus Environment

Page 7 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
5
6
7
10
11
12
13
hERG PC (% inh @ 3µM)
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
56 %
19 %
78 %
33 %
cLogP
1.62
1.51
2.83
1.82
3
8
9
10
hERG PC (% inh @ 3µM)
cLogP
22%
13 %
27 %
19 %
1.92
1.81
3.13
2.12
The synthetic route towards the synthesis of compounds 1, 2, and 4 ꢀ 7 has been described before.8 The
synthesis of CF2ꢀ and CF3ꢀsubstituted analogs is depicted in Schemes 1 and 2. The previously described
synthesis route towards the 2ꢀCF3ꢀ1,4ꢀoxazine derivatives involved the introduction of the CF3
substituent by addition of RuppertꢀPrakash reagent (TMSCF ) to diketo oxazine 11, and subsequent
3
removal of the hydroxyl group in 12 via chlorination to 13, followed by a zinc mediated reductive
dehalogenation towards 14 (Scheme 1). Thionation of the amide with P2S5, followed by aminolysis of
the thioamide, delivered the corresponding amidine derivative 15. This was next converted to the aniline
16, using a copperꢀcatalyzed reaction employing sodium azide, which was further elaborated to final
products 17 via amide formation with the corresponding carboxylic acids using the coupling reagent 4ꢀ
12
(4,6ꢀdimethoxyꢀ1,3,5ꢀtriazinꢀ2ꢀyl)ꢀ4ꢀmethylꢀmorpholinium chloride (DMTMM).
Recently, an
alternative route for the multigram synthesis of compounds with this warhead has been published.13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 38
Alternatively, Suzukiꢀcoupling reaction of bromoarene 15 with 5ꢀpyrimidinylboronic acid or Eꢀ2ꢀ(5ꢀ
cyanopyridinꢀ2ꢀyl)vinyl)boronic acid pinacol ester led to the analogues 18 and 19, respectively. Buchwaldꢀ
coupling between aniline 16 and 2ꢀbromoꢀ3ꢀcyanopyridine yielded the bisaryl aniline derivative 20.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1. Synthesis of 2ꢀCF3ꢀ1ꢀ4ꢀoxazinesa
O
O
CF3
CF3
Cl
O
CF3
O
O
OH
R
R
R
R
N
H
F
F
O
O
N
N
F
F
O
N
H
O
H
H
b
a
c
Br
Br
98%
41%
99%
Br
Br
13
11
12 (6R/6S 3:1)
14 (6R/6S 4:1)
O
CF3
O
CF3
O
R
R
CF3
R
d,e
R
f
g
R
R
F
F
F
H2N
H2N
H2N
N
N
N
54%
90%
12-62%
2 steps
Br
15
H2N
HN
R
O
16
17
h
85%
i
26%
j
11%
O
R
R
CF3
O
CF3
O
CF3
R
R
R
N
R
N
F
F
H2N
N
H2N
F
H2N
HN
N
N
NC
N
N
NC
18b
19
20
a Reagents: (a) TMSCF3, TBAT, THF, 0 °C to rt, 20 min; (b) SOCl2, CH2Cl2, 0 °C to rt; (c) Zn, AcOH, 100 °C,
20 min; (d) P2S5, THF, 70 °C, 2ꢀ4 h; (e) 7N NH3 in MeOH, 80 °C, 2 h; (f) NaN3, CuI, DMEDA, Na2CO3, DMSO,
ACS Paragon Plus Environment

Page 9 of 38
Journal of Medicinal Chemistry
110 °C 4h; (g) RCOOH, DMTMM, MeOH, 0 °Cꢀrt, 2ꢀ24 h; (h) 5ꢀpyrimidinylboronic acid, Pd(PPh3)4, NaHCO3
sat. aq. 1,4 dioxane, 80 °C, 2 h; (i) (E)ꢀ(2ꢀ(5ꢀcyanopyridinꢀ2ꢀyl)vinyl)boronic acid pinacol ester, Pd(PPh3)4,
NaHCO3 sat. aq. 1,4ꢀdioxane, 140 °C, 0.5 h, µw; (j) 2ꢀbromoꢀ3ꢀcyanopyridine, dppf, Pd2(dba)3, Cs2CO3, 1,4ꢀ
dioxane, 160 °C, 1 h, µw. b Compound 18 was prepared as a racemate, starting from racemic 15.
1
2
3
4
5
6
7
8
9
Over the course of the project, the transformation from 13 to 14 had been carried out numerous times
with significant amounts of elimination product 21 observed. (Scheme 2). After optimization, the ratio
14 to 21 was found to be de dependent on the zinc source and reaction conditions. Thus, the use of fresh
metallic zinc dust in acetic acid at 100 °C was selective for the reductive dehalogenation (14, Scheme 1),
whereas using zincꢀcopper couple in acetic acid at room temperature resulted in almost quantitative
formation of the elimination product 21 (Scheme 2).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 2. Synthesis of 2ꢀCF2ꢀ1,4ꢀoxazinesa
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 38
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a Reagents: (a) ZnꢀCu, AcOH, rt, 16 h; (b) P2S5, THF, 70 °C, 2ꢀ4 h; (c) 7N NH3 in MeOH, 80 °C, 2 h; (d) NaN3,
CuI, DMEDA, Na2CO3, DMSO, 110 °C; (e) 5ꢀmethoxyꢀ2ꢀpyrazinecarboxylic acid, DMTMM, MeOH, 0 °C, 2ꢀ6
h; (f) H2, Pd/C, thiophene, EtOAc, rt, 16 h.
The difluoroalkene intermediate 21 was also elaborated into the corresponding BACE1 inhibitor 24 in a
similar reaction sequence as described for 17. Moreover, hydrogenation of the double bond in 24
yielded the CHF2 substituted analog 25 as a single diastereomer. The stereochemistry of the CF2
substituent in 25 was confirmed via 1HꢀNOESY and 19Fꢀ1H NOESY experiments indicating a perfect
fit with MOEꢀcalculated distances. Profiling of these derivatives allowed a match pair comparison of
additional small changes in substitution pattern of the 1,4ꢀoxazines, and the results are shown in Table 3.
Table 3. profiles of F, CF3, CF2= and CHF2 analogs.a
ACS Paragon Plus Environment

Page 11 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
7
10
24
25
BACE1 IC50 nM
12.0
13.2
>10
29.5
27.7
>10
19
151.4
35.1
>10
n.d.
216
86.3
2.5
11.4
4.9
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
hAβ42 cell IC50 nM
hERG IC50 µM
>10
24
hERG PC (% inh. @ 3 µM) 33
Papp +Elacridar (nm/s)
Papp ꢀElacridar (nm/s)
Papp ratio
201
160
137
1.2
166
48
176
1.1
4.4
28
3.5
Fu,b (%)
2.0
1.7
4.7
Fu, plasma mouse (%)
pKa
15.5
7.8
14.6
7.9
27.9
8.4
7.8
Aβ42 reduction in mice (%) 39, 65, 62
after 1, 2 and 4 h at 10
23, 37, 29
6, 7
(2, 4 h)
11, 38, 40
mg/kg p.o.
Kp (4 h)
0.60 ± 0.10 1.81 ± 0.93 0.97 ± 0.15
0.09 ± 0.01 0.23 ± 0.12 0.11 ± 0.02
0.40 ± 0.05
0.07 ± 0.01
Kp u,u (4h)
a. See supplementary information for assay details. The IC50 values for the enzymatic BACE1 and
hAß42 cellular assays represent the mean values of at least two independent experiments.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 38
A 12ꢀfold decrease in potency was observed for the CF2ꢀolefinic amidine 24 when compared to the 2ꢀFꢀ
substituted oxazine 7 (BACE1 IC50 of 12 nM for 7 vs 151 nM for 24): Furthermore compound 24
showed only a modest Aβ42 reduction in mice at 10 mg/kg p.o. While the cellular activity for the CHF2ꢀ
analog 25 improved 3ꢀfold when compared to 7, an increased efflux ratio was also measured in the
LLCꢀMDR1 cell line, resulting in a relatively low total and free brain/plasma ratio in mouse (Kp and
Kp u,u, of 0.4 and 0.07, respectively). These observations are most likely pKaꢀdriven. More importantly,
the improvement on hERG patchclamp inhibition was most significant with the CF3ꢀanalog 10.
Together with the better combination of brain penetration and significant in vivo efficacy, this prompted
us to further explore the CF3ꢀsubstituted morpholine series. An overview of the various CF3ꢀsubstituted
analogs and their primary in vitro activity is given in Table 4.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Compounds 3 and 8ꢀ10 represent a set of standard variations of Bꢀrings, widely used in the BACE1
inhibitor field.5 They all showed good enzymatic and cellular potency, and minimal inhibition in the
hERG patch clamp assay (13 to 27% inhibition at 3 µM, Table 2). Compound 8 suffered from low
metabolic stability and the same low bioavailability as its 2ꢀfluoroꢀ1,4ꢀoxazine analog 5, and was
therefore not considered further. Analogs bearing substituents that extend deeper into the P3 pocket
from the heteroaromatic Bꢀring, such as 26 and 27, showed good potency, although not improved over 3
and 8ꢀ10. Polar substituents as the methoxyethoxy chain in 26, led to poor metabolic stability, whereas
compounds with an increasingly lipophilic substituent suffered from a relative loss in cellular potency,
as illustrated by going from 26 to 27 to 28. Interestingly, while oxygen linked extensions were well
tolerated in terms of enzymatic potency, nitrogen linked extensions (29 ꢀ 31) resulted in loss of activity,
as best illustrated by comparison of the trifluoro ethoxy substituted 27 with the almost inactive nitrogen
linked analog 29. To further investigate the influence of the P3 targeting moiety, the amide linker was
omitted or replaced. Compounds bearing a bisaryl group, as exemplified by 18, were significantly less
potent, most likely due to a suboptimal filling of the P3 pocket. Replacement of the amide linker by a
trans double bond (19), which should project the heteroaryl Bꢀring to a similar position as the amide
ACS Paragon Plus Environment

Page 13 of 38
Journal of Medicinal Chemistry
linker, also resulted in a considerable loss in potency, possibly due to the loss of the Hꢀbond interaction
between the amide NH and Gly230 in BACE1. Indeed, the use of a NH linker, able to make Hꢀbond
interaction with Gly230, resulted in reasonably potent analogs as exemplified by 20. The overall profile
of 20 was however suboptimal, with high metabolic instability in human and mouse liver microsomes
(36% and 100% metabolized, respectively, after 15 minutes incubation).
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 4. Inhibitory activity and metabolic stability of 2ꢀCF3ꢀ1,4ꢀoxazine analogs.a
compound
Structure
BACE1
IC50 nM
hAβ42
cell IC50 metabolized metabolized
nM @ 15 min) @ 15 min)
hLM (%
mLM (%
22.2
6.0
6.4
7.2
2
13
38
48
19
59
3
28.5
20.4
29.5
40.7
38
0
8
9
27.7
12.5
2
10
26
56
24.5
20.9
n.d.
n.d.
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 38
30.2
49.0
n.d.
n.d.
n.d.
n.d.
28
29
1
2
3
4
5
6
7
8
>10000
4500
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3020
457
603
115
n.d.
n.d.
n.d.
n.d.
30
31
18b
19
1862
2344
631
251
7
51
n.d.
n.d.
129
40.3
36
100
20
The data represent the mean values of at least two independent experiments. b Compound
a.
18 was tested as a racemate.
To select the best compound for further progression, compounds 3, 9 and 10 were tested for their human
hepatotoxicity potential, as drug induced liver injury led to the termination of the BACE1 inhibitor
LY2886721 after Phase 1 clinical trials. 14 The high content screening (HCS) multiparametric
ACS Paragon Plus Environment

Page 15 of 38
Journal of Medicinal Chemistry
cytotoxicity assay is based on the simultaneous measurement of 6 key cell health indicators associated
with nuclear morphology, plasma membrane integrity, mitochondrial function and cell proliferation. It
has been shown to be a sensitive test for the assessment of the hepatotoxic potential of chemical and
pharmaceutical compounds.15 Inꢀhouse validation of the HCS cytotoxicity assay based on a EC20 value
of 30µM for the lowest toxic concentration, demonstrated a 67% sensitivity (33% false negatives) and a
100% specificity (no false positives) for the identification of severely hepatotoxic compounds. While 30
µM is considerably above the expected human efficacy and exposure levels, this cutꢀoff value provides
an empirically validated method to select the analog with the lowest likelihood of hepatotoxicity. For 3,
9 and 10, the lowest toxic concentrations were determined to be 31.5 µM, 13.6 µM, and 15.6 µM,
respectively. This led to the selection of compound 3 as the preferred candidate to advance towards
further profiling and in vivo testing.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The metabolic stability was assessed more extensively across species and showed that compound 3 had
acceptable stability in microsomes, as reflected by the in vitro predicted hepatic clearance, using the
wellꢀstirred model (human, rat, dog, mouse: 9.8, 32.3, 15.9, 60.6 mL/min/kg, respectively).16 Also, low
plasma protein binding was observed (Fu human, dog, mouse, rat: 36%, 26%, 23%, 26%), combined
with a moderate free fraction in the brain (Fu,b rat: 3.0%). Compound 3 demonstrated a very weak
inhibition potential towards cytochrome P450, reflected in IC50 values >30µM for all tested isoforms
(1A2, 2C8, 2C9, 2C19 and 3A4) and no significant evidence of time dependent inhibition (TDI) against
3A4 up to 30 µM.
In terms of enzymatic activity, 3 was tenꢀfold more potent towards BACE1 than BACE2, inhibition of
which may lead to hair depigmentation.17 To further examine the selectivity profile of 3, the compound
was evaluated at Cerep for the inhibition of the aspartate proteases cathepsin D and E, as well as 50
other receptors, ion channels or transporter targets. All tested targets showed less than 50% inhibition at
a dose of 10µM, except for kappa (76%) and mu (91%) opiate receptors. The higher inhibition of the
latter two did however not result in any functional effect.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 38
Compound 3 was tested in dose response in our dog PK/PD model. In single and repeated dose studies,
doses up to 20 mg/kg were very well tolerated in dogs, with doseꢀrelated decreases in Aβ42 in CSF. The
decreases lasted up to 49 h after dosing, corresponding to a halfꢀlife of 12 h for this compound.
1
2
3
4
5
6
7
8
9
In Figure 2, the effect on Aβ42 levels in CSF and corresponding compound levels are shown after 1 and
6 days of treatment. The compound was well tolerated, without observations of changes in behavior and
appearance. The doseꢀrelated decrease in CSF Aβ peptides was slightly more pronounced on day 6
compared to day 1. The data of all dog PK/PD experiments were used to determine the plasma EC50 to
lower Aβ42 in dog CSF for 3. This resulted in a value of 105 ng/mL, or 249 nM total plasma
concentration.10 This corresponds to a free concentration in dog of 65 nM, which is 3ꢀfold above the in
vitro IC50 (22 nM).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Beagle dog plasma levels of 3 and effect on CSF Aβ42 levels on day 1 and 6, after repeated
oral dosing, (fasted state, once daily dosing). (Avg + SEM, n=6/dose group) (MS: MesoScale).
The pharmacokinetic profile and oral bioavailability of 3 were investigated in mouse, rat and dog
following intravenous and oral administration (Table 5).
ACS Paragon Plus Environment

Page 17 of 38
Journal of Medicinal Chemistry
Table 5. Pharmacokinetic parameters for 3 (single dose; n=3)
1
2
3
4
5
6
7
8
9
parameter
Mouse Rat
Dog
1/10
7.7
Humana
Dose (mg/ kg) iv/po
CL (mL/min/kg) iv
LM CL (ml/min/kg)
Fu, plasma (%)
Vdss (L/kg) iv
LBF (%) iv
2.5/10 2.5/10
69
47
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
60.6
23
32.3
26
15.9
26
9.8
36
7.6
~50
1.4
333
1
11
7.6
11.5
~25
25
~67
3.9
160
2
~25
12
T1/2 (h) iv
Cmax (nM) po
Tmax (h) po
1,092
2
AUC0ꢀinf (ng.h/mL) po
F% po
1,334
45
1,220
36
15,433
71
54
a.
In vivo Human PK data are predicted values derived by allometric scaling.
The clearance data (CL) are in line with the in vitro predictions: while the systemic clearance in mouse
and rat was moderate to moderately high, it was low in dog. The volume of distribution (Vdss) on the
other hand, was large across the species, indicating there is significant tissue distribution, which results
in a halfꢀlife (T1/2) of 12h for the dog. Following oral dosing, the compound was rapidly absorbed in
mouse, but slowly in rat and dog. Given the high systemic exposure (Cmax and AUC), the bioavailability
was acceptable across species and the balance between clearance (% liver blood flow (LBF)) and
bioavailability indicated that absorption was complete.
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
These PK parameters were used for interspecies allometric scaling, resulting in a predicted low
clearance in human (25% LBF), with a large volume of distribution (11.5 L/kg) and a long halfꢀlife (25
h). Assuming an estimated bioavailability of 54% in human (based on the average bioavailability in
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 38
preclinical species), a once daily dose of 150 mg was predicted to maintain plasma concentrations of
100 ng/mL, required for achieving a sustained reduction of at least 50% of CSF Aßꢀlevels and building
a maximum concentration of 175 ng/mL at this repeated dose.
1
2
3
4
5
6
7
8
9
In the hERG potassium ion channel binding assay, compound 3 showed less than 50% binding up to the
highest dose (10 ꢁM ) tested (IC50 > 10 ꢁM). This correlated with a relatively low functional hERG
inhibition in the hERG patch clamp assay (22% inhibition at 3 ꢁM). To assess the cardiovascular safety
more profoundly, the compound was subjected to a pharmacology study in anesthetized female guinea
pigs. In this model, prolongation of the QT interval was measured after administration of 3 by i.v.
infusion in pentobarbital anesthetized animals (cumulatively dosed up to 10 mg/kg i.v.). At the topꢀdose,
at an exposure of 11.500 ng/mL, a slight QTcB prolongation was observed (+9% versus +7% with vehicle).
With a free plasma concentration in guinea pig determined at 37%, this corresponds to a free
concentration of 10.1 ꢁM, confirming the lack of significant findings in the hERG in vitro assays below
10 ꢁM. No relevant effect on QTcB was observed at an exposure of 5,235 ng/mL (4.60 ꢁM free), hence
the reduced hERG inhibition was translated into an improved No Observed Adverse Effect Level
(NOAEL). Based on these data, 3 has a potential to prolong the QTc interval only at exposures over 65ꢀ
fold the predicted human Cmax (175 ng/mL, 0.15 ꢁM free) , or a safety margin for QTc prolongation of
30ꢀfold when using total concentrations and 31ꢀfold when using free concentrations. The key data and
margins are tabulated in Table 6 together with the available data for 2. To assess the improvement in
CV safety profile of 3 over 2, we compared safety margins based on dog efficacy levels, since human
PK and dose prediction have not been performed for 2. Typically, a safety margin of minimally 30ꢀfold
is considered optimal. This margin is met for 3, irrespective of the method of calculation. Based on the
dog EC50 data, a 6ꢀ to 8ꢀfold margin improvement has been achieved for 3 over 2.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 6. Comparison of cardiovascular safety data and margins for 2 and 3
ACS Paragon Plus Environment

Page 19 of 38
Journal of Medicinal Chemistry
2
3
22.2 nM
6.0 nM
1
2
3
4
5
6
7
8
BACE1 IC50
7.2 nM
8.3 nM
8.3 µM
56 %
hAβ42 cell IC50
hERG binding IC50
hERG PC (% inh @ 3 µM)
>10 µM
9
22 %
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
Fu, plasma dog, guinea pig, human
LOAEL for QTcB prolongation
42%, 41%, 43%
26%, 37%, 36%
Total conc
Free conc
475 ng/mL
0.55 µM
11500 ng/mL
10.1 µM
NOAEL for QTcB prolongation
Total conc
Free conc
170 ng/mL
0.20 µM
5235 ng/mL
4.60 µM
Human Cmax (predicted for 50% Aβ
reduction in CSF)
Total conc.
Free conc.
n.d.
n.d.
175 ng/mL
0.15 µM
EC50 in dog (50% Aβ reduction in CSF
Total conc
20 ng/mL
0.023 µM
105 ng/mL
0.065 µM
Free conc
Safety margin: NOAEL vs human Cmax
Total conc.
n.d.
n.d.
30ꢀfold
31ꢀfold
Free Conc.
Safety margin: NOAEL vs dog EC50
Total conc
8.5ꢀfold
8.7ꢀfold
50ꢀfold
71ꢀfold
Free conc
56
57
58
59
60
.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 38
In terms of mutagenic potential, 3, as well as its main metabolites, was clean in vitro both in
nonmammalian cell systems (AMES II) 18 as in mammalian cell systems (micronucleus test, MNT).
Also, no reactive metabolite formation was observed using glutathioneꢀ or CNꢀtrapping assays. The
aniline 16, a major metabolite, was tested separately in both AMES II and MNT, and also found to be
clean.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The tissue distribution of 3 was investigated in male Sprague Dawley rat following p.o. administration
of a nominal dose of 10 mg/kg. Concentrations were particularly high in kidney, liver and lung,
compared to plasma, with ratios of 19, 31 and 169, respectively. Given these relatively high tissue
distribution levels in rat, 3 was evaluated for its potential to induce phospholipidosis in human THPꢀ1
monocytes, using a fluorescence assay.19 Concentrations up to 200µM were tested, where the compound
induced a 2ꢀfold increase in fluorescence at 21 µM, indicating a weak potential for phospholipidosis.
Compound 3 was subsequently progressed into 1ꢀmonth repeated dose oral, good laboratory practices
(GLP) toxicity studies in rat and dog with a 1ꢀmonth recovery period to investigate the reversibility of
any possible toxicological effect. In rat, the compound was well tolerated at doses of 25 and 75
mg/kg/day. At a dose of 75 mg/kg/day, an increase in cholesterol and a minimal to slight increase in
foamy macrophages in the lung, suggestive of phospholipidosis, were noted in female rats. At 225
mg/kg/day, histological changes indicative for phospholipidosis were more pronounced and present in
various tissues, including kidney and lung, in addition to transient effects on body weight, food
consumption, and clinical pathology. The 75 mg/kg/day was considered to be the No Observed Adverse
Effect Level (NOAEL) with exposures for male and female rats indicated in Table 6. With a pKa lower
than 8 and a cLogP lower than 2, 3 does not have the characteristics of a typical phospholipidosisꢀ
inducer,20 as confirmed by the only weak potential to induce phospholipidosis in the in vitro assay. An
explanation for the phospholipidosis observed in the toxicology studies probably lies in the tissue
distribution of the compound, with high tissue/plasma ratios especially in those tissues where
indications of phospholipidosis were observed. The phospholipidosis was found to be reversible during
ACS Paragon Plus Environment

Page 21 of 38
Journal of Medicinal Chemistry
a 1ꢀmonth recovery period following the 1ꢀmonth treatment. In addition, even in the high dose group, no
signals of tissue damage of toxicity were observed which could be linked to the phospholipidosis.
Therefore, phospholipidosis was not considered as a showꢀstopper to move into clinical trials.
1
2
3
4
5
6
7
8
9
In dog, 3 was dosed orally at 20, 70 and 250 mg/kg/day. During the first week of dosing, the dogs were
fed immediately after dosing. At 70 and 250 mg/kg/day, this resulted in central nervous system (CNS)
symptoms (ataxia, decreased general activity and/or tremors). From the fifth dose onwards, dogs were
fasted up to 4 h after dosing and this resulted in lower exposures, and consequently in the disappearance
of CNS symptoms. A minimal decrease in red blood cell parameters was seen in males dosed at 250
mg/kg/day. There were no relevant changes at the end of the 1ꢀmonth recovery period. Based on these
findings, the NOAEL for 3 in dog was 70 mg/kg/day with the exposures given in Table 7. No
depigmentation was observed in either the rat or dog toxicity studies. The 10ꢀfold selectivity of BACE1
over BACE2 may be advantageous here, but a longer treatment period will be required to exclude
depigmentation upon chronic BACE1 and BACE2 inhibition.17
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 7. NOAEL after a 1ꢀmonth repeated dose GLP toxicity study in rat and dog
Male
Female
Cmax
Dose
Cmax
AUC 0ꢀ24h
AUC 0ꢀ24h
Rat
75 mg/kg/day
1490 ng/mL 13663 ng.h/mL 2723 ng/mL
1973 ng/mL 26319 ng.h/mL 2083 ng/mL
38846 ng.h/mL
23066 ng.h/mL
Dog 70 mg/kg/day
CONCLUSIONS
Replacement of the 2ꢀfluoro substituent in 2 with a trifluoromethyl group resulted in improvement on
hERG inhibition. While 2ꢀfluoroꢀ1,4ꢀoxazine amidine 2 was more potent in vivo (EC50 20 ng/mL vs 105
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 38
ng/mL for 3), the CF3ꢀoxazine 3 has an improved cardiovascular profile, with a sufficient safety margin
of at least 30ꢀfold of the NOAEL over the human predicted dose. Also, concerns around the potential
chemical instability of the anomericꢀlike fluorine in 2 were mitigated. Compound 3 was considered
sufficiently safe in toxicity assessment studies and is now positioned to move into clinical trials.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
EXPERIMENTAL SECTION
Chemistry
All reactions were carried out employing standard chemical techniques under inert atmosphere. Solvents
used for extraction, washing, and chromatography were HPLC grade. Unless otherwise noted, all
reagents were purchased from SigmaꢀAldrich or Acros Organics and were used without further
1
1
purification. All final compounds were characterized by H NMR and LC/MS. H Nuclear Magnetic
Resonance spectra were recorded on Bruker spectrometers: 360 MHz, DPXꢀ400 MHz and AVꢀ500
MHz. For the 1H spectra, all chemical shifts are reported in part per million (δ) units, and are relative to
13
the residual signal at 7.26 and 2.50 ppm for CDCl3 and DMSO, respectively. C chemical shifts are
reported as δ values in ppm relative to the residual solvent peak (CDCl3 = 77.16). All final compounds
were confirmed to be >95% pure via HPLC methods. All the LC/MS analyses were performed using an
Agilent G1956A LC/MS quadrupole coupled to an Agilent 1100 series liquid chromatography (LC)
system consisting of a binary pump with degasser, autosampler, thermostated column compartment and
diode array detector. The mass spectrometer (MS) was operated with an atmospheric pressure electroꢀ
spray ionization (APIꢀES) source in positive ion mode. The capillary voltage was set to 3000 V, the
fragmentor voltage to 70 V and the quadrupole temperature was maintained at 100 °C. The drying gas
flow and temperature values were 12.0 L/min and 350 °C, respectively. Nitrogen was used as the
nebuliser gas, at a pressure of 35 psig. Data acquisition was performed with Agilent Chemstation
software. Analyses were carried out on a YMC pack ODSꢀAQ C18 column (50 mm long x 4.6 mm I.D.;
3 ꢁm particle size) at 35 °C, with a flow rate of 2.6 mL/min. A gradient elution was performed from 95%
(water + 0.1% formic acid)/5% acetonitrile to 5% (water + 0.1% formic acid)/95% acetonitrile in 4.8
ACS Paragon Plus Environment

Page 23 of 38
Journal of Medicinal Chemistry
min; the resulting composition was held for 1.0 min; from 5% (water + 0.1% formic acid)/95%
acetonitrile to 95% (water + 0.1% formic acid)/5% acetonitrile in 0.2 min. The standard injection
volume was 2 ꢁL. Acquisition ranges were set to 190ꢀ400 nm for the UVꢀPDA detector and 100ꢀ1400
m/z for the MS detector. Optical rotations measurements were carried out on a 341 PerkinElmer
polarimeter in the indicated solvents. Melting points were determined with a DSC823e (MettlerꢀToledo)
and were measured with a temperature gradient of 30 ºC/min. The reported values are peak values. The
synthesis and characterization of intermediates 11ꢀ16 and compounds 1, 2, 6, 7, and 10 have been
described before.8 The following section comprises the synthetic procedures and analytical data for all
other final compounds reported in this manuscript.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Nꢀ{3ꢀ[(2R,3R)ꢀ5ꢀAminoꢀ3ꢀmethylꢀ2ꢀ(trifluoromethyl)ꢀ3,6ꢀdihydroꢀ2Hꢀ1,4ꢀoxazinꢀ3ꢀyl]ꢀ4ꢀ
fluorophenyl}ꢀ5ꢀcyanopyridineꢀ2ꢀcarboxamide (3).
5ꢀCyanoꢀ2ꢀpyridinecarboxylic acid (1.56 g, 10.5 mmol) and 4ꢀ(4,6ꢀdimethoxyꢀ1,3,5ꢀtriazinꢀ2ꢀyl)ꢀ4ꢀ
methyl morpholinium chloride (3.72 g, 12.6 mmol) were dissolved in MeOH (400 mL) at rt. After
stirring for 5 min at rt, the mixture was cooled on an iceꢀbath and a solution of aniline 168 (3.40 g, 10.5
mmol) in MeOH (10 mL) was added. The mixture was stirred at 0 °C for 6 h, then stirred at rt
overnight. The mixture was partitioned between a sat. aq. NaHCO3 solution and DCM. The organic
layer was dried (MgSO4), filtered, and concentrated in vacuo. The crude product was purified by flash
column chromatography (silica; 7 N solution of ammonia in MeOH/DCM 0/100 to 5/95). The desired
fractions were collected and the solvents evaporated in vacuo and subsequently triturated with
diisopropylether/isopropanol, filtered, and dried in vacuo to yield 3 as a white solid (1.83 g, 41% yield).
1H NMR (400 MHz, CDCl3): δ (ppm) 1.56 (s, 3H), 4.10 (d, J = 15.9 Hz, 1H), 4.23 (d, J = 15.9 Hz, 1H),
4.54 (q, J = 8.1 Hz, 1H), 7.13 (dd, J = 12.0, 8.8 Hz, 1H), 5.86 (s, 2H), 7.86−7.79 (m, 1H), 8.18 (dd, J =
7.2, 2.8 Hz, 1H), 8.28 (dd, J = 8.2, 0.8 Hz, 1H), 8.58 (dd, J = 8.2, 2.0 Hz, 1H), 9.20 (dd, J = 2.0, 0.8 Hz,
13
1H), 10.77 (s, 1H). C NMR (150 MHz, CDCl3): δ (ppm) 28.6 (d, JC−F = 3.8 Hz), 55.2 (d, JC−F = 3.6
Hz), 58.5, 72.5 (dq, JC−F = 27.3, 5.6 Hz), 112.4, 116.0, 116.2 (d, JC−F =25.1 Hz), 120.4 (d, JC−F = 8.7
Hz), 120.9 (d, JC−F = 4.7 Hz), 122.3, 124.1 (q, JC−F = 286.1 Hz), 130.8 (d, JC−F = 14.0 Hz), 133.2 (d, JC−F
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 38
= 2.3 Hz), 141.3, 150.7, 152.4, 154.6, 157.7 (d, JC−F = 242.7 Hz), 159.9. [α]20 = ꢀ31.6 (c = 0.14 in
D
1
2
3
4
5
6
7
8
MeOH); Melting point 240 °C, LRMS (ESI) m/z: [M + H]+ calcd for C19H16F4N5O2, 422.1; found,
422.1. Anal. calcd for C19H15F4N5O2: C, 54.16%; H, 3.59%; N, 16.62%. Found: C, 54.23%; H, 3.36%;
N, 16.54%.
9
Nꢀ{3ꢀ[(2R,3R)ꢀ5ꢀAminoꢀ2ꢀfluoroꢀ3ꢀmethylꢀ3,6ꢀdihydroꢀ2Hꢀ1,4ꢀoxazinꢀ3ꢀyl]ꢀ4ꢀfluorophenyl}ꢀ5ꢀ
cyanoꢀ3ꢀfluoropyridineꢀ2ꢀcarboxamide (4).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Starting from (5R,6R)ꢀ5ꢀ(5ꢀaminoꢀ2ꢀfluorophenyl)ꢀ6ꢀfluoroꢀ5ꢀmethylꢀ5,6ꢀdihydroꢀ2Hꢀ1,4ꢀoxazinꢀ3ꢀ
amine8 (73 mg, 0.3 mmol) and 5ꢀcyanoꢀ3ꢀfluoropyridineꢀ2ꢀcarboxylic acid (50 mg, 0.3 mmol), and
following the same procedure as for 3, the corresponding 4 was obtained (52 mg, 44 % yield). 1H NMR
(360 MHz, CDCl3) δ 1.65 (s, 3 H) 4.04 (d, J = 15.4 Hz, 1H) 4.27 (d, J = 15.7 Hz, 1H) 6.04 (d, J = 52.7
Hz, 1H) 7.09 (dd, J = 11.3, 8.8 Hz, 1H) 7.46 (dd, J = 6.8, 2.7 Hz, 1H) 7.85 ꢀ 7.96 (m, 2H) 8.70 (d, J =
1.8 Hz, 1H) 9.53 ꢀ 9.71 (br. s., 1H). LCꢀMS m/z 390 [M + H]+.
Nꢀ{3ꢀ[(2R,3R)ꢀ5ꢀAminoꢀ2ꢀfluoroꢀ3ꢀmethylꢀ3,6ꢀdihydroꢀ2Hꢀ1,4ꢀoxazinꢀ3ꢀyl]ꢀ4ꢀfluorophenyl}ꢀ3ꢀ
chloroꢀ5ꢀcyanopyridineꢀ2ꢀcarboxamide (5).
Starting from (5R,6R)ꢀ5ꢀ(5ꢀaminoꢀ2ꢀfluorophenyl)ꢀ6ꢀfluoroꢀ5ꢀmethylꢀ5,6ꢀdihydroꢀ2Hꢀ1,4ꢀoxazinꢀ3ꢀ
amine8 (0.10 g, 0.42 mmol) and 3ꢀchloroꢀ5ꢀcyanopyridineꢀ2ꢀcarboxylic acid (83 mg, 0.46 mmol), and
following the same procedure as for 3, the corresponding 5 was obtained as a solid after trituration with
heptane (50 mg, 30% yield). 1H NMR (500 MHz, CDCl3) δ ppm 1.65 (s, 3H), 4.04 (d, J = 15.6 Hz, 1H),
4.27 (d, J = 15.6 Hz, 1H), 6.05 (d, J = 52.0 Hz, 1H), 7.10 (dd, J = 11.3, 9.0 Hz, 1H), 7.40 (dd, J = 6.8,
2.5 Hz, 1H), 7.94 (dt, J = 8.5, 3.3 Hz, 1H), 8.13 ꢀ 8.19 (m, 1H), 8.74 (d, J = 1.2 Hz, 1H), 9.66 (br s, 1H).
LCꢀMS m/z 406 [M + H]+; [α]20D = +75.6 (c = 0.52 in DMF); mp = 125.5° C.
Nꢀ{3ꢀ[(2R,3R)ꢀ5ꢀAminoꢀ3ꢀmethylꢀ2ꢀ(trifluoromethyl)ꢀ3,6ꢀdihydroꢀ2Hꢀ1,4ꢀoxazinꢀ3ꢀyl]ꢀ4ꢀ
fluorophenyl}ꢀ3ꢀchloroꢀ5ꢀcyanopyridineꢀ2ꢀcarboxamide (8).
Starting from aniline 16 (0.16 g, 0.47 mmol) and 3ꢀchloroꢀ5ꢀcyanopyridineꢀ2ꢀcarboxylic acid (85 mg, 0.47
mmol), following the same procedure as for 3, the corresponding 8 was obtained (85 mg, 40% yield). 1H
NMR (360 MHz, CDCl3) δ ppm 1.68 (s, 3H), 1.71 ꢀ 1.95 (m, 2H), 4.24 (s, 2H), 4.66 (q, J = 8.4 Hz, 1H),
ACS Paragon Plus Environment

Page 25 of 38
Journal of Medicinal Chemistry
7.07 (dd, J = 11.5, 9.0 Hz, 1H), 7.87 (dd, J = 6.6, 2.9 Hz, 1H), 8.06 (ddd, J = 8.8, 4.0, 2.9 Hz, 1H), 8.17
(d, J = 1.8 Hz, 1H), 8.73 (d, J = 1.5 Hz, 1H), 9.76 (s, 1H). LCꢀMS m/z 456 [M + H]+; [α]20D = ꢀ33.8 (c =
0.59 in DMF).
1
2
3
4
5
6
7
8
9
Nꢀ{3ꢀ[(2R,3R)ꢀ5ꢀAminoꢀ3ꢀmethylꢀ2ꢀ(trifluoromethyl)ꢀ3,6ꢀdihydroꢀ2Hꢀ1,4ꢀoxazinꢀ3ꢀyl]ꢀ4ꢀ
fluorophenyl}ꢀ5ꢀchloropyridineꢀ2ꢀcarboxamide (9).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Starting from aniline 16 (0.25 g, 0.86 mmol) and 5ꢀchloropyridineꢀ2ꢀcarboxylic acid (135 mg, 0.86
mmol), following the same procedure as for 3, the corresponding 9 was obtained as a solid after
1
trituration with diisopropylether (230 mg, 62% yield). H NMR (360 MHz, CDCl3) δ 1.69 (s, 3H) 4.19 ꢀ
4.31 (m, 2H) 4.28 ꢀ 4.54 (m, 2H) 4.66 (q, J = 8.1 Hz, 1H) 7.06 (dd, J = 11.7, 8.8 Hz, 1H) 7.88 (dd, J =
8.4, 2.6 Hz, 1H) 7.91 (dd, J = 6.8, 2.7 Hz, 1H) 7.99 ꢀ 8.08 (m, 1H) 8.25 (d, J = 8.4 Hz, 1H) 8.56 (d, J =
2.2 Hz, 1H) 9.88 (br. s, 1H). LCꢀMS m/z 431 [M + H]+; [α]20D = ꢀ36.8 (c = 0.23 in MeOH); mp = 218.3°
C.
(5RS,6RS)ꢀ5ꢀ(2ꢀFluoroꢀ5ꢀ(pyrimidinꢀ5ꢀyl)phenyl)ꢀ5ꢀmethylꢀ6ꢀ(trifluoromethyl)ꢀ5,6ꢀdihydroꢀ2Hꢀ
1,4ꢀoxazinꢀ3ꢀamine (18).
Racemic bromide 15 (100 mg, 0.28 mmol), 5ꢀpyrimidinylboronic acid (70 mg, 0.56 mmol), and
tetrakis(triphenylphosphine)palladium (33 mg, 0.028 mmol) were dissolved in a mixture of saturated aq.
NaHCO3 solution (5 mL) and 1,4ꢀdioxane (4 mL). The mixture was flushed with nitrogen, and then
heated at 80 °C for 2 h. The mixture was allowed to come to rt, water was added, and the mixture was
extracted with DCM. The organic layer was separated, washed with brine, dried (MgSO4), filtered, and
concentrated in vacuo. The crude product was purified by flash column chromatography (silica; 7 N
solution of ammonia in MeOH/DCM 0/100 to 10/90). The desired fractions were collected and the
1
solvents evaporated in vacuo to give 18 (85 mg, 85% yield). H NMR (400 MHz, DMSOꢀd6) δ ppm
1.59 (s, 3H), 4.06 ꢀ 4.29 (m, 2H), 4.65 (q, J = 8.5 Hz, 1H), 5.95 (br s, 2H), 7.32 (dd, J = 12.0, 8.5 Hz,
1H), 7.78 (ddd, J = 8.4, 4.5, 2.6 Hz, 1H), 8.21 (dd, J = 7.3, 2.5 Hz, 1H), 9.06 (s, 2H), 9.20 (s, 1H). LCꢀ
MS m/z 355 [M + H]+.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
6ꢀ((E)ꢀ3ꢀ((2R,3R)ꢀ5ꢀAminoꢀ3ꢀmethylꢀ2ꢀ(trifluoromethyl)ꢀ3,6ꢀdihydroꢀ2Hꢀ1,4ꢀoxazinꢀ3ꢀyl)ꢀ4ꢀ
fluorostyryl)nicotinonitrile (19).
Page 26 of 38
1
2
3
4
5
6
7
8
9
Bromide 15 (123 mg, 0.35 mmol), (E)ꢀ(2ꢀ(5ꢀcyanopyridinꢀ2ꢀyl)vinyl)boronic acid pinacol ester (89 mg, 0.35
mmol), and tetrakis(triphenylphosphine)palladium (40 mg, 0.035 mmol) were dissolved in a mixture of
saturated aq. NaHCO3 solution (0.4 mL) and 1,4ꢀdioxane (4.9 mL). The mixture was flushed with
nitrogen, and then heated at 140 °C for 30 min. under microwave irradiation. The mixture was allowed
to come to rt, water was added, and the mixture was extracted with DCM. The organic layer was
separated, washed with brine, dried (MgSO4), filtered, and concentrated in vacuo. The crude product
was purified by flash column chromatography (silica; 7 N solution of ammonia in MeOH/DCM 0/100 to
4/96). The desired fractions were collected and the solvents evaporated in vacuo to give 19 (36 mg, 26%
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
yield). H NMR (360 MHz, CDCl3): δ ppm 1.69 (d, J = 0.7 Hz, 3H), 4.24 (s, 2H), 4.64 (q, J = 8.4 Hz,
1H), 7.05 (dd, J = 11.7, 8.4 Hz, 1H), 7.12 (d, J = 16.1 Hz, 1H), 7.42 (d, J = 8.4 Hz, 1H), 7.51 (ddd, J =
8.4, 4.8, 2.6 Hz, 1H), 7.78 (d, J = 16.1 Hz, 1H), 7.88 (dd, J = 8.4, 2.2 Hz, 1H), 8.15 (dd, J = 7.7, 2.2 Hz,
1H), 8.82 (d, J = 2.2 Hz, 1H). LCꢀMS m/z 405 [M + H]+.
2ꢀ((3ꢀ[(2R,3R)ꢀ5ꢀAminoꢀ3ꢀmethylꢀ2ꢀ(trifluoromethyl)ꢀ3,6ꢀdihydroꢀ2Hꢀ1,4ꢀoxazinꢀ3ꢀyl]ꢀ4ꢀ
fluorophenyl}ꢀamino))nicotinonitrile (20).
Aniline 16 (0.358 g, 1.23 mmol) was dissolved in 1,4ꢀdioxane (15 mL). 2ꢀBromoꢀ3ꢀcyanopyridine (191
mg, 1.05 mmol), cesium carbonate (0.80 g, 2.46 mmol), 1,1′ꢀbis(diphenylphosphino)ferrocene (103 mg,
0.18 mmol) were added and the mixture stirred under argon atmosphere for a few minutes.
Tris(dibenzylideneꢀacetone)dipalladium(0) (56 mg, 0.062 mmol) was then added. The reaction tube was
sealed and the mixture was stirred at 160 ºC for 1 h. under microwave irradiation. After cooling, the
reaction mixture was diluted with DCM and filtered over dicalite. The filtrate was concentrated in vacuo.
The crude product was purified by flash column chromatography (silica; 7 N solution of ammonia in
MeOH/DCM 0/100 to 5/95). The desired fractions were collected and concentrated. This crude was
further purified by preparative SFC on (Chiralpak Diacel AS 20x250mm). Mobile phase (CO2, MeOH
1
with 0.2% iPrNH2) to yield 20 (0.052 g, 11% yield). H NMR (360 MHz, CDCl3): δ ppm 1.67 (s, 3H)
ACS Paragon Plus Environment

Page 27 of 38
Journal of Medicinal Chemistry
4.14 ꢀ 4.34 (m, 4H) 4.65 (d, J = 8.42 Hz, 1H) 6.78 (dd, J = 7.50, 4.94 Hz, 1H) 6.98 ꢀ 7.08 (m, 2H) 7.71
(dd, J = 6.59, 2.93 Hz, 1H) 7.78 (dd, J = 7.68, 1.83 Hz, 1H) 7.94 (ddd, J = 8.78, 4.03, 2.93 Hz, 1H) 8.37
(dd, J = 5.12, 1.83 Hz, 1H). LCꢀMS m/z 394 [M + H]+; [α]20D = ꢀ21.3 (c = 0.36 in DMF).
1
2
3
4
5
6
7
8
9
(5R)ꢀ5ꢀ(5ꢀBromoꢀ2ꢀfluorophenyl)ꢀ6ꢀ(difluoromethylene)ꢀ5ꢀmethylmorpholinꢀ3ꢀone (21).
(5R)ꢀ5ꢀ(5ꢀBromoꢀ2ꢀfluorophenyl)ꢀ6ꢀchloroꢀ5ꢀmethylꢀ6ꢀ(trifluoromethyl)morpholinꢀ3ꢀone (13) 8
(7 g, 17.9 mmol) and zinc copper couple (8.55 g, 66.3 mmol) were stirred in acetic acid (420 mL) at rt
for 16 h. The reaction mixture was filtered, washed with DCM and concentrated in vacuo. Ammonium
hydroxide solution (28% in water) and DCM were added and the mixture was stirred at rt for 1 h. The
organic layer was separated and the aqueous layer was extracted with DCM. The combined organic
layers were dried (MgSO4), filtered and evaporated in vacuo to yield 21 (6 g, 99% yield) as a white
powder. 1H NMR (360 MHz, DMSOꢀd6): δ ppm 1.73 (d, J = 5.1 Hz, 3H), 4.27ꢀ4.44 (m, 2H), 7.26 (dd, J
= 11.7, 8.8 Hz, 1H), 7.55 (dd, J = 7.3, 2.6 Hz, 1H), 7.60ꢀ7.67 (m, 1H), 9.07 (s, 1H). LCꢀMS m/z 337 [M
+ H]+.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(5R)ꢀ5ꢀ(5ꢀbromoꢀ2ꢀfluorophenyl)ꢀ6ꢀ(difluoromethylene)ꢀ5ꢀmethylꢀ5,6ꢀdihydroꢀ2Hꢀ1,4ꢀoxazinꢀ3ꢀ
amine (22).
P2S5 (5.95 g, 26.8 mmol) was added to a solution of 21 (6 g, 17.9 mmol) in THF (145 mL) at rt. The
reaction mixture was stirred at 70 ºC for 90 minutes. Then the mixture was cooled to rt, filtered off and
the organic solvent evaporated in vacuo to yield crude (5R)ꢀ5ꢀ(5ꢀbromoꢀ2ꢀfluorophenyl)ꢀ6ꢀ
(difluoromethylene)ꢀ5ꢀmethylmorpholinꢀ3ꢀthione (5.9 g). This was dissolved in 7N ammonia in MeOH
(390 mL) and the reaction mixture was stirred at 80 ºC for 2 h. The solvent was evaporated and the
crude product purified by column chromatography (silica; 7 N solution of ammonia in MeOH/DCM
0/100 to 5/95). The desired fractions were collected and concentrated in vacuo to yield 22 (4.04 g, 72%
yield). 1H NMR (360 MHz, CDCl3): δ ppm 1.72 (d, J = 4.0 Hz, 3H), 4.29 (s, 2H), 6.90 (dd, J = 11.3, 8.8
Hz, 1H), 7.35 (ddd, J = 8.6, 4.2, 2.6 Hz, 1H), 7.53 (dd, J = 7.3, 2.6 Hz, 1H). LCꢀMS m/z 336 [M + H]+.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
(5R)ꢀ5ꢀ(5ꢀAminoꢀ2ꢀfluorophenyl)ꢀ6ꢀ(difluoromethylene)ꢀ5ꢀmethylꢀ5,6ꢀdihydroꢀ2Hꢀ1,4ꢀoxazinꢀ3ꢀ
amine (23)
Page 28 of 38
1
2
3
4
5
6
7
8
9
Compound 22 (3.6 g, 10.7 mmol) was mixed with NaN3 (1.75 g, 26.9 mmol), CuI (2.56 g, 13.4 mmol)
and Na2CO3 (2.28 g, 21.5 mmol) in DMSO (153 mL) and the reaction mixture was degassed. After that,
N,N’ꢀdimethylethylenediamine (2 mL, 18.8 mmol) was added and the mixture was heated at 110 °C
until completion of the reaction, about 3 h. The reaction mixture was concentrated in vacuo. 7N
ammonia in MeOH was added and the mixture was stirred overnight. The precipitate formed was
filtered off and the filtrate was concentrated in vacuo. The crude product was purified by column
chromatography ((silica; eluent: 7 M solution of ammonia in MeOH/DCM 0/100 to 30/70). The desired
fractions were collected and concentrated in vacuo to yield 23 (1.52 g, 52% yield). 1H NMR (360 MHz,
DMSOꢀd6): δ ppm 1.53 (d, J = 4.4 Hz, 2H), 4.07ꢀ4.29 (m, 1H), 4.90 (s, 1H), 5.90 (s, 1H), 6.34ꢀ6.45 (m,
1H), 6.60 (dd, J = 7.0, 2.9 Hz, 1H), 6.73 (dd, J = 11.7, 8.8 Hz, 1H). LCꢀMS m/z 272 [M + H]+.
Nꢀ{3ꢀ(3R)ꢀ(5ꢀAminoꢀ2ꢀ(difluoromethylene)ꢀ3ꢀmethylꢀ3,6ꢀdihydroꢀ2Hꢀ1,4ꢀoxazinꢀ3ꢀyl)ꢀ4ꢀ
fluorophenyl}ꢀ5ꢀmethoxypyrazineꢀ2ꢀcarboxamide (24)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5ꢀMethoxypyrazineꢀ2ꢀcarboxylic acid (0.22 g, 1.42 mmol) was dissolved in MeOH (30 mL) and
DMTMM (0.46 g, 1.55 mmol) was added. After stirring the mixture for 5 minutes, a solution of aniline
23 (0.35 g, 1.29 mmol) in MeOH (20 mL) was added at 0 ºC, and the mixture was stirred for 16 h at rt.
The solvent was evaporated in vacuo. The crude material was purified by flash column chromatography
(silica; eluent: 7 M solution of ammonia in MeOH/DCM 0/100 to 5/95). The desired fractions were
collected and concentrated in vacuo. The residue was suspended from diisopropylether/heptanes,
1
filtered and dried under high vacuum to yield 24 (0.266 g, 51% yield) as a white solid. H NMR (360
MHz, DMSOꢀd6): δ ppm 1.62 (d, J = 3.7 Hz, 2H), 4.02 (s, 2H), 4.24 (d, J = 13.2 Hz, 2H), 5.98 (br s,
2H), 7.04ꢀ7.16 (m, 1H), 7.75ꢀ7.84 (m, 1H), 7.96 (dd, J = 7.3, 2.6 Hz, 1H), 8.41 (d, J = 1.1 Hz, 1H), 8.89
(d, J = 1.1 Hz, 1H), 10.57 (s, 1H). LCꢀMS m/z 408 [M + H]+; [α]20D = +156.8 (c = 0.36 in DMF).
ACS Paragon Plus Environment

Page 29 of 38
Journal of Medicinal Chemistry
Nꢀ{3ꢀ[(2R,3R)ꢀ5ꢀAminoꢀ2ꢀ(difluoromethyl)ꢀ3ꢀmethylꢀ3,6ꢀdihydroꢀ2Hꢀ1,4ꢀoxazinꢀ3ꢀyl]ꢀ4ꢀ
fluorophenyl}ꢀ5ꢀmethoxypyrazineꢀ2ꢀcarboxamide (25).
1
2
3
4
5
6
7
8
9
Compound 24 (0.154 g, 0.378 mmol) was dissolved in EtOAc (5 mL) and palladium on carbon (10%)
(0.04 g, 0.038 mmol) and thiophene (0.4% solution in THF, 0.5 mL, 0.026 mmol) were added. The
mixture was hydrogenated at rt and atmospheric pressure for 16 h. The catalyst was filtered off and the
solvents evaporated in vacuo. The crude product was purified by flash column chromatography (silica;
eluent: 7 M solution of ammonia in MeOH/DCM 0/100 to 2/98). The desired fractions were collected
and concentrated in vacuo. The residue was suspended from diisopropylether, filtered and dried under
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
high vacuum to yield 25 (0.067 g, 43% yield). H NMR (360 MHz, DMSOꢀd6): δ ppm 1.52 (s, 3H),
3.93 ꢀ 4.01 (m, 1H), 4.02 (s, 3H), 4.10 (d, J = 15.9 Hz, 1H), 4.15 (d, J = 15.9 Hz, 1H), 5.64 (td, J = 54.0,
4.4 Hz, 1H), 5.80 (br. s, 2H), 7.11 (dd, J = 12.1, 8.8 Hz, 1H), 7.80 (ddd, J = 8.8, 4.1, 2.7 Hz, 1H), 8.04
(dd, J = 7.3, 2.8 Hz, 1H), 8.42 (d, J = 1.3 Hz, 1H), 8.88 (d, J = 1.3 Hz, 1H), 10.44 (s, 1H). LCꢀMS m/z
410 [M + H]+; mp = 227.3° C.
Nꢀ{3ꢀ[(2R,3R)ꢀ5ꢀAminoꢀ3ꢀmethylꢀ2ꢀ(trifluoromethyl)ꢀ3,6ꢀdihydroꢀ2Hꢀ1,4ꢀoxazinꢀ3ꢀyl]ꢀ4ꢀ
fluorophenyl}ꢀ5ꢀ(2ꢀmethoxyethoxy)pyrazineꢀ2ꢀcarboxamide (26).
Starting from aniline 16 (0.25 g, 0.86 mmol) and 5ꢀ(2ꢀmethoxyethoxy)pyrazineꢀ2ꢀcarboxylic acid (cas
710322ꢀ69ꢀ3; 135 mg, 0.86 mmol), following the same procedure as for 3, the corresponding 26 was
1
obtained as a solid after trituration with diisopropylether (230 mg, 62% yield). H NMR (360 MHz,
CDCl3): δ ppm 1.68 (s, 3H), 3.46 (s, 3H), 3.76 ꢀ 3.83 (m, 2H), 4.23 (s, 2H), 4.34 (br s, 2H), 4.57 ꢀ 4.69
(m, 3H), 7.05 (dd, J =1 1.5, 9.0 Hz, 1H), 7.89 (dd, J = 6.8, 2.7 Hz, 1H), 8.03 (dt, J = 8.8, 3.5 Hz, 1H),
8.21 (s, 1H), 8.99 (d, J = 0.7 Hz, 1H), 9.55 (s, 1H). LCꢀMS m/z 472 [M + H]+; [α]20D = ꢀ17.5 (c = 0.32
in DMF).
Nꢀ{3ꢀ[(2R,3R)ꢀ5ꢀAminoꢀ3ꢀmethylꢀ2ꢀ(trifluoromethyl)ꢀ3,6ꢀdihydroꢀ2Hꢀ1,4ꢀoxazinꢀ3ꢀyl]ꢀ4ꢀ
fluorophenyl}ꢀ5ꢀ(2,2,2ꢀtrifluoroethoxy)pyrazineꢀ2ꢀcarboxamide (27).
Starting from aniline 16 (0.10 g, 0.34 mmol) and 5ꢀ(2,2,2ꢀtrifluoroethoxy)ꢀpyrazineꢀ2ꢀcarboxylic acid
(cas 1174323ꢀ36ꢀ4; 76 mg, 0.34 mmol), following the same procedure as for 3, the corresponding 27
ACS Paragon Plus Environment